Aeronomy of The Middle Atmosphere

Aeronomy of the Middle Atmosphere
ATMOSPHERIC AND OCEANOGRAPHIC SCIENCES LIBRARY
VOLUME 32
Editors
Lawrence A. Mysak, Department of Atmospheric and Oceanographic Sciences,

McGill University, Montreal, Canada
Kevin Hamilton, International Pacific Research Center, University of Hawaii,
Honolulu, HI, U.S.A.
Editorial Advisory Board
L. Bengtsson Max-Planck-Institut für Meteorologie, Hamburg, Germany

A. Berger Université Catholique, Louvain, Belgium
P.J. Crutzen Max-Planck-Institut für Chemie, Mainz, Germany
J.R. Garratt CSIRO, Aspendale, Victoria, Australia
G. Geernaert DMU-FOLU, Roskilde, Denmark
M. Hantel Universität Wien, Austria
A. Hollingsworth European Centre for Medium Range Weather Forecasts, Reading,
UK
H. Kelder KNMI (Royal Netherlands Meteorological Institute), De Bilt,
The Netherlands
T.N. Krishnamurti The Florida State University, Tallahassee, FL, U.S.A.
P. Lemke Alfred-Wegener-Institute for Polar and Marine Research,
Bremerhaven, Germany
P. Malanotte-Rizzoli MIT, Cambridge, MA, U.S.A.
S.G.H. Philander Princeton University, NJ, U.S.A.
D. Randall Colorado State University, Fort Collins, CO, U.S.A.
J.-L. Redelsperger METEO-FRANCE, Centre National de Recherches
Météorologiques, Toulouse, France
R.D. Rosen AER, Inc., Lexington, MA, U.S.A.
S.H. Schneider Stanford University, CA, U.S.A.
F. Schott Universität Kiel, Kiel, Germany
G.E. Swaters University of Alberta, Edmonton, Canada
J.C. Wyngaard Pennsylvania State University, University Park, PA, U.S.A.
The titles published in this series are listed at the end of this volume.
Aeronomy of the Middle
Atmosphere
Chemistry and Physics of the
Stratosphere and Mesosphere
Third revised and enlarged edition
Guy P. Brasseur
Max Planck Institute for Meteorology,
Hamburg, Germany
and
National Center for Atmoshpheric Research,
Boulder, CO, U.S.A.
and
Susan Solomon
NOAA Aeronomy Laboratory,
Boulder, CO, U.S.A.
A C.I.P. Catalogue record for this book is available from the Library of Congress.
Library of Congress Cataloging-in-Publication Data
ISBN-10 1-4020-3284-6 (HB)

ISBN-13 978-1-4020-3284-4 (HB)
ISBN-10 1-4020-3824-0 (e-book)
ISBN-13 978-1-4020-3824-1 (e-book)
Published by Springer,
P.O. Box 17, 3300 AA Dordrecht, The Netherlands.
www.springeronline.com
Printed on acid-free paper
All Rights Reserved

© 2005 Springer
No part of this work may be reproduced, stored in a retrieval system, or transmitted
in any form or by any means, electronic, mechanical, photocopying, microfilming, recording
or otherwise, without written permission from the Publisher, with the exception
of any material supplied specifically for the purpose of being entered
and executed on a computer system, for exclusive use by the purchaser of the work.
Printed in the Netherlands.

Contents
Preface ix
1. THE MIDDLE ATMOSPHERE AND ITS EVOLUTION 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . 1
1.2 Evolution of the Earth’s Atmosphere . . . . . . . . . . 2
1.3 Anthropogenic Perturbations . . . . . . . . . . . . . . 4
2. CHEMICAL CONCEPTS IN THE ATMOSPHERE 11

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . 11
2.2 Energy Levels of Molecules . . . . . . . . . . . . . . . 12
2.3 Thermodynamic Considerations . . . . . . . . . . . . . 21
2.4 Elementary Chemical Kinetics . . . . . . . . . . . . . 24
2.5 Heterogenous and Multiphase Reactions . . . . . . . . . 34
2.6 Photolysis Processes . . . . . . . . . . . . . . . . . . 40
2.7 Excited Species in the Middle Atmosphere . . . . . . . . 44
3. DYNAMICS AND TRANSPORT 51

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . 51
3.2 Structure of the Atmosphere and Some Observed
Dynamical Characteristics . . . . . . . . . . . . . . . 53
3.3 Fundamental Description of Atmospheric Dynamics . . . . 59
3.4 Atmospheric Waves . . . . . . . . . . . . . . . . . . 74
3.5 Effects of Dynamics on Chemical Species: Transport . . . 84
3.6 Dynamics and Meridional Transport in Two
Dimensions: A Conceptual View . . . . . . . . . . . . . 92
3.7 The Importance of Wave Transience and Dissipation . . 112
3.8 Vertical Transport above the Mesopause . . . . . . . . 125
3.9 Models of the Middle Atmosphere . . . . . . . . . . . 129
v
vi AERONOMY OF THE MIDDLE ATMOSPHERE
4. RADIATION 151
4.1 Introduction . . . . . . . . . . . . . . . . . . . . 151
4.2 Definitions . . . . . . . . . . . . . . . . . . . . . 156
4.3 Extraterrestrial Solar Radiation . . . . . . . . . . . . 159
4.4 The Attenuation of Solar Radiation in the Atmosphere . 169
4.5 Radiative Transfer . . . . . . . . . . . . . . . . . 180
4.6 The Thermal Effects of Radiation . . . . . . . . . . . 206
4.7 Photochemical Effects of Radiation . . . . . . . . . . 218
5. COMPOSITION AND CHEMISTRY 265

5.1 General . . . . . . . . . . . . . . . . . . . . . . 265
5.2 Oxygen Compounds . . . . . . . . . . . . . . . . . 271
5.3 Carbon Compounds . . . . . . . . . . . . . . . . . 292
5.4 Hydrogen Compounds . . . . . . . . . . . . . . . . 307
5.5 Nitrogen Compounds . . . . . . . . . . . . . . . . 327
5.6 Halogen Compounds . . . . . . . . . . . . . . . . 358
5.7 Stratospheric Aerosols and Clouds . . . . . . . . . . 386
5.8 Generalized Ozone Balance . . . . . . . . . . . . . . 401
6. OZONE PERTURBATIONS 443

6.1 Introduction . . . . . . . . . . . . . . . . . . . . 443
6.2 The Photochemically-Controlled Upper Stratosphere
and Mesosphere: 25-75 km . . . . . . . . . . . . . . 444
6.3 Lower Stratospheric Ozone Depletion: Observations
and Explanations . . . . . . . . . . . . . . . . . . 459
6.4 Summary and Outlook . . . . . . . . . . . . . . . 501
7. THE IONS 533

7.1 Introduction . . . . . . . . . . . . . . . . . . . . 533
7.2 Formation of Ions in the Middle Atmosphere . . . . . . 539
7.3 Positive Ion Chemistry . . . . . . . . . . . . . . . 552
7.4 Negative Ion Chemistry . . . . . . . . . . . . . . . 570
7.5 Effect of Ionic Processes on Neutral Constituents . . . . 581
7.6 Radio Waves in the Lower Ionosphere . . . . . . . . . 586
APPENDICES 599
Appendix 1: Physical Constants and Other Data . . . . . . 599
Appendix 2: Conversion Factors and Multiplying Prefixes . . 603
CONTENTS vii
Appendix 3: Rate Coefficients for Second-Order Gas-Phase

Reactions . . . . . . . . . . . . . . . . . . . . . . 607
Appendix 4: Rate Coefficients for Gas-Phase Associating
Reactions . . . . . . . . . . . . . . . . . . . . . . 611
Appendix 5: Surface Reaction Probability . . . . . . . . . 613
Appendix 6: Atmospheric Profiles . . . . . . . . . . . . 615
Figure Acknowledgements 625
Color S ection 627
Index 637
Preface
Since the industrial revolution, the chemical composition of the

atmosphere has changed at a rate unprecedented in recent history. The
concentration of carbon dioxide has increased by about 25% during
the 20th century as a result of fossil fuel combustion, and is expected
to further increase substantially over the next several decades. The
level of methane is more than a factor 2 higher than a century ago,
due in large part to agricultural practices. The aerosol load of the
atmosphere has been perturbed by such practices as the intensive use
of coal as a primary energy source. The release of nitrogen oxides,
carbon monoxide and hydrocarbons, as a result of fossil fuel and biomass
burning has led to enhanced abundances of ozone and other oxidants
in the troposphere. Organic halogenated compounds including the
industrially manufactured chlorofluorocarbons have become a dominant
source of ozone-depleting chlorine and bromine in the stratosphere. The
science of the ozone layer and its interactions with halogenated chemicals
are the focus of this book.
These human-induced changes have led to significant perturbations
in the Earth’s global environment, and will further affect it in the
decades ahead. Among these perturbations are the degradation of air
quality with associated health effects, the acidification of precipitation
with potential consequences for ecosystems, changes in the oxidizing
power of the atmosphere and hence on the ability of the atmosphere
to eliminate pollutants, changes in the Earth’s climate in response to
modified absorption and emission of radiation by atmospheric trace
constituents, and stratospheric ozone depletion leading to enhanced
levels of harmful solar ultraviolet radiation at the Earth’s surface.
Ozone plays a central role in the Earth’s environment. By absorbing
DNA damaging ultraviolet light originating from the Sun, it protects
living organisms including humans from lethal effects. It also strongly
affects the thermal structure of the atmosphere, and maintains a
dynamically stable layer, –the stratosphere–, between approximately 10
and 50 km altitude. Ozone, a greenhouse gas and a powerful oxidant,
also influences the Earth’s climate and plays a key role in the ability
ix
x AERONOMY OF THE MIDDLE ATMOSPHERE
of the atmosphere to destroy several chemical compounds, including

many primary or secondary pollutants.
The depletion of the stratospheric ozone layer captured the attention
of scientists, policy makers, and the public, particularly in the 1970s,
1980s, and 1990s. Active research quickly provided an improved
understanding of the mechanisms involved in the formation and fate
of stratospheric ozone. Ozone depletion emerged as one of the most
important environmental issues of the 20th century as the evidence grew
for substantial human influences on the ozone abundances over much
of the globe. An international environmental agreement (The Montreal
Protocol and its amendments and adjustments) now governs the global
use of ozone-depleting halocarbons. In the 21st century, interest remains
strong in the science, the history, and the future of the stratospheric
ozone layer.
The purpose of this volume is to provide to graduate students and
research scientists a comprehensive view of the chemical, dynamical and
radiative processes that affect ozone and other chemical compounds in
the middle atmosphere. The title of this volume –Aeronomy of the
Middle Atmosphere– has been chosen to highlight the topics covered in
this book. The word “Aeronomy” is defined as the science dealing with
planetary atmospheres with reference to their chemical composition,
physical properties, relative motions and reactions to radiation from
outer space (Chambers Dictionary). The field of aeronomy is therefore
highly interdisciplinary, drawing on the fields of chemistry, physics,
fluid dynamics, meteorology, statistics, mathematics, engineering, etc.
The middle atmosphere is defined as the region extending from the
tropopause (approximately 12 km) to the homopause (about 100 km
altitude).
The book is intended to provide an overview in a manner
understandable to persons familiar with college level chemistry and
physics. After a general introduction presented in Chapter 1, Chapter
2 reviews basic concepts from physical chemistry, which are of relevance
to atmospheric studies. Chapter 3 presents a highly simplified view of
dynamical and transport processes above the tropopause, and Chapter
4 summarizes important aspects of radiative transfer in relation to
the energy budget and photolytic processes in the middle atmosphere.
Chapter 5 presents an overview of the key chemical processes, which
influence the chemical composition of the middle atmosphere, while
Chapter 6 discusses human-induced perturbations affecting ozone and
other compounds. The chapter also presents a detailed discussion of
ozone depletion, particularly the formation of the spectacular Antarctic
PREFACE xi
ozone hole is provided in this chapter. Finally, Chapter 7 provides an

overview of ionization processes in the mesosphere and stratosphere.
Although the discussion focuses on the atmospheric layers ranging from
10 to 100 km, we have sometimes found it necessary to discuss processes
occurring in the troposphere (surface to about 12 km altitude) and
thermosphere (region above 100 km) in order to explain the roles of
these neighboring regions.
Due to the large number of topics to be addressed in the present
volume, we have found it impossible to provide a complete review of
the available literature. We anticipate that the interested reader will
regard the discussions as a starting point and find additional references
by consulting publications cited here. The periodic international
assessments of the state of the ozone layer, published by the World
Meteorological Organization and the United Nations Environment
Program, provide additional information.
When we wrote the first edition of this book in the early 1980’s,
we did not anticipate that the question of stratospheric ozone would
soon receive so much scientific attention. The discovery of the Antarctic
ozone hole sparked the rapid development of exciting scientific programs
and new research understanding. Since the publication of the early
editions of this book, new measurement techniques (including space
platforms and instruments), more accurate laboratory methods, and
more sophisticated atmospheric models have become available. Our
understanding of the middle atmosphere, including the role of chemical
transformations on the surfaces of liquid or solid particles and the
complexity of the air motions, has improved dramatically. More
importantly, the discovery and explanation of the Antarctic ozone hole
has shown that human activities can lead to major changes in the
chemical composition of the atmosphere not just locally and regionally,
but at the global scale. The new edition of the present volume attempts
to synthesize some of the most recent findings.
The review of critical portions of the manuscript by colleagues has
certainly improved the quality of this volume. Material provided to us
has also been of great help. We particularly thank S. Chabrillat, D.
Edwards, R. Garcia, M. Giorgetta, H. Graf, C. Granier, M. Hagan, J.
Holton, I. Karol, K. Kodera, E. Kopp, S. Madronich, E. Manzini, D.
Marsh, M. Mlynczak, J. Orlando, T. Peter, W. Randel, A. Smith, Stan
Solomon, X. Tie, G. Tyndall, S. Walters, X. Zhu. We deeply appreciate
the assistance of R. Terrell Bailey, Paula Fisher, and Debe Fisher for
the preparation of the camera-ready manuscript, of C. Granier for her
proofreading of the manuscript, and of the Imaging and Design Center
xii AERONOMY OF THE MIDDLE ATMOSPHERE
at the National Center for Atmospheric Research (NCAR) for their

expert figure drafting. We also acknowledge the Atmospheric Chemistry
Division at NCAR for supporting the preparation of this book.
Guy P. Brasseur
Susan Solomon
Boulder, Colorado
March, 2005
Chapter 1
THE MIDDLE ATMOSPHERE

AND ITS EVOLUTION
1.1 Introduction
Our atmosphere is the medium for life on the surface of the planet, and
the transition zone between Earth and space. Humankind has always
been interested in the characteristics, manifestations and perturbations
of the atmosphere — in its changing weather patterns, and its brilliant
sunsets, rainbows, and aurorae.
The lower part of the atmosphere has been continuously studied
over many years through meteorological programs. The development
of rocket and satellite technology during the past 40 years has also led
to the investigation of the upper atmosphere, establishing a new field of
research.
The intermediate region which extends from about 10 to 100 km
altitude is sometimes called the middle atmosphere. This region is
somewhat less accessible to observation and has only been systematically
studied for the past 25 to 30 years. The purpose of this volume is to
outline some of the factors which control the behavior of this layer of
the atmosphere, a region which is particularly vulnerable to external
perturbations such as solar variability and volcanic eruptions or the
emission of anthropogenic material, either at the surface or at altitude.
One of the most important chemical constituents in the middle
atmosphere is ozone, because it is the only atmospheric species that
effectively absorbs ultraviolet solar radiation from about 250-300 nm,
protecting plant and animal life from exposure to harmful radiation.
Therefore, the stability of the ozone layer (located near 15 to 25 km) is
a central part of the study of the middle atmosphere.
In the study of planetary atmospheres it has been customary
to distinguish between thermodynamic and dynamic aspects, which
constitute a portion of the field of meteorology, and the chemical and
photochemical aspects, which are part of the domain of aeronomy.
However, the interactions between these different disciplines play an
1
2 AERONOMY OF THE MIDDLE ATMOSPHERE
important role, particularly in the altitude region which we shall address

here, so that an effort to examine the coupled dynamical, chemical, and
radiative problem must be made in order to understand the observed
variations in atmospheric chemical constituents.
1.2 Evolution of the Earth’s Atmosphere
The terrestrial atmosphere has evolved considerably over the course

of time. The details of this evolution, however, are not well established,
and not all scholars agree on how it occurred. Reviews of our current
understanding of the atmospheric evolution are given, for example, by
Walker (1977) and Kasting (1993).
The universe was created some 15-20 billion years ago when an
enormous explosion of a dense center of matter took place (the big bang).
The Earth was formed approximately 4.6 billion years ago by accretion of
small celestial bodies (planetisimals) that had condensed from the solar
nebula during its cooling. The gases present at the early stage of the
Earth’s formation, which could have formed the primordial atmosphere,
rapidly dissipated before they could be retained by a strong gravitational
field. Instead, a secondary atmosphere was produced by the gradual
release of gases from the Earth’s interior (which was extremely warm due
to radiative decay of uranium, thorium, and potassium contained in the
planetisimals). Support for this idea stems in part from the depletion
of inert gases such as He, Ne, Ar, and Kr on Earth, relative to their
observed abundance in the solar system. Since there is no known process
which could remove these very unreactive species from the atmosphere,
their low abundance suggests that they never represented as large a
fraction of our atmosphere as they do elsewhere in our solar system
(e.g., Moulton, 1905).
It is believed that the secondary atmosphere contained large quantities
of carbon dioxide (CO2 ), water vapor (H2 O), nitrogen (N2 ) and perhaps
hydrogen chloride (HCl) and trace amounts of other substances. As the
planet cooled, water vapor condensed, initiating a vigorous hydrological
cycle and producing large oceans. The sediments of the early Earth
were probably produced by weathering of the primordial igneous rock
minerals by dissolved carbon, nitrogen, chlorine, and sulfur acidic rain
water.
An important question related to the evolution of the Earth’s climate
is the so-called “faint young sun paradox”. Like other stars, the Sun
evolves with time, so that the radiative energy it provides was probably
25-30% smaller 4 billion years ago than it is today. In spite of the small
THE MIDDLE ATMOSPHERE AND ITS EVOLUTION 3
energy available, the planet’s average temperature was always higher

than the water frost point (based on the analysis of geological records). It
is probably the presence of high quantities of CO2 in the atmosphere (on
the order of 100 times the present value) and perhaps of other greenhouse
gases (i.e., methane, ammonia) at the early stage of the Earth’s evolution
that prevented the planet from being frozen (a condition required for life
to develop).
As time evolved and the energy provided by the sun increased,
the gradual removal of carbon dioxide from the atmosphere became
critical to avoid a “runaway greenhouse effect” (with extremely high
surface temperatures) as observed on Venus. This removal of CO2
was accomplished by weathering of calcium silicate (CaSiO3 ) minerals
by acidic CO2 -rich rainwater, leading to the formation of limestone
(CaCO3 ).
The evolution of atmospheric CO2 on geological time scales became
regulated by the storage of limestone (and of dead organic carbon which
appeared later as life developed) in the interior of the Earth’s crust
(as a result of plate tectonic motions) and the volcanic ejection of
CO2 produced by metamorphism in the Earth’s sediments (chemical
activation of limestone and dead organic carbon in high pressure and
high temperature environments).
The first atmospheric molecules of oxygen (O2 ) were probably formed
as a result of water vapor photolysis. This source, however, is insufficient
to account for the large percentage of oxygen found in the present
atmosphere. The evolution of oxygen was rather closely tied to the
evolution of life. Several hypotheses have been presented to describe the
emergence of life on the planet some 3.5-4.0 billion years ago. One of
them, supported by the historical laboratory experiment of Miller (1953)
and Miller and Urey (1959), suggests that organic molecules such as
amino acids and sugar (which may be the precursors of living matter)
could have been formed by spark discharges (i.e., lightning flashes) in
a mixture of gases containing methane, ammonia, hydrogen, and water
molecules. Organic molecules could also have been injected from space
by bombardment of micro-meteorites which include molecules containing
carbon, hydrogen, oxygen, and nitrogen. A more plausible hypothesis is
that life began in specialized regions of the ocean (e.g., in volcanic vents)
where high concentrations of sulfide may have favored the formation
of organic molecules. Some of the first living creatures survived by
fermentation. These were followed by chemoautotrophs which obtained
energy from chemical reactions, and finally photosynthetic organisms
began to produce oxygen in much larger amounts, eventually increasing
the oxygen content to contemporary levels. The evolution from an
oxygen-free, reducing atmosphere to an oxidizing atmosphere (one that
can sustain life for higher organisms) has been a special characteristic
of the Earth. Today, the most abundant gases in the atmosphere of our
neighboring planets, Mars and Venus, are not nitrogen (N2 ) and oxygen
(O2 ) as on Earth, but carbon dioxide (CO2 ).
Present photosynthetic life is protected from harmful solar radiation
by oxygen and ozone. Thus we must ask how these original creatures
survived and evolved to the present state even as they formed the
protective shield which their descendants would enjoy. It is possible
that they were a form of algae, protected by liquid water from the sun’s
rays. It has also been suggested that primitive microbes could have been
protected by layers of purine and pyrimidine bases, which absorb in the
ultraviolet range (Sagan, 1973).
In conclusion, it is clear that the evolution of the Earth’s atmosphere
depended on many factors, including its albedo, the biosphere (plant
and animal life), the oceans, the sun, and the composition of the solid
Earth. Physical processes such as volcanic eruptions together with
biological activity at the surface have contributed to the evolution of
the composition of the terrestrial atmosphere and of the Earth’s climate
over geological time. Lovelock and Margulis (1974), for example, have
argued that the Earth’s environment is controlled primarily by the biota
in a way that is favorable to life. Thus, the stability of our atmosphere
is a complex function of both organic and inorganic processes. Today,
many of these processes are disrupted by the human species. In the
rest of this chapter and throughout this volume, we discuss the natural
processes which control the present atmosphere, and assess the possible
impact which human activities may have upon it.
1.3 Anthropogenic Perturbations
The possible modification of the chemical composition of the

atmosphere, and the attendant climatic effects, is a problem which
attracts the attention of the international scientific community, decision
makers, and the public. Some of the processes which may produce these
modifications will be briefly identified in this section. A more detailed
analysis will be presented in Chapter 6.
We first consider the effects of the agricultural revolution and ask
whether the modification of the Earth’s ecosystems may influence the
atmosphere. In tropical regions, for example (Brazil, Central Africa,
Southeast Asia), large regions of forest and savannah are burned for
agricultural reasons. It has been estimated that 160000 km2 of wooded
land are destroyed each year (e.g., Kandel, 1980). The combustion
produces numerous chemical species, such as CO2 , CO, CH4 (and many
other organic compounds), H2 , N2 , N2 O, NO, NO2 , COS, CH3 Cl,
CH3 Br, CH3 I, HCN, CH3 CN, as well as aerosol particles (e.g., black
carbon; see Crutzen et al., 1979; Andreae and Merlet, 2001). Some of
these compounds eventually reach the middle atmosphere, where they
can influence the budgets of several minor species, particularly ozone.
Further, the introduction of modern agricultural techniques, especially
the intensive use of nitrogen fertilizers, has altered the natural nitrogen
cycle by increasing the fixation of this element in the form of ammonia,
amino acids, and nitrates. During nitrification and denitrification, part
of the nitrogen is emitted to the atmosphere in the form of N2 O rather
than N2 . N2 O provides the principal source of NO in the middle
atmosphere and accelerates the destruction of ozone in the stratosphere
(see Chapter 5).
The rate of nitrogen fixation on the Earth’s surface is not known
with great accuracy. The contributions of natural biological fixation,
as well as the effects of lightning and combustion must be considered.
It is estimated, however (Vitousek, 1994), that the annual amount of
nitrogen fixed naturally by the biosphere is between 50 to 200 Tg N/yr,
while fixation by combustion represents about 20 to 40 Tg/yr. The
contribution from nitrogen fertilizers probably exceeds 80 Tg N/yr,
and the fixation by legume crops is of the order of 30-50 Tg N/yr.
An additional 2-20 Tg N/yr are fixed by lightning in thunderstorms
and 20 Tg N/yr in internal combustion engines. Since human-induced
nitrogen fixation has increased with time, the effect of this artificial
source on the ozone layer (pointed out by Crutzen, 1974, and McElroy
et al., 1976) could become significant in the future.
The possible effects of nitrogen oxides on the middle atmosphere have
been very actively studied since Crutzen (1970) and Johnston (1971)
indicated that the injection of large quantities of these species in aircraft
exhaust might alter the protective ozone layer. Later, calculations
showed that the effect introduced by such aircraft depends directly
on the flight altitude: An injection of nitrogen oxide could contribute
to ozone production at altitudes below about 15 km (Brasseur et al.,
1998; IPCC, 1999), but to ozone destruction if the craft flew at higher
altitudes. It is estimated that current commercial aircraft release
approximately 0.5 Tg N/yr (a third of which is being released in the
stratosphere) and that the fleet of 500 high-speed civil transport aircraft
projected in the early 1970s would have injected 0.1 Tg N/yr (mostly as
NO) near 15-22 km altitude. Another source of nitrogen oxides which
perturbed the atmosphere in the 1950s and 1960s is the explosion of
powerful atomic bombs. Approximately 0.15 to 0.55 Tg N were injected
into the stratosphere during the Soviet tests in the fall of 1962. This
figure is a significant fraction of the 1 Tg N produced naturally from

N2 O oxidation.
The effects of industrial activities, especially the emission of gaseous
effluents and their possible effects on climate, have also attracted
considerable attention. The most well known problem is certainly that
of CO2 , but the cycling of this gas between the atmosphere, biosphere,
hydrosphere (clouds and oceans), and the lithosphere (solid Earth)
remains insufficiently well quantified. The emission of CO2 as a result
of combustion of fossil fuels (coal, oil) appears to be responsible for
the increase in the CO2 concentration which has been observed in the
atmosphere. Between 1960 and 1990 the total amount of CO2 in the
atmosphere increased from 2.53 to 2.81 × 1012 tons. Such an increase
influences the thermal budget of the atmosphere, both by cooling of the
upper layers due to an increase in the emission of infrared radiation to
space, and through heating in the lower layers due to absorption of the
infrared emission from the Earth’s surface and lower atmosphere (the
so-called greenhouse effect). Global models indicate that the globally
averaged surface temperature of the Earth should increase by 1.4 to
5.8◦ C over the period 1990-2100, depending on the adopted scenario for
future emissions (IPCC, 2001). Such changes would greatly influence the
polar ice caps and global precipitation processes. The exact magnitude
of the effect of increased greenhouse gas emissions is difficult to establish
because of the complicated feedbacks that exist between the atmosphere,
oceans, and biosphere. Observations indicate that a systematic increase
in the Earth’s mean surface temperature has occurred in the twentieth
century. Since the secular temperature variation is also dependent on
the aerosols released by combustion processes and perhaps on solar
variability, this trend cannot be unambiguously attributed entirely to
CO2 and other greenhouse gases.
Changes in the atmospheric temperature alter the rates of many
photochemical processes and therefore the concentrations of minor
species. For example, the increase in CO2 between 30 and 50 km
should lead to a small increase in the ozone content, according to current
models. Thermal perturbations can also influence atmospheric dynamics
(see Chapter 3).
The emission of chlorine and bromine compounds also is a matter
of interest, because stratospheric ozone is easily destroyed by these
halogen species (Stolarski and Cicerone, 1974; Molina and Rowland,
1974; Wofsy et al., 1975). Chlorine and bromine atoms are present in a
variety of industrial products dispersed in the atmosphere, but many of
these are rapidly destroyed at low altitudes and therefore do not reach
the middle atmosphere where the bulk of the atmospheric ozone layer
resides. Some industrial compounds do, however, have sufficiently long
lifetimes to represent a potential perturbation. This is the case, for

example, for halogenated hydrocarbons such as the chlorofluorocarbons
(CFCs), and specifically for trichlorofluoromethane (CFCl3 or CFC-
11), dichlorodifluoromethane (CF2 Cl2 or CFC-12), carbon tetrachloride
(CCl4 ), and methylchloroform (also called trichloroethane or CH3 CCl3 ).
CFCs are very stable compounds and, hence, have been used as
refrigerants, solvents, foam-blowing agents and, until the late 1970s, as
aerosol spray propellants. Methyl chloride (CH3 Cl) is the major natural
chlorocarbon of relevance for the stratosphere; it is produced in seawater
and released by biomass burning. As these long-lived organic molecules
penetrate into the stratosphere, they are photodecomposed and release
ozone-destroying agents. It is estimated that in the early 1990s when
the emissions of industrially produced CFCs represented approximately
1 Tg/yr, about 85% of the stratospheric chlorine was of anthropogenic
origin. There is strong evidence today that the ozone depletion observed
in the 1980s and 1990s, and specifically the formation of a large “ozone
hole” in Antarctica each spring since about the early 1980s, has been
caused by the rapid increase in the chlorine load of the stratosphere.
The CFCs are also believed to be responsible for the ozone depletion
recorded in the Arctic during many winters since the early 1990s.
International agreements (Montreal Protocol in 1987 and subsequent
amendments), as well as national regulations, have strongly limited the
production and the use of the CFCs. These chemical compounds have
been gradually replaced by partially halogenated hydrocarbons, and
specifically by hydrochlorofluorocarbons (HCFCs) and hydrofluorocar-
bons (HFCs). These alternative products are relatively easily destroyed
in the troposphere and hence their lifetimes are substantially shorter
than those of the CFCs (typically 1-10 yrs as opposed to 10-100 yrs). The
ozone depletion potential of the HCFCs is about an order of magnitude
smaller than that of the fully halogenated halocarbons. HFCs are not
a threat to ozone because they do not contain any chlorine or bromine
atoms.
Bromine, whose potential to destroy ozone (per atom) is about
50 times higher than that of chlorine, is released in the stratosphere
by photodecomposition of methyl bromide and other brominated organic
compounds, including the so-called halons (e.g., CF2 ClBr or Halon-1211,
and CF3 Br or Halon-1301). CH3 Br is produced by biological activity
in the ocean, biomass burning, automobile exhaust, and agricultural
usage (i.e., as a pre-planting and post-harvesting fumigant). The
anthropogenic source of CH3 Br represents about 40% of the total
emission. Halons, whose lifetime is several decades, are used in fire
extinguishers. Their production is now banned owing to their harmful
effects on stratospheric ozone.
Numerical models have been used to predict the potential ozone

depletion in response to the emission of halocarbons, based on different
plausible scenarios. All of these models indicate that the time required
for the middle atmosphere to respond to surface emissions of these
halocarbons is very long (several decades). In particular, even with the
measures taken to reduce or phase-out the emissions of the CFCs and
other halocarbons, it is expected that the Antarctic ozone hole will be
observed each spring (September-October) at least until the year 2040.
It should also be noted that the halocarbons are active in the infrared,
and contribute to the greenhouse effect.
The analysis of perturbations to the middle atmosphere must also
include natural processes, such as the effects of volcanic eruptions,
which produce large quantities of fine particles as well as water vapor
and SO2 , which eventually produces H2 SO4 and sulfate aerosols. The
amount of gas injected, the composition and the maximum altitude
of injection vary with the intensity of the eruption. Such events can
alter the budgets of some atmospheric constituents and are clearly
reflected in the middle atmospheric aerosol content. Particles also
provide sites for surface reactions to occur. Such heterogeneous reactions
may activate chlorine and enhance the depletion of ozone by industrially
manufactured halocarbons.
We have also noted the important role of the sun in establishing
the thermal and photochemical conditions in the Earth’s environment.
Variations in the solar output are caused by its 27-day rotation period,
the 11-year solar cycle, and by solar flares. Each of these processes may
account for part of the natural variability of the atmosphere.
The effects of human activities on the composition of the atmosphere
have been clearly detected on local, regional, and global scales.
Agricultural development and industrial activities are likely to influence
terrestrial ecosystems as well as the climate system. It is therefore
important to understand the behavior of the atmosphere, which is so
vulnerable to external perturbations and so essential to the climatic and
photochemical environment on Earth.
References
Andreae, M.O., and P. Merlet, Emission of trace gases and aerosols from biomass
burning. Glob Biogeochem Cycles: 15, 955, 2001.
Brasseur, G.P., R.A. Cox, D. Hauglustaine, I. Isaksen, J. Lelieveld, D.H. Lister, R.
Sausen, U. Schumann, A. Wahner, and P. Wiesen, European scientific assessment of
the atmospheric effects of aircraft emissions. Atmos Env: 32, 2327, 1998.
Crutzen, P.J., The influence of nitrogen oxide on the atmospheric ozone content. Quart
J Roy Met Soc: 96, 320, 1970.
Crutzen, P.J., Estimates of possible variations in total ozone due to natural causes and
human activity. Ambio: 3, 201, 1974.
Crutzen, P.J., L.E. Heidt, J.P. Krasnec, W.H. Pollock, and W. Seiler, Biomass burning
as a source of the atmospheric gases CO, H2 , N2 O, NO, CH3 Cl, and COS. Nature:
282, 253, 1979.
IPCC (Intergovernmental Panel on Climate Change), Aviation and the Global
Atmosphere. J. Penner et al. (eds.), Cambridge University Press, 1999.
IPCC (Intergovernmental Panel on Climate Change), Climatic Change 2001: The
scientific basis. J.T. Houghton, Y. Ding, D.J. Griggs, M. Noguer, P.J. Van de Linden,
X. Da, K. Maskell, and C.A. Johnson (eds.), Cambridge University Press, 2001.
Johnston, H.S., Reduction of stratospheric ozone by nitrogen oxide catalysts from
supersonic transport exhaust. Science: 173, 517, 1971.
Kandel, R.S., Earth and Cosmos. Pergamon Press, 1980.
Kasting, J.F., Earth’s early atmosphere. Science: 259, 920, 1993.
Lovelock, J.E., and L. Margulis, Atmospheric homeostatis by and for the atmosphere:
The gaia hypothesis. Tellus: 26, 2, 1974.
McElroy, M.B., J.W. Elkins, S.C. Wofsy, and Y.L. Yung, Sources and sinks for
atmospheric N2 O. Rev Geophys Space Phys: 14, 143, 1976.
Miller, S.L., A production of amino acids under possible primitive earth conditions.
Science: 117,7 528, 1953.
Miller, S.L., and H.C. Urey, Organic compound synthesis on the primitive earth. Science:
130, 245, 1959.
Molina, M.J., and F.S. Rowland, Stratospheric sink for chlorofluoromethanes: Chlorine
atom catalyzed destruction of ozone. Nature: 249, 810, 1974.
Moulton, F.R., On the evolution of the solar system. Astrophys J: 22, 165, 1905.
Sagan, C., Ultraviolet selection pressure on the earliest organisms. J Theo Bio: 39, 195,
1973.
Stolarski, R.S., and R.J. Cicerone, Stratospheric chlorine: A possible sink for ozone.
Can J Chem: 52, 1610, 1974.
Vitousek, P.M., Beyond global warming: Ecology and global change. Ecology: 75, 1861,
1994.
Walker, J.C.G., Evolution of the Atmosphere. McMillan Pub., 1977.
Wofsy, S.C., M.B. McElroy, and Y.L. Yung, The chemistry of atmospheric bromine.
Geophys Res Lett: 2, 215, 1975.
Chapter 2
CHEMICAL CONCEPTS IN
THE ATMOSPHERE
2.1 Introduction
Almost all of the constituents that are present in the middle

atmosphere undergo chemical and photochemical processes, which
greatly affect their spatial and temporal distributions. Chemical
reactions can occur in the gas phase, in the condensed phase, or at
the interface between these phases. The energy provided by the Sun can
break the chemical bonds of many species, producing reactive fragments
which interact with other compounds. Most of these fragments are free
radicals (particles containing one or more unpaired electrons). Gas
phase reactions typically involve at least one “free radical” species,
while condensed phase reactions can include molecules only. Reactions
occurring in the middle atmosphere are discussed in detail in Chapter 5.
Rate constants for chemical reactions are measured in the laboratory.
Quantitative values of these rate constants are regularly compiled by
the Jet Propulsion Laboratory (see e.g., JPL, 2002). The absorption of
solar energy also plays an important role in the thermal budget and
dynamical properties of the middle atmosphere. To understand the
aeronomy of the middle atmosphere, we must first understand some
of the basic principles of spectroscopy, photochemistry, kinetics and
thermodynamics. In this chapter, we attempt to outline how these enter
into the chemistry of the middle atmosphere, and present examples of
their use. The discussion presented here will be very elementary; the
reader may wish to examine the bibliography for detailed treatments.
See, for example, the comprehensive textbook by Finlayson-Pitts and
Pitts (1999). Some useful definitions are provided in Box 2.1.
11
2.2 Energy Levels of Molecules
The motions of molecules can be decomposed into translational,

rotational and vibrational motions. Radiation, once absorbed by a
molecule, affects only its rotational and vibrational states. Therefore,
translation motions will not be further considered here, even though we
should note that once absorbed, the incoming energy can be converted
into translational energy (kinetic energy) and hence transformed into
heat through collisions with the medium. The absorption of radiation
can also lead to changes in the electron distribution and hence in
the electronic state of the molecules. Transitions between rotational,
vibrational and electronic states are quantified, and obey certain
selection rules, which can be derived from quantum mechanics principles
(e.g., Steinfeld, 1978). These selection rules describe the ease with which
transitions occur. It is common to refer to transitions as “forbidden”
or “allowed”. In practice, even highly “forbidden” transitions occur
under certain circumstances, although much more slowly than “allowed”
transitions. Therefore, we will refer to transitions here as “unfavorable”
or “allowed”.
The energy associated with transitions between electronic states is
relatively large, of the order of electron-volts (eV), and the photons
capable of inducing such transitions are generally in the ultraviolet
and the visible. The energy associated with vibrational transitions is
significantly smaller, and corresponds to the near-infrared (wavelengths
of typically 10 µm). In the case of rotational transitions, the energy
involved is even smaller, and corresponds to the far-infrared and
microwave regions (wavelengths of the order of 10 mm).
Box 2.1 Some Useful Definitions

(Adapted from Ebbing, 1990)
Photon Quantum of electromagnetic radiation with energy proportional
to the observed frequency ν of the light: E = hν (where h
is Planck’s constant).
Atom An extremely small particle of matter that retains its identity
during chemical reactions.
Nucleus Central core of an atom containing one or more positive charges.
Proton A particle found in the nucleus of an atom. Its electric charge is
equal in magnitude, but opposite in sign, to that of an electron
and its mass is 1836 times that of the electron.
Neutron A particle found in the nucleus of an atom. Its mass is almost
identical to that of a proton. It has no electric charge.
CHEMICAL CONCEPTS IN THE ATMOSPHERE 13
Box 2.1 (Continued)

Electron A light particle (mass = 9.1096 × 10−31 kg) that carries a
unit negative charge (1.6022 × 10−19 C) and exists in
the region around the positively-charged nucleus.
Element A substance that cannot be decomposed by any chemical
reaction into simpler substances. A type of matter composed
of one kind of atom.
Molecule A group of atoms that are tightly connected by attractive
forces (that are chemically bonded together).
Isomers Compounds of the same molecular formula but with different
arrangements of atoms.
Covalent A chemical bond formed by the sharing of a pair of electrons
Bond between atoms in a molecule.
Free Highly reactive species (e.g., H, OH, NO, NO3 , Cl) with
Radical one or more unpaired electrons
Mole The quantity of a given substance that contains as many
molecules as the number of atom in exactly 12g of carbon
–12. The amount of substance containing 6.0222137 × 1023
molecules.
Compound A substance composed of two or more elements chemically
combined in fixed proportions.
Ion An electrically-charged particle obtained from an atom or a
chemically-bonded group of atoms by adding or removing
one or more electrons.
Plasma An electrically neutral gas made up of ions and electrons.
Gas The form of matter that is an easily compressible fluid.
Vapor The gaseous state of matter.
Liquid The form of matter that is a relatively incompressible fluid;
it has a fixed volume, but no fixed shape.
Solid The form of matter that is characterized by rigidity; it is
relatively incompressible and has a fixed shape and volume.
Condensation The changing of a gas to either the liquid or the solid state.
Vaporization The changing of a solid or a liquid to the vapor state.
Sublimation The changing of a solid to the vapor state.
Fusion The changing of a solid to the liquid state.
Aerosols Liquid droplets or solid particles dispersed throughout a gas.
Nucleation The formation of particles in a supersaturated vapor (or in
the air).
Coagulation A process in which particles grow by collisions.
Diffusion A process whereby a gas spreads out through another gas to
occupy the space with uniform partial pressure.
Chemical A change in which one or more kinds of matter are trans-
Reaction formed into a new kind of matter or several new kinds of
matter. It consists of the rearrangement of atoms present in
the reacting substances to give new chemical combinations
characterizing the substances formed by the reaction.
Box 2.1 (Continued)

Chemical The state reached by a reaction mixture when the rate of
Equilibrium forward reaction becomes equal to the rate of reverse
reaction.
Exothermic A chemical reaction or physical change in which heat is
Process released.
Endothermic A chemical reaction or physical change in which heat is
Process absorbed.
Internal The sum of the kinetic and the potential energies of the
Energy (U) particles making up a system.
Enthalpy (H) An extensive property of a substance that can be used to
obtain the heat absorbed or released by a chemical reaction
or physical change at constant pressure. It is defined as the
sum of the internal energy (U) and the product of the
pressure and the volume of the system (PV) H = U + PV.
Entropy (S) A thermodynamic quantity that is a measure of the
randomness or disorder of a system.
Free A thermodynamic quantity defined by G = H – TS where
Energy (G) H is the enthalpy, S the entropy, and T the absolute
temperature.
Oxidation A chemical reaction in which a compound or a radical
loses electrons.
Reduction A chemical reaction in which a compound or a radical gains
electrons.
Catalyst A substance that increases the rate of reaction without
being consumed in the overall reaction.
Figure 2.1.a presents a typical potential energy diagram for a

hypothetical diatomic molecule XY. The bond between the two atoms X
and Y can be viewed as an oscillating spring, in which the internuclear
distance varies with a frequency that is determined by the strength of the
bond and the masses of the atoms. In the idealized case of a harmonic
oscillator, the restoring force acting on the atoms is proportional to the
displacement from the equilibrium position located at the minimum of
the potential energy curve, which is represented by a parabola. Quantum
mechanics (Schrödinger’s
¨ equation solved for the harmonic oscillator)
shows that, in this case, the vibrational energy is quantized with energy
levels equally spaced. The corresponding quantum numbers v are equal
to 0 (fundamental state), 1, 2, 3, etc. Transitions can only modify the
energy level by one quantum at the time. The selection rule for the
harmonic oscillator is therefore ∆v = ±1. For a non-idealized oscillator,
the curve representing the potential energy becomes asymmetric and is
Figure 2.1.a. Typical potential energy diagram for a hypothetical diatomic

molecule XY. The lower curve corresponds to an anharmonic oscillator, showing
a strong repulsive branch for small internuclear distances and an attractive
branch allowing, however, the dissociation of the molecule, when the molecular
bond is stretched beyond a given internuclear distance. Electronic states
corresponding to XY and XY∗ exhibit maximum stability at the minimum
of the potential energy curve, so that this state represents a stable bound
configuration of the molecule. In these bound states, vibrational levels ((v=0,
1, 2, etc.) are represented by horizontal lines. Rotational levels are not shown.
The electronically excited XY∗∗ state, represented by the uppermost curve, is
repulsive for all values of the internuclear distance, and is thus unstable.
often called the Morse curve (see Figure 2.1.a). The potential energy
increases sharply towards smaller internuclear distances, as charge
repulsion becomes more intense. For large internuclear distances, the
molecular bond is stretched and can break; in this case, the molecule
dissociates. In the non-idealized case, weak overtone transitions with
∆v = ±2, ±3, etc. are possible and the vibrational energy levels are
not equidistant anymore, but are more closely spaced with increasing
quantum number v. Note, in addition, that vibrational transitions occur
only if the electric field of the radiation can interact with an oscillating
electric field produced by the dipole moment of the vibrating molecule.
Only heteronuclear diatomic molecules such as NO or CO produce such
an oscillating dipole moment, while homonuclear species like O2 and N2
do not undergo vibrational transitions. Since the energy corresponding
to a vibrational transition is relatively large compared to the thermal
energy, most molecules present in the atmosphere are found to be in
their lowest vibrational state (v = 0). Vibrationally excited species
can be produced by photon absorption, collision with an atom or a
molecule, or as a result of a chemical reaction (see e.g., reaction 2.95).

De-excitation can result from a collision or from spontaneous radiative
emission from level u to level . The probability of this latter process
is provided by the Einstein coefficient Au, , whose value is determined
in the laboratory. Vibrational energy transitions play a particularly
important role in the thermal balance of the middle atmosphere, because
they are the fundamental processes responsible for most of the infrared
cooling occurring there (see Chapter 4).
Diatomic molecules with a permanent dipole moment (i.e., heteronu-
clear molecules) produce an oscillating electric field during their rotation,
resulting in light absorption and rotational transitions. These transitions
occur between quantified energy levels (with the corresponding quantum
numbers J equal to 0, 1, 2, 3, etc.). In the case of an idealized rigid rotor,
the selection rule for such transitions is given by ∆J = ±1. Since the
energy corresponding to rotational transitions is small compared to that
of vibrational transitions, and to the thermal energy of the molecules, the
population of the rotational energy is not confined to the fundamental
state, but is generally significant for levels above J = 0.
Since molecules vibrate and rotate simultaneously, combined
vibration-rotation transitions (and spectra) need to be considered. In
the idealized harmonic oscillator, rigid rotor case, the only significant
processes to be considered are the transitions between vibrational level
v = 0 and level v = 1, together with a number of rotational transitions.
An example of such system is shown in Figure 2.1.b. The intensity
of the corresponding lines observed in the absorption spectra reflects
the relative population of the rotational energy levels. The transitions
associated with the selection rule ∆J = +1 (i.e., found on the high
energy side of the spectrum) produce the so-called R branch, while the
transitions associated with ∆J = −1 (low energy side of the spectrum)
lead to the so-called P branch.
In the case of polyatomic molecules, such as CO2 , H2 O, O3 , etc.,
the principles discussed above still apply, but the spectra become more
complex. Polyatomic molecules do not rotate only about one single axis,
but about three mutually perpendicular axes. In addition, the number
of vibrational degrees of freedom is also increased.
Vibrational quanta are distributed among the available vibrational
modes of the molecule, which correspond physically to stretching and
bending the chemical bonds of the molecule. These are conventionally
referred to as ν1 , ν2 , etc.
Figure 2.2 presents a schematic diagram of the possible vibrational
modes for diatomic, linear triatomic, and bent triatomic molecules. The
vibrational term symbol provides an ordered list of quantum numbers
Figure 2.1.b. Schematic diagram of the energy levels and vibration-rotation

transitions (after Herzberg, 1950). The lower lines (v = 0) correspond to the
fundamental vibrational state with different rotational states (J = 0, 1, 2, 3, 4).
The first vibrationally excited state (v = 1) with related rotational states
(J = 0 to 4) are shown by the upper lines. The transitions corresponding to
the P and R branches are indicated. Transition ∆J = 0 is not allowed in the
case of the idealized rigid rotor, harmonic oscillator for a linear molecule, but
can be observed (Q-branch) in other cases. The notations are from Herzberg
(1950).
Figure 2.2. Vibrational modes of diatomic and triatomic molecules.

for each mode. For example, (010) means that the quantum number is
zero for ν1 , one for ν2 and zero for ν3 . As we shall see in Chapter 4, the
15µm ν2 band of CO2 plays a dominant role in cooling throughout the
middle atmosphere.
The absorption of visible and ultraviolet radiation can lead to
electronic transitions. The curves labelled XY∗ and XY∗∗ in Figure 2.1.a
correspond to electronically excited states of the hypothetical molecule
XY considered earlier. Electronically excited states may be repulsive
at any internuclear distance (as in the case of XY∗∗ ), or may have a
stable bound configuration (XY∗ ) like the ground state, with associated
vibrational and rotational levels. If an electronic transition occurs along
with simultaneous vibrational and rotational transitions, the selection
rules are not as restricted as those described above. Transitions from
the fundamental vibrational level v = 0 to high levels (v > 1) becomes
possible. Rotational transitions with unchanged quantum number J
(∆J = 0) are observed and produce a Q-branch in the corresponding
spectrum.
The electronic states of molecules and atoms are expressed by
quantum numbers, which represent the total orbital angular momentum
and the net spin of the electrons. The corresponding symbols and the
selection rules for electronic transitions are discussed in Sections 2.2.1
and 2.2.2, respectively. Different electronic configurations of the same
particle generally have very different reactive properties; for example, the
excited O1 D atom behaves very differently upon collision with a water
molecule than does the ground O3 P atom. Photochemical processes
involving absorption and emission of photons induce transitions from
one state to another, and these occur more readily between certain
electronic states than others. The transitions between electronic states,
which result from absorption of relatively high energy photons, serve two
important purposes in the middle atmosphere. They represent the major
source of heating, as will be discussed in Chapter 4, and they initiate
photochemical processes (e.g., photodecomposition or photolysis and
photoionization) which play a major role in establishing the distributions
of minor constituents (this is described in detail in Chapters 5 and 7).
2.2.1 Term Symbols
In general, term symbols indicate both spin angular momentum (S

for atoms, Σ for diatomic molecules) and orbital angular momentum (L
for atoms, Λ for diatomic molecules). The term symbol is written 2S+1 L
for atoms and 2Σ+1 Λ for molecules. We will not concern ourselves here
with the derivation of L and Λ. Part of the reason for this is the fact
that the selection rules involving orbital angular momentum are often
not strictly followed in actual transitions; the restrictions regarding spin
are the ones which tend to be most rigidly followed in practice, and we
shall concentrate only on these. For increasing L from 0 to 6, the term
takes on the values S, P, D, F, G, H. For increasing Λ, the Greek symbols
Σ, Π, ∆, and Φ are used. See, for example, Karplus and Porter (1970).
The spin terms can be readily evaluated given knowledge of basic
atomic and molecular orbital theory (see Moore, 1962; Karplus and
Porter, 1970). The ground state of the oxygen atom is conventionally
described as
(1s)2 (2s)2 (2p1 )2 (2p2 )1 (2p3 )1
The superscripts denote the number of electrons in each atomic

orbital. Electrons must be paired (i.e., of equal and opposite spin)
in atomic orbitals containing two electrons. Paired electrons do
not contribute to the atomic or molecular spin angular momentum.
Hund’s rule of maximum multiplicity states that in the lowest energy
configuration, the electrons must be spread over as many available
orbitals of equal energy as possible, in order to maximize the spin
multiplicity. Since three 2p orbitals are available in which 4 electrons
must be distributed, the lowest electronic state therefore has two
unpaired electrons in each of two 2p orbitals. Each unpaired electron
contributes 12 to the spin angular momentum. Thus 2S + 1 = 3 for the
oxygen atom, and the term symbol is 3 P.
The oxygen molecule has the ground configuration
(σg 1s)2 (σu 1s)2 (σg 2s)2 (σu 2s)2 (σg 2p)2 (πu 2p)4 (πg 2p)2
The two πg 2p electrons are unpaired, so that 2Σ + 1 = 3, and the

ground state is X3 Σ−g . The X is used to denote that this is the ground
state of the molecule; first and second excited states are indicated by
A, B, or a, b, etc. The subscripts “g” (from the German “gerade” =
even) and “u” (“ungerade” = uneven) used for homonuclear diatomic
molecules (e.g., O2 ) refer to cases where the wavefunction representing
the molecular state keeps the same sign or changes sign, respectively,
through the center of symmetry of the molecule. The superscripts
“+” and “–” apply only to Σ states. Symbol “+” is used when the
wavefunction is unaltered by reflection through a plane passing through
the 2 nuclei of the molecule. Symbol “–” is used if the wavefunction
changes sign. More details regarding the significance of the term symbols
can be obtained, for example, in Herzberg (1950).
Polyatomic molecules have different term symbols, but again the spin
multiplicity is given by the left-hand superscript. For example, the
ground state of the ozone molecule is denoted by the term 1 A1 . Note that
these terms should not be confused with the excited states of diatomic
molecules, which can also contain A, B, etc. in their term symbols.
2.2.2 Selection Rules for Electronic Radiative Processes
The importance of the term symbols lies in their use in predicting the
ease with which photon absorption
A + hν → A∗ (2.1)
and radiative decay can occur.
A∗ → A + hν (2.2)
The selection rules for radiative processes permit us to evaluate the ease
or difficulty with which a particular transition can proceed. The basic
rules are as follows:
∗ Spin is unchanged in an allowed transition, for both atoms and
molecules. This is generally the most important consideration in

establishing the transition probability.
∗ L or Λ changes of 1 or zero are allowed, except in the case of atoms
if 0 → 0 (S → S). Changes of L or Λ of more than one (i.e., D → S)

are unfavorable.
∗ For molecules with g and u subscripts, g → u transitions are allowed,
but u → u or g → g transitions are difficult.

∗ For Σ states, Σ+ → Σ+ and Σ− → Σ− transitions are allowed, while
Σ+ → Σ− transitions are unfavorable.
Examples of transitions which are of importance in the middle

atmosphere are, for example,
O2 (1 ∆g ) → O2 (3 Σ−
g ) + hν (2.3)
This process is unfavorable because it involves a change in spin, a change
in Λ of two, and a g → g transition. Thus its probability is relatively
low, which is reflected in a time constant for radiative decay of almost an
hour. Similarly, the cross section for production of O2 (1 ∆g ) by photon
absorption from the ground state
O2 (3 Σ−
g ) + hν → O2 ( ∆g )
1
(2.4)
is small.
2.3 Thermodynamic Considerations
2.3.1 Enthalpy, Entropy, and Free Energy
The basic principles governing the reactivity of chemical substances

lie in the thermodynamic properties of these species. In order to evaluate
the feasibility of chemical processes, the enthalpy of reaction must be
examined. The enthalpy of formation of chemical constituents or “heat
content” represents the energy required to make and break the chemical
bonds which compose that substance, starting only from the elements in
their most stable form (e.g., O2 , N2 , S, etc.), which are conventionally
assigned an enthalpy of formation of zero. Consider, for example, the
formation of two ground state oxygen atoms, starting from an oxygen
molecule. The energy required to break the O2 bond is about 120
kcal/mole at room temperature, so that the enthalpy of formation of
each atom is about 60 kcal/mole (see Table 2.1). Let us examine a
general reaction:
A+B→C+D (2.5)
The enthalpy change associated with such a reaction is given by the
difference between the enthalpy of formation of the products and the
enthalpy of formation of the reactants:
∆H0R = ∆H0f (C) + ∆H0f (D) − ∆H0f (A) − ∆H0f (B) (2.6)
where the superscript 0 refers to the standard state, which is 298.15 K
by convention. If ∆HR0 is negative, heat is liberated by the reaction,
and it is called exothermic. If ∆HR0 is positive, heat must be
supplied in order for the reaction to proceed, and it is then called
endothermic. Exothermic reactions can proceed spontaneously in the
atmosphere, while endothermic processes require additional energy,
and this necessary external energy may not be available. Exothermic
reactions may nevertheless be non-spontaneous.
It must be noted that ∆H is a function of temperature. Tabulations
of ∆H generally refer to ∆H298.15K . The temperature dependence of
∆H is given by
T
∆HT = ∆H0298.15K + Cp dT (2.7)
298.15K
where T is the temperature and Cp is the heat capacity at constant

pressure (cal/deg mol). This correction is generally small over the range
of atmospheric temperatures so that the tabulated values can be applied
as a first approximation.
Entropy may also influence the spontaneity of chemical reactions.

In general terms the entropy represents the disorder or randomness
associated with a particular process. The Gibbs free energy of reaction
includes both the enthalpy and entropy associated with chemical
processes, and is defined as
∆G0 = ∆H0 − T∆S 0 (2.8)
and the ∆G0R is defined similarly to ∆HR0 :
∆G0R = ∆G0f (C) + ∆G0f (D) − ∆G0f (A) − ∆G0f (B) (2.9)
In the case of an elementary reaction (i.e., reaction without energy
barrier associated with the formation of an intermediate complex), if
∆G0R is negative, then the reaction can proceed spontaneously at the
temperature under consideration. Values of ∆G 0 , ∆H 0 , S0 and Cp0 for
most atmospheric species can be found in chemical handbooks.
One of the most important uses of thermodynamic data is the
evaluation of the feasibility of proposed chemical reactions, which have
perhaps never been studied in the laboratory. Let us imagine, for
example, that the reaction of NO3 with H2 O might be an important
source of HNO3 :
NO3 + H2 O → HNO3 + OH (2.10)
Is this reaction possible? This question can be answered immediately
by examining the Gibbs free energy of reaction:
∆G0R = ∆G0HNO3 + ∆G0OH − ∆G0NO3 − ∆G0H2 O (2.11)
which equals +17.8 kcal/mole. Therefore, the reaction is thermody-
namically impossible without an external source of energy; we need not
await laboratory studies to rule out the possibility of its occurrence in
the atmosphere.
Another important use of thermodynamic data in atmospheric
applications is the evaluation of the exothermicity of chemical reactions
which may play a role in the thermal budget of the atmosphere.
Consider, for example, the addition of atomic oxygen to molecular
oxygen:
O + O2 + M → O3 + M (2.12)
where M can be any arbitrary atom or molecule (see the discussion of
termolecular reactions below). We may write
∆H0R = ∆H0O3 + ∆H0M − ∆H0O − ∆H0O2 − ∆H0M
(2.13)
∆H0R = ∆H0O3 − ∆H0O
Adopting numerical values from thermodynamic tables, one finds ∆HR0
= 34.1 – 59.5 = –25.4. Thus this reaction is exothermic, liberating
25.4 kcal per mole. We shall see later in Chapter 4 that this process
can serve as an important source of heat in the upper part of the middle
atmosphere.
An endothermic reaction can proceed if the required energy is
provided by some external source, such as a photon. The energy of
a photon of wavelength λ is given by
hc
E= (2.14)
λ
where h = 6.62 × 10−34 Js is Planck’s constant, and c = 2.998 × 108
m s−1 is the speed of light in a vacuum. Other commonly used units
are the frequency, ν, which is related to λ by
c
ν= (2.15)
λ
and the wavenumber ν̄, given by
1 ν
ν̄ = = (2.16)
λ c
Thus consider, for example, for the photolysis of molecular oxygen to
form two O3 P atoms:
O2 + hν → 2O(3 P) (2.17)
hc hc
∆HR = 2(59.55) − = 119.10 kcal/mole − (2.18)
λ λ
This process requires that at least 119.10 kcal/mole, or 8.27 ×
10−19 J/molecule, be provided by the photon. This corresponds to a
wavelength λ ≤ 240 nm.
2.3.2 Solubility of Gases in Liquid Particles
The atmosphere is not only a mixture of gases. It also contains a

variety of tiny liquid or solid particles, commonly referred to as aerosols.
Atmospheric particulate matter may consist of a large variety of species;
in the lower stratosphere, however, the most abundant aerosol particles
are composed of highly concentrated aqueous sulfuric acid droplets. In
polar regions during winter, very sparse clouds, called polar stratospheric
clouds (PSCs) are also observed.
The chemical composition of stratospheric particles at a given altitude
is governed by local thermodynamic conditions, and specifically is a
complex function of temperature and water vapor concentration. The
solubility of gas-phase compounds into liquid aerosol particles
X(gas) X(aqueous) (2.19)
is governed by an equilibrium usually expressed as

[X(aqueous)]
Hx = (2.20)
px
where [X(aqueous)] is the concentration (moles per liter) of X dissolved
in the aqueous phase and px (atm) is the partial pressure of X in the gas-
phase. The equilibrium constant Hx is called Henry’s law constant and
is usually expressed in M atm−1 , where M (= mole/) is the molarity.
Henry’s law constants are measured in the laboratory for different
temperatures. It can be shown that the solubility of gases increases with
decreasing temperature. If a dissolved compound dissociates in water
(e.g., acids), the total solubility of the gas is increased above what would
result from the physical solubility alone. In this case, the solubility is
usually expressed as an effective Henry’s law constant which varies with
the pH of the solution and the equilibrium constants associated with the
dissociation mechanisms involved.
2.4 Elementary Chemical Kinetics
The preceding section describes the thermodynamics associated with

determining whether a reaction can proceed spontaneously at a given
temperature. The next question one might wish to pose is how fast
such a reaction will occur. That is the concern of chemical kinetics.
The concentrations of almost all atmospheric chemical species depend
critically on reaction kinetics, which determines the rates at which these
constituents are produced or destroyed. The observed density of most
chemical species in the atmosphere is dependent on the balance between
the rates of photochemical production and destruction, and the rate of
atmospheric transport processes. The rate at which a chemical reaction
occurs varies considerably from one reaction to another and must be
determined in the laboratory. For a reaction between species A, B, C,
. . ., expressed as
aA + bB + cC + . . . → mM + nN + . . . (2.21)
where coefficients a, b, c, etc. are the smallest possible integers that
provide the correct stoichiometry, the reaction rate R can be expressed
by the general expression
R = f [(A), (B), (C), . . . ] (2.22)
where (A) represents the concentration of species A. As shown by
experimental studies, this function may be complex, but may often be
of the form
R = k(A)α (B)β (C)γ . . . (2.23)
where k is the rate coefficient or rate constant for this particular

reaction. For many chemical processes, factor k varies with temperature.
Exponents α, β, γ . . . are independent of concentration and time, and are
referred to as the order of the reaction with respect to reactants A, B,
C, . . ., respectively. The overall order of the reaction is given by the sum
α + β + γ + . . . Note that α, β, γ, . . . are not always the stoichoimetric
coefficients a, b, c, . . . as shown by laboratory experiments. Different
experimental methods have been developed to derive reaction rates, but
will not be presented here. The parameters describing the kinetics of
the reactions are generally derived by measuring the concentration of
the reactants or products as a function of time. The rate of change
of a reactant or of a product is related to the rate of reaction R
(molec cm−3 s−1 ) by
1 d(A) 1 d(B) 1 d(M) 1 d(N)
R=− =− = ... = = + . . . (2.24)
a dt b dt m dt n dt
For example, for reaction
(k); O + O → O2
where k is the corresponding rate constant, the reaction rate is R = k
(O)2 and the changes with time in the concentration of O and O2 are
given by
d(O)
= −2k(O)2 (2.25a)
dt
d(O2 )
= k(O)2 (2.25b)
dt
Reactions involving decomposition of a single particle are called first
order or unimolecular; reactions involving the collision between two
particles are second order, two body, or bimolecular; processes involving
the collision between three particles are referred to as third order, three
body, or termolecular.
For the special case of a unimolecular (first order) process:
(ku ); A→C+D (2.26)
the rate is given by
d(A)
R=− = ku (A) (2.27a)
dt
where ku is the reaction rate constant (expressed in s−1 for a
unimolecular process) and (A) denotes the concentration of species A
(molec cm−3 ). Rearranging this equation, we can write
d(A)
= −ku dt (2.27b)
(A)
or
ln (A) = −ku t + C (2.27c)
where C is the constant of integration. Assuming that initially (A) =
(A)o at time t = 0, the evolution with time of number density (A) can
be expressed as
(A) = (A)o e−ku t (2.27d)
The time required for the concentration of A to decrease to 1/e of its
initial value, if this were the only reaction process occurring, is therefore
equal to 1/ku . This defines the chemical lifetime (or e-folding lifetime)
of species A. The concept of lifetime is an important one in atmospheric
applications. If, for example, we wish to know whether an atmospheric
species is likely to be directly affected by transport processes, we
can compare its chemical lifetime to the time scale appropriate to
transport. If chemistry is much faster than transport for that particular
species, then the direct effects of transport can be neglected to a first
approximation. This will be discussed in more detail in Chapters 3 and
5.
Referring to the general bimolecular equation A + B → products, the
rate at which such a reaction proceeds is described by
d(A) d(B)
R=− =− = kb (A) (B) (2.28)
dt dt
where kb represents the reaction rate constant (expressed in cm3
molec−1 s−1 for a two body reaction).
In the case of a termolecular process
A + B + C → products (2.29)
the reaction rate is expressed as
d(A)
R=− = kt (A)(B)(C) (2.30)
dt
where the rate coefficient is now expressed in cm6 molec−2 s−1 . Very
often, one of the reactants is not directly involved in forming the reaction
products, but assists in promoting the reaction by transferring energy
through collisions with the other reactants. This reactant, conveniently
termed M, is called the third body and in most cases represents an O2 or
N2 molecule.
Second and third order reactions (A + B → products and A + B +
C → products are commonly reduced to a pseudo first order form to
examine the chemical lifetime. For example,
(A) ≈ (A)o e−kb (B)t or (A) ≈ (A)o e−kt (B)(C)t (2.31)
and the lifetime is:

1 1
τA = or . (2.32)
kb (B) kt (B)(C)
These expressions can be used only if the concentrations of species B and
C can be considered constant over the time τA . This is not always the
case, since B and C may also be undergoing production and destruction
reactions that introduce a time dependence in their densities over the
time τA .
In a system like the atmosphere, in which numerous gases are
interacting simultaneously, chemical change usually occurs through
a network of interacting steps (reaction mechanism). In this case,
examination of the lifetime and the rate of change of a species (e.g.,
species A) must involve the sum of all of its possible production and loss
processes, i.e., assume the following reactions
(k1 ); A + B → Products (2.33a)
(k2 ); A + C + M → Products (2.33b)

(k3 ); A + F → Products (2.33c)
(k4 ); G + H → A + Products (2.33d)
d(A)
= −k1 (A)(B) − k2 (A)(C)(M) − k3 (A)(F) + k4 (G)(H) (2.34)
dt
1
τA = (2.35)
k1 (B) + k2 (C)(M) + k3 (F)
If this lifetime is short compared to transport times and if the densities
of B, C, F, G, and H are not changing over the time scale considered,
then photochemical stationary state (also called steady state) can be
assumed; this implies

that A is in

instantaneous balance as determined
by its sources ( Pi ) and sinks ( Li ):
i i
d (A)
=0= Pi − Li (A) (2.36)
dt i i

Pi
(A) = i (2.37)
i Li
In the particular case described by mechanism (2.33),
k4 (G)(H)
(A) = (2.38)
k1 (B) + k2 (C)(M) + k3 (F)
In some cases, the effective rate of a reaction can be increased by
a so-called catalytic cycle of intermediate reactions. Such a cycle is
initiated by a substance that is not consumed in the overall process,

but is constantly regenerated. This is the case, for example, for the
destruction of ozone
O3 + O → 2 O2 (2.39)
which is accelerated in the presence of a catalytic agent X (which can
be H, OH, NO, Cl, Br) through the following catalytic cycle
X + O3 → XO + O2 (2.40a)
XO + O → X + O2 (2.40b)
Overall: O3 + O → 2 O2
Other cycles including more than two reactions are known to destroy
ozone efficiently (see Chapter 5). Catalytic cycles can be very efficient,
even if the concentration of the catalyst X is several orders of magnitude
smaller than that of ozone. The efficiency of such a cycle regarding the
destruction of ozone is given by the number of times the cycle repeats
itself before the catalyst is eventually lost, usually by conversion to a
stable reservoir species (e.g., NO and NO2 are converted into HNO3 ).
The rate limiting step for the destruction of ozone in a cycle like (2.40)
is provided by the reaction that ultimately determines the rate at which
ozone is destroyed. Such a step can be identified, provided that all
reactions involved are known.
2.4.1 Collision Theory of Bimolecular Reactions
Simple collision theory provides a useful conceptual framework to

understand two body reactions. It cannot be applied to first and third
order reactions without considerable modifications.
Let us again examine the reaction of A and B. Clearly the reaction
cannot occur more rapidly than the rate at which A and B collide. If
we assume that the particles behave as hard spheres, then the collision
rate constant between these two particles is defined by
kAB = σvr (2.41)
where σ is the collision cross section and vr is the relative particle speed
(thermal velocity). In the hard sphere approximation, the collision cross
section is given by
σ = π (rA + rB )2 (2.42)
where rA and rB are the radii of particles A and B. For typical molecular
radii, the value of σ is about 10−15 cm2 . The mean relative speed, vr is
1
8kT (mA + mB ) 2
vr = (2.43a)
πmA mB
where mA and mB are the respective molecular masses of A and B,
and k is Boltzmann’s constant. For typical masses, the mean relative
speed is about 4 × 104 cm s−1 , and the value of kAB is about 4 ×
10−11 cm3 molec−1 s−1 . We can also see that the temperature dependence
1
of the rate constant thus defined is T 2 . Note that if the mass of one
of the colliding particles is much larger than that of the other particle
(e.g., mB >> mA ), the velocity vr can be approximated as
1
8kT 2
vr = (2.43b)
π mA
This expression is typically used to describe the collisions between
molecules (mass mA ) with aerosol particles (mass mB ).
Unfortunately, the actual interactions between reacting gases are
considerably more complex than this and not all two-body rate constants
are equal to 4 × 10−11 (in fact few of them do conform to this simple
description). The reasons for this lie in the differences between the
actual interactions between reacting molecules and the “hard sphere”
assumption made above. In general, particles form an intermediate
species during reaction, and this intermediate is often called an activated
complex. Indeed, several detailed theories of chemical interactions
achieve successful descriptions of certain processes simply by considering
the energy modes available to the activated complex (e.g., RRK theory,
see Rice and Ramsperger, 1927; Weston and Schwarz, 1972). Here
we simply note that formation of such a complex generally involves
an enthalpy of reaction, such that the overall process can be crudely
represented as shown in Figure 2.3. In general, a certain “energy of
activation” is associated with forming the activated complex. If the
reactants have sufficient energy to surmount the barrier involved in
complex formation, then products will be formed.
The energy distribution of particles is often expressed as a Maxwell-
Boltzmann distribution:
3/2
dn∗ m 2
= 4π c2 e−(mc /2+p )/kT dc (2.44)
no 2πkT
(see, e.g., Castellan, 1971) where p is the potential energy of the
particle, and where dn∗ is the number of particles per cubic centimeter,
which have velocities from c to c + dc, and m is the particle mass. Here
k is the Boltzmann constant. The exponential term accounts for both
the kinetic and potential energies of the particle. For a given potential
energy, the kinetic energy distribution depends only on temperature T.
Given the form of this expression and the reaction barrier discussed
above, it is not surprising that the experimentally derived kinetic rate
expression for many processes is:
kreaction = A e (−Eact /RT) (2.45)
In this expression, called the Arrhenius equation, Eact represents the
activation energy, R is the gas constant, and A is the “pre-exponential”
factor (assumed to be constant over a limited temperature range). In this
case, we would then expect the exponential term to denote the fraction
of colliding molecules which have the energy required to surmount the
barrier, while A represents roughly the collision rate constant, kAB
as presented in Equation (2.41). Once again, we find experimentally
that this is not always the case. One simple explanation for this is
that only certain orientations are conducive to reactions, especially for
complex molecules. Thus we can introduce a “steric factor” to represent
the probability that the molecules will be in the proper geometric
arrangement when they collide. Finally, another area of possible failure
in the hard sphere model lies in the true nature of the approaching
particles: They are not hard spheres, rather, they exert attractive forces
on one another. In the case of ion-molecule reactions, the attractive
forces are coulombic, and exert an influence even at relatively large
distances. Therefore these reactions often proceed very much faster than
values derived from simple hard sphere collision theory. Even for neutral
unpolarized molecules, London forces produce an interaction potential.
Thus the interactions between particles become multi-dimensional and
quite complex. Molecular scattering experiments, in which reaction
kinetics are examined as functions of energy and orientation, have led to
considerable progress in our study of these processes. See, for example,
Johnston (1966), Weston and Schwarz (1972) and Smith (1980).
2.4.2 Unimolecular Reactions
The special case of a unimolecular reaction

(ku ); A→B+C (2.46)
deserves a brief discussion. Under what circumstances does a molecule
decompose into its constituent parts? Clearly this type of process cannot
be understood in the context of collision theory as discussed above for
bimolecular processes.
The first successful theory of unimolecular reactions (2.46) was
proposed by Lindemann in 1922. He introduced the idea that apparently
unimolecular reactions were really the result of the following processes:
(kact ); A + M → A∗ + M activation (2.47a)
Figure 2.3. Enthalpy variation during an exothermic reaction. EACT represents

the activation energy of the reaction and ∆HR the enthalpy or heat released
by the reaction.
(kd ); A∗ + M → A + M deactivation (2.47b)

(kr ); A∗ → B + C reaction of energized molecule (2.47c)
The rate of decomposition of A (which is equal to the rate of formation
of B or C) is

d(A)
− = kr (A∗ ) (2.48)
dt
The steady state concentration of A∗ is derived as follows:
d(A∗ )
= 0 = kact (A)(M) − [kd (M) + kr ] (A∗ ) (2.49)
dt
kact (A)(M)
(A∗ ) = (2.50)
kr + kd (M)
so that

d(A) kact (A)(M)
= − kr (2.51)
dt kr + kd (M)
and the effective first order reaction rate ku (sec−1 ) is given by
kr
ku = kact (M) (2.52)
kr + kd (M)
Thus A is conceived to be transformed to an “energized” state, A∗ ,
by collision. A∗ can then be deactivated by collision, or it may
decompose. At sufficiently high pressures (high pressure limit), the rate
of deactivation is much greater than the rate of decomposition; the rate
limiting step is then Equation (2.47c) and the reaction will be found to
be independent of pressure.
d(A) kact (A)
kd (M) >> kr ; = −kr (2.53)
dt kd
At very low pressures (low pressure limit), the observed rate is
proportional to the pressure, since the rate limiting step is given by
the production of the energized molecules by Equation (2.47a).
d(A)
kd (M) << kr ; = −kact (A)(M) (2.54)
dt
Between these two extremes, the reaction is neither first nor second
order, and the pressure dependence must be carefully parameterized.
The thermal decomposition of N2 O5 ,
N2 O5 → NO2 + NO3 (2.55)
is an important stratospheric reaction which is of intermediate order at
stratospheric pressures.
2.4.3 Termolecular Reactions
Reactions involving three particles

(kt ); A+B+M→C+M (2.56)
do not generally occur via a single collision involving all three particles
at once; indeed, the probability of such an event is quite small. In
general a termolecular reaction can be understood in the context of
activated complex theory as discussed earlier. The initial step involves
an interaction of two particles:
(kact ); A + B → C∗ (2.57a)
which must be followed by deactivation
(kd ); C∗ + M → C + M (2.57b)
in order that the energized molecule not decompose back to reactants:
(kuni ); C∗ → A + B (2.57c)
Thus this type of process is the reverse of unimolecular decay. The rate
of formation of the product, C, is given by
d(C) kact (A)(B)
= kd (M)(C∗ ) = kd (M) (2.58)
dt kd (M) + kuni
(in which steady state is assumed for C∗ as shown above). Thus, the
effective rate constant can be expressed in terms of the elementary steps
kd (M)
kt = kact (2.59)
kd (M) + kuni
If k0 = kact kd /kuni is the termolecular rate coefficient at low pressure

(expressed in cm6 molecule−2 s−1 ) and k∞ = kact is the bimolecular
high pressure limiting rate coefficient (expressed in cm3 molecule−1 s−1 ),
the effective rate constant can be expressed as
k∞ k0 (M)
kt = (2.60)
k0 (M) + k∞
At low pressure when kuni >> kd (M) (e.g., in most of the stratosphere),
the apparent kinetic rate depends linearly on pressure (third-order
reaction)
kt = k0 (M) (2.61a)
while at high-pressure when kuni << kd (M) (e.g., in the troposphere),

it becomes independent of pressure (second-order reaction)
kt = k∞ (2.61b)
As in the case of unimolecular processes, it is certainly possible for a

reaction to be neither second nor third order, and a more detailed model
of the pressure dependence must be used. For example, the apparent
rate constant can be expressed as a function of the air density (M) and
the temperature T by the empirical Troe’s relation (Troe, 1979)
k∞ k0 (M) 2 −1
kt = Fc {1+[log10 (k0 (M)/k∞ )] } (2.62)
k0 (M) + k∞
where k0 and k∞ can be expressed as a function of temperature by

T −n
k0 (T) = k0 (300K) (2.63a)
300

T −m
k∞ (T) = k∞ (300K) (2.63b)
300
Parameters k0 and k∞ at 300 K as well as n and m are derived from
laboratory data. The broadening factor Fc is often taken as equal to
0.6.
Since kuni is generally larger than kd (M), termolecular processes often
are found to exhibit negative energies of activation. Physically, this may
be thought of as related to the degree of energization of C∗ . At higher
temperatures, the C∗ intermediate will be formed with higher mean
internal energy, and fewer of them will live long enough for the rate
limiting stabilizing reaction, kd , to occur.
2.5 Heterogeneous and Multiphase Reactions

Reactions taking place on the surface of solid or liquid particles and
inside liquid droplets play an important role in the middle atmosphere,
especially in the lower stratosphere where sulfate aerosol particles and
polar stratospheric clouds (PSCs) are observed. The nature, properties
and chemical composition of these particles are described in Chapters 5
and 6. Several parameters are commonly used to describe the uptake
of gas-phase molecules into these particles: (1) the sticking coefficient s
which is the fraction of collisions of a gaseous molecule with a solid
or liquid particle that results in the uptake of this molecule on the
surface of the particle; (2) the accommodation coefficient α which is the
fraction of collisions that leads to incorporation into the bulk condensed
phase, and (3) the reaction probability γ (also called the reactive uptake
coefficient) which is the fraction of collisions that results in reactive
loss of the molecule (chemical reaction). Thus, the accommodation
coefficient α represents the probability of reversible physical uptake of
a gaseous species colliding with a surface, while the reaction probability
γ accounts for reactive (irreversible) uptake of trace gas species on
condensed surfaces. This latter coefficient represents the transfer of a
gas into the condensed phase and takes into account processes such as
liquid phase solubility, interfacial transport or aqueous phase diffusion,
chemical reaction on the surface or inside the condensed phase, etc.
The loss rate of an atmospheric molecule resulting from the uptake
by liquid or solid particles is generally expressed by a pseudo-first-order
equation
d(X)
= −kX (X) (2.64)
dt
if (X) is the atmospheric concentration of molecule X. Neglecting gas-
phase diffusion, the rate coefficient kX (s−1 ) is given by the product of
the collision frequency (between a gas-phase molecule and a particle),
and the reaction probability γ
γcX A
kX = (2.65)
4
where A is the surface area density (expressed, for example, in cm2
cm−3 ) provided by the particles present in the atmosphere and cX is the
mean thermal velocity of the molecule (see Eq. (2.43b)), with
1
8kT 2
cX = (2.66)
πmX
if T is the temperature, mX the mass of molecule X, k the Boltzmann
constant and π = 3.14159. Box 2.2 provides more information on the
methods used to calculate kX .
Box 2.2 Modeling Heterogeneous Processes in the Stratosphere

Liquid and solid particles observed in the atmosphere are generally
a mixture of water, sulfuric acid and nitric acid with potentially the
presence of traces of other chemical compounds like HCl (see Table 5.9 in
Section 5.7). The determination of the pseudo first-order rate coefficient
kX (see Eq. (2.65)) for uptake by such particles in the stratosphere requires
an accurate estimate of the surface area density available and of the
reaction probability involved.
The mean surface area density A can be calculated if the size
distribution of the particles present in the atmosphere is known. For
spherical particles
∞
dn
A= 4πr2 dr
0 dr
if dn represents the number of particles whose radius is in the range [r, r +
dr]. Typical values of A are 1 µm2 per cm3 of air in the case of background
sulfuric aerosols (10 to 100 times more after a large volcanic eruption),
1 µm2 per cm3 in the case of nitric acid trihydrates, 10 µm2 cm−3 for
liquid ternary solutions and 10-100 µm2 per cm3 in the case of ice particles.
The reaction probability γ depends in general on the physical nature
and chemical composition of the particle and on temperature. The
composition of the liquid or solid particles is derived from thermodynamic
considerations. The value of γ for different conditions is determined in the
laboratory. In the case of liquid sulfuric acid aerosols, for most reactions,
γ depends on the weight percentage W (wt%) of H2 SO4 in the aerosol.
W can be expressed as a function of temperature T(K) and water vapor
pressure pH2 O (hPa) by (Steele and Hamill, 1981; Hanson et al., 1994)
(−14.458 + 0.62456Z) T + 3565
W (%wt H2 SO4 ) =
44.777 + 1.3204Z − 0.19988T
with
Z = ln pH2 O
When HCl molecules are dissolved in the aerosol, they can contribute
to the uptake take of other molecules, and the γ coefficient is modified as
a function of the concentration of HCl dissolved in the solution [HCl](M)
and hence of the partial pressure of HCl in the atmosphere pH2 O and
temperature T(K) since
[HCl] = HHCl pHCl
if HHCl (M atm−1 ) is the Henry’s law constant which can be expressed as
HHCl = 10(15.514−0.1791W)
Box 2.2 (Continued)

Heterogeneous reaction rates on particles in polar stratospheric clouds
(PSCs) are more difficult to estimate because of the uncertainties in
the type of PSC particles present in the lower stratosphere. The
particle composition, volume density and radius must be derived from
a thermodynamic model (e.g., Carslaw et al., 1994; 1997). Condensation
of HNO3 occurs when the partial pressure of nitric acid in air exceeds
the equilibrium vapor pressure peq
HNO3 . For example, in the case of nitric
acid trihydrates (NAT), if partial pressures are expressed in Torr and the
temperature T in Kelvin, we have (Hanson and Mauersberger, 1988)
log peq
HNO3 = (−2.7836 − 0.00088T ) log pH2 O
+ 38.9855 − 11397/T + 0.009179T
if pH2 O is the partial pressure of water vapor in the atmosphere.
Similarly, ice particles may form when the partial pressure of
atmospheric water vapor becomes supersaturated relative to ice (i.e., as
temperatures drop below the freezing point). peq
H2 O (here in hPa) over ice
can be expressed as a function of temperature T (in Kelvin) as
eq
log pH2 O = −2663.5/T + 12.537
Sedimentation of PSC particles can be an effective process for removing

water and nitrogen compounds from the atmosphere (dehydration and
denitrification), and has to be taken into account when modeling PSC
effects. The velocity at which particles sediment increases with the size
of these particles. If the size distribution of the particles follows a log-
normal law, the sedimentation flux for a given chemical compound can be
expressed as (Considine et al., 2000)

Φ = n w0 exp 8 ln2 σ
where n is the number density of the sedimenting molecules in the
condensed phase, w0 is the fall velocity derived for a radius equal to the
median radius of the size distribution, and σ is the standard deviation of
the log-normal distribution.
Different expressions relating the reaction probability γ to other

parameters of the system have been proposed in the case of liquid
particles (see e.g., Schwartz, 1986; Hanson and Ravishankara, 1994; Hu
et al., 1995; Hanson, 1997a, b; Seinfeld and Pandis, 1998). Adopting the
scheme shown in Figure 2.4 to account for the transfer of molecules from
the gas phase to the surface (adsorption with a corresponding transfer
coefficient noted kads ), from the surface to the gas phase (desorption
noted kdes ), from the surface into the bulk of the particle (solvation
noted ksol ), and from the bulk to the surface (ks ), Hanson (1997b) has
Figure 2.4. Reactive uptake of a gas-phase molecule by a liquid particle.
derived the following expression

1 1 1
= + ks
(2.67)
γ s Γbulk + Γsurf
ks +(kR DL )0.5
in which s represents the sticking coefficient, DL is the diffusion
coefficient in the solution, kR is the first order reactive loss coefficient in
the bulk of the particle, and

4RTHX (kR DL )0.5 kR 0.5
Γbulk = f a (2.68a)
cX DL
4ks kads
Γsurf = (2.68b)
cX kdes
represent the effects of bulk and surface reactions, respectively. In
these expressions, HX is the Henry’s law coefficient which relates
the equilibrium concentration of molecule X in the gas phase to the
concentration of the same species in the liquid phase. R is the gas
constant, T is the temperature, and ks is the first-order loss rate
coefficient for molecule X on the surface of the particle. Factor

f a (kR /DL )0.5 , where a is the radius of the particles, accounts for the
sphericity of the particles. The value of this function is given by
f (ζ) = coth (ζ) − ζ −1 (2.69)
where ζ = a (kR /DL )0.5 is a dimensionless parameter which describes
the relative rates of chemical reaction and diffusion.
Assuming that the transfer rates between the surface and the bulk of
the particles are fast, and noting that the sticking coefficient s is related
to the accommodation coefficient α by
ksol
α=s (2.70)
ksol + kdes
the expression for the reaction probability becomes (Hanson and
Ravishankara, 1994)
1 1 1
= + (2.71)
γ α Γbulk + Γsurf
By analogy with an electrical circuit, the resistance to uptake (1/γ) can
thus be represented by 2 resistors in series 1/α and 1/(Γbulk + Γsurf ).
In the case of solid particles (e.g., ice), reactive uptake is driven
primarily by chemical reactions at the surface since bulk diffusion is
generally too slow to allow for solubility in the condensed phase.
Recent laboratory studies have shown that the reaction probability γ
can depend strongly on the physical nature, the chemical composition
of the aerosols, and hence on the atmospheric temperature and
concentration of water vapor and other chemical compounds (e.g., HCl).
Figure 2.5 shows the reaction probability as a function of temperature
(and equivalently of the weight percentage of sulfuric acid) for the
following key stratospheric heterogeneous processes on liquid sulfate
aerosols:
N2 O5 (g) + H2 O () → 2 HNO3 (2.72)
ClONO2 (g) + H2 O () → HOCl(g) + HNO3 (2.73a)
ClONO2 (g) + HCl () → Cl2 (g) + HNO3 (2.73b)
HOCl (g) + HCl () → Cl2 (g) + H2 O (2.74a)
BrONO2 (g) + H2 O () → HOBr + HNO3 (2.74b)
where g and refer to the gas and liquid phases, respectively. The
values of γ displayed in Figure 2.5 refer to conditions characteristic of
the lower stratosphere (JPL, 2000). Such reactions, which also occur in
the presence of other types of particles (including PSC particles) convert
relatively inert halogen compounds such as HCl and ClONO2 into
gaseous products (Cl2 , HOCl, etc.) which are rapidly photolyzed into
Figure 2.5. Reactive uptake coefficient (γ) for several stratospheric

heterogeneous processes as a function of temperature and aerosol sulfuric acid
weight percentage. Values established for atmospheric conditions representative
of the lower stratosphere: pressure of 50 hPa, 5 ppmv H2 O, 2 ppbv HCl and
0.1 ppbv ClONO2 . Aerosol diameter of 10−5 cm. After JPL (2000).
reactive halogen radicals that efficiently destroy ozone (see Chapters 5

and 6).
Heterogeneous reactions are often complex and nonlinear processes.
For example, in the case of the heterogeneous conversion of ClONO2 ,
the total (observed) reaction probability is determined by reactions with
water molecules (Reaction 2.73a) and with HCl (Reaction 2.73b). The
total reactive uptake coefficient γ varies, therefore, with the amount
of HCl dissolved in the particles, and hence with the concentration
of gaseous HCl present in the atmosphere (Hanson and Ravishankara,
1994). Figure 2.6 shows the contribution of the two reactions as a
function of the atmospheric HCl partial pressure, and specifically
Figure 2.6. Reactive uptake coefficient (γ) for the heterogeneous conversion of
ClONO2 on 55.6 wt% sulfuric acid aerosol particles. The triangles represent the
measurements of Hanson and Ravishankara (1994). The solid line represents
the parameterization of γ as a function of the atmospheric partial pressure of
HCl. The dashed lines represent the fraction of ClONO2 reacting with water
molecules (γ H2 O ), with HCl in the bulk of the aerosol (ΓHCl
b ), and with HCl on
the surface of the particles (ΓHCl
s ). A HCl mixing ratio of 2 ppbv at a pressure
level of 50 hPa (lower stratosphere) corresponds to a HCl partial pressure of
10−10 atm. From Peter (1997).
the fraction of ClONO2 reacting with HCl in the bulk (ΓHCl b ) and on
the surface (ΓHCl
s ) of a sulfate aerosol particle (see Eq. 2.71). Note
that the reaction pathway with H2 O dominates for low concentrations
of HCl, but drops as the reaction with HCl becomes dominant. In this
latter case, the probability for the reaction to occur on the surface of
the aerosol is often larger than in the bulk of the particle. As expected,
the total γ coefficient tends to saturate at a value close to unity for high
values of the HCl partial pressure.
2.6 Photolysis Processes
Photochemical processes in the atmosphere are initiated by the

absorption of a photon (quantum of electromagnetic energy) by a
molecule XY, leading to promotion of the molecule to an electronically-
excited state (conventionally denoted by the superscript *):
XY + hν → XY∗ (2.75)
The excited state is usually unstable, and is therefore followed by

dissociation
XY∗ → X + Y (2.76a)
by ionization
XY∗ → XY+ + e− (2.76b)
by fluorescence
XY∗ → XY + hν (2.76c)
or by intramolecular energy transfer
XY∗ → XY∗∗ (2.76d)
The excited state molecule can also be quenched by a third body M
XY∗ + M → XY + M (2.77a)
or react with a molecule AB with intermolecular energy transfer.
XY∗ + AB → products (2.77b)
The quantum yield of a given pathway following the light absorption
is defined as the number of reactant molecules which decompose along
this pathway relative to the number of photons absorbed. The sum of
quantum yields for all possible pathways is then unity. In this section, we
focus on photodissociation processes (also called photolysis), as described
by Reactions (2.75) and (2.76a).
Figure 2.7 presents a typical potential energy diagram for a diatomic
molecule, where possible photon-induced transitions are indicated.
At atmospheric temperatures, most molecules are in their ground
vibrational state. Possible transitions are represented by the vertical
lines labeled 1 and 2, for example. Transition 1 leads from the ground
state to an unbound state, and thus a continuum of wavelengths near
the energy of 1 are possible. The spectrum should therefore consist
of a smooth continuum at wavelengths above the dissociation limit.
Transition 2, on the other hand requires that absorption into a specific
vibrational level of the upper electronic state occur. It is therefore
quantized at nearly the energy defined by 2 (differences in rotational
energy levels, which are too small to show in the figure, will lead to
a band of possible energies about 2). The spectrum corresponding to
these transitions will be banded at discrete wavelengths. Such spectra
are exhibited by oxygen in the Herzberg continuum and Schumann-
Runge bands and continuum. An approximate potential energy diagram
for O2 is shown in Figure 2.8, where these transitions are indicated.
The Schumann-Runge bands correspond to transitions from the ground
3 Σ− to the bound portion of the B3 Σ− state, and the Schumann-
g u
Runge continuum involves the same electronic states, but at shorter
wavelengths above the dissociation limit. The Herzberg continuum

involves excitation from the ground state to the repulsive A3 Σ+
u state.
Figure 2.7. Typical potential energy curves for a diatomic molecule.
Figure 2.8. Potential energy diagram for the principal states of the oxygen
molecule. Adapted from Gilmore (1964).
Figure 2.9. Potential energy curves for a predissociative process in a diatomic

molecule.
If the upper state is bound, as it is for transition 2 in Figure 2.7,

can such a transition lead to dissociation? Dissociation can often
occur in such a case through curve crossing onto a predissociating state.
Figure 2.9 presents a corresponding potential energy curve showing this
phenomenon. The excited state, XY∗ , which is initially produced,
can cross into XY∗∗ by a non-radiative transition or by collision-
induced crossing. From the predissociating state, ground state atoms
are eventually produced. The indicated transition for the Schumann-
Runge bands as depicted in Figure 2.8 involves crossing into the
repulsive 3 Πu state leading to dissociation. Nitric oxide photolysis in the
middle atmosphere occurs largely through predissociation. Polyatomic
molecules have more complex potential surfaces in three or more degrees
of freedom, and an increased probability of possible predissociating
surfaces.
The rate at which molecule XY is photodecomposed is expressed as
d (XY)
= −Jxy (XY) (2.78)
dt
where Jxy is the unimolecular rate constant for photolytic loss of XY.
Jxy , which is also called the photolysis frequency (expressed in s−1 ), is
determined by the number of photons available (the solar actinic flux),
the ability of the molecule to absorb these photons (the absorption cross-
section), and the probability that the photon absorption leads to the
decomposition of the molecule (quantum yield for photodissociation).
The product of these 3 factors is integrated over all wavelengths
contributing to the photolysis of the molecule (see Chapter 4 for
more details). The absorption cross sections and quantum yields are
determined in the laboratory. In many cases, these quantities vary with
temperature. In some cases, the quantum yield is a function of pressure.
2.7 Excited Species in the Middle Atmosphere
Photolysis reactions produce a number of electronically excited

species, (e.g., O(1 D)) which play an important role in the middle
atmosphere. Some of these constituents are important because they are
very reactive and can initiate chains of important free radical chemistry.
Examples of such processes include
O3 + hν → O∗3 → O (1 D) + O2 (1 ∆g ) (2.79)
O (1 D) + H2 O → 2OH (2.80)
which provides the major source of reactive hydrogen containing radicals
in the stratosphere and troposphere. This will be discussed in much
more detail in Chapter 5. Another excited species, N(2 D), produces
large amounts of NO in the lower thermosphere (see Chapters 5 and 7)
via
N (2 D) + O2 → NO + O (2.81)
A significant aspect of excited state chemistry, especially in the
mesosphere and lower thermosphere, is the importance of fluorescence
and chemiluminescence. Fluorescence is the rapid re-emission of
absorbed photons:
A + hν → A∗ (2.82)
A∗ → A + hν (2.83)
In the process of chemiluminescence, the emitting particle is excited by
chemical reactions rather than direct photon absorption:
A + B → C∗ + D (2.84)
C∗ → C + hν (2.85)
Excited species can lose their energy by collision (or quenching) with
other molecules. However, in the upper atmosphere, where the density
of air molecules is relatively small and more high energy photons are
available to produce excited species, many constituents fluoresce or
chemiluminesce. Thus, photons of specific wavelengths are released,
producing characteristic airglow emissions. Many of these emissions
can also be used to deduce information about the concentrations of
atmospheric species. Table 2.1 presents a summary of some of the more
Table 2.1 Emissions of Some Excited Species in the Middle Atmosphere
Lower Excited Radiative∗ λ Name

state state lifetime (s) (nm)
O (3 P) O (1 D) 110 630 Red line

O (1 D) O (1 S) 0.74 557.7 Green line
O2 (X3 Σ−
g) O2 (a1 ∆g ) 2.7 (3) 1270 Infrared
atmospheric
bands
O2 (X3 Σ−
g) O2 (b1 Σ+
g) 12 761.9 Atmospheric
bands
O2 (X3 Σ−
g) O2 (A3 Σ+
u) 1 260-380 Herzberg
bands
OH (X2 Π)v=0,1,.. OH (X2 Π)v=9,8,.. 6 (–2) < 2800.7 Meinel bands
N (4 S) N (2 D) 9.36 (4) 520
N (4 S) N (2 P) 12 346.6
N2 (X1 Σ+
g) N2 (A3 Σ+
u) 2 200-400 Vegard-Kaplan
bands
NO (X2 Π) NO (A2 Σ+ ) 2 (–7) 200-300 γ bands
∗ read 2.7 (3) as 2.7 × 103 .
important emitting species in the middle atmosphere. Plate 1 shows

an aurora produced by the radiative emission of excited oxygen atoms
(green and red light).
Note the very long radiative lifetimes of O (1 D), O2 (1 ∆g ), and
N (2 D). Transitions of these species to their respective ground states
involve violation of the spin conservation rule, and are therefore
relatively slow. The O (1 S) → O (1 D) transition, which produces
the green line, is spin allowed, but corresponds to ∆L of 2, and is
therefore somewhat unfavorable. In contrast, the NO (A2 Σ+ ) → NO
(X2 Π) transition is fully allowed and the radiative lifetime of this state
is correspondingly short.
The green line of atomic oxygen is believed to arise mainly through
the reaction
O + O + M → O∗2 + M (2.86)
followed by
O∗2 + O → O (1 S) + O2 (2.87)
and finally
O (1 S) → O (1 D) + hν (557.7 nm) (2.88)
commonly referred to as the Barth mechanism (Barth and Hildebrand,

1961; see the review by Bates, 1981). Here the symbols O and O2 refer
to oxygen atoms and molecules in their fundamental states 3 P and 3 Σ,
respectively. The green line emission provides a useful means of deducing
the atomic oxygen density near 100 km, as shown for example by Slanger
and Black (1977).
O2 (1 ∆g ) is produced in the sunlit mesosphere principally via
O3 + hν → O (1 D) + O2 (1 ∆g ) (2.89)
and to a lesser extent from
O (1 D) + O2 → O2 (1 Σ+
g)+O (2.90)
or
O2 + hν → O2 (1 Σ+
g) (2.91)
followed by
g ) + M → O2 ( ∆g ) + M
O2 (1 Σ+ 1
(2.92)
O2 (1 ∆ g) is in turn destroyed either by radiative decay (1.27µm),

O2 1
∆g → O2 3
Σ−
g + hν (1.27µm) (2.93)
or by quenching:
O2 (1 ∆g ) + O2 → 2O2 (2.94)
If the rates of all of these processes are known, then ozone densities
can be derived from measurements of the infrared atmospheric volume
emission rate (e.g., Evans et al., 1968; Thomas et al., 1983). This
provides a useful means of measuring mesospheric ozone abundances
(Mlynczak and Olander, 1995). Dayglow production mechanisms by
oxygen compounds are shown schematically in Figure 2.10 (Mlynczak
et al., 1993).
The OH Meinel bands are generated between 80 and 100 km altitude
by the reaction between atomic hydrogen and ozone
H + O → OH† (v = 9, 8, 7, 6) + O
3 (2.95)
2
followed by chemiluminescence (radiative emission)

OH† v → OH† v + hν (2.96)
or by quenching with M (N2 , O2 or O)
OH† + M → OH + M (2.97)
of the vibrationally excited hydroxyl radical (v is the vibrational
quantum number). The Einstein A-coefficients associated with the
OH† vibrational emissions (and hence the corresponding radiative
lifetimes) have been measured in the laboratory (Nelson et al., 1990).
Figure 2.10. Oxygen dayglow production mechanism in the middle atmosphere.

The photolysis of ozone (JHar ) in the Hartley band (λ < 320 nm) leads to the
formation of the excited oxygen atom O (1 D) and molecule O2 (1 ∆g). The
photolysis of molecular oxygen in the Schumann-Runge continuum (JSRC at
λ < 175 nm) and at Lyman α (JLα at λ = 121.6 nm) leads to the formation of
2 electronically excited oxygen atoms O (1 D). ) De-activation is associated with
radiative emissions (Ai ) and quenching (Qi ). The emission A1 produces a red
line at 630 nm (see Table 2.1),) A2 produces atmospheric bands at wavelengths
near 762 nm, and A3 (Reaction 2.93) produces infrared bands at 1.27µm. The
quenching processes correspond to the following reactions: 2.90 for Q1 and Q4 ,
2.92 for Q2 , and 2.94 for Q3 . Radiative excitation (g) at 762 nm is represented
by reaction 2.91 (from Mlynczak et al., 1993).
Uncertainties regarding the quenching lifetimes and the rate constants

for reactions of OH† with atomic oxygen remain fairly large (Mlynczak
et al., 1998), and make the interpretation of the airflow feature
difficult (Llewellyn et al., 1978). Note that the hydroxyl airglow
reduces the amount of energy potentially available for heat from the
exothermic reaction between ozone and atomic hydrogen. Measurements
of the hydroxyl airglow can be used to derive the atomic hydrogen
concentration (if the ozone concentration is known, see Mlynczak et al.,
1998) and the (OH rotational) temperature near the mesopause (Viereck
and Deehr, 1989; Lowe and Turnbull, 1995).
Emissions from N (2 D) and N (2 P) are particularly important in
the aurora, where they are produced primarily through photoelectron
impact on N2 (see Chapter 5 and 7). As mentioned above, the reactions
of N (2 D) or N (2 P) provide an important source of NO in the lower
thermosphere, and this emission feature can indicate the yield of excited
nitrogen production in auroral processes (see, for example, Frederick and
Rusch, 1977; Zipf et al., 1980).
The fluorescence of NO in the γ bands represents a useful method
of NO measurement in the lower thermosphere, as first pointed out by
Barth (1966).
Excited state chemistry is one of the most important aspects of

classical chemistry as applied to the middle atmosphere. Indeed, as
shown in Figures 2.7, 2.9, and 2.10, all photolysis reactions involve at
least temporary production of electronically excited species; thus the
photochemistry of the atmosphere is closely tied to the excited states of
a number of important species. More detail regarding the chemistry of
some of these states will be presented in Chapter 5.
References
Barth, C.A., and A.F. Hildebrand, The 5577 Å airglow emission mechanism.
J Geophys Res: 66, 985, 1961.
Barth, C.A., Nitric oxide in the upper atmosphere. Ann Geophys: 22, 198, 1966.
Bates, D.R., The green light of the night sky. Planet Space Sci: 29, 1061, 1981.
Carslaw, K.S., B.P. Luo, S.L. Clegg, T. Peter, P. Brimblecombe, and P.J. Crutzen,
Stratospheric aerosol growth and HNO3 gas phase depletion from coupled HNO3 and
water uptake by liquid particles. Geophys Res Lett: 21, 2479, 1994.
Carslaw, K.S., T. Peter, and S. L. Clegg, Modeling the composition of liquid
stratospheric aerosols. Rev Geophys: 35, 125, 1997.
Castellan, G.W., Physical Chemistry. Addison-Wesley, 1971.
Considine, D.B., A.R. Douglass, P.S. Connell, D.K. Kinnison, and D.A. Rotman, A
polar stratospheric cloud parameterization for the global modeling initiative three-
dimensional model and its response to stratospheric aircraft. J Geophys Res: 105,
3955, 2000.
Ebbing, D.D., General Chemistry. 3rd ed. Houghton Mifflin Co., 1990.
Evans, W.F.J., D.M. Hunten, E.J. Llewellyn, and A. Vallance-Jones, Altitude profile
of the infrared atmospheric system of oxygen in the dayglow. J Geophys Res: 73,
2885, 1968.
Finlayson-Pitts, B.J., and J.N. Pitts, Jr., Chemistry of the Upper and Lower
Atmosphere, Theory, Experiments, and Applications. Academic Press, 1999.
Frederick, J.E., and D.W. Rusch, On the chemistry of metastable atomic nitrogen in
the F-region deduced from simultaneous satellite measurements of the 5200 Å airglow
and atmospheric composition. J Geophys Res: 82, 3509, 1977.
Gilmore, F.R., Potential energy curves for N2 , NO, O2 , and corresponding ions.
RAND corporation memorandum R-4034-PR, June, 1964.
Hanson, D.R., and K. Mauersberger, Laboratory studies of the nitric acid trihydrate:
Implications for the south polar stratosphere. Geophys Res Lett: 15, 855, 1988.
Hanson, D.R., and A.R. Ravishankara, Reactive uptake of ClONO2 onto sulfuric acid
due to reaction with HCl and H2 O. J Phys Chem: 98, 5728, 1994.
Hanson, D.R., A.R. Ravishankara, and S. Solomon, Heterogeneous reactions in
sulfuric acid aerosols: A framework for model calculations. J Geophys Res: 99,
3615, 1994.
Hanson, D.R., Reaction of N2 O5 with H2 O on bulk liquids and on particles and the
effect of dissolved HNO3 , Geophys Res Lett: 24, 1087, 1997a.
Hanson, D.R., Surface-specific reactions on liquid. J Phys Chem: 101, 4998, 1997b.
Herzberg, G., Spectra of Diatomic Molecules. D. Van Nostrand Co., 1950.
Hoffman, D.J., and S. Solomon, Ozone destruction through heterogeneous chemistry
following the eruption of El Chichón. J Geophys Res: 94, 5029, 1989.
Hu, J.H., Q. Shi, P. Davidovits, D.W. Worsnop, M.S. Zahniser, and C.E. Kolb,
Reactive uptake of Cl2 (g) and Br2 (g) by aqueous surfaces as a function of Br− and
I− ion concentrations: The effect of chemical reaction at the interface. J Phys Chem:
99, 8768, 1995.
Jet Propulsion Laboratory (JPL), Chemical Kinetics and Photochemical Data for
Use in Atmospheric Studies, Evaluation 14: NASA Jet Propulsion Laboratory
Publication 02-25, 2002.
Johnston, H. S., Gas Phase Reaction Rate Theory. Ronald Press, 1966.
Karplus, M., and R.N. Porter, Atoms and Molecules: An Introduction for Students
of Physical Chemistry. W.A. Benjamin, Inc., 1970.
Llewellyn, E.J., B.H. Long, and B.H. Solheim, The quenching of OH∗ in the
atmosphere. Planet Space Sci: 26, 525, 1978.
Lowe, R.P., and D.N. Turnbull, Comparison of ALOHA-93, ANLC-93 and ALOHA-
90 observations of the hydroxyl rotational temperature and gravity wave activity.
Mlynczak, M.G., S. Solomon, and D.S. Zaras, An updated model for the O2 (a1 ∆g )
concentrations in the mesophere and lower thermosphere and implications for remote
sensing of ozone at 1.27µm. J Geophys Res: 98, 18,639, 1993.
Mlynczak, M.G., and D.S. Olander, On the utility of the molecular oxygen dayglow
as proxies for middle atmosphere ozone. Geophys Res Lett: 22, 1377, 1995.
Mlynczak, M.G., D.K. Zhou, and S.M. Adler-Golden, Kinetic and spectroscopic
requirements for the inference of chemical heating rates and atomic hydrogen densities
from OH Meinel band measurements. Geophys Res Lett: 25, 647, 1998.
Moore, W.J., Physical Chemistry. Prentice Hall, 1962.
Mozurkiewicz, M., and J.G. Calvert, Reaction probability of N2 O5 on aqueous
aerosols. J Geophys Res: 93, 15,889, 1998.
Nelson, D.D., Jr., A. Schiffman, D.J. Nesbitt, J.J. Orlando, and J.B. Burkholder,
H + O3 Fourier-transform infrared emission and laser absorption studies of OH
(X2 π) radical: An experimental dipole moment function and state-to-state Einstein
A coefficients. J Chem Phys: 93, 7003, 1990.
Peter, T., Microphysics and heterogeneous chemistry of polar stratospheric clouds.
Ann Rev Phys: 48, 785, 1997.
Rice, O.K., and H.C. Ramsperger, Theories of unimolecular gas reactions at low
pressures. J Am Chem Soc: 49, 1617, 1927.
Schwartz, S.E., Mass transport considerations pertinent to aqueous-phase reactions
of gases in liquid-water clouds, in Chemistry of Multiphase Atmospheric Systems. W.
Jaeschke, ed. Springer-Verlag, 1986.
Seinfeld, J.H., and S.N. Pandis, Atmospheric Chemistry and Physics — From Air
Pollution to Climate Change. J. Wiley and Sons, 1998.
Slanger, T.G., and G. Black, O1 S in the lower thermosphere — Chapman vs. Barth.
Planet Space Sci: 25, 79, 1977.
Smith, I.W.M., Kinetics and Dynamics of Elementary Gas Reactions. Butterworths,
1980.
Steele, H.M., and P. Hamill, Effects of temperature and humidity on the growth and
optical properties of sulfuric acid-water droplets in the stratosphere. J Atmos Sci:
12, 517, 1981.
Steinfeld, J.I., Molecules and Radiation: An Introduction to Modern Molecular
Spectroscopy. MIT Press, 1978.
Thomas, R.J., C.A. Barth, G.J. Rottman, D.W. Rusch, G.H. Mount, G.M. Lawrence,
R.W. Sanders, G.E. Thomas, and L.E. Clemens, Ozone density in the mesosphere (50-
90 km) measured by the SME limb scanning near infrared spectrometer. Geophys Res
Lett: 10, 245, 1983.
Troe, J., Predictive possibilities of unimolecular rate theory. J Phys Chem: 83, 114,
1979.
Viereck, R.A., and C.S. Deehr, On the interaction between gravity waves and the
OH Meinel (6-2) and the O2 atmospheric (0-1) bands in the polar night airglow. J
Geophys Res: 94, 5397, 1989.
Weston, R.E., and H.A. Schwarz, Chemical Kinetics. Prentice Hall, 1972.
Zipf, E.C., P.J. Espy, and C.F. Boyle, The excitation and collisional deactivation of
metastable N2 P atoms in auroras. J Geophys Res: 85, 687, 1980.
Chapter 3
DYNAMICS AND TRANSPORT
3.1 Introduction
The distribution of most chemical species in the middle atmosphere

results from the influences of both dynamical and chemical processes.
In particular, when the rates of formation and destruction of a chemical
compound are comparable to the rate at which it is affected by
dynamical processes, then transport plays a major role in determining
the constituent distribution. In an environment like the Earth’s
atmosphere, air motions, and hence transport of chemical species, are
strongly constrained by density stratification (gravitational force) which
resists vertical fluid displacements, and the Earth’s rotation (Coriolis
force) which is a barrier against meridional displacements. Geophysical
fluid dynamics describes how atmospheric motions are produced within
these constraints.
Transport occurs at a variety of spatial and temporal scales. In
atmospheric models, it is customary to distinguish between motions
that can be resolved on the numerical grid of the model (often
called large-scale advection) and sub-grid scale processes (such as
molecular diffusion or turbulent mixing) that are often parameterized
by diffusion coefficients. In turn, the distributions of certain chemical
species, particularly ozone, can influence the radiative budget of the
atmosphere, affecting temperatures and flow patterns. Therefore the
study of atmospheric chemistry intersects greatly with those of fluid
dynamics and meteorology. In this chapter, the structure of the
middle atmosphere will be discussed, and the transport processes in the
stratosphere, mesosphere, and lower thermosphere will be described. In
our description of atmospheric motions, we have not sought to provide a
detailed account of the dynamic meteorology of the middle atmosphere.
For a comprehensive description of middle atmosphere dynamics, the
reader is referred to the excellent books by Andrews et al. (1987), Holton
(1992) and James (1994). See also the chapter by Garcia et al. in the
book by Brasseur et al. (1999) and the papers by Shepherd (2000; 2003).
51
The present chapter is directed towards an understanding of the effects

of atmospheric dynamics on chemical constituents.
The sections of this chapter deal with the following elements of
atmospheric dynamics: vertical structure of the atmosphere (Section
3.2), fundamental equations of atmospheric motions (Section 3.3),
transport of chemical constituents and the relative importance of
dynamical and chemical effects on photochemical species (Section 3.4),
atmospheric waves (Section 3.5), the mean meridional circulation and
the use of the transformed Eulerian formalism to illustrate the roles of
mean meridional and eddy transports (Section 3.6), the important role
of wave transience and dissipation (Section 3.7), vertical transport by
molecular diffusion in the thermosphere (Section 3.8), and finally, models
of the middle atmosphere (Section 3.9).
(mi) (km)
0.0001
100
Thermosphere
os 60
0.001 90
Mesopause 50 80
0.01
GEOMETRIC HEIGHT
70
PRESSURE (mb)
0.1 40
Mesosphere 60
50
1 Stratopause 30
re
atu
er 40
mp Ozone Region
Te
20
10 30
Stratosphere
Cirrus
20
100 10
Mt. Everest
Tropopause 8848 m
10
Cumulonimbus
Troposphere
oss
1000
180 200 220 240 260 280 300 (deg K)
-140 -100 -60 -20 20 60 100 (deg F)

TEMPERATURE
1013/fig 1.5/8.98/dpm
Figure 3.1. Schematic representation of the thermal structure of the atmosphere

with its different layers.
DYNAMICS AND TRANSPORT 53
3.2 Structure of the Atmosphere and Some Observed

Dynamical Characteristics
The Earth’s atmosphere is commonly described as a series of layers

defined by their thermal characteristics (Figure 3.1). Specifically, each
layer is a region where the change in temperature with respect to altitude
has a constant sign. The layers are called “spheres” and the boundary
between connecting layers is the “pause”. The lowest layer, called the
troposphere, exhibits generally decreasing temperatures with increasing
altitudes up to a minimum called the tropopause. The temperature
and location of the tropopause vary with latitude and season (see
Box 3.1). At the equator, its mean altitude is located near 16 km,
and the corresponding temperature is about 190 K, while in polar
regions its elevation is only about 8 km, and the mean temperature
roughly 220 K (see Figure 3.2). Above the tropopause, the stratosphere
exhibits increasing temperatures with altitude up to a maximum of
about 270 K at the level of the stratopause located near 50 km.
At still higher altitudes, the temperature again decreases up to 85-
100 km, where another temperature minimum is found. This layer is
called the mesosphere and its upper boundary is the mesopause (see
Box 3.2). In all these layers, the major constituents, N2 and O2 ,
make up about 80 and 20%, respectively, of the total number density,
so that the mean molecular weight of air varies little with altitude.
Because of this common feature, the three layers are collectively referred
to as the homosphere. Figure 3.2 presents a contour plot of the
observed meridional distribution of the zonal mean temperature in the
homosphere in January, showing some of these features in more detail.
The region located above the mesopause is called the thermosphere.
The temperatures there increase very rapidly with altitude and can reach
500 to 2000 K above 500 km, depending on the level of solar activity. The
composition of the thermosphere is very different from that of the lower
regions due to diffusive separation between light and heavier compounds
(see Section 3.8). As a result, the relative abundance of atomic oxygen
increases substantially with height and becomes comparable to and even
greater than those of O2 and N2 above about 130 km. The abundances
of O2 and N2 decrease as a result of rapid photodissociation. Atomic
hydrogen (the lightest chemical element) becomes the most abundant
compound in the upper thermosphere. In contrast to the homosphere,
the mean molecular weight of air in this region varies with altitude; for
this reason the region above 100 km is also called the heterosphere.
As was mentioned in Chapter 2, the major source of heat in the
middle atmosphere is provided by absorption of ultraviolet radiation,
particularly by ozone and to a lesser extent by molecular oxygen
Figure 3.2. Zonally averaged temperature (K) from the surface to

approximately 120 km altitude in January, based on Fleming et al. ((1988). )
Note the temperature minimum (less
( than 200 K)) at the tropical tropopause, the
temperature maximum (280
( K)) at the summer stratopause and the temperature
minimum (less than 140 K) at the summer mesopause. The height of the
mesopause increases from approximately 90 km in summer to 100 km in winter.
(see Chapter 4). Radiative cooling occurs through infrared emission

associated with the vibrational relaxation of CO2 , H2 O, and O3 .
Figure 3.3 presents a model estimate of the rate of net radiative heating
in the middle atmosphere. The large radiative heating rates found in the
stratosphere are due primarily to the large amounts of ozone found there.
The observed increase in temperature with altitude in the stratosphere
is a result of heating by ozone, illustrating the important relationship
of atmospheric chemical composition to the radiation budget and the
thermal structure of the atmosphere.
Comparison of Figures 3.2 and 3.3 reveals several interesting features.
For example, although the radiative heating rate at the summer
mesopause is large, the temperatures observed there (120-160 K) are
much lower than those found in the winter hemisphere (200-240 K).
The tropical tropopause, as we have already noted, is much colder
than its counterpart at middle and high latitudes, although no dramatic
variation in the radiative heating rate occurs there. Both cases illustrate
the importance of dynamical effects in establishing the temperature
structure of the middle atmosphere. Specifically, an air parcel displaced
adiabatically upward undergoes expansion and cooling while adiabatic
Figure 3.3. Net radiative heating rate associated with (1) absorption of
ultraviolet radiation by molecular oxygen in the upper mesosphere and
thermosphere, and by ozone in the stratosphere and mesosphere, and (2)
emission of infrared radiation by atmospheric CO2 , O3 , and H2 O. Values given
in K/day
/ and positive in the summer hemisphere ((net diabatic heating)) and
negative (net diabatic cooling) in the winter hemisphere. From London (1980).
compression and heating accompanies downward displacements. These

principles will be described in more detail below in connection with the
thermodynamic and momentum equations, but we note at this point
that mean upward motions are characteristic of the summer mesopause
and of the tropical tropopause.
One of the most important atmospheric dynamical quantities is the
zonal wind (wind speed in the longitudinal direction). Wind velocities
are conventionally positive for eastward winds (also called westerlies)
and negative for westward winds (easterlies). They are mostly derived
from the observed thermal structure of the atmosphere, although local
values can be provided by radiosonde and radar measurements. To date,
very few attempts have been made to directly measure the atmospheric
wind components from space. The High Resolution Doppler Imager
(HRDI) on board the Upper Atmosphere Research Satellite (UARS)
has provided information on horizontal winds in the mesosphere/lower
thermosphere (50-115 km) and in the stratosphere (10-40 km) by
observing the Doppler shifts in the emission lines of an O2 atmospheric
band (Hays et al., 1993; Ortland et al., 1996; Burrage et al., 1996).
Figure 3.4 shows the wind fields recovered from HRDI at an altitude
of 25 km on March 8, 1994 (Ortland et al., 1996). The Wind Imaging
Interferometer (WINDII) on the same satellite has provided the neutral
winds in the 80-300 km altitude range from measurements of the Doppler
shift in the O(1 S) green line emission (Shepherd et al., 1993; Gault et al.,
1996).
Figure 3.5 presents a typical latitude height distribution of mean zonal
winds, averaged over longitude and time. In the lower stratosphere, the
zonal circulation is an extension of the tropospheric flow characterized by
the two jets centered at 30-40 degrees latitude, with wind speeds almost
twice as large in winter as in summer. In most parts of the stratosphere
(above 20 km) and the mesosphere, the prevailing mean zonal winds are
directed eastward (westerlies) during winter and westward (easterlies)
during summer, with maximum velocities of the order of 60-75 m/s near
70 km. Above this altitude the zonal flow becomes less intense and its
direction even reverses in the extratropics above 85-95 km. Note also
the strong westerly winds in the winter polar stratosphere which form
the so-called polar vortex and partly isolate the polar region from low
latitude influences. The zonal mean flow is strongly influenced by the
seasonal variation in the solar heating.
Horizontal wind field at 25 km altitude measured by the

UARS/HRDI instrument on 8 March 1994. From Ortland et al. (1996).
Figure 3.5. Longitudinally averaged zonal wind (m/s) from the surface to
approximately 120 km altitude in January, based on Fleming et al. ((1988). )
Positive winds are blowing eastwards (westerlies)
( ) and negative winds are
blowing westwards (easterlies). Note the presence of the 2 tropospheric
subtropical jets. Stratospheric winds are primarily westerlies in winter and
easterlies in summer. A reversal of the winds is observed above approximately
90 km altitude.
Vertical and meridional winds in the stratosphere can be estimated

from theoretical studies, which are usually based on solutions to the
equations of atmospheric dynamics, as will be described below. In the
mesosphere, however, the meridional wind (wind speed in the latitudinal
direction) has been measured at particular points, usually with radar
techniques (see Figure 3.6 from Nastrom et al., 1982) and derived from
space observations (see, e.g., Lieberman et al., 2000). Meridional winds
are denoted by a positive velocity when they are directed from the south
to the north. In the vertical, wind speeds are positive when they are
directed upwards.
It should be emphasized that Figure 3.5 presents the mean zonal wind,
that is, the wind speed averaged over longitude and time. In addition,
local variations in the wind speed and direction are observed. These
dynamical disturbances are associated with the presence of atmospheric
waves (see Section 3.4). The propagation of these waves through the
atmospheric medium is determined by the basic state of the atmospheric
fluid; at the same time, the dissipation of the waves tends to modify
the basic state itself. Beside radiatively forced seasonal variations,
Figure 3.6. Velocity of meridional winds observed in the mesosphere at Poker

Flat, Alaska (m s−1 ). The air flow at 88 km is directed southwards during
summertime. From Nastrom et al. (1982).
the atmosphere exhibits periodic or quasi-periodic cycles resulting from

the interactions of atmospheric waves with the mean flow; examples
are provided by the quasi-biennial oscillation (QBO) in the lower
stratospheric equatorial zonal wind (whose period varies from 22 to 34
months), and of the semi-annual osciallation (SAO) which is observed
near the equatorial stratopause and mesopause.
In addition to large-scale advection, another possible mechanism
for transport of chemical species is provided by irreversible mixing,
sometimes referred to as eddy diffusion. The only true diffusion in
the atmosphere occurs at the molecular level, via the random motion
of atoms and molecules (see Section 3.8). This becomes important as
a transport process only in the thermosphere. At lower altitudes, all
transport results from motions of parcels of air rather than individual
particles. However, the concept of eddy diffusion is a useful way to
describe the mixing effects of fluid motions occurring on time and
spatial scales much smaller than advection. This process is formally
represented as turbulent mixing, modeled after the behavior of diffusion.
In this chapter we are primarily interested in gradients of photochemical
constituents which occur over spatial scales of the order of kilometers
in the vertical, and thousands of kilometers in the horizontal directions.
Motions which occur over spatial scales much smaller than these may be
understood as mixing processes rather than organized motion over the
relevant distance.
3.3. Fundamental Description of Atmospheric

Dynamics
3.3.1 The Primitive Equations
To describe the theoretical dynamical and thermal behavior of the

atmosphere, the fundamental equations of fluid mechanics must be
employed. In this section these equations are presented in a relatively
simple form. A more conceptual view will be presented in Section 3.6.
The circulation of the Earth’s atmosphere is governed by three basic
principles: Newton’s laws of motion, the conservation of energy, and the
conservation of mass. Newton’s second law describes the response of a
fluid to external forces. In a frame of reference which rotates with the
Earth, the first fundamental equation, called the momentum equation,
is given by:
dv 1 xv = g +
+ ∇p + 2Ω F (3.1)
dt ρ
where v represents the velocity vector of an air parcel, p is pressure, ρ
is the mass per unit volume, Ω is the angular rotation rate of the Earth,
g is the effective gravitational acceleration (including the centrifugal
acceleration owing to the rotation of the Earth), F is the frictional force,
and t is time. The terms in the above equation each correspond to a
different force acting on the fluid; these are the pressure gradient force

(1/ρ∇p), the Coriolis force (expressed by the vector product 2Ω xv),
the
gravitational (g)
and frictional (F) forces (these conventionally represent
the force per unit mass in meteorological studies). d/dt is the total
derivative (following the motion). Note that the Coriolis force is an
apparent force which arises because the frame of reference used to
describe the motion is rotating. See, for example, Holton (1992) and
Andrews et al. (1987) for a detailed physical explanation of each of these
terms.
The second fundamental equation expresses conservation of energy
(first law of thermodynamics):
dT 1 dp
cp − =Q (3.2)
dt ρ dt
where T is the temperature (K), cp represents the specific heat of
air at constant pressure and Q is the net heating rate per unit mass
(due to physical processes such as radiative effects or release of latent
heat). The first two terms express the relationship between pressure and
temperature, which describes expansion cooling or compression heating
associated with adiabatic (energy conserving) processes taking place in
a compressible fluid. The remaining term on the right-hand side of the
equation represents diabatic processes involving heating or cooling of

the parcel. In the stratosphere and mesosphere, this rate is principally
due to the difference between two terms: i) heating by absorption of
solar ultraviolet radiation by ozone and, in the upper mesosphere and
lower thermosphere by molecular oxygen, and ii) cooling by infrared
emission by carbon dioxide, ozone, water vapor, and other radiatively
active molecules. It therefore depends strongly on the atmospheric
distributions of these minor constituents. A detailed account of the
radiation and thermal budgets is presented in Chapter 4.
The third fundamental equation expresses conservation of mass (and
is called the continuity equation):
dρ · v = 0
+ ρ∇ (3.3a)
dt
In equations (3.1), (3.2) and (3.3), the total time derivative (also
called the material derivative along the flow) can be written:
d ∂
= + v · ∇ (3.4)
dt ∂t
The material derivative is the total time rate of change following the
motion. By combining (3.3) and (3.4) one finds an alternative form of
the continuity equation
∂ρ
+ ∇ · (ρv)
=0 (3.3b)
∂t
The first term on the left-hand side of this equation represents the local
time rate of change (at a fixed point in space) and the second term
describes the advective transport of mass by the velocity v.
The net radiative heating (Q) is the only source of external forcing
in the energy conservation equation. If we assume that Q = 0 and
consider a frame of reference which moves with a moving air parcel, we
can examine the thermodynamics of adiabatic air parcel displacements.
In this case, equation (3.2) becomes
dT 1 dp
cp = (3.5)
dt ρ dt
The ideal gas law (also called the equation of state) can be used to relate
p and T:
p = ρRT (3.6)
where R is the gas constant for air. Substituting for ρ using (3.6), we
obtain
dT R dp
= (3.7)
T cp p
Figure 3.7. Zonally averaged distribution of the potential temperature

((unbroken lines, in K)) from the surface to approximately 30 km altitude
(10 hPa). The isolines for the absolute temperature (dashed lines, in K)
are also shown. The tropopause is represented by the dotted line. Note
that isentropes corresponding to potential temperatures larger than 380 K
are located exclusively in the stratosphere (an area called the overworld). Air
parcels located in the lowermost stratosphere between the tropopause and the
380 K isentrope (an area called the middle world) are susceptible to crossing
the tropopause when adiabatically transported, and entering the troposphere
(also called the underworld). From Holton et al. (1995), based on Appenzeller
(1994).
which can be integrated to yield

T = Apκ (3.8)
where A is a constant and κ = R/cp = 2/7 ≡ 0.286 for dry air. Noting
that the ratio T/pκ remains constant under adiabatic conditions, one
can define the potential temperature Θ
κ
Θ = T [p0 /p] (3.9)
which represents the temperature that would be attained by an air
parcel if it were adiabatically compressed or expanded starting from
a temperature T and pressure p to a nominal pressure p0 . p0 is
usually taken as 1000 hPa, representative of the pressure at sea level.
Θ is therefore a conservative quantity in any dry adiabatic air parcel
displacement, and is sometimes used to evaluate air parcel trajectories
(see, e.g., Danielsen, 1961). Figure 3.7 shows the zonally averaged
distribution of the potential temperature along with temperature
contours. From Equation (3.9) it can be seen that in an isothermal
2
atmosphere the potential temperature varies with height as p− 7 .
The thermodynamic equation (3.2) can be expressed using the
potential temperature Θ rather than the absolute temperature T:
κ
dΘ po Q
= ≡q (3.10)
dt p cp
This form allows for ready evaluation of the influence of external heating
(Q) on Θ, and expresses that, in the absence of heat input or loss [Q =
0], the potential temperature is conserved.
Newton’s second law, relation (3.1), is generally written as three
scalar equations, often expressed in spherical coordinates. Given the
independent variables:
λ the longitude in the easterly direction
φ the latitude in the northerly direction
zg the geometric altitude
we define
r = a + zg as the distance from the center of the Earth
(a is the Earth’s radius).
u = r cos φ dλ/dt as the zonal component of the wind
velocity v measured towards the east.
v = r dφ/dt as the meridional component of v measured
towards the north.
w = dzg /dt as the vertical component of v.
These components of motion are the scalar products of v with the unit
vectors in the zonal, meridional, and vertical directions, respectively.

i = Ω xr
r|
|Ωx
j = kxi
k = r
r
In the spherical coordinate system, the components of the acceleration
dv/dt are not simply given by the derivatives of the components of v
because the derivatives of the Cartesian unit vectors i,j, and k must be
considered. Figure 3.8 shows a schematic representation of the adopted
coordinate system.
The equations of motion then become:
du tan φ 1 1 ∂p
= uv − uw + 2Ω v sin φ − 2Ω w cos φ − + Fλ
dt r r ρr cos φ ∂λ
(3.11a)
dv tan φ 2 1 1 ∂p
=− u − uw − 2Ω u sin φ − + Fφ (3.11b)
dt r r ρr ∂φ
dw 1 1 1 ∂p
= u2 + v2 + 2Ω u cos φ − g − + Fz (3.11c)
dt r r ρ ∂zg
Figure 3.8. Representation of the adopted coordinate system.
with the material derivative (or derivative following the motion) defined
as
d ∂ u ∂ v ∂ ∂
= + + +w (3.11d)
dt ∂t r cos φ ∂λ r ∂φ ∂zg
and where Fλ , Fφ , and Fz are the 3 components of the dissipative force
F.
The last of these three momentum equations can be greatly simplified
by retaining only the two dominant terms:
1 ∂p
+g=0 (3.12)
ρ ∂zg
This simplification is commonly called the hydrostatic approximation.
The approximation states that, in the vertical direction, the most
important forces acting on a parcel of air are the vertical pressure
gradient and gravity. This approximation is common in dynamical
models, although it neglects some phenomena such as sound waves, and
is not appropriate for high resolution models. The ideal gas law can be
used to substitute for ρ in the above equation to obtain
dp dzg
=− (3.13)
p Hg
where Hg is called the atmospheric scale height. Hg is given by
RT kT
Hg = = (3.14)
g mg
if m is the molecular mass of air and k is the Boltzmann’s constant.

Relation (3.13) can be used to characterize the vertical structure of the
atmosphere in a convenient form. For example, the vertical distributions
of the pressure and air density can be represented by, respectively
⎡ ⎤
zg
⎢ dh ⎥
p (zg ) = po exp ⎣− ⎦ (3.15a)
Hg (h)
zg,o
and
⎡ ⎤
zg
no To ⎢ dh ⎥
n (zg ) = exp ⎣− ⎦ (3.15b)
T(z) Hg (h)
zg,o
where the subscript o denotes that the corresponding value is defined at

some reference altitude zg,0 . If the scale height (and the temperature)
vary little with altitude, one can write

zg − zg,o
p (zg ) ≈ po exp − (3.16a)
Hg
and

zg − zg,o
n (zg ) ≈ no exp − (3.16b)
Hg
The atmospheric scale height is roughly 7 ± 1 km in the homosphere. It
represents the altitude which corresponds to a reduction by 1/e in the
pressure or the air concentration. The distribution of any atmospheric
constituent i can also be characterized by its concentration, ni (zg ), or its
scale height, Hi (zg ). When the scale height of a minor species is equal to
the atmospheric scale height, Hg (z), the constituent is well mixed and
its mixing ratio (or mole fraction) Xi = ni /n is constant with altitude
(n is the total air density). It can be shown that Hg represents twice the
distance through which atoms or molecules having the equipartition of
translational energy, 1/2 kT, can rise in the vertical direction against the
force of gravity (Campbell, 1977). Table 3.1 presents a possible reference
atmosphere for mid-latitudes, illustrating the variation of temperature,
pressure, air density, potential temperature, and scale height with
geometric altitude.
The decrease of temperature with increasing altitude is called the
lapse rate. Combining the hydrostatic equation and the expression for
potential temperature, we can evaluate the dry adiabatic lapse rate (Γd ),
i.e., the temperature decrease which would be associated with a vertical
adiabatic displacement. Logarithmic differentiation of equation (3.9)
Table 3.1 Example of typical values of physical parameters below

120 km at middle latitudes. From the U. S. Standard Atmosphere (1976)
Scale Potential
Altitude Temperature Height Pressure Concentration Temperature(2)
(km) (K) (km) (hPa) (cm−3 ) (K)
0 288 8.4 1013.3 2.55 (19)(1) 288

5 256 7.5 540.5 1.53 (19) 307
10 223 6.5 265.0 8.61 (18) 327
15 217 6.4 121.1 4.04 (18) 398
20 217 6.4 55.3 1.85 (18) 498
25 222 6.5 25.5 8.33 (17) 636
30 227 6.6 12.0 3.83 (17) 807
35 237 6.9 5.7 1.74 (17) 1043
40 250 7.3 2.3 6.67 (16) 1426
45 264 7.7 1.5 4.12 (16) 1702
50 271 7.9 8.0 (-1) 2.14 (16) 2091
55 261 7.6 4.3 (-1) 1.19 (16) 2405
60 247 7.2 2.2 (-1) 6.45 (15) 2757
65 233 6.8 1.1 (-1) 3.42 (15) 3170
70 220 6.4 5.2 (-2) 1.71 (15) 3709
75 208 6.1 2.4 (-2) 8.36 (14) 4374
80 198 5.8 1.1 (-2) 4.03 (14) 5205
85 189 5.5 4.5 (-3) 1.72 (14) 6416
90 187 5.5 1.8 (-3) 6.98 (13) 8250
95 188 5.5 7.6 (-4) 2.93 (13) 10613
100 195 5.7 3.2 (-4) 1.19 (13) 14098
105 209 6.1 1.5 (-4) 5.20 (12) 18767
110 240 7.0 7.1 (-5) 2.14 (12) 26690
115 300 8.8 4.0 (-5) 9.66 (11) 39313
120 360 10.5 2.5 (-5) 5.03 (11) 53963
(1) read 2.55 (19), for example, as 2.55 × 1019 .
(2) calculated with p0 = 1013 hPa.
holding Θ constant yields:

1 dT κ dp
= (3.17)
T dzg p dzg
Substituting for dp/dzg from equation (3.13), one obtains

dT g
Γd = − = = 10K/km (3.18)
dzg Θ=ct cp
If potential temperature varies with height, then the actual lapse rate
(Γ = − dz
dT
g
) will differ from the dry adiabatic lapse rate. This difference
indicates the importance of diabatic processes, such as the absorption

of ultraviolet radiation by ozone, for example. The difference between
the actual and the dry adiabatic lapse rates provides a measure of the
tendency of a displaced air parcel to return to its original position. The
static stability of the atmosphere is often characterized by the buoyancy
or Brunt-Vïs¨
V al
V¨ ä¨ frequency N defined as
−g
N2 = [Γ − Γd ] (3.19a)
T
or, in terms of potential temperature,
∂ (ln Θ)
N2 = g (3.19b)
∂zg
If N2 > 0 (Γ < Γd ), i.e., if the actual temperature T increases

with height or decreases less rapidly than the adiabatic lapse rate,
the atmosphere is stably stratified. Such conditions apply in most
regions of the atmosphere and specifically in the stratosphere where the
temperature increases with height. Therefore, an air parcel lifted from
its equilibrium position tends to sink back to its original position. N can
be regarded as the frequency of the oscillation movement produced by
the Archimedean restoring force that applies to the air parcel. Typical
values of the buoyancy period (1/N) are 5 minutes in the stratosphere
and 10 minutes in the troposphere. The occasional breakdown of static
stability in the troposphere (for example, when the Earth’s surface is
becoming warm during summer months) generates convection with rapid
vertical exchanges of heat and mass.
As can be seen from the hydrostatic equation (3.12), there exists a
single-valued monotonic relationship between pressure and geometric
height in a static atmosphere. Therefore, meteorologists often use
the pressure (rather than the altitude) as the independent vertical
coordinate. It is also convenient to use the log-pressure altitude

p
z = −H ln (3.20)
ps
as the vertical coordinate because it scales nearly linearly with geometric

height zg . (It would scale exactly in an isothermal atmosphere). In this
expression, H is a constant scale height, often chosen to be 7 km, and ps
a reference pressure, often chosen to be 1000 hPa. When this definition
is adopted, the equations of motion as well as the thermodynamic
and continuity equations assume simpler forms, and their solution is
facilitated.
It is also convenient to introduce the geopotential, Φ, defined as the

work required to raise a unit mass from mean sea level to height zg
zg
Φ= gdz (3.21)
0
Using the hydrostatic approximation, one derives the following

expression
RT RT
dΦ = gdzg = −RTd ln p = dzg = dz (3.22)
Hg H
which states that the rate of change of the geopotential relative to the
log-pressure altitude is proportional to temperature.
By neglecting minor terms and noting that the depth of the
atmosphere is small compared to the radius of the Earth (r a), the
equations of motion become in spherical log-pressure coordinates:

du u tan φ 1 ∂Φ
− f− v+ = Fλ (3.23a)
dt a a cos φ ∂λ

dv u tan φ 1 ∂Φ
+ f+ u+ = Fφ (3.23b)
dt a a ∂φ
∂Φ RT
− =0 (3.23c)
∂z H
where f = 2Ω sin φ is the Coriolis parameter (vertical component of the
Earth rotation vector with the angular velocity Ω = 7.292 × 10−5 s−1 ).
Equations (3.23a) and (3.23b) are statements of Newton’s second law of
motion applied to a fluid continuum. They are often referred to as the
momentum equations because they describe the change of momentum
per unit mass along the two horizontal coordinate directions. Equation
(3.23c) is the hydrostatic equation expressed in terms of the geopotential
and in log pressure coordinates, and is equivalent to equation (3.12). The
law of conservation of energy is given by
dT κT Q
+w = (3.24a)
dt H cp
or equivalently by
dΘ
=q (3.24b)
dt
where, as stated in (3.10),
κ
p0 Q
q≡
p cp
Mass conservation (continuity) is expressed by

1 ∂u 1 ∂ (v cos φ) 1 ∂ρ0 w
+ + =0 (3.25)
a cos φ ∂λ a cos φ ∂φ ρ0 ∂z
where ρ0 = ρs exp (−z/H), and ρs is a reference density at the surface
(z = 0). The operator d( )/dt in spherical coordinates is again defined
by (3.11d). Equations (3.23) through (3.25) are called the primitive
equations because they contain almost no approximations, except for
those of hydrostatic balance, the neglect of Coriolis acceleration terms
due to the horizontal component of the Earth’s rotation vector, and the
substitution of the Earth’s radius a for variable r.
These equations define the basis for a theoretical treatment of the
atmospheric circulation. Together with suitable boundary and initial
conditions they are used in primitive equation models of the atmosphere
(see Section 3.9).
It can be easily deduced from equations (3.10) and (3.20) that the
potential temperature increases with the pressure-altitude z as

κz
Θ = T exp (3.26)
H
Because, in a stably stratified atmosphere, the potential temperature
increases monotonically with height, it is often used as “the vertical
coordinate” in atmospheric problems. Such a system is called the
isentropic coordinate system because a constant Θ surface corresponds
to a surface of constant entropy S (Θ) = cp ln Θ + constant. In this
framework, the vertical velocity
dΘ
Θ̇ = =q (3.27)
dt
is proportional to the diabatic heating rate Q. The “horizontal”
momentum equations are expressed by

dΘ u u tan φ ∂Ψ
− f+ v+ = Fλ (3.28a)
dt a a cos φ∂λ

dΘ v u tan φ ∂Ψ
+ f+ u+ = Fφ (3.28b)
dt a a∂φ
where Ψ = cp T + Φ is the Montgomery stream function, and where the
material derivative is now expressed as
dΘ ∂ u ∂ v ∂ ∂
= + + + Θ̇ (3.29)
dt ∂t a cos φ ∂λ a ∂φ ∂Θ
The hydrostatic relation takes the form
∂Ψ T
= cp (3.30)
∂Θ Θ
and the mass conservation equation is given by

∂σ ∂ (σu) 1 ∂ ∂
+ + (σv cos φ) + σ Θ̇ = 0 (3.31)
∂t a cos φ∂λ a cos φ ∂φ ∂Θ
where
1 ∂p
σ=− (3.32)
g ∂Θ
is called the pseudodensity.
The conceptual advantage of the isentropic system is that, in the
absence of diabatic heating (Θ̇ = 0), the motion becomes two-
dimensional: it has no “vertical” component and air parcels are displaced
along “horizontal” isentropic surfaces. In this case, isopleths (surfaces
of constant mixing ratios) for long-lived tracers tend to be parallel to
isentropic surfaces. In the presence of diabatic processes with “cross
isentropic” motions, these isopleths tend to slope slightly across the
isentropes. In the stratosphere, diabatic processes are sufficiently slow
that air parcels remain close to their isentropes for typically 1-2 weeks.
Isentropic coordinates are therefore convenient to analyze motions and
tracer transport over such time scales. In the interpretation of the
results, however, it should be remembered that isentropic surfaces move
(often significantly) relative to geometric or pressure height.
3.3.2 The Geostrophic Approximation
Scale analysis of the primitive momentum equations, (3.23a,b),

reveals that, for typical wind speeds and spatial scales observed in the
atmosphere, the time derivative and curvature terms in these equations
are small compared to the pressure gradient and Coriolis accelerations
(see Holton, 1992, for details). In particular, if we scale the time
derivative term in equation (3.23a) as follows:
du U U2
≈ ≈ (3.33)
dt (L/U) L
where U is a typical zonal wind speed (order of 10-100 m/s for the
stratosphere) and L is a typical length scale (order of 1000-10,000 km),
the Coriolis acceleration is
2Ω sin φ u ≈ fU (3.34)
where f is the Coriolis parameter (roughly 10−4 s−1 except near the
equator). The ratio of the inertial (du/dt) to Coriolis accelerations
defines the Rossby number ((Ro ) of the flow, given by

U2 /L U
Ro = = (3.35)
fU fL
Substituting the numbers given above, we obtain Ro ≈ 10−1 , indicating
that the inertial term (du/dt) is significantly smaller than the Coriolis
acceleration in the momentum equation for the case considered here. It
can be shown that for this case, the only term in equation (3.23a) which
is of comparable magnitude to the Coriolis term is the one representing
the horizontal pressure gradient force, 1/(a cos φ) ∂Φ/∂λ, and indeed
these two terms are nearly in balance with one another. It can also be
shown that Ro ≤ 0.1 is a good assumption for all scales of motion larger
than about 1000 km, except very near the equator, where f → 0.
Retaining only the Coriolis and pressure gradient terms in (3.23a) and
(3.23b), we obtain the geostrophic wind equations:
1 ∂Φ
fu = − (3.36a)
a ∂φ
1 ∂Φ
fv = (3.36b)
a cos φ ∂λ
These equations show that stratospheric winds blow approximately
parallel to the contours of the geopotential field. An alternative
version of the geostrophic balance can be obtained by substituting
the hydrostatic approximation as given in (3.23c) into (3.36a,b), after
differentiating with respect to z:
∂u R 1 ∂T
f =− (3.37a)
∂z H a ∂φ
∂v R 1 ∂T
f = (3.37b)
∂z H a cos φ ∂λ
These are the thermal wind equations, which state that the vertical shear
of the horizontal wind field is proportional to the horizontal temperature
gradient. These equations illustrate that the wind structure is strongly
coupled to the temperature gradient, underscoring the close relationship
between thermodynamic conditions and advective motion. Note, for
example, that the zonal wind, u, as shown in Figure 3.4, increases
with altitude in the presence of an equatorward temperature gradient
(Figure 3.2) in both the northern and southern hemispheres. It decreases
for a poleward temperature gradient. Thus the region of maximum zonal
wind (the “jet”) near the polar night region can be understood to be a
manifestation of geostrophic balance as expressed by the thermal wind
equation.
Although the geostrophic wind equations are an excellent approxi-

mation of the horizontal wind field in atmospheric motion systems of
scales greater than about 1000 km, they do not contain time derivatives
(tendencies), and are thus diagnostic, not prognostic (predictive)
equations. However, by performing a more detailed scale analysis, it
is possible to derive a system of prognostic equations which still embody
the geostrophic balance as the lowest order approximation, but include
in certain terms the possibility of departures from geostrophy. These
are known as the quasi-geostrophic equations. There is no single quasi-
geostrophic system, since the form of the equations depends, among
other things, on the horizontal scale of the motions of interest compared
to the Earth’s radius.
3.3.3 The Potential Vorticity
For a large class of scale of motions, the primitive equations (which are
rather expensive to solve numerically) can be replaced by approximate
equations for the vorticity or the potential vorticity. Relative vorticity
can be thought of as a measure of the rotation of the fluid with respect to
a coordinate sytem fixed to the Earth. Its vertical component is defined
in spherical coordinates as
1 ∂v 1 ∂ (u cos φ)
ξ= − (3.38)
a cos φ ∂λ a cos φ ∂φ
By cross-differentiating and combining the primitive equations (3.23a
and 3.23b) and using the continuity equation (3.25), one obtains the
vorticity equation

d (ξ + f) 1 ∂u 1 ∂ (v cos φ)
= (ξ + f) +
dt a cos φ ∂λ a cos φ ∂φ
1 ∂Fφ 1 ∂ (Fλ cos φ)
+ − (3.39)
a cos φ ∂λ a cos φ ∂φ
The sum ξ + f is called the absolute vorticity. It represents the spin
of air parcels relative to a coordinate system fixed to the Earth and
of the planetary vorticity (represented by the Coriolis parameter f),
which accounts for the fact that the coordinate system is rotating at
angular vorticity Ω. As shown by equation (3.39), even in the absence
of frictional forces, the absolute vorticity is not a conserved quantity.
Its tendency is proportional to the horizontal component of the wind
convergence. If this component is positive, absolute vorticity filaments
will gather closer together, increasing the magnitude of the air parcel
spin by decreasing the cross-sectional area of the vortex tubes (see e.g.,
Pedlosky, 1987).
A more general form of the vorticity is provided by Ertel’s potential
vorticity (Rossby, 1940; Ertel, 1942). In the log-pressure coordinate
system, this quantity is defined by
1 ∂Θ
P= (ξ + f) (3.40a)
ρs ∂z
and is therefore the product of the absolute vorticity and the vertical
gradient of the potential temperature. The potential vorticity is a
measure of the vertical component of angular momentum of fluid
elements. In the stratosphere, the value of P on an isentropic surface
increases monotonically from the South Pole to the North Pole as the
Coriolis factor f increases with latitude. The value of P is generally
negative in the Southern hemisphere and positive in the Northern
hemisphere. In the vicinity of the tropopause (where the vertical
gradient of the potential temperature is large), a strong vertical gradient
is also observed in the potential vorticity.
As indicated earlier, atmospheric motions are often advantageously
represented in an isentropic coordinate system. In this case, Ertel’s
potential vorticity is defined by
1
P = (ξ + f) (3.41)
σ
where σ is the “pseudo-density” (see expression (3.32)). Starting from
the primitive equations (3.28a-b), one can show that P satisfies the
following equation

dΘ P 1 ∂ (Fλ cos φ) ∂Fφ ∂Q ∂v ∂Q ∂u
= − + − + cos φ
dt σa cos φ ∂φ ∂λ ∂λ ∂Θ ∂φ ∂Θ
∂Q ∂P
+P −Q (3.42)
∂Θ ∂Θ
where the material derivative in the isentropic system is defined by
(3.29). This expression shows that for frictionless and adiabatic
conditions (Fλ = Fφ = 0; Q = 0), Ertel’s potential vorticity is a
conserved quantity following the motion.
Apart from the absolute vorticity (ξ+f), the Ertel’s potential vorticity
varies in the vertical as Θ9/2 (see Lait, 1994). It is therefore useful to
define a modified potential vorticity

Θ −9/2
Π=P (3.40b)
Θ0
which removes the quasi-exponential variation with height while
preserving the conservation properties of P. (Θ0 is a reference potential
temperature chosen, for example, as 420 K, see Lait, 1994). Figure 3.9
740
590
500 16
6
380
-8 8
-2
2
Figure 3.9. Zonally averaged distribution of the modified potential vorticity

(unbroken lines in 10−4 K m2 s−1 kg−1 ) from the surface to approximately 30 km
altitude (10
( hPa). ) The potential temperature (K)
( ) is represented by dashed lines
and the tropopause by a dotted line. From Appenzeller (1994).
shows the zonally averaged distribution of the modified potential

vorticity superimposed on the potential temperature for January.
The concept of potential vorticity is of central importance in fluid
dynamics and in meteorology. When the powerful constraint of vorticity
conservation can be applied, the physical interpretation of atmospheric
motions becomes more accessible (Hoskins et al., 1985). The fact that
in many situations dissipative processes do not play a major role on
advective time scales makes this quantity very useful as a tracer of
atmospheric motions (see Danielsen, 1961; McIntyre and Palmer, 1983).
Figure 3.10 presents an example of an observationally derived potential
vorticity distribution on the 850 K isentropic surface. Potential vorticity
is often expressed in 10−6 K m2 kg−1 s−1 , referred to as PV units (Holton
et al., 1995). This figure shows that, as expected, the potential vorticity
tends to increase with latitude. Potential vorticity contours, however,
are affected by planetary-scale waves. The behavior of long-lived
chemical tracers is very similar to that of potential vorticity. Knowledge
of potential vorticity distributions from meteorological analyses can
therefore be used to extrapolate limited observations of chemical tracers
to a broad region of the atmosphere (see e.g., Coffey et al., 1999).
Potential vorticity has, however, an important property not shared by
chemical tracers: It is not only advected by the flow, but it actually
determines the distribution of the flow itself. Changes in potential
vorticity, therefore, induce changes in the wind and temperature fields.
Figure 3.10. Global distribution of the potential vorticity on the 850 K

isentropic surface during a wintertime planetary wave event. The shaded region
over the Pacific ocean with a weak gradient is characterized by nonlinear wave
dissipation and strong quasi-horizontal mixing. This region is referred to as the
“surf zone”. Wind vectors are also indicated and provide information about
large-scale transport. Courtesy of A. O’Neill, University of Reading, UK.
Finally, it is interesting to note that, since potential vorticity is

a conserved quantity, air parcels that are displaced to an area with
different PV values develop a differential motion which tends to restore
them to their original PV area. Thus, strong PV gradients act as
dynamical barriers to transport. Barriers against horizontal mixing are
observed in the lower stratosphere near the polar vortex and at the
boundary between tropical and extratropical air masses. Wave events
distort PV surfaces often leading to the formation of thin filaments which
are pulled away from the barriers.
3.4 Atmospheric Waves
Waves are key dynamical features of the atmosphere. They can

be defined as propagating disturbances of material contours whose
acceleration is balanced by a restoring force. This force can be produced

by buoyancy in a stably stratified fluid (e.g., gravity waves) or by the
Coriolis force (e.g., inertial waves). In the first case, the force acts
to oppose vertical displacements, while in the second case, it opposes
horizontal displacements.
Different types of waves can be distinguished: resonant modes
produced by small random forcing are referred to as free modes. The
observed 2, 5, and 16-day large-scale atmospheric waves belong to this
category. Waves excited at the Earth’s surface, for example by air flow
over topography, by land/ocean temperature contrasts or by convective
systems, are called forced modes. Some such waves can propagate
to high altitudes where they dissipate and transfer momentum to the
atmospheric circulation. Waves are often refracted in the atmosphere,
and a refractive index which determines the ray path along which the
wave propagates can be defined. Large-scale waves that propagate
vertically as well as horizontally are called global modes, while waves
that remain trapped in the tropics and propagate only in the zonal
and vertical directions are referred to as equatorial modes. If the waves
propagate over several scale heights in the vertical direction, they are
said to be internal modes. Conversely, if they are rapidly attenuated as
they propagate upwards, they are called external modes.
Waves are characterized by their amplitudes and their phases, as well
as by their phase speeds (which describe the rates at which the crest and
the trough of the waves propagate) and their group speeds (which are a
measure of the rate at which the energy of the disturbance propagates).
The evolution of disturbances on a given basic atmospheric state
is described by the wave momentum, continuity, and thermodynamic
equations. These are obtained by replacing all fields ψ in the primitive
equations by the sum of the basic state ψ and a wave perturbation
of ψ (see Section 3.6.1), and by subtracting the equations applied
to the basic state. The basic state is often taken to be the zonally
averaged state of the atmosphere. For small-amplitude disturbances,
second-order terms can be neglected and waves are approximated as
linear. In the case of large disturbances, the nonlinear terms in the
wave equations become significant and nonlinear wave-wave interactions
become important. The solution of the wave equations leads to
dispersion equations which express the conditions under which a wave
can propagate in the atmosphere.
Gravity waves are oscillations with relatively short horizontal
wavelengths (typically 10-1000 km) that arise in a stably stratified
fluid when air parcels are being displaced vertically. These waves are
produced by air flow over mountains (orographic waves) or by other (not
well identified) non-orographic sources such as thunderstorms, frontal
Figure 3.11. Approximate altitude profiles of the mean zonal winds in winter
(left panel) and summer (right panel). The permitted phase speeds for the
propagation of gravity waves and their breaking levels are also shown. Adapted
from Lindzen (1981).
systems or instabilities. The propagation of gravity waves through the

atmosphere depends on the wind distribution and thermal structure,
which varies markedly with season and the static stability. Specifically
it can be shown that, when the phase speed c of a wave is equal to the
zonal wind speed u, the wave is absorbed. The point where this occurs
is referred to as the critical level for that particular wave. Figure 3.11
shows how this absorption effect influences gravity wave propagation
at mid-latitudes for different seasons. Specifically, owing to the strong
filtering by passage through the stratosphere, mesospheric gravity waves
are predominantly westward propagating waves in winter and eastward
propagating waves in summer. In the absence of dissipation, the
amplitude of gravity waves grows as the inverse square of the density.
This can be understood by considering that for kinetic energy
1
E = ρo |v (z)|2 (3.43)
2
to remain constant, the amplitude of the wave must grow as
1
|v (z)| = A [ρ0 (z)]− 2 (3.44a)
where A is a constant. Since ρ0 (z) varies with altitude as ρs exp(−z/H),

then

A z
|v (z)| = exp (3.44b)
ρs 2H
i.e., the wave amplitude grows exponentially with height. At some
altitude, the gravity wave will have grown so large that its temperature
perturbation will produce a superadiabatic lapse rate and become
convectively unstable. At this point, the wave is said to “break”. Note
that if, in addition to buoyancy, the Coriolis force acts as a restoring
force, the waves are referred to as inertia-gravity waves. This effect
becomes significant if the horizontal wavelength of the waves is larger
than approximately 300 km.
Gravity waves provide the major source of dynamical variability
in the mesosphere. The presence of gravity waves has been inferred
from oscillations observed in airglow emissions (Walterscheid et al.,
1987; Makhlouf et al., 1995; 1998; Swenson and Gardner, 1998).
Lidar observations (see Figure 3.12 and also Hauchecorne et al.,
1987) also show, for example, that such waves can create alternating
layers with near adiabatic lapse rate temperature profiles and strong
temperature inversions. These waves can also affect some chemical
species, particularly in the mesosphere and lower thermosphere. In
addition, as first pointed out by Leovy (1966), photochemistry can damp
or enhance the amplitude of gravity waves and even destabilize them in
the vicinity of the mesopause (see also Xun Zhu and Holton, 1986; Xu,
1999).
Figure 3.12. Sequence of temperature profiles measured by Lidar at Bicarosse,

France (from 19:00 to 23:00) showing the presence of gravity waves in the
mesosphere. From Hauchecorne et al. (1987).
Rossby waves are planetary-scale waves that owe their existence to

meridional gradients in potential vorticity. These waves are westward
propagating relative to the mean flow. The waves with the largest
amplitudes, however, are quasi-stationary because they are produced
by stationary sources including large-scale orography and land-sea
contrasts. These Rossby waves propagate upwards and equatorwards
in the middle atmosphere. It can be shown, however, that upward
propagation requires that the zonal wind be westerly (from west to
east) with a velocity smaller than a critical value. This critical value
is a function of the horizontal wavenumber of the wave and is typically
100 m/s in the case of wavenumber 1. On the basis of this criterion
(called the Charney-Drazin criterion), Rossby waves propagate into the
stratosphere primarily during winter, which explains the quasi-zonal
symmetric nature of the dynamical fields during summer, when wave
propagation is much reduced. Figure 3.13 shows the seasonal difference
in stratospheric meteorological fields. Note, however, that free Rossby
waves with smaller amplitudes do propagate in the summer and may
cause substantial mixing in the lower stratosphere at high latitudes
(see Wagner and Bowman, 2000). The presence of such waves could
explain the variability of ozone observed at mid- to high latitudes
during the summer in the altitude range of 20-30 km (see Hoppel et al.,
1999). In practice, the Charney-Drazin criterion is satisfied only for
wavenumbers 1-3, so that only the longest Rossby waves penetrate in
the extratropical winter stratosphere. An extension of the tropospheric
weather systems (wavenumbers 5 to 15) can, however, be observed in
the lower stratosphere.
When the planetary waves reach environments for which nonlinear
effects can no longer be neglected, wave breaking may occur. In this
case, the reversible undulation of material contours (mixing ratios, PV)
described by linear wave theory is replaced by rapid and irreversible
deformation of these contours with the generation of small-scale motions
in situ. A simple mathematical treatment of the wave equation for the
quasi-geostrophic potential vorticity shows that this situation occurs
in the vicinity of a critical surface along which the phase velocity
of the wave is equal to the zonal mean wind (u = c). When
wave breaking occurs, considerable deformation of material contours
(potential vorticity, trace constituents) takes place, producing a cascade
of gradually smaller scale motions, and eventually irreversible mixing
of air parcels along isentropic surfaces. Regions where large scale
mixing associated with Rossby wave breaking is observed (mid-latitude
winter hemisphere) are referred to as the “surf zone” (McIntyre and
Palmer, 1983; 1984) and are characterized by weak horizontal gradients
in conservative tracers. The surf zone is bounded by narrow regions of
strong PV gradients which act as dynamical barriers: the polar vortex

and the subtropical barrier. The existence of a surf zone between 25 and
45◦ S is refected in the distribution of nitrous oxide shown in Figure 3.25.
The surf zone is also apparent over the Pacific Ocean in Figure 3.10 where
the potential vorticity is very uniform.
234
3
Figure 3.13. Geopotential height ((km)) and temperature ((K)) on the 10 hPa
surface in the stratosphere during Northern Hemisphere winter (18 February
1999 — left) and summer (9 July 1999 — right). Planetary-scale waves
are visible in the winter hemisphere with the polar vortex displaced towards
northern Europe and a “high” visible over Western Canada and the Aleutian
Islands. The temperature is lowest over the north Atlantic and is highest over
Siberia. This situation is representative of a wavenumber 1 event. In the
summer hemisphere, planetary-scale wave activity is low and the geopotential
height and temperature fields are nearly symmetric around the North Pole
(Courtesy of W. Randel, NCAR, 2001).
The change of sign of the Coriolis parameter at the equator leads

to a specific class of large-scale waves that are generally trapped
laterally in the tropics, but propagate vertically and zonally. Kelvin
waves, which propagate eastward with a phase velocity of typically
30 m s−1 , are characterized by a vertical wavelength of typically
10 km and wavenumbers of 1-3. Mixed Rossby-gravity waves propagate
westward, with vertical wavelengths of the order of 6-8 km, and with 4-5
wavenumbers in the zonal direction. The horizontal structure of these
types of waves (geopotential and winds) is shown in Figure 3.14.
In the equatorial upper stratosphere and lower mesosphere, low
zonal wavenumber “pancake” structures have frequently been observed
(Hitchman et al., 1987), and are suggestive of inertially unstable motions.
The sources and dynamics of such disturbances need to be elucidated.
Finally, the tides observed in the mesosphere and thermosphere should
be mentioned. Atmospheric solar tides are global-scale waves with
periods that are harmonics of a 24-hour day. These oscillations are
Figure 3.14. Geopotential and wind structure of an equatorial Kelvin wave

(upper panel) and mixed Rossby-gravity wave (lower panel).
driven primarily by the periodic absorption of solar ultraviolet energy

by ozone in the middle atmosphere and of solar infrared energy by water
vapor in the troposphere. An additional thermal driver is provided
by latent heat which is stored as water vapor, transported within the
troposphere and released when the vapor precipitates.
Migrating tides are those waves which propagate westward with
the apparent motion of the Sun. They have significant diurnal and
semi-diurnal components. Non-migrating tidal components are also
harmonics of the solar day, but they may be stationary, or propagate
either eastward or westward. Their dominant source is provided by
latent heat release in the troposphere, and thus related to meteorological
processes in the lower atmosphere. A small semi-diurnal tide is also
forced by the gravitational attraction of the Moon.
The propagation of atmospheric tides has been described by Lindzen
(1967), Chapman and Lindzen (1970), Lindzen and Blake (1970), Forbes
(1982a, b), Hagan (2000) and many others. These waves have also been
observed from space instruments such as HRDI and WINDII on board
the Upper Atmosphere Research Satellite (UARS; see e.g., Hays et al.,
1994). Figures 3.15a-b show the signature of the tides in the zonally
averaged winds measured at 12:00 h local solar time (McLandress et al.,
1996). Tide ampitudes increase with height in the mesosphere due to
decreasing air density. Figure 3.15c shows the amplitude of the diurnal
tide in the meridional wind at 95 km and 20◦ N between 1992 and 1995
(Burrage et al., 1996). There is some indication that the tide amplitude
is modulated by the quasi-biennial oscillation (see Section 3.7). Large
tidal variations are also observed in the temperature (1 K at 80 km,
10 K at 100-120 km, and 100 K at 150 km). The tidal oscillation in the
vertical wind produces strong variations in the concentration of chemical
species that exhibit large vertical concentration gradients (e.g., atomic
oxygen, nitric oxide).
The amplitude of the diurnal tide is largest around 30◦ latitude in
the case of the horizontal wind components. In the case of temperature,
pressure and the vertical wind component, the tidal perturbation is
largest at the equator. Figure 3.15d shows the diurnal vertical wind
perturbation (migrating tide) in April, when the tidal amplitude is
largest. The semi-diurnal component is observed at all latitudes above
80 km altitude. Ground-based and satellite instruments, as well as
dynamical models, show that tides often govern the dynamics of the
upper mesosphere and lower thermosphere. For example, numerical
experiments (see, e.g., Liu and Hagan, 1998) show that the tidal wind
may have a strong impact on gravity wave breaking and hence on local
heating/cooling of the mesosphere. Tidal dampening occurs above 80 km
altitude, where it provides a source of momentum and causes turbulence.
Figure 3.15a,b. Zonally averaged wind velocities for the equinox period
(February 12 to May 3, 1993) at 12:00 h local solar time derived from the
combined observations of HRDI and WINDII on board the UARS satellite.
Zonal (a) and meridional (b) wind velocities (m/s) are shown between 65 and
140 km altitude. Dashed lines correspond to westward winds (easterlies) for the
zonal component and to southward winds for the meridional component. The
signature of solar tides is visible. Adapted from McLandress et al., 1996.
Figure 3.15c. Amplitude of the meridional wind variation at 95 km and 20◦ N

between October 1991 and April 1995 produced by the diurnal component of
the migrating tides. From Burrage et al., 1996.
Figure 3.15d. Calculated diurnal vertical wind perturbation (m/s) at 12 LT

associated with the migrating tide in April. Courtesy of M. Hagan, NCAR,
2001.
3.5 Effects of Dynamics on Chemical Species:

Transport
3.5.1 The Continuity Equation
The density of a chemical compound at a given point of the

atmosphere is derived from the continuity equation. The equation can
be understood by noting that the change in concentration results from
local sources (e.g., chemical reactions) and from transport processes.
The tendency associated with transport is given by the convergence
(or divergence) of the concentration of this chemical compound within
an infinitely small volume element surrounding the point under
consideration. Thus, for a chemical species i, the mass density ρi at
time t is governed by the following equation
∂ρi
+ ∇ · (ρiv) = Si (3.45a)
∂t
Note that the source term Si (expressed here in mass per unit volume and
time) is often expressed as the difference between rates of photochemical
production Pi and loss Li . Introducing the mass fraction or mass mixing
ratio µi = ρi /ρ where ρ is the air mass density, and making use of the
mass conservation equation for air (3.3b), it is easy to show that equation
dµi Si
= (3.45b)
dt ρ
is equivalent to (3.45a). Relation (3.45b) expresses that, in the absence
of chemical processes, the mixing ratio of a chemical tracer in an air
parcel is conserved along the parcel trajectory. Just as in the case of
the potential temperature and potential vorticity, tracer mixing ratios
can be used to characterize air motions in the atmosphere if dissipative
(chemical) processes are weak.
The continuity equations can also be expressed in terms of number
density ni = ρi /m and mole fraction or volume mixing ratio Xi = ni /n,
where n = ρ/m is the air number density, m = M/NA is the molecular
mass of air, NA is Avogadro’s number (6.02 × 1026 molecules per mol)
and M is the molar mass of air. In the homosphere, where M 29
kg/mol is approximately constant,
∂ni
+ ∇ · (niv) = Si (3.45c)
∂t
or
dXi Si
= (3.45d)
dt n
where the chemical source term Si is now expressed in particles per

unit volume and time (m−3 s−1 or cm−3 s−1 ), with the total air number
density n expressed in m−3 (or cm−3 ).
In spherical coordinates, these equations are expressed as
∂ni 1 ∂ (ni u) 1 ∂ (ni v cos φ) ∂ (ni w)
+ + + = Si (3.46a)
∂t a cos φ ∂λ a cos φ ∂φ ∂z
and
∂Xi u ∂Xi v ∂Xi ∂Xi Si
+ + +w = (3.46b)
∂t a cos φ ∂λ a ∂φ ∂z n
3.5.2 Time Constants for Transport
In Chapter 2, the time constant appropriate to photochemical

processes (ττchem ) was discussed, and it was shown that this time constant
can be readily derived from knowledge of the rate of loss of chemical
species. The time constants for dynamical effects on chemical species
are somewhat more difficult to evaluate.
If we consider only the effects of advection and chemistry, the
continuity equation for a chemical species i can be written as:
∂ni ∂ ∂ ∂
= Si − (ni w) − (ni v) − (ni u) (3.47)
∂t ∂z ∂y ∂x
where ni is the number density of species i, and where, for simplicity,
we have used Cartesian coordinates. To derive the time constant for
dynamical processes (ττdyn ), some assumption must be made regarding
the gradient of compound i. If we assume, for example, that the vertical
distribution of chemical species i can be described by a scale height, Hi :
ni = ni,o exp (−z/Hi ) (3.48)
where z is the altitude and ni,o is the density of i at a reference altitude,
then
∂ni 1 1
=− ni,o exp (−z/Hi ) = − ni (3.49)
∂z Hi Hi
Consider, for the purpose of illustration, transport only by vertical
winds:
∂ni ∂ w
= − (ni w) = ni (3.50)
∂t ∂z Hi
(assuming that w is constant over the spatial distance Hi ). By
integration,
w
ni,t = ni,t o exp ( t) (3.51)
Hi
so that the time required for species i to change by 1/e relative to

its initial value, ni,t o is given by Hi /w. This defines a time constant
for transport by the vertical winds. Assuming a scale height of about
5 km for chemical constituents and approximate values for the zonally
averaged mean vertical winds as derived from models, the time constant
for vertical exchanges is of the order of months in the stratosphere and
days in the upper mesosphere.
Time constants for transport by the meridional and zonal winds can
be derived in similar fashion, but in these directions it is somewhat
more difficult to characterize the typical gradients of chemical species.
Assuming for these horizontal changes a typical length scale of about
1000 km in the zonal and meridional directions, the time constant for
transport by the zonal winds is of the order of days throughout the
middle atmosphere, while that for mean meridional transport is of the
order of months in the stratosphere and days in the upper mesosphere.
When considering the formulation adopted in a simple one-
dimensional model, in which the mean vertical flux is represented by
“eddy diffusivity”

∂ni ∂ ∂ (ni /n)
= Si − −Kz (3.52)
∂t ∂z ∂z
(where Kz is the one-dimensional diffusion coefficient), the time constant

associated with vertical exchanges is given by H2 /Kz , where H is
the atmospheric scale height. It is of the order of months to years
in the lower stratosphere, and days in the mesosphere. Figure 3.16
presents an altitude profile of the approximate time constant for one-
dimensional eddy diffusion, as well as the chemical time constants of
the Ox (O + O3 ), HOx (H+OH+HO2 ), and NOy (N + NO + NO2
+ NO3 + HNO4 + 2×N2 O5 + HNO3 ) chemical families, and water
vapor. As a general rule, wherever the time constant for the dynamical
process under consideration is comparable to the chemical time constant,
then the effects of transport and chemistry will both be important
in determining the constituent’s density. For example, in the one-
dimensional framework, we can see that vertical eddy transport would
be important for NOy at all the altitudes shown, but not for HOx below
about 85 km. This is discussed more explicitly below.
Three different cases can be identified to characterize the effects of the
competition between dynamics and chemistry in determining constituent
distributions:
Figure 3.16. Vertical profiles of the chemical lifetimes of Ox , HOx , H2 O, and

NOy . Typical constants for vertical exchanges (ττD ) are also shown.
• τchem << τdyn . Under these circumstances, the species in question

will be in photochemical equilibrium (see Eq. 2.30), and the effects of
dynamical processes will not be directly important. Strong spatial
gradients often characterize the distribution of photochemically-
driven species. Dynamics may still be important in an indirect fashion
through the effects of temperature or coupling between chemical
species. For example, a very short lived species, such as HOx can be
produced or destroyed by reaction with a longer lived species which
does depend on transport (for example, H2 O, see Figure 3.16). Thus
an indirect coupling can be important even for very short-lived species.
• τchem >> τdyn . In this case the effect of dynamics is to reduce
substantially the gradient in the mixing ratio of chemical compounds.
For example, the chemical lifetime of N2 O is of the order of years in
the lower stratosphere, while the dynamical time constant associated
with the zonal wind is of the order of days. Thus the effect of
the zonal wind is to make the N2 O mixing ratio very uniform in
longitude. Moderate perturbations to the zonal wind speed will not
dramatically influence the constituent density. The major species, O2
and N2 , are so long lived in the middle atmosphere that they are
thoroughly mixed by atmospheric motions, and typical spatial and
temporal variations in transport do not influence them. This is not
the case in the thermosphere, where their chemical lifetimes are much
shorter.
• τchem ≈ τdyn . Under these circumstances, the effect of dynamics

can be quite large, and the species distribution depends critically on
both dynamics and chemistry. Chemical processes tend to introduce
or enhance spatial gradients in the distribution of the tracers, while
dynamics (e.g., mixing) tends primarily to reduce these gradients.
The time constants for meridional and vertical transport, are, for
example, comparable to the photochemical lifetime of N2 O in the
stratosphere, so that transport in the meridional plane is expected
to be quite important in determining its density, in contrast to zonal
transports, as discussed above.
The special case of zonal asymmetries should also be mentioned. The

vector wind is generally aligned along latitude circles in summer, but in
winter the influence of planetary waves can cause the vector wind to flow
across latitude lines (see Figure 3.13). Under these zonally asymmetric
conditions, the local wind speed in the meridional direction is much
greater than the mean meridional wind. Indeed, winds of zonal speeds
(tens of meters per second) can flow across latitude lines, and any species
exhibiting a latitude gradient will be affected.
3.5.3 The Age of Air
Long-lived trace gases of tropospheric origin penetrate into the

stratosphere, primarily in the tropics. If the mixing ratio of an inert
tracer increases with time at the tropical tropopause, its mixing ratio at
any point of the middle atmosphere will also increase, but with a certain
time lag. An estimate of this lag provides information on the rate at
which transport takes place. Observations of tracers such as CO2 , SF6 ,
and chlorofluorocarbons, whose concentration has been systematically
increasing with time, have shown that the time required to transport
air parcels from the equatorial tropopause to the mid- and high latitude
stratosphere is typically 3-7 years (Bischof et al., 1985; Schmidt and
Khedim, 1991; Pollock et al., 1992; Boering et al., 1996; Elkins et al.,
1996, etc.).
If large-scale advection alone was responsible for the transport of air
parcels from the tropopause to any location in the middle atmosphere,
the calculation of the transit time between these two locations would be
straightforward. Air parcels, however, do not maintain their integrity
over long time scales, because they are rapidly distorted by wave-induced
mixing processes into many smaller-scale components. Therefore, the
transit time from any point P0 to point P in the atmosphere cannot
be characterized by a single quantity, but rather by the statistical

distribution of the transit times for all material elements that comprise
the air parcel. Thus, for conserved tracers that enter the stratosphere at
the equatorial tropopause (point P0 ) with a time-varying mixing ratio
X (P0 , t), the mixing ratio at some point P can be expressed by adding
the contributions of all irreductible material elements which reach point
P at time t after having traversed different paths with different transit
times. Therefore (see Hall and Plumb, 1994)
t
X(P, t) = X(P0 , t − t ) G(P, P0 , t ) dt (3.53)
0
where G represents the Green’s function of the differential operator

governing the transport from P0 to P. The Green’s function can be
interpreted as the distribution of transit times and is therefore called
the age spectrum (Kida, 1983; Hall and Plumb, 1994; Hall and Waugh,
1997). Its value cannot be directly measured, but it can be estimated as
a function of time at any point P by applying a pulse (X(P0 , t) = δ(t)
in Eq. (3.53)) in the tracer mixing ratio at the equatorial tropopause.
The mean age Γ of middle atmosphere air is defined as the first moment
of the age spectrum:
∞

Γ(P, P0 ) = t G(P, P0 , t)dt (3.54)
0
Figure 3.17 shows an estimate of the mean age of air as a function of

altitude and latitude as derived from aircraft observations of CO2 and
SF6 . It suggests that the mean age in the tropics, which increases with
height, exhibits little vertical variation above 25 km altitude, where it
can be 4 or more years. At mid-latitudes, Γ increases monotonically
from the tropopause to the upper stratosphere and reaches 7 or more
years about 30 km altitude.
It is straightforward to show that, for purely advective motions
(without mixing), the age of air at a given point is simply the time by
which the mixing ratio at this point lags the value of the mixing ratio at
the equatorial tropopause. In a more general case, when advection and
mixing take place, the equivalence between age and lag time does not
hold, except in the long-term limit (t → ∞) when the mixing ratio of
the tracer at the equatorial tropopause increases linearly with time (see
Hall and Prather, 1993; Hall and Plumb, 1994). The measurement of
time lags for species such as CO2 or SF6 , whose mixing ratios increase
quasi-linearly in time (neglecting seasonal variations), provides a good
estimate of the age of middle atmosphere air.
Figure 3.17. Meridional distribution of the mean age of stratospheric air

(schematic representation of the annual mean expressed in years). From Hall
et al., (1999).
3.5.4 Tracer Correlations
As indicated earlier (see Section 3.3), in the absence of strong

dissipative processes (i.e., diabatic heating, friction), quantities such
as the potential temperature Θ and the potential vorticity P are often
sufficiently conserved to act as tracers of atmospheric motions. Long-
lived chemical compounds such as nitrous oxide (N2 O), methane (CH4 ),
chlorofluorocarbons (CFCs), and even, under many circumstances, ozone
(O3 ), can also be regarded as excellent tracers of motion, especially in the
lower stratosphere. Because of its long photochemical lifetime, nitrous
oxide is a suitable reference for comparing measurements of a variety of
trace gases (Lowenstein et al., 1989; Chang et al., 1996a; 1996b).
Observations made in the lower stratosphere have shown many cases
of strong quasi-linear correlations between the mixing ratio of different
long-lived species. These correlations are represented by scatter plots
as shown in Figure 3.18 (Chang et al., 1996a; 1996b). Such diagrams
suggest that, if the observed mixing ratios of 2 compounds collapse
to a compact curve, it is possible to deduce the mixing ratio of
one of them from the measurement of the mixing ratio of the other.
Furthermore, the departure from the correlation curve is an indication
that chemical processes have affected the corresponding airmass during
its displacement in the atmosphere. Such departure is clearly seen in the
Figure 3.18. Observed mixing ratio of several chemical compounds (O3 , NOy ,
SF6 , CH4 , CFC-12, HCl) represented as a function of the observed mixing ratio
of N2 O. These scatterplots are established with measurements made from the
ER-2 high altitude aircraft and during the spaceborne ATMOS experiment.
From Chang et al. (1996a, 1996b).
case of ozone if measurements are performed in air masses that have been
travelling in the polar regions during springtime when massive ozone
destruction takes place (e.g., Antarctic ozone hole). The compactness of
the correlations also breaks down for many species in the tropical upper
stratosphere (where lifetimes become short) and for ozone in the middle
and upper stratosphere.
Plumb and Ko (1992) have shown that, for two chemical compounds
whose local photochemical lifetimes are longer than the quasi-horizontal
transport time constant, the slope of the correlation curve is provided
by the ratio between the net fluxes of the two species through their
respective mixing ratio isopleths (see also Murphy and Fahey, 1994).
In addition, they calculated that for two long-lived source gases (e.g.,
N2 O, CH4 , etc.) that are emitted at the surface and are photochemically
destroyed in the stratosphere, the slope dX2 /dX1 of the correlation curve
is related to the ratio of the atmospheric photochemical lifetimes τ1 /ττ2
of the 2 species. For steady-state conditions:
τ1 dX2 X1 (0)
· (3.55)
τ2 dX1 X2 (0)
where X1 (0) and X2 (0) represent the mixing ratio of compounds 1 and
2 in the very low stratosphere (or in the troposphere). Measurements of
correlations between chemical species provide a powerful analysis tool
to establish the origin and fate of air masses in the stratosphere and in
the troposphere.
3.6 Dynamics and Meridional Transport in Two

Di mensions: A Conceptual View
3.6.1 Zonal Means and Eddies
The description of dynamical processes is often facilitated by

distinguishing between zonal mean motions and fluctuations about the
zonal mean (called the eddies). As indicated in Section 3.4, these
fluctuations can be of varied spatial scale — from a few meters to
thousands of kilometers. In the stratosphere, observed wave structure
indicates that most of the waves are of large scales, of zonal wavenumber
3 or less, as discussed earlier. Note that the use of the word “eddy” in the
context of atmospheric dynamics is thus rather different than the idea of
a whirlwind, which is probably a more common usage of the same word
(Oxford English Dictionary). Any atmospheric quantity ψ(λ, ϕ, z) can
be expressed as the sum of the zonal mean:
2π
1
ψ (ϕ, z; t) = ψ (λ, ϕ, z; t) dλ (3.56)
2π
0
and the eddy term, ψ, as follows:
ψ (λ, ϕ, z; t) = ψ (ϕ, z; t) + ψ (λ, ϕ, z; t) (3.57)
where λ, ϕ, z, and t represent longitude, latitude, height, and time,
respectively. Using (3.56) and (3.57), one can derive from the primitive
equations (3.23a), (3.25) and (3.24b), the zonal mean momentum
equation
∂u v ∂ (u cos φ) ∂u
+ +w − fv
∂t a cos φ ∂φ ∂z
1 ∂ 1 ∂ w

= Fλ − u v cos 2
φ − ρ 0 u (3.58)
a cos2 φ ∂φ ρ0 ∂z
the continuity equation
1 ∂ (v cos φ) 1 ∂
+ (ρ0 w) = 0 (3.59)
a cos φ ∂φ ρ0 ∂z
and the thermodynamic equation
∂Θ v ∂Θ ∂Θ
+ +w
∂t a ∂φ ∂z

1 ∂ 1 ∂
=q− v Θ cos φ + ρ0 w Θ (3.60)
where
κ
p0 Q
q=
p cp
is proportional to the zonally averaged net heating rate Q.

The zonally averaged continuity equation for chemical species i
expressed in mixing ratio Xi becomes
∂Xi v ∂Xi ∂Xi

+ +w
∂t a ∂φ ∂z

1 ∂ 1 ∂
= si − v Xi cos φ + ρ0 w Xi (3.61)
where
Si
si = (3.62)
n
is the chemical source term (m−3 s−1 ) divided by the total air density
n (m−3 ). The eddy terms in the right-hand side of the momentum,
thermodynamic, and chemical continuity equations are known as the
eddy flux divergences. Note that equations (3.58) to (3.60) do not
constitute a closed system; they contain products of eddy quantities
(such as v Θ ) which are not calculated within this system. These eddy
terms can be an important part of the heat and momentum budgets.
For example, Figure 3.19 shows the relative magnitudes of the most
important terms in the thermodynamic equation as a function of latitude
at 10 hPa. Note that the terms w∂Θ/∂z and 1/a cos φ ∂/∂φ(v Θ ) cos φ
are larger than the diabatic heating rate and are nearly in balance with
one another.
Figure 3.19. Magnitude of the terms in the thermodynamic equation as

a function of latitude at 10-30 hPa during winter. Note that the largest
contributions to the heat budget are provided by adiabatic compression and
wave driven meridional heat transport. From O’Neill (1980).
3.6.2 Description of the Mean Meridional Stratospheric

Circulation
Some of the earliest proposed descriptions of stratospheric transport

were based on observations of chemical species. The first observations
of stratospheric water vapor densities showed that the stratosphere
was extremely dry, exhibiting mixing ratios of the order of a few
parts per million by volume, in marked contrast to the troposphere,
where water vapor abundance reaches a few percent. Brewer (1949)
suggested that the dryness of the stratosphere was determined primarily
by condensation and that the water vapor content of an air parcel
rising from the troposphere to the stratosphere would therefore be
determined by the lowest temperature experienced by the parcel, which
would normally correspond to the tropopause. He also noted that
the tropopause temperatures in the tropics were low enough to yield
stratospheric water vapor densities as low as those observed, while the
middle and high latitude tropopause was much too warm to explain
the observed dryness. Thus he suggested a circulation exhibiting rising
motion only in the tropics, and descending motion at extra-tropical
latitudes. Dobson (1956) suggested a qualitatively similar transport
pattern based on observations of stratospheric ozone.
Murgatroyd and Singleton (1961) presented the first calculated
mean circulation for the middle atmosphere by using an analysis of
radiative heating rates. They derived a circulation much like the
one suggested by the tracer studies discussed above. However, they
noted that discrepancies existed in the angular momentum budget of
the derived circulation, which were presumably due to eddy transport
processes. The meridional circulation derived from that study is shown
in Figure 3.20.
A few years later, observational studies using the angular momentum
and/or heat budgets led to mean circulations of the lower stratosphere
based on meteorological data, (specifically, observations of the dominant
eddy terms in Eqs. 3.58 and 3.60; e.g., Reed et al., 1963; Julian and
Labitzke, 1965; Vincent, 1968). These studies produced circulations
which were quite different from the one derived by Murgatroyd and
Singleton or those deduced from tracer studies (see e.g., Figure 3.21).
In particular, the existence of rising motions (the “Ferrell cell”) was
suggested as occurring during winter at high latitudes (60◦ ) in the
troposphere and lower stratosphere.
Figure 3.20. Meridional circulation based on the study of Murgatroyd and

Singleton (1961). From Murgatroyd (1971).
Mass streamfunction representing the Eulerian mean circulation

in January calculated by Becker and Schmitz (2002). Values are expressed in
109 kg/s.
Which depiction of the mean meridional circulation is correct,

Figure 3.20 or Figure 3.21? The answer is that they both are, but that
they represent different quantities. To a first approximation, Figure 3.20
represents the zonally averaged motion of air parcels (net transport in
the meridional plane) while Figure 3.21 represents the transport by the
zonally averaged circulation. As we will show below, the net motion
in the case described by Murgatroyd and Singleton is the sum of the
zonal mean circulation shown in Figure 3.21 and the transport by eddy
motions. The differences between the two circulations are a result of the
zonal averaging procedure applied in the presence of wave disturbances
(the “eddies”). It can be shown that, for waves that are steady (no
change in amplitude with time) and conservative (no dissipation or wave
breaking), the mean meridional eddy heat transport and the adiabatic
heating/cooling associated with the mean vertical motion cancel exactly
(Boyd, 1976; Andrews and McIntyre, 1976; see also Figure 3.19). This
property of Eulerian zonal averages is known as the non-acceleration
theorem and is of great importance in understanding the evolution of
zonally averaged fields of wind and temperature in the presence of wave
motions.
A physical understanding of how the cancellation between the eddy

heat transport and the wave induced mean meridional circulation arises
can be gained from Figure 3.22. This figure depicts in schematic
form the structure of a vertically propagating planetary wave which is
assumed to be steady and conservative. Streamlines of the flow in the
latitude-longitude plane are shown as solid lines that meander north and
south. Horizontal velocity vectors (heavy arrows) and vertical velocity
vectors (dashed lines) are shown at selected points along the streamlines.
Temperature perturbations T associated with the wave are denoted by
the letters W (warm) and C (cold). The projection onto the height-
latitude plane of the parcel trajectories following a given streamline are
shown as ellipses, with the arrows indicating the sense of the parcel
motion (it may be useful to imagine these projections as those of a three-
dimensional elliptical corkscrew). Note, first of all, that for the wave
depicted here the horizontal velocity and the temperature perturbations
are in phase (i.e., positive velocities coincide with positive temperature
perturbations, and vice versa). Thus, the average heat flux, v T taken
along any of the latitude circles φ1,3,5 is positive. It is largest at φ3
because the wave has maximum amplitude in mid-latitudes. As a result,
there is a divergence of zonally averaged eddy heat flux, v T at φ2
and a convergence at φ4 . These are balanced by compression heating
(sinking motion) and expansion cooling (rising motion), respectively,
and so one obtains negative w at φ2 and positive w at φ4 (dashed
Figure 3.22. Schematic diagram of the idealized planetary wave structure and
the resulting eddy transport of heat. Adapted from Matsuno (1980).
arrows). This is the “Ferrell cell” pattern characteristic of the Eulerian

meridional circulation in winter. However, since the projections of the
parcel trajectories are closed ellipses, there is in fact no net parcel motion
in the latitude-height plane as a result of these processes, and the Ferrell
cell arises purely as a result of the zonal averaging procedure at the
fixed points φ1,3,5 . This is essentially the state of affairs described by
the non-interaction theorem. A detailed explanation of the relationship
between Eulerian eddy heat fluxes, meridional circulations, and parcel
trajectories is given by Matsuno (1980).
The foregoing arguments demonstrate qualitatively that as long
as wave motion is steady and conservative, a net meridional mass
circulation takes place only when the zonally averaged diabatic heating
(q in Eq. 3.60) is different from zero. The large-scale planetary waves,
which dominate the wave structure of the stratosphere probably conform
to this description at least in part. Therefore no net transport is
produced in the meridional plane through the effects of these eddies.
Thus the apparent discrepancy between the mean circulations derived by
Brewer, Dobson, and Murgatroyd and Singleton, versus those computed
by the later studies is reconciled when it is understood that the former
studies represent net transport, while the latter correspond to a mean
circulation only. The meridional transport produced by this latter
circulation is cancelled to a large degree by eddy transports, when these
terms are evaluated in the standard fashion using a reference frame which
is fixed in latitude and height.
The eddy-mean flow cancellation represents a difficulty for two-
dimensional descriptions of stratospheric transport. Perhaps most
importantly, it suggests that in the Eulerian framework, the mean and
eddy transports are intimately coupled, and that a proper representation
of atmospheric transport must employ a consistent set (see, e.g.,
Harwood and Pyle, 1975; Rood and Schoeberl, 1983). Further, it
suggests that the use of eddy diffusion coefficients to characterize
all types and scales of eddy processes must be questioned, since the
stratospheric eddies appear to be at least partly steady and conservative
(i.e., non-diffusive in character, see e.g., Matsuno, 1980).
3.6.3 Lagrangian Mean, Diabatic, and Transformed

Eule rian Mean Circulations
The difficulties associated with the eddy-mean flow cancellation have

motivated the search for a mathematical formalism to describe the
zonally averaged structure of the atmosphere in such a way as to
provide a more meaningful separation between the eddies and the mean
meridional circulation. Andrews and McIntyre (1978) introduced the
so-called Lagrangian mean averaging, wherein horizontal averages are
computed not along latitude circles, but along a path that follows the
wave trajectory in latitude and longitude. Referring again to Figure 3.22,
it should be clear that if v is evaluated along any of the wave streamlines
shown, then no eddy fluxes appear. Although the Lagrangian mean
formalism is mathematically elegant and physically meaningful, it suffers
from technical problems that limit its applicability to situations where
wave amplitudes are very large (McIntyre, 1980).
A more practical approach is simply to derive the net mean circulation
directly from the diabatic heating rate, q, as suggested by Dunkerton
(1978). From the thermodynamic equation (3.60) in which it is assumed
that ∂Θ/∂t and (v/a)(∂Θ/∂φ) are small (see Figure 3.19), and that the
vertical eddy heat fluxes are negligibly small, one can deduce that
∂Θ 1 ∂
w q− v Θ cos φ (3.63)
∂z a cos φ ∂φ
This expression reveals that for small diabatic effects (q 0), the eddy
heat flux divergence is in close balance with the vertical advection of the
mean potential temperature (non-acceleration theorem).
A transformation can be introduced to represent the net mean
meridional and vertical velocities v ∗ and
w∗
1 ∂ ρ0 v Θ
v∗ = v − (3.64a)
ρ0 ∂z ∂Θ/∂z

∗ 1 ∂ v Θ cos φ
w =w+ (3.64b)
a cos φ ∂φ ∂Θ/∂z
It defines the transformed Eulerian mean (TEM) or residual circulation.
When these definitions are applied to the simplified form (3.63) of
the thermodynamic equation, the resulting circulation is called the
diabatic circulation and the meridional and vertical velocities will be
noted (vD , wD ) rather than (v ∗ , w ∗ ). In this case, the simplified
thermodynamic equation and the steady state approximation for the
continuity equation become
∂Θ
wD =q (3.65)
∂z
1 ∂ 1 ∂
(vD cos φ) + (ρ0 wD ) = 0 (3.66)
Note that the eddy heat flux does not appear in the transformed
thermodynamic equation (3.65). From these equations, vD and wD may
be derived directly from a knowledge of the diabatic heating rate. We
emphasize that this is, in fact, just what Murgatroyd and Singleton
(1961) did when they ignored the eddy heat flux. See also Dunkerton
(1978).
It is important to note that the diabatic circulation can be
estimated from the thermodynamic and continuity equations only if
the temperature distribution is known a priori. However, a fully
self-consistent mean meridional circulation can be obtained when the
momentum budget is considered together with the thermodynamic
and continuity equations. When transformations (3.64a, b) are
applied to the zonal mean equations (3.53-3.60), the solution of the
following momentum, continuity, and thermodynamic equations defines
the transformed Eulerian mean (TEM) or residual circulation (here
expressed in log-pressure coordinates):
∂u v ∗ ∂u cos φ ∂u ∇·E
+ + w∗ − fv ∗ = + Fλ ≡ Gu (3.67)
∂t a cos φ ∂φ ∂z ρ0 a cos φ
1 ∂v ∗ cos φ 1 ∂
+ (ρ0 w ∗ ) = 0 (3.68)
∂Θ v∗ ∂Θ ∂Θ
+ + w∗ = q + GΘ (3.69)
∂t a ∂φ ∂z
where

1 ∂ ∂Θ/a∂φ
GΘ = − ρ0 v Θ + w Θ (3.69a)
ρ0 ∂z ∂Θ/∂z
represents the contribution of the eddies to the temperature tendency.
This term is generally small and often ignored (although the contribution
involving w Θ can be important for gravity waves). The forcing term
Gu appearing in the zonal momentum equation includes the divergence
·E
of the Eliassen flux ∇ as well as any additional dissipative force (Fλ ).
The components of the Eliassen-Palm (EP) flux are

∂u v Θ
Eφ = −ρ0 a cos φ u v − (3.70a)
∂z ∂Θ/∂z

1 ∂u cos φ v Θ
Ez = ρ0 a cos φ f− − u w (3.70b)
a cos φ ∂φ ∂Θ/∂z
For planetary waves in the stratosphere, quasi-geostrophic scaling can
be applied, in which case expressions (3.70a) and (3.70b) reduce to
Eφ = −ρ0 a cos φ u v (3.71a)
and
v Θ
Ez = ρ0 a cos φ f (3.71b)
∂Θ/∂z
Thus, the meridional component of the EP flux is proportional to the
momentum flux u v while the vertical component is proportional to the
meridional heat flux v Θ .
Note that, aside from the contribution of GΘ in the thermodynamic
equation, which vanishes exactly for steady conservative waves (Andrews
and McIntyre, 1978), wave forcing in the TEM system is limited to the
Eliassen-Palm divergence in the zonal momentum equation (assuming
Fλ 0). The important role played by the momentum budget will
be discussed in more detail in Section 3.7. Figure 3.23 shows the
transformed Eulerian mean streamlines from the model of Becker and
Schmitz (2002). As the transformed Eulerian mean circulation is
very similar to the diabatic circulation, upward transport by TEM
motions occurs primarily where the net diabatic heating (Q) is positive,
while downward transport is found in regions where Q is negative. A
Figure 3.23. Mass streamfunction for the transformed Eulerian mean

circulation (109 kg/s) calculated by Becker and Schmitz (2002) for the month of
January. Compare this residual circulation with the Eulerian mean circulation
shown in Figure 3.21.
generalization of the TEM circulation which describes the mass flux

across an arbitrary iso-surface (provided that this quantity defining this
iso-surface varies monotonically with height) is given by Juckes (2001).
As in the case of the potential temperature, the Eulerian mean and
eddy components of meridional transport for the mixing ratio of chemical
species tend to cancel, and the continuity equation for species i, written
as a function of the TEM winds becomes
∂Xi v∗ ∂Xi ∂Xi

+ + w∗ = si + GX (3.72)
∂t a ∂φ ∂z
where GX expresses the contribution of the eddies through the

correlations between the wave components of the mixing ratio and wind
velocities. It is written as
1 ∂ ∗ 1 ∂ ∗

GX = v X cos φ + ρ0 w X
where the net eddy flux components are expressed by (Garcia and
Solomon, 1983)
∗ v Θ ∂X
v X = v X −
∂Θ/∂z ∂z
∗ v Θ ∂X
w X = w X +
∂Θ/∂z ∂y
For inert tracers, and for conservative and steady waves, the GX
term vanishes identically, and the zonally averaged distribution X is
determined solely by the residual circulation (v∗ , w∗ ). If the chemical
lifetime of the chemical compound is comparable to the physical scale of
the wave, and if that lifetime is comparable to the time it takes for an
air parcel to move through the wave, then the effects of production
and loss along wave trajectories cannot be ignored. Under these
special circumstances, the wave can move air parcels into different
photochemical environments (regions of different chemical lifetimes),
introducing chemical eddy transport. This effect can be parameterized
in terms of wave structure and the photochemical lifetime of the
species considered (see Plumb, 1979). The importance of this effect
has been shown, for example, by Garcia and Hartmann (1980), Pyle
and Rogers (1980), Matsuno (1980), Hartmann (1981), Strobel (1981),
and Kawahira (1982). In practical applications, the net eddy flux
components are often written as

∗
∂X ∂X
v X = − Kyy + Kyz
∂y ∂y

∗ ∂X ∂X
w X = − Kzy + Kzz
∂y ∂z
where the eddy diffusion coefficients Kij can be expressed as a function

of the chemical lifetime of the trace constituents and of the time needed
for an air parcel to move through a disturbance. One of the species most
strongly influenced by chemical eddy transports is stratospheric ozone
in high latitude winter. Figure 3.24 shows an estimate of the influence
of chemical eddy transport on ozone as calculated with the model of
Garcia and Solomon (1983).
The conceptual advantage of the diabatic or transformed Eulerian
mean circulations is that the eddy-mean flow cancellation is avoided,
and a first approximation to the net meridional and vertical transport
needed for atmospheric studies is therefore readily obtained. The most
important features of this transport description are that air generally
enters the lower stratosphere in the tropics and exits at middle and
high latitudes in both hemispheres. In the upper stratosphere and
mesosphere, air flows from the summer hemisphere to the winter
hemisphere, and downward near the winter pole. This transport
description is qualitatively very similar to that proposed by Brewer based
Figure 3.24. Calculated ozone distribution (in ppmv) for Northern Hemisphere
winter solstice with and without chemical eddy transport. From Garcia and
Solomon (1983).
Figure 3.25. Schematic representation of the wave-driven transport circulation

(solid lines with arrows) in the winter stratosphere. The shaded area represents
the surf zone where meridional mixing associated with wavebreaking is taking
place. Light dashed lines represent isentropes. The heavy dashed line shows a
constant mixing ratio surface of a long-lived tracer. The dotted line represents
the tropopause. Adapted from Holton and Alexander (2000).
on early observations of water vapor. Indeed, this very simple picture

provides a reasonable qualitative description of transport of chemical
species, at least in the monthly or seasonal mean, and can successfully
explain many of the observed features of the distributions of chemical
constituents.
The formalism of the transformed Eulerian mean circulation shows
that meridional transport in the middle atmosphere is generated
primarily by non-local momentum forcing associated with wave
dissipation. This forcing, represented by the Eliassen-Palm flux
divergence in equation (3.67), acts as an extratropical pump producing
strong upward air motions in the tropics and downward “return”
flows at high latitudes (Figure 3.25). This circulation is most intense

in winter when planetary waves are penetrating in the stratosphere
(see Section 3.4). In the mesosphere, the meridional flow generated
primarily by gravity wave breaking is directed from the summer to the
winter pole. By continuity, the flow is directed upward in the summer
hemisphere and downward in the winter hemisphere (see Figure 3.33 in
Section 3.7). Such cross-isentropic motions are accompanied by diabatic
heating/cooling and tend to steepen the slope of atmospheric tracers.
At the same time, mixing is produced as a result of planetary wave
dissipation, primarily along quasi-horizontal isentropic surfaces. This
process tends to reduce the slope of quasi-inert atmospheric tracers.
The balance between these two competing processes affects the compact
relationship between the mixing ratio of two long-lived species noted in
some of the scatterplots in Figure 3.18. The slope of the mixing ratio
isopleths for different long-lived constituents is determined essentially
by transport processes, and is therefore nearly the same for all these
compounds.
The relative importance of slope-steepening diabatic motions and
slope-flattening quasi-horizontal mixing can be deduced from the
examination of Figure 3.26 (see also Plate 2). Figure 3.26 represents
the zonally averaged distribution of hydrogen fluoride (HF) as measured
from space. The source region of this long-lived compound is located in
Figure 3.26. Zonally averaged distribution of hydrogen fluoride (HF) mixing

ratio (in ppbv) measured by the HALOE instrument on UARS. The heavy solid
arrows denote the mean meridional transport and the horizontal arrows show
the location of strong quasi-horizontal mixing by planetary waves. Courtesy of
W. Randel, NCAR, 2001.
the stratosphere, and its mixing ratio below 50 km therefore increases

with height. The minimum mixing ratio in the tropics is associated
with strong upward motions while the high concentrations at high
latitudes result from downwelling, primarily inside the wintertime
polar vortex. Quasi-horizontal mixing at mid-latitude in the winter
hemisphere explains the nearly flat mixing ratio isopleths observed
between 30◦ S and 50◦ S.
3.6.4 Dynamical Barriers against Meridional Transport
Observations of numerous tracers reveal sharp horizontal concentra-

tion gradients near 30◦ latitude in both hemispheres, suggesting that
the tropical stratosphere is fairly isolated from extratropical regions
(McIntyre, 1990). Plumb and Ko (1992) noted that the compactness of
the tracer relationships depicted in the scatterplots (Figure 3.18) breaks
down when considering mixing ratios observed in the tropical upper
stratosphere. The classic picture of stratospheric transport has been,
therefore, significantly refined in recent years, and, for example, Plumb
(1996) proposed replacing the notion of a global diffuser model by a
“tropical pipe model” represented by the lower panel of Figure 3.27. In
this model, the abundance of chemical compounds within the “tropical
pipe” is determined primarily by a balance between upwelling and local
chemical destruction or production.
Several observations confirm the quasi-isolation of the tropical
stratosphere (in addition to the sharp meridional gradients affecting
long-lived tracers near 20-30◦ latitude). These include the formation
of a tropical reservoir of radioactive material after nuclear explosions
in the atmosphere (Feely and Spar, 1960) and of aerosols after volcanic
eruptions (Trepte and Hitchman, 1994; Grant et al., 1996). Another
indication is provided by the annual variation in tropical water vapor in
the lower stratosphere, specifically the tape recorder effect (Mote et al.,
1995; see Figure 5.24b). Because air passing through the equatorial
tropopause tends to be “freeze dried” to a water saturation mixing
ratio corresponding to the tropopause temperature (“cold trap”), and
since this temperature is affected by a strong annual cycle, the water
influx (and hence the H2 O mixing ratio near the equator) is minimum
during the Northern Hemisphere early spring (February-March) and
maximum six months later. The fact that this water vapor signal is
pulled upward by rising motions and remains confined within the tropical
pipe provides evidence for the existence of strong dynamical barriers
against meridional mixing at the edge of the tropical region.
Figure 3.27. Schematic representation of the global diffuser model ((upper panel))
and tropical pipe model (lower panel). Gray arrows denote meridional transport
by the transformed Eulerian mean circulation while the heavy solid arrows show
quasi-horizontal mixing by large scale waves. The vertical lines in the lower
panel represent dynamical barriers against meridional transport in the tropics.
From Plumb and Ko (1992).
Randel et al. (1993) have shown that planetary-scale deformations of

the barrier, shaped as tongues of tropical air can extend out deep into
the mid-latitudes (Figure 3.28 and Plate 3). Occasionally, such tongues
become unstable and break, implying subsequent irreversible mixing in
the extratropics. This process provides a potential mechanism for quasi-
isentropic transport of chemical compounds across the tropical barrier
(Boering et al., 1996). The timescale for detrainment from the tropics
is estimated to be less than 6 months in the lower stratosphere (Volk
et al., 1996). Similarly, studies that have attempted to assess the degree
of isolation of the tropics suggest that the timescale for mixing into the
tropics (from mid-latitudes) is close to 1 year.
The polar vortex (near 60◦ latitude, above approximately 16 km) is
another dynamical barrier against meridional transport. The isolation
of the polar regions, and specifically of the Antarctic lower stratosphere
Figure 3.28. Sequence showing the evolution of a tongue of nitrous oxide

extending from the tropics into mid-latitude regions and subsequent irreversible
mixing. Nitrous oxide mixing ratio ((ppbv)) is shown on the 1100 K isentropic
surface (near 38 km altitude or a pressure surface of 5 hPa) for 6-10 September,
1992, based on the measurements of UARS/CLAES. From Randel et al. (1993).
during winter, is an important factor in explaining the formation of the

ozone hole in spring. Limited mixing of air across the vortex occurs
primarily when large-scale Rossby waves irreversibly pull filaments of
polar air from the edge of the winter vortex (See Plate 4). Air
is stripped from the vortex edge, eroding the vortex and steepening
the potential vorticity gradient at the edge of the vortex (Juckes
and McIntyre, 1987). Modeling studies (Waugh and Plumb, 1994;
Waugh et al., 1994) suggest that air is ejected from the vortex as long
and narrow filaments (see Figure 3.29). These fine-scale structures,
also observed in the lower stratosphere from high altitude aircraft
(Newman et al., 1996), are stretched around the polar vortex and
progressively mixed (after typically 20-25 days) with surrounding mid-
latitude air. Below approximately 16 km (potential temperature of
400 K), the vortex edge barrier becomes weaker, and mixing reaches high
latitudes. As Rossby waves are generally less intense in the Southern
Hemisphere, the Antarctic polar vortex is more stable than the Arctic
vortex, and Antarctic air is considerably more isolated from mid-latitude
influences than Arctic air. The large planetary wave disturbance and the
associated weak ozone hole observed in Antarctica during September
2002 correspond to an exceptional situation. Meridional transport out
of the polar vortex regions is thought to affect significantly the mid-
latitude ozone budget in the lower stratosphere.
Figure 3.30 presents a schematic view of the meridional circulation
in the lower stratosphere, including the roles of the tropical and polar
dynamical barriers. This figure also shows that extratropical cross-
tropopause exchanges (Box 3.1) involve wave-driven cross-isentropic
diabatic descent from the “overworld” (the region where isentropic
surfaces lie entirely in the stratosphere) to the “lowermost stratosphere”
(the region where isentropes cross the tropopause). Once air and
chemical constituents have reached the lowermost stratosphere following
diabatic descent, they can be adiabatically transported along isentropes
(followed by irreversible mixing) into the troposphere (“the underworld”;
see Haynes et al., 1991; Holton et al., 1995). Thus, on the average,
the rate at which a stratospheric inert tracer is transported through
the tropopause is equal to the rate at which it is transported by
the “extratropical pump” across the upper boundary of the lowermost
stratosphere. The tropopause exchanges are often linked to strong
disturbances (e.g., “cutoff lows”) leading to the formation of tongues
of stratospheric air (with high PV and high ozone concentrations) which
can penetrate deep into the troposphere.
The details of such mass exchanges involve dynamical processes at
medium and small scales. For example, in the vicinity of extratropical
cyclones (spatial scale of typically 1000 km), cold stratospheric air is
Figure 3.29. Fine-scale structure of a tracer in the vicinity of the polar

vortex from 6 January to 16 January 1992 on the 850 K isentropic surface.
The calculation made with NMC-analyzed winds show that, as a result of
Rossby wave breaking, air is ejected from the polar vortex as long filamentary
structures. This process transfers vortex air into mid-latitudes. From Waugh
et al.(1994).
transported equatorwards at the west side of the depression (within

tropopause folds) and is gradually mixed with tropospheric air as a
Figure 3.30. Schematic representation of the atmospheric circulation (arrows)

and associated quasi-horizontal mixing between the surface and the middle
stratosphere. Mixing processes leading to stratosphere-troposphere exchanges
are also represented. The heavy vertical lines denote dynamical barriers against
meridional transport. Note the large-scale ascent in the tropical stratosphere
above intense convective systems in the tropical troposphere, and large scale
descent associated with the polar vortex during winter (WMO, 1999).
result of mechanical and radiative dissipation (turbulence and irregular

distribution of ozone and water vapor). Part of the descending
air may return to the stratosphere, so that the net stratosphere to
tropospheremass flux is not straightforward to quantify. Convective
activity in the vicinity of such storms, which tend to mix tracers in the
vicinity of the tropopause, makes this estimate even more difficult. At
smaller spatial scales (typically 100 km), turbulence and mixing caused
by gravity wave breaking events or strong wind shear (e.g., clear air
turbulence) also contribute to cross-tropopause exchanges.
Transport from the troposphere to the stratosphere occurs primarily
in the tropics and is associated with the upward branch of the
Brewer-Dobson circulation. Folkins et al. (1999) argue that the
top of the tropospheric Hadley circulation in the tropics occurs at
14 km, i.e., well below the tropopause, and that a barrier to vertical
mixing therefore exists in the tropical tropopause layer (TTL) 2 or
3 km below the thermal tropopause. Air injected above this barrier
subsequently participates in generally slow vertical ascent into the
stratosphere. Small-scale exchanges also take place at mid-latitudes
through filamentary structures that are drawn poleward in relation with
anticyclonic circulation in the upper troposphere (Chen, 1995).
Box 3.1 The Tropopause

The surface separating the troposphere from the stratosphere is
called the tropopause. It is associated with the minimum in the
vertical temperature profile observed near 7-16 km altitude. The World
Meteorological Organization (WMO) defines the tropopause level as the
lower boundary of a layer in which the vertical decrease in temperature is
less than 2 K/km for a depth of at least 2 km.
The stratosphere is characterized by a rapid increase with height of
the potential temperature in response to the strong heating produced
by ozone absorption of solar ultraviolet radiation. As a result, the
stratosphere is stably stratified with very slow vertical exchanges. At the
same time, the troposphere, especially in the tropics, is easily subject to
destabilization caused primarily by the solar heating of the Earth’s surface,
producing powerful deep convection. The transition between these two
regions is complex and not fully understood. The temperature minimum
at the tropical tropopause (located near 16 km altitude) is believed to
be produced by the wave-driven upward motion in this region. In the
extratropics, the tropopause is located at considerably lower altitudes
(12 km at mid- latitudes and 8 km in the polar regions). Its height is
determined primarily by the dynamical heating associated with baroclinic
waves and cyclonic activity.
Rather than being an infinitely thin surface, the tropopause should
be regarded as a highly dynamic layer over which the mixing ratios of
chemical compounds (such as water vapor or ozone) change rapidly from
their tropospheric to their stratospheric values. This layer behaves as a
quasi-material entity: in the tropics, it nearly coincides with the isentropic
surface of 380 K. In the extratropics, it corresponds to the iso-surface of 2-
4 potential vorticity units, and includes a mixture of air with tropospheric
and stratospheric characteristics (e.g., tracer mixing ratios).
3.7 The Importance of Wave Transience and

Dissipa tion
Equations (3.67) to (3.69) show that, in a hypothetical case where
the effects of wave transience and dissipation are ignored (Gu and GΘ =
0) in both the thermodynamic and momentum equations and steady
state is assumed, the atmosphere would eventually reach a radiative
equilibrium, corresponding to a temperature distribution similar to that
displayed in Figure 3.31. In this case, the residual circulation (v∗ , w∗ )
would vanish and the temperatures would be quite different from those
observed, particularly at the tropical tropopause and in the mesosphere.
At the same time, zonal winds would be considerably faster than are
observed with a particularly strong polar vortex during winter.
Figure 3.31. Meridional distribution of the radiative equilibrium temperature

calculated by Fels et al. (1985).
What is lacking in this picture is the mechanism for deceleration

of the zonal winds provided by the Eliassen-Palm divergence or any
other dissipation force (Leovy, 1964; Schoeberl and Strobel, 1978;
Holton and Wehrbein, 1980). Assuming that wave transience and
dissipation can be neglected in the thermodynamic (but not the zonal
momentum) equation, and considering the dominant terms in the
momentum equation (3.67), we can write
−fv ∗ = Gu (3.73)
if Gu accounts for momentum forcing produced by wave breaking
and other dissipative effects. The TEM circulation as well as the
related meridional transport in the middle atmosphere is therefore driven
primarily by wave dissipation. Our ability to derive the strength of the
mean meridional circulation in a rotating frame of reference depends
consequently on the accuracy to which transient and dissipative eddies
(and specifically the Eliassen-Palm flux divergence ∇ E)
are determined.
From equation (3.73), from the continuity equation (3.45), and for an
upper boundary condition ρ0 w → 0 as z → ∞, one easily derives
∞
∗ −1 ∂ 1
w (φ, z) = ρ0 z Gu φ, z cos φ dz (3.74)
ρ0 (z) a cos φ ∂φ f z
This expression describes the so-called downward control principle
(Haynes et al., 1991). It shows that in the extratropics (f = 0),
the vertical velocity w ∗ (and hence the vertical transport) at a given
altitude is determined only by wave dissipation above this height. Thus,

as suggested by the concept of the extratropical fluid pump described
earlier, the TEM circulation at a given atmospheric level is driven by the
zonal momentum force associated with wave dissipation above this level.
Specifically, the strength of the mass exchanges between the stratosphere
and the troposphere can be explained by nonlocal wave forcing in the
middle atmosphere rather than local dynamical or thermodynamical
processes.
An accurate representation of wave drag is essential for chemical
modeling because chemical species are advected by the mean meridional
circulation and are also transported by the irreversible mixing which
accompanies wave drag. Modeling studies (e.g., Schoeberl and Strobel,
1978; Holton and Wehrbein, 1980) sometimes use a “Rayleigh”
coefficient to crudely parameterize the effects of wave dissipation. In this
case, a deceleration is assumed to act linearly on u, and is substituted
for the wave drag term Gu as follows:
∂u
− fv ∗ = Gu ≡ −kr u (3.75)
∂t
where kr is the Rayleigh friction coefficient. Although the formulation
in equation (3.75) must be regarded as no more than a crude
parameterization, the dynamic and thermal structure obtained by
models that include Rayleigh friction is in broad agreement with
observations as long as k−1
r is of the order of 100 days in the stratosphere
and a few days in the mesosphere. The higher values of kr required
for the mesosphere as compared to the stratosphere highlights the
importance of gravity wave dissipation above the stratopause.
In the thermosphere, where ions (charged particles) represent a
substantial fraction of the atmospheric fluid, dissipative forces involve
the effects of the electromagnetic field and of momentum transfer
between ions and neutral particles. As a result, the dissipative forces
cannot be expressed anymore by a simple Rayleigh friction coefficient,
but by a tensor λij (Dickinson et al., 1975; 1977; Hong and Lindzen,
1976; Kasting and Roble, 1981) characterizing an anisotropic drag:
Gu = − (λxx u + λxy v) (3.76a)

Gv = − (λyx u + λyy v) (3.76b)
Two effects must be considered: At altitudes sufficiently high for the

collisions between ions and neutral particles to become infrequent, and
hence the ions and the neutrals to behave as two independent fluids
(ions move preferentially along magnetic field lines), neutral particles
that cross field lines suffer momentum loss: this dissipation is called ion
drag. Hong and Lindzen (1976) have shown that the ion drag coefficients
can be expressed as
λxx = αi (z)
λxy = αi (z) sin2 I
where factor αi (z) can be expressed as a function of the ion number

density Ni (z) by
5 × 10−10 Ni (z)
αi (z) = 2 s−1
1 + (νi /ωi )
where νi /ωi is the ratio between the ion/neutral collision frequency νi

and the ion gyrofrequency ωi .
The second effect to be considered is the deflection of the weakly
charged air flow by the electromagnetic field through the Lorentz force.
The corresponding drag coefficients are (Hong and Lindzen, 1976)

νi
λxy = λyx = αi (z) sin I .
ωi
In the above expressions, I is the dip angle of the magnetic field. For
magnetic poles coincident with the geographic poles, tan I = –2 tan φ,
where φ is the geographic latitude. λij coefficients become significant
only above 100 km altitude. The time constant (1/λxx ) associated with
ion drag is typically 1 day at 120 km and about 1 hour at 250 km. The
Lorentz effect (λxy ) plays a role only in the lower thermosphere (110 to
150 km).
At lower altitudes, except at the tropical tropopause and in
the winter hemisphere, the stratosphere may be considered to be
approximately in radiative equilibrium, while the mesosphere is very
far from these conditions. The most dramatic manifestation of the
importance of meridional motions for the temperature structure of the
middle atmosphere occurs at the summer polar mesopause, where the
temperature (about 130 K) is the lowest observed anywhere on Earth,
even though the region receives substantially more solar radiation than
does the winter mesopause (Box 3.2).
Box 3.2 The Mesopause

The transition between the mesosphere and the thermosphere,
called the mesopause, is characterized by a temperature minimum whose
location varies significantly with latitude and season (see Figure 3.32). In
the tropics during all seasons, and in the winter mid- and high-latitudes,
this transition takes place at approximately 100 km altitude. At the winter
pole, the mesopause temperature is typically 180–200 K. At mid- and high
latitudes during summer, the mesopause is located near 85 km, and the
temperature at the pole reaches values as low as 120-140 K (coldest place
in the Earth’s atmosphere where polar mesospheric clouds are occasionally
observed).
Figure 3.32. Zonally averaged temperature in the vicinity of the

mesopause (65 to 105 km) observed in July by the HRDI instrument on
board the Upper Atmosphere Research Satellite (UARS). From Ortland
et al., 1997.
The structure of the mesopause (see Smith, 2004) depends on a complex

interplay between dynamics, radiative transfer, and photochemistry.
Although, on the average, radiative processes dominate the energy
budget in the region surrounding the mesosphere, adiabatic heating and
cooling mechanisms associated with large scale advection by the general
circulation, transport of heat by molecular diffusion and by turbulence,
energy deposition by gravity waves and occasionally heat transport
resulting from convective instability must be taken into account. Heat
release by exothermic reactions (chemical heating) as well as energy loss
to space by radiative emissions, primarily by carbon dioxide, also play key
roles in the energetics of the region.
Box 3.2 (Continued)

Interestingly, the mesopause can be regarded as a layer separating two
distinct dynamical regions of the atmosphere: the thermosphere, where
atmospheric processes are governed primarily by external solar influences,
and the mesosphere, which is influenced primarily by internal variability
produced by upward propagating waves.
Recent studies have suggested that the mesopause may have been
perturbed in response to human-induced emissions of chemical com-
pounds, including carbon dioxide and methane. Although a cooling of
the mesosphere has been reported, not enough data is currently available
to quantify any change that could have occurred in the vicinity of the
mesopause during recent decades.
It has long been suspected that wave drag in the mesosphere results
from the dissipation of gravity waves (e.g., Lindzen, 1967). Radar
observations (e.g., Fritts and Vincent, 1987) support the view that the
momentum flux associated with these waves accounts for the momentum
budget of the mesosphere. Several parameterizations to represent these
effects have been developed (Lindzen, 1981; Weinstock, 1982; Fritts and
Lu, 1993; Medvedev and Klaasen, 1995; Hines, 1997a,b) and used in
atmospheric models. Lindzen (1981), for example, assumes that, above
the breaking level, gravity waves produce strong turbulent diffusion and
further growth of their amplitude is prevented. This assumption leads
to expressions for the wave drag and the turbulent diffusion that are
strong functions of the Doppler shift, u − c, the difference between the
zonally averaged wind (u) and the phase speed (c) of the wave:
k(u − c)3
Gu = − (3.77)
2HN
k(u − c)4
Kz = (3.78)
2HN3
In these expressions, k is the horizontal wavenumber of the gravity wave,
H is the atmospheric scale height, and N is the Brunt-Vaisala frequency
(see e.g., Eqs. 3.19a-b, and Holton, 1992). Equation (3.77) is a more
satisfactory description of the effect of wave drag in the mesosphere than
is the simple Rayleigh friction used in Equation (3.75), since it is based
on a simple but physically motivated description of the wave breaking
process. Equation (3.77) can produce positive or negative acceleration
of the zonal wind, depending on the sign of u − c. Figure 3.33 shows a
typical meridional distribution of the forcing produced on the zonal mean
velocity as a result of gravity wave breaking. The resulting meridional
circulation consists of a single cell directed from the summer to the
Figure 3.33. Meridional distribution of the zonal momentum forcing associated

with gravity wave breaking (m s−1 /day) as calculated by the 2-D model of
Brasseur et al. (2000). The forcing is positive in the summer hemisphere
and negative in the winter hemisphere, so that gravity wave breaking tends
to decelerate the zonal flow in both seasons. The pole-to-pole wave driven
circulation in the mesosphere (solid lines with arrows) is schematically
represented.
winter poles. Air is therefore drawn upward toward the mesopause in

summer and downward away from the mesopause in winter. As a result,
the mesosphere is kept away from thermodynamic equilibrium, with very
low temperatures at the summer mesopause (130 K) and relatively high
temperature at the winter mesopause (210 K).
It should be noted that, since the zonal wind distribution varies con-
siderably with season, equations (3.77)-(3.78) introduce the possibility of
seasonal variability of wave drag and diffusion. Holton (1982; 1983) has
shown that the filtering produced by stratospheric winds (see Section 3.4
and Figure 3.11) does indeed result in seasonal differences in mesospheric
wave drag. Wave breaking produces a westward force in winter and an
eastward force in summer, leading in both seasons to a deceleration of the
zonal wind in good agreement with observations. Clearly, if wave drag
varies seasonally, so too must the mixing produced by wave breaking, a
phenomenon that has consequences on the vertical transport of chemical
species (like water vapor) in the mesosphere (see Garcia and Solomon,
1985; Holton and Schoeberl, 1988).
Gravity wave breaking also affects the heat budget of the mesosphere
indirectly through changes in the meridional circulation induced by
momentum deposition. Direct thermal effects have also the potential to
play a significant role (see e.g., Huang and Smith, 1991; Lieberman et al.,
2000) through conversion of turbulent energy to heat (see Eq. 3.69a)
(1) 1 ∂
GΘ = ρ0 w Θ
ρ0 ∂z
and, above the gravity wave breaking level, by transport of heat through
induced turbulence

(2) 1 ∂ ∂Θ
GΘ = ρ0 Kh
ρ0 ∂z ∂z
(1) (2)
In these expressions, GΘ and GΘ represent additional dissipative terms
to be added to the thermodynamic equation (3.69). Kh represents the
vertical eddy diffusion coefficient for heat, which is related to the vertical
eddy diffusion coefficient for momentum and tracers (Eq. 3.78) by the
Prandtl number Pr:
Kz
Kh =
Pr
Strobel et al. (1985) and Strobel (1989) showed that tracer and heat
transport in the mesosphere were best reproduced for a Prandtl number
of 3. Huang and Smith (1991) suggested the choice of a Prandtl number
of 10, which implies that gravity wave breaking has a small direct effect
on the thermal budget of the mesosphere. Gavrilov and Roble (1994)
(1) (2)
showed that gravity wave dissipation GΘ and turbulent heating GΘ
are nearly in equilibrium near 100 km altitude.
The origin and magnitude of wave drag in the stratosphere is also an
important question for stratospheric dynamics. Geller (1983) suggested
that gravity wave breaking may provide much of the frictional dissipation
in the stratosphere as well as in the mesosphere. However, Rossby wave
transience and dissipation must also be important, at least in winter.
Figure 3.34 shows a climatological view of the Eliassen-Palm (EP) flux
and its divergence produced by the absorption of Rossby waves during
the month of January. The EP flux is directed upward and equatorward.
Wave propagation into the stratosphere is observed primarily in the
winter hemisphere and the momentum forcing on the mean flow is largest
between 30◦ N and 70◦ N. As indicated earlier (McIntyre and Palmer,
1983; 1984), Rossby waves tend to break in the vicinity of the critical
surface (u = c), where a substantial EP flux convergence and hence a
strong wave-mean flow interaction take place (see Figure 3.34). Such a
situation is accompanied by strong quasi-horizontal mixing on isentropic
surfaces over distances of typically several thousands of kilometers (see
also Figure 3.10). The mechanism leading to Rossby wave instability
is not, as in the case of gravity waves, the presence of vertical static
-3
-1
-1
Figure 3.34. Eliassen-Palm (EP) flux (arrows) and its divergence (dashed lines)
represented as a function of altitude and latitude for January. The EP flux
propagates in the stratosphere only during winter. The EP flux divergence is
associated with Rossby wave transience and dissipation. Courtesy of W. Randel,
NCAR (2001).
instability, but a reversal of the potential vorticity gradient on an

isentropic surface.
In late winter and spring within the space of a few days, large
temperature increases are occasionally observed to occur in the lower
stratosphere, accompanied by cooling in the mesosphere and upper
stratosphere (e.g., Labitzke, 1980). Under these circumstances, the
stratopause may be found to be as low as 35 km altitude. The
polar night jet decreases in strength and on certain occasions even
reverses (breakdown of polar vortex). Matsuno (1971) and Holton (1976)
have simulated such sudden stratospheric warmings by considering the
interaction of large scale waves that propagate from the troposphere with
winds in the stratosphere and mesosphere. Matsuno (1971), Matsuno
and Nakamura (1979), Dunkerton et al. (1981), and Palmer (1981),
have further emphasized that sudden warmings are caused by strong
planetary wave drag due to wave transience and critical level absorption.
As a result of the decelerated flow, the propagation of planetary waves
is modified and produces an even stronger deceleration of the mean
zonal wind at a lower altitude. At the same time, a mean meridional
circulation is induced with strong downward motion at high latitude
and upward motion in the tropics, producing the observed high latitude
warming as a result of adiabatic compression. Sudden warmings have
significant effects on the annual cycle of total ozone, and on many
other constituents. Observations (e.g., Naujokat and Labitzke, 1993)
show that the wind deceleration associated with momentum convergence
often starts in the upper stratosphere and propagates downward. A
sudden warming is considered to be major when the zonal wind reverses
in the region poleward of 60◦ and above the height corresponding to
10 hPa. Extreme warming events are occasionally observed in the upper
stratosphere where the temperature (which is typically 240 K at 40 km
in January) can reach more than 320 K at certain locations near the
pole (see e.g., von Zahn et al., 1998).
In the tropics, where the dynamics is determined by the vanishing
Coriolis force and its maximum change with latitude, the zonal mean
circulation in the middle atmosphere is characterized by oscillatory
equatorial zonal wind systems, including super-rotating westerly jets,
and a spectrum of transient vertically propagating waves, some of
them equatorially trapped. At these latitudes, the zonal forces exerted
by dissipating equatorial eastward and westward waves can efficiently
accelerate (decelerate) zonal currents or jets and indirectly cause
secondary meridional circulations. Existing zonal jets in turn alter
the propagation and dissipation conditions for waves emerging from
the troposphere, leading to a strong vertically varying wave mean-flow
interaction and dynamical vertical coupling of the atmospheric layers.
The most prominent examples of oscillatory low latitude wind systems
are the quasi-biennial oscillation (QBO) in the equatorial stratosphere
and the semi-annual oscillation (SAO) at the equatorial stratopause
(Figure 3.35). The quasi-biennial oscillation (QBO) (Veryard and
Ebdon, 1961; Reed et al., 1961) is characterized by the downward
propagation of westerly and easterly jets from the upper stratosphere
to the lower stratosphere near 70 hPa, almost without loss in amplitude,
at a rate of typically 1 km per month (Figure 3.36). The period of the
QBO varies between 22 and 34 months (28 months on average), and the
amplitude of the westerly and easterly jets is typically 15 m/s and -30
m/s, respectively. The QBO is the main cause of interannual variance
of the zonal wind in the equatorial stratosphere. Detailed reviews are
given by Dunkerton and Delisi (1985) and Baldwin et al. (2001).
Figure 3.35. Time-height cross section of the zonal wind near the equator
highlighting the quasi-biennial oscillation between 16 and 36 km and the semi-
annual oscillation in the vicinity of the stratopause. Solid lines are drawn at
intervals of 10 ms−1 . Shaded areas denote flow from west to east (westerlies).
Other areas correspond to westward winds. The mean value of the wind has
been filtered out. From Wallace (1973).
Figure 3.36. Time height cross-section of the zonal wind at Singapore between
1992 and 1997 with the quasi-biennial oscillation between the pressure levels of
100 and 10 hPa. Contour intervals 10 m/s. Shaded areas represent eastward
(or westerly) winds. Courtesy of B. Naujokat, Free University of Berlin.
Theoretical studies (Lindzen and Holton, 1968; Holton and Lindzen,

1972; Dunkerton, 1997) as well as laboratory simulations of the analogue
of large scale geophysical flows (Plumb and McEwan, 1978) suggest that
the origin of the QBO lies in the vertical transport of momentum by
certain types of equatorial waves and by gravity waves. Although the
QBO is primarily a tropical phenomenon, it has substantial impact in
the extratropics, including on the breakdown of the polar vortices in
spring (Holton and Austin, 1991; O’Sullivan and Young, 1992). By
filtering tropical waves that propagate upward, the QBO also affects
the variability of the mesosphere. Atmospheric tracers including water
vapor, methane, and ozone respond to the circulation changes (see Plate
5). The importance of the coupling between the QBO anomalies in ozone
concentration, stratospheric heating rates, temperature and winds has
been reviewed by Hasebe (1994).
The semi-annual oscillation (SAO), which is observed in the zonal
mean flow near the equatorial stratopause with maximum westerlies at
the equinox and maximum easterlies at the solstices (see Figure 3.35),
is also believed to be of dynamical (rather than radiative) origin.
Unlike the QBO, the SAO is locked to the annual cycle. The westerly
accelerations in March and September are believed to be caused by
Kelvin and high frequency gravity waves, while the easterly acceleration
in June and December is produced by the advection of easterly
momentum across the equator from the summer hemisphere and involve
EP flux convergence produced by Rossby wave breaking. The QBO
signal is evident below 36 km, while the SAO signal is dominant in the
upper stratosphere. A separate SAO observed near the mesopause is
probably caused by gravity wave breaking with alternate easterly and
westerly forcing around 80-90 km altitude. The signal of the SAO is also
observed in the concentrations of chemical compounds.
Finally, increasing attention has recently been given to an important
circulation anomaly observed in the Northern Hemisphere (see Box 3.3)
and characterized by a nearly symmetric oscillation in temperature,
wind, and pressure between the Arctic and mid-latitudes (Thompson
and Wallace, 1998). This annular mode of oscillation in the climate
system is called the Arctic oscillation (AO), and extends from the surface
to the upper stratosphere. The AO affects the strength and location of
the Arctic vortex, with associated effects on stratospheric ozone (Wallace
and Thompson, 2002).
Box 3.3 Stratospheric Dynamics, Weather, and Climate

There have been long debates in the scientific community regarding
the possible influence of stratospheric dynamics on weather and climate
at the Earth’s surface. Dynamical links between the stratosphere and
the troposphere were suggested by Charney and Drazin (1961) and
Matsuno (1970) several decades ago, and more recent investigations have
provided strong evidence of the relationship between the tropospheric
and stratospheric circulations, and specifically between the strength
of the stratospheric winter vortex and the structure of tropospheric
planetary waves (Boville, 1984; Kodera, et al., 1991; 1996; Baldwin et al.,
1994; Perlwitz and Graf, 1995; Perlwitz et al., 2000). Links between
stratospheric and tropospheric circulations have been studied in relation
with large volcanic eruptions (Graf et al.., 1993; 1994; Kodera, 1994),
solar-terrestrial relationships (Geller and Alpert, 1980; Labitzke and van
Loon, 1988; Kodera and Yamazaki, 1990), and middle atmosphere cooling
(Kodera and Yamazaki, 1994; Shindell et al., 2001.) The discovery of
certain recurrent spatial patterns in the dynamics of the atmosphere,
called modes of variability, provides some evidence for dynamical couplings
between the middle and lower atmosphere. These modes can often
be regarded as the signature of dynamical interactions and feedback
mechanisms in the Earth system.
One of the patterns that has recently received wide attention is
the Northern Hemisphere annular mode (NAM), also called the Arctic
Oscillation (AO), a distinctive annular feature in the sea-level pressure
deviation near 55◦ N (Thompson and Wallace, 1998; 2000). A similar
mode (SAM) is also observed in the Southern Hemisphere. These
patterns correspond to the annular modes provided by a set of empirically
determined orthogonal functions (EOF) that can be derived from the
observed dynamical fields. The NAM’s index is a normalized value of the
NAM expansion coefficient: a positive value corresponds to below-normal
sea-level surface pressures in the Arctic with strong eastward winds around
50-60◦ N, especially across the North Atlantic, and higher than normal
temperatures in the Northern Hemisphere. Under those conditions,
Northern Europe during wintertime is subject to precipitations, while the
southern part of the continent is sunny. A negative index is suggestive of
weak surface eastward winds over the Atlantic, with slowly propagating
weather systems; winter cold air masses are expanding from Russia
towards Northern Europe (with occasionally extremely cold conditions),
and storms are developing in the Mediterranean. As the contrasts in
meteorological variables appear to be most coherent over the Atlantic
sector, the Northern Hemisphere annular mode is often referred to as
the North Atlantic Oscillation (NAO). Fragmentary evidence of this
oscillation was already appearing in the beginning of the 20th century
(Exner, 1913; Walker, 1924; Walker and Bliss, 1932).
Box 3.3 (Continued)

Analyses of meteorological fields by Baldwin and Dunkerton (1999,
2001) have highlighted a statistically significant relationship between
the strength of the stratospheric winter polar vortex and the Northern
Hemisphere annular mode: a strengthening of the polar vortex tends to
be associated with a positive NAM index at the Earth’s surface, while
a weakening of the polar night jet induces negative NAM indices. A
statistical analysis based on several wintertime events shows that the
signal produced by a reduction in the strength of the polar vortex at 30 km
propagates down and reaches the troposphere with a time lag of several
weeks (Baldwin and Dunkerton, 2001). Although downward propagation
(within a timeframe of 5-10 days) of middle atmosphere planetary-scale
perturbations (such as those resulting from stratospheric warming events)
to the lower stratosphere and perhaps the upper troposphere are well-
known features (see e.g., Matsuno, 1971; Kodera and Kuroda, 2000a,
b), the recent analyses of Baldwin and Dunkerton (2001) suggest that
the stratospheric signal penetrates as far down as the surface, and with a
phase delay of typically 30-40 days. Links between stratospheric dynamics
and tropospheric weather patterns offer therefore hope for improved long-
range weather forecasts (Wallace and Thompson, 2002).
Trends in the past 30 years toward higher positive NAM indices
(Thompson et al., 2000) have favored milder and wetter winters in
Northern Europe together with a strengthening of the polar night jet
and reduced ozone abundances in the stratosphere at high latitudes.
Stratospheric zonal wind anomalies produced by changes in solar activity,
by the equatorial QBO, by volcanic eruptions, by stratospheric ozone
depletion and by the increase in carbon dioxide concentration could trigger
or modulate internal modes of stratospheric variability and generate
similar patterns that propagate downward into the troposphere.
The observed trends in the NAM towards higher indices may have
resulted from human-induced changes in the temperature structure of
the lower stratosphere in response to greenhouse gas emissions and ozone
depletion (Shindell et al., 2001). The radiative effects of solar activity
and of volcanic eruptions may also have produced NAM- like signatures
detectable at the Earth’s surface. The response of the Earth system
to climate forcing may therefore involve changes in particular dynamical
modes, and hence the human influence on climate at the Earth surface
may occur in part by way of the stratosphere.
3.8 Vertical Transport above the Mesopause
Above the mesopause (85-95 km altitude), the air density starts to

decrease more slowly with height than at lower altitudes, suggesting that
the atmospheric scale height is progressively increasing. This change

results primarily from a dramatic temperature increase in the lower
thermosphere as well as a simultaneous decrease with height in the mean
molecular mass of air. In the vicinity of the homopause (transition
between the homosphere and the heterosphere near 100 km), vertical
mixing (which leads to a uniform mixing ratio of major gases such as
N2 or O2 below 100 km) is gradually replaced by diffusive separation of
atmospheric constituents according to their molecular mass. As a result,
the lightest compounds (such as the hydrogen and helium atoms) are
transported upward and their relative abundances increase with altitude.
Conversely, the heaviest species (such as N2 or O2 ) are transported
downwards and, as a result, their mixing ratio decreases very rapidly
with height. Figure 3.37 shows an example of the vertical profiles derived
for several compounds between the surface and 1000 km altitude.
The thermosphere is also characterized by large molecular viscosity
(which affects the circulation of the air) and by rapid thermal conduction
(which is the principal transport process for thermal energy above
100 km altitude). Under these conditions, the momentum and
thermodynamic equations ((3.1) and (3.2)) need to be slightly modified
by the inclusion of additional terms. For a mixture of gases, they can
Figure 3.37. Vertical distribution of the concentration of several chemical

species in the thermosphere. The importance of diffusive separation is visible.
From Richmond (1983).
be expressed as
dv 1
× v = g + µ ∇2 v +
+ ∇p + 2Ω Fi
dt ρ ρ i
and
dT 1 dp 1
cp − + ∇·E=Q
dt ρ dt ρ
with most of the symbols defined in Section 3.3.1. In the above
expressions, µ is the viscosity coefficient and Fi accounts for the external
is the heat flow
forces (e.g., ion drag) applying to i-type particles. E
vector, which is proportional to the temperature gradient
= −λ ∇
E T
where λ is the thermal conductivity coefficient. Expressions for µ and λ
can be derived from the gas kinetics theory. Semi-empirical expressions
of the form
µ = A Ts
λ = B Ts + CT + D
can also be applied for a mixture of gases (see e.g., Banks and Kockarts,
1973), with s usually chosen as equal to 0.69, and A, B, C, and D varying
with the relative abundance of the major thermospheric compounds.
Vertical transport of chemical compounds in the thermosphere is
governed primarily by molecular diffusion. The vertical flux of chemical
species associated with molecular diffusion can be derived from the
momentum equations applied to thermospheric conditions (Gombosi,
1998). Considering elastic collisions between particles, but ignoring all
reactive collisions, and taking into account the effect of thermal diffusion,
the vertical “molecular diffusion velocity” for species i at geometric
altitude zg can be expressed as (Banks and Kockarts, 1973)

1 dni 1 dT 1
wD,i = −Di + (1 + αT ) + (3.79)
ni dzg T dzg Hi
where ni is the concentration of minor constituent i, Hi = kT/mi g
is the scale height of this species and mi its molecular mass. αT is
the thermal diffusion factor (–0.40 for helium, –0.25 for hydrogen, 0
for heavier particles) and Di the molecular diffusion coefficient. This
coefficient is related to the collision frequency between species i and all
other molecules and atoms. Its value can be derived from the classic
gas kinetic theory (Chapman and Cowling, 1970). An approximate
expression for Di (in cm2 s−1 ) is provided by Banks and Kockarts (1973):
1 1
1 1 2 T2
Di = 1.52 × 10 18
+ (3.80)
mi m n
where the molecular mass of species i (mi ) and the mean molecular
air mass (m) are expressed in amu, the temperature (T) is expressed
in Kelvin and the total air density (n) is expressed in cm−3 . Since
it is inversely proportional to the air density, the molecular diffusion
coefficient Di increases quasi-exponentially with height with values of
the order of (1-5) × 104 cm2 s−1 at 80 km and (1-5) × 107 cm2 s−1 at
120 km.
As indicated earlier, vertical exchanges associated with turbulent
mixing such as that caused by gravity wave breaking below the
homopause are often represented as eddy diffusion. When expressed
in terms of number density in an atmosphere following hydrostatic
equilibrium conditions, the effective vertical “eddy diffusion velocity”
becomes

1 dni 1 dT 1
wE,i = −Kz + + (3.81)
ni dzg T dzg Hg
where Kz is the vertical eddy diffusion coefficient and Hg is the

atmospheric scale height. In the vicinity of the homopause, the effects
of both eddy and molecular diffusion have to be taken into account, and
the total vertical diffusion velocity is
wi = wE,i + wD,i (3.82)
In the absence of significant chemical production or destruction mech-

anisms, a chemical compound follows diffusive equilibrium conditions
(wi = 0), and the vertical profile of its density can be expressed as
(Banks and Kockarts, 1973)

zg
T0 1 Λi (z ) dz
neq
i (zg ) = ni,0 exp − +
T (zg ) zg,0 Hi (z ) Hg (z ) 1 + Λi (z )
zg
αT 1 dT
− dz (3.83)
zg,0 1 + Λi (z ) T (z ) dz

where ni,0 and T0 are the number density of species i and temperature at
altitude zg,0 , respectively, and Λi = Kz /Di is the ratio between eddy and
molecular diffusion coefficients. In the homosphere where Λi → ∞, the
vertical profile of an inert gas follows the hydrostatic law (corresponding
to perfect mixing)

zg
T0 dz
neq
i (zg ) = ni,0 exp − (3.84a)
T (zg ) zg,0 Hg (z )
while in the heterosphere, where Λi → 0, the diffusive equilibrium

condition becomes

1+αT
z
T0 dz
neq
i (zg ) = ni,0 exp − (3.84b)
T (zg ) zg,0 Hi (z )
In both cases, the number density of inert (or quasi-inert) species i
decreases exponentially with height. However, in the homosphere, the
rate of decrease is characterized by the atmospheric scale height Hg (and
is therefore identical for all inert gases), while in the heterosphere, it is
related to the scale height Hi and hence the mass mi of species i (and
is therefore specific to each compound). As a result, the density of the
lightest particles decreases the least rapidly with height (see Figure 3.37).
Note that in numerical models of the middle atmosphere, the upper
boundary condition is often prescribed by the maximum diffusion flux
Φmax
i = ni wimax where wimax is given by (Banks and Kockarts, 1973;
Gombosi, 1998)

Di mi
wimax = 1− (3.85)
Hg m
In the case of the hydrogen atom (mi = 1) whose density is of the
order of 108 cm−3 at 100 km, the upward vertical flux at this height is
typically a few times 108 cm−2 s−1 . The diffusive velocity wimax increases
with height. Above 500 km, the mean free path of atoms become so
long that collisions can be ignored: particles follow individual balistic
trajectories. If the kinetic energy of atoms becomes larger than the
potential energy required for these particles to be lifted out of the Earth’s
gravitational fields, these atoms escape to space. The probability that
an atom reaches the escape velocity (typically 11 km s−1 at 500 km) is,
however, extremely low, except in the case of hydrogen and, to a lesser
extent, helium. The loss of hydrogen to space is compensated by the
continuous production of H by water vapor photolysis above 60-70 km
altitude.
3.9 Models of the Middle Atmosphere
3.9.1 General Circulation Models
Three-dimensional models which provide solutions to some form of the

primitive equations outlined in Section 3.3 are called general circulation
models (GCMs). These models can provide insight on the coupling
between dynamical and radiative processes in the atmosphere. They
resolve large-scale waves and synoptic eddies, and include state-of-the-
art parameterizations to represent unresolved (small-scale) processes.

Deficiencies in model simulations often lead to the identification of key
processes that have been omitted or inadequately treated in numerical
models. In the most recent climate system models, atmospheric GCMs
coupled to ocean circulation models, ice models, and continental surface
models attempt to simulate the evolution of the entire Earth system and
its response to various types of climate forcings. A long-term objective
of GCMs is to quantify natural variability in the atmosphere and to
assess the response of the atmosphere and to perturbations of natural
or of human origin.
Ideally, the solution of the primitive equations for specified external
constraints (e.g., the solar irradiance at the top of the atmosphere) and
appropriate boundary conditions (e.g., observed sea-surface tempera-
ture) should provide a comprehensive representation in space and time
of the atmospheric dynamical system. In practice, however, limitations
in computer capabilities impose limits on the spatial resolution of
these models, so that small-scale processes, rather than being explicitly
reproduced, must be parameterized. The uncertainties associated
with these physical parameterizations (e.g., boundary layer exchanges,
convection, clouds, gravity wave breaking, etc.) often limit the overall
accuracy in the model results.
General circulation models of the middle atmosphere cannot ignore
the dynamical and other physical processes occuring in the troposphere,
due to the strong coupling between these two atmospheric regions.
Middle atmosphere GCMs generally extend therefore down to the
Earth’s surface. An exception is provided by the so-called middle
atmosphere mechanistic models in which the effects of tropospheric
processes are represented by lower boundary conditions in the vicinity
of the tropopause (e.g., 200 or 300 hPa). These models are suitable for
performing sensitivity tests and studying specific physical or chemical
processes in the stratosphere and mesosphere.
The first general circulation model extending to the stratosphere
was developed by Smagorinsky et al. (1965) at the Geophysical Fluid
Dynamics Laboratory (GFDL) (Princeton, New Jersey, USA). This
model covered only the Northern Hemisphere and included 9 vertical
levels, among which 3 were in the stratosphere. Manabe and Hunt
(1968) used a similar model, but included 18 levels up to the altitude of
37.5 km. After these early attempts to represent dynamical fields above
the tropopause, many other middle atmosphere GCMs were developed
and used to address scientific questions (Fels et al., 1980; Simmons and
Strufing,
¨ 1983; Mechoso et al., 1985; Dequ´
´ e et al., 1994; Boville, 1995;
Hamilton et al., 1995; Kitoh et al., 1995; Chiba et al., 1996; Beagley
et al., 1997; Langematz and Pawson, 1997; Manzini et al., 1997; Rind
et al., 1998; Swinbank et al., 1998). Many middle atmosphere models

extend to the mesopause, or to even higher altitudes.
The three-dimensional equations of atmospheric dynamics are
generally solved using different numerical approaches: In grid point
models (see e.g., Arakawa, 1966), the variables are defined by their
values at particular points spaced in longitude, latitude, and altitude
(model grid). The derivatives appearing in the equations are replaced
by finite differences. The stability and accuracy generally depends on
the spatial resolution of the model (size of the spacing between grid
points) and on a time step chosen to solve the equations (typically 10-30
minutes in most models). In spectral models (see e.g., Bourke, 1972), the
horizontal distribution of the meteorological fields is expressed in terms
of truncated spherical harmonic expansions (which can be regarded as
the superposition of atmospheric waves). This method has the advantage
of allowing exact calculation of the horizontal derivatives. The horizontal
resolution of the model is provided by the number of waves retained in
the harmonic expansions. This number is typically 21, 30, or 42 in most
current middle atmosphere spectral models.
Figure 3.38 shows the zonal mean wind velocity calculated for January
conditions by a number of middle atmosphere GCMs (Pawson et al.,
2000), together with a reference climatology (Randel, 1992). The degree
of success in reproducing the observed winds depends strongly on the
formulation of dissipative processes including diabatic heating (radiative
transfer) and momentum deposition (gravity wave parameterization).
In the case of gravity waves, the simplest parameterization is provided
by the introduction of a Rayleigh friction term in the momentum
equation (see Eq. (3.75)). More elaborate approaches are based on
the recent parameterizations of Lindzen (1981), Medvedev and Klaasen
(1995), or Hines (1997a,b), which account for orographic (zero phase
speed) and non-orographic (nonzero phase speed) gravity waves (see
Section 3.4). In some models, the temperature calculated in the polar
winter stratosphere is colder than observed, and hence the polar jet is
too strong. This cold polar bias is significantly reduced in the lower
stratosphere when the effects of orographic gravity waves are included
in the model (Boville, 1995).
3.9.2 Chemical-Transport Models
Chemical-transport models (CTMs) simulate the formation, trans-

port, and destruction of chemical compounds in the atmosphere. These
models can be directly coupled to the general circulation models
Comparison of the zonal wind calculated for January by several

general circulation models of the middle atmosphere with climatological values.
From Pawson et al. (2000).
described in the previous section and run “on-line”. In this case, the
three-dimensional transport of chemical tracers is calculated using the
winds derived at each timestep by the GCM, and diabatic heating

and cooling rates calculated in the GCM make use of the calculated
distributions of radiatively active compounds (ozone, water vapor,
carbon dioxide, etc.) in the CTM. “Off-line” chemical-transport models
are uncoupled from the GCMs. The winds and temperature required to
calculate the distribution of chemical species are provided at regular time
intervals (e.g., every 6 or 12 hours) from pre-calculated dynamical fields
or from meteorological analyses based on observations. In the latter
case, the distributions of calculated chemical tracers can be compared
to the distributions observed during specific time periods.
The first attempts to simulate the three-dimensional global distribu-
tion of chemically active species in the stratosphere have been based
on the “on-line” approach. Hunt (1969) introduced ozone in the early
GCM developed at GFDL, and studied the transport of this compound,
using a simple parameterization for its photochemistry. Somewhat more
sophisticated coupled models for ozone and related compounds have
been developed by Cunnold et al. (1975) and Schlesinger and Mintz
(1979).
The role of dynamics in controlling the distribution of idealized
tracers (with extremely simple chemistry) has been investigated in some
detail by Mahlman and Moxim (1978), Levy et al. (1979), Mahlman
and Umscheid (1983), Sassi et al. (1993), and many others. Levy
et al. (1979), for example, studied the behavior of a long-lived compound
like N2 O, including its zonal mean concentration and the longitudinal
deviation in the concentration. More recent chemical-transport models
(e.g., Grose et al., 1987; Rose and Brasseur, 1989; Austin et al., 1992;
Pitari et al., 1992; Chipperfield et al., 1993; Lefèvre
f et al., 1994; Rasch
et al., 1995; de Grandpré et al., 1997; Brasseur et al., 1997) have included
relatively comprehensive representations of the photochemistry in the
middle atmosphere. State-of-the-art CTMs account typically for 150-
250 chemical (gas phase and heterogeneous) reactions affecting 50-100
chemical species.
Resolved transport by the atmospheric circulation (advection) is
simulated through different numerical methods (see Box 3.4). The
performance of transport algorithms is measured by evaluating their
accuracy, stability, and computational efficiency. In addition, important
properties for the adopted method are its ability to conserve mass, to
preserve the shape of the solution (by avoiding overshoots or undershoots
near regions of strong gradients), to be local (so that the solution at a
given point is not strongly influenced by the field far from this point), and
to advect any perturbation downwind only. Spectral transport methods,
for example, are computationally efficient, but non-local. Most Eulerian
algorithms require small timesteps. In Lagragian schemes (where a
Box 3.4 Numerical Algorithms to Solve the Transport Equation

Advective transport is described by partial differential equations. In
the simple one-dimensional (1-D) case where the wind u is directed along
direction x, the evolution with time of a conserved quantity ψ (e.g., tracer
mixing ratio, potential temperature, potential vorticity) is expressed as
∂ψ ∂ (uψ)
+ =0 (1)
∂t ∂x
If the wind velocity u is a constant and the initial distribution of the
transported quantity at time t = 0 is ψ0 (x), the analytic solution of (1)
at location x and time t is provided by:
ψ (x, t) = ψ0 (x − ut) (2)

In the more general case where the wind velocity varies with space and
time, the solution of (1) is derived by numerical algorithms. Rood, (1987)
has reviewed several practical methods. None of these algorithms can be
regarded as the single universal method that is recommended in all cases.
The choice of the method will usually result from a compromise between
different requirements (e.g., accuracy, efficiency, etc.). These include the
accuracy of the solution, the stability of the method (which often imposes
a restriction on the timestep), and the efficiency of the algorithm (limited
computer resources). Additional criteria should be met: conservation (no
mass gain or loss), transportivity (perturbations should be advected only
downwind), locality (solutions at a given point should not be substantially
affected by the field far from the point), monotonicity (no overshoots or
undershoots should be produced by the algorithm). If ψ represents the
concentration of a tracer, it should remain positive in the entire model
domain.
In Eulerian schemes, the solution of (1) is provided on fixed grid points
(i−1, i, i+1, . . .) at time intervals (n−1, n, n+1, . . .). In this case, Equation
(1) is replaced by a discretized form such as
∆t
ψin+1 = ψin − (uψ)i+ 1 − (uψ)i− 1 (3)
∆x 2 2
where the flux (uψ) is evaluated at the right (i + 12 ) and left (i − 12 )

boundaries of the numerical box. Methods differ by the way these fluxes
are evaluated. Some of them are diffusive (e.g., the upstream scheme)
or produce noise near strong gradients in the advected field (e.g., the
leapfrog scheme). In order to overcome these limitations Smolarkiewicz
(1983) introduced an “antidiffusion velocity” in the upstream algorithm
and Bott (1989), rather than considering the value of ψ to be constant over
individual grid-boxes, fitted the function by a polynomial within each box.
Prather (1986) represented ψ by a second-order polynomial and assumed
Box 3.4 (Continued)

that the zeroth order moment (mean value of ψ within the grid cell),
first order moment (slope of ψ) and second-order moment (curvature of
ψ) are conserved during an advective step. His method provides very
little numerical diffusion, but requires considerable amounts of computer
memory to store various moments.
A practical limitation of the Eulerian schemes is that the convergence
of the numerical integration is achieved only if the adopted timestep ∆t ≤
C∆x/u, where the constant C is generally equal to unity. This criterion,
called the Courant-Fredrichs-Lewy (CFL) condition, imposes timesteps
that are generally much smaller than the physical timescales of interest.
In Lagrangian algorithms, the displacement of distinct air parcels (in
which the tracer is assumed to be homogeneously mixed) is calculated
as a function of time. If a sufficiently large number of parcels are
considered, the evolution of a tracer over a global or regional domain can
be determined. The accumulation of errors in the determination of the
parcel location and the lack of mixing allowed between neighboring parcels
provides a major limitation of the method, especially if the integration is
carried on over long periods of time. Lagrangian methods are, however,
often used in conjunction with field observations to assess the evolution
of chemical species within a limited number of air parcels along their
trajectory during a few days.
The semi-Lagrangian method (Robert, 1981; Staniforth and Coté, ´ 1991)
determines through a backward Lagrangian trajectory calculation the
origin at time t − ∆t of air parcels which have reached the model grid
points at time t. The value of ψ at these original points is determined
by interpolation of the solution at the grid points at time t − ∆t. The
accuracy of the method depends greatly on the type of interpolation used.
A major limitation of the method is that it is intrinsically nonconservative.
The flux-form algorithm of Lin and Rood (1996) is a combination of
semi-Lagrangian and finite volume methods. It is mass conservative.
Advantages of the semi-Lagrangian method are the circumvention of the
CFL stability condition and, at the same time, the convenience of a regular
mesh discretization.
large number of distinct air parcels with prescribed mixing ratios are
followed along the wind trajectories), the accumulation of errors in
the determination of the parcel trajectory over long periods of time
introduces a limitation when applied to chemical-transport models. The
semi-Lagrangian approach (Robert, 1981), in which the trajectory of air
parcels is derived over 1 time-step and interpolations are made on the
Eulerian grid of the model, is not mass-conserving. Conservative flux-
form semi-Lagrangian methods, however, have been recently developed
(Lin and Rood, 1996).
Unresolved transport, such as mixing processes associated with gravity

wave breaking in the mesosphere, is generally expressed as vertical
eddy diffusion (see Eq. (3.52)) with the corresponding coefficient being
determined through parameterization algorithms.
A large aspect of model development is the evaluation and verification
of the results, mainly through extensive comparisons with observations.
Global distributions of chemical species observed from space have been
very useful in testing these models. Models are also used increasingly
to “assimilate” observational data. The purpose is to combine our “a
priori” knowledge of the atmosphere (model) with observations that are
usually irregularly distributed in space and time to obtain an optimal
representation of the atmospheric system.
In many cases, a full three-dimensional treatment may not be
necessary to resolve many of the important features of the distributions
of chemical species. The time constant associated with transport by
the zonal wind is relatively rapid, of the order of days. Many chemical
species with which we shall be concerned have time constants much
longer than this, and may therefore be considered to be “well mixed” in
the zonal direction. In view of the computational limitations associated
with global three-dimensional models, and the relatively small impact
of zonal transport on longer lived species, two-dimensional (zonally
averaged) models are often used to describe the global distribution of
chemical species in the middle atmosphere, and the response of the
atmosphere to external perturbations. Different approaches can be
used to derive the zonal mean of the quantities involved. As shown in
Section 3.5.3, it is often advantageous to write the zonal mean primitive
equations and continuity equations in the Transformed Eulerian Mean
(TEM) framework. Most current models (e.g., Garcia and Solomon,
1983; Brasseur et al., 1990; Fleming et al., 1999) use this formulation.
References
Andrews, D.G., and M.E. McIntyre, Planetary waves in horizontal and vertical shear:
The generalized Eliassen-Palm relation and the zonal mean acceleration. J Atmos Sci:
33, 2031, 1976.
Andrews, D.G., and M.E. McIntyre, An exact theory of nonlinear waves on a
Lagrangian mean flow. J Fluid Mech: 89, 609, 1978.
Andrews, D.G., J.R. Holton, and C.B. Leovy, Middle Atmosphere Dynamics.
Academic Press, 1987.
Appenzeller, C., Wave developments on surface fronts and stratospheric intrusions.
Ph.D. thesis, 1117 pp., ETH, Zurich, Switzerland, 1994.
Arakawa, A., Computational design for long-term numerical integration of the

equations of fluid motions: Two-dimensional incompressible flow. J Comput Phys: 1,
119, 1966.
Austin, J., N. Butchard, and K.P. Shine, Possibility of an Arctic ozone hole in a
doubled-CO2 climate. Nature: 360, 221, 1992.
Baldwin, M.P., X. Cheng, and T.J. Dunkerton, Observed correlations between winter-
mean tropospheric and stratospheric circulation anomalies. Geophys Res Lett: 2,
1141, 1994.
Baldwin, M.P., and T.J. Dunkerton, Propagation of the Arctic Oscillation from the
stratosphere to the troposphere. J Geophys Res: 104, 30,937, 1999.
Baldwin, M.P., and T.J. Dunkerton, Stratospheric harbingers of anomalous weather
regimes. Science: 294, 581, 2001.
Baldwin, M.P., L.J. Gray, T.J. Dunkerton, K. Hamilton, P.H. Haynes, W.J. Randal,
J.R. Holton, M.J. Alexander, I. Hirota, T. Horinouchi, D.B.A. Jones, J.S. Kinnersley,
C. Marquardt, K. Sato, and M. Takahashi, The quasi-biennial oscillation. Rev
Geophys: 39, 179, 2001.
Banks, P.M., and G. Kockarts, Aeronomy. Academic Press, 1973.
Beagley, S.R., J. de Grandpré,
´ J.N. Koshyk, N.A. McFarlane, and T.G. Shepherd,
Radiative-dynamical climatology of the first-generation Canadian middle atmosphere
model. Atmos Ocean: 35, 293, 1997.
Becker, E., and G. Schmitz, Energy deposition and turbulent dissipation owing to
gravity waves in the mesosphere. J Atmos Sci: 59, 54, 2001.
Bischof, W., R. Borchers, P. Fabian, and B.C. Kruger, Increased concentration and
vertical distribution of carbon dioxide in the stratosphere. Nature: 316, 708, 1985.
Boering, K.A., S.C. Wofsy, B.C. Daube, H.R. Schneider, M. Loewenstein, J.R.
Podolske, and T.J. Conway, Stratospheric mean ages and transport rates from
observations of atmospheric CO2 and N2 O. Science: 274, 1340, 1996.
Bott, A., A positive definite advection scheme obtained by nonlinear normalization
of the advective fluxes. Mon Weath Rev: 117,7 1006, 1989
Bourke, W., An efficient, one-level primitive equation spectral model. Mon Weath
Rev: 100, 683, 1972.
Boville, B.A., The influence of the polar night jet on the tropospheric circulation in
a GCM. J Atmos Sci: 41, 1132, 1984.
Boville, B.A., Middle atmosphere version of CCM2 (MACCM2): Annual cycle and
interannual variability. J Geophys Res: 100, 9017, 1995.
Boyd, J.P., The noninteraction of waves with the zonally-averaged flow on a spherical
Earth, and the interrelationship of eddy fluxes of energy, heat and momentum. J
Atmos Sci: 33, 2285, 1976.
Brasseur, G.P., M.H. Hitchman, S. Walters, M. Dymek, E. Falise, and M. Pirre,
An interactive chemical dynamical radiative two-dimensional model of the middle
atmosphere. J Geophys Res: 95, 5639, 1990.
Brasseur, G.P., X.X. Tie, P.J. Rasch, and F. Lefèvre,
f
f` A three-dimensional simulation
of the Antarctic ozone hole: Impact of anthropogenic chlorine on the lower
stratosphere and upper troposphere. J Geophys Res: 102, 8909, 1997.
Brasseur, G.P., J.J. Orlando, and G.S. Tyndall, Atmospheric Chemistry and Global
Change. Oxford University Press, 1999.
Brasseur, G.P., A.K. Smith, R. Khosravi, T. Huang, and S. Walters, Natural
and human-induced perturbations in the middle atmosphere: A short tutorial, in:
Atmospheric Science Across the Stratopause, Geophys Monograph 123: 7

7. American
Geophysical Union, 2000.
Brewer, A.W., Evidence for a world circulation provided by measurements of helium
and water vapor distribution in the stratosphere. Quart J Roy Meteorol Soc: 75, 351,
1949.
Burrage, M.D., W.R. Skinner, D.A. Gell, P.B. Hays, A.R. Marshall, D.A. Ortland,
A.H. Manson, S.J. Franke, D.C. Fritts, P. Hoffman, C. McLandress, R. Niciejewski,
F.J. Schmidlin, G.G. Shepherd, W. Singer, T. Tsuda, and R.A. Vincent, Validation of
mesosphere and lower thermosphere winds from the high resolution Doppler images
on UARS. J Geophys Res: 101, 10,365, 1996.
Campbell, I.M., Energy and the Atmosphere. John Wiley and Sons, 1977.
Chang, A.Y., R.J. Salawitch, H.A. Michelsen, M.R. Gunson, M.C. Abrams, R.
Zander, C.P. Rinsland, M. Loewenstein, J.R. Podolske, M.H. Proffitt, J.J. Margitan,
D.W. Fahey, R.-S. Gao, K.K. Kelly, J.W. Elkins, C.R. Webster, R.D. May, K.R.
Chan, M.M. Abbas, A. Goldman, F.W. Irion, G.L. Manney, M.J. Newchurch, and
G.P. Stiller, A comparison of measurements from ATMOS and instruments aboard
the ER-2 aircraft: Tracers of atmospheric transport. Geophys Res Lett: 23, 2389,
1996a.
Zander, C.P. Rinsland, J.W. Elkins, G.S. Dutton, C.M. Volk, C.R. Webster, R.D.
May, D.W. Fahey, R.-S. Gao, M. Loewenstein, J.R. Podolske, R.M. Stimpfle, D.W.
Kohn, M.H. Proffitt, J.J. Margitan, K.R. Chan, M.M. Abbas, A. Goldman, F.W.
Irion, G.L. Manney, M.J. Newchurch, and G.P. Stiller, A comparison of measurements
from ATMOS and instruments aboard the ER-2 aircraft: Halogenated compounds.
Geophys Res Lett: 23, 2393, 1996b.
Chapman, S., and T.G. Cowling, The Mathematical Theory of Non-uniform Gases.
3rd ed., Cambridge University Press, 1970.
Chapman, S., and R.S. Lindzen, Atmospheric Tides. D. Reidel, The Netherlands,
1970.
Charney, J.G., and P.G. Drazin, Propogation of planetary-scale disturbances from
the lower into the upper atmosphere. J Geophys Res: 66, 83, 1961.
Chen, P., Isentropic cross-tropopause mass exchange in the extratropics. J Geophys
Res: 100, 16,661, 1995.
Chiba, M., K. Yamazaki, K. Shibata, and Y. Kuroda, The description of the MRI
atmospheric spectral GCM (MRI-GSPM) and its mean statistics based on a 10-year
integration. Pap Meteorol Geophys: 47,
7 1, 1996.
Chipperfield, M.P., D. Cariolle, P. Simon, R. Ramaroson, and D.J. Lary, A three-
dimensional modeling study of trace species in the Arctic lower stratosphere during
winter 1989-1990. J Geophys Res: 98, 7199, 1993.
Coffey, M.T., W.G. Mankin, and J.W. Hannigan, A reconstructed view of polar
stratospheric chemistry. J Geophys Res: 104, 8295, 1999.
Crutzen, P.J., A two-dimensional photochemical model of the atmosphere below
55 km: Estimates of natural and man-caused perturbations due to NOx , in Proc
Fourth Conf on CIAP, DOT-TSC-OST-38, 1975.
Cunnold, D.M., F. Alyea, N. Phillips, and R.G. Prinn, A three-dimensional
dynamical-chemical model of atmospheric ozone. J Atmos Sci: 32, 170, 1975.
Danielsen, E.F., Trajectories: Isobaric, isentropic, and actual. J Meteorol: 18, 479,
1961.
de Grandpre,
´ J., J.W. Sandilands, J.C. McConnell, S.R. Beagley, P.C. Croteau, and
M.Y. Danilin, Canadian middle atmosphere model: Preliminary results from the
chemical transport module. Atmos Ocean: 35, 385, 1997.
Dequ
´ é, M., C. Dreveton, A. Braun, and D. Cariolle, The ARPEGE/IFS atmospheric
model: A contribution to the French community climate modelling. Clim Dyn: 10,
249, 1994.
Dickinson, R.E., E.C. Ridley, and R.G. Roble, Meridional circulation in the
thermosphere, I. Equinox conditions. J Atmos Sci: 32, 1737, 1975.
Dickinson, R.E., E.C. Ridley, and R.G. Roble, Meridional circulation in the
thermosphere. J Atmos Sci: 34, 178, 1977.
Dobson, G.M.G., Origin and distribution of polyatomic molecules in the atmosphere.
Proc Roy Soc Lond A: 236, 187, 1956.
Dunkerton, T., On the mean meridional mass motions of the stratosphere and
mesosphere. J Atmos Sci: 35, 2325, 1978.
Dunkerton, T., C.P.F. Hsu, and M.E. McIntyre, Some Eulerian and Lagrangian
diagnostics for a model stratospheric warming. J Atmos Sci: 38, 819, 1981.
Dunkerton, T.J., The role of gravity waves in the quasi-biennial oscillation. J Geophys
Res: 102, 26,053, 1997.
Dunkerton, T. J., and D. P. Delisi, Climatology of the equatorial lower stratosphere.
J Atmos Science: 42, 376, 1985.
Elkins, J.W., D.W. Fahey, J.M. Gilligan, G.S. Dutton, T.J. Baring, C.M. Volk, R.E.
Dunn, R.C. Myers, S.A. Montzka, P.R. Wamsley, A.H. Hayden, J.H. Butler, T.M.
Thompson, T.H. Swanson, E.J. Dlugokencky, P.C. Novelli, D.F. Hurst, J.M. Lobert,
S.J. Ciciora, R.J. McLaughlin, T.L. Thompson, R.H. Winkler, P.J. Fraser, L.P. Steele,
and M.P. Lucarelli, Airborne gas chromatograph for in situ measurements of long-
lived species in the upper troposphere and lower stratosphere. Geophys Res Lett: 23,
347, 1996.
Ertel, H., Ein neuer hydrodynamischer Wirbelsatz. Meteorol Z: 59, 271, 1942.
Exner, F.M., Sitzungberichte der Mathematisch-Naturwissenschaftlichen Klasse der
Akad. Wissenschaften, 122 Abt. 2a, 2. Halbband 6(10), 1165, 1913.
Feely, H.W., and J.S. Spar, Tungsten-185 from nuclear bomb tests as a tracer for
stratospheric meteorology. Nature: 188, 1062, 1960.
Fels, S.B., J.D. Mahlman, M.D. Schwarzkopf, and R.W. Sinclair, Stratospheric
sensitivity to perturbations in ozone and carbon dioxide: Radiative and dynamical
response. J Atmos Sci: 37,7 2265, 1980.
Fels, S.B., Radiative-dynamical interactions in the middle atmosphere. Adv Geophys:
28A, 277, 1985.
Fleming, E.L., S. Chandra, M.R. Schoeberl, and J.J. Barnett, Monthly mean global
climatology of temperature, wind, geopotential height, and pressure for 0-120 km.
NASA Tech Memo 100687, 85 pp., 1988.
Fleming, E.L., C.H. Jackman, R.S. Stolarski, and D.B. Considine, Simulation of
stratospheric tracers using an improved empirically-based two-dimensional model
transport formulation. J Geophys Res: 104, 23,911, 1999.
Folkins, I., M. Loewenstein, J. Podolske, S. Oltmans, and M. Proffitt, A barrier to
vertical mixing at 14 km in the tropics: Evidence from ozonesondes and aircraft
measurements. J Geophys Res: 104, 22,095, 1999.
Forbes, J.M., Atmospheric tides: 1. Model description and results for the solar diurnal
tides. J Geophys Res: 87,
7 5222, 1982a.
Forbes, J.M., Atmospheric tides: 2. Solar and lunar semidiurnal components. J

Geophys Res: 87,
7 5241, 1982b.
Fritts, D.C., and R.A. Vincent, Mesospheric momentum flux studies at Adelaide,
Australia: Observations of gravity wave/tidal interaction model. J Atmos Sci: 44,
605, 1987.
Fritts, D.C., and W. Lu, Spectral estimates of gravity wave energy and momentum
fluxes, II. Parameterization of wave forcing and variability. J Atmos Sci: 50, 3695,
1993.
Garcia, R.R., and D.L. Hartmann, The role of planetary waves in the maintenance of
the zonally averaged ozone distribution of the upper stratosphere. J Atmos Sci: 37,
7
2248, 1980.
Garcia, R.R., and S. Solomon, A numerical model of the zonally averaged dynamical
and chemical structure of the middle atmosphere. J Geophys Res: 88, 1379, 1983.
Garcia, R.R., and S. Solomon, The effect of breaking gravity waves on the dynamics
and chemical composition of the mesosphere and lower thermosphere. J Geophys Res:
90, 3850, 1985.
Gault, W.A., G. Thuillier G.G. Shepherd, S.P. Zhang, R.H. Wiens, W.E. Ward, C.
Tai, B.H. Solheim, Y.J. Rochon, C. McLandress, C. Lathuillere, V. Fauliot, M. Hersé,
C.H. Hersom, R. Gattinger, L. Bourg, M.D. Burrage, S.J. Franke, G. Hernandez, A.
Manson, R. Niciejewski, and R.A. Vincent, Validation of O(1 S) wind measurements
by WINDII: The wind imaging interferometer on UARS. J Geophys Res: 101, 10,405,
1996.
Gavrilov, N.M., and R.G. Roble, The effect of gravity waves on the global mean
temperature and composition structure of the upper atmosphere. J Geophys Res: 99,
25,773, 1994.
Geller, M.A., and J.C. Alpert, Planetary wave coupling between the troposphere and
the middle atmosphere as a possible sun-weather mechanism. J Atmos Sci: 37,7 1197,
1980.
Geller, M.A., Dynamics of the middle atmosphere. Space Sci Rev: 34, 359, 1983.
Gombosi, T.I., Physics of the Space Environment. Cambridge University Press, 1998.
Graf, H.-F., I. Kirchner, A. Robock, and I. Schult, Pinatubo eruption winter climate
effects: Model versus observations. Climate Dyn: 9, 81, 1993.
Graf, H.-F., J. Perlwitz, and I. Kirchner, Northern hemisphere tropospheric
midlatitute circulation after violent volcanic eruptions. Contrib Atmos Phys: 67,
7
3, 1994.
Grant, W.B., E.V. Browell, C.S. Long, L.L. Stowe, R.G. Grainger, and A. Lambert,
Use of volcanic aerosols to study the tropical stratospheric reservoir. J Geophys Res:
101, 3973, 1996.
Grose, W.L., J.E. Nealy, R.E. Turner, and W.T. Blackshear, Modeling the transport
of chemically active constituents in the stratosphere, in Transport Processes in the
Middle Atmosphere. G. Visconti, and R. Garcia, eds., 229, D. Reidel, Massachusetts,
1987.
Hagan, M.E., Modeling atmospheric tidal propagation across the stratopause, in:
Atmospheric Science Across the Stratopause. Geophysical Monograph: 123, 177,
American Geophysical Union, 2000.
Hall, T.M., and M.J. Prather, Simulations of the trend and annual cycle in
stratospheric CO2 . J Geophys Res: 98, 10,573, 1993.
Hall, T.M., and R.A. Plumb, Age as a diagnostic of stratospheric transport. J

Geophys Res: 99, 1059, 1994.
Hall, T.M., and D.W. Waugh, Timescales for the stratospheric circulation derived
from tracers. J Geophys Res: 102, 8991, 1997.
Hall, T.M., et al., 2. Transport experiments, in Models and Measurements
Intercomparison II.
I J.H. Park, M.K.W. Ko, C.H. Jackman, R.A. Plumb, J.A. Kaye,
and K.H. Sages, eds., NASA/TM-1999-209554, 1999.
Hamilton, K., R.J. Wilson, J.D. Mahlman, and L.J. Umscheid, Climatology of the
GFDL SKYHI general circulation model. J Atmos Sci: 52, 44, 1995.
Hare, F.K., and B.W. Boville, The Polar Circulation. Tech. Note 70, World
Meteorological Organization, Geneva, 1965.
Harwood, R.S., and J.A. Pyle, A two-dimensional mean circulation model for the
atmosphere below 80 km. Quart J Roy Meteorol Soc: 101, 723, 1975.
Hartmann, D.L., Some aspects of the coupling between radiation, chemistry, and
dynamics in the stratosphere. J Geophys Res: 86, 9631, 1981.
Hasebe, F., Quasi-biennial oscillations of ozone and diabatic circulation in the
equatorial stratosphere. J Atmos Sci: 51, 729, 1994.
Hauchecorne, A., M.L. Chanin, and R. Wilson, Mesospheric temperature inversion
and gravity wave breaking. Geophys Res Lett: 14, 933, 1987.
Haynes, P.H., C.J. Marks, M.E. McIntyre, T.G. Shepherd, and K.P. Shine, On the
“downward control” of extratropical diabatic circulation by eddy-induced mean zonal
forces. J Atmos Sci: 48, 651, 1991.
Hays, P.B., V.J. Abreu, M.E. Dobbs, D.A. Gell, H.J. Grassl, and W.R. Skinner,
The high-resolution Doppler imager on the Upper Atmosphere Research Satellite. J
Geophys Res: 98, 10,713, 1993.
Hays, P.B., D.L. Wu, and the HRDI Science Team, Observations of the diurnal tide
from space. J Atmos Sci: 51, 3077, 1994.
Hines, C.O., Doppler spread parameterization of gravity wave momentum deposition
in the middle atmosphere, Part 1: Basic formulation. J Atmos Solar Terr Phys: 59,
371, 1997a.
Hines, C.O., Doppler spread parameterization of gravity wave momentum deposition
in the middle atmosphere, Part 2: Broad and qausi-monochromatic spectra and
implementation. J Atmos Solar Terr Phys: 59, 387, 1997b.
Hitchman, M.H., C.B. Leovy, J.C. Gille, and P.L. Bailey, Quasi-stationary zonally
assymmetric circulations in the equatorial lower mesosphere, J Atmos Sci: 44, 2219,
1987.
Holton, J.R., and R.S. Lindzen, An updated theory for the quasi-biennial cycle of the
tropical stratosphere. J Atmos Sci: 29, 1076, 1972.
Holton, J.R., A semi-spectral numerical model for wave, mean flow interactions in
the stratosphere: Application to sudden stratospheric warmings. J Atmos Sci: 33,
1639, 1976.
Holton, J.R., and W.M. Wehrbein, The role of forced planetary waves in the annual
cycle of the zonal mean circulation of the middle atmosphere. J Atmos Sci: 37,
7 1968,
1980.
Holton, J.R., The role of gravity wave induced drag and diffusion in the momentum
budget of the mesosphere. J Atmos Sci: 39, 791, 1982.
Holton, J.R., The influence of gravity wave breaking on the general circulation of the
middle atmosphere. J Atmos Sci: 40, 2497, 1983.
Holton, J.R., and M.R. Schoeberl, The role of gravity wave generated advection and
diffusion in transport of tracers in the mesosphere. J Geophys Res: 93, 11,075, 1988.
Holton, J.R., and J. Austin, The influence of equatorial QBO on sudden stratospheric
warmings. J Atmos Sci: 48, 607, 1991.
Holton, J.R., Introduction to Dynamic Meteorology. 3rd ed. Academic Press, 1992.
Holton, J.R., P.H. Haynes, M.E. McIntyre, A.R. Douglass, R.B. Rood, and L. Pfister,
Stratosphere-troposphere exchange. Rev Geophys: 33, 403, 1995.
Holton, J.R., and M.J. Alexander, The role of waves in the transport circulation of
the middle atmosphere, in: Atmospheric Science Across the Stratopause, Geophys
Monograph: 123, 21, American Geophysical Union, 2000.
Hong, S.-S., and R.S. Lindzen, Solar semi-diurnal tide in the thermosphere. J Atmos
Sci: 33, 135, 1976.
Hoppel, K.W., K.P. Bowman, and R.M. Bevilacqua, Northern hemisphere summer
ozone variability observed by POAMII. Geophys Res Lett: 26, 827, 1999.
Hoskins, B.J., M.E. McIntyre, and A.W. Robertson, On the use and significance of
isentropic potential vorticity maps. Quart J Roy Meteorol Soc: 111, 877, 1985.
Huang, T.Y.W., and A.K. Smith, The mesospheric diabatic circulation and the
parameterized thermal effect of gravity wave breaking on the circulation. J Atmos
Sci: 48, 1093, 1991.
Hunt, B.G., and S. Manabe, Experiments with a stratospheric general circulation
model, II. Large scale diffusion of tracers in the stratosphere. Mon Weath Rev: 96,
503, 1968.
Hunt, B., Experiments with a stratospheric general circulation model, III. Large scale
diffusion of ozone including photochemistry. Mon Weath Rev: 97, 7 287, 1969.
James, I.N., Introduction to Circulatory Atmospheres, Cambridge University Press,
1994.
Juckes, M.N., and M.E. McIntyre, A high-resolution, one layer model of breaking
planetary waves in the stratosphere. Nature: 328, 590, 1987.
Juckes, M., A generalization of the transformed Eulerian-mean meridional circulation.
Quart J Roy Meteorol Soc: 127, 7 147, 2001.
Julian, P.R., and K.B. Labitzke, A study of atmospheric energetics during the
January-February 1963 stratospheric warming. J Atmos Sci: 22, 597, 1965.
Kasting, J.F., and R.G. Roble, A zonally averaged chemical-dynamical model of the
lower thermosphere. J Geophys Res: 86, 964l, 1981.
Kawahira, K., A two-dimensional model for ozone changes by planetary waves in the
stratosphere, I. Formulation and the effect of temperature waves on the zonal mean
ozone concentration. J Meteorol Soc Japan: 60, 1058, 1982.
Kida, H., General circulation of air parcels and transport characteristics derived from
a hemispheric GCM, Part 1. A determination of advective mass flow in the lower
stratosphere. J Meteorol Soc Japan: 61, 171, 1983.
Kitoh, A., Y. Noda, Y. Nikaidou, T. Ose, and T. Tokioka, AMIP simulations of the
MRI GCM. Pap Meteor Geophys: 45, 121, 1995.
Kodera, K., and K. Yamazaki, Long-term variation of upper stratospheric circulation
in the Northern hemisphere in December. J Meteorol Soc Japan: 68, 101, 1990.
Kodera, K., M. Chiba, K. Yamazaki, and K. Shibata, A possible influence of the polar
night stratospheric jet on the subtropical tropospheric jet. J Meteorol Soc Japan: 69,
715, 1991.
Kodera, K., Influence of volcanic eruptions on the troposphere through stratospheric

dynamical processes in the Northern hemisphere winter. J Geophys Res: 99, 1273,
1994.
Kodera, K., and K. Yamazaki, A possible influence of recent polar stratospheric
coolings on the troposphere in the Northern hemisphere winter. Geophys Res Lett:
21, 809, 1994.
Kodera, K., M. Chiba, H. Koide, A. Kitoh, and Y. Nikaidou, Interannual variability
of the winter stratosphere and troposphere in the Northern hemisphere. J Meteorol
Soc Japan: 74, 365, 1996.
Kodera, K., and Y. Kuroda, Tropospheric and stratospheric aspects of the Arctic
oscillation. Geophys Res Lett: 27,
7 3349, 2000a.
Kodera, K., and Y. Kuroda, A mechanistic model study of slowly propagating coupled
stratosphere-troposphere variability. J Geophys Res: 105, 12,361, 2000b.
Lait, L., An alternative form of potential vorticity. J Atmos Sci: 12, 1754, 1994.
Labitzke, K., Climatology of the stratosphere and mesosphere. Philos Trans R Soc
London: 296, 7, 1980.
Labitzke, K., and H. van Loon, Association between the 11-year solar cycle, the QBO
and the atmosphere, Part I: The troposphere and the stratosphere in the northern
hemisphere in winter. J Atmos Terr Phys: 50, 197, 1988.
Langematz, U., and S. Pawson, The Berlin troposphere-stratosphere-mesosphere
GCM: Climatology and annual cycle. Quart J Roy Meteorol Soc: 123, 1075, 1997.
Lefevre,
f
f` F., G.P. Brasseur, I. Folkins, A.K. Smith, and P. Simon, The chemistry of
the 1991-92 stratospheric winter: Three-dimensional model simulation. J Geophys
Res: 99, 8183, 1994.
Leovy, C., Simple models of thermally driven mesospheric circulation. J Atmos Sci:
21, 327, 1964.
Leovy, C., Photochemical de-stabilization of gravity wave near the mesopause. J
Atmos Sci: 23, 223, 1966.
Levy, H., J.D. Mahlman, and W.J. Moxim, A preliminary report on the numerical
simulation of the three-dimensional structure and variability of atmospheric N2 O.
Lieberman, R.S.,A.K. Smith, S.J. Franke, R.A. Vincent, J.R. Isler, A.H. Manson, C.E.
Meek, G.J. Fraser, A. Fahrutdinova, T. Thayaparan, W. Hocking, J. MacDougall,
T. Nakamura, and T. Tsuda, Comparison of mesospheric and lower thermospheric
residual wind with High Resolution Doppler Imager, medium frequency, and meteor
radar winds. J Geophys Res: 105, 27,023, 2000.
Lin, S.J., and R.B. Rood, Multidimensional flux-form semi-Lagrangian transport
schemes. Mon Weath Rev: 124, 2046, 1996.
Lindzen, R.S., Thermally driven diurnal tide in the atmosphere. Quart J Roy Meteorol
Soc: 93, 18, 1967.
Lindzen, R.S., and J.R. Holton, A theory of the quasi-biennial oscillation. J Atmos
Sci: 22, 341, 1968.
Lindzen, R.S., and D. Blake, Mean heating of the thermosphere by tides. J Geophys
Res: 75, 6868, 1970.
Lindzen, R.S., Turbulence and stress owing to gravity wave and tidal breakdown. J
Geophys Res: 86, 9707, 1981.
Liu, H.L., and M.E. Hagan, Local heating/cooling of the mesosphere due to gravity
wave and tidal coupling. Geophys Res Lett: 25, 2941, 1998.
Loewenstein, M., J.R. Podolske, K.R. Chan, and S.E. Strahan, Nitrous oxide as a
dynamical tracer in the 1987 Airborne Antarctic Ozone Experiment. J Geophys Res:
94, 11,598, 1989.
London, J.L., Radiative energy sources and sinks in the stratosphere and mesosphere,
Proc NATO Advanced Study Institute on Atmospheric Ozone FAA-EE-80-20, NTIS,
Springfield, Va., 1980.
Mahlman, J.D., and W.J. Moxim, Tracer simulation using a global general circulation
model: Results from a mid-latitude instantaneous source experiment. J Atmos Sci:
35, 1340, 1978.
Mahlman, J.D., and L.J. Umscheid, Dynamics of the middle atmosphere: Successes
and problems of the GFDL “SKIHI” general circulation model, in Proc of the U.S.-
Japan Seminar on Middle Atmosphere Dynamics, Terra Scientific Pub., Tokyo, 1983.
Makhlouf, U.B., R.H. Picard, and J.R. Winick, Photochemical-dynamical modeling
of the measured response of airglow to gravity waves, 1. Basic model for OH airglow.
J Geophys Res: 100, 11,289, 1995.
Makhlouf, U.B., R.H. Picard, J.R. Winick, and T.F. Tuan, A model for the response
of the atomic oxygen 557.7 nm and the OH Meinal airglow to atmospheric gravity
waves in a realistic atmosphere. J Geophys Res: 103, 6261, 1998.
Manabe, S., and B.G. Hunt, Experiments with a stratospheric general circulation
model, I. Radiative and dynamic aspects. Mon Weath Rev: 96, 477, 1968.
Manzini, E., N.A. McFarlane, and C. McLandress, Impact of the Doppler-spread
parameterization on the simulation of the middle atmosphere circulation using the
MA/ECHAM4 general circulation model. J Geophys Res: 102, 25,751, 1997.
Matsuno, T., Vertical propagation of stationary planetary waves in the Northern
hemisphere. J Atmos Sci: 27,
7 871, 1970.
Matsuno, T., A dynamical model of the stratospheric sudden warming. J Atmos Sci:
28, 1479, 1971.
Matsuno, T., and K. Nakamura, The Eulerian and Lagrangian mean meridional
circulations in the stratosphere at the time of a sudden warming. J Atmos Sci:
36, 640, 1979.
Matsuno, T., Lagrangian motion of air parcels in the stratosphere in the presence of
planetary waves. Pure Appl Geophys: 118, 189, 1980.
Matsuno, T., Quasi-geostrophic motions in the equatorial area. J Meteorol Soc Japan:
44, 25, 1986.
McIntyre, M.E., Towards a Lagrangian mean description of stratospheric circulations
and chemical transports. Philos Trans R Soc London: 296, 129, 1980.
McIntyre, M.E., and T.N. Palmer, Breaking planetary waves in the stratosphere.
Nature: 305, 593, 1983.
McIntyre, M.E., and T.N. Palmer, The “surf zone” in the stratosphere. J Terr Phys:
46, 825, 1984.
McIntyre, M.E., Middle atmosphere dynamics and transport: Some current challenges
to our understanding, in Dynamics, Transport, and Photochemistry of the Southern
Hemisphere, A. O’Neill (ed.), Kluwer, Dordrecht, 1990.
McLandress, C., G.S. Shepherd, B.H. Solheim, M.D. Burrage, P.B. Hays, and W.R.
Skinner, Combined mesosphere/thermosphere winds using WINDII and HRDI data
from the Upper Atmosphere Research Satellite. J Geophys Res: 101, 10,441, 1996.
Mechoso, C.R., K. Yamazaki, A. Kitoh, and A. Arakawa, Numerical forecasts of

atmospheric warming events during the winter of 1979. Mon Weath Rev: 113, 1015,
1985.
Medvedev, A.S., and G.P. Klaassen, Vertical evolution of gravity wave spectra and
the parameterization of associated wave drag. J Geophys Res: 100, 25,841, 1995.
Mote, P.W., K.H. Rosenlof, J.R. Holton, R.S. Harwood, and J.W. Waters, Seasonal
variations of water vapor in the tropical lower stratosphere. Geophys Res Lett: 22,
1093, 1995.
Murgatroyd, R.J., and F. Singleton, Possible meridional circulations in the
stratosphere and mesosphere. Quart J Roy Meteorol Soc: 87,
7 125, 1961.
Murgatroyd, R.J., Dynamical modelling of the stratosphere and mesosphere, in
Mesospheric Models and Related Experiments. G. Fiocco, ed., Reidel Publishing Co.,
Dordrecht, 1971.
Murphy, D.M., and D.W. Fahey, An estimate of the flux of stratospheric reactive
nitrogen and ozone into the troposphere. J Geophys Res: 99, 5325, 1994.
Nastrom, G.D., B.B. Balsley, and D.A. Carter, Mean meridional winds in the mid
and high latitude mesosphere. Geophys Res Lett: 9, 139, 1982.
Naujokat, B., and K. Labitzke (eds.), Collection of Reports on the Stratospheric
Circulation during Winters 1974-75 to 1991-92. Report 1, Sci Comm on Sol Terr
Phys, University of Illinois, 1993.
Newman, P.A., L.R. Lait, M.R. Schoeberl, M. Seablom, L. Coy, R. Rood, R.
Swinbank,M. Profitt, M. Loewenstien, J.R. Podolske, J.W. Elkins, C.R. Webster,
R.D. May, D.W. Fahey, G.S. Dutton and K.R. Chan, Measurement of polar vortex
air in the midlatitudes. J Geophys Res: 101, 12,879, 1996.
O’Neill, A., Dynamical processes in the stratosphere: Wave motion, in Proc NATO
Advanced Study Institute on Atmospheric Ozone, FAA-EE-80-20, NTIS, Springfield,
Va., 1980.
Ortland, D.A., W.R. Skinner, P.B. Hays, M.D. Burrage, R.S. Lieberman, A.R.
Marshall, and D.A. Gell, Measurement of stratospheric winds by the high resolution
Doppler imager. J Geophys Res: 101, 10,351, 1996.
Ortland, D.A., P.B. Hays, W.R. Skinner, and J.-H. Yee, Remote sensing of mesosphere
temperature and O2 (1 Σ) band volume emission rates with the higher resolution
Doppler imager. J Geophys Res: 103, 1821, 1997.
O’Sullivan, D., and R.E. Young, Modeling the quasi-biennial oscillation effect on the
winter stratospheric circulation. J Atmos Sci: 49, 2437, 1992.
Palmer, T. N., Diagnostic study of a wavenumber 2 stratospheric sudden warming in
a transformed Eulerian mean formalism. J Atmos Sci: 38, 544, 1981.
Pawson, S., K. Kodera, K. Hamilton, T.G. Shepherd, S.R. Beagley, B.A. Boville,
J.D. Farrara, T.D.A. Fairlie, A. Kitoh, W.A. Lahoz, U. Langematz, E. Manzini, D.H.
Rind, A.A. Scaife, K. Shibata, P. Simon, R. Swinbank, L. Takacs, R.J. Wilson, J.A.
Al-Saadi, M. Amodei, M. Chiba, L. Coy, J. de Grandpré, ´ R.S. Eckman, M. Fiorino,
W.L. Grose, H. Koide, J.N. Koshyk, D. Li, J. Lerner, J.D. Mahlman, N.A. McFarlane,
C.R. Mechoso, A. Molod, A. O’Neill, R.B. Pierce, W.J. Randel, R.B. Rood, and F.
Wu, The GCM-reality intercomparison project for SPARC (GRIPS): Scientific issues
and initial results. Bull Amer Meteorol Soc: 81, 781, 2000.
Pedlosky, J., Geophysical Fluid Dynamics, Springer-Verlag, 1987.
Perlwitz, J., and H.-F. Graf, The statistical connection between tropospheric and
stratospheric circulation of the Northern hemisphere in winter. J Clim: 8, 2281,
1995.
Perlwitz, J., H.-F. Graf, and R. Voss, The leading variability mode of the coupled
troposphere-stratosphere winter circulation in different climate regimes. J Geophys
Res: 105, 6915, 2000.
Pitari, G.S., S. Palermi, G. Visconti, and R. Prinn, Ozone response to a CO2 doubling:
Results from a stratospheric circulation model with heterogeneous chemistry. J
Geophys Res: 97, 7 5953, 1992.
Plumb, R.A., and A.D. McEwan, The instability of a forced standing wave in a viscous
stratified fluid: A laboratory analogue of the quasi-biennial oscillation. J Atmos Sci:
35, 1827, 1978.
Plumb, R.A., Eddy fluxes of conserved quantities by small amplitude waves. J Atmos
Sci: 36, 1699, 1979.
Plumb, R.A., and M.K.W. Ko, Interrelationships between mixing ratios of long-lived
stratospheric constituents. J Geophys Res: 97,
7 10,145, 1992.
Plumb, R.A., A “tropical pipe” model of stratospheric transport. J Geophys Res:
101, 3957, 1996.
Pollock, W.A., L.E. Heidt, R.A. Lueb, J.F. Vedder, M.J. Mills, and S. Solomon, On
the age of stratospheric air and ozone depletion potentials in the polar regions. J
Geophys Res: 97,7 12,993, 1992.
Prather, M.J., Numerical advection by conservation of second order moments. J
Geophys Res: 91, 6671, 1986.
Pyle, J. A., and C. F. Rogers, Stratospheric transport by stationary planetary waves
— The importance of chemical processes. Quart J Roy Meteorol Soc: 106, 421, 1980.
Randel, W.J., Global Atmospheric Circulation Statistics, 1000-1 mb. NCAR Tech
Note, TN-366+STR, 1992.
Randel, W.J., J.C. Gille, A.E. Roche, J.B. Kumer, J.L. Mergenthaler, J.W. Waters,
E.F. Fishbein, and W.A. Lahoz, Stratospheric transport from the tropics to middle
latitudes by planetary-wave mixing. Nature: 365, 533, 1993.
Rasch, P.J., B.A. Boville, G.P. Brasseur, A three-dimensional general circulation
model with coupled chemistry for the middle atmosphere. J Geophys Res: 100, 9041,
1995.
Reed, R.J., W.J. Campbell, L.A. Rasmusson, and D.G. Rogers, Evidence of a
downward propagating annual wind reversal in the equatorial stratosphere. J Geophys
Res: 66, 813, 1961.
Reed, R.J., J.L. Wolfe, and H. Nishimoto, A spectral analysis of the energetics of the
stratospheric sudden warming of early l957. J Atmos Sci: 20, 256, 1963.
Reed, R.J., The quasi-biennial oscillation of the atmosphere between 30 and 50 km
over Ascension Island. J Atmos Sci: 22, 331, 1965.
Reed, R.J., and K.E. German, A contribution to the problem of stratospheric diffusion
by large scale mixing. Mon Weath Rev: 93, 313, 1965.
Richmond, A.D., Thermospheric dynamics and electrodynamics, in Solar Terrestrial
Physics, R.L. Corovilano and J.M. Forbes, eds. D. Reidel, Dordrecht, 1983.
Rind, D., R. Suozzo, N.K. Balachandran, A. Lacis, and G. Russell, The GISS global
climate-middle atmosphere model, Part I: Model structure and climatology. J Atmos
Sci: 45, 329, 1988.
Rind, D., D. Shindell, P. Lonergan, and N.K. Balachandran, Climate change in the
middle atmosphere, Part III. The doubled CO2 climate revisited. J Clim: 11, 876,
1998.
Robert, A., A stable numerical integration scheme for the primitive meteorological
equations. Atmos Ocean: 19, 35, 1981.
Rood, R.B., and M.R. Schoeberl, A mechanistic model of Eulerian, Lagrangian mean,
and Lagrangian ozone transport by steady planetary waves. J Geophys Res: 88, 5208,
1983.
Rood, R.B., Numerical advection algorithms and their role in atmospheric transport
and chemistry models. Rev Geophys: 25, 71, 1987.
Rose, K., and G. P. Brasseur, A three-dimensional model of chemically active trace
species in the middle atmosphere during disturbed winter conditions. J Geophys Res:
94, 16,387, 1989.
Rossby, C.-G., Planetary flow patterns in the atmosphere. Quart J Roy Meteorol Soc:
66, Suppl 68, 1940.
Sassi, F., R.R. Garcia, and B.A. Boville, The stratopause semiannual oscillation in
the NCAR Community Climate Model. J Atmos Sci: 50, 3608, 1993.
Schlesinger, M.E., and Y. Mintz, Numerical simulation of ozone production, transport
and distribution with a global atmospheric general circulation model. J Atmos Sci:
36, 1325, 1979.
Schmidt, U., and A. Khedim, In situ measurements of carbon dioxide in the winter
Arctic vortex and at mid-latitude: An indicator of the age of stratospheric air.
Schoeberl, M.R., and D.F. Strobel, The zonally averaged circulation of the middle
atmosphere. J Atmos Sci: 35, 577, 1978.
Shepherd, T.G., The middle atmosphere. J Atmos Sol-Terr Phys: 62, 1587, 2000.
Shepherd, T.G., Large-scale atmospheric dynamics for atmospheric chemists. Cehm
Rev: 103, 4509, 2003.
Shepherd, G.G., G. Thuillier, W.A. Gault, B.H. Solheim, C. Hersom, J.M. Alunni,
J.-F. Brun, S. Brune, P. Charlot, L.L. Cogger, D.-L. Desaulniers, W.F.J. Evans, R.L.
Gattinger, F. Girod, D. Harvie, R.H. Hum, D.J.W. Kendall, E.J. Llewellyn, R.P.
Lowe, J. Ohrt, F. Pasternak, O. Peillet, I. Powell, Y. Rochon, W.E. Ward, R.H.
Wiens, and J. Wimperis, WINDII, the wind imaging interferometer on the upper
atmosphere research satellite. J Geophys Res: 98, 10,725, 1993.
Shindell, D.T., G.A. Schmidt, M.E. Mann, D. Rind, and A. Waple, Solar forcing of
regional climate change during the Maunder minimum. Science: 294, 2149, 2001.
Simmons, A.J., and R. Strüfing,
¨ Numerical forecasts of stratospheric warming events
using a model with a hybrid vertical coordinate. Quart J Roy Meteorol Soc: 109, 81,
1983.
Smagorinsky, J., S. Manabe, and J.L. Holloway, Numerical results from a nine-level
general circulation model of the atmosphere. Mon Weath Rev: 93, 727, 1965.
Smith, A.K., Physics and chemistry of the mesopause region. J Atm Sol Terr Phys:
66, 839, 2004.
Smolarkiewicz, P.K., A simple positive definite advection scheme with small implicit
diffusion. Mon Weath Rev: 111, 479, 1983.
Staniforth, A., and J. Coté,
´ Semi-Lagrangian integration schemes for middle
atmosphere models—A review. Mon Weath Rev: 119, 2206, 1991.
Strobel, D.F., Parameterization of linear wave chemical transport in planetary
atmospheres by eddy diffusion. J Geophys Res: 86, 9806, 1981.
Strobel, D.F., J.P. Apruzese, and M.R. Schoeberl, Energy balance constraints on
gravity wave induced eddy diffusion in the mesophere and lower thermosphere. J
Geophys Res: 90, 13,067, 1985.
Strobel, D.F., Constraints on gravity wave induced diffusion in the middle
atmosphere. Pure Appl Geophys: 130, 533, 1989.
Swenson, G.R., and C.S. Gardner, Special section: The 1993 airborne lidar and
observations of Hawaiian airglow airborne noctilucent cloud campaigns–preface. J
Geophys Res: 103, 6249, 1998.
Swinbank, R., W.A. Lahoz, A. O’Neill, C.S. Douglas, A. Heaps, and D. Podd, Middle
atmosphere variability in the UK Meteorological Office Unified Model. Quart J Roy
Meteorol Soc: 124, 1485, 1998.
Thompson, D.W.J., and J.M. Wallace, The Arctic oscillation signature in the winter
geopotential height and temperature fields. Geophys Res Lett: 25, 1297, 1998.
Thompson, D.W.J., and J.M. Wallace, Annular modes in the extratropical circulation,
Part I. Month-to-month variability. J Clim: 13, 1000, 2000.
Thompson, D.W.J., J.M. Wallace, and G.C. Hegerl, Annular modes in the
extratropical circulation, Part II. Trends. J Clim: 13, 1018, 2000.
Trepte, C.R., and M.H. Hitchman, Tropical stratospheric circulation deduced from
satellite aerosol data. Nature: 335, 626, 1992.
U.S. Standard Atmosphere 1976. National Oceanic and Atmospheric Administration,
National Aeronautics and Space Administration, and United States Air Force,
Washington, D.C., 1976.
Veryand, R.G., and R.A. Ebdon, Fluctuations in tropical stratospheric winds.
Meteorol Mag: 90, 125, 1961.
Vincent, D.G., Meridional circulation in the northern hemisphere lower stratosphere
during 1964 and 1965. Quart J Roy Meteorol Soc: 94, 333, 1968.
Volk, C.M., J.W. Elkins, D.W. Fahey, R.J. Salawitch, G.S. Dutton, J.M. Gilligan,
M.H. Proffitt, M. Loewenstein, J.R. Podolske, K. Minschwaner, J.J. Margitan, and
K.R. Chan, Quantifying transport between the tropical and mid-latitude stratosphere.
Science: 272, 1763, 1996.
von Zahn, U., J. Friedler, B. Naujokat, V. Langematz, and K. Krüger,¨ A note on
record-high temperature at the northern polar stratopause in winter 1997-98. Geophys
Res Lett: 25, 4169, 1998.
Vupputuri, R.K., The structure of the natural stratosphere and the impact of
chlorofluoromethanes on the ozone layer investigated in a 2-D time dependent model.
Pure Appl Geophys: 117,
7 448, 1979.
Wagner, R.E., and K.P. Bowman, Wavebreaking and mixing in the Northern
hemisphere summer stratosphere. J Geophys Res: 105, 24,799, 2000.
Walker, G.T., Correlations in seasaonal variations of weather. Mem Ind Meteorol
Dept: 25, 275, 1924.
Walker, G.T., and E.W. Bliss, World weather, Mem Roy Meteorol Soc: 4, 53, 1932.
Wallace, J.M., General circulation of the tropical lower stratosphere. Rev Geophys
Space Phys: 11, 191, 1973.
Wallace, J.M., and D.W.J. Thompson, Annular modes and climate prediction. Phys
Today: 28, 2002.
Walterscheid, R.-L., G. Schubert, and J.M. Straus, A dynamical-chemical model of
wave-driven fluctuations in the OH nightglow. J Geophys Res: 92, 1241, 1987.
Waugh, D.W., and R.A. Plumb, Contour advection with surgery: A technique for
investigating fine scale structure in tracer transport. J Atmos Sci: 51, 530, 1994.
Waugh, D.W., R.A. Plumb, R.J. Atkinson, M.R. Schoeberl, L.R. Lait, P.A. Newman,
M. Loewenstein, D.W. Toohey, L.M. Avallone, C.R. Webster, and R.D. May,
Transport out of the lower stratospheric Arctic vortex by Rossby wave breaking.
J Geophys Res: 99, 1071, 1994.
Weinstock, J., Nonlinear theory of gravity waves: Momentum deposition, generalized
Rayleigh friction, and diffusion. J Atmos Sci: 39, 1698, 1982.
World Meteorological Organization (WMO), Scientific Assessment of Ozone
Depletion: 1998, Global Ozone Research and Monitoring Project- Report No. 44,
Geneva, 1999.
Xu, J., The influence of photochemistry on gravity waves in the middle atmosphere.
Earth Planets Space: 51, 855, 1999.
Xun, Z., and J.R. Holton, Photochemical damping of inertio-gravity waves. J Atmos
Sci: 43, 2578, 1986.
Chapter 4
RADIATION
4.1 Introduction
The atmosphere can be considered to be a mixture of gases and

particles exposed to the electromagnectic energy of the sun. An
understanding of the dynamical and photochemical processes that occur
in this environment requires evaluation of atmospheric radiative transfer.
For example, the rate of reaction between two constituents often
depends on the local temperature (see Chapter 2), which is affected by
absorption, scattering, and emission of solar and terrestrial radiation.
Further, solar radiation of particular energies can dissociate and ionize
atmospheric molecules, producing reactive ions and radicals which, in
turn, participate in many important chemical processes.
Radiation is characterized by a spectrum of energies. Spectral
variables can be wavelength λ, frequency ν = c/λ (where c = 3 ×
108 m s−1 is the speed of light in vacuum), or wavenumber ν = 1/λ
(see Figure 4.1). The energy E of a photon is given by
hc
E = hν = = hcν (4.1)
λ
where h = 6.626 × 10−34 J s is the Planck constant. Table 4.1 presents
the nomenclature associated with various parts of the electromagnetic
spectrum. In the region of X-rays, ultraviolet (UV), and visible,
wavelength λ is usually expressed in nanometers (1 nm = 10−9 m), while,
in the infrared (IR), λ is often expressed in micrometers or microns
(1 µm = 10−6 m). Radiation frequency is commonly expressed in s−1
or Hz and wavenumber in cm−1 . Table 4.1 also provides typical values
for the magnitude of solar variability in each spectral region, and brief
comments on how radiation interacts with the atmosphere.
As discussed in Chapter 2, atmospheric molecules are characterized by
discrete rotational and vibrational energy states, and by their electronic
configurations (see Figure 4.1). The absorption of photons can induce
transitions between ground and excited states. The difference between
rotational energy levels are much smaller than those of the vibrational
151
Table 4.1 Regions of the Electromagnetic Spectrum (Thomas and Stamnes,

1999)
Subregion Range Solar Comments

Variability
X rays λ < 10 nm 10-100% Photoionizes all thermo-
sphere species.
Extreme UV 10 < λ < 100 nm 50% Photoionizes O2 and N2 .
Photodissociates O2 .
Far UV 100 < λ < 200 nm 7-80% Dissociates O2 . Discrete
electronic excitation of
atomic resonance lines.
Middle UV, 200 < λ < 280 nm 1-2% Dissociates O3 in intense
or UV-C Hartley bands. Potent-
ially lethal to biosphere.
UV-B 280 < λ < 320 nm < 1% Some radiation reaches
surface, depending on
O3 optical depth.
Responsible for skin
erythema.
UV-A 320 < λ < 400 nm <1% Reaches surface. Benign
to humans. Scattered
by clouds, aerosols, and
molecules.
Visible 400 < λ < 700 nm ≤ 0.1% Absorbed by ocean, land.
Scattered by clouds,
aerosols, and molecules.
Primary energy source
for biosphere and
climate system.
Near IR 0.7 < λ < 3.5 µm Absorbed by O2 , H2 O,
CO2 in discrete
vibrational bands.
Thermal IR 3.5 < λ < 100 µm Emitted and absorbed by
surfaces and IR-
active gases.
states. The separation between electronic levels is in general much

larger than the separation between vibrational levels. For processes
which do not result in continuum ionization or photodissociation of the
molecule, only the photons of the specific frequencies corresponding to
transitions between energy states of the molecule can be absorbed; the
corresponding spectrum therefore appears as a series of distinct lines
which are often part of absorption bands.
Due to the relatively high energies needed to dissociate or ionize
atmospheric gases, photochemistry is mainly initiated by less than
one per cent of the solar photons, more specifically, those whose
characteristic wavelength is in the X-ray, ultraviolet, or, for certain
molecules, in the visible region.
RADIATION 153
Figure 4.1. Spectral regions and their effect on molecules: from left to right:
ionization, dissociation, vibration, and rotation.
Solar radiation also produces thermal effects. Solar energy is absorbed

primarily by ozone in the stratosphere and mesosphere, and by molecular
oxygen in the upper mesosphere and lower thermosphere. Photolysis
of ozone yields some thermal energy directly (through formation of
translationally hot products), along with some chemical energy. The
latter is converted to thermal energy through chemical reactions (e.g.,
recombination of atomic oxygen in the presence of a third body) which
follow the photodissociation of O2 and O3 . CO2 also plays an important
role, rivaling O3 and O2 in the mid-mesosphere by absorbing solar
radiation in the near-infrared bands at 2.0, 2.7, and 4.3 µm. This heating
by absorption of solar ultraviolet radiation is balanced by cooling related
to infrared emission by CO2 (15 µm band), and, to a limited extent, by
ozone (9.6 µm band) and H2 O (6.7 µm band and rotational bands).
Thus, the thermal structure depends on an equilibrium between these
processes as well as other phenomena such as the production of latent
heat in the troposphere and the dissipation of atmospheric waves in the
mesosphere and thermosphere.
In the study of radiation affecting the middle atmosphere it is
often convenient to distinguish between two distinct spectral regions
(Figure 4.2): (1) wavelengths shorter than 3.5 µm, which represent
radiation of solar origin; and (2) wavelengths longer than 3.5 µm, which
is made of radiation of both terrestrial and atmospheric origin. For
wavelengths longer than 3.5 µm, the solar flux can be neglected relative
to the radiation emitted by the Earth’s surface. This regime is therefore
quite different from shorter wavelengths because of the sharp limit
imposed by the spectral function of the terrestrial emission temperature.
The penetration of solar radiation into the Earth’s atmosphere
depends on atmospheric absorption. Since the absorption coefficients of
Figure 4.2. Spectral distribution of solar and terrestrial radiation. Atmospheric

absorption at different wavelengths. After Iribane and Cho (1980).
minor constituents are functions of wavelength, the penetration depth

varies according to the shape of the absorption spectrum (Figure 4.3).
For wavelengths below 100 nm, radiation is almost completely
absorbed above 100 km by molecular and atomic oxygen and to a lesser
extent by molecular nitrogen. X-rays of wavelength less than 1 nm can,
however, penetrate to the middle atmosphere. These X-rays mainly
photoionize, or indirectly dissociate the major species, N2 and O2 , and
are responsible for certain sporadic ionospheric perturbations. These
aspects of the photochemistry will be discussed in Chapter 7.
At wavelengths longer than 100 nm, solar ultraviolet radiation
photodissociates atmospheric molecules. The solar Lyman-α line at
RADIATION 155
Depth of penetration of solar radiation as a function of wavelength.

Altitudes correspond to an attenuation of 1/e. The principle absorbers and
ionization limits are indicated.
121.6 nm is very intense and is situated in an optical window, that

is, a spectral region where absorption by the atmosphere is relatively
weak. Therefore, it can penetrate into the upper part of the middle
atmosphere, where it effectively dissociates water vapor, carbon dioxide,
and methane. Further, it photoionizes nitric oxide, providing the
principal source of ionization in the D-region (see Chapter 7). Finally,
the Lyman-α line is a significant source of heat in the mid-mesosphere.
At longer wavelengths, the solar spectrum is subdivided into regions of
absorption by the principal absorbing species, O2 and O3 . The first
of these molecules absorbs only the radiation at wavelengths shorter
than about 240 nm. The latter, abundant in the stratosphere, absorbs
primarily between 200 and 300 nm, but also to some extent in the visible
and even in the infrared. Even so, in the visible (wavelengths greater
than 310 nm), the greater part of the solar photons reach the troposphere
and the surface. Thus, in this region the effects of molecular scattering
and cloud and surface albedo must be considered. Table 4.2 indicates
the different spectral regions which must be considered when addressing
the photochemical effects of solar radiation on the neutral atmosphere.
Detailed treaments of radiative transfer in the Earth’s atmosphere
are given by Chandrasekhar (1950), Goody (1964), Liou (1980, 2002),
Andrews et al. (1987), Goody and Yung (1989), and Thomas and
Stamnes (1999). An excellent treatment of non-local thermodynamic
equilibrium (non-LTE) radiative transfer in the atmosphere is provided

by Lopez-Puertas and Taylor (2001).
Table 4.2 Spectral Regions of Photochemical Importance in the Atmosphere
Wavelength Atmospheric Absorbers
121.6 nm Solar Lyman α line, absorbed by O2 in the mesosphere;

no absorption by O3 .
130 to 175 nm O2 Schumann Runge continuum. Absorption by O2 in
the thermosphere. Can be neglected in the mesosphere
and stratosphere.
175 to 200 nm O2 Schumann Runge bands. Absorption by O2 in the
mesosphere and upper stratosphere. Effect of O3 can
be neglected in the mesosphere, but is important in the
stratosphere.
200 to 242 nm O2 Herzberg continuum. Absorption by O2 in the
stratosphere and weak absorption in the mesosphere.
Absorption by the O3 Hartley band is also important;
both must be considered.
200 to ≈310 nm O3 Hartley band. Absorption by O3 in the stratosphere
leading to the formation of O (1 D).
≈310 to 400 nm O3 Huggins bands. Absorption by O3 in the strato-
sphere and troposphere leads to the formation of O (3 P).
400 to 850 nm O3 Chappuis bands. Absorption by O3 in the tropo-
sphere induces photodissociation even at the surface.
4.2 Definitions
Before examining in detail the specific effects of solar and terrestrial

radiation on the middle atmosphere, we present a few definitions relative
to radiative transfer in the atmosphere.
The radiation field is defined by the monochromatic radiance Lν ,
which represents the amount of energy dE in frequency interval dν
traversing a surface dS per time dt in a pencil of solid angle dω inclined
at an angle θ relative to the normal to the surface (see Figure 4.4a). We
can then write
dE
Lν (r, ω ) = (4.2)
dν dt dS dω cos θ
This quantity is often expressed in Wm−2 sr−1 Hz−1 , and is defined at
each point r in space and for each direction ω of the pencil of light.
RADIATION 157
Figure 4.4. (a) Geometry of a pencil of light in solid angle dω traversing a

unit of surface dS . θ is the inclination of the beam relative to the surface. ((b))
Definition of the azimuthal (φ) and zenith (θ) angles. (c) Balance of energy
fluxes for a thin layer between altitude z and z + δz.
If Lν is not a function of direction, the field of radiance is said to be

isotropic; if Lν is not a function of position in space, the field is said to
be homogeneous.
In order to evaluate the rate of radiative heating due to absorption
of energy by a unit volume, the net flux density traversing the surface
must be determined. The energy flux density (W m−2 Hz−1 ) across the
surface at angle ω due to all the pencils of light of direction ω , called
the irradiance, is given by the integral:

Fν (r, ω ) = Lν (r, ω ) cos(ω , ω )dω (4.3)
4π
In atmospheric problems, we are most often concerned with the energy
transfer between horizontal layers, and can therefore write:
2π 1
Fν (z) = dφ µLν (µ, φ)dµ (4.4)
0 −1
where µ = cos θ, and φ are the parameters describing the zenith and
azimuthal directions, respectively, of the propagation (see Figure 4.4b).
In the case of a plane parallel atmosphere at altitude z, the net
irradiance can be separated into its upward propagating and downward
propagating components:
2π 1
Fν ↑ (z) = dφ µLν (z; µ, φ)dµ (µ > 0) (4.5)
0 0
2π −1
Fν ↓ (z) = dφ µLν (z; µ, φ)dµ (µ < 0) (4.6)
0 0
such that the net irradiance is given by
Fν (z) = Fν ↓ (z) − Fν ↑ (z) . (4.7)
The radiance L (r, ω ) and irradiance F (r, ω ) over a frequency interval
∆ν is derived by integration over the frequency

L (r, ω ) = Lν (r, ω ) dν (4.8)
∆ν

F (r, ω ) = Fν (r, ω ) dν (4.9)
∆ν
where L (r, ω ) and F (r, ω ) are expressed in W m−2 sr−1 and W m−2 ,
respectively.
To determine the heating rate at the layer between altitude z and z +
δz, the energy balance at each boundary of the layer must be solved (see
Figure 4.4c). As the thickness δz approaches zero, the energy absorbed
per unit volume is given by the net flux divergence dF/dz. Therefore,
the diabatic heating rate (expressed as the variation of the temperature
in the layer per unit time) is given by
−1 dF g dF
Q= = (4.10)
ρcp dz cp dp
RADIATION 159
where cp is the specific heat at constant pressure, ρ is the total air

density, g is the acceleration of gravity and p is the pressure. Q is often
expressed in units of K/s or K/day.
In order to study the photochemical effects of radiation, we define the
actinic flux Φ (r) as the amount of energy (W m−2 Hz−1 ) penetrating a
unit volume of space at point r with

Φν (r) = Lν (r; ω )dω (4.11)
4π
Φ(r) is therefore 4π times the so-called mean radiance. Thus, in

this definition, each photon arriving at point r is given equal weight
regardless of the direction of propagation. The dissociation or ionization
of atmospheric molecules occurs regardless of the direction of the
incident photon, and its value is therefore proportional to the actinic
flux (not the irradiance). In the case of a plane stratified atmosphere,
one obtains:
2π 1
Φν (z) = dφ Lν (z; µ, φ) dµ (4.12)
0 −1
The actinic flux can also be expressed as the number of photons

penetrating a unit volume of space at point r per unit time (photons
m−2 s−1 Hz−1 ). In this case, it is given by
Φν (r)
qν (r) = (4.13)
hν
Note that the spectral density of any quantity X can be expressed not
only as a function of frequency Xν = dX/dν but also as a function of
wavelength Xλ = dX/dλ or of wavenumber Xν = dX/dν where λ = c/ν
and ν = 1/λ = ν/c.
4.3 Extraterrestrial Solar Radiation
4.3.1 The Solar Atmosphere
The production of radiation by the Sun depends strongly on the

physical and chemical structure of the solar atmosphere. Therefore, in
order to understand the nature and variations of the radiation incident
on the Earth, we briefly present some of the important aspects of
the Sun’s atmosphere, whose structure is schematically represented in
Figures 4.5 and 4.6.
Figure 4.5. Schematic diagram of the solar atmosphere with its different layers
and physical characteristics.
The Sun is primarily composed of hydrogen and helium, along with

smaller amounts of heavier elements such as calcium, iron, magnesium,
aluminum, nickel, etc. The temperature in its interior is believed to be as
high as 2×107 K, due to a chain of nuclear reactions which convert H into
He. This energy is radiated to the upper convective levels, undergoing
a series of absorption and emission processes.
Most of the energy reaching the Earth’s atmosphere originates from a
relatively thin layer (about 1000 km thick) called the photosphere. This
layer defines the visible volume of the Sun, and although the entire star is
RADIATION 161
Figure 4.6. Vertical distribution of temperature (T) and density (ρ) at the
Sun’s surface and its atmosphere (left), which determines the solar irradiance
spectrum (right) between 1 and 105 nm. From Lean, 1997.
in the gaseous state, the base of the photosphere is generally considered

to be the “surface” of the Sun. Its effective temperature is about
5800 K. Observations indicate that the brightness of the photosphere
is not continuous, but is characterized by granules which are uniformly
distributed over the solar disk. These granules are probably associated
with strong convective processes in and below the photosphere.
Transient phenomena such as sunspots and faculae appear in the
photosphere, and influence the variability in solar emission at short
wavelengths. Sunspots are relatively dark regions with a temperature of
about 3000 K and a typical diameter of under 50000 km. These spots are
usually grouped in certain places, called active regions on the solar disk.
The lifetime of a sunspot is variable (from a few days to a few months).
Faculae are bright events generally associated with active regions, at
least at middle and low solar latitudes. They also appear independently
of active regions at high solar latitudes. Observations indicate a fairly
regular variation in the occurrence of active regions, with a periodicity
of about 11 years, which is called the solar cycle. The Sun’s rotation on
its own axis takes place on an apparent timescale of 27 days, so that the
exposure of the Earth to the sunspots also has a 27 day cycle.
The emission of the photosphere is a continuum, superimposed with
relatively dark lines called the Fraunhofer lines. These are produced by
selective absorption and reemission of radiation in the upper photosphere
where the temperature is as low as 4000 to 5000 K. Some of these lines
also originate from higher levels, but their effective emission temperature
is lower than the kinetic temperature.
The layer above the photosphere extends to 5000 to 10000 km,
and is called the chromosphere. This layer can sometimes be seen
during total solar eclipses. Its temperature is 105 – 106 K at the
upper levels. Radiation originating from the chromosphere is composed
predominantly of emission lines (H, He, Ca), and the visible emission is
weak.
The region above the chromosphere is called the corona, which extends
outward for several solar diameters. Its temperature is about 106 K.
Several emission lines are associated with this region. Its free electrons
scatter photospheric light.
Active regions in these upper layers are characterized by variations in
chromospheric plages, spicules, prominences, solar flares, and enhanced
coronal emission. Plages are bright areas in the chromosphere, and
are usually observed by their intense calcium K line emission. They
usually precede sunspots and last after the sunspots disappear. Spicules
are small protuberances which appear continuously at the top of the
chromosphere, even in quiet regions. Their lifetime is only a few minutes.
Prominences are large, rather stable clouds of bright gas in the upper
part of (and sometimes above) the chromosphere. Solar flares are intense
eruptive phenomena occurring in the active regions of the chromosphere.
They are accompanied by a rapid increase in brightness and an intense
enhancement in the emission of ionizing radiation (EUV and X-rays).
Flares last from a few minutes to more than an hour, and are divided
into different classes in order of increasing importance to the Earth’s
atmosphere (class 1 or 3). The most intense flares are accompanied
by ejection of large quantities of high energy particles (e.g., protons,
electrons, and alpha particles).
The Sun is also a source of radio waves. These emissions vary with
the solar cycle, and are enhanced (radio burst) during chromospheric
or coronal events. Since these emissions can easily be recorded (e.g., at
10.7 cm or 2.8 GHz), they are often used as an indicator of solar activity.
4.3.2 The Solar Constant
The solar constant is defined as the total radiative energy flux outside
the Earth’s atmosphere. This parameter is used to characterize the total
solar radiation input. The magnitude of this constant is estimated by
Brusa and Frohlich (1982) to be 1367 W m−2 , at 1 AU (astronomical
unit), which corresponds to an effective solar temperature of 5780 K.
RADIATION 163
4.3.3 The Sun as a Black Body
If we assume for the purpose of illustration that the Sun behaves as

a black body, the energy flux ME (W m−2 ) emerging from the surface
of the sun is proportional to the fourth power of the temperature, as
indicated by the Stefan-Boltzmann law:
ME = σT4 (4.14)
where σ = 5.67 × 10−8 W m−2 K−4 . The theoretical value of the solar
constant is obtained by multiplying the solar emission by a dilution
factor βs which accounts for the Earth-Sun distance:
SC = βs ME (4.15)
with
R2s
βs = (4.16)
r2s
Here Rs represents the radius of the Sun and rs the Sun-Earth distance.
Due to the annual variation of rs , the mean dilution factor βs is 2.164 ×
10−5 , while the maximum value attained in January βsmax is 2.235 ×
10−5 and the minimum value occurring in July is βsmin is 2.093 × 10−5 .
The energy input to the Earth from the Sun thus varies annually by
6.6 percent.
The spectral distribution of the exitance emitted by a black body, Mν
is given by Planck’s law:
2πhν 3 1
−2 −1
Mν = W m Hz (4.17a)
c2 ehν/kT − 1
where k = 1.38 × 10−23 J K−1 is Boltzmann’s constant and h = 6.63 ×
10−34 J s is Planck’s constant. It is also often expressed in terms of
wavelength (µm) as
c1 1
Mλ = 5 c /λT W m−2 µm−1 (4.17b)
λ e2 −1
where c1 = 2πhc2 = 3.74 × 108 Wm−2 µm4 and c2 = hc/k = 1.44 ×
104 µm K are the first and second radiation constants, respectively. Thus,
if one approximates the Sun as a black body, the number of photons of
frequency between ν and ν + dν (the actinic flux) impinging on the top
of the atmosphere per unit surface and per unit time is
βs Mν dν 2πβs ν 2 dν
qν dν = = (4.18a)
hν c 2 e hν/kT −1
The corresponding flux of photons between wavelength λ and λ + dλ
(expressed for example in photons cm−2 s−1 µm−1 ) is
c1 βs dλ
qλ dλ = /λT
(4.18b)
4
λ e2 c −1
where c1 = 2πc = 1.88 × 1023 cm−2 s−1 µm3 . The total number of
photons of frequencies between ν1 and ν2 is therefore given by:
ν2
2πβs ν2
q (ν1 − ν2 ) = 2 dν (4.19)
c ehν/kT − 1
ν1
In the ultraviolet where the factor (hν/kT) is much greater than one:
3 2 ν1
2πβs kT −hν/kT hν
q (ν1 − ν2 ) = 2 e +1 +1 (4.20)
c h kT ν2
A better description of the solar spectrum is obtained by adopting an
equivalent temperature which varies with frequency. Such an equivalent
temperature corresponds to the temperature of an effective emission
layer in the solar atmosphere at the frequency. A complete description
has to account for the solar emission lines in the visible, near ultraviolet,
and far ultraviolet regions (see below).
4.3.4 The Observed Solar Spectrum
Since this volume focuses on the terrestrial atmosphere below 100 km,
we shall concern ourselves only with the radiation penetrating to or
below that level (i.e., the ultraviolet, visible and infra-red, as well as
X-rays of wavelengths shorter than 1 nm).
The visible and infrared portion of the solar spectrum is essentially
a continuum with the highest emission taking place near 500 nm.
Emission lines superimposed on the continuum become increasingly
more pronounced in the ultraviolet at wavelengths less than 300 nm.
At wavelengths shorter than 208 nm, a sharp decrease in the solar
flux occurs. This feature is associated with the Al I ionization edge.
Other less pronounced edges with their related continua are due to other
ionized elements in the solar atmosphere, such as H, Mg, Si, Fe, and
C. Below the aluminum edge, the importance of the solar emission lines
increases rapidly, while the absorption lines disappear from the spectrum
below 150 nm. At wavelengths shorter than 140 nm, the emission by
chromospheric and coronal lines begins to dominate the emission in the
continuum. The source of the solar emission thus changes from the
photosphere to the chromosphere as the wavelength decreases from 300
to 120 nm. As shown in Figure 4.7, the effective emitting layer of the
radiation moves up with shorter wavelength while the absorption in the
solar atmosphere increases. The minimum in the brightness temperature
between 180 and 150 nm characterizes the transition region between the
photosphere and the chromosphere. The variation in the brightness
RADIATION 165
Figure 4.7. Observed continuum brightness temperature of the sun as a function

of wavelength and corresponding relative height above τ500nm = 1 of the solar
emission layers ((upper panel).
) Relative contributions to the opacity at this
height (lower panel). After Vernazza et al.(1976).
temperature (6000 K above 300 nm, 5000 K at 200 nm, 4500 K between
170 and 130 nm) is reflected in the flux values at different wavelengths.
The solar spectrum has been the subject of numerous rocket
experiments undertaken since the 1940s. The first spectrum measured
above the ozone layer by a V-2 rocket dates from October 1946 (Baum
et al., 1946), and not until the 1950s was a solar spectrum observed
from an altitude of 100 km (Johnson et al., 1952). Today the solar
irradiance is routinely observed by spectrometers on board balloons,
aircraft, or spacecraft. During the 1990s, for example, the solar UV flux
was measured almost continuously between 120 and 400 nm by the Solar
Ultraviolet Spectral Irradiance Monitor (SUSIM) and the Solar/Stellar
Irradiance Comparison Experiment (SOLSTICE), both on board the
Upper Atmosphere Research Satellite (UARS) (see e.g., Woods et al.,

1996; 1998).
The flux generated by the nuclear burning core of the Sun changes
significantly as a function of time. Variations in the solar irradiance,
as well as changes in sunspots, faculae and plages, are associated with
the 11-year solar cycle, and hence with periodic changes in the magnetic
field of the Sun (see e.g., Nesme-Ribes et al., 1996). Over the last 2 solar
cycles, the variation in the Sun’s total radiative energy has been of the
order of 0.1% (Willson and Hudson, 1991; Lee et al., 1995). Variations
as a function of wavelength are more difficult to estimate because they
rely primarily on intermittant measurements by instruments that are
difficult to calibrate. Variability has been shown to be largest at
shortest wavelengths, corresponding to the emission from the corona
and the upper chromosphere. Large increases in the solar flux at short
wavelengths are associated with solar flares.
X-rays of wavelengths shorter than 1 nm constitute an important
source of ionization in the D-region, becoming the dominant source
during high levels of solar activity. The value of the corresponding flux
varies considerably: between 0.1 and 0.8 nm, it has been noted that
variations as large as a factor of 1000 may occur over the solar cycle.
Similarly, important fluctuations are observed over a solar rotation
(27 days) and substantial differences occur from one solar rotation
to another. During a solar flare, the X-ray irradiance may become
10000 times greater than that observed during quiet periods.
The flux in the Extreme Ultraviolet (EUV) portion (10-100 nm) of
the solar spectrum (see Figure 4.8a) directly affects the chemistry and
dynamics of the thermosphere. Its value, and hence the ionization rate in
the upper atmosphere, varies substantially with solar activity. Richards
et al. (1994) have developed a model that provides the extra-terrestrial
solar actinic flux from 5 to 105 nm as a function of a solar proxy P which
is the average between the instantaneous radiowave flux F10.7 emitted
by the Sun at 10.7 cm and its average value F10.7A over 81 days. The
10.7 cm flux is routinely monitored and reported in units of 10−22 W
m−2 Hz−1 ; its value ranges typically from 70 (solar minimum) to 200
(solar maximum), and is highly correlated with other indicators of solar
activity (e.g., sunspot number).
In the spectral region near 120 nm, the solar hydrogen emission line at
Lyman α represents an important source of ionization and dissociation.
During quiet periods, it contains more energy than the rest of the
spectrum at shorter wavelengths. The total flux of this line as well
as its shape vary with solar activity. The actinic flux varies between a
minimum of about 3.0 × 1011 photons cm−2 s−1 for quiet solar activity
and a maximum of about 6.0 × 1011 during high solar activity.
RADIATION 167
Figure 4.8a. Solar irradiance spectrum at wavelengths less than 200 nm

measured on 3 November 1994 with emphasis on the Extreme and Far
Ultraviolet regions. The solar bright lines are labeled. From Woods et al.(1998).
The solar flux in the 120-420 nm spectral interval plays a key role in
the photochemistry of the middle atmosphere. It contains radiation
emitted from the chromosphere (temperature of 104 to 106 K) and
photosphere (temperature of approximately 5700 K) in both lines and
continuum. Figure 4.8b presents the solar spectrum measured on
29 March 1992 by SUSIM with a 1.1 nm spectral resolution (Brueckner
et al., 1993) and by SOLSTICE at 0.1 to 0.25 nm spectral resolution
(Rottman et al., 1993).
Analyses of observed data suggest that, over the 11-year solar cycle,
the solar flux varies by a factor of 2 at Lyman-α (see Figure 4.9), 20%
at 140-155 nm, 15% at 160 nm, 10% at 170-190 nm, and 8% at 200 nm.
The variability above the Al I edge at 208 nm is extremely difficult to
measure since its magnitude is smaller than instrumental error. It is at
most a few percent between 210 and 300 nm, (3-4% at 250 nm, 0.5-1%
at 270 nm), and an estimate by Lean (1997) of the 11-year variability in
the Sun’s spectral irradiance between 0 and 400 nm is given in Figure
4.10.
Changes in the solar irradiance with a 27-day period (apparent
rotation period of the Sun) are also observed. The amplitude of the
variability decreases with wavelength in a manner similar to that shown
in Figure 4.10., but with reduced amplitudes.
Figure 4.8b. Solar irradiance spectrum between 119 and 420 nm measured
on March 29, 1992 by the SUSIM and SOLSTICE spaceborne spectrometers
on board the Upper Atmosphere Research Satellite (UARS). From Woods
et al. (1996).
Figure 4.9. Evolution with time of the solar irradiance at Lyman α, based on
measurements by the Atmospheric Explorer-E, the Solar Mesosphere Explorer,
and the SOLSTICE instrument on board the Upper Atmosphere Research
Satellite (UARS), after correction by Woods and Rottman (1997).
RADIATION 169
Figure 4.10. Ratio between the solar irradiance at solar maximum and
minimum activity conditions ((11-year solar cycle) as a function of wavelength
(10-400 nm), as estimated by Lean (1997).
4.4 The Attenuation of Solar Radiation in the

Atmosphere
4.4.1 Absorption
As solar photons penetrate into the earth’s atmosphere, they undergo

collisions with atmospheric molecules and are progressively absorbed
and scattered. The probability of absorption by a molecule depends
on the nature of the molecule and the wavelength of the incoming
photon. An effective absorption cross section, σa (λ), can be defined
for each photochemical species. This quantity is expressed in cm2 , and
is independent of the concentration of the species under consideration.
The Beer-Lambert law describes the absorption of a ray of incident
intensity Iλ passing through an infinitesimally thin layer ds (see
Figure 4.11). The variation of intensity is given by
d Iλ = −ka (λ) Iλ ds (4.21)
where ka is the absorption coefficient (expressed for example in cm−1 ).

This coefficient is proportional to the concentration n (cm−3 ) of the
absorbing medium (gas or particles) and is related to the effective
absorption cross section σa (cm2 ) by the expression
ka (λ) = σa (λ)n (4.22)

Figure 4.11. Absorption of solar radiation in an atmospheric layer of unit area.
Integrating (4.21) and substituting (4.22) we find

⎛ ⎞

Iλ (s) = I0,λ exp ⎝− σa (λ)n(s)ds⎠ (4.23)
s
where I0,λ is the flux at s = 0. The optical thickness or optical depth
over the length s is defined as

τa (λ, s) = σa (λ)n(s)ds = ka (λ, s)ds (4.24)
s s
and the corresponding transmission is
Tr (λ, s) = exp (−ττa (λ, s)) (4.25)
(Note that the optical depth is defined here along the line of sight, and
not, as in some texts, as the vertical thickness.)
Solar radiation penetrates the atmosphere at an angle of incidence
which depends on the local time, season, and latitude. The cosine of the
local solar zenith angle (χ) is given by
cos(χ) = cos(φ) cos(δ) cos(H) + sin(φ) sin(δ) (4.26)
where φ is the latitude, δ is the solar declination angle, which depends
on season (±23.5◦ for solstice conditions, 0◦ for equinox, for example),
and H is the hour angle, which is 0◦ for local noon, and increments
RADIATION 171
by 15◦ for each hour from noon. The infinitesimal distances along the
optical path ds and along the vertical dz are related by
ds = dz sec χ = dz/µ0 (4.27)
where µ0 = cos χ. Further, the concentration n(z) is often assumed to
vary exponentially with altitude z according to a scale height H (n0 is
the concentration at z = 0; see Chapter 3):
n(z) = n0 exp (−z/H) (4.28)
such that, for a medium containing only one well-mixed absorbing gas
(e.g., O2 ), the variation of monochromatic solar radiation with altitude
in a plane-parallel atmosphere (neglecting the curvature of the Earth)
can be written (omitting the index λ):
⎛ ⎞
∞
I (z) = I (∞) exp ⎝−sec χ σa n0 e−z/H dz ⎠ (4.29)
z

= I (∞) exp −sec χσa n0 H e−z/H (4.30)
where I (∞) represents the solar intensity outside the earth’s
atmosphere. The rate of formation of ions, the rates of photodissociation
and production of heat, are all directly linked to the rate of energy
deposition in the atmosphere by absorption. This last quantity can be
written as
dI dI
r=− = cos χ (4.31)
ds dz

z
= σa n0 I (∞) exp − + τ0 exp(−z/H) (4.32)
H
where
τ0 = σa n0 H sec χ (4.33)
represents the optical depth of the entire atmosphere for zenith angle χ.
The rate of energy deposited in the atmosphere exhibits a maximum at
the altitude
zm = H ln τ0 = H ln(σa n0 H sec χ). (4.34)
For an overhead sun, this maximum is located at the altitude:
z0 = H ln(σa n0 H) (4.35)
The rate of energy deposited at altitude
zm = z0 + H ln sec χ (4.36)
is thus equal to

z0
rm = σa n0 I (∞) cos χ exp −1 − . (4.37)
H
For an overhead sun, one obtains

z0
r0 = rm (χ = 0) = σa n0 I (∞) exp −1 − (4.38)
H
Making use of the dimensionless variable
z − z0
Z= (4.39)
H
the rate of energy deposition in a medium whose concentration decreases
exponentially with height, becomes
r
= exp (1 − Z − sec χ exp (−Z)) (4.40)
r0
which is represented in Figure 4.12. It should be noted that the
penetration of solar radiation and its interaction with the atmosphere
leads to the formation of layers whose characteristic altitude zm is
independent of the intensity of the radiation, but strongly dependent
on the physical characteristics of the atmospheric medium (nature and
concentration of the absorbing gas), the solar zenith angle, and the
wavelength of the radiation (because of the spectral distribution of
the absorption coefficient). Thus for polychromatic radiation for which
Figure 4.12. Appearance of a characteristic layer due to absorption of solar

energy. Distribution obtained by Chapman theory for different values of the
solar zenith angle.
RADIATION 173
regions of strong absorption by one or more absorbing gases occur at

different wavelengths, several absorbing layers are found at different
altitudes.
The theoretical framework just presented was first suggested by
Chapman (1931). This theory provides an explanation for the behavior
of the layers of ionization in the thermosphere or of photodissociation
in the middle atmosphere. The production of ozone through
photodissociation of molecular oxygen exhibits a maximum near 45 km
dependent on the insolation. The rate of heating through absorption of
ultraviolet radiation by ozone similarly leads to a maximum near the
stratopause. Numerous examples of such layers can be found in the
neutral and ionized atmosphere.
In practice, however, the analytic calculations by Chapman should
be replaced by an expression that accounts for the combined effects of
multiple absorbing gases. The Beer-Lambert law for the actinic flux qλ
can also be written:
qλ (z, χ) = qλ (∞) exp (−ττa (λ, z, χ)) = qλ (∞)Tr (λ, z, χ) (4.41)
where Tr = exp (−ττa ) is the transmission function, but in this case

the value of the optical depth results from the sum of several terms
which vary with the wavelength of incident radiation and the column
amount of absorbing gases. For example, in the middle atmosphere,
ultraviolet radiation is absorbed by molecular oxygen, or ozone, or both
(see Table 4.1), such that one generally writes:
⎡∞ ⎤
∞
τa (λ, z, χ) = secχ ⎣ σa (O2 , λ)(O2 )dz + σa (O3 , λ)(O3 )dz⎦ (4.42)
z z
When the solar zenith angle exceeds 75 degrees (sunrise or sunset), the
effect of the earth’s curvature can no longer be neglected and the secant
must be replaced by a more complicated function (Chapman function),
which depends not only on solar zenith angle χ but also on the altitude
where the absorption occurs (see Box 4.1). The effect of refraction may
also be important, especially for long optical paths and for radiation
penetrating deep into the atmosphere. Figure 4.13b provides an estimate
of the relative contributions of absorption by O2 and O3 for different
wavelengths. The cross sections of each of these gases must be multiplied
by their respective column abundances to compare their absorption
effects. The effect of molecular oxygen dominates for wavelengths less
than 220 nm but ozone absorption dominates near 250 nm and in the
visible.
Box 4.1 Atmospheric Sphericity

The calculation of the optical depth

τ (s) = σa n(s)ds (1)
s
requires an integration of the absorption along the optical path s. In
contrast to the troposphere, where inhomogeneities (e.g., clouds) are
often large, the middle atmosphere is generally stratified and more
homogeneous. In this case, the plane-parallel approximation is generally
satisfied, except for direct solar radiation at larger zenith angles. Under
twilight conditions (χ > 75◦ ), the sphericity of the Earth must be taken
into account by noting that the local solar zenith angle or its cosine
value µ0 vary along the optical path. Thus, under these conditions, the
optical depth between a point at altitude z and the top of the atmosphere
(∞) along the oblique path must be computed using the more complete
expression
∞
σa n(z )dz
τ (z, ∞) = (2a)
z µ0 (z )
if
1 a + z
= 1/2 (2.b)
µ0 (z ) 2 2 2
(a + z ) − (a + z) sin χ
where a is the Earth’s radius and χ the solar zenith angle at altitude z.
For the derivation of Eq. (2.b), see Figure 4.13a below.
Expressions similar to (4.42), in which the sec χ factor is replaced by
the Chapman function Ch(z,χ) are often used to account for the Earth’s
curvature. If the absorption cross section σa does not vary along the
optical path,
∞
Ch(z, χ) = z n(z )dz /µ0 (z ) (3)
∞
z n(z )dz
This function, also called the airmass factor represents the ratio of the
total amount of absorbant along the optical path versus the amount of
total absorbant in the vertical. For an isothermal atmosphere in which
light refraction can be neglected, and assuming that the absorbing gas
has a constant scale height H, the value of the Chapman function can be
estimated by the following expressions (Smith and Smith, 1972):
π πx 1
2
Ch(x,χ ≤ ) exp y2 erfc (y)
2 2
π πx 1 ! 1
2
Ch(x, χ > ) (sin χ) 2 exp [x (1 − sin χ)] (4a)
2 2 "
1
− exp y2 erfc (y) (4b)
2
RADIATION 175
Box 4.1 (Continued)

#
where x = (a + z) H, a is the Earth’s radius, z the altitude and H is the
atmospheric scale height, and where y = [x/2]1/2 | cos χ| and erfc (y) =
1 − erf (y) is the complimentary error function. For χ = π2 , (sunrise
1
or sunset), Ch (x, π2 ) = ( πx
2 ) . This function is equal to about 38 in
2
the middle atmosphere. An empirical approximation to the Chapman

function is given by the following expression:
35
Ch (χ > 85◦ ) ≈ $ (5)
1224 · cos2 χ + 1
For absorbants whose distribution is not horizontally homogeneous, or
do not vary exponentially with height, or if the absorption coefficient is not
constant, the approximations discussed above are not valid. In this case,
the calculation of the the optical depth must consider the full geometry
of the problem.
• For solar zenith angles smaller than 90◦ , the optical depth can be
computed by Eqs. (2a) and (2b)
• For solar zenith angles larger than 90◦ , the optical depth is given by
z ∞
σa n(z ) σa n(z )
τ (z, ∞) = 2 dz + dz (6)
z0 µ0 (z ) z µ0 (z )
where 1/µo (z) is given by (2b) and where
z0 = a(sin χ − 1) + z sin χ (7)

is the lowest altitude encountered by the optical beam. Note that
in expression (6), the dependence of the density n with latitude and
longitude, which can be important, has been omitted.
• For a zenith angle of 90◦ , because of a discontinuity in expression (2b)

when values of z become close to the altitude z, the integration is
performed along the optical path by noting that the increment ∆s
within a vertical layer ∆z is given by (see Figure 4.13a within the
box):
1/2
∆s = (a + z + ∆z )2 − (a + z)2 sin2 χ
1/2 (8)
− (a + z )2 − (a + z)2 sin2 χ
Limb sounding retrievals use the approach to perform integrations along

the line of sight.
Box 4.1 (Continued)
Figure 4.13a. Geometry for calculating the path length in a spherical

atmosphere. Factor µ0 (z ) = cos χ is given by the ratio between the
% &1/2
distances AP and OP , with AP = (a + z )2 − (a + z)2 sin2 χ
and OP = a + z , if a is the Earth’s radius, χ and χ the solar zenith
angle at altitudes z and z , respectively. The path length increment
∆s is the difference between distances AP and AP . The value of
AP is given above. The value of AP is given by the same expression
in which z is replaced by z +∆z . Adapted from Lopez-Puertas and
Taylor (2001)
Figure 4.13b. Relative contributions of ozone and molecular oxygen to the

vertical optical depth of the atmosphere. Courtesy of Sasha Madronich
(NCAR).
RADIATION 177
In the visible region, the absorption by nitrogen dioxide can also

contribute to the optical depth, especially in the lower stratosphere and
troposphere. Moreover, molecular oxygen has two weak bands in the red
region of the solar spectrum near 0.7 µm.
In the near infrared, the absorption of solar radiation is due chiefly to
vibrational and rotational transitions of several atmospheric molecules.
The most important absorber is water vapor with several absorbing
bands between 0.8 and 3.2 µm. Carbon dioxide also exhibits absorption
bands such as the 2.0 µm band, the 2.7 µm band (which overlaps with
the 2.7 µm band of water vapor) and the 4.3 µm band, located in a
spectral region where both solar and terrestrial radiation is weak. Since
these bands consist of many narrow lines, the transmission function
corresponding to a given spectral interval cannot be described in terms of
the Beer-Lambert law and empirical formulas must be used in practical
applications.
4.4.2 Scattering by Molecules and Aerosol Particles
In the optically denser layers of the atmosphere, solar radiation

undergoes multiple scattering due to the presence of large concentrations
of air molecules and aerosol particles. Scattering is a physical process by
which a particle in the path of an electromagnetic wave abstracts energy
from this incident wave and re-radiates that energy in all directions. The
optical depth describing the attenuation of radiation is thus the result
of two contributions (absorption and scattering):
τ (λ, z, χ) = τa (λ, z, χ) + τs (λ, z, χ) (4.43)
where
∞
τs (λ, z, χ) = sec χ ks (λ, z )dz (4.44)
z
or, in the case of scattering by air molecules M,
τs (λ, z, χ) = sec χ σs (λ) n (M; z) H
ks being the scattering coefficient, σs (λ) being the corresponding
effective cross section, n(M;z) is the total concentration at altitude z
and H is the atmospheric scale height. The albedo for single scattering
can be defined as
ks ks
Ω= = (4.45)
ka + ks k
which describes the relative importance of scattering versus the total
attenuation. The parameter k is called the extinction coefficient. The
atmospheric particles responsible for scattering cover a range of sizes

from gas molecules (≈ 10−8 cm) to large raindrops and hail particles
(≈cm). The relative intensity of the scattered light depends strongly on
the ratio of the particle size to the wavelength of the incident radiation.
When this ratio is small, the scattered light is distributed equally into
the forward and backward directions (Rayleigh scattering).
The scattering of unpolarized light by air molecules is generally
described by Rayleigh theory. In this case, the scattering coefficient
is expressed as a function of wavelength λ by (see e.g., Penndorf, 1957)
8π 3 (n2λ − 1)2
ks = D (4.46)
3 λ4 n(M)
where nλ represents the index of refraction of air and D is the
depolarization factor which expresses the influence of molecular
anisotropy. For air, a value of D of 1.06 is recommended (Penndorf,
1957; Elterman, 1968), although the depolarization is dependent on
wavelength (Bates, 1984). Adopting an analytic expression for the
wavelength dependence of the refractive index, the Rayleigh scattering
cross section can be written as (adapted from Nicolet, 1984):
4.02 × 10−28
σs = cm2 (4.47)
λ3.677+0.389λ+0.094/λ
where λ is in µm.
Aerosols (solid or aqueous particles in the atmosphere) also interact
with the radiation field. However, Rayleigh theory can no longer be
applied to these particles, because the size of the aerosols is of the same
order of magnitude or larger than the wavelength of the incident light.
The basic theory for the study of the scattering of light by aerosols
was presented by Mie (1908). This theory generally assumes an ensemble
of identical spherical particles and requires considerable computing
resources for its solution. Modifications to allow for cylindrical and
ellipsoidal particles have also been developed. In practice, simplified
formulas are often used (see, e.g., Van de Hulst, 1957). The total
extinction coefficient at a particular wavelength can be determined by
integration over all particle radii r
∞
dn
k(λ) = πr2 QE (λ, r, nλ ) dr (4.48)
0 dr
where dn/dr (particles cm−3 µm−1 ) represents the size distribution of
the particles, and QE (λ, r, nλ ) is the efficiency factor for attenuation of
radiation at wavelength λ by a particle of radius r and index of refraction
nλ . QE is often expressed as the sum of an absorption and a scattering
efficiency factor. An example of a stratospheric particle size distribution
measured at Laramie, Wyoming, by Dehsler et al. (1993) on October 2,
RADIATION 179
1991 (22 km altitude) is shown in Figure 4.14a. In this case, most of

the particles are believed to be sulfate aerosol with a peak in the size
distribution near 0.07 µm.
Figure 4.14. (a) Size distribution

of aerosols measured on October
2, 1991 at 22 km over Laramie,
Wyoming (Dehsler et al., 1993).
(b) Real and imaginary parts of
the refraction index as a func-
tion of wavelength (0-25 µm) for
sulfate aerosols (72% H2 SO4 by
weight). (c) Extinction, absorp-
tion, and scattering coefficients
(km−1 ) of aerosols for the par-
ticle size distribution shown as a
function of wavelength in Panel
a and the indices of refraction
displayed in Panel b. (d) Single
scattering albedo and asymmetry
factor shown as a function of
wavelength for aerosols as de-
scribed in the previous panels.
Courtesy of S. Massie, NCAR.
Another type of particle is also present as shown by the secondary peak

near 0.5 µm. Note that atmospheric particles such as those present
in polar stratospheric clouds, and ice crystals in the tropical lower
stratosphere, have large radii and/or non-spherical shapes.
The real and imaginary parts of the index of refraction calculated for
the sulfate aerosols (72% H2 SO4 by weight) are shown as a function
of wavelength in Figure 4.14b. (The real part of the refractive index
determines the phase speed of a wave in a given medium while the
imaginary part is a measure of the attenuation of the wave by this
medium.) For wavelengths less than 3 µm, the imaginary part of
the index is negligible, suggesting that the stratospheric sulfate aerosol
primarily scatters light in the visible. The extinction, absorption,
and scattering coefficients (expressed in km−1 ) for the particle size
distribution given in Figure 4.14a and for the indices of refraction given
in Figure 4.14b are presented in Figure 4.14c (assuming that the particles
are spherical and applying Mie theory). The total extinction is primarily
due to absorption for wavelengths greater than 3 µm, and primarily due
to scattering for wavelengths shorter than 2 µm. Also notice that the
extinction at visible wavelengths (less than 1µm) is 10 to 30 times larger
than the extinction in the infrared. Finally, Figure 4.14d shows the
wavelength dependence of the single scattering albedo (see Eq. 4.45)
and of the asymmetry factor g defined by
+1
g= p(Θ) cos Θ d cos Θ (4.49)
−1
where Θ is the angle between the direction of the incident light beam and
the direction of the scattered beam. The phase function p(Θ) measures
the intensity pattern of light scattered off of the aerosol particle for each
angle Θ. When the asymmetry factor is negative, most of the light
is scattered in the backward direction, while particles with positive g
scatter light primarily in the forward direction.
4.5 Radiative Transfer
4.5.1 The Radiative Transfer Equation
The radiation field in the atmosphere is determined from the equation

of radiative transfer (Chandrasekhar, 1950; Kourganoff, 1952; Sobolev,
1963; Lenoble, 1977), which is an expression of the energy balance in
each unit volume of the atmosphere, including absorption, scattering,
and emission. In the case of a horizontally stratified medium, the
following expression can be used to describe the radiative transfer in
RADIATION 181
a layer bounded by two infinite parallel planes (Lenoble, 1977):

µdLν (z; µ, φ)
= −kν (z) [Lν (z; µ, φ) − Jν (z; µ, φ)] (4.50)
dz
where Lν (z; µ, φ) is the spectral density of the radiance, kν (z) is
the extinction coefficient and Jν (z;µ, φ) is called the source function
at altitude z and for the direction µ, φ. Jν expresses the incoming
radiation due either to scattering from all other directions (µ , φ ) of
solar radiation, or to thermal emission by atmospheric molecules. The
importance of these contributions varies with the spectral domain under
consideration. When the source function can be neglected (Jν = 0), a
simple exponential extinction expression is obtained. It should also be
noted that in many studies, the altitude variable z is replaced by the
equivalent variable τν , which represents the optical depth for vertically
incident radiation (dττν = – kν dz). In this case, the radiative transfer
equation can be written as:
µdLν (ττν ; µ, φ)
= Lν (ττν ; µ, φ) − Jν (ττν ; µ, φ) (4.51)
dττν
When the atmosphere is assumed to be stratified and plane-parallel,
it is common to introduce the radiance Lν ↑ and Lν ↓ describing upward
and downward propagating photons, as was presented previously for the
flux (see Eqs. 4.39 and 4.40). In this case, integrating (Eq. 4.51), we
obtain:
Lν ↑ (ττν ; µ, φ) =Lν ↑ (ττν,1 ; µ, φ) e(τν,1 −τν )/µ
τν,1

1
− Jν (ττν ; µ, φ) e(τν −τν )/µ dττν (4.52)
µ
τν
for µ > 0, and
Lν ↓ (ττν ; µ, φ) =Lν ↓ (0; µ, φ) e−τν /µ
τν
1
+ Jν (ττν ; µ, φ) e(τν −τν )/µ dττν (4.53)
µ
0
for µ < 0 where the source function Jν depends on the radiance Lν ,
at least when scattering cannot be neglected (see Section 4.5.2). To
solve these equations, certain boundary conditions must be specified.
For example, outside of the atmosphere one can assume that the only
downward propagating radiation is that emitted by the Sun. At lower
altitudes (surface or clouds) (ττν = τν,1 ) one can assume that short
wavelength incident radiation is isotropically reflected by the surface of
albedo A (0.0<A<1.0). A more accurate treatment allows for an angular
dependence of A. For long wave radiation, it is usually assumed that the
Earth’s surface emits approximately as a black body at temperature Ts .
Note that the Earth’s surface infrared and microwave emissivity is a

function of vegetation, snow cover, soil moisture, sea conditions, etc.
4.5.2 Solution of the Equation of Radiative transfer for

Wavelengths Less than 3.5 µm: Multiple Scattering
As indicated previously, shortwave radiation below 3.5 µm originates

from the Sun. When the direct flux penetrates into the atmosphere, it is
progressively attenuated by absorption (ττa ) and scattering (ττs ). In this
case, the source function must account for the diffuse radiation scattered
into the beam from other directions (µ , φ ) as well as the direct sunlight
scattered into the direction of the pencil of radiation. If Φν (∞) is the
solar irradiance at the top of the atmosphere and (µo , φo ) the direction
of the sun, the source function is given by (Lenoble, 1977):
2π 1
Ων (ττν )
Jν (ττν ; µ, φ) = dφ pν (ττν ; µ, φ; µ , φ )Lν (ττν ; µ , φ )dµ
4π
0 −1
Ων (ττν )
+ pν (ττν ; µ, φ; µo , φo )Φν (∞)e−τν /µo
4π
(4.54)
where Ω(ττν ) is the albedo for single scattering (see Eq. 4.45) and
pν (ττν ; µ, φ; µ , φ ) is the phase function defining the probability that a
photon propagating in the direction (µ , φ ) is scattered in the direction
(µ, φ). Thus, the first term in equation (4.54) represents multiple
scattering effects, while the second term represents the contribution of
first-order scattering.
If the atmosphere is perfectly clear (neglecting the presence of solid
and liquid particulates), the scattering is due only to air molecules
and Rayleigh theory can be applied. The corresponding attenuation
coefficient is given by the equation (4.46), and the phase function is
given by:
3
p(Θ) = (1 + cos2 Θ) (4.55)
4
where
1 1
cos Θ = µµ + (1 − µ2 ) 2 (1 − µ 2 ) 2 cos(φ − φ ) (4.56)
Θ being the angle between incident and scattered light.
The scattering by solid and liquid particulates constitutes a more
difficult problem (see, e.g., McCartney, 1976; Bohren and Huffman,
1983). If the particles may be assumed spherical, Mie theory can be
invoked. In this case, for aerosols of radius r, the phase function is given
RADIATION 183
by
λ2
p(r, Θ) = (S1 S1∗ + S2 S∗2 ) (4.57)
2πks (r)
where the coefficients are given by the following expressions:
∞

2n + 1 an dPn1 P1
S1 (Θ) = + bn n (4.58a)
n=1
n(n + 1) dΘ sin Θ
∞

2n + 1 P1 dP 1
S2 (Θ) = an n + bn n (4.58b)
n=1
n(n + 1) sin Θ dΘ
Pn1 is the associated Legendre polynomial and an and bn are coefficients
depending on the index of refraction and on the Mie parameter αa =
2πr/λ. S ∗ is the complex conjugate of S and ks is the scattering
coefficient
∞
2πr2
ks (r) = 2 (2n + 1)(a2n + b2n ) (4.59)
αa n=1
The number of significant terms in this series is of the order of 2αa + 3.
In the presence of an ensemble of aerosols characterized by a
distribution function f (ra ), the phase function p(Θ) is deduced from
a function related to a specific radius ra by the relation
∞
p(ra , Θ)σ(ra ) f (ra ) dra
0
p(Θ) = ∞ (4.60)
σ(ra ) f (ra ) dra
0
Mie theory can be extended to non-spherical particles (such as ice
crystals, for example) but it is often sufficient to assume that randomly
oriented irregular particles scatter light in the same way that spherical
particles do.
Several analytic methods have been proposed to solve the equation
of radiative transfer in an absorbing and scattering atmosphere, but
they can only be applied for the most simple cases. To obtain
quantitative solutions, numerical methods are generally used, such as the
Monte-Carlo method, DART method, iterative Gauss, discrete ordinate
method, etc. A complete summary of these techniques is provided by
Lenoble (1977), and a detailed discussion of multiple scattering processes
in plane parallel atmospheres is given in the book by Liou (2002).
Because multiple scattering, surface reflection, clouds, and aerosols
have a significant effect on radiative intensities at photodissociative
wavelengths and consequently on the composition of the middle
atmosphere (see, e.g., Luther et al., 1978), simplified radiative transfer
schemes are often used in photochemical models. In the two-stream

methods, for example, (Meador and Weaver, 1980; Toon et al., 1989),
the diffuse radiation is divided into upward-propagating and downward-
propagating components. One commonly-used example is the delta-
Eddington method (Joseph et al., 1976). These methods allow for
at least approximate representations of the non-isotropic scattering by
aerosol and cloud particles, and are typically accurate to about 5% or
better in the stratosphere, and to about 10-20% in the troposphere
(Petropavlóvskikh,
´ 1995). Multi-stream methods, in which the angular
distribution of the diffuse radiation is computed in greater detail (e.g.,
the discrete ordinates method of Stamnes et al., 1988) are more accurate,
but also considerably more computationally intensive.
The model of Meier et al. (1982) and Nicolet et al. (1982) solves
a form of the radiative transfer equation in which the radiance is
integrated over all solid angles, and provides the solar actinic flux Φ
at wavelength λ or the enhancement factor or source function. The
latter is defined by S = Φ/Φ∞ , where Φ∞ is the solar actinic flux at
the top of the Earth’s atmosphere. A schematic diagram of their model
(which assumes that the phase function for scattering is isotropic) is
presented in Figure 4.15. It can be shown that the calculated source
function S at altitude z results from four contributions: 1) the direct
solar flux, 2) the multiply-scattered flux, 3) the ground reflection of
Figure 4.15. Schematic diagram of the scattering model by Meier et al.(1982).

RADIATION 185
Figure 4.16. Stratospheric enhancement factor calculated by Meier et al.(1982)

for different values of the albedo.
the multiply-scattered flux, and 4) the ground reflection of the direct

flux. The mean enhancement factor in the stratosphere between 300
and 800 nm for different values of the albedo and for an overhead sun
is shown in Figure 4.16. The “error bars” indicate the variability of
the S factor between the tropopause and the stratopause, suggesting
that for λ > 400 nm a single enhancement factor can be applied
to the entire atmosphere with a good accuracy. The S factor as a
function of altitude for wavelength intervals from 240 to 700 nm is
shown in Figure 4.17a, b, and c for different cases. Figure 4.17a
shows the results ignoring the effect of multiple scattering and reflection
at ground level. In this case, the enhancement factor is simply the
transmission factor which characterizes the attenuation by absorption
as well as by Rayleigh and Mie scattering. In the visible part of the
spectrum, the attenuation occurs near ground level, while in the UV
region, the flux is reduced by the presence of ozone at higher altitudes.
The effect of multiple scattering, neglecting ground albedo, appears in
Figure 4.17b. The contribution of the scattered light to the radiation
field is large except at short wavelengths where pure absorption is strong.
At 350 to 400 nm, for example, the scattering provides an additional
irradiance of about 50 percent. Finally, Figure 4.17c shows the S factor
when both multiple scattering and ground albedo (A = 0.5) are taken
Figure 4.17. Upper left panel (a)

Stratospheric enhancement fac-
tor as a function of altitude calcu-
lated by Meier et al.(1982). Ab-
sorption only Upper right panel
(b) Same as Panel A but for
absorption and multiple scatter-
ing. Lower left panel (c) Same
as Panel A but for absorption,
multiple scattering, and surface
albedo (A = 0.5).
into account. The enhancement factor is now greater than 2 even at

higher altitudes for radiation of wavelength longer than 320 nm. The
multiple scattering and albedo will therefore considerably affect the
photodissociation of molecules whose absorption cross sections are large
above 300 nm (O3 , NO2 , N2 O5 , NO3 , ClONO2 ). Figure 4.18 shows
the scattered radiance measured in the upper stratosphere between 175
and 325 nm for two viewing elevation angles. The actinic flux spectrum
recorded above the atmosphere by the spaceborne SUSIM instrument is
also shown. The effects of ozone absorption in the ozone Hartley band
(220-290) and of O2 absorption in the Schumann-Runge bands below
200 nm are clearly seen in the observed radiance. Radiances measured
at negative viewing elevation angles are larger in spectral regions with
small absorption optical depths (195-220 nm and > 285 nm) and smaller
in regions where absorption dominates. Scattered solar ultraviolet
radiation is an essential component of stratospheric photochemistry
RADIATION 187
Figure 4.18. Observed scattered radiance at 46.3 km for viewing elevation

angles at 14.6◦ and −3.1◦ along with SUSIM measurements of the solar
irradiance above the atmosphere by the Solar Ultraviolet Spectral Irradiance
Monitor (SUSIM; van Hoosier et al., 1988). The solar zenith angle at the
time of the scattered radiance measurements during a 1983 balloon flight over
Palestine, Texas, is 28◦ . From Minschwaner et al. (1995).
(Herman and Mentall, 1982a; Kylling et al., 1993; Minschwaner et al.,

1995) as scattered radiation in many cases dominates the actinic flux and
has an important effect on the photolysis of key chemical species (e.g.,
O3 , NO2 , HNO3 ). The measurement of scattered and absorbed solar
radiation can provide information on the abundance of several chemical
compounds (e.g., ozone, aerosols in the atmosphere) (see Box 4.2).
The presence of clouds in the troposphere modifies somewhat the
radiation field in the stratosphere (Lacis and Hansen, 1974) by altering
the albedo and introducing a highly scattering medium. Because the
reflectance properties of clouds vary considerably with cloud type,
numerical models dealing with this problem have to define statistical
properties of the cloud distribution.
Box 4.2 Space Observations of Atmospheric Compounds by Solar

Absorption Techniques
The atmospheric abundance of some chemical compounds that absorb
solar radiation (e.g., ozone) can be measured from space through different
techniques.
In the solar occultation method a space-borne detector (e.g.,
photomultiplier tube, photolytic array detector, Fourrier transformed
interferometer) points towards the Sun, and during brief periods (sunrise,
sunset), when the optical path penetrates into the atmosphere (see
Figure 4.19a), measures the attenuation of the solar radiation by the
absorbing compounds. The intensity at frequency ν detected by the
spacecraft is
+∞
Iν = I∞,ν exp −σν n(s) ds (4.61)
−∞
if I∞,ν is the intensity of the exo-atmospheric solar radiation (measured
by the spacecraft when the optical path of the detected radiation does
not penetrate into the atmosphere), σν is the (known) absorption cross
section of the absorbant at frequency ν, and n(s) is the number density of
the absorbing compound along optical paths. (For simplicity, we consider
only one absorbing gas).
Figure 4.19a. Geometry of solar occultation. McCormick et al.(1979)

Assuming that this absorbant is uniformly distributed in each atmospheric
layer (and hence that its concentration varies only with height), the
integrated concentration N along the optical path with tangent height
r0 (slant column density) can be derived from the observation of Iν and
Iν,∞ through Eq. (4.61), when
RADIATION 189
Box 4.2 (Continued)
+∞ ∞
r n(r)dr
N (r0 ) = n(s) ds = 2 ' (4.62)
−∞ r0 r2 − r20
if n(r) represents the vertical profile of the concentration (note that r
and r0 are defined from the center of the earth). The inversion of
Equation (4.62) to derive the vertical profile n(r) is often performed using
an “onion peeling” method in which the atmosphere is divided into a
finite number of vertical layers, and the concentration of the absorbing
compound is calculated successively from the top of the atmosphere to
the bottom layer. Note that in some cases absorbing constituents are not
uniformly distributed in the layers, especially if their concentrations are
strongly affected by photochemical conditions. In this case, a correction
must be applied (see e.g., Newchurch et al., 1996).
One of the advantages of the solar occultation method is that the
concentrations are derived from the measurement of a ratio of 2 fluxes, and
therefore are not substantially affected by instrument calibration errors or
solar spectral features. Because of the observing geometry involved, this
technique provides good vertical resolution. The major limitation results
from the limited number of observations due to the sunrise or sunset
contraints. Better coverage can be obtained by considering, in addition,
lunar and stellar occultations.
Assuming that the column N(z) above altitude z is proportional to the
pressure at level z,
N(z) = c p(z) (4.63)
and noting that further attentuation occurs as the scattered light
propagates towards the top of the atmosphere, the radiance Iν measured
by the spacecraft is proportional to
∞
Iν = Iν,∞ p(z) exp [−2σν c p(z)] dz (4.64)
0
It can easily be shown that the largest contribution to this integral is
provided by the atmospheric layer at pressure
p = [2σν c]−1 (4.65)
Thus, by choosing a set of wavelengths for which the cross section σν is
notably different, the mole fraction of the absorbant (proportional to c)
can be retrieved at different altitudes. This method is commonly used
to measure the vertical distribution of ozone in the stratosphere on the
global scale. The vertical resolution, however, is not better than several
kilometers.
The total ozone column abundance can be obtained by measuring the
differential absorption of reflected near-UV radiance at two distinct wave-
Box 4.2 (Continued)

lengths (corresponding to strong and weak absorption by ozone,
respectively) (see Bhartia et al., 1984).
Figure 4.19b. Vertical profiles of chemical compounds retrieved from

satellite observations based on occultation methods. Upper Panel:
Nighttime ozone number density (cm−3 ) measured on 24 March 2002 from
the tropopause to the mesopause levels (15◦ N, 115◦ E) by the GOMOS
instrument on board the ENVISAT spacecraft (stellar occultation).
Courtesy of J.L. Bertaux and A. Hauchecorne, Service d’Aéronomie
´ du
CNRS, France. Lower Panel: Water vapor mixing ratio (ppmv) between
the surface and 50 km altitude (33◦ N, 125◦ W) measured by SAGE II
on 11 January 1987 (solar occultation). Courtesy of M. Geller, State
University of New York.
RADIATION 191
4.5.3 Solution of the Radiative Transfer Equation at

Wavelengths Longer than 3.5 µm: Absorption and Emission
of Infrared Radiation
In the infrared region above 3.5 µm, the contribution from terrestrial
radiation dominates the solar input. Significant thermal emission and
absorption by atmospheric gases and aerosols take place in this spectral
region. Scattering can usually be neglected. The detailed study of
radiative transfer requires complete knowledge of the important bands
and a compilation of a large body of spectroscopic data (see, e.g.,
Rothman et al., 1998, and references therein).
Figure 4.20 shows the infrared emission spectrum recorded by a down-
looking spaceborne instrument. Also shown are the black body curves
for several temperatures. The deviations in the observed spectrum
from black body behavior can be attributed to the nonisothermal
character of the atmosphere and the variation of the atmospheric opacity
with wavelength (or wavenumber). Highest emissions correspond to
infrared fluxes originating from the warm surface of the Earth in quasi-
transparent spectral regions (called windows). Lowest spectral emissions
correspond to spectral regions where absorption by atmospheric
molecules is strong, and hence the observed radiation originates from
the cold layers of the upper troposphere.
Figure 4.20. Thermal emission spectrum of the Earth recorded by the IRIS
Michelson interferometer on board the Nimbus 4 satellite over the Sahara
region. Strong absorption by CO2 (15 µm or 667 cm−1 ), and ozone (9.6 µm
or 1041 cm−1 ) and water bands are clearly visible. From Hanel and Courath
(1970).
The radiatively active trace gases, which contribute most to

atmospheric absorption, and hence play a key role in the energy budget
of the middle atmosphere, are carbon dioxide, ozone, and water vapor.
Other gases, whose sources are partly related to anthropogenic activity
(such as CH4 , N2 O, and chlorofluorocarbons) also contribute to the
radiation budget.
Carbon dioxide is a linear molecule which has a relatively simple
absorption spectrum. One of the strongest bands is the ν2 fundamental
at 15 µm, which contributes significantly to the energy budget of the
atmosphere. The ν3 band at 4.3 µm is also strong and rivals heating by
ozone in the middle mesosphere. Additional weak bands centered in the
12-18 µm, 10 µm, and 7.6 µm regions must be considered in the CO2
climate problem, as discussed by Kiehl and Dickinson (1987).
Ozone is a non-linear molecule, and possesses strong rotational
structure as well as 3 fundamental vibration bands, ν1 , ν2 , and ν3 at
9.066, 14.27, and 9.597 µm, respectively. The ν2 band overlaps the
15 µm CO2 band, but the ν1 and ν3 transitions form the important
9.6 µm band. Another strong ozone band at 4.7 µm is located in a
spectral region with weak solar and terrestrial irradiances.
Water vapor is a non-linear molecule with a complex vibration-
rotation spectrum. The ν2 fundamental band centered at 6.25 µm is
overlaid with a series of rotational transitions to make a broad intense
band centered at 6.3 µm. A wide pure rotation band extends from
about 18 to beyond 100 µm with varying intensity. The ν1 and ν3
fundamentals at 2.74 and 2.66 µm absorb a significant amount of solar
energy, particularly in the troposphere. Table 4.3 lists the principal
spectral bands which are important for climate studies, including the
bands from radiatively active constituents which play a secondary
role in the radiation budget. Figure 4.21 shows the absorption of
radiant energy by the gases which contribute significantly to radiative
transfer in the atmosphere (see also Plate 6). This figure shows that
each gas has a particular absorption signature which is wavelength
dependent. The observation of absorbed or emitted terrestrial radiation
in these wavelength bands can yield important information about the
distributions of these gases in the atmosphere.
Throughout most of the middle atmosphere, the primary energy
loss occurs via radiative emission in the ν2 band of CO2 at 15 µm.
The contribution of this band to the total cooling has been studied
in detail below 100 km (López-Puertas
´ et al., 1992; Mertens et al.,
1999; Mlynczak et al., 1999). Dickinson (1973) and Ramanathan (1976)
indicated the importance of the CO2 hot bands in the calculation of the
cooling rate. In the region of the atmosphere where the conditions of
local thermodynamic equilibrium (LTE) apply, (that is when collisions
RADIATION 193
Table 4.3 Important infrared atmospheric bands (From WMO, 1982)
Trace gas Band Band center Band intensity

cm−1 cm−2 atm−1
(296K)
CO2 ν2 (15 µm) 667 ≈220

CO2 ν3 (4.3 µm) 2348 ≈2440
O3 ν3 (9.6 µm) 1041 312
H2 O rotation 0–1650 1306
H2 O ν2 (6.3 µm) 640–2800 257
CH4 ν4 (7.66 µm) 1306 134
N2 O ν1 (7.78 µm) 1285 218
N2 O ν2 (17 µm) 589 24
CFCl3 ν4 (11.8 µm) 846 1813
CFCl3 ν1 + 2ν2 1085, 2144 679
CF2 Cl2 ν8 (10.9 µm) 915 1161
CF2 Cl2 ν1 (9.13 µm) 1095 1141
CF2 Cl2 ν6 (8.68 µm) 1152 777
Figure 4.21. Spectral distribution of the absorption by several radiatively active

gases (from the Earth’s surface to infinity). The aggregate spectrum due to all
atmospheric gases is shown in the bottom panel (adapted from Shaw, 1953).
are sufficiently frequent that the energy levels are populated according
to a Boltzmann distribution, see Eq. 2.44) then the source function is:
Jν (ττν ) = Bν (T(ττν )) (4.66)
where Bν (T (ττν )) is given by Planck’s law:
2hν 3 1
Bν (T(ττν )) = 2 hν/kT (4.67)
c e −1
The radiance over most of the middle atmosphere (below 75 km)
can thus be derived from the transfer equation (called Schwarzchild’s
equation, see e.g., Liou, 2002):
∂Lν (ττν ; µ, φ)
µ = Lν (ττν ; µ, φ) − Bν (ττν ) (4.68)
∂ττν
In other words, assuming azimuthal symmetry (no dependence of
the radiance on φ) and making use of equations (4.52) and (4.53), the
upward and downward components of the radiance will be evaluated
from equations of the following type:
z
∂Tr,ν (z , z; µ)
Lν ↑ (z; µ) = Bν (Ts )Tr,ν (z, 0; µ) + Bν (z ) dz (4.69a)
∂z
0
and
∞
∂Tr,ν (z , z; µ)
Lν ↓ (z; µ) = − Bν (z ) dz (4.69b)
∂z
z
where Ts is the temperature of the Earth’s surface and z is the altitude,
while ⎛ ⎞
z
Tr,ν (z, zo ; µ) = exp ⎝− kν dz /µ⎠ = exp (−σν N) (4.70)
zo
is the transmission function at frequency ν (see Eq. 4.25) for an optical
path from zo to z with inclination µ. The quantity N represents the
integrated amount of absorbing gas between altitudes z0 and z, and σν
is the effective absorption cross section at frequency ν. Equation (4.69a)
provides the formalism that is central to infrared remote sounding of the
atmosphere from space (see Box 4.3).
The solution of equations (4.69a,b) is complex since the absorption
spectrum exhibits structure characterizing the numerous rotation and
vibration-rotation lines of atmospheric molecules. The absorption
coefficient or the transmission function Tr,ν varies therefore considerably
over small regions of frequency (or wavenumber) and in principle a line-
by-line integration is required. Figure 4.22 shows, for example, the
vertical transmittance through the atmosphere at the edge of the strong
15 µm CO2 bands (Edwards, 1992).
RADIATION 195
Figure 4.22. An illustration of the spectral fine structure found in atmospheric

absorption bands. This example shows a line-by-line calculation of the vertical
transmittance through the atmosphere at the edge of the strong 15 µm CO2
bands (Edwards, 1992).
Box 4.3 Remote Sounding of Atmospheric Temperature and

Chemical Composition from Space
By measuring the spectral distribution of the upwelling infrared
radiation emitted by the Earth and its atmosphere, spaceborne sensors
can provide information on the vertical temperature profile and on the
atmospheric abundance of radiatively active trace gases. When local
thermodynamic equilibrium conditions apply, the radiance received by
a detector with spectral response function ϕν over frequency interval ∆ν
and viewing vertically downwards is given by (see Eq. 4.69a)
∞
L∆ν = B∆ν (Ts ) Tr,∆ν (0, ∞) + B∆ν (z) w∆ν (z) dz (4.71)
0
where Tr,∆ν (z1 , z2 ) is the optical transmission (for frequency interval ∆ν)
between levels z1 , and z2 , B∆ν (Ts ) is the Planck function in the same
spectral interval corresponding to the surface temperature Ts and B∆ν
(z) is the Planck function calculated at each atmospheric altitude z for air
temperature profile T(z). The weighting function

∂Tr,ν (z, ∞)
w∆ν (z) = ϕν dν (4.72)
∆ν ∂z
Box 4.3 (Continued)

determines the contribution of the emission by different atmospheric layers
to the radiance measured at the top of the atmosphere in the spectral
interval ∆ν. The response function of the instrument is such that

ϕν dν = 1 (4.73)
∆ν
Thus, in principle if the atmospheric abundance of the absorbing
constituents, the response function of the instrument and the surface
temperature are known, the temperature profile can be retreived by
inverting equation (4.71). Similarly, the concentration of the absorbing
species can be derived for a known temperature profile. If the absorption
results from a uniformly mixed compound such as CO2 (constant mixing
ratio X), the monochromatic transmission can be expressed as
Tr,ν (z, ∞) = exp [−σν Xn(z)H] (4.74)
where σν is the absorption cross section at frequency ν, n(z) is the altitude-
dependent air density and H is the atmospheric scale height (assumed
to be constant). The corresponding weighting function at frequency ν
(assuming ϕν = 1)
wν (z) = σν Xn(z) exp [−σν Xn(z)H] (4.75)
reaches a maximum at the altitude zmax where σν Xn(zmax )H = 1,
and the radiance observed at this frequency by the nadir-viewing
spaceborne sensor will be representative of the atmospheric conditions
(e.g., temperature) in the vicinity of this altitude. The vertical profile of
the temperature or of radiatively active gases can therefore be obtained
by measuring the upwelling radiance at different frequencies (using, for
example, a multi-channel radiometer). The vertical resolution of these
profiles is, however, severely limited by the width of the weighting function
(usually more than one atmospheric scale height). It can be somewhat
improved by adding observing channels, but remains constrained by the
fact that overlapping weighting functions do not provide independent
information. In the case of chemical compounds in trace quantities,
nadir soundings provide in most cases only the column abundance, and
very little information on the vertical distribution. Increased vertical
resolution can be obtained by using radiometers with narrow fields of
view scanning through the limb of the planet. Limb observations of
atmospheric emissions against the cold space background have provided
unique information (high vertical resolution, near global coverage) on the
abundance of many chemical species in the stratosphere and mesosphere.
The determination of the transmission function versus wavenumber

depends strongly on knowledge of the detailed structure of the
RADIATION 197
absorption spectrum, and in particular, the shape of the absorption lines.

(For reviews on the subject, see Mitchell and Zemansky, 1934; Penner,
1959; Goody, 1964; Kondratyev, 1969; Liou, 2002; Gordley et al.,
1994; Marshall et al., 1994.) In the absence of molecular collisions and
molecular motions, the natural width of a spectral line is very narrow and
is determined by the radiative lifetime of the excited state. This width
is called the “natural line broadening” and is a direct consequence of
the Heisenberg uncertainty principle. It is considerably smaller than the
Doppler and Lorentz broadening due, respectively, to the thermal motion
of the molecules and the collisions between them. In the troposphere and
lower stratosphere the width of the lines of the absorbing molecules such
as carbon dioxide and water vapor is determined essentially by collision.
In this case, one can assume a Lorentz line shape:
SγL
σν̄ = (4.76)
π (ν̄ − νν̄o )2 + γL2
where σν̄ represents the effective absorption cross section and ν̄ is the
wavenumber, following the convention of spectroscopists. γL indicates
the half width at half maximum of the line (expressed in wavenumber),
ν̄o is the midpoint of the line and
∞
S= σν̄ dν̄ (4.77)
0
is the integrated intensity of the line. Since the molecular collision
frequency varies with pressure and temperature, one may write, (see,
e.g., Tiwari, 1978):

p To α
γL = γL,o (4.78)
po T
where γL,o represents the width of the line at standard pressure po and
temperature To . For most gases, γL,o is of the order of 0.1 cm−1 . The
exponent α is approximately equal to 0.75 for CO2 , 0.64 for H2 O, and
0.76 for O3 . At very high altitudes, collisions become much less frequent,
and the effect of pressure can be neglected, which results essentially in
a Doppler linewidth. In this case, the profile can be represented as a
Gaussian:
S ν̄ − νν̄o 2
σν̄ = √ exp − (4.79)
γD π γD
where γD is the half width of the line measured at 1/e of the maximum
intensity. (Note the difference in the definition of γD and γL .) A typical
value of γD (which is independent of pressure, though dependent on
temperature and wavenumber) is 10−3 cm−1 . In the transition region
between the Lorentz and Doppler regimes, the lineshape is determined
by both broadening processes. If these are assumed to be independent,

they may be combined to give the Voigt profile (Mitchell and Zemansky,
1934; Penner, 1959; Armstrong, 1967):

+∞
a S exp (−y2 )
σν = dy (4.80)
γD π 3/2 a2 + (v − y)2
−∞
where a = γL /γD , v = (ν̄ − νν̄o )/γD , and y = (ν̄ − νν̄o )/γD . The
Voigt profile is a good approximation to the observed line shape. The
Voigt function (4.80) has no known analytic solutions, so that numerical
techniques are required for its evaluation (see e.g., Humlicek, 1982). At
high pressure, when a → ∞, and at low pressure, when a → 0, the
Voigt profile approaches the Lorentz and Doppler profiles, respectively.
Figure 4.23 shows the different line shapes.
To determine the upward flux F ↑ and the downward flux F ↓ in a
one dimensional representation, an integration of Eqs. (4.69a,b) must be
performed over all solid angles according to Eq. (4.5) and Eq (4.6). It
is convenient to introduce the “flux transmissivity”
1
T∗r (z , z) =2 Tr (z , z;µ)µdµ (4.81)
0
which provides a weighted average of the contribution of light
Figure 4.23. Lorentzian, Doppler, and Voigt spectral line profiles for
approximately equal half widths and intensities. (From Andrews et al.., 1987).
RADIATION 199
propagating along various inclination angles. It can be shown (see

Rodgers and Walshaw, 1966; Goody and Yung, 1989) that the scattered
transmission is approximately equivalent to the propagation of a
collimated beam at about a 53 degree angle (µ = 35 ). In other words,
the calculation of T∗r can be performed with an air mass whose depth is
5
3 its vertical thickness. Thus
3
T∗r (z ; z) ≈ Tr (z , z; µ = ) (4.82)
5
(Note that Apruzese et al. (1982) have shown that a value of 1.81
(versus 1.66) is more relevant for 1/µ in the mesosphere and lower
thermosphere.) One then finds
z ∗ (z , z)
∂Tr,ν
∗
Fν̄ ↑ (z) = πBν̄ (Ts )Tr,ν (zo , z) + πBν̄ (z ) dz (4.83a)
∂z
0
∞ ∗ (z, z )
∂Tr,ν
Fν̄ ↓ (z) = − πBν̄ (z ) dz (4.83b)
∂z
z
Thus, in theory, by knowing the location, strength, and shape of all
absorption lines, the irradiance at any wavelength (or frequency) can
be derived, and the total irradiance corresponding to an entire band
could be obtained by a lengthy integration over the spectrum. In order
to simplify the computation, and to derive the irradiance over a larger
spectral range, the previous equations can be integrated over an interval
δν̄ which is sufficiently narrow that the Planck function can be assumed
constant (δν̄ < 50cm−1 ), but containing the largest possible number of
emission lines (δν̄ > 5cm−1 ). In this case, the irradiance for interval i
is given by the following expression (after integration by parts of the
expression above):
z
dBi (z ) ∗
Fi ↑ (z) = πBi (z) − π Tr,i (z , z)dz (4.84a)
dz
0
∞
∗ dBi (z ) ∗
Fi ↓ (z) = −πBi (z = ∞)Tr,i (z, ∞) + πBi (z) + π Tr,i (z , z)dz
dz
z
(4.84b)
where Bi (z) represents the Planck function in the frequency interval i
for the temperature of the air and the altitude z, and where

1
T∗r,i (z , z) = T∗r (z , z)dν̄ (4.85)
δν̄
δν̄
is an effective transmission function corresponding to interval i. From

Eq. (4.10), one can derive that the heating rate corresponding to spectral
interval i at altitude z:
∗
1 dTr,i
Qi = −πBi (z = ∞) (z, ∞)
ρcp dz
∞ (4.86)
∂T∗r,i (z , z) dBi (z )
+ π dz
0 ∂z dz
where the first term, called the cool-to-space component, accounts for
the emission of radiation from any level z out to space, and the second
term accounts for exchange of radiation between any level z and levels
above and below this level of interest.
The determination of the effective transmission Eq. (4.85) requires, in
principle, a line by line integration including all the complex structures
of the various absorbing lines and their overlap. Such a calculation
is potentially very accurate and capable of including all the relevant
radiative transfer physics, although it may require a considerable amount
of computer time. As with all schemes, there remains the uncertainty
associated with the underlying spectral line data. However, because
they perform the spectral integration directly, line-by-line models may
be used as a reference against which the accuracy of more approximate,
but faster, techniques may be measured (Ellingson et al., 1991). They
can also be used to provide input parameterizations to an approximate
method in which the effective transmission function for a band or a
portion of a band is expressed in terms of global parameters such as
the mean line strength, the mean spacing between the individual lines,
and the mean line width. Models have been developed to treat either
sections of bands (narrow band models) or entire bands (wide band
models). See Tiwari (1978) and Goody and Yung (1989) for excellent
reviews of these models. The choice of any particular formulation of
the transmission function is dictated by the degree to which the model
approximates experimental data. The spectral parameters are usually
adjusted to optimize the fit to observed absorption.
Narrow band models parameterize the transmission for wavenumber
intervals δν of typically 5 to 20 cm−1 . The narrow band model of
Elsasser (1942) represents the spectrum by a series of regularly spaced
Lorentz lines of the same size and intensity. This model is best applied to
linear triatomic molecules such as CO2 and N2 O. The model by Goody
(1964) is based on the idea that the lines are randomly spaced over
a particular wavelength interval, with some exponential distribution of
line strength. This model can readily be applied to water vapor and
to carbon dioxide. If an exponential distribution of the line intensities
RADIATION 201
is assumed (i.e., the probability of finding a line intensity Sj in a given

intensity range decreases exponentially with the value of Sj ), the effective
transmission function in the spectral interval δν̄ is given by
−1/2
S S
Tr,δν̄ = exp − w 1+ w (4.87)
d δν̄ πγ δν̄
where S (cm−2 atm−1 ) is the mean strength of the lines for the interval
δν̄, γ the mean half width (cm−1 ), and d the mean line spacing, w
the amount of absorber (atm cm−1 ). Malkmus (1967) introduced a
distribution of the lines in S−1
j . The statistical model then yields
1/2
−(S/d)δν̄ S
Tr,δν̄ = exp 1 + 4w −1 (4.88)
2(Sπγ)δν̄ πγ δν̄
Because lines in vibration-rotation bands are arranged neither at regular
nor at random intervals, other models have been suggested, such
as the random Elsasser model, which is constructed by the random
superposition of several different Elsasser bands. Each of the superposed
bands may have different average line intensities, line widths, and line
spacing. Such a model provides a more accurate representation of the
absorption for bands whose lines have relatively constant intensities over
a narrow band region, but different intensities from one of these regions
to another. These models were originally developed for Lorentz lines,
but have also been extended for the Doppler and Voigt line shapes (see
Tiwari, 1978).
The above formulations of the effective transmission function have
been derived for homogeneous (constant pressure and temperature)
optical paths. In the atmosphere, the pressure and the temperature vary
with altitude, and the value of the spectral parameters (line width, line
strength) should be adjusted accordingly. Curtis (1952) and Godson
(1953) presented an approximation in which the transmission of an
inhomogeneous atmospheric path is replaced by the transmission of a
homogeneous path with an effective line strength
w
1
S̄ = S (T) dw (4.89)
w
0
and an effective half width of the line
w w
γ̄ = S (T)γ(p, T)dw / S (T)dw (4.90)
0 0
after which the band models for the homogeneous case can be applied.
This approximation provides fairly good results for species whose optical
mass decreases with pressure (CO2 , H2 O), but is rather poor for
constituents such as ozone, whose molar fraction changes with altitude

(Liou, 1980).
There are other techniques that rely on a reordering of the
transmittance integrand. As a function of wavenumber, the integrand
of
1
Tr,δν = e−kν w dν (4.91)
δν δν
shows a fast spectral variation requiring it to be sampled at many
wavenumber points. For example, a representative lower stratosphere
15 µm CO2 Voigt line halfwidth has a value of about 5 × 10−3 cm−1 . If
we assume that a minimum of two wavenumber points per halfwidth are
required to model the line profile, then on the order of 40000 wavenumber
points would be needed to cover 100 cm−1 over the strong fundamental
band and the resulting integration would be computationally expensive.
However, if we know the fraction f(k) of the integration range δν in which
the absorption coefficient lies between k and k + ∆k, it is equivalent to
write the band integrated transmittance for a homogeneous path using
a k-distribution as ∞
1
Tr,δν = f (k) e−kw dk, (4.92)
δν 0
although it is usually more convenient to switch to a cumulative g-
distribution function
k

g(k) = f k dk , (4.93)
0
such that the band transmittance integral may be written
1
Tr,δν = e−g(k)w dg (4.94)
0
Because g(k) is a monotonically increasing function, it may be evaluated
with a few tens of g-points in the range zero to unity, thereby allowing
the integral to be evaluated quickly (Lacis and Oinas, 1991).
This method essentially calculates a single equivalent pseudo-line over
a pseudo-frequency parameter g from the rapidly varying line spectrum
in the wavenumber range δν. In this sense, it retains some of the
characteristics of a monochromatic line-by-line calculation. If a good
correlation exists between the line centers and shapes of spectral lines in
different layers of a multi-layered atmosphere, i.e., the mapping of ν to
g is essentially the same in all layers, then the g-distributions of several
layers may be added together to allow for multi-layer radiative transfer
calculations. The conditions under which this correlated-k method is
valid are discussed by Goody et al. (1989) and Fu and Liou (1992).
An advantage of the k-distribution technique over band models is that
scattering problems can be addressed directly.
RADIATION 203
Narrow band methods, although reasonably accurate when compared

to line-by-line integrations, are still too cumbersome for most climate
model applications. Wide band models are particularly useful when the
spectral variation of the flux is not an important consideration. In
this case, the flux equations can be integrated over a whole spectral
band ∆ν (such as the 15 micron band system of CO2 ). A number of
approaches exist for constructing the effective transmission T∗∆ν or the
corresponding absorptance A∆ν = 1 – T∗∆ν as a function of the optical
path (see e.g., Cess and Ramanathan (1972) and Ramanathan (1976)).
Band models are useful for practical applications but have some basic
limitations; for example, they are only capable of yielding low spectral
resolution and they do not always account for the effects of band wings.
Kiehl and Ramanathan (1983) have shown, for example, that in the case
of the CO2 band at 15 µm the Goody and Malkmus formulation can only
yield accurate results if the spectral interval δν̄ is less than 10 cm−1 ,
and that the wide band models are probably more suitable for radiative
transfer formulations in the middle atmosphere. With the development
of fast computer systems, however, band models are becoming obsolete
for many applications (e.g., retrieval of temperature and chemical
constituents from satellite measurements) and are progressively being
replaced by more accurate line-by-line integrations.
The previous discussion is based on the assumption that collisional
excitation and de-excitation is the dominant process determining the
populations of the vibrational-rotational levels of all radiatively active
molecules. This is usually the case in the lower atmosphere, where
the pressure is high and molecular collisions are frequent. The relative
populations of the upper and lower states of a vibrational-rotational
transition are then described by the Boltzmann distribution at the
local kinetic temperature T, the gas is considered to be in local
thermodynamic equilibrium (LTE), and Kirchoff’s law can be applied
locally. If we consider a two-level system, then the relative populations
of the lower and upper states, n0 and n1 , respectively, are given by

n1 g1 −hν
= exp (4.95)
n0 g0 kT
where g0 and g1 are the level statistical weights, and ν is the transition
frequency. The source function is then given by Planck’s law. In
LTE, the relaxation time due to collisions is considerably shorter than
the radiative lifetime of the excited vibrational-rotational states. The
radiative lifetimes with which we are concerned are of the order of 0.1 sec
(0.7 sec for the CO2 15 µm band, 0.08 sec for the ozone 9.6 µm band,
and 0.05 sec for the H2 O 6.3 µm band; (Rothman et al., 1998)). The
relaxation time at ground level is determined by thermal collisions with
molecular oxygen and nitrogen, and is considerably shorter. In the

case of the CO2 15 µm band, it is of the order of 2.5 × 10−5 sec in
the temperature range 150-210 K with an uncertainty of about 50%
(Simpson et al., 1977). Since this relaxation time varies inversely with
the pressure, the radiative and relaxation time constants become equal
at about 68 km, assuming a scale height of 7 km.
To determine the altitude at which LTE actually breaks down for
each state of a given molecule, one must consider all the pumping
mechanisms that affect the state in question. These may include; (1)
absorption of radiation originating in warmer parts of the atmosphere
at lower altitude, (2) absorption of solar radiation in the near-IR, (3)
chemical pumping when the molecule is formed in a vibrationally-excited
state, (3) collisions where thermal energy produces a transition, and
(4) collisions involving exchange of vibrational quanta. Processes (1)
and (2) may be particularly important when an absorption band has
high optical depth. Breakdown of LTE occurs at about 90 km for the
15 µm CO2 ν2 fundamental band, at about 70 km for the 9.5 µm O3 ν3
fundamental band, and about 60 km for the 6.3 µm H2 O ν2 fundamental
band. It should be noted that hot bands and isotope bands often depart
from LTE much lower down in the stratosphere. Following Kuhn and
London (1969), we consider the case of the CO2 15 µm fundamental
band where the partition of population between the upper and lower
states is determined by radiative and thermal collisional processes. In
the middle atmosphere, these must be considered together in a balance
equation to give
n0 B01 L∆ν − n1 A10 − n1 B10 L∆ν + pt n1 − lt n1 = 0 (4.96)
where A10 , B10 , and B01 are the Einstein coefficients for spontaneous
emission, induced emission, and absorption, respectively. The thermal
collisional deactivation and activation specific loss rates are lt and pt ,
respectively, and the mean radiation field averaged over the band is given
by

1
L∆ν = Lν kν dνdω (4.97)
4πS ω ∆ν
This allows us to write the population ratio between the states as
n1 B01 Lν + pt
= (4.98)
n0 A10 + B10 Lν + lt
A general form of the source function may be written

−1
2hν 3 g1 n0
Jν = − 1 (4.99)
c2 g0 n1
which reduces to the Planck function under LTE conditions when n1 /n0
is given by the Boltzmann equation. The line intensity S can also be
RADIATION 205
defined in the general case as

hν n0 g0 n1
S = 2 B01 1− (4.100)
c N g1 n0
Therefore, using the expression for n1 /n0 , we are able to adapt the
radiative transfer equations described in the previous sections to the
general case where LTE can no longer be assumed (Edwards et al., 1993).
Defining
lt
E= [1 − exp(−hν/kT)] (4.101)
A10
and making use of the fundamental relations between the Einstein
coefficients, and the fact that pt /lt must be given by the Boltzmann
distribution (because this ratio alone determines n1 /n0 under LTE
conditions), the non-LTE source function may be written
Lν + EBν
Jν = (4.102)
1+E
The more general, multi-level case, is described by López-Puertas
´
et al. (1986). In the lower atmosphere where collisions dominate,
E → ∞, Jν → Bν , and LTE conditions apply. At high altitude where
the population of the energy levels is determined by radiative processes
only, E → 0, and the source function becomes Jν = Lν . This case is
referred to as monochromatic radiative equilibrium (MRE) in which, for
each frequency, the absorbed radiant energy is balanced by the emitted
radiant energy.
The importance of non-LTE processes in stellar atmospheres was
first pointed out by Milne (1930), and the first method for treating
the problem in the terrestrial mesosphere was presented by Curtis and
Goody (1956). Since that time, these questions have been addressed
by Houghton (1969), Dickinson (1984), López-Puertas
´ et al. (1986),
Wintersteiner et al. (1992) and Shved et al. (1998), among others.
Crutzen (1970) pointed out the possible importance of the excitation of
the CO2 ν2 fundamental band due to collisions with atomic oxygen and
the effect that this might have on the cooling of the upper mesosphere
and lower thermosphere. Until recently, there was considerable doubt
as to the rate constant for this process. López-Puertas
´ et al. (1992)
obtained a value of between 3 and 6 × 10−12 cm3 s−1 from measurements
of the Atmospheric Trace Molecules Spectroscopy (ATMOS) instrument
aboard the Spacelab 3 shuttle mission, an order of magnitude faster than
previous, widely-used values, and this was supported by the laboratory
measurements of Shved et al. (1991). This results in a cooling rate at
100 km nearly ten times greater than earlier calculations due to the fact
that the higher collision rate helps maintain the population of the upper
level of the fundamental band much closer to LTE. Accuracy in the

calculations of heating and cooling rates in the mesosphere and lower
thermosphere remains limited by uncertainties in the relaxation rates
and their temperature dependencies, together with uncertainties in the
kinetic temperature and the concentrations of collision partners.
4.6 The Thermal Effects of Radiation
Figure 4.24 summarizes the different major contributors to solar heat-

ing and terrestrial cooling calculated by solving the radiative transfer
equations and applying equation (4.10) to derive heating/cooling rates.
This figure shows that on the average, the major contribution to solar
heating is provided by the absorption of solar ultraviolet radiation by
ozone in the stratosphere and mesosphere and by molecular oxygen in
the lower thermosphere. Thermal infrared cooling is associated primarily
with the presence of carbon dioxide, ozone, and water vapor, and is a
strong function of temperature.
The thermal effects of solar and atmospheric radiation have been
studied using mathematical models, generally by solving the radiative
transfer equations which we have just described (e.g., Ramanathan,
1976). The most elaborate models, particularly those with fine spectral
resolution, require rather large computing resources, and cannot be
coupled with chemical or dynamical models using present computers
unless they are greatly simplified.
In many cases, only the mean vertical temperature distribution
is desired. The distribution obtained by assuming local radiative
equilibrium between the energy absorbed and the energy emitted at
a given altitude yields a very large vertical gradient in the troposphere
(Manabe and Moller, 1961). It becomes necessary to consider convection
and vertical transport of non-radiative energy. To resolve this problem,
one-dimensional radiative convective models, such as that first presented
by Manabe and Strickler (1964) are widely used. In such models
the lapse rate is adjusted to a representative value whenever the
computed value exceeds this level. This adjustment is performed until
an equilibrium is reached in each atmospheric layer. The net incoming
solar energy flux is thus balanced by the net outgoing long wave energy
flux at the top of the atmosphere. At the Earth’s surface, a balance
must exist between the net energy gain by radiation and the loss of
heat by convective transfer into the atmosphere. Whenever the lapse
rate is subcritical, the model assumes that pure radiative equilibrium
is achieved. Such model techniques require an iterative procedure
RADIATION 207
Figure 4.24. Vertical distribution of solar short wave heating rates by O3 , O2 ,

NO2 , H2 O, CO2 , and of terrestrial long wave cooling rates by CO2 , O3 , and
H2 O. From London (1980).
in which the layers of the atmosphere are treated repeatedly until these
conditions are satisfied.
Middle atmosphere models calculate the spatial and temporal
distribution of the net heating rate Q and the temperature not only
as a function of altitude but also as a function of latitude, and even
of longitude (local time). Such studies consider the multidimensional
transport of heat and the solution of a thermodynamic equation like
the one shown in Equation (3.10). The effect of waves should also
be considered. Gravity wave dissipation, for example, may play an
important role in the mesospheric heat budget. In the multidimensional
models, the radiative scheme is often simplified and parameterized; the
most simple approach is to assume the “cool-to-space approximation”,
in which it is assumed that exchange of heat between layers can be
neglected in comparison to propagation out to space.
4.6.1 Heating Due to Absorption of Radiation
The absorption of ultraviolet radiation by ozone in the Huggins and

Hartley bands constitutes the principal source of heat in the stratosphere
and mesosphere. The heating rate reaches 10 K/day near the stratopause
on the average, with a maximum of about 15 K/day near the summer
pole. The effect of the Huggins bands in the visible region becomes
important in the lower stratosphere, where the resulting heating rate is

nearly 1 K/day. These numbers are obviously related to the amount of
ozone present, and a change in ozone density would lead to a variation
in the stratospheric and mesospheric temperatures, as well as possible
changes in the locations of the stratopause and mesopause.
The primary source of heat in the thermosphere is provided by
the absorption of solar EUV radiation (10-100 nm). At high
latitudes, particle precipitation and Joule heating can become dominant
contributions. Above about 75 or 80 km, absorption in the Schumann-
Runge continuum of molecular oxygen contributes to the heating of
the atmosphere and begins to play the dominant role in the lower
thermosphere. At 100 km, the mean heating rate is about 10 K/day,
but large variations occur with latitude and season. In the middle and
upper mesosphere, the absorption of solar radiation at Lyman α by O2
also contributes significantly to the heat budget.
The absorption by molecular oxygen in the Herzberg continuum only
contributes slightly to the heating rate in the middle atmosphere, and
can be neglected relative to the effects of ozone in the stratosphere.
There are, however, some O2 bands at wavelengths as long as 760 nm in
the visible. These bands contribute 5-15% of the total heating between
60 and 80 km (Mlynczak and Marshall, 1996).
The meridional distribution of the heating rate due to O2 and O3 ,
shown in Figure 4.25, is based on observations by the Upper Atmosphere
Research Satellite (UARS) during the period November 19-December 19,
1991 (Mlynczak et al., 1999). The corresponding values are determined
by an expression of the form (4.10), i.e.,

cos χ dFs,ν
Q= dν (4.103)
ρcp dz
ν
where the incident solar irradiance is given by
Fs,ν (z, χ) = Fs,ν (∞) exp
⎛ ⎞
∞ ∞
(4.104)
− ⎝σ(O3 ) n(O3 )dz + σ (O2 ) n(O2 ) dz ⎠ sec χ
z z
considering only the effects of absorption by ozone and oxygen. One
may then write:
⎛ ⎞

1 ⎝
Q= n(O3 ) σ (O3 )Fs,ν dν + n(O2 ) σ (O2 ) Fs,ν dν ⎠ (4.105)
ρcp
ν ν
A parameterization of the atmospheric heating rate by O3 and O2 in
different spectral regions has been suggested by Schoeberl and Strobel
(1978), and is presented in Table 4.4.
RADIATION 209
Figure 4.25. Meridional distribution of the atmospheric heating rate (K/day)

between 100 and 0.1 mb during November 19-December 19, 1991 as derived
from measurements by UARS. Effects of absorption of radiation by ozone and
oxygen. After Mlynczak et al.(1999).
Relations (4.103) and (4.105) assume that the incoming solar energy
absorbed by O2 and O3 is entirely converted into heat, and that this
process occurs at the location where the photon absorption takes place.
In reality, the processes involved are more complex. Absorption of solar
radiation often leads to the dissociation of the absorbing molecule, and
the release of heat occurs as the photolysis products recombine or react
with other chemical species (exothermic reactions). Garcia and Solomon
(1983), Apruzese et al. (1984), and Brasseur and Offermann (1986)
noted that the recombination of atomic oxygen provides an important
heating source for the upper mesosphere and lower thermosphere. The
hypothesis behind equations (4.103) and (4.105), which is valid in
the lower atmosphere where collisions are frequent and recombinations
are fast, must therefore be rejected at high altitudes, where the
recombination of atomic oxygen can take place several thousands of
kilometers away from the location where the original photon was
absorbed. In addition, collisions are so infrequent that energy gained
by the photolysis products can be partly released by radiative emission
rather than through quenching. The primary energy source remains
the absorption of solar ultraviolet light by O2 and O3 . A fraction of
the heating is provided (as kinetic energy) through collisions of the
photolysis products with other chemical species; it can be quantified
Table 4.4 Parameterization of Atmospheric Heating Rates (from

Schoeberl and Strobel, 1978)
Spectral region Parameterization
Qc
Chappuis bands = 1.05 × 10−15 exp (−2.85 × 10−21 N3 )
(O3 )
QHu 1
Huggins bands = 4.66 × 103
(O3 ) N3
−7.8 × 102 exp (−1.77 × 10−19 N3 )

−3.88 × 103 exp (−4.22 × 10−18 N3 )
QHa
Hartley region = 4.8 × 10−14 exp (−8.8 × 10−18 N3 )
(O3 )
QSRB 1
Schumann Runge =
bands (O2 ) 0.67N2 + 3.44 × 109 (N2 )1/2
QSRB
If N2 < 1018 cm2 , = 2.43 × 10−19
(O2 )
QSRC 1
Schumann Runge = 0.98 exp (−2.9 × 10−19 N2 )
continuum (O2 ) N2
−0.55 exp (−1.7 × 10−18 N2 )

−0.43 exp (−1.15 × 10−17 N2 )
(Note that N2 and N3 represent the total slant column abundances (molec
cm−2 ), and (O2 ) and (O3 ) the number densities (molec cm−3 ), of O2 and
O3 , respectively. Heating rates are in degrees K/sec.)
as the difference between the energy of the initial incoming photon and
the energy required to photodissociate the absorbing molecule (O2 or
O3 ). Chemical heating is released by exothermic reactions in which
the photolysis products are involved. The most important of these
reactions as well as their rate constants and heat of reactions are listed in
Table 4.5. As noted by Crutzen (1971), Brasseur and Offermann (1986),
and Mlynczak and Solomon (1991), in addition to the recombination of
oxygen atoms, several reactions involving members of the odd-hydrogen
family affect the energy budget of the mesophere.
RADIATION 211
Table 4.5 Reactions, Rate Constants and Heat of Reactions. (From Mlynczak
and Solomon, 1991)
Reaction Rate Constant Heat of Reaction

(kcal/mole)
O + O + M → O2 + M 4.7 × 10−33 (300/T)2 cm6 s−1 –119.40
O + O2 + M → O3 + M 6.0 × 10−34 (300/T)2.4 cm6 s−1 –25.47
O + O3 → O2 + O2 8.0 × 10−12 exp (–2060/T) cm3 s−1 –93.67
H + O2 + M → HO2 + M 5.7 × 10−32 (300/T)1.6 cm6 s−1 –49.10
H + O3 → OH∗ + O2 1.4 × 10−10 exp (–470/T) cm3 s−1 –76.90
OH + O → H + O2 2.2 × 10−11 exp (120/T) cm3 s−1 –16.77
HO2 + O → OH + O2 3.0 × 10−11 exp (200/T) cm3 s−1 –53.27
The energy released by chemical heating is generally lower than the

energy required to dissociate the absorbing molecules. For example,
since the photolysis products of O2 and O3 can be excited species
including O(1 D) and O2 (1 ∆), the amount of energy available for heat is
reduced by the spontaneous emissions of O2 (1 ∆) (and of O2 (1 Σ) excited
by energy transfer from O(1 D)). Mlynczak and Solomon (1993) have
estimated that the heating efficiency associated with the absorption of
solar ultraviolet radiation by ozone is typically 0.95 at 60 km, 0.85 at
70 km, and 0.7 between 70 and 100 km, due to energy loss by O2 (1 ∆)
emission at 1.27 µm. Energy can also be lost by spontaneous emission
of the vibrationally excited OH∗ radical. A detailed evaluation of the
heating efficiency in the middle atmosphere is given by Mlynczak and
Solomon (1993). Chemical heating due to exothermic chemical reactions
as shown in Table 4.5 is typically 1-15 K/day between 80 and 100 km
altitude.
The heating due to absorption in the 2.7 and 4.3 µm bands of carbon
dioxide, and the absorption of ultraviolet by water vapor, play minor
roles in the heating budget. However, the H2 O bands in the visible
and near infrared cannot be neglected in the troposphere, where the
water vapor content is much higher than in the middle atmosphere.
The absorption of solar radiation by NO2 in the region between 300
and 600 nm should also be considered. The mean heating rate depends,
however, on the distribution of NO2 concentration and must also include
the effects of multiple scattering and reflection by the surface and by
clouds. Its contribution can be considerable in the lower stratosphere in
summer, where NO2 is probably present in large amounts. Figure 4.26
shows various contributions to the solar heating rate at equinox (44◦ N)
in the stratosphere and lower mesosphere.
Figure 4.26. Contribution to the stratospheric solar heating ((between 100 and
0.1 hPa)) of UV and near IR absorption by O2 , ozone absorption (Hartley,
Huggins, and Chappuis bands), ) CO2 near infrared absorption, and NO2
absorption. From Mertens et al.(1999).
4.6.2 Cooling by Radiative Emission
The radiative infrared cooling produced by atmospheric carbon

dioxide, water vapor, and ozone, and obtained by solving the radiative
transfer equation (see e.g., Eq. (4.84a,b) and Eq. (4.86)) is represented
in Plate 7 as a function of wavenumber and atmospheric pressure
(corresponding to an altitude ranging from the surface to approximately
60 km). The calculations (Clough and Iacono, 1995) were made
for mid-latitude summer conditions. The 15µm CO2 band (centered
at 667 cm−1 ) produces a strong stratospheric cooling, which reaches
a maximum around 45-50 km (1 hPa pressure level), where the
temperature reaches a high value. A local warming, however, associated
with this band, is calculated at the tropopause. The 9.6 µm ozone band
(centered at 1043 cm−1 ) also contributes to the cooling of the middle and
upper stratosphere, but leads to a warming of the lower stratosphere.
Water vapor contributes to upper tropospheric cooling, especially in the
spectral region near 300 cm−1 . Thermal infrared heating (CO2 , O3 )
and cooling (H2 O) near the tropopause tends to largely cancel, so that
RADIATION 213
changes in greenhouse gas concentrations lead to significant radiative

effects in this region of the atmosphere.
Two-dimensional distributions of the radiative cooling rate can be
obtained when similar calculations are repeated for different latitudes
and seasons. Figure 4.27a shows, for example, the meridional
distribution of the 15 µm CO2 cooling calculated by Mlynczak
et al. (1999) between 100 and 0.1 hPa, on the basis of measurements
by the Upper Atmosphere Research Satellite (UARS) for the period
of November 19-December 19, 1991. A precise calculation must include
the contribution of different isotopes of CO2 , even those which constitute
only 1.5% of the atmospheric abundance of this species. The effect of the
fundamental and hot bands of these isotopes is large in the mesosphere
and upper stratosphere, where a corresponding cooling rate of about
2 K/day is calculated (Williams and Rodgers, 1972).
An analysis based on the work of Dopplick (1972), Williams
and Rodgers (1972), Kuhn and London (1969), and López-Puertas
´
et al. (1992a, b), shows that the maximum cooling rate by CO2 is found
at the winter mesopause, where the temperature is relatively high. Flux
convergence leads to a slight heating near 80 km in summer, that is,
at relatively cold temperatures. The results in the lower thermosphere
are subject to some inaccuracy as a result of the uncertainties in the
relaxation times of vibrational levels.
The second contribution to infrared energetics of the middle
atmosphere is associated with radiative transfer by ozone at 9.6 µm.
This contribution leads to substantial cooling near the stratopause,
but some significant heating in the lower stratosphere. In the lower
stratosphere, the radiative transfer at 9.6 µm leads to a net heating
because ozone absorbs some of the terrestrial emission at this frequency
at altitudes below the ozone layer. The presence of high clouds in the
troposphere can significantly affect the magnitude of ozone longwave
heating, however. Figure 4.27b shows the meridional distribution of the
cooling by O3 based on observations by UARS (Mlynczak et al., 1999).
Finally, the meridional distribution of the infrared cooling due to
the presence of water vapor in the middle atmosphere is shown in
Figure 4.27c (Mlynczak et al., 1999). The longwave emission from
the rotational bands of H2 O in the 250-500 cm−1 range, including the
emission from the upper troposphere (Harries, 1996), is very important
in determining the outgoing longwave radiation.
The infrared cooling in the middle atmosphere (e.g., by the 15 µm
CO2 band) can generally be calculated from the “cool-to-space” term of
Figure 4.27. (a) Merid-

ional distribution of the at-
mospheric cooling rate due
to infrared emission by CO2
(15 µm) between 100 and
0.1 mb during November 19-
December 19, 1991 as de-
rived from measurements by
UARS. (b) Meridional distri-
bution of the infrared cooling
rate by ozone at 9.6 µm
derived from measurements
by UARS during the Novem-
ber 19-December 19, 1991
period. (c) Meridional dis-
tribution of infrared cooling
by water vapor. Values are
in K/day. After Mlynczak
et al. (1999).
RADIATION 215
equation (4.86)
1 dTi∗
Qi = − πBi (∞) (z, ∞) . (4.106)
ρcp dz
If we assume that the temperature T relaxes to some reference
temperature To , then to a first order

∂Bi
Bi (T) = Bi (To ) + (T − To ) (4.107)
∂T To
and
Q(T) = Q(To ) + α (T − To ) , (4.108)
which is the expression for the so-called Newtonian cooling. Coefficient
α defined as

∂B ∂T∗
α = −ρcp (z, ∞) (4.109)
∂T To ∂z
determines a Newtonian relaxation time or a radiative time constant
τrad = 1/α. This time (Figure 4.28) is of the order of 6 days in the
upper stratosphere, 20 days near 30 km altitude, and 100 days or more
in the lower stratosphere (15-20 km). Mlynczak et al. (1999) found that
the radiative relaxation time had increased substantially in the lower
stratosphere in the early 1990s due to the presence of aerosols produced
by the eruption of Mt. Pinatubo. Transport of heat can only occur
if the dynamic time constant, τdyn is less than or comparable to the
radiative time constant, τrad . In the troposphere, where τdyn < τrad , the
temperature field is very dependent on transport, while in the middle
atmosphere, where τrad << τdyn , the thermal structure is strongly
affected by the difference between absorption of ultraviolet radiation
and emission in the infrared. Where the radiative lifetime is less than
a few days, the diurnal variation in solar energy input produces only
small diurnal temperature variations. Variations in the temperature
structure induce modifications in the concentration of chemical species,
particularly ozone, and as a result, influence the transfer of radiation.
Ghazi et al. (1979) showed that this effect leads to a reduction in the
radiative lifetime in the upper stratosphere, and that the total relaxation
coefficient can be written:
α = αNC + αPH (4.110)
i.e., as the sum of the Newtonian cooling and photochemical relaxation
coefficients. Note that the Newtonian cooling approximation requires
that the exchange term in Equation (4.86) can be neglected, that the
temperature perturbations be small, and the temperature effects in the
transmission can be ignored. The approximation is therefore not valid
Figure 4.28. Globally averaged infrared radiative relaxation time deduced from
data provided by UARS and LIMS. Note the increase in the lower stratospheric
relaxation time in response to enhanced aerosol load after the eruption of Mt.
Pinatubo (UARS measurements). From Mlynczak et al. (1999).
in the lower stratosphere and in the troposphere where the exchange of

radiation between layers plays an important role.
4.6.3 Global Energy Budget
Figure 4.25 presents the contribution of the different radiative

processes discussed earlier to the globally averaged heating rate between
the surface and 100 km altitude. In the stratosphere and lower
mesosphere (below 60-70 km), the contribution of radiative processes
dominates the global energy budget. Although dynamics can affect
the energy budget on the local scale (especially during wintertime),
from a global (average) point of view, heating is provided primarily
by ozone absorption of solar ultraviolet radiation in the Hartley band,
and cooling results from the emission of infrared radiation by carbon
dioxide, ozone, and water molecules. Thus, under these conditions,
RADIATION 217
the globally averaged vertical temperature profile can be determined

from radiative equilibrium considerations. In the viscously dominated
thermosphere, above approximately 120-130 km, the energy budget is
also relatively simple: the solar heating due to EUV absorption by O2 ,
N2 , and O (with an efficiency depending on the chemical composition)
is balanced primarily by molecular thermal conductivity. Dynamical
processes provide a significant fraction of the heating above 160 km
altitude.
Figure 4.29. Major contri-

bution to the globally aver-
aged energy budget (K/day)
in three different regions of
the middle and upper at-
mosphere (from the upper
to the lower panel: 90-
210 km, 60-120 km, and 10-
90 km), as derived for June
by the extended Canadian
Middle Atmosphere Model
(CMAM). The global av-
eraged dynamical heating
term represents the sum
of heat advection, grav-
ity wave heating eddy mix-
ing, convective adjustment,
and other dissipative terms.
From Fomichev et al. (2002).
In the transition region between the mesosphere and thermosphere

(60-130 km), the situation is more complex. Diabatic heating results
primarily from the absorption of solar radiation by ozone (Hartley band)
and molecular oxygen (Schumann-Runge continuum), and cooling from
the emission of infrared radiation (15 micron CO2 band). Although, on
the average, these radiative processes dominate the energy balance in
this region of the atmosphere, other mechanisms are also of importance.
These include chemical heating (see Section 4.6.1), which provide a
significant contribution near the mesopause, heat deposition from the
overlying thermosphere by molecular conduction, adiabatic heating by
the atmospheric general circulation, conversion of kinetic energy into
heat from the dissipation of horizontal winds, heating associated with
gravity wave breaking, as well as vertical heat transport by eddy mixing
and sporadically by convective adjustment. In the thermosphere, Joule
heating (kinetic energy dissipated by ion drag — see Chapter 3) and
radiative cooling (which maximizes around 140-150 km) associated with
the 5.3 micron emission by nitric oxide (Kockarts, 1980) also play a
significant role. The major components of the energy budget in different
regions of the atmosphere, as calculated by the model of Fomichev
et al. (2002), are shown in Figure 4.29.
4.7 Photochemical Effects of Radiation
4.7.1 General
Absorption of ultraviolet and visible photons by atmospheric

molecules can induce transitions into electronically excited states, which
may then photodissociate (see Chapter 2). The dissociation products
play a crucial role in the photochemistry of the atmosphere.
In atmospheric problems, it is necessary to estimate the quantitative
value of the photodissociation rate of a molecule A. This is given by:
d (A)
= −JA (A) (4.111)
dt
where n(A) represents the concentration (cm−3 ) of the molecule and
JA the photodissociation probability or coefficient. This coefficient is
also called the photodissociation frequency and is expressed in s−1 . The
inverse of J represents the lifetime of the molecule against photolysis.
For a wavelength interval dλ, the photodissociation coefficient of
molecule A is proportional to the actinic flux qλ dλ, the absorption cross
section σA (λ), and the photodissociation quantum efficiency A (λ). The
actinic flux qλ dλ represents the number of photons per unit area and
RADIATION 219
unit time in the wavelength interval [λ to λ + dλ] and is related to the

energetic flux over the same wavelength interval dΦ = Φλ dλ (where Φ
is defined by Equation (4.12)) by
Φλ λΦλ
qλ = = (4.112)
hν hc
qλ is expressed, for example, in photons cm−2 s−1 nm−1 . When the entire
portion of the solar spectrum over which the molecule can dissociate is
considered (λx − λy ), the coefficient is given by (see also Chapter 2):
λy
JA (z; χ) = A (λ)σA (λ) qλ (λ; z; χ) dλ (4.113)
λx
The spectral distribution of the absorption cross section and the
quantum efficiency of atmospheric molecules is measured in the
laboratory and can vary with temperature. In most cases, the
quantum efficiency is nearly unity but may be less, especially near
the dissociation limit. This limit (maximum wavelength or minimum
frequency) corresponds to the minimum energy required to dissociate
the molecule (see, e.g., Eq. (2.18) in the case of O2 photolysis).
In most cases, the spectral variation of the cross section is sufficiently
gradual that one can integrate Equation (4.113) by simple methods. In
a few cases (Schumann Runge bands of O2 and δ bands of NO) the
spectrum contains a multitude of narrow lines and more sophisticated
methods are needed. A more detailed discussion occurs in Section 4.7.3.
4.7.2 Absorption Cross Sections of the Principal

Atmospheric Molecules
In this Section, we present a brief overview of the measured absorption

cross sections and photodissociation quantum yields which are used for
the calculation of the photolysis rates for the key molecules of the middle
atmosphere. More details can be found in the evaluation of chemical
and photochemical data periodically published by the Jet Propulsion
Laboratory.
4.7.2.1 Molecular Nitrogen

The photodissociation of the most abundant gas in the atmosphere
plays very little role in atmospheric chemistry below 100 km because
the absorption of dissociating radiation by N2 is very weak. However,
production of atomic nitrogen from N2 photolysis via predissociation
does occur following absorption into the Lyman-Birge-Hopfield bands
(a1 πg − X1 Σg transition). The coefficient of photodissociation is about

5 × 10−12 s−1 at the top of the atmosphere and decreases rapidly,
becoming negligible near the mesopause. The solar CIII line at 97.7 nm
plays an important role in this process (see Richards et al., 1981).
4.7.2.2 Molecular Oxygen

The photodissociation of molecular oxygen produces oxygen atoms,
which play a central role in middle atmosphere chemistry. A potential
diagram for the principal electronic states of the O2 molecule was
presented in Chapter 2, and some of the important transitions between
them were mentioned. From the aeronomic viewpoint, the most
important transitions are the X3 Σ− g − A Σu which constitutes the
3 +
Herzberg system (forbidden transition resulting in weak absorption

from 185 to 242 nm) and the X3 Σ− 3 −
g − B Σu which constitutes the
Schumann Runge system. The latter is characterized by a banded
structure from 175 to 200 nm and a continuum from 137 to 175 nm.
The first of these two transitions is a predissociation, that is, a non-
radiative transition to a number of repulsive π states; it thus leads to
the formation of two oxygen atoms in the ground 3 P state. In the second
case (λ < 175nm) one of the atoms is formed in the excited 1 D state. At
shorter wavelengths (λ < 137nm) some diffuse bands occur, but these are
interspersed with a series of windows, which allow some wavelengths to
penetrate relatively deeply into the atmosphere. One of these windows
coincides with the solar Lyman α line, which plays an important role
in atmospheric chemistry. Figure 4.30 shows the general shape of the
spectral distribution of the absorption cross section. At wavelengths less
than 102.8 nm, photoionization can occur.
The absorption coefficient of molecular oxygen has been the subject
of numerous laboratory investigations. In the Herzberg continuum (195-
242 nm), early laboratory measurements indicated an absorption cross
section less than 1.5×10−23 cm−2 , but exhibited significant discrepancies
in the absolute value. Part of the difficulty associated with measuring
this cross section in the laboratory arises from the formation of the Van
Der Waal’s molecule, O4 , for which observational evidence has even been
found in the atmosphere (Perner and Platt, 1980). Balloon observations
of the solar irradiance in the stratosphere led to inferences of the O2
absorption cross section in-situ (Frederick and Mentall, 1982b; Herman
and Mentall, 1982b; Anderson and Hall, 1983), and these have resulted
in values somewhat lower than the early laboratory investigations below
210 nm. The lower cross sections (Figure 4.31) have been confirmed
by later laboratory measurements (Cheung et al., 1986; Jenouvrier
et al., 1986; Yoshino et al., 1988). Note that the Herzberg continuum
RADIATION 221
Figure 4.30. Spectral distribution of the absorption cross section of molecular

oxygen.
Figure 4.31. Absorption cross section of molecular oxygen in the Herzberg

continuum between 194 and 240 nm (From Yoshino et al., 1988).
underlies the Schumann-Runge band region below 205 nm, where it

makes an important contribution to the total O2 cross section (between
195 and 205 nm). The absorption of solar radiation in this spectral
region has important implications in middle atmosphere chemistry,
particularly for the distributions of N2 O and chlorofluorocarbons in the
stratosphere (Froidevaux and Yung, 1982; Brasseur et al., 1983). The O2
photodissociation frequency corresponding to the Herzberg continuum

is about 10−9 s−1 for zero optical depth. In spite of the weakness of
this transition, it becomes dominant in the stratosphere (Figure 4.34).
The rate of ozone production in the stratosphere, therefore, depends
critically on the value of the absorption cross sections in this region.
In the region of the Schumann-Runge bands (175-205 nm), molecular
oxygen is predissociated and provides an important source of O(3 P). The
absorption cross section varies by about 5 orders of magnitude between
175 and 205 nm. The calculation of the photodissociation rate requires
a detailed study of the absorption cross section. Experimental results
obtained in this region have been applied to the atmospheric problem by
several investigators who have proposed various methods to parameterize
the solar absorption and the O2 photolysis in the spectral range
corresponding to these bands (e.g., Fang et al., 1974; Kockarts, 1976;
Allen and Frederick, 1982; Gijs et al., 1996). Laboratory measurements
by Yoshino et al. (1983) and Lewis et al. (1986; 1994) have led to
improved values for Schumann-Runge band spectral parameters, most
notably oscillator strengths and predissociation widths of rotational
lines. These data have been used by Murtagh (1988), Nicolet and Kennes
(1989), Minschwaner et al. (1992), and Kockarts (1994) to construct
temperature dependent line-by-line representations of the Schumann-
Runge bands. Figure 4.32 (right panel) shows the variation of the
absorption cross section of O2 from 175 nm to 205 nm as calculated
by Kockarts (1994). Detailed calculations indicate a photodissociation
frequency of about 10−7 s−1 at zero optical depth (Nicolet et al., 1989).
The determination of the O2 photodissociation frequency in the
Schumann-Runge bands is obtained from a computation including the
complex rotational structure of the band system. A major difficulty
in the calculation of Schumann-Runge band photolysis in the middle
atmosphere, in addition to requirements for high spectral resolution,
arises from the temperature dependence of absorption cross sections.
Polynomial expressions have been derived to reproduce the temperature
variation provided by the line-by-line calculations (Minschwaner et al.,
1992); this kind of approach, however, is generally not feasible to be
implemented in detailed middle atmosphere models, but can be used to
develop more efficient broad-band parameterizations (see Section 4.7.3).
Because of its strong absorption (σ = 1.5 × 10−17 cm2 at 140 nm), the
Schumann-Runge continuum (130-183 nm; see Figure 4.32, left panel)
is only important in the thermosphere, and from the point of view of
photolysis, plays a dominant role only above 90 km. Each photolysis
event leads to the production of two oxygen atoms, one in the ground
3 P state and the other in the excited 1 D state. As shown by Hudson
et al. (1966) and by Lean and Blake (1981), the temperature dependence
RADIATION 223
ABSORPTION CROSS SECTION
ABSORPTION CROSS SECTION

O2
10-17 10-18
10-19
(cm2 molecule-1)
(cm2 molecule-1)
10-20
10-18 10-21
10-22
10-23
10-19
10-24
130 140 150 160 170 175 180 185 190 195 200 205
WAVELENGTH (nm) WAVELENGTH (nm)
1796/fig 2.6a & 2.6b/8.98/dpm
Figure 4.32. Left: Absorption spectrum of O2 in the Schumann-Runge

continuum (from Watanabe et al., 1953). Right: Molecular oxygen absorption
cross sections for a temperature of 220 K in the wavelength region of the
Schumann-Runge bands ((from Kockarts, 1994, using the technique described
by Minschwaner et al., 1993). The band structure is clearly visible and the
effect of the underlying Herzberg continuum is apparent at wavelengths greater
than 190 nm. Kockarts’ approximation can be safely used down to a solar
attenuation of the order of 10−10 .
of the cross section becomes noticeable for wavelengths longer than

160 nm and is mainly due to the redistribution of the population
of rotational states. The Schumann-Runge continuum adjoins the
Schumann-Runge band system at 175 nm and underlies the bands up to
183 nm. This continuum therefore contributes to the absorption cross
section minima in the band system. The photodissociation frequency
in the Schumann-Runge continuum at zero optical depth is (2.5-2.8)
×10−6 s−1 (Zhu et al., 1999), but only approximately 1 × 10−8 s−1 at
90 km for an overhead sun; this illustrates the large optical depth
at these wavelengths even near the top of the middle atmosphere. A
parameterization of the photodissociation frequency as a function of the
O2 column abundance is provided by Zhu et al. (1999).
At wavelengths less than 130 nm (see Figure 4.33), the most important
contribution to oxygen photolysis comes from the solar Lyman α line
at 121.6 nm (σ = 10−20 cm2 ), which penetrates to the mesosphere. The
corresponding photolysis rate at zero optical depth is about 3×10−9 s−1 ,
but its value can change strongly with solar activity. Figure 4.34 shows
the contribution of each spectral region to the photolysis rate of O2 for
an overhead sun. Note the importance of the Herzberg continuum below
65 km (stratosphere and mesosphere), the Schumann-Runge bands from
65 to 90 km (mesosphere), and the Schumann-Runge continuum in the
thermosphere. The Lyman α line plays a secondary role in oxygen
Figure 4.33. Spectral distribution of the absorption cross section of oxygen

from 105 to 135 nm. The cross section exhibits a minimum for the wavelength
corresponding to the solar Lyman α line. After Watanabe et al., (1953).
Figure 4.34. Contribution of each spectral region to the photodissociation of

molecular oxygen as a function of altitude.
photolysis, but becomes much more important for other atmospheric

constituents.
4.7.2.3 Ozone
The energy required to dissociate the ozone molecule being only of
25.4 kcal, (corresponding to a wavelength threshhold of 1180 nm) a large
RADIATION 225
fraction of the solar radiation from the near infrared to the ultraviolet
can, in principle, lead to the photolysis of O3 . It is, however, primarily
in 3 specific spectral regions that the absorption by ozone takes place
(Figure 4.35): the Hartley band (200-310 nm), the Huggins bands (310-
350 nm), and the Chappuis bands (410-750 nm).
The primary absorption by ozone occurs in the Hartley band. The
absorption cross section maximizes at about 250 to 260 nm, where
σ = 10−17 cm2 . The absorption in this band has been studied since
the beginning of the century. The temperature dependence of the
absorption is negligible for wavelengths shorter than about 260 nm.
Vigroux (1953, 1969), however, observed a weak variation of the cross
section with temperature at longer wavelengths, which is probably due
to the superposition of strong predissociating bands on the continuum.
These bands were observed as early as 1929 by Lambrey and Chalonge.
Around 300 nm, the Hartley band becomes weak, and from 310
to 350 nm it blends with the temperature dependent Huggins bands.
Figure 4.35 shows the importance of this temperature sensitivity,
especially at long wavelengths. A careful calculation is required to
account for the details of the spectral structure as a function of
temperature.
Figure 4.35. Spectral distribution of the ozone cross section in the Hartley band
(200-300 nm), Huggins bands (310-350 nm), and Chappuis bands (410-750 nm).
Ozone also absorbs in the visible region via the Chappuis bands
(Figure 4.35). This spectral regime contributes significantly to the
photodissociation of ozone and plays a dominant role in the lower
stratosphere and troposphere (z < 25 km).
Finally, at wavelengths less than 200 nm, the absorption by ozone is
related to the existence of large bands superimposed on a continuum.
This spectral region plays a very minor role in the photochemistry of
the ozonosphere because these wavelengths are absorbed by molecular
oxygen at altitudes far above the ozone layer. It should be noted,
however, that in the region of the O2 Herzberg continuum both ozone
and oxygen contribute to absorption of radiation.
A large number of laboratory measurements have been reported for
the absorption cross sections of ozone (e.g., Inn and Tanaka, 1953;
Vigroux, 1969; Molina and Molina, 1986; Daumont et al., 1992; Yoshino
et al., 1993), and are the basis for the recommended values presented by
WMO (1985) or JPL (1997). High resolution measurements from 195
to 345 nm, and at different temperatures, have been made by Malicet
et al. (1995).
Table 4.6 Theoretical Limits Corresponding to Different Photolysis

Products (nm)
O2 (3 Σg ) O2 (1 ∆g ) O2 (1 Σg +) O2 (3 Σu +) O2 (3 Σ−
u)
O (3 P) 1180 590 460 230 170

O (1 D) 410 310 260 167 150
O (1 S) 234 196 179 129 108
The photochemistry of ozone is reviewed by Wayne (1987) and

Steinfeld et al. (1987). The photodissociation products, O and O2 , are
found in different energy states depending directly on the energy of the
incident photon. Table 4.6 indicates the theoretical limits corresponding
to different dissociation products. Some transitions violate the spin rules
and therefore play a minor role compared to the allowed transitions in
the same spectral region. For wavelengths greater than approximately
320 nm, the following reaction occurs:
JO 3 ; O3 + hν → O(3 P) + O2 (3 Σ−
g) (4.114a)
and, for wavelengths less than 305 nm, the spin allowed fragmentation
channel
J∗O3 ; O3 + hν → O(1 D) + O2 (1 ∆g ) (4.114b)
RADIATION 227
leads mainly to the formation of the electronically excited atom O(1 D).
The formation of O(1 S) is possible for λ < 196 nm.
The calculation of the rate of formation of the O(1 D) atom and the
O2 (1 ∆g) molecule requires exact knowledge of the quantum efficiency
for channel (4.114b), especially near the “fall off” region in the vicinity of
310 nm. Several measurements of the quantum yield for O1 D formation
as a function of wavelength and temperature have been reported in
the literature (e.g., Lin and De More, 1973; Moortgat et al., 1977;
Ball et al., 1993; 1995; Takahashi et al., 1996; 1998), suggesting the
following picture: At wavelengths less than 305 nm, the O(1 D) yield is
approximately constant and equal to 0.9-0.95. In the vicinity of 310 nm,
it decreases rapidly with wavelength, but remains different from zero
at 335 nm (Silvente et al., 1997; Michelsen et al., 1994). In addition
to the spin-allowed photolysis (4.114b) of vibrationally excited ozone
molecules, the spin-forbidden dissociation process

O3 + hν(λ ≤ 411nm) → O(1 D) + O2 3
Σ−
g (4.114c)
is believed to contribute to the formation of O(1 D) at wavelengths larger

than 310 nm (see Figure 4.36, upper panel, and Takahashi et al., 1998).
Figure 4.36 (lower panel) illustrates how the O1 D yield varies with
temperature for a number of selected wavelengths in the 305-325 nm
range.
Figure 4.37 shows the relative contribution of each band to the ozone
photolysis frequency as a function of altitude. It should be noted that
the value of the ozone photodissociation coefficient depends critically
on the absorption by ozone itself, introducing a nonlinear coupling as a
function of altitude. The calculation of JO3 must include the effects of
molecular scattering and albedo (see Figure 4.38).
4.7.2.4 Water Vapor

The absorption spectrum of water vapor consists of a continuum from
145 to 186 nm, diffuse bands from 69 to 145 nm, and a continuum below
69 nm. A maximum in the intensity of a diffuse band (121.8 nm) is
located in the vicinity of the Lyman-α line (121.6 nm). Figure 4.39
shows the shape of the absorption cross section of water vapor above
120 nm. Measurements of this cross section were reported by Watanabe
and Zelikoff, 1953, Thompson et al. (1963), Laufer and McNesby (1965),
Schurgers and Welge (1968), Yoshino et al. (1996b), and Cantrell
et al. (1997). In particular, the value of σ (H2 O) at Lyman α has been
determined to be about 1.4 ×10−17 cm2 . Another noteworthy feature is
the rapid decrease in the cross section longward of 175 nm. Beyond this
wavelength, the penetration of solar radiation is determined by the O2
Figure 4.36. Top panel: Contributions of different processes to the O(1 D)

quantum yield as a function of wavelengths between 305 and 329 nm. The solid
and broken lines represent the quantum efficiencies measured at 295 K and
227 K, respectively. Shaded region I shows the contribution of the hot band
excitation process leading to the formation of O(1 D) via Reaction (4.114b) at
295 K. Region II (in black) represents the contribution of the spin-forbidden
dissociation (4.114c). Region III corresponds to the O(1 D) formation via
Reaction (4.114b) following excitation of ozone at ground vibration state.
Bottom panel: O(1 D) quantum efficiency as a function of temperature at
305, 308, 310 and 312 nm. Laboratory measurements are compared to
recommendations (dashed lines) by JPL (1997). From Takahashi et al. (1998).
RADIATION 229
50
40
ALTITUDE (km) Chappuis
λ<240nm
total
30
Huggins
20
Hartley
10
10-6 10-5 10-4 10-3 10-2
O3 PHOTODISSOCIATION COEFFICIENT (s-1)

2300.fig 2.22.cc
Figure 4.37. Contributions of different absorption bands to the total

photodissociation frequency of ozone for overhead sun. Adapted from Nicolet
(1980).
Figure 4.38. Relative effects of scattering on the photolysis rate of ozone from
the ground to 60 km for several solar zenith angles. Surface albedo = 0.3.
Schumann-Runge bands. Special attention is therefore required when

calculating JH2 O below about 60 km (see Section 4.7.3).
Figure 4.40 shows the contributions of the Lyman α line and
the Schumann-Runge bands to the vertical distribution of JH2 O .
Photodissociation in the middle atmosphere occurs at wavelengths
above about 100 nm, while photoionization occurs for wavelengths
less than 98 nm. In the mesosphere and thermosphere, water vapor
is photodissociated mainly by the solar Lyman α line, leading to a
substantial fraction of the hydrogen atom production at these altitudes.
In the lower mesosphere and upper stratosphere, the continuum

absorption of water vapor occurs in the domain of the O2 Schumann
Runge bands. The photolysis frequency at zero optical depth varies
from 4.0 to 6.5 × 10−6 s−1 depending on the level of solar activity.
Figure 4.39. Spectral distribution of the absorption cross section of water vapor
((H2 O),
) methane (CH4 ), hydrogen peroxide (H2 O2 ), and the peroxy radical
(HO2 ).
Figure 4.40. Contributions of different spectral regions to the photodissociation

of water vapor for an overhead sun.
RADIATION 231
The products of water photolysis vary with the energy of the incoming
radiation. At wavelengths larger than 147 nm (continuum), they are
primarily hydrogen atoms and hydroxyl radicals in their fundamental
electronic state (X2 π). At wavelengths shorter than 147 nm (diffuse
bands), hydrogen molecules together with oxygen atoms are believed to
be produced with a quantum efficiency between 105 and 145 nm which
could vary from 2.5 to 23% according to Rebbert et al. (1972) and is
estimated as 11% according to Stief et al. (1975). At wavelengths less
than 136 nm, the OH radical produced could also be in its electronically
excited A2 Σ+ state, with a yield of perhaps 10% (Nicolet, 1978). Thus,
one has to consider, in the mesosphere, 3 channels that are initiated by
the solar Lyman-α line:

H2 O + hν (λ = 121.6 nm) → H + OH X2 π (4.115a)

→ H2 + O 1
D (4.115b)

→ H + OH A2 Σ+ (4.115c)
with very uncertain quantum efficiencies. In the stratosphere, the major

channel is provided by

H2 O + hν (λ < 200 nm) → H + OH X2 π (4.115d)
4.7.2.5 Hydrogen Peroxide (H

H2 O2 )
The absorption cross section of hydrogen peroxide is shown in
Figure 4.39. In the spectral domain which is important in the
stratosphere, absorption cross section measurements have been made by
Schurgens and Welge (1968) from 120 to 200 nm, Holt et al. (1948) and
Holt and Oldenberg (1949) from 185 to 253 nm, and Urey et al. (1929)
from 215 to 380 nm. More recent measurements include those of Lin
et al. (1978), Molina and Molina (1981), Nicovich and Wine (1988), and
Vaghjiani and Ravishankara (1989). The cross sections vary slightly
with temperature. The photodissociation quantum efficiency is believed
to be unity.
4.7.2.6 Methane (CH

H4 )
Methane only photodissociates in the upper part of the middle
atmosphere because its absorption cross section becomes very weak at
wavelengths longer than 145 nm (Figure 4.39). The most intense part of
the spectrum is located below 130 nm, where the cross section is about
1.9 ×10−17 cm2 . The Lyman α line dominates the photolysis rate with
a mean value at zero optical depth of about

∞
JCH 4
(Lyα) = 5.5 × 10−6 s−1
Like water vapor, the photolysis rate varies over the solar cycle due to
the variation in Lyman α flux. The different spin-allowed dissociation
pathways after excitation at the Lyman α wavelength are
CH4 + hν(Lyα) → CH3 + H (39%) (4.116a)
→ CH2 + H2 (52%) (4.116b)
→ CH2 + H + H (1%) (4.116c)
→ CH + H + H2 (8%) (4.116d)
with the relative contribution of each channel quoted in parentheses and
based on the recommendation by Brownword et al. (1997). These quoted
efficiencies are different from the values suggested earlier by Slanger and
Black (1982).
4.7.2.7 Carbon Dioxide (CO2 )

The spectral distribution of the absorption cross section of CO2
(Figure 4.41) has been measured by several experimenters (Inn et al.,
1953, from 106 to 175 nm; Ogawa, 1971, from 175 to 195 nm; Shemansky,
1972, above 195 nm; Lewis and Carter, 1983, from 120 to 195 nm;
Yoshino et al., 1996a, from 118 to 175 nm). Carbon dioxide dissociates
only at relatively high altitudes, where solar Lyman α and Schumann
Runge continuum radiation is present. The corresponding photolysis
rates at the top of the atmosphere are (Nicolet, 1980):
J∞
CO2 (Lyα) = 2.20 × 10
−8 −1
s
J∞
CO2 (SRC) = 9.28 × 10
−8 −1
s
and, for T = 300K,
J∞
CO2 (SRB) = 1.90 × 10
−9 −1
s
These values are also subject to solar activity variations, especially
the contribution of Lyman α and the Schumann Runge continuum.
The complete calculation of the photolysis rate including absorption
by molecular oxygen indicates that CO2 photolysis is due primarily to
Lyman α in the upper mesosphere, and to the Schumann Runge bands
in the mesosphere and stratosphere. Due to the strong temperature
dependence of the CO2 cross section in the O2 Schumann Runge bands
region (DeMore and Patapoff, 1972), the value of JCO2 decreases at lower
temperatures (see Nicolet, 1980).
RADIATION 233
Figure 4.41. Spectral distribution of the absorption cross section of carbon

dioxide for three temperatures: 200 K, 300 K, and 370 K. From Lewis and
Carter (1983).
4.7.2.8 Formaldehyde (CH

H2 O)
The absorption cross section of formaldehyde (Figure 4.42) has been
studied by several investigators since the experiments of Henri and Schou
(1928), in particular by Bass et al. (1980), Moortgat et al. (1980; 1983),
and Cantrell et al. (1990). It has long been recognized (Herzberg, 1931;
Norrish and Kirkbride, 1932) that the photolysis of CH2 O can lead to
two different product paths:
(JH−CHO ); CH2 O + hν → H + CHO (4.117a)
(JH2 −CO ); CH2 O + hν → H2 + CO (4.117b)
The relative importance of these processes has been studied by several

investigators. According to Moortgat et al. (1983), the quantum yield
for the production of H2 and CO is pressure and temperature dependent
for wavelengths longer than 330 nm. Figure 4.42 also shows the quantum
efficiency associated with each process.
4.7.2.9 Nitric Oxide (NO)

The absorption spectrum of nitric oxide was observed more than a
century ago, but the atmospheric importance of this chemical species
has been stressed only in recent decades. Cieslik and Nicolet (1973)
and Cieslik (1977), for example, used the measurements of Miescher and
colleagues at the University of Basel to calculate the photolysis frequency
of NO in the middle atmosphere. High resolution measurements of
Figure 4.42. Spectral distribution of the formaldehyde cross section and

quantum yields for the H + HCO and H2 + CO channels.
nitric oxide absorption have also been reported by Murray et al. (1994).
The ultraviolet absorption spectrum of NO exhibits four band systems
(Marmo, 1953) defined by the following transitions (cf. Figure 4.43):
γ bands : A2 Σ → X2 π
β bands : B2 Π → X2 π
δ bands : C2 Π → X2 π
bands : D2 π → X2 π
Calculation of the photolysis rate of nitric oxide must include the
predissociation process which occurs at wavelengths less than 192 nm.
In particular, predissociation occurs (Figure 4.43) in the δ bands, the
β bands (v > 6) and the bands (v > 3). Callear and Pilling (1970
a,b) indicate that emission dominates predissociation in the bands. At
zero optical depth, the contribution of the γ bands is small compared to
those of the δ and β bands.
In the mesosphere and stratosphere, the β bands of quantum number

v greater than 9, are almost totally attenuated and the other β bands
contribute less than 10 percent to the photolysis rate. Thus, only the
two bands δ (0-0) at λ ≈ 190.9 nm and δ (1-0) at λ ≈ 182.7 nm
need be considered in calculating the photolysis rate in the middle
atmosphere. JNO has been studied by Cieslik and Nicolet (1973), Park
(1974), Frederick and Hudson (1979), Nicolet and Cieslik (1980), and
Minschwaner and Siskind (1993). It should be noted that the derived
values are sensitive to the different values assumed for the oscillator
RADIATION 235
Figure 4.43. Energy levels of nitric oxide and corresponding transitions
strength and the solar flux. Minschwaner and Siskind (1993) indicate
that for zero optical depth
JNO (δ [0 − 0]) = 2.6 × 10−6 s−1
and
JNO (δ [1 − 0]) = 3.2 × 10−6 s−1
The calculation of the photolysis rate at all atmospheric levels requires
detailed knowledge of the solar spectrum. The primary attenuation
of the solar flux in the δ (0-0) band is due to the (5-0) Schumann-
Runge band of O2 ; the attenuation in the δ (1-0) band is due to the
(9-0) and (10-0) bands of O2 . Thus, the determination of JNO requires
consideration of the ensemble of rotational lines of each NO and O2 band
system. Further, the effect of temperature is important in the intensities
of both the NO and O2 bands. The vertical temperature structure must
therefore be considered for aeronomic studies. The attenuation of the
δ (1-0) band is more rapid than that of the δ (0-0) band; thus the
photolysis rate of NO in the stratosphere and lower mesosphere depends
Figure 4.44. Detailed structure of the (5-0)) band of the Schumann-Runge

system of O2 in the spectral range of the δ (0-0) band of nitric oxide. From
Cieslik and Nicolet (1973).
critically on the behavior of the δ (0-0) band and its attenuation by

the (5-0) band of O2 . Figure 4.44 shows the detailed structure of the δ
(0-0) band of nitric oxide together with that of the (5-0) band of the O2
Schumann-Runge system.
Frederick et al. (1983) have shown that the inclusion of the opacity
provided by nitric oxide itself reduces the calculated dissociation at
lower altitude. Since thermospheric NO is known to increase with
increasing solar activity, a modulation of the solar penetration with
the solar cycle should occur at the wavelengths which photodissociate
NO. The direction of this change is, however, opposite to that of the
solar irradiance, which increases with solar activity. For example,
Minschwaner and Siskind (1993) found that for solar zenith angles
greater than 60◦ , JNO below 100 km may be smaller during solar
maximum compared to solar minimum (see Minschwaner and Siskind,
1993). However, the self absorption is sensitive to the δ-band
predissociation widths which are highly uncertain.
4.7.2.10 Nitrous Oxide (N

N2 O)
The absorption spectrum of N2 O has been measured several times
(e.g., Selwyn et al., 1977; Hubrich and Stuhl, 1980; Mérienne
´ et al.,
1990) since the detection of this molecule by Leifson in 1926. The
absorption coefficient above 260 nm has been shown to be very small,
which implies that N2 O must be quite inert in the troposphere. The
absorption spectrum is shown in Figure 4.45.
RADIATION 237
Figure 4.45. Spectral distribution of the absorption cross section for various
nitrogen oxides at 220 K.
As shown by Johnston and Selwyn (1975), the cross section of

N2 O varies strongly with temperature. The quantum yield for
photodissociation is unity, and the products are N2 and O(1 D). The
atmospheric photolysis rate comes predominantly from the absorption
of solar radiation in the O2 Herzberg continuum and Schumann Runge
bands. J∞ N2 O ≈ 9 × 10
−7 s−1 for λ > 175 nm.
4.7.2.11 Nitrogen Dioxide (NO2 )

The absorption spectrum of NO2 has been studied in various spectral
domains since its detection by Bell in 1885 (see, e.g., Bass et al.,
1976; Davidson et al., 1988; and Schneider et al., 1987). Figure 4.45
shows the shape of the absorption cross section of NO2 from 190 to
410 nm. The photolysis quantum efficiency decreases with wavelength
near the photolysis limit. JPL (1997) recommends quantum yields
of 1.0 at wavelengths less than 285 nm, 0.997 at 300 nm, 0.987 at
350 nm, 0.695 at 400 nm, 0.018 at 420 nm, and 0.0 beyond 424 nm.
Because the absorption cross sections vary slightly with temperature,
the photodissociation rate decreases by about 10 percent when the
temperature changes from 300 to 235 K. It should be noted that this
spectral region is quite sensitive to the effects of molecular scattering
and albedo (Figure 4.46a,b) so that the precise value of JNO2 depends
strongly on geographic conditions and on the albedo of the troposphere.
Figure 4.46. ((a)) Vertical distribution of the NO2 photolysis rate for several
values of the solar zenith angle (absorption of direct solar radiation only). (b)
Relative effect of scattering on the photolysis rate of NO2 . Earth albedo = 0.3.
4.7.2.12 Nitrogen Trioxide (NO3 )

The spectroscopy of NO3 has been reviewed by Wayne et al. (1991).
The absorption spectrum of this radical (see Figure 4.45) exhibits several
predissociating bands first observed in 1962 by Ramsay. Laboratory
studies reviewed by Johnston et al. (1996) provided quantitative values
of the photodissociation parameters. The quantum efficiency for each of
the two possible pathways
NO3 + hν(λ < 580 nm) → NO2 + O (4.118a)
and
NO3 + hν(λ > 580 nm) → NO + O2 (4.118b)
RADIATION 239
is still not completely established, despite measurements by Graham and

Johnston (1978), Magnotta and Johnston (1980), and more recently by
Orlando et al. (1993). Laboratory studies indicate, however, that the
first process is dominant and that it may even occur at wavelengths
longer than the limit of 580 nm at elevated temperatures. The quantum
yield of the second channel is no more than 25 percent from 520 to
640 nm, and is very small at longer wavelengths. The values of the
photolysis frequencies determined by Johnston et al. (1996) can be used
for most atmospheric applications:
JaNO3 = 0.156 s−1
JbNO3 = 0.0201 s−1
4.7.2.13 Dinitrogen Pentoxide (N

N2 O5 )
The absorption cross sections of N2 O5 (Figure 4.45) were first
measured by Jones and Wulf (1937). Since then, other measurements
have been performed by Johnston and Graham (1974), Graham (1975),
Yao et al. (1982), and by Harwood et al. (1993). The latter studies
found significant temperature dependence at wavelengths longer than
about 280 nm.
4.7.2.14 Nitrous Acid (HNO2 )

The absorption spectrum of HNO2 (Figure 4.47) has been studied
by Stockwell and Calvert (1978), and by Vasudev (1990) and Bongartz
et al. (1991). More detailed experimental work is needed, but it is clear
that the photolysis of HNO2 is fast, even at low altitudes.
4.7.2.15 Nitric Acid (HNO3 )

The absorption by HNO3 (Figure 4.47) is a continuum from 220
to 300 nm, and was observed as early as 1943 by Dalmon. Johnston
and Graham (1973) and Biaumé´ (1973) measured the absorption cross
section from 180 to 330 nm. The data obtained by these studies are in
general agreement between 205 and 300 nm. Burkholder et al. (1993)
noted that discrepancies remain below 205 nm and above 300 nm. In
addition, they recommended that the temperature dependence observed
for wavelengths greater than 280 nm be taken into account when
calculating JHNO3 . Johnston et al. (1974) recommend a quantum yield
value of approximately 1 for the OH + NO2 channel. However, at
193 nm, Turnipseed et al. (1992) report this quantum yield to be 0.33 at
193 nm, with the O + HNO2 channel being important at this wavelength.
Figure 4.47. Spectral distribution of the absorption cross section for HNO2 ,
HNO3 , and HNO4 at 220 K.
The quantum yield for OH production could be as low as 0.75 at 248 nm
(Schiffman et al., 1993). More detailed investigations are needed.
4.7.2.16 Peroxynitric Acid (HO2 NO2 or HNO4 )

The absorption spectrum of HO2 NO2 (Figure 4.43) was observed by
Graham et al. (1978) and by Cox and Patrick (1979). These two studies
are in good agreements from 205 to 260 nm, but at 195 nm the results
of Graham are twice as large as those of Cox and Patrick. Molina and
Molina (1981) presented a spectral distribution of the absorption cross
section which exhibits much smaller values than the previous studies
at all wavelengths. The cross sections are temperature independent
between 253 K and 298 K (Singer et al., 1989). The quantum yield
of the photolysis was measured at 248 nm by McLeod et al. (1988).
They report an efficiency of one third for the OH + NO3 channel and
two thirds for HO2 + NO2 .
4.7.2.17 The Halocarbons

Since the important effects of chlorocarbon compounds on atmo-
spheric ozone were pointed out by Molina and Rowland (1974) and
by Rowland and Molina (1975), the absorption spectra of the various
chlorocarbons that provide the source of chlorine to the atmosphere have
RADIATION 241
Figure 4.48. Upper left panel

(a) Spectral distribution of the
absorption cross sections of CCl4 ,
CFCl3 (CFC-11), CF2 Cl2 (CFC-
12),
) and CH3 Cl at T=220 K.
Lower left panel (b) Spectral dis-
tribution of the absorption cross
section of CH3 CCl3 , CFC-113,
CFC-114, CFC-115 at T=220 K.
Upper right panel (c) Spectral
distribution of the absorption
cross section of several hydro-
genated chlorofluorocarbons
(HCFC-22, HCFC-123, HCFC-
124, HCFC-141b, and HCFC-
142b) at T = 220 K.
been measured by several groups, including Rowland and Molina (1975),

Hubrich et al. (1977), Chou et al. (1977; 1978), Van Laethem-Meurée
et al. (1978a,b), Hubrich and Stuhl (1980), and Simon et al. (1988).
Spectral distributions of the cross sections for several chlorofluorocar-
bons (CFCs) and hydrogenated chlorofluorocarbons (HCFCs) are given
in Figure 4.48a,b,c.
4.7.2.18 Methyl Bromide (CH

H3 Br), Bromoform (CHBr3 ), and
Halons
The measurements of the CH3 Br absorption cross sections (Fig-
ure 4.49a by Robbins (1976), Molina et al. (1982), and Gillotay and
Simon (1988) are in very good agreement. In the case of CHBr3 , the
measurements by Gillotay et al. (1989) and Moortgat et al. (1993) are
in excellent agreement in the spectral region where they overlap. The
Figure 4.49. (a) Spec-

tral distribution of the ab-
sorption cross sections of
methyl bromide (CH3 Br) and
methyl bromoform (CHBr3 )
(T=220 K). (b) Spectral dis-
tribution of the absorption
cross sections of halons–1211
and –1301 (T=220 K).
cross sections for halon 1211 (CF2 ClBr) and halon 1301 (CF3 Br) are
given in Figure 4.49b.
4.7.2.19 Chlorine and Bromine Oxides

The absorption cross section of chlorine monoxide ClO is characterized
by a continuum (which is independent of temperature) at wavelengths
less than 265 nm and by a band structure (which is extremely
temperature dependent) at higher wavelengths.
The different laboratory measurements (see e.g., Watson, 1977) are
usually in good agreement. A detailed theoretical study of ClO was
presented by Langhoff et al. (1977a,b). The photodissociation of ClO
accounts for 2 to 3 percent of the total chemical destruction of this
radical (by atomic oxygen and nitric oxide).
RADIATION 243
The spectrum of chlorine dioxide (OClO) is characterized by a well

developed progression of bands covering parts of the ultraviolet and
visible regions (280 to 480 nm). Cross sections were measured by Wahner
et al.(1987) and by Hubinger and Nee (1994). The lifetime of OClO
against photolysis is only a few seconds.
The absorption cross sections of the chlorine peroxyl radical (ClOO)
have been measured between 220 and 280 nm by Mauldin et al.(1992).
The maximum absorption takes place around 246 nm.
The absorption cross section of BrO is represented by a banded
spectrum in the 290-380 nm range with the strongest absorption
feature around 338 nm (Wahner et al., 1988). Values are temperature
dependent. The lifetime associated with the photodecomposition of BrO
is of the order of 20 seconds at the Earth’s surface (solar zenith angle of
30◦ ) (Cox et al., 1982).
The spectrum of OBrO has a remarkable similarity with that of OClO
with highly structured absorption bands. (Rattigan et al., 1994).
4.7.2.20 Dichlorine Peroxide (Cl2 O2 )

The absorption cross sections of Cl2 O2 (or ClOOCl) have been
measured by different investigators (e.g., DeMore and Tschuikow-Roux,
1990; Cox and Hayman, 1988; Permien et al., 1988; Burkholder et al.,
1990; Molina et al., 1990; and Huder and DeMore, 1995) between 190
and 450 nm (see Figure 4.50). Values beyond 360 nm, however, are very
uncertain. The quantum yield for the photodissociation is unity for the
Cl + ClOO channel (Molina et al., 1990).
Figure 4.50. Spectral distribution of the absorption cross sections of various

inorganic chlorine compounds (T=220 K).
4.7.2.21 Chlorine Nitrate (ClONO2 ) and Bromine Nitrate

(BrONO2 )
The absorption cross sections of ClONO2 (Figure 4.50) have been
measured by Molina and Molina (1979) and by Burkholder et al. (1994).
The values are in good agreement and exhibit a temperature variation.
The nature of the photo-fragments has been investigated in several
laboratory studies. As concluded by JPL (1997), the quantum efficiency
Φ1 for the following channel
ClONO2 + hν → Cl + NO3 (4.119a)
is equal to approximately 0.6 for wavelengths less than 308 nm, and to
1.0 for wavelengths longer than 364 nm. Between these two boundaries,
the efficiency for this channel is expressed by 7.143 × 10−3 λ − 1.60, if λ is
the wavelength expressed in nm. The quantum yield Φ2 for the second
channel
ClONO2 + hν → ClO + NO2 (4.119b)
is given by Φ2 = 1 − Φ1 .
Bromine nitrate possesses a similar spectrum (Figure 4.51) to that
of chlorine nitrate. It is, however, shifted toward slightly longer
wavelengths such that the photolysis rate of BrONO2 is faster than
that of ClONO2 . Like chlorine nitrate, the absorption cross sections are
temperature dependent. They have been measured in the laboratory
by Spencer and Rowland (1978) and more recently by Burkholder
et al. (1995). Nickolaisen and Sander (1996) have shown that the
quantum yields should be approximately 0.71 for the
BrONO2 + hν → BrO + NO2 (4.120a)
channel, and 0.29 for
BrONO2 + hν → Br + NO3 (4.120b)
channel.
4.7.2.22 Hydrochloric (HCl), Hydrobromic (HBr), and

Hydrofluoric (HF) acids
The absorption cross sections (Figure 4.50) for hydrochloric acid (or
hydrogen chloride) HCl were measured by Vodar (1948), Romand and
Vodar (1948), and Romand (1949), and later by Inn (1975). The latter
measurements extend from 140 to 220 nm, leading to a photodissociation
frequency of about 3 × 10−6 s−1 at zero optical depth. To calculate JHCl
in the atmosphere, the Schumann Runge bands and continuum must be
considered, as well as the Herzberg continuum.
Measurements of the spectrum of hydrobromic acid or hydrogen
bromide HBr have been presented by Goodeve and Taylor (1935), and
RADIATION 245
Figure 4.51. Spectral distribution of the absorption cross sections of various

inorganic bromine compounds (T=220 K).
Romand (1949). JHBr at zero optical depth is about 5 × 10−6 s−1 , but
it is comparable to the rate of reaction with the OH radical in the
stratosphere.
The absorption spectrum of hydrofluoric acid or hydrogen fluoride HF
indicates that its photolysis can be completely neglected in the middle
atmosphere.
4.7.2.23 Hypochlorous (HOCl) and Hypobromous (HOBr) acids

The spectral distribution of the absorption cross section of HOCl
was measured by Molina and Molina (1978), Knauth et al. (1979),
Mishalanie et al. (1986), Permien et al. (1988), and Burkholder (1993).
Discrepancies exist between these measurements, but the values of
Burkholder (Figure 4.50) are recommended. The quantum yield for the
Cl + OH channel is close to unity.
The absorption cross sections of HOBr were measured by Orlando
and Burkholder (1995), Deters et al. (1996), Benter et al. (1995), and
Rattigan et al. (1996). The measurements agree reasonably well between
240 and 400 nm, except in the case of Rattigan et al., where the measured
cross sections are roughly 50% larger between 300 and 400 nm. Their
spectrum (Figure 4.51) also exhibits an absorption tail beyond 400 nm,
so that the photolysis rate is a factor 3 higher than if calculated with
the cross sections derived from the other studies.
4.7.2.24 BrCl
Absorption cross sections for BrCl (Figure 4.51) have been measured
by Maric et al. (1994). They vary with temperature. The resulting
photolysis rate is very rapid because of the large cross section at visible
wavelengths.
4.7.3 Numerical Calculation of Photodissociation

Coeffi cients
The photodissociation frequencies for several important trace species

are shown in Figure 4.52 for specific conditions (winter at 40◦ latitude,
solar maximum). The determination of these photolysis rates J at
altitude z for solar zenith angle χ is accomplished by the numerical
methods described below.
a) When the absorption spectrum is a continuum, thus varying slowly
with wavelength, the spectrum is divided into discrete intervals and a
mean value of σi and i is assigned to each. Equation (4.113) can then
be written as a sum of the intervals i of wavelength δλi :

J (z,χ) = i σi Qi (z, χ) (4.121)
i
Figure 4.52. Photodissociation rates of several important atmospheric species

for winter, solar maximum, 40◦ latitude. From the model of Garcia and
Solomon (1983).
RADIATION 247
where

Qi = qλ dλ (4.122)
δλi
is the solar actinic flux within spectral interval i at altitude z and solar
zenith angle χ.
b) The absorption spectra of some molecules exhibit a structure of
discrete narrow bands. Examples include NO and O2 in the Schumann
Runge bands (175-205 nm). In these cases, it is not possible to determine
a mean value σi of the cross section for use in a broad spectral interval.
The generalized expression for atmospheric absorption Eq. (4.113) must
then be considered, and an integration performed over spectral intervals
smaller than the width of the bands. Such a calculation is often
computationally expensive, and approximate methods are often used
to reduce computation time. Two aspects must be considered: on one
hand, the rate of photolysis of O2 must be computed, and on the other,
the transmission of solar radiation including O2 absorption must be
considered in order to calculate the photodissociation coefficients in the
Schumann Runge bands for constituents such as H2 O, CO2 , N2 O, HNO3 ,
etc. NO must be treated separately because of the very narrow width
of its lines (Doppler widths of about 0.1 cm−1 ). The Schumann Runge
region can be analyzed by dividing it into 20 intervals beginning at each
O2 rotational band. The first band, (19–0) is found at 175.345 nm, while
the last occurs at 204.08 nm. Other intervals can be adopted, notably
the classical division in 500 cm−1 bands.
One method which leads to a simple treatment of the penetration of
radiation in the Schumann Runge bands and the photolysis rate of O2
involves establishing an equivalent cross section of O2 in the interval
δλi and then applying the methods for continuum absorption. Thus,
for each interval δλi over which the quantum yield is assumed to be
unity and the ozone transmission Tr,O3 is not varying significantly with
wavelength, the O2 photolysis frequency

JO2 (δλi ; z, χ) = σO2 (λ) q∞ (λ) Tr,O2 (λ; z, χ) Tr,O3 (λ; z, χ) dλ
δλi
(4.123)
can be approximated as
JO2 (δλi ; z, χ)
(4.124)
σeff,i (δλi ; z, χ) Qi,∞ (δλi ) TSRB
r,O2 (δλi ; z, χ) Tr,O3 (δλi ; z, χ)
where the effective O2 cross section for wavelength interval δλi is
δλi σO2 (λ) q∞ (λ) Tr,O2 (λ; z, χ) dλ
σeff,i (δλi ; z, χ) = (4.125)
δλi q∞ (λ) Tr,O2 (λ; z, χ) dλ
the effective O2 transmission in a δλi Schumann-Runge band interval is

δλi q∞ (λ) Tr,O2 (λ; z, χ) dλ
TSRB
r,O2 ,i (δλi ; z, χ) = (4.126)
δλi q∞ (λ) dλ
and the solar actinic flux at the top of the atmosphere, integrated over
wavelength interval δλi is

Qi,∞ (δλi ) = q∞ (λ) dλ (4.127)
δλi
Note that if the solar flux q∞ (λ) varies slowly within interval δλi , the
effective O2 transmission TSRB
r,O2 ,i is equivalent to the reduction factor Ri
introduced by Kockarts (1976).
For compounds X other than O2 (or NO) for which the absorption
cross section varies slowly with wavelength (continuum) over the
Schumann-Runge region (e.g., H2 O, CH4 , CO2 ), the photodissociation
frequency over interval δλi can be derived from
JX (δλi ; z, χ) = σX (δλi ) Qi,∞ (δλi ) TSRB
r,O2 ,i (δλi ; z, χ) Tr,O3 ,i (δλi ; z, χ)
(4.128)
The solar flux at altitude z and for solar zenith angle χ, integrated over
wavelength interval δλi is simply
Qi (δλi ; z, χ) = Qi,∞ (δλi ) TSRB
r,O2 ,i (δλi ; z, χ) Tr,O3 ,i (δλi ; z, χ) (4.129)
Since a detailed treatment of radiative transfer in the Schumann-
Runge bands is often computationally prohibitive when applied to
photochemical models, the effective O2 cross section σeff and atmo-
spheric transmission TSRBr,O2 are often pre-calculated through detailed
line-by-line integrations, using expressions similar to those shown above.
Parameterizations have also been developed (Hudson and Mahle, 1972;
Kockarts, 1976; 1994; Allen and Frederick, 1982; Nicolet and Kennes,
1989; Gijs et al., 1997) in which these quantities are expressed as a
function of the O2 slant column
∞
N (O2 ; z, χ) = sec χ (O2 ) dz (4.130)
z
if (O2 ) is the O2 number density at altitude z . Kockarts (1994), for
example, expressed effective parameters through polynomial expansions
that account for temperature variations, while Gijs et al. (1997) express
the effective O2 cross section for each chosen spectral interval by
ln [σeff (N (O2 ))] = A (N (O2 )) [T (N (O2 )) − 220 K] + B (N (O2 ))
(4.131)
where T(N(O2 )) is the local temperature at the altitude corresponding
to the O2 slant column N(O2 ). Parameters A and B are expressed as a
function of ln (N(O2 )) using Chebeshev polynomial fits. The effective
RADIATION 249
O2 transmission is then derived as a function of N(O2 )) by noting that

dTSRB
O2
= −σeff (O2 ) (4.132)
dN(O2 )
An alternative procedure for calculating O2 transmission and
photolysis in the Schumann-Runge bands involves the use of opacity
distribution functions (Fang et al., 1974; Minschwaner et al., 1993). The
method is analogous to the k-distribution techniques used in infrared
radiative transfer (e.g., Lacis and Oinas, 1991). A single set of O2 cross
sections are used for both transmission (or solar actinic flux Qi ) and
photolysis frequency for O2 within a given vibrational band i:

6
Qi (z, χ) = Qi (∞)e−ττi (O3 ;z,χ) wj e−σj (O2 ) N(O2 ) (4.133)
j=1

6
−τi (O3 ;z,χ)
Ji (O2 ; z, χ) = Qi (∞)e wj σj (O2 )e−σj (O2 ) N(O2 ) (4.134)
j=1
where σj are determined by the distribution of cross section magnitudes

in each band and wj are weighting factors. In Zhu et al. (1999),
the weighting factors are determined by Gauss-Legendre quadratures,
which reduces Qi and Ji to 4-term summations. One advantage of
this formulation is that it permits calculation of overlapping of complex
spectra, such as occurs in the NO δ bands, through the use of spectral
mapping methods (Minschwaner and Siskind, 1993). This method has
also been adopted for use in calculating H2 O photolysis in the Schumann-
Runge bands (Siskind et al., 1994).
In the spectral intervals close to the solar Lyman-α line, the solar
flux as well as the O2 absorption cross section vary relatively rapidly
with wavelength, and parameterizations similar to those used for the
Schumann-Runge intervals have been used (Chabrillat and Kockarts,
1997).
References
Allen, M., and J.E. Frederick, Effective photodissociation cross sections for molecular
oxygen and nitric oxide in the Schumann-Runge Bands. J Atmos Sci: 39, 2066, 1982.
Anderson, G.P., and L.A. Hall, Attenuation of solar irradiance in the stratosphere:
Spectrometer measurements between 191 and 207 nm. J Geophys Res: 88, 6801-6806,
1983.
Academic Press, 1987.
Apruzese, J.P., M.R. Schoeberl, and D.F. Strobel, Parameterization of IR cooling in

a middle atmosphere dynamics model, 1. Effects of the zonally averaged circulation.
J Geophys Res: 87,
7 8951, 1982.
Apruzese, J.P., D.F. Strobel, and M.R. Schoeberl, Parameterization of IR cooling
in the middle atmosphere dynamics model, 2. Non LTE radiative transfer and the
globally averaged temperature of the mesosphere and lower thermosphere. J Geophys
Res: 89, 4917, 1984.
Armstrong, B.H., Spectrum line profiles: The Voigt function. J Quant Spectrosc
Radiat Transfer: 7
7, 61, 1967.
Ball, S.M., G. Hancock, I.J. Murphy, and S.P. Rayner, The relative quantum yields
of O2 (A1 ∆G) from the photolysis of ozone in the wavelength range 270 nm ≤ λ ≤
329 nm. Geophys Res Lett: 20, 2063, 1993.
Ball, S.M., G. Hancock, and F. Winterbottom, Product channels in the near-UV
photodissociation of ozone. Faraday Discuss: 100, 215, 1995.
Bass, A.M., A.E. Ledford, and A.H. Laufer, Extinction coefficients of NO2 and N2 O4 .
J Res Natl Bur. Stand SecA: 80, 143, 1976.
Bass, A.M., L.C. Glasgow, C. Miller, J.P. Jesson, and D.L. Filkin, Temperature
dependent cross sections for formaldehyde [CH2 O]: The effect of formaldehyde on
stratospheric chlorine chemistry. Planet Space Sci: 28, 675, 1980.
Bates, D.R., Rayleigh scattering by air. Planet Space Sci: 32, 785, 1984.
Baum, W.A., F.S. Johnson, J.J. Obserly, C.C. Rockwood, C.V. Strain, and R. Tousey,
Solar ultraviolet spectrum to 88 kilometers. Phys Rev: 70, 781, 1946.
Benter, T., C. Feldmann, U. Kirchner, M. Schmidt, S. Schmidt, and R.N. Schindler,
UV/VIS absorption spectra of HOBr and CH3 OBr; Br(2P3/2) atom yields in the
photolysis of HOBr. Ber Bunsenges Phys Chem: 99, 1144, 1995.
Bhartia, P.K., K.F. Klenk, A.J. Fleig, C.G. Wellemeyer, and D. Gordon,
Intercomparison of Nimbus 7 Solar Backscattered Ultraviolet ozone profiles with
rocket, balloon, and umkehr profiles. J Geophys Res: 89, 5227, 1984.
Bhartia, P.K., R.D. McPeters, C.L. Mateer, L.E. Flynn, and C. Wellemeyer,
Algorithm for the estimation of vertical ozone profiles from the backscattered
ultraviolet technique. J Geophys Res: 101, 18,793, 1996.
Biaumé,
´ F., Nitric acid vapour absorption cross section spectrum and its
photodissociation in the stratosphere. J Photochem: 2, 139, 1973.
Bohren, C.F., and D.R. Huffman, Absorption and Scattering of Light by Small
Particles. J. Wiley and Sons, 1983.
Bongartz, A., J. Kames, F. Welter, and V. Schurath, Near UV absorption cross
sections and trans/cis equilibrium of nitrous oxide. J Phys Chem: 95, 1076, 1991.
Brasseur, G., and D. Offermann, Recombination of atomic oxygen near the
mesopause: Interpretation of rocket data. J Geophys Res: 91, 10,818, 1986.
Brasseur, G., A. De Rudder, and P.C. Simon, Implication for stratospheric
composition of a reduced absorption cross section in the Herzberg continuum of
molecular oxygen. Geophys Res Lett: 10, 20, 1983.
Brownword, R.A., M. Hillenkamp, T. Laurent, R.K. Vasta, H.-R. Volpp, and J.
Wolfrum, Quantum yield for H atom formation in the methane dissociation after
photoexcitation at the Lyman- α (121.6 nm) wavelength. Chem Phys Lett: 266, 259,
1997.
RADIATION 251
Brueckner, G.E., K.L. Edlow, L.E. Floyd, J.L. Lean, and M.E. van Hoosier, The
solar ultraviolet spectral irradiance monitor (SUSIM) experiment on board the Upper
Atmosphere Research Satellite. J Geophys Res: 98, 10,695, 1993.
Brusa, R.W., and C. Frohlich, Recent solar constant determinations from high
altitude balloons, Paper presented at the Symposium on the Solar Constant and
the Spectral Distribution of Solar Irradiance, IAMAP Third Scientific Assembly,
Published by the Radiation Commission, Boulder, Colo, USA, 1982.
Burkholder, J.B., Ultraviolet absorption spectrum of HOCl. J Geophy Res: 98, 2963,
1993.
Burkholder, J.B., J.J. Orlando, and C.J. Howard, Ultraviolet absorption cross section
of Cl2 O2 between 210 and 410 nm. J Phys Chem: 94, 687, 1990.
Burkholder, J.B., R.K. Talukdar, A.R. Ravishankara, and S. Solomon, Temperature
dependence of the HNO3 UV absorption cross section. J Geophys Res: 98, 22,937,
1993.
Burkholder, J.B., R.K. Talukdar, and A.R. Ravishankara, Temperature dependence
of the ClONO2 UV absorption spectrum. Geophys Res Lett: 21, 585, 1994.
Burkholder, J.B., A.R. Ravishankara, and S. Solomon, UV visible and IR absorption
cross sections of BrONO2 . J Geophys Res: 100, 16,793, 1995.
Callear, A.B., and M.J. Pilling, Fluorescence of nitric oxide, 6. Predissociation and
cascade quenching in NO D2 Σ+ (v = 0) and NO C2 Π (v = 0), and the oscillator
strengths of the σ (0,0) and (0,0) bands. Trans Faraday Soc: 66, 1886, 1970a.
Callear, A.B., and M.J. Pilling, Fluorescence of nitric oxide, 7. Quenching rates of
NO C2 π (v = 0), and its rate of radiation to NO A2 Σ+ , energy transfer efficiencies
and mechanisms of predissociation. Trans Faraday Soc: 66, 1618, 1970b.
Cantrell, C.A., J.A. Davison, A.H. McDaniel, R.E. Shetter, and J.G. Calvert,
Temperature dependent formeldehyde cross sections in the near ultraviolet spectral
region. J Phys Chem: 94, 3902, 1990.
Cantrell, C., A. Zimmer, and G.S. Tyndall, Absorption cross sections for water vapor
from 183 to 193 nm. Geophys Res Lett: 24, 2195, 1997.
Cess, R.D., and V. Ramanathan, Radiative transfer in the atmosphere of Mars and
that of Venus above the cloud deck. J Quant Spectrosc Radiat Transfer: 12, 933,
1972.
Chabrillat, S., and G. Kockarts, Simple parameterization of the absorption of the
solar Lyman-α line. Geophys Res Lett: 24, 2659, 1997. Correction: Geophys Res
Lett: 25, 79, 1998.
Chandrasekhar, S., Radiative Transfer. Oxford University Press, 1950 (Reprinted by
Dover Publ., 1960).
Chapman, S., The absorption and dissociative or ionizing effect of monochromatic
radiations in an atmosphere on a rotating earth. Proc Phys Soc: 43, 483, 1931.
Cheung, A.S.-C., K. Yoshino, W.H. Parkinson, S.L. Guberman, and D.E. Freeman,
Absorption cross section measurements of oxygen in the wavelength region 195-241 nm
of the Herzberg continuum. Planet Space Sci: 34, 1007, 1986.
Chou, C.C., W.S. Smith, H. Vera Ruiz, K. Moe, G. Crescentini, J.J. Molinar, and F.S.
Rowland, The temperature dependence of the ultraviolet absorption cross sections of
CCl2 F2 and CCl3 F, and their stratospheric significance. J Phys Chem: 81, 1977.
Chou, C.C., R.J. Milstein, W.S. Smith, H. Vera Ruiz, M.J. Molinar, and F.S. Row-
land, Stratospheric photodissociation of several saturated perhalo chlorofluorocarbon
compounds in current technological use (Fluorocarbons –13, –113, –114, and –115).
J Phys Chem: 82, 1, 1978.
Cieslik, S., Détermination
´ exp´
érimentale des forces d’oscillateur des bandes β, γ, δ,
et ξ de la molecule
´ NO. Bull Cl Sci Acad Roy Belg: 63, 884, 1977.
Cieslik, S., and M. Nicolet, The aeronomic dissociation of nitric oxide. Planet Space
Sci: 21, 925, 1973.
Clough, S.A., and M.J. Iacono, Line by line calculation of the atmospheric fluxes and
cooling rates: Application to carbon dioxide, ozone, methane, nitrous oxide, and the
halocarbons. J Geophys Res: 100, 16,519, 1995.
Cox, R.A., and K. Patrick, Kinetics of the reaction HO2 + NO2 (+ M) → HO2 NO2
using molecular modulation spectrometry. Int J Chem Kinetics: 11, 635, 1979.
Cox, R.A., D.W. Sheppard, and M.P. Stevens, Absorption coefficients and kinetics of
the BrO radical using molecular modulation. J Photochem: 19, 189, 1982.
Cox, R.A., and G.D. Hayman, The stability and photochemistry of dimers of the ClO
radical and implications for the Antarctic ozone depletion. Nature: 332, 796, 1988.
Crutzen, P., Comment on paper “Absorption and emission by carbon dioxide in the
mesosphere”, by J.T. Houghton. Quart J Roy Meteorol Soc: 96, 767, 1970.
Crutzen, P.J., Energy conversions and mean vertical motions in the high latitude
summer mesosphere and lower thermosphere, in Mesospheric Models and Related
Experiments. G. Fiocco (ed.), D. Reidel Publishing Co., Dordrecht, Holland, 1971.
Curtis, A.R., Discussion of a statistical model for water vapour absorption. Quart J
Roy Meteorol Soc: 78, 638, 1952.
Curtis, A.R., and R.M. Goody, Thermal radiation in the upper atmosphere. Proc
Roy Soc: A236, 193, 1956.
Daumont, D., J. Brion, J. Charbonnier, and J. Malicet, Ozone UV spectroscopy, 1.
Absorption cross sections at room temperatures. J Atmos Chem: 15, 145, 1992.
Davidson, J.A., C.A. Cantrell, A.H. McDaniel, R.E. Shetter, S. Madronich, and
J.G. Calvert, Visible-ultraviolet absorption cross sections for NO2 as a function of
temperature. J Geophys Res: 93, 7105, 1988.
DeMore, W.B., and M. Patapoff, Temperature and pressure dependence of CO2
extinction coefficients. J Geophys Res: 77,
7 6291, 1972.
DeMore, W.B., and E. Tschuikow-Roux, Ultraviolet spectrum and chemical reactivity
of the ClO dimer. J Phys Chem: 94, 5856, 1990.
Deshler, T., B.J. Johnson, and W.R. Rozier, Baloon-borne measurements of Pinatubo
aerosol during 1991 and 1992 at 41◦ N, vertical profile size distribution and volatility.
Deters, B., J.P. Burrows, S. Himmelmann, and C. Blindauer, Gas phase spectra of
HOBr and Br2 O and their atmospheric significance. Ann Geophys: 14, 468, 1996.
Dickinson, R.E., Method of parameterization for infrared cooling between the
altitudes of 30 and 70 km. J Geophys Res: 78, 4451, 1973.
Dickinson, R.E., Infrared radiative cooling in the mesosphere and lower thermosphere.
J Atmos Terr Phys: 46, 995, 1984.
Dopplick, T.G., Radiative heating of the global atmosphere. J Atmos Sci: 29, 1278,
1972.
Edwards, D.P., GENLN2: A general line-by-line atmospheric transmittance and
radiance model. Version 3.0 description and users guide, NCAR Technical Note,
RADIATION 253
NCAR/TN-367+STR, National Center for Atmospheric Research, Boulder, Colo.,

USA, 1992.
Edwards, D.P., M. Lopez-Puertas,
´ and M.A. L´ opez-Valverde,
´ Non-local thermody-
namic equilibrium studies of the 15-µm bands of CO2 for atmospheric remote sensing.
J Geophys Res: 98, 14,955, 1993.
Ellingson, R.G., J. Ellis, and S. Fels, The intercomparison of radiation codes used in
climate models: Long wave results. J Geophys Res: 96, 8929, 1991.
Elsasser, W.M., Heat transfer by infrared radiation in the atmosphere. Harvard
Meteorol. Stud: 6, Harvard Univ. Press, Cambridge, Mass., USA, 1942.
Elterman, L., UV, Visible and IR Attenuation for Altitudes to 50 km. AFCRL Report
68-0153, Environ. Res. Papers, Bedford, Mass, USA, 1968.
Fang, T.M., S.C. Wofsy, and A. Dalgarno, Capacity distribution functions and
absorption in Schumann-Runge bands of molecular oxygen. Planet Space Sci: 22,
413, 1974.
Fomichev, V.I., W.E. Ward, S.R. Beagley, C. McLandress, J.C. McConnell, N.A.
McFarlane, and T.G. Shepherd, Extended Canadian Middle Atmosphere Model:
Zonal-mean climatology and physical parameterizations. J Geophys Res: 107,
7 DIO,
doi:10.1029/2001JD000479, 2002.
Frederick, J.E., and R. D. Hudson, Predissociation of nitric oxide in the mesosphere
and stratosphere. J Atmos Sci: 36, 737, 1979.
Frederick, J.E., and J.E. Mentall, Solar irradiance in the stratosphere: Implication
for the Herzberg continuum absorption of O2 . Geophys Res Lett: 9, 461, 1982.
Frederick, J.E., R.B. Abrams, and P.J. Crutzen, A potential mechanism for coupling
thermospheric variations to the mesosphere and the stratosphere. J Geophys Res: 88,
3829, 1983.
Froidevaux, L., and Y.L. Yung, Radiation and chemistry in the stratosphere:
Sensitivity to O2 cross sections in the Herzberg continuum. Geophys Res Lett: 9,
854, 1982.
Fu, Q., and K.N. Liou, On the correlated k-distribution method for radiative transfer
in nonhomogeneous atmospheres. J Atmos Sci: 49, 2139, 1992.
Garcia, R.R., and S. Solomon, A numerical model for the zonally averaged dynamical
Ghazi, A.V., V. Ramanathan, and R.E. Dickinson, Acceleration of upper
stratospheric radiative damping: Observational evidence. Geophys Res Lett: 6, 437,
1979.
Gijs, A., A. Koppers, and D.P. Murtagh, Model studies of the influence of O2
photodissociation parameterizations in the Schumann-Runge bands on ozone related
photolysis in the upper atmosphere. Ann Geophys: 14, 68, 1997.
Gillotay, D., and P.C. Simon, Ultraviolet absorption cross sections of methyl bromide
at stratospheric temperatures. Ann Geophys: 6, 211, 1988.
Gillotay, D., A. Jenouvrier, B. Coquart, M.F. Merienne, and P.C. Simon, Ultraviolet
absorption cross sections of bromoform in the temperature range 295-210 K. Planet
Space Sci: 37,
7 1127, 1989.
Godson, W.L., The evaluation of infrared radiative fluxes due to atmosphere water
vapour. Quart J Roy Meteorol Soc: 79, 367, 1953.
Goodeve, C.F., and A.C.W. Taylor, The continuous absorption spectrum of hydrogen
bromide. Proc Roy Soc: A152, 221, 1935.
Goody, R.M., Atmospheric Radiation, I. Theoretical Basis. Oxford at the Clarendon

Press, 1964.
Goody, R.M., and Y.L. Yung, Atmospheric Radiation: Theoretical Basis. Oxford
University Press, 1989.
Goody, R.M., R. West, L. Chen, and D. Crisp, The correlated k-method for radiation
calculations in nonhomogeneous atmospheres. J Quant Spectrosc Radiat Transfer:
42, 539, 1989.
Gordley, L.L., B.T. Marshall, and D.A. Chu, LINEPAK: Algorithms for modeling
spectral transmittance and radiance. J Quant Spectrosc Radiat Transfer: 52, 563,
1994.
Graham, R.A., Photochemistry of NO3 and the kinetics of the N2 O5 – O3 system,
PhD Thesis, University of California, Berkeley, Calf, USA, 1975.
Graham, R.A., and H.S. Johnston, The photochemistry of NO3 and the kinetics of
the N2 O5 – O3 system. J Phys Chem: 82, 254, 1978.
Graham, R.A., A.M. Wier, and J.A. Pitts, Ultraviolet and infrared cross section of
gas phase HO2 NO2 . Geophys Res Lett: 5, 909, 1978.
Hanel, R.A., and B.J. Courath, Thermal Emission Spectra of the Earth and
Atmosphere Obtained from the Nimbus 4 Michelson Interferometer Experiment.
NASA Report X-620-70-244, 1970.
Harwood, M.H., R.L. Jones, R.A. Cox, E. Lutman, and O.V. Rattigan, Temperature-
dependent absorption cross sections of N2 O5 . J Photochem Photobiol A: Chem: 73,
167, 1993.
Harries, J., The greenhouse earth: A view from space. Quart J Roy Meteorol Soc:
122, 799, 1996.
Henri, V., and S.A. Schou, Struktur und Akitivierung der Molekel des Formaldehyds,
eine Analyse auf Grund des ultrvioletten Absorption-Spektrums des Dampfes. Zeit
Phys: 49, 774, 1928.
Herman, J.R., and J.E. Mentall, The direct and scattered solar flux within the
stratosphere. J Geophys Res: 87,
7 1319, 1982a.
Herman, J.R., and J.E. Mentall, O2 absorption cross section (187-225 nm) from
stratospheric solar flux measurements. J Geophys Res: 87,
7 8967, 1982b.
Herzberg, G., Ultraviolet absorption spectra of acetylene and formaldehyde. Trans
Faraday Soc: 27,
7 378, 1931.
Holt, R.B., and O. Oldenberg, Role of hydrogen peroxide in the thermal combination
of hydrogen and oxygen. J Chem Phys: 17, 7 1091, 1949.
Holt, R.B., C.K. McLane, and O. Oldenberg, Ultraviolet absorption spectrum of
hydrogen peroxide. J Chem Phys: 16, 225, 1948. Erratum: J Chem Phys: 16, 638,
1948.
Houghton, J.T., Absorption and emission by carbon dioxide in the mesosphere. Quart
J Roy Meteorol Soc: 95, 1, 1969.
Hubinger, S., and J.B. Nee, Photoabsorption spectrum for OClO between 125 and
470 nm. Chem Phys: 181, 247, 1994.
Hubrich, C., and F. Stuhl, The ultraviolet absorption of some halogenated methanes
and ethanes of atmospheric interest. J Photochem: 12, 93, 1980.
Hubrich, C., C. Zetsch, and F. Stuhl, Absorptionsspektren von halogenierten
Methanen im Bereich von 275 bis 160 nm bei Temperaturen von 298 und 208 K.
Ber Bunsenges Phys Chem: 81, 437, 1977.
RADIATION 255
Huder, K.J., and W.B. DeMore, Absoprtion cross sections of the ClO dimer. J Phys
Chem: 99, 3905, 1995.
Hudson, R.D., and S.H. Mahle, Photodissociation rates of molecular oxygen in the
mesosphere and lower thermosphere. J Geophys Res: 77,
7 2902, 1972.
Hudson, R.D., V.L. Carter, and J.A. Stein, An investigation of the effect of
temperature on the Schumann-Runge absorption continuum of oxygen, 1580-910 A.
J Geophys Res: 71, 2295, 1966.
Humlicek, J., Optimized computation of the Voigt and complex probability functions.
J Quart Spectrosc Radiat Transfer: 27,
7 437, 1982.
Inn, E.C.Y., Absorption coefficient of HCl in the region 1400 to 2200 A. J Atmos Sci:
32, 2375, 1975.
Inn, E.C.Y., and Y. Tanaka, Absorption coefficient of ozone in the ultraviolet and
visible regions. J Opt Soc Amer: 43, 8760, 1953.
Inn, E.C.Y., K. Watanabe, and M. Zelikoff, Absorption coefficients of gases in the
vacuum ultraviolet: 3. CO2 . J Chem Phys: 21, 1648, 1953.
Iribarne, J.V., and H.R. Cho, Atmospheric Physics, D. Reidel, Dordrecht, The
Netherlands, 1980.
Jenouvrier, A., B. Coquart, and M.F. Merienne-Lafore, New measurements of the
absorption cross sections in the Herzberg continuum of molecular oxygen in the region
between 205 and 240 nm. Planet Space Sci: 34, 253, 1986.
Jet Propulsion Laboratory (JPL), Chemical Kinetics and Photochemical Data for Use
in Stratospheric Modeling, W.B. DeMora, S.P. Sander, D.M. Golden, R.F. Hampson,
M.J. Kurylo, C.J. Howard, A.R. Ravishankara, C.E. Kolb, and M.J. Molina, eds.,
JPL Publication 97-4, 1997.
Johnson, F.S., J.D. Porcell, R. Tousey, and K. Watanabe, Direct measurements of
the vertical distribution of atmospheric ozone to 70 km altitude. J Geophys Res: 57,
7
157, 1952.
Johnston, H.S., and R.A. Graham, Gas-phase ultraviolet spectrum of nitric acid
vapor. J Chem Phys: 77,
7 62, 1973.
Johnston, H.S., and R.A. Graham, Photochemistry of NOx compounds. Canad J
Chem: 52, 1415, 1974.
Johnston, H.S., and G. Selwyn, New cross sections for the absorption of near
ultraviolet radiation by nitrous oxide (N2 O). Geophys Res Lett: 2, 549, 1975.
Johnston H.S., S.G. Chang, and G. Whitten, Photolysis of nitric acid vapor. J Phys
Chem: 78, 1, 1974.
Johnston, H.S., H.F. Davies, and Y.T. Lee, NO3 photolysis product channels:
Quantum yields from observed energy thresholds. J Phys Chem: 100, 4713, 1996.
Jones, E.J., and O.R. Wulf, The absorption coefficient of nitrogen pentoxide in the
ultraviolet and the visible absorption spectrum NO3 . J Chem Phys: 5, 873, 1937.
Joseph, J.H., W.J. Wiscombe, and J.A. Weinman, The delta-Eddington approxima-
tion for radiative flux transfer. J Atmos Sci: 33, 2452, 1976.
Kiehl, J.T., and V. Ramanathan, CO2 radiative parameterization used in climate
models: Comparison with narrow band moels and with laboratory data. J Geophys
Res: 88, 5191, 1983.
Kiehl, J.T., and R.E. Dickinson, A study of the radiative effects of enhanced
atmospheric CO2 and CH4 on early earth surface temperatures. J Geophys Res:
92, 2991, 1987.
Knauth, H.-D., H. Alberti, and H. Clausen, Equilibrium constant of the gas reaction
Cl2 + H2 O ultraviolet spectrum of HOCl. J Phys Chem: 83, 1604, 1979.
Kockarts, G., Absorption and photodissociation in the Schumann-Runge bands of
molecular oxygen in the terrestrial atmosphere. Planet Space Sci: 24, 589, 1976.
Kockarts, G., Nitric oxide cooling in the terrestrial thermosphere. Geophys Res Lett:
7, 137, 1980.
7
Kockarts, G., Penetration of solar radiation in the Schumann-Runge bands of
molecular oxygen, A robust approximation. Ann Geophys, 1994.
Kondratyev, K.Y., Radiation in the Atmosphere. Academic Press, 1969.
Kourganoff, V., Basic Methods in Transfer Problems. Oxford University Press, 1952.
Kuhn, W.R., and J. London, Infrared radiative cooling in the middle atmosphere
(30-110 km). J Atmos Sci: 26, 189, 1969.
Kylling, A., K. Stamnes, R.R. Meier, and D.E. Anderson, The 200-300 nm radiation
field within the stratosphere: Comparison of models with observation. J Geophys
Res: 98, 2741, 1993.
Lacis, A.A., and J.E. Hansen, A parameterization for the absorption of solar radiation
in the earth’s atmosphere. J Atmos Sci: 31, 118, 1974.
Lacis, A.A., and V. Oinas, A description of the correlated k distribution method for
modeling nongray gaseous absorption, thermal emission, and multiple scattering in
vertically inhomogeneous atmospheres. J Geophys Res: 96, 9027, 1991.
Langhoff, J.R., J.P. Dix, J.O. Arnold, R.W. Nicholls, and L.L. Danylewych,
Theoretical intensity parameters for the vibration-rotation bands of ClO. J Chem
Phys: 67,
7 4306, 1977a.
Langhoff, S.R., R.L. Jaffe, and J.O. Arnold, Effective cross sections and rate constants
for predissociation of ClO in the earth’s atmosphere. J Quant Spectrosc Radiat
Transfer: 18, 227, 1977b.
Laufer, A.M., and J.R. McNesby, Deuterium isotope effect in vacuum ultraviolet
absorption coefficients of water and methane. Canad J Chem: 43, 3487, 1965.
Lean, J., The sun’s variable radiation and its relevance for earth. Annu Rev Astron
Astrophys: 35, 33, 1997.
Lean, J.L. and A.J. Blake, The effect of temperature on thermospheric molecular
oxygen absorption in the Schumann-Runge continuum. J Geophys Res: 86, 211,
1981.
Lee, R.B., M.A. Gibson, R.S. Wilson, and S. Thomas, Long-term solar irradiance
variability during sunspot cycle 22. J Geophys Res: 100, 1667, 1995.
Leifson, S.W., Absorption spectra of some gases and vapors in the Schumann region.
Astrophys J: 63, 73, 1926.
Lenoble, J., Standard procedures to compute atmospheric radiative transfer in a
scattering atmosphere, I.A.M.A.P., National Center for Atmospheric Research,
Boulder, Colo., USA, 1977.
Lewis, B.R., and J.H. Carter, Temperature dependence of the carbon dioxide
photoabsorption cross section between 1200 and 1700 Å. J Quant Spectros Radiat
Transfer: 30, 297, 1983.
Lewis, B.R., L. Berzins, and J.H. Carter, Oscillator strengths for the Schumann-
Runge bands of O2 . J Quant Spectrosc Radiat Transfer: 36, 209, 1986.
RADIATION 257
Lewis, B.R., S.T. Gibson, and P.M. Dooley, Fine structure dependence of
predissociation linewidth in the Schumann-Runge bands of molecular oxygen. J Chem
Phys: 100, 6993, 1994.
Lin, C.-L., and W.B. DeMore, O(1 D) production in ozone photolysis near 3100 A. J
Photochem: 2, 161, 1973.
Lin, C.L., N.K. Rohatgi, and W.B. Demore, Ultraviolet absorption cross sections of
hydrogen peroxide. Geophys Res Lett: 5, 113, 1978.
Liou, K.-N., An Introduction to Atmospheric Radiation, Academic Press, 1980.
Liou, K.-N., An Introduction to Atmospheric Radiation, Second Edition, Academic
Press, 2002.
London, J., Radiative energy sources and sinks in the stratosphere and mesophere,
Proc of the NATO Advanced Institute on Atmospheric Ozone, A.C. Aikin, ed., U.S.
Dept. of Transportation, FAA-EE-80-20, FAA, Washington, D.C., USA, 1980.
López-Puertas,
´ M., and F.W. Taylor, Non-LTE Radiative Transfer in the Atmosphere,
World Scientific, 2001.
Lopez-Puertas,
´ M., R. Rodrigo, A. Molina, and F.W. Taylor, A non-LTE radiative
transfer model for infrared bands in the middle atmosphere, I. Theoretical basis and
application to CO2 15-µm bands. J Atmos Terr Phys: 48, 729, 1986.
Lopez-Puertas,
´ M., M.A. López-Valverde,
´ C.P. Rinsland, and M.R. Gunson,
Analysis of the upper atmosphere CO2 (ν2 ) vibrational temperatures retrieved from
ATMOS/Spacelab 3 observations. J Geophys Res: 97, 7 20,469, 1992a.
Lopez-Puertas,
´ M., M.A. López-Valverde,
´ and F.W. Taylor, Vibrational temperatures
and radiative cooling of the CO2 15µm bands in the middle atmosphere. Quart J
Roy Meteorol Soc: 118, 499, 1992b.
Luther, F.M., D.J. Wuebbels, W.H. Duewer, and J.C. Chang, Effect of multiple
scattering on species concentrations and model sensitivity. J Geophys Res: 83, 3563,
1978.
Magnotta, F., and H.S. Johnston, Photodissociation quantum yields for the NO3 free
radical. Geophys Res Lett: 7,
7 769, 1980.
Malkmus, W., Random Lorentz band model with exponential-tailed S-1 line intensity
distribution function. J Opt Soc Amer: 57,
7 323, 1967.
Malicet, J., D. Daumont, J. Charbonnier, C. Parisse, A. Chakir, and J. Brion, Ozone
UV spectroscopy, 2. Absorption cross sections and temperature dependence. J Atmos
Chem: 21, 263, 1995.
Manabe, S., and F. Moller, On the radiative equilibrium and heat balance of the
atmosphere. Mon Wea Rev: 89, 503, 1961.
Manabe, S., and R.F. Strickler, Thermal equilibrium of the atmosphere with a
convective adjustment. J Atmos Sci: 21, 361, 1964.
Maric, D., J.P. Burrows, and G.K. Moortgat, A study of the UV-visible absorption
spectra of Br2 and BrCl. J. Photochem Photobiol A: Chem: 83, 179, 1994.
Marmo, F.F., Absorption coefficients of nitrogen oxide in the vacuum ultraviolet. J
Opt Soc Amer: 43, 1186, 1953.
Marshall, B.T., L.L. Gordley, and D.A. Chu, BANDPAK: Algorithms for modeling
broadband transmission and radiance. J Quant Spectrosc Radiat Transfer: 52, 581,
1994.
Mauldin III, R.L., J.B. Burkholder, and A.R. Ravishankara, A photochemical,
thermodynamic and kinetic study of ClOO. J. Phys Chem: 96, 2582, 1992.
McCartney, E.J., Optics of the Atmosphere: Scattering by Molecules and Particles,

Wiley and Sons, 1976.
McCormick, M.P., P. Hamill, T.J. Pepin, W.P. Chu, T.J. Swissler, and L.R.
McMaster, Satellite studies of the stratospheric aerosol. Bull Am Meteoroleor Soc:
60, 1038, 1979.
McLeod, H., G. P. Smith, and D. M. Golden, Photodissociation of pernitric acid at
248 nm. J Geophys Res: 93 3813, 1988.
Meador, W.E., and W.R. Weaver, Two-stream approximations to radiative transfer
in planetary atmospheres: A unified description of existing methods and a new
improvement. J Atmos Sci: 37,
7 630, 1980.
Meier, R.R., D.E. Anderson, Jr., and M. Nicolet, Radiation field in the troposphere
and stratosphere from 240 to 1000 nm, I. General analysis. Planet Space Sci: 30,
923, 1982.
Merienne,
´ M.F., B. Coquart, and A. Jenouvrier, Temperature effect of the ultraviolet
absorption of CFCl3 , CF2 Cl2 and N2O. Planet Space Sci: 38, 617, 1990.
Mertens, C.J., M.G. Mlynczak, R.R. Garcia, and R.W. Portmann, A detailed
evaluation of the stratospheric heat budget, 1. Radiation transfer. J Geophys Res:
104, 6021, 1999.
Michelsen, H.H., R.J. Salawitch, P.O. Wennberg, and J.G. Anderson, Production of
O(1 D) from photolysis of O3 . Geophys Res Lett: 21, 2227, 1994.
Mie, G., Beitrage zur optik trueber Medien, Speziell koloidaller metaloesungen Ann
der Phys: 25, 377, 1908.
Milne, E.A., Handbuch der Astrophysik, 3, Part I. 1930 (Reprinted in “Selected Papers
on the Transfer of Radiation”, Dover, 1966).
Minschwaner, K., and D.E. Siskind, A new calculation of nitric oxide photolysis in
the stratosphere, mesosphere, and lower thermosphere. J Geophys Res: 98, 20,401,
1993.
Minschwaner, K., G.P. Anderson, L.A. Hall, and K. Yoshino, Polynomial coefficients
for calculating O2 Schumann-Runge cross sections at 0.5 cm−1 resolution. J Geophys
Res: 97,
7 10,103, 1992.
Minschwaner, K., R.J. Salawitch, and M.B. McElroy, Absorption of solar radiation
by O2 : Implications for O3 and lifetimes of N2 O, CFCl3 , and CF2 Cl2 . J Geophys
Res: 98, 10,543, 1993.
Minschwaner, K., R.J. Thomas, and D.W. Rusch, Scattered ultraviolet radiation in
the upper stratosphere, I. Observations. J Geophys Res: 100, 11,157, 1995.
Mishalanie, E.A., J.C. Rutkowski, R.S. Hutte, and J.W. Birks, Ultraviolet absorption
spectrum of gaseous hypochlorous acid. J Phys Chem: 90, 5578, 1986.
Mitchell, A.C.G., and W.M. Zemansky, Resonance Radiation and Excited Atoms.
Harvard Univ. Press, Cambridge, Mass., USA, 1934 (Reprinted 1961).
Mlynczak, M.G., and B.T. Marshall, A reexamination of the role of solar heating
in the O2 atmospheric and infrared atmospheric bands. Geophys Res Lett: 23, 657,
1996.
Mlynczak, M.G., and S. Solomon, On the efficiency of solar heating in the middle
atmosphere. Geophys Res Lett: 18, 1201, 1991.
Mlynczak, M.G., and S. Solomon, A detailed evaluation of the heating efficiency in
the middle atmosphere. J Geophys Res: 98, 10,517, 1993.
RADIATION 259
Mlynczak, M.G., C.J. Mertens, R.R. Garcia, and R.W. Portman, A detailed
evaluation of the stratospheric heat budget, 2. Global radiation balance and diabatic
circulations. J Geophys Res: 104, 6039, 1999.
Molina, M.J., and F.S. Rowland, Stratospheric sink for chlorofluoromethanes:
Chlorine atom-catalyzed destruction of ozone. Nature: 249, 810, 1974.
Molina, L.T., and M.J. Molina, Ultraviolet spectrum of HOCl. J Phys Chem: 42,
2410, 1978.
Molina, L.T., and M.J. Molina, Chlorine nitrate ultraviolet absorption spectrum at
stratospheric temperatures. J Photochem: 11, 139, 1979.
Molina, L.T., and M.J. Molina, UV absorption cross sections of HO2 NO2 vapor. J
Photochem: 15, 97, 1981.
Molina, L.T., and M.J. Molina, Absolute absorption cross sections of ozone in the
185 to 350 nm wavelength range. J Geophys Res: 91, 14,501, 1986.
Molina, L.T., M.J. Molina, and F.S. Rowland, Ultraviolet absorption cross sections
of several brominated methanes and ethanes of atmospheric interest. J Phys Chem:
86, 2672, 1982.
Molina, M.J., A.J. Colussi, L.T. Molina, R.N. Schindler, and T.L. Tso, Quantum yield
of chlorine atom formation in the photodissociation of chlorine peroxide (ClOOCl) at
308 nm. Chem Phys Lett: 173, 310, 1990.
Moortgat, G.K., E. Kudszus, and P. Warneck, Temperature dependence of O(1 D)
formation in the near UV photolysis of ozone. J Chem Soc, Faraday Trans: 11, 73,
1216, 1977.
Moortgat, G.K., W. Klippel, K.H. Mobius, W. Seiler, and P. Warneck, Laboratory
measurements of photolytic parameters for formaldehyde. Rep. FAA-EE-80-47,
Federal Aviation Administration, Washington, D.C., USA, 1980.
Moortgat, G.K., W. Seiler, and P. Warneck, Photodissociation of HCHO in air: CO
and H2 2 quantum yield at 220 and 300 K. J Chem Phys: 78, 1185, 1983.
Moortgat, G.K., R. Meller, and W. Schneider, Temperature dependance (256-296 K)
of the absorption cross sections of bromoform in the wavelength range 285-360 nm,
in Proceedings of NATO workshop on ”The Tropospheric Chemistry in the Polar
Regions”, Wolfville, Canada, 23-28 August 1992, NATO ASI Series. H. Niki and
K.H. Becker (eds.), Springer Verlag, Berlin, 359-369, 1993.
Murray, J.E., K. Yoshino, J. R. Esmond, W. H. Parkinson, Y. Sun, and A. Dalgarno,
Vacuum ultraviolet Fourier transform spectroscopy of the δ(0,0) and β(7,0) bands of
NO. J Chem Phys: 101, 62, 1994.
Murtagh, D.P., The O2 Schumann-Runge system: New calculations of photodissoci-
ation cross sections. Planet Space Sci: 36, 819, 1988.
Nesme-Ribes, E., S.L. Baliunes, and D. Sokoloff, The stellar dynamo. Sci Amer: 51,
1996.
Newchurch, M.J., M. Allen, M.R. Gunson, R.J. Salawitch, G.B. Collins, K.H. Huton,
M.M. Abbas, M.C. Abrams, A.Y. Chang, D.W. Fahey, R.S. Gao, F.W. Irion, M.
Loewenstein, G.L. Manney, H.A. Michelsen, J.R. Podolske, C.P. Rinsland, and R.
Zander, Stratospheric NO and NO2 abundances from ATMOS solar-occultation
measurements. Geophys Res Lett: 23, 2373, 1996.
Nickolaisen, S.L., and S.P. Sander, Pressure dependent yields and product branching
ratios in the broadband photolysis of chlorine nitrate. J Phys Chem: 100, 10,165,
1996.
Nicolet, M., Etude des réactions

´ chimiques de l’ozone dans la stratosphere. Royal
Meteorological Institute of Belgium, 1978.
Nicolet, M., The chemical equations of stratospheric and mesopheric ozone, Proc of
the NATO Advanced Institute on Atmospheric Ozone, A.C. Aikin, ed., U.S. Dept. of
Transportation, FAA-EE-80-20, FAA, Washington, D.C., USA, 1980.
Nicolet, M., On the molecular scattering in the terrestrial atmosphere: An empirical
formula for its calculation in the homosphere. Planet Space Sci: 32, 1467, 1984.
Nicolet, M., and S. Cieslik, The photodissociation of nitric oxide in the mesophere
and stratosphere. Planet Space Sci: 28, 105, 1980.
Nicolet, M., and R. Kennes, Aeronomic problems of molecular oxygen photodissocia-
tion, IV. Photodissociation frequency and transmittance in the spectral range of the
Schumann-Runge bands. Planet Space Sci: 37, 7 459, 1989.
Nicolet, M., R.R. Meier, and D.E. Anderson, Radiation field in the troposphere and
stratosphere, II. Numerical analysis. Planet Space Sci: 30, 935, 1982.
Nicolet, M., S. Cieslik, and R. Kennes, Aeronomic problems of molecular oxygen
photodissociation, V. Predissociation in the Schumann-Runge bands of oxygen.
Planet Space Sci: 37,
7 427, 1989.
Nicovich, J.M., and P.H. Wine, Temperature-dependent absorption cross section of
the hydrogen peroxide vapor, J Geophys Res: 93, 2417, 1988.
Norrish, R.G.W., and F.N. Kirkbride, Primary photochemical processes, I. The
decomposition of formaldehyde, J Chem Soc: 1, 1518, 1932.
Ogawa, M., Absorption cross sections of O2 and CO2 continua in the Schumann-
Runge and far-UV regions, J Chem Phys Lett: 9, 603, 1971.
Orlando, J.J., and J.B. Burkholder, Gas phase UV visible absorption spectra of HOBr
and Br2 O, J Phys Chem: 99, 1143, 1995.
Orlando, J.J., G.S. Tyndall, G.K. Moortgat, and J.G. Calvert, Quantum yields for
NO3 photolysis between 570 and 635 nm, J Phys Chem: 97,7 10,996, 1993.
Park, J.H., The equivalent mean absorption cross sections for the O2 Schumann-
Runge bands: Application to the H2 O and NO photodissociation rates. J Atmos Sci:
312, 1893, 1974.
Penndorf, R., Tables of the refractive index for standard air and the Rayleigh
scattering coefficient for the spectral region between 0.2 and 20.0 µm and their
application to atmospheric optics. J Opt Soc Amer: 47,
7 176, 1957.
Penner, S.S., Quantitative Molecular Spectroscopy and Gas Emissivities. Addison-
Wesley, Reading, Mass, USA, 1959.
Permien, T., R. Vogt, and R.N. Schindler, Mechanisms of gas phase-liquid phase
chemical transformations. Air Pollution Report 17, R.A. Cox, ed., Environmental
Research Program of the CEC, Brussels, 1988.
Perner, D., and U. Platt, Absorption of light in the atmosphere by collision pairs of
oxygen (O2 )2 . Geophys Res Lett: 7,
7 1053, 1980.
Petropavlóvskikh,
´ I., Evaluation of Photodissociation Coefficient Calculations for
Use in Atmospheric Chemical Models. Univ. of Brussels and National Center for
Atmospheric Research Cooperative Ph.D. Thesis, NCAR/CT-159, 1995.
Ramanathan, V., Radiative transfer within the earth’s troposphere and stratosphere:
A simplified radiative-convective model. J Atmos Sci: 33, 1330, 1976.
Rattigan, O.V., R.L. Jones, and R.A. Cox, The visible spectrum of gaseous OBrO.
Chem Phys Lett: 230, 121, 1994.
RADIATION 261
Rattigan, O.V., D.J. Lary, R.L. Jones, and R.A. Cox, UV-visible absorption cross
sections of gaseous Br2 O and HOBr. J Geophys Res: 101, 23,021, 1996.
Rebbert, R.E., R.L. Lilly, and P. Ausloos, Abstract of papers, 164th National Meeting,
Association Chemical Society, New York, August 1972.
Richards, P.G., D.G. Torr, and M.A. Torr, Photodissociation of N2 : A significant
source for thermospheric atomic nitrogen. J Geophys Res: 86, 1495, 1981.
Richards, P.G., J.A. Fennelly, and D.G. Torr, EUVAC: A solar EUV flux model for
aeronomic calculations. J Geophys Res: 99, 8981, 1994.
Robbins, D.E., Photodissociation of methyl chloride and methyl bromide in the
atmosphere. Geophys Res Lett: 3, 213, 1976. Erratum: Ibid, 757, 1976.
Rodgers, C.D., and C.D. Walshaw, The computation of infra-red cooling rate in
planetary atmospheres. Quart J Roy Meteorol Soc: 92, 67, 1966.
Romand, J., Absorption ultraviolette dans la région
´ of Schumann-Runge, Etude de
ClH, BrH et IH gazeux. Ann Phys, Paris: 4, 527, 1949.
Romand, J., and B. Vodar, Spectre d’absorption de l’acide chlorhydrique gazeux dans
´
la region de Schumann. Canad Roy Acad Sci Paris: 226, 238, 1948.
Rothman, L.S., C.P. Rinsland, A. Goldman, S.T. Massie, D.P. Edwards, J.-M. Flaud,
A. Perrin, C. Camy-Peyret, V. Dana, J.-Y. Mandin, J. Schroeder, A. McCann, R.R.
Gamache, R.B. Wattson, K. Yoshino, K.V. Chance, K.W. Jucks, L.R. Brown, V.
Nemtschinov, and P. Varanasi, The HITRAN molecular spectroscopic data base and
HAWKS: 1996 Edition. J Quant Spectrosc Radiat Transfer: 60, 665, 1998.
Rottman, G.J., T.N. Woods, and T.P. Sparn, Solar stellar irradiance comparison
experiment: Instrument design and operation. J Geophys Res: 98, 10,667, 1993.
Rowland, F.S., and M.J. Molina, Chlorofluoromethanes in the environment. Rev
Geophys Space Phys: 13, 1, 1975.
Schiffman, A., D.D. Nelson, Jr., and D.J. Nesbitt, Quantum yields for OH production
from 193 and 248 nm photolysis of HNO3 and H2 O2 . J Chem Phys: 98, 6935, 1993.
Schneider, W.F., A.K. Moortgat, G.S. Tyndall, and J.P. Burrows, Absorption cross
sections of NO2 in the UV and visible region (200-700 nm) at 298K. J of Photochem
Photobiol: 40, 195, 1987.
Schoeberl, M.R., and D.F. Strobel, The zonally averaged circulation of the middle
atmosphere. J Atmos Sci: 35, 577, 1978.
Schurgers, M., and K.H. Welge, Absorptionskoeffizient von H2 O2 und N2 H4 zwischen
1200 und 2000 A. Zeit Naturforsch: 23A, 1508, 1968.
Selwyn, G., J. Podolske, and H.S. Johnston, Nitrous oxide ultraviolet absorption
spectrum at stratospheric temperatures. Geophys Res Lett: 4, 427, 1977.
Shaw, J., Solar Radiation. Ohio J Sci: 53, 258, 1953.
Shemansky, D.E., CO2 extinction coefficient 1700-3000Å. J Chem Phys: 56, 1582,
1972.
Shved, G.M., L.E. Khvorostovskaya, I.Y. Potekhin, A.I. Demyanikov, A.A. Kutepov,
and V.I. Fomichev, Measurement of the quenching rate constant of CO2 (011 0)-O
collisions and its significance for the thermal regime and radiation in the lower
thermosphere, Atmos Oceanic Phys: 27, 7 295, 1991.
Shved, G.M., Kutepov, A.A., and Ogivalov, V.P., Non-local thermodynamic
equilibrium in CO2 in the middle atmosphere, I. Input data and populations of the
ν3 mode manifold states. J Atmos Solar Terr Phys: 60, 289, 1998.
Silvente, E., R.C. Richter, M. Zheng, E.S. Saltzman, and A.J. Hynes, Relative
quantum yields for O1 D production in the photolysis of ozone between 301 and
336 nm: Evidence for the participation of a spin forbidden channel. Chem Phys
Lett: 264, 309, 1997.
Simon, P.C., D. Gillotay, N. Vanlaethem-Meurée,
´ and J. Wisemberg, Ultraviolet
absorption cross sections of chloro- and chlorofluoro-methane at stratospheric
temperatures. J Atmos Chem: 77, 107, 1988.
Simpson, C.J.S.M., P.D. Gait, and J.M. Simmie, The vibrational deactivation of the
bending moe of CO2 by O2 and by N2 . Chem Phys Lett: 47,7 133, 1977.
Singer, R.J., J.N. Crowley, J.P. Burrows, W. Schneider, and G.K. Moortgat,
Measurement of the absorption cross section of peroxynitric acid between 210 and
330 nm in the range 253-298 K. J Photochem Photobiol: 48, 17, 1989.
Siskind, D.E., K. Minschwaner, and R.S. Eckman, Photodissociation of oxygen and
water vapor in the middle atmosphere: Comparison of numerical methods and impact
on modeled ozone and hydroxyl. Geophys Res Lett: 21, 863, 1994.
Slanger, T.G., and G. Black, Photodissociative channels at 1216 Å for H2 O, NH3 ,
and CH4 . J Chem Phys: 77,
7 2432, 1982.
Smith, F.L., III, and C. Smith, Numerical evaluation of Chapman’s grazing incidence
integral Ch (X, χ). J Geophys Res: 77,
7 3592, 1972.
Sobolev, V.V., A Treatise of Radiative Transfer. D. Van Nostrand, 1963.
Spencer, J.E., and F.S. Rowland, Bromine nitrate and its stratospheric significance.
J Phys Chem: 82, 7, 1978.
Stamnes, K., S.C. Tsay, W. Wiscombe, and K. Jayaweera, Numerically stable
algorithm for discrete-ordinate-method radiative transfer in multiple scattering and
emitting layers. Appl Opt: 27,
7 2502, 1988.
Steinfeld, J. I., S.M. Adler-Golden, and J. W. Gallagher, Critical survey of data on
the spectroscopy and kinetics of ozone in the mesosphere and thermosphere. J Phys
Chem: 16, 911, 1987.
Stief, L.J., W.A. Payne, and R.B. Klemm, A flash-photolysis-resonance fluorescence
study of the formation of O (1 D) in the photolysis of water and the reaction of O
(1 D) with H2 , Ar and He. J Chem Phys: 62, 4000, 1975.
Stockwell, W.R., and J.C. Calvert, The near ultraviolet absorption spectrum of
gaseous HONO and N2 O4 . J Photochem: 8, 193, 1978.
Takahashi, K., Y. Matsumi, and M. Kawasaki, Photodissociation processes of ozone
in the Huggins band at 308-326 nm: Direct observation of O (1 D2 ) and O (3 Pj)
products. J Phys Chem: 100, 4084, 1996.
Takahashi, K., N. Taniguchi, Y. Matsumi, M. Kawasaki, and M.N.R. Ashfold,
Wavelength and temperature dependence of the absolute O(1 D) production yield
from the 305-329 nm photodissociation of ozone. J Chem Phys: 108, 7161, 1998.
Thomas, G.D., and K. Stamnes, Radiative Transfer in the Atmosphere and Ocean.
Cambridge University Press, 517 pp., 1999.
Thompson, B.A., P. Harteck, and R.R. Reeves, Jr., Ultraviolet absorption coefficients
of CO2 CO, O2 , H2 O, N2 O, NH3 , NO, SO2 and CH4 between 1850 and 4000 A. J
Geophys Res: 68, 6431, 1963.
Tiwari, S.N., Models for infrared atmospheric radiation. Adv Geophys: 20, 1, 1978.
Toon, O.B., C.P. McKay, T.P. Ackerman, and K. Santhanam, Rapid calculation
of radiative heating rates and photodissociation rates in inhomogeneous multiple
scattering atmospheres. J Geophys Res: 94, 16,287, 1989.
RADIATION 263
Turnipseed, A.A., G.L. Vaghjiani, J.E. Thompson, and A.R. Ravishankara,

Photodissociation of HNO3 at 193, 222, and 248 nm: Products and Quantum Yields.
J Chem Phys: 96, 5887, 1992.
Urey, H.C., L.C. Dawsey, and F.O. Rice, The absorption spectrum and decomposition
of hydrogen peroxide by light. J Amer Chem Soc: 51, 1371, 1929.
Vaghjiani, G.L., and A.R. Ravishankara, Absorption cross sections of CH3 OOH,
H2 O2 and D2 O2 vapors between 210 and 365 nm at 297 K. J Geophys Res: 94,
3487, 1989.
Van de Hulst, H.C., Light Scattering by Small Particles. Wiley, 1957.
Van Hoosier, M.E., J.-D.F. Bartoe, G.E. Brueckner, and D.K. Prinz, Absolute solar
spectral irradiance 120 nm-400 nm (Results from the Solar Ultraviolet Irradiance
Monitor — SUSIM — experiment on board Spacelab 2). Astrophys Lett Commun:
27,
7 163, 1988.
Van Laethem-Meuree, ´ N., J. Wisemberg, and P.C. Simon, Absorption des
chloromethanes
´ dans l’ultraviolet: Mesures des sections efficaces d’absorption en
fonction de la température.
´ Bull Acad Roy Belgique Cl Sci: 64, 34, 1978a.
Van Laethem-Meuree, ´ N., J. Wisemberg, and P.C. Simon, Influence de la temp´ érature
sur les sections efficaces d’absorption des chlorofluorométhanes
´ dans l’ultraviolet. Bull
Acad Roy Belgique Cl Sci: 64, 42, 1978b.
Vasudev R., Absorption spectrum and solar photodissociation of gaseous nitrous acid
in the actinic wavelength region. Geophys Res. Lett: 17,
7 2153, 1990.
Vernazza, J., E.H. Avrett, and R. Loeser, Structure of the solar chromosphere, II.
The underlying photosphere and temperature minimum region. Astrophys J: 30, 1,
1976.
Vigroux, E., Contribution expérimentale
´ de l’absorption de l’ozone. Ann Phys, Paris:
8, 709, 1953.
Vigroux, E. Coefficients d’absorption de l’ozone dans la bande de Hartley. Ann
Geophys: 25, 169, 1969.
Vodar, M.B., Spectre d’absorption ultraviolet du gaz chlorhydrique et courbe
d’energie potentielle de l’état
´ excité de la molecule ClH. J Phys Rad: 9, 166, 1948.
Wahner, A., G.S. Tyndall, and A.R. Ravishankara, Absoprtion cross sections for
OClO as a function of temperature in the wavelength range 240-480 nm. J Phys
Chem: 91, 2734, 1987.
Wahner, A., A.R. Ravishankara, S.P. Sander, and R.R. Friedl, Absorption cross
section of BrO between 312 and 385 nm at 298 and 223 K. Chem Phys Lett: 152,
507, 1988.
Watanabe, K., and M. Zelikoff, Absorption coefficient of water vapor in the vacuum
ultraviolet. J Opt Soc Amer: 43, 753, 1953.
Watanabe, K., E.C.Y. Inn, and M. Zelikoff, Absorption coefficients of oxygen in the
vacuum ultraviolet. J Chem Phys: 21, 1026, 1953.
Watson, R.T., Rate constants of ClO of atmospheric interest. J Phys Chem Ref Data:
6, 87, 1977.
Wayne, R.P., The photochemistry of ozone. Atmos Environ: 21, 1683, 1987.
Wayne, R.P., I. Barnes, P. Biggs, J.P. Burrows, C.E. Canosa-Mas, J. Hjorth, G.
LeBras, G.K. Moortgag, D. Perher, G. Poulet, G. Restelli, and J. Sidebottom, The
nitrate radical: Physics, chemistry and the atmosphere. Atmos Environ: 25A, 1,
1991.
Williams, A.P., and C.D. Rodgers, Radiative transfer by the 15 micron CO2 band in
the mesophere. Proc Intl Radiation Symposium, Sendai, Japan, 26 May-2 June 1972.
Wilson, R.C., and H.S. Hudson, The sun’s luminosity over a complete solar cycle.
Nature: 351, 42, 1991.
Wintersteiner, P.P., R.H. Picard, R.D. Sharma, J.R. Winick, and R.A. Joseph, Line-
by-line radiative excitation model for the nonequilibrium atmosphere: Application to
CO2 15 µm emission. J Geophys Res: 97, 7 18,083, 1992.
World Meteorological Organisation (WMO), The Stratosphere 1981: Theory and
Measurements. Report No. 11, Geneva, Switzerland, 1982.
World Meteorological Organisation (WMO), Atmospheric Ozone 1985. Global Ozone
Research and Monitoring Project, Report No. 16, Geneva, Switzerland, 1985.
Woods, T.N., and G.J. Rottman, Solar Lyman α irradiance measurements during two
solar cycles. J Geophys Res: 102, 8769, 1997.
Woods, T.N., D.K. Prinz, G.J. Rottman, J. London, P.C. Crane, R.P. Cebula, E.
Hilsenrath, G.E. Brueckner, M.D. Andrews, O.R. White, M.E. VanHoosier, L.E.
Floyd, L.C. Herring, B.G. Knapp, C.K. Pankratz, and P.A. Reiser, Validation of
the UARS solar ultraviolet irradiances: Comparison with the ATLAS 1 and 2
measurements. J Geophys Res: 101, 9541, 1996.
Woods, T.N. G.J. Rottman, S.M. Bailey, S.C. Solomon, and J. Worden, Solar extreme
ultraviolet irradiance measurements during solar cycle 22. Solar Physics: 177,
7 133,
1998.
Yao, F., I. Wilson, and H. Johnston, Temperature dependent ultraviolet absorption
spectrum for dinitrogen pentoxide. J Phys Chem: 86, 3611, 1982.
Yoshino, K., D.E. Freeman, J.R. Esmond, and W.H. Parkinson, High resolution
absorption cross section measurements and band oscillator strengths of the (1,0)-
(12,0) Schumann-Runge bands of O2 . Planet Space Sci: 31, 339, 1983.
Yoshino, K., A.S.-C. Cheung, J.R. Esmond, W.H. Parkinson, D.E. Freeman, and S.L.
Guberman, Improved absorption cross sections of oxygen in the wavelength region
205-240 nm of the Herzberg continuum. Planet Space Sci: 36, 1469, 1988.
Yoshino, K., J.R. Esmond, D.E. Freeman, and W.H. Parkinson, Measurements of
absolute absorption cross sections of ozone in the 185 to 254 nm wavelength region
and the temperature dependence. J Geophys Res: 98, 5205, 1993.
Yoshino, K., J.R. Esmond, Y. Sun, and W.H. Parkinson, Absorption cross section
measurements of carbon dioxide in the wavelength region 118.7-175.5 nm and the
temperature dependence. Spectrosc Ra: 55, 53, 1996a.
Yoshino, K., J.R. Esmond, W.H. Parkinson, K. Ito, and T. Matsui, Absorption cross
section measurements of water vapor in the wavelength region 120 to 188 nm. Japan
Chem Phys: 211, 387, 1996b.
Zhu, X., J.-H. Yee, S.A. Lloyd, and D.F. Storbel, Numerical modeling of chemical-
dynamical coupling in the upper stratosphere and mesosphere. J Geophys Res: 104,
23,995, 1999.
Chapter 5
COMPOSITION AND CHEMISTRY
5.1 General
The chemical composition of the atmosphere near the Earth’s

surface is relatively well known due to numerous measurements of the
concentration and variability of many of the important gases. The
local surface sources of atmospheric species are related to biological
and microbiological activity, volcanic eruptions, and human activities.
Some of the species emitted in the troposphere are also destroyed in
the troposphere by biological, chemical, and physical processes. Others
are transported upward, eventually reaching the stratosphere, where
many can undergo chemical transformations. For most constituents,
natural sources are larger than those of anthropogenic origin. However,
for compounds such as methane (CH4 ), carbon monoxide (CO), nitric
oxide (NO), sulfur dioxide (SO2 ), and nearly all halocarbons, the
anthropogenic production has become larger than the natural emissions.
Some minor constituents are produced or destroyed in the atmosphere
by photochemical reactions. The altitude at which these processes occur
depends on the penetration depth of the solar radiation which initiates
the photochemistry.
In this chapter, the chemical and photochemical processes relevant to
the principal constituents of the middle atmosphere will be discussed.
Knowing the distribution and behavior of the various constituents,
global atmospheric budgets can be derived. The budgets of atmospheric
methane, carbon monoxide, molecular hydrogen, nitrous oxide, nitrogen
oxides, halocarbons, inorganic chlorine, bromine and fluorine, as well
as sulfur compounds and ozone, will be discussed in detail. Global
perturbations of the chemical composition in the middle atmosphere,
including the formation of a springtime ozone hole over Antarctica, will
be discussed in Chapter 6. The chemistry of ions will be addressed in
Chapter 7.
The terrestrial atmosphere is made up of a multitude of chemical
constituents, the most abundant being molecular nitrogen (N2 ) and
oxygen (O2 ). The relative concentrations of these compounds are
265
essentially constant in the homosphere, as was mentioned in Chapter 3.

Noble gases (Argon, Neon, Helium, Krypton, Xenon) are also insensitive
to chemical processes. The mean atmospheric mole fraction (also
called the mixing ratio) of well-mixed gases and the corresponding total
atmospheric mass content for dry air are given in Table 5.1 below:
Table 5.1 Mean Atmospheric Mole Fraction and Total Mass Content
Compound Mole fraction Total Mass Content (kg)
Nitrogen (N2 ) 78.084 × 10−2 3.9 × 1018

Oxygen (O2 ) 20.946 × 10−2 1.2 × 1018
Carbon dioxide (CO2 ) 370 × 10−6 7.5 × 1014
Argon (Ar) 9340 × 10−6 6.6 × 1016
Neon (Ne) 18.18 × 10−6 6.5 × 1013
Helium (He) 5.239 × 10−6 3.7 × 1012
Krypton (Kr) 1.14 × 10−6 1.7 × 1013
Xenon (Xe) 0.086 × 10−6 2.0 × 1012
Air 1 5.1 × 1018
In the middle atmosphere, molecular nitrogen is particularly stable

since it cannot be photodissociated below the mesopause. On the
other hand, the photodissociation of molecular oxygen can occur at
altitudes as low as 20 km. Where transport processes can replace
the photodissociated molecules, their abundances remain constant, but
as photodissociation rates increase at higher altitudes, mixing ratios
begin to decline. Oxygen photolysis initiates a series of reactions which
determine the chemistry of the “oxygen atmosphere”; these will be the
subject of the next section.
The abundance of atmospheric species (see Box 5.1) is variable in
space and time, and can be calculated from the continuity equation,
which expresses mass conservation in an infinitesimal volume element.
For each chemical compound i, continuity can be equivalently expressed
by the flux form
∂ρi
+ ∇ · (ρi v) = S̃i (5.1a)
∂t
or by the advective form
∂µi S̃i
+ v · ∇µi = (5.1b)
∂t ρ
where ρi is the mass density (expressed for example as kg of constituent
i per cubic meter of air), µi = ρi /ρ is the mass fraction (expressed for
example as kg of constituent i per kg of air), v is the wind velocity
COMPOSITION AND CHEMISTRY 267
Box 5.1 Definitions and Units

Number density: Number of particles (e.g., atoms, molecules,
or aerosol particles) of substance per unit volume of air expressed, for
example, in particles m−3 (SI) or particles cm−3 (cgs).
Mass density: Mass of substance per unit volume of air expressed,
for example, in kg m−3 (SI) or in g cm−3 .
Mole fraction: Number density of substance divided by number
density of air (dimensionless variable). The mole fraction is generally
expressed in cmol/mol, mmol/mol, µmol/mol, nmol/mol, and pmol/mol,
which corresponds to mole fractions of 10−2 , 10−3 , 10−6 , 10−9 , and
10−12 , respectively. The terms parts per million by volume (ppmv), parts
per billion per volume (ppbv), or parts per trillion per volume (pptv) are
often being used to express mixing ratios by volume. Volume mixing ratio
is equivalent to mole fraction only for an ideal gas.
Mass fraction: Mass density of substance divided by mass density
of air (dimensionless variable). The pertinent unit for mass mixing ratio
is kg/kg, but this quantity is often expressed as g/kg or g/g. Abundances
expressed in µg/g, ng/g, and pg/g correspond to mass fractions of 10−6 ,
10−9 , and 10−12 , respectively. The terms ppmm, ppbm, and pptm are
often used to characterize mixing ratios by mass.
Column abundance: Vertically integrated number density of
substance (expressed in particles m−2 (SI) or cm−2 (cgs)). In the case
of ozone, the column abundance is often expressed in Dobson units (DU).
One Dobson unit corresponds to the height (in 10−3 cm) of an ozone
column if the gas were at standard temperature and pressure. It is
equivalent to 2.687 × 1016 molecules cm−2 .
Particle or mass flux: Number of particles or mass of substance
crossing a unit surface per unit time. The appropriate units (SI) are
particles m−2 s−1 or kg m−2 s−1 , but particles cm−2 s−1 or g cm−2 s−1
are often used. Total mass flux of a substance (emitted, for example, at
the Earth’s surface) is often expressed in Tg yr−1 (1012 g yr−1 ).
Atmospheric lifetime: Average time that a substance spends in
the atmosphere. Atmospheric residence time is a synonymous concept,
expressed in days, months, or years. If the removal of the substance is
directly proportional to its mass density, the lifetime is also a measure of
the time needed to reduce the mass density of the compound by a fraction
e(2.718), if there are no sources.
vector, S̃i is the net chemical source term (expressed for example by kg
of species i per cubic meter of air and per second), ρ is the air mass
density (kg m−3 ) and t is time (s). Noting that the total derivative is
given by
d ∂
= +v·∇
dt ∂t
equation (5.1b) can also be expressed as
dµi S̃i
= (5.1c)
dt ρ
which shows that, in the absence of chemical source or sink, the mass
fraction is a conserved quantity. The continuity equation can also
be expressed in terms of number density ni (expressed for example in
number of particles i per cubic meter of air):
∂ni
+ ∇ · (ni v) = Si (5.1d)
∂t
where Si is the net source term, now expressed for example in particles
per cubic meter of air and per second. Note that, as the loss term for
species i is often proportional to its concentration ni , the net source term
is often expressed as
Si = Pi − Li ni
where Pi is the gross production and Li is the loss coefficient. In the
homosphere (below 90-100 km altitude), Eq. (5.1d) can be written in
terms of mole fraction or volume mixing ratio Xi = ni /n, and
dXi Si
= (5.1e)
dt n
if n is the air number density, Thus, in regions where the mean molecular
mass of air is constant, the mole fraction is also a conserved quantity in
the absence of chemical conversion processes. When chemical reactions
take place, the source terms for species i such as Si depend on the
concentration of other compounds nj , and the single continuity equation
has to be replaced by a system of N equations, if N is the number
of interacting chemical species. The product φi = ni v represents the
flux of constituent i which can be represented in a one-, two-, or three-
dimensional space. In numerical models, continuity equations must be
solved simultaneously for all of the many species under consideration
with appropriate initial and boundary conditions. There are a variety of
numerical methods which can be used to provide solutions to this coupled
set of equations (see, e.g., Brasseur and Madronich, 1992). Algorithms
frequently used to solve a system of nonlinear chemical equations
(continuity equations in which the transport terms are neglected) are
briefly discussed in Box 5.2.
Box 5.2. Numerical Solutions for Chemical Continuity

Equations
A variety of numerical methods can be used to solve the non-linear

differential equation
∂y
= S (t, y) (5.2)
∂t
which represents the continuity equation (5.1). In this equation, y, a
vector of N elements (y1 , . . . , yN ), represents the density or the mixing
ratio of N chemically coupled compounds. Vector S (t, y) accounts for
transport, chemical production and destruction processes at time t, and
is called the source term. In many cases, when the chemical lifetimes of the
N chemical species (which are provided by the inverse of the eigenvalues
λi of the Jacobian matrix ∂S/∂y) vary by several orders of magnitude, the
system is said to be “stiff”, and appropriate numerical algorithms must
be used.
The performance of numerical methods for chemical continuity
equations is generally characterized in terms of accuracy, stability, degree
of mass conservation, and computational efficiency. The simplest of such
methods is provided by the forward Euler or fully explicit scheme, by
which the solution yn+1 at time tn+1 is given by
yn+1 = yn + ∆t S (tn , yn ) (5.3)
where tn = t0 + n∆t (t0 is the initial time), ∆t = tn+1 − tn is the
time step. This solution, in which the source term is evaluated at time
tn , is represented strictly in terms of known quantities. The algorithm
is therefore straightforward, but rarely practical (except for long-lived
species), because numerical stability requires that the time step ∆t be
constrained by the following condition:
2
∆t ≤ = 2ττmin
λmax
where λmax is the largest eigenvalue λj of the Jacobian matrix, and τmin
the smallest chemical lifetime of the species involved.
The Eulerian backward or fully implicit method

yn+1 = yn + ∆t S tn+1 , yn+1 (5.4)
is unconditionally stable for any ∆t, so that the time step may be selected
for accuracy considerations. For stiff systems, very small time steps may
be required, which introduces a computational constraint that may be
as severe as that of explicit methods. Since, in general, the source term
S includes nonlinear terms, the solution cannot be deduced immediately,
and equation (5.4) is frequently solved by iterative methods. Prominent
among the iterative algorithms is the Newton-Raphson method in which
the solution at time tn+1 for iteration (m + 1) is obtained from
Box 5.2 (Continued)

n+1
y(m+1) n+1
= y(m) − J−1 G y(m)
n+1
(5.5)
where G is an N-valued
vector function
n+1 n+1
G y =y − yn − ∆t S tn+1 , yn+1 (5.6)
and J is the Jacobian matrix whose elements are Jij = ∂Gi /∂yj . The
iteration converges to the solution for time steps smaller than a problem-
dependent upper bound.
Another approach is to linearize the source term
S (t, y) = P (t, y) − L (t, y) y (5.7)
where P represents the production and transport rates and L the loss
coefficient. The solution at time tn+1 for iteration (m + 1) can easily be
derived from
n+1
y(m+1) = yn + ∆t Pn+1
(m)
− Ln+1 yn+1
(m) (m+1)
(m = 0, 1, 2, ...) (5.8)

where Pn+1(m)
= P tn+1 , y n+1
(m)
and L n+1
(m)
= L tn+1 , y n+1
(m)
. This
approach is computationally far less expensive than the Newton-Ralpson
iteration since only a diagonal, rather than a full matrix needs to be solved
at each iteration. In both algorithms, the initial guess for the solution at
time tn+1 is the solution at time tn :
n+1
y(0) = yn
Another algorithm is provided by the semi-analytic algorithm
(Hesstvedt et al., 1978) in which the solution of the iith component of
Eq. (5.2) with (5.7) is given by
n n
n+1 Pi n Pi
yi = + yi − exp (−Ln i ∆t) (5.9)
Li Li
where Pi /Li represents the solution at steady-state. This scheme, which
represents the (exact) analytic solution for a scalar linear equation with
constant coefficients, is stable. For a system of coupled equations,
the accuracy of the solution is improved by an iterative procedure.
This computationally efficient algorithm is, however, inherently non-
conservative, although rather accurate solutions are obtained in many
cases.
Single step methods, such as the Eulerian forward and backward
algorithms, can be regarded as particular cases of multistep methods
(Byrne and Hindmarsh, 1975). One of them, developed by Gear (1971) is
particularly well adapted to “stiff” systems. In this algorithm, the solution
to (5.2) is approximated by
Box 5.2 (Continued)

K
yn+1 = αk yn−k + ∆tβ0 S tn+1 , yn+1 (5.10)
k=0
where coefficients αk , β0 , and K, the order of the method (up to
6), are selected to maximize stability and accuracy. Elegant solver
packages build around Gear’s method continuously and automatically
adjust the time step and the order of the method to meet user-specified
error tolerances. Gear’s method is extremely accurate, robust and
stable, but computationally very expensive (since solutions at K previous
integration times need to be stored, and numerous iterations and matrix
manipulations need to be performed), and hence is not practical for multi-
dimensional codes.
The recent application of sparse-matrix techniques combined with
computer optimization (vectorization) techniques has, however, improved
the speed of Gear’s code substantially, so that this advanced algorithm
can now be used to study complex problems in multi-dimensional models
(see e.g., the SMVGEAR package developed by Jacobson (1995; 1998)
and Jacobson and Turco, 1994).
Finally, in order to reduce the “stiffness” of the chemical schemes
under consideration in atmosphere models, fast-reacting compounds can
be grouped into chemical families whose lifetimes are relatively long.
Examples of such families are the odd oxygen, odd hydrogen, odd
nitrogen and odd chlorine families discussed in the present chapter.
Time-dependent solutions for the concentration of the “families” are
found by using simple implicit or even explicit algorithms with relatively
large timesteps. When the chemical species belonging to a given
family cycle rapidly among each other, the concentrations of individual
family members can be computed by assuming photochemical equilibrium
conditions between these family members. The advantage of this method
is that it is relatively fast since large time steps can be used; the
disadvantage is that care must be exercised in choosing the group and
assuring that the equilibrium assumptions are valid. In many cases, e.g.,
during nighttime, the partitioning between family members cannot be
derived from equilibrium considerations.
5.2 Oxygen Compounds
Ozone plays a key role in the chemical and radiative budgets of the
middle atmosphere. By absorbing solar energy at wavelengths shorter
than approximately 320 nm, it protects the biosphere from harmful solar
radiation, and specifically from DNA damage in living cells, from animal
and human skin damage, including skin cancer, and from eye cornea and
lens damage.
Ozone was discovered in 1839 at the University of Basel, Switzerland,
by the German chemist C.F. Schönbein,
¨ who had detected the odor of a
gas produced in the laboratory without being able to identify its chemical
structure. Schönbein
¨ named this gas ozone, from the Greek word óζiν
(ozein, to smell). In 1845, M. de la Rive in Geneva, Switzerland,
indicated that this gas was most likely produced by the transformation
of oxygen, and in 1863, J.L. Soret, also in Geneva, suggested that ozone
was made of 3 oxygen atoms.
Early measurements of atmospheric ozone were provided by chemical
analysis (starting in 1853 by Schönbein
¨ himself in Vienna, Austria, in
1858 by A. Houzeau in Rouen, France, and systematically from 1877
to 1907 by A. Levy at the Parc Montsouris in Paris). Spectroscopic
observations were also made, involving the contributions of A. Cornu
(1878 in France), J. Chappuis (1880 in France), W.N. Hartley (1880/81
in the United Kingdom), W. Huggins (1890 in the UK), A. Fowler
and R.J. Strutt (1917 in England), C. Fabry and H. Buisson (1920
in France), F.W.P. Götz¨ (1924 in Switzerland), and G.M.B. Dobson
(starting in 1924 in the UK). Today, the most accurate measurements
of the vertically integrated ozone concentration (also called the ozone
column abundance) are still provided by UV spectrometers similar to
the instrument developed by Dobson in 1928.
The first theory for the formation and destruction of stratospheric
ozone involving only oxygen reactions was presented by the British
geophysicist S. Chapman at a scientific conference held in France in 1929
(Chapman, 1930). Additional mechanisms were later invoked to account
for the destruction of ozone in the mesosphere and stratosphere (see
e.g., Nicolet 1975a), including the contribution of hydrogen compounds
(Bates and Nicolet, 1950), of nitrogen oxides (Crutzen, 1970; Johnston,
1971), of chlorine species (Stolarski and Cicerone, 1974), and of
bromine compounds (Wofsy et al., 1975). The role of industrially
produced chlorofluorocarbons as a major source of stratospheric chlorine,
and hence as a threat for ozone, was pointed out by Molina and
Rowland (1974a,b). In the 1980s, record low ozone concentrations
were reported in Antarctica (Chubachi, 1985; Farman et al., 1985)
during springtime. The cause of this ozone hole was attributed to
the presence of anthropogenic chlorine in the stratosphere, enhanced by
surface chemistry on polar stratospheric clouds (Solomon et al., 1986;
Molina and Molina, 1987).
The next section presents the major photochemical processes affecting
the formation and destruction of ozone in a “pure oxygen atmosphere”.
The following sections describe the roles of other chemical species in the
budget of stratospheric and mesospheric ozone.
5.2.1 Pure Oxygen Chemistry
The photodissociation of molecular oxygen by ultraviolet radiation

at wavelengths less than 242.4 nm produces atomic oxygen (Chapman,
1930):
(J2 ); O2 + hν → O + O (5.11)
These atoms may recombine in a three body process:
(k1 ); O + O + M → O2 + M (5.12)
or they may react with molecular oxygen to produce ozone:
(k2 ); O + O2 + M → O3 + M (5.13)
Reaction (5.12) plays a significant role only in the thermosphere, where
the concentration of atomic oxygen is large (i.e., 1012 cm−3 at 100 km).
Reaction (5.13) is the only known mechanism in the atmosphere that
leads to the formation of an ozone molecule. Ozone can recombine with
atomic oxygen according to the reaction
(k3 ); O + O3 → 2O2 (5.14)
which leads to the formation of two oxygen molecules. This reaction,
which is strongly temperature dependent, provides a loss for O and O3 .
The photodissociation of ozone leads to formation of oxygen atoms (see
Chapter 4), in either their ground state (3 P)
(J3 ); O3 + hν (λ ≥ 320 nm) → O2 (3 Σ− 3
g ) + O( P) (5.15)
or in their first excited state (1 D)
(J∗3 ); O3 + hν (λ ≤ 320 nm) → O2 (1 ∆g ) + O(1 D) (5.16)
In the upper part of the middle atmosphere, some O(1 D) atoms can also
be produced by molecular oxygen photolysis:
(J∗2 ); O2 + hν (λ < 175.9 nm) → O(1 D) + O(3 P) (5.17)
When molecular oxygen is in the excited 1 ∆g state, it can be deactivated
by collision with ground state oxygen molecules O2 (3 Σ− g ):
(k6 ); O2 (1 ∆g ) + O2 (3 Σ− 3 −
g ) → 2O2 ( Σg ) (5.18)
or it can relax radiatively by emitting a photon in the infrared at
λ = 1.27µm
(A1.27 ); O2 (1 ∆g ) → O2 (3 Σ−
g ) + hν (5.19)
The observation of this emission can be used to deduce the concentration

of O2 (1 ∆g ) and O3 (see, e.g., Evans et al., 1968; Thomas et al., 1983;
Mlynczak and Marshall, 1996; Mlynczak, 1999). O(1 D) is rapidly
quenched by collision with N2 or O2 :
(k4a ); O(1 D) + N2 → O(3 P) + N2 (5.20)
(k4b ); O(1 D) + O2 → O(3 P) + O2 (1 Σ+

g) (5.21)
and, in the thermosphere, by collision with ground state atomic oxygen
(k4c ); O(1 D) + O(3 P) → 2 O2 (5.22)
It can also react with ozone
(k5 ); O(1 D) + O3 → 2O2 (5.23)
and other chemical compounds, including water vapor and nitrous oxide
(see following sections). Because it is produced almost exclusively by
photolysis reactions, the concentration of O(1 D) is strongly dependent
on local photochemical conditions. Figure 5.1 shows the distribution of
O(1 D) calculated for several solar zenith angles.
5.2.2 The Odd Oxygen Family and Oxygen Partitioning
It is useful to examine the photochemistry of the pure oxygen system

described in the previous section to provide a framework to which other
processes may be added and to illustrate the concept of chemical families.
Concentration of O(1 D) atoms at various solar zenith angles. From

Nicolet (1971).
Stratospheric chemical cycle affecting odd oxygen species in the

stratosphere. The numbers in boxes represent concentrations (cm−3 ) calculated
at 25 km altitude while the numbers associated with the arrows account for
the reaction fluxes ((cm−3 s−1 ) between different compounds (24 ( hour global
average conditions). Note that the figure extends beyond the simple “pure
oxygen chemistry” case and that NO2 is identified as an odd-oxygen reservoir.
(See following sections for more details; from Zellner, 1999).
Considering the reactions given in Section 5.2.1, and omitting

transport, the continuity equations for O3 , O(3 P), O(1 D), and O2 (1 ∆g )
can be written as follows:
d(O3 )
+ (J3 + J3∗ )(O3 ) + k3 (O)(O3 ) + k5 (O(1 D))(O3 )
dt
= k2 (M)(O2 )(O) (5.24)
d(O)
+ 2k1 (M)(O)2 + k2 (M)(O2 )(O) + k3 (O3 )(O)
dt
= 2J2 (O2 ) + J2∗ (O2 ) + J3 (O3 ) + k4a (N2 )(O1 D)
+ k4b (O2 )(O1 D) (5.25)
d(O1 D)
+ [k4a (N2 ) + k4b (O2 ) + k5 (O3 )] (O1 D)
dt
= J3∗ (O3 ) + J2∗ (O2 ) (5.26)
and
d(O2 1∆g )
+ A1.27 (O2 1∆g ) + k6 (O2 )(O2 1∆g ) = J3∗ (O3 ) (5.27)
dt
where (O2 ) represents the concentration of ground state O2 (3 Σ−

g ) and
3
(O) is the density of O( P). A similar equation could be written for
ground state molecular oxygen, but would not be independent since
total oxygen is always conserved:
3d(O3 ) 2d(O2 ) 2d(O2 1∆g ) d(O) d(O1 D)
+ + + + =0 (5.28)
dt dt dt dt dt
The photochemical lifetimes of these species as determined only by the
above reactions are:
1
τO3 = ∗ ≈ 2000 s at 30 km (5.29)
J3 + J3 + k3 (O) + k5 (O 1 D)
1
τO = ≈ .04 s at 30 km (5.30)
k2 (O2 )(M) + k3 (O3 ) + 2k1 (M)(O)
1
τO1 D = ≈ 10−8 s at 30 km (5.31)
k4a (N2 ) + k4b (O2 ) + k5 (O3 )
1
τO2 (1 ∆g ) = ≈ 1 s at 30 km (5.32)
k6 (O2 ) + A1.27
The chemical loss of O2 (1 ∆g ) produces a ground state oxygen
molecule. Since O2 is a permanent gas always present at a
constant mixing ratio in the middle atmosphere, this process is not
photochemically important. Further, the lifetime of O2 (1 ∆g ) is never
more than its radiative lifetime of about an hour, so that photochemical
steady state (see Eqs. (2.37) and (5.27)) may be assumed for this species:
J3∗ (O3 )
(O2 1∆g ) ≈ (5.33)
A1.27 + k6 (O2 )
(See also, Thomas et al., 1983, for more details regarding O2 (1 ∆g )
chemistry).
The O3 , O, and O(1 D) equations are a coupled set of partial
differential equations whose time constants differ by many orders of
magnitude. Such a set of equations is called a “stiff” system (see
Box 5.1). It cannot be easily solved by simple methods, unless a timestep
of the order of the smallest lifetime is used for the numerical integration.
In view of this problem, it is very useful to define chemical families,
whose lifetimes can be very much longer than those of the constituent
members. Adding equations (5.24), (5.25), and (5.26) one obtains an
equation describing the behavior of the sum of O(3 P), O(1 D), and O3 ,
which is generally referred to as the odd oxygen family, Ox :
(Ox ) = (O3 P) + (O1 D) + (O3 ) . (5.34)
One can then write:
d(Ox )
+ 2k3 (O3 )(O) + 2k5 (O(1 D))(O3 ) + 2k1 (M)(O)2
dt
= 2(J2 + J2∗ )(O2 ) = 2JO2 (O2 ) (5.35)
where JO2 = J2 + J2∗ is the total photolysis rate of O2 . The chemical

lifetime of the odd oxygen family is then given by
(Ox )
τOx ≈
2k1 (M)(O)2 + 2k3 (O)(O3 ) + 2k5 (O1 D)(O3 ) (5.36)
≈ weeks at 30 km and 1 year at 20 km
Figure 5.3 presents an altitude profile of the photochemical lifetimes
of the Ox family (including important reactions with hydrogen, nitrogen,
and chlorine containing species, which will be introduced later), O3 , and
O, as well as an estimate of the approximate time constants associated
with transport by the winds and a typical vertical diffusive lifetime.
In this and other figures involving lifetimes to be presented in this
chapter, τD represents the vertical diffusive lifetime (mean vertical
mixing), assuming a vertical scale height of 5 km. τu , τu∗ and τw∗
represent the time constants for transports by the zonal, meridional,
Figure 5.3. Daytime photochemical lifetimes of Ox , O3 , and O, and

characteristic transport lifetimes for mid-latitude winter, from the model of
Garcia and Solomon (see Garcia and Solomon, 1994, and references therein).
and vertical winds at middle latitudes, assuming characteristic scales

of 1000, 1000, and 5 km, respectively. These values are based on an
updated version of the model by Garcia and Solomon (1983). We will
discuss the implications of the magnitudes of the chemical and dynamical
time constants in determining the behavior of these species below.
The definition of the odd oxygen family clearly produces a substantial
increment in the photochemical time constant of the equation to be
considered, enabling us to solve it more readily. The very fast reactions,
such as
O3 + hν → O(1 D) + O2 (5.16)
O( D) + O2 → O + O2 ( Σg )
1 1
(5.21)
O + O2 + M → O3 + M (5.13)
now produce only an exchange, or partitioning, among members of

the family, but do not appear as production or loss terms in the
continuity equation for the family. The use of chemical families is useful,
however, only if the time constants characterizing exchange between
family members are short compared to the dynamical time constants.
Since the net transport of a chemical species depends in part upon its
gradient, any species whose lifetime is sufficiently long that transport
may play a role in establishing its distribution should be transported
according to its own individual gradient. The family definitions may
also depend on the problem under consideration. For example, if the
detailed photochemistry of the rapidly changing solar illumination found
at twilight is of special interest, then the families used may be different
from those appropriate to a study of variations on a seasonal time scale.
The use of chemical families allows a clearer distinction to be drawn
between reactions which represent net and gross production and loss
terms over the time scale considered. For example, the photolysis of
O3 (Reactions (5.15) and (5.16)) is a gross, but not a net loss term for
ozone at altitudes below about 80 km over time scales longer than a
few seconds, because nearly all the oxygen atoms which are formed by
this reaction rapidly reform ozone (Reaction 5.13), implying no net loss.
On the other hand, when ozone reacts with atomic oxygen to form O2
(Reaction 5.14), a net loss of ozone occurs over an extended time scale.
Further, since a typical time scale for meridional transport is of the order
of months, it is convenient to consider odd oxygen, rather than ozone
or atomic oxygen, since it is only odd oxygen which has a sufficiently
long lifetime to be influenced by meridional transport processes. The
importance of transport is often more easily seen when families are
considered.
This analysis allows us to obtain the concentrations of the odd oxygen

family, but it must also be partitioned into its constituent parts. This
can be done by examining the photochemical equilibrium expressions for
certain members of the family. If the family is composed of N members,
we must write photochemical equilibrium expressions for the shortest
lived N − 1 members of the family and use these to establish ratios
between family members. Below the mesopause, the photochemical time
constant for atomic oxygen is relatively short and one can assume that
it is essentially in photochemical equilibrium. Similar conditions apply
to the excited O(1 D) atom. Neglecting minor terms, Eq. (5.26) becomes
d(O1 D)
≈ 0 ≈ J3∗ (O3 ) + J2∗ (O2 ) − [k4a (N2 ) + k4b (O2 )] (O1 D) (5.37)
dt
or
J3∗ (O3 ) + J2∗ (O2 ) J3∗ (O3 )
(O1 D) ≈ ≈ (5.38)
k4a (N2 ) + k4b (O2 ) k4a (N2 ) + k4b (O2 )
The partitioning between O(1 D) and ozone is thus expressed by the
following ratio:
(O1 D) J3∗
≈ (5.39)
(O3 ) k4a (N2 ) + k4b (O2 )
Similarly, Eq. (5.25) becomes
d(O3 P)
≈ 0 ≈ J3 (O3 ) + [k4a (N2 ) + k4b (O2 )] (O1 D) + 2JO2 (O2 )
dt
− [k2 (M)(O2 ) + k3 (O3 )] (O3 P)
(5.40)
Substituting for O(1 D) from the expression above, the concentration of
O(3 P) at equilibrium is
JO3 (O3 ) + 2JO2 (O2 ) JO3 (O3 )

(O3 P) ≈ ≈ (5.41a)
k2 (M)(O2 ) + k3 (O3 ) k2 (M)(O2 )
and the ratio between O(3 P) and ozone concentration is
(O3 P) JO 3
≈ (5.41b)
(O3 ) k2 (M)(O2 )
The partitioning between the members of the Ox family is given by
(O3 ) 1
= 1+(O3 P) (O1 D) (5.41c)
(Ox ) +(O3 ) (O3 )
and
(O) (O3 ) (O)
= × (5.41d)
(Ox ) (Ox ) (O3 )
In the stratosphere ozone is much more abundant than atomic oxygen

(O(3 P) and O(1 D); see Table 5.2), so that (Ox ) ≈ (O3 ). Neglecting
Reaction (k1 ) and (k5 ) in Eq. (5.35) and assuming that d(O3 P)/dt
and d(O1 D)/dt ≈ 0, the odd oxygen equation can be written (using
Eq. 5.41a):
d(Ox ) 2k3 JO3 (O3 )2
+ = 2JO2 (O2 ) (5.42)
dt k2 (M)(O2 )
In the upper stratosphere and lower mesosphere, odd oxygen (and ozone)
is short-lived, and is approximately in photochemical equilibrium, so
that this may be rewritten as:

k2 JO2 1/2
(O3 )eq = (M)(O2 )2 (5.43)
k3 JO 3
for pure oxygen chemistry. Below about 25 km the photochemical
lifetime of odd oxygen becomes comparable to or longer than the
transport time scales (as shown in Figure 5.3), so that both chemistry
and dynamics influence its concentration. This will be discussed in
more detail below. Table 5.2 presents typical values of the daytime
distribution of oxygen species in the atmosphere (calculated with
hydrogen, nitrogen and chlorine reactions taken into account).
In the upper part of the mesosphere and in the thermosphere, atomic
oxygen becomes more abundant than ozone. At these heights, ozone
is short-lived due to rapidly increasing photolysis rates, and, during
daytime, is in photochemical equilibrium with atomic oxygen. The
lifetime of atomic oxygen becomes longer and longer with increasing
altitude due to its slower recombination rate (see Eqs. (5.12) and (5.13))
as the air pressure decreases (see Figure 5.3). It can be deduced from
Eq. (5.24) that when d(O3 )/dt=0,
k2 (M)(O2 )(O) k2 (M)(O2 )(O)

(O3 ) = ≈ (5.44)
JO3 + k3 (O) + k5 (O(1 D)) JO 3
and from (5.38), setting d(Ox )/dt ≈ d(O)/dt,

d(O) k3 k2 (M)(O2 )
+ 2 k1 (M) + (O)2 = 2JO2 (O2 ) (5.45)
dt JO 3
The photochemical time constant τ (O) for atomic oxygen is about
4 hours at 70 km, a day near 80 km and about a week at 100 km.
Therefore, the distribution of atomic oxygen above the mesopause is
dependent on dynamical conditions. It does not exhibit much diurnal
variation above about 85 km, since its lifetime exceeds a day. At lower
altitudes, on the other hand, atomic oxygen disappears rapidly after
sunset, forming ozone through the recombination reaction (5.13). In
Table 5.2 Example of the Vertical Distributions of Oxygen Com-

pounds in the Middle Atmosphere
Altitude (O2 ) (O3 ) (O 3 P) (O 1 D)
(km) (cm−3 ) (cm−3 ) (cm−3 ) (cm−3 )
0 5.3 (18)∗ 8.0 (11) 1.0 (3) 2.7 (–3)

5 3.2 (18) 7.7 (11) 4.6 (3) 9.0 (–3)
10 1.8 (18) 7.5 (11) 1.2 (4) 1.8 (–2)
15 8.8 (17) 2.2 (12) 1.3 (5) 1.2 (–1)
20 3.8 (17) 4.8 (12) 1.6 (6) 8.9 (–1)
25 1.7 (17) 4.9 (12) 8.6 (6) 4.6 (0)
30 7.8 (16) 3.1 (12) 3.3 (7) 2.0 (1)
35 3.6 (16) 1.3 (12) 1.2 (8) 6.7 (1)
40 1.6 (16) 3.9 (11) 3.8 (8) 1.4 (2)
45 8.0 (15) 1.2 (11) 1.1 (9) 2.0 (2)
50 4.2 (15) 4.1 (10) 2.0 (9) 1.9 (2)
55 2.3 (15) 1.5 (10) 2.7 (9) 1.5 (2)
60 1.2 (15) 5.7 (9) 3.3 (9) 1.1 (2)
65 6.3 (14) 2.2 (9) 3.8 (9) 7.5 (1)
70 3.2 (14) 6.0 (8) 3.5 (9) 4.0 (1)
75 1.6 (14) 1.3 (8) 2.8 (9) 1.9 (1)
80 7.1 (13) 9.3 (7) 1.0 (10) 3.0 (1)
85 4.8 (13) 1.7 (8) 1.1 (11) 1.2 (2)
90 2.3 (13) 4.0 (7) 4.5 (11) 1.0 (2)
95 1.0 (13) 5.7 (6) 8.6 (11) 2.0 (2)
100 4.6 (12) 2.5 (6) 1.1 (12) 1.0 (3)
∗ 5.3 (18) denote 5.3 × 1018 molecules/cm3 .
the middle and lower stratosphere, the abundance of atomic oxygen is

very small compared to that of ozone, and the lifetime of ozone is greater
than a day. Therefore, relatively small diurnal variations are expected to
occur in stratospheric ozone concentrations. In the mesosphere, atomic
oxygen densities are comparable to, and even greater than those of ozone
(see Table 5.2 and Figure 5.4), so that the recombination reaction leads
to ozone increases throughout the mesosphere at the beginning of the
night (and decreases after sunrise).
5.2.3 Observations of Ozone
Most of the ozone molecules in a vertical column are found at altitudes

ranging from 10 to 35 km (see Table 5.2). Measurements of the total
ozone abundance can be performed from the ground using ultraviolet
absorption techniques (see e.g., Dobson, 1963) or from space. Total
Calculated vertical distributions of ozone and atomic oxygen at

noon and midnight. From the model of Garcia and Solomon (1983).
ozone has been systematically observed for many years and at numerous
locations. Therefore, the general spatial and temporal behavior of
this gas in the atmosphere is very well documented (Dütsch, ¨ 1970;
1980; London, 1980). For example, it is well known that ozone
is characterized by latitudinal and seasonal variations that include a
maximum abundance in the region of the least production (see, for
example, Dobson, 1963). Figure 5.5 shows the total ozone column versus
month for several locations in the northern hemisphere. These values
are expressed in Dobson units (DU), which correspond to the height
(in millicentimeter) that the ozone column would have if all the gas were
at standard temperature and pressure (1 DU = 10−3 atm. cm = 2.69 ×
1016 molecules cm−2 ). It should be noted that ozone is most abundant
at higher latitudes during most seasons. This increase with latitude is
most pronounced in winter and spring. The seasonal variations only
become large at latitudes poleward of 30 degrees north. At 80◦ N, for
example, the relative variation in total ozone is about 50 percent over
six months.
The global distribution of total ozone as a function of latitude and
time deduced from the observations by the Total Ozone Mapping
Spectrometer (TOMS) on board the Nimbus 7 satellite is presented
in Figure 5.6. This figure shows that the ozone column abundance
is typically 260-270 DU in the tropics. The peak total ozone
Figure 5.5. Mean seasonal variation of the total ozone abundance at different
latitudes in the northern hemisphere. From Dobson (1963).
Figure 5.6. Variation of total ozone (DU) with latitude and season in the year
1990, as measured by the TOMS instrument.
abundance (typically 400 DU) appears in the Arctic in late winter.

The maximum observed in the southern hemisphere (approximately
360 DU) is somewhat smaller, and located at lower latitudes (60-65◦ S)
and earlier in the season compared to the maximum occurring in the
northern hemisphere. The very low ozone column abundance observed
from September to November over Antarctica since the late 1970s (see
Chapter 6 and Plates 8 and 9) is referred to as the ozone hole. Ozone
columns as low as 100 DU are observed during the Antarctic springtime.
During the 1990s, low ozone concentrations have also been recorded in
the Arctic spring. These anomalous ozone levels are associated with
cold temperatures in the polar lower stratosphere and the activation of
chlorine on surfaces of polar stratospheric cloud particles (see Chapters 2
and 6).
Since the chemical lifetime of odd oxygen in the lower stratosphere
is long compared to the time constants associated with transport
by the mean meridional circulation (Figure 5.3), transport by mean
meridional advection can play an important role in determining the
distribution of ozone. Much of the hemispheric asymmetry in total
ozone distributions as shown in Figure 5.6 is generally presumed to be
due to differences in the stratospheric dynamics of the two hemispheres.
As first pointed out based on empirical evidence by Dobson (1930),
the primary mechanism accomplishing this transport can be simply
understood to a first approximation in terms of the downward and
poleward net transport associated with mean motion in the transformed
Eulerian or diabatic framework, as discussed in Chapter 3. This implies
that the slopes of the mean ozone mixing ratio surfaces are somewhat
steeper than the slopes of the isentropic surfaces represented by potential
temperatures, as pointed out by Newell (1964), and discussed in more
detail by Tung (1982).
A more detailed understanding of the behavior of ozone requires
examination of its altitude profile, so that the regions responsible for
the observed total column variations can be identified. Figure 5.7
shows some typical vertical distributions of ozone for spring and fall
at different latitudes of the Northern Hemisphere (Dütsch,
¨ 1980). It is
clear that the changes in total column abundance are due to variations
found in the ozone concentration between the tropopause and 20 km
altitude. Dobson (1963) suggested that this behavior is related to
ozone transport in cyclones and anticyclones (e.g., weather systems) that
propagate into the lower stratosphere. Experimental support for this
view comes from the observation of close correlations between high total
ozone abundances and low pressure systems, as shown, for example, in
Figure 5.8. As a result of these processes, large local fluctuations in total
ozone are often observed, with amplitudes sometimes as large as those
Figure 5.7. Mean vertical distribution of the ozone partial pressure ((10−4 hPa)
for different latitudes of the Northern Hemisphere in April and October (D¨ utsch,
¨
1980).
Figure 5.8. Relationship between the variations in total ozone and the altitudes
of the tropopause and the 200 hPa level. From Dobson (1963).
Figure 5.9. Evolution of the total ozone column abundance during 1994 at
Uccle, Belgium (De Muer, private communication).
related to the seasonal variations (see Figure 5.9). Dynamical studies

suggest that transport above 100 hPa can be broadly understood in
terms of a residual Eulerian mean motion in the middle stratosphere
(see Section 3.6), but that considerable dispersion related to cyclone
propagation dominates the behavior for lower altitudes (see, e.g., Holton
et al., 1995).
It is sometimes stated that ozone is produced where its mixing ratio
maximizes in the upper stratosphere of the tropics and transported
downward and poleward to higher latitudes, thus producing the observed
structure in the ozone column. But closer examination reveals a different
picture. Figure 5.3 shows that the lifetime of odd oxygen is less
than a month above about 30 km in the sunlit atmosphere, decreasing
to only about a day near 40 km. The tropical ozone mixing ratio
maximum occurs between about 30 and 40 km (see Figure 5.10), in
the heart of this region of short lifetime. Under such conditions, ozone
is essentially in photochemical equilibrium, with rapid production being
balanced by rapid destruction — it does not live long enough to be
transported away from this photochemically-controlled zone. Since the
time scale for meridional overturning of the stratosphere is slow (order
of years) compared to these chemical phenomena, the tropics cannot
serve as the source region for the large amounts of total ozone found in
Figure 5.10. Zonally averaged ozone mixing ratio (ppmv) in the stratosphere
((15-55 km)) in January. Climatological values derived from observations by the
Halogen Occultation Experiment ((HALOE)) and the Microwave Limb Sounder
((MLS)) on board the Upper Atmosphere Research Satellite (UARS). (Courtesy
of W. Randel, NCAR).
the extratropical lower stratosphere, which dominate the total column

(see Figure 5.7; also Ko et al., 1989; Perliski et al., 1989).
The primary region that supplies ozone to the extratropical lower
stratosphere is a narrow “transition” zone as shown in Figure 5.11.
Air flowing across this zone moves through rapidly enough to retain
relatively high mixing ratios compared to those found at lower levels.
The transition zone slopes upward in the winter hemisphere, due to
latitude-dependent changes in the propagation of the solar radiation
that drives photochemistry and determines the lifetime. Thus the source
for the extratropical ozone column lies not in the tropics, but in the
meridional flow poleward and downward from this transition zone — a
zone of competition between transport and chemistry.
Satellite observations have added much to our understanding of the
morphology of atmospheric ozone, both in terms of its altitude profile
and total column density (see Box 5.3). For example, early observations
by the Backscatter Ultraviolet Spectrometer (BUV) on board Nimbus 4,
as well as by the Limb Infrared Monitor of the Stratosphere (LIMS), and
the Solar and Backscatter Ultraviolet Spectrometer (SBUV) on board
Nimbus 7 led to the first global view of the distribution of ozone
LOG10 OX PHOTOCHEMICAL LIFETIME (sec)

116
8
106
7
Dynamical
96 Control
6
86
ALTITUDE (km)
76 4
Chemical
66 Control 6
7
56 8
46
5
36
26 Dynamical
Control
16
-80 -60 -40 -20 0 20 40 60 80
1138 fig 13.10/1.95 dpm
SUMMER LATITUDE WINTER
Figure 5.11. Photochemical lifetime of the Ox ((O + O3 ) family ((logarithmic

scale of values expressed in seconds). In the region of chemical control, the
photochemical production and destruction are nearly in balance and steady
state conditions can be adopted to derive the concentrations of O3 and O. In
the dynamically controlled regions, photochemical processes play a secondary
role and Ox (primarily O3 in the stratosphere and O in the thermosphere)
is primarily influenced by transport processes. In the shaded region, transport
and photochemistry play an equally important role. From Garcia and Solomon,
1985.
in the stratosphere. Satellite observations of ozone were also provided by

the Stratospheric Aerosol and Gas Experiments (SAGE I and II) and,
in the mesosphere, by the Solar Mesospheric Explorer (SME). More
recently, an operational version of SBUV (SBUV-2) has been collecting
ozone data, while the Upper Atmosphere Research Satellite (UARS)
has provided measurements of ozone and a range of related chemical
constituents in the stratosphere and mesosphere. Figure 5.10 presents
a climatological (zonal mean) distribution of the ozone mixing ratio
in January from 100 hPa (approximately 16 km) to 0.3 hPa (55 km)
based on the observations of 2 UARS instruments. It is interesting to
note that the maximum mixing ratio (9.6 ppmv) is located at 10 hPa
(30 km) in the equatorial region, while in the lower stratosphere, the
ozone abundance increases with latitude due to poleward transport (as
indicated previously in terms of the observed total column abundances).
Box 5.3. Platforms to Observe the Chemical Composition of the

Middle Atmosphere
The earliest observations of the chemical composition of the middle
atmosphere were performed primarily by balloon-borne (stratosphere) and
rocket-borne (mesosphere) instrumentation. In recent decades, however,
systematic observations have been made by space-borne sensors, and
information on chemical processes in the upper troposphere and lower
stratosphere has been provided through chemical measurements made
from aircraft platforms (e.g., from high-altitude aircraft such as the ER-
2, the WB-57 and the Geofysika). Measurements from ground stations,
combined with satellite observations, have provided accurate estimates of
long-term ozone trends in the different regions of the world.
Satellite observations provide information on the global distribution
and seasonal variation of chemical compounds. Measurements of vertically
or horizontally integrated radiances (atmospheric absorption of solar
radiation or emission of atmospheric radiation) over a given spectral range
(UV-visible, infrared, microwave) are inverted by appropriate algorithms
to derive the concentration of selected chemical species. Examples of
space instruments include TOMS (Total Ozone Mapping Spectrometer,
measuring the ozone column abundances), SBUV (Solar Backscatter
Ultraviolet, measuring the ozone column and the vertical distribution of
ozone), SAGE (Stratospheric Aerosol and Gas Experiments, measuring
the vertical distributions of ozone, nitrogen dioxide, water vapor and
aerosols), GOME (Global Ozone Monitoring Experiment, measuring the
vertical distribution of ozone, nitrogen dioxide and other minor species).
Instruments on board the Upper Atmosphere Research Satellite (UARS)
included HALOE (Halogen Occultation Experiment, measuring ozone,
NO, NO2 , HCl, HF, CH4 , and H2 O by infrared solar occultation), MLS
(Microwave Limb Sounder, measuring ozone, ClO, H2 O, and HNO3 by
a microwave emission technique) and CLAES (Cryogenic Limb Array
Etalon Spectrometer, measuring ozone, NO, NO2 , HNO3 , N2 O, N2 O5 ,
ClONO2 , CFCl3 , CF2 Cl2 , CH4 , H2 O, and aerosols using an infrared
limb emission technique). The vertical distribution of many chemical
compounds including key species belonging to the nitrogen and chlorine
families was provided by the ATMOS instrument (Fourier Transformed
infrared technique for absorbed solar infrared radiation) on board the
Space Shuttle. The CRISTA (Cryogenic Infrared Spectrometers and
Telescopes for the Atmosphere) and MAHRSI (Middle Atmosphere High
Resolution Spectrograph Investigation), also on the Space Shuttle, have
provided high-resolution distributions of several chemical compounds in
the stratosphere and a near-global distribution of OH in the mesosphere,
respectively. Additional information has been obtained from MIPAS (IR
limb sounder), SCIAMACHY (UV-visible and near-IR nadir and limb
viewer) and GOMOS on board the ENVISAT space platform. Obser-
Box 5.3 (Continued)

vations of important trace constituents at high spatial resolution in the
stratosphere and upper troposphere should be provided by HIRDLS
(High Resolution Dynamics Limb Sounder), TES (Tropospheric Emission
Spectrometer), OMI (Ozone Monitoring Instrument), and an improved
version of MLS (Microwave Limb Sounder) on the Aura platform.
One of the great advantages of space observations is that they provide

global information on spatial and short-term temporal (e.g., synoptic
scale) variations. Figure 5.12, which presents the ozone mixing ratio
observed by LIMS at 10 hPa on 24 January 1979, shows that planetary
waves, which propagate into the middle atmosphere during winter,
perturb significantly the distribution of this chemical compound. The
displacement from the North Pole of ozone-poor airmasses, caused by
planetary wave disturbances, and the strong poleward flow of subtropical
ozone (which propagates counterclockwise around the vortex) are typical
of winter situations. If planetary waves break, (for example, during
a stratospheric warming event), irreversible transport takes place, and
90°E
120°E 60°E
8
150°E 30°E
180°W
150°W 30°W
20°N
120°W 60°W
90°W
Figure 5.12. Ozone mole fraction (ppmv) on the isobaric surface of 10 hPa
measured by the LIMS instrument (Nimbus 7) on 24 January, 1979. Note the
intrusion of subtropic ozone into the polar region during a planetary scale wave
disturbance. Courtesy of L. Lyjak and J. Gille, NCAR.
contributes to the ozone maximum observed in late winter/early spring

at high latitudes. Similar disturbances are less common in the Southern
Hemisphere where a minimum in the ozone abundance prevails over the
pole during the entire winter. The rapid decrease in the abundance of
lower stratospheric ozone observed in September and October is caused
by photochemical destruction inside the Antarctic vortex and results
from the presence of anthropogenic chlorine compounds (see Chapter 6);
it is facilitated by the stability and the strength of the vortex, which does
not allow ozone to be easily transported south of 60◦ S.
Ozone observations near 90 km available from the Solar Mesosphere
Explorer are characterized by strong latitudinal gradients and large local
variations. Barth et al. (1983) suggested that this may result from
downward transport of oxygen from the thermosphere because of the
long Ox lifetime at these altitudes, as shown in Figure 5.3.
To conclude this discussion, we turn to the question of longer term
variations in ozone. Ramanathan (1963), Angell and Korshover (1964;
1973), and Shah (1967) were among the first to note the existence of an
oscillatory component in the total ozone abundance which is correlated
with the behavior of the zonal winds in the tropical lower stratosphere.
These winds change speed and direction with a period between 24
and 33 months; this is the quasi-biennial oscillation (QBO) noted in
Chapter 3. Figure 5.13 shows an analysis of ozone QBO based on
13 years of TOMS observations and Plate 6 ozone anomalies associated
with QBO in the lower tropical stratosphere. Another question related
to long-term variability is that of a possible relationship between
Figure 5.13. Global patterns of ozone QBO deduced from observations by

TOMS. Contour intervals correspond to three Dobson units with solid lines for
positive anomalies and dashed lines for negative values. From Yang and Tung
(1995).
ozone and the 11 year solar cycle. This has long been the subject
of vigorous controversy (Willett, 1962; London and Haurwitz, 1963;
Willett and Prohaska, 1965; Paetzold et al., 1972; Angell and Korshover,
1973). Observational studies suggest a connection between the chemical
and thermal structure of the middle atmosphere and variations in the
solar ultraviolet flux. This is certainly true for short-term (27-day)
variations (associated with the apparent solar rotation), as revealed by
analyses of satellite data (Gille et al., 1984; Hood, 1987; Keating et al.,
1987; Hood et al., 1991; Hood and Jirikowic, 1991) and model studies
(Eckman, 1986; Brasseur et al., 1987; Summers et al., 1990; Brasseur,
1993). Convincing statistics are more difficult to obtain for the 11-year
period, although more recent satellite data analyses suggest a correlation
of stratospheric ozone with the solar cycle (Chandra, 1991; Hood and
McCormack, 1992; Hood et al., 1993). Model studies (Brasseur, 1993;
Huang and Brasseur, 1993) imply that the variation in total ozone over
the solar cycle is probably less than two percent. Detection of such a
small variation is difficult because of the natural variability in total ozone
due to its dependence on dynamical conditions, especially at middle and
high latitudes (Figures 5.5 and 5.9).
Finally, and perhaps more importantly, statistical analyses of ground-
based and satellite ozone observations suggest that, on the global scale,
the abundance of stratospheric ozone has decreased significantly over
the 1980s and 1990s. The downward trend is most marked at mid- and
high latitudes, and is believed to result largely from the production of
industrially manufactured chlorofluorocarbons. This question will be
further discussed later in this chapter and in Chapter 6.
5.3 Carbon Compounds
The most abundant carbon-containing compound in the stratosphere

and mesosphere is carbon dioxide (CO2 ). By interacting with infrared
radiation, this gas plays an important role in the thermal budget of the
atmosphere, and the 30% increase in its concentration resulting mainly
from fossil fuel burning has provided a significant forcing to the climate
system of about 1.5 Wm−2 (IPCC, 2001). Carbon dioxide does not
play any substantial role in the chemistry of the atmosphere except in
the lower thermosphere, where its photolysis is an important source of
carbon monoxide (CO). This latter gas, which is also released at the
Earth’s surface by incomplete combustion (pollution) and is partially
transported to the stratosphere, is converted to CO2 by reaction with
the hydroxyl radical (OH).
Methane (CH4 ), another important carbon-containing species and

also a radiatively important gas of the atmosphere, is released at the
surface as a result of biogenic processes in oxygen-deficient environments
such as wetlands. It is also emitted to the atmosphere in connection
with mining activities and natural gas consumption. The oxidation of
methane produces a number of partially oxygenated organic compounds
such as formaldehyde (CH2 O) and methyl peroxide (CH3 OOH), as well
as carbon monoxide and carbon dioxide. This degradation mechanism
tends to destroy ozone and OH radicals in NO-poor air masses and
to produce ozone and OH in NO-rich environments. The reaction
of methane with chlorine atoms converts reactive chlorine atoms into
stable hydrogen chlorine (HCl) and hence influences the impact of man-
made halocarbons on stratospheric ozone (see Section 5.6.2 and 5.8).
Methane is also a source of water vapor in the middle atmosphere, and
the increase in its surface emissions associated with agricultural and
industrial activities has probably enhanced the abundance of H2 O in
the stratosphere and mesosphere.
The following sub-sections present the budget of carbon-containing
species and describe the photochemical processes affecting these species
in the middle atmosphere.
5.3.1 Methane
The chemistry of atmospheric methane in the middle atmosphere was

first considered by Bates and Nicolet (1950). Its destruction is due to
reactions by the OH radical:
(c2 ); CH4 + OH → CH3 + H2 O (5.46)
by excited oxygen O(1 D):
(c1a ); CH4 + O(1 D) → CH3 + OH (5.47a)
(c1b ); CH4 + O(1 D) → CH2 O + H2 (5.47b)
(c1c ); CH4 + O(1 D) → CH3 O (or CH2 OH) + H (5.47c)
or by chlorine atoms:
(d5 ); CH4 + Cl → CH3 + HCl (5.48)
Branching ratios for (5.47a), (5.47b), and (5.47c) are approximately 75%,
5%, and 20%, respectively. Above the stratopause, it is also necessary
to introduce the photolysis of CH4 , particularly by the Lyman α line of
the Sun:
(JCH4 ); CH4 + hν → CH3 + H (5.49a)
→ CH2 + H2 (5.49b)
→ CH2 + H + H (5.49c)
→ CH + H2 + H (5.49d)
The quantum efficiencies for these different channels vary with
wavelength. The values quoted in Chapter 4 (see also Brownsword et al.,
1997) for Lyman α excitation need to be confirmed. CH2 produced by
two of these channels reacts rapidly with O2 to form CH2 O + O, HCO +
OH, CO + H2 O and CO2 + H2 (or H + H). The photolysis of methane
produces atomic and molecular hydrogen. However, since the amount
of methane in the upper part of the middle atmosphere is small, CH4
photolysis can generally be neglected in the global atmospheric methane
budget.
Since there is no source of methane in the atmosphere, the
vertical distribution of CH4 results from an equilibrium between its
photochemical destruction and transport upward from the surface. The
continuity equation can therefore be written:
∂(CH4 )
+ ∇ · φCH4 + c1 (O1 D) + c2 (OH) + d5 (Cl) + JCH4 (CH4 ) = 0
∂t
(5.50)
where c1 = c1a + c1b .
Figure 5.14 shows the photochemical lifetime for atmospheric methane
as a function of altitude along with the time constants associated with
transport by the winds and vertical mixing. Since the stratospheric
lifetime of this compound against photochemical destruction is the same
order of magnitude as the transport time, it provides an excellent tracer
to study transport processes.
After the spectroscopic detection of atmospheric methane by Migeotte
(1948), methane has been observed many times and its vertical
distribution has been characterized at various latitudes by optical
methods, or by cryogenic collection and laboratory analysis (balloons,
rockets, space shuttle) (see Ehhalt, 1974; Ehhalt et al., 1974; Ackerman
et al., 1977; Bush et al., 1978; Fabian et al., 1979; Gunson et al., 1990;
and Figure 5.15). These data show that methane is well mixed in the
troposphere with a current mixing ratio of approximately 1.75 ppmv,
while above the tropopause its concentration decreases rapidly. Methane
has also been measured by the Stratospheric And Mesospheric Sounder
(SAMS) on board Nimbus 7, and, more recently, by UARS (HALOE
and CLAES instruments), so that the global distribution of atmospheric
methane can now be characterized, and used to deduce information
about transport processes. Figure 5.16 shows the mixing ratio of
methane measured in the stratosphere and mesosphere by the HALOE
Figure 5.14. Photochemical lifetime of CH4 , and the time constants associated
with transport processes.
Figure 5.15. Vertical distribution of the methane mixing ratio (ppmv).

Observations at various latitudes and one-dimensional model calculations (from
Ehhalt and Tonnissen, 1980.)
Figure 5.16. Methane mixing ratio (ppmv) measured by HALOE (UARS)

during the period 21 September-15 October 1992 and represented as a function
of latitude and pressure (Russell et al., 1993).
instrument on board UARS between 21 September and 15 October 1992

(Russell et al., 1993). The high values in the tropical lower and middle
stratosphere reflect rapid upward transport from the tropospheric source
region in the rising branch of the Hadley cell. The propagation of these
high values into the northern hemisphere also indicates the direction of
flow (see, e.g., Figure 3.26 and Plate 2).
As we have already noted, CH4 is not photochemically produced in the
atmosphere. Rather, the biosphere and lithosphere provide the natural
sources of atmospheric methane. Most of the biological production is
attributed to the effects of anaerobic bacteria such as methanobacterium,
methanococcus, and methanosarcina. These are primarily found in
alkaline regions, of relatively high temperature and enriched in organic
materials. Peat bogs, swamps, ponds, rice paddies, and humid tundra
produce large amounts of methane. Table 5.3 presents an estimate of the
magnitude of these sources, and shows that the most likely natural and
anthropogenic sources of methane represent 155 Tg/yr and 360 Tg/yr,
respectively. A large fraction of the anthropogenic sources is associated
with agricultural processes (cattle, rice cultivation, biomass burning in
the tropics) and waste management.
Human production of methane is also associated with industrial and
mining activities. Comparison of the methane isotopic ratio (13 C/12 C,
Table 5.3 Global Budget for Methane (Tg/year)1
Sources Range Likely

Natural
Wetlands
Tropics 30-80 60
Northern Latitudes 20-60 40
Others 5-10 10
Termites 10-50 20
Oceans 5-50 10
Freshwater 1-25 5
Others 8-13 10
..................................................................
Total Natural 155
Anthropogenic
Fossil Fuel Related 100
Coal Mines 15-45
Natural Gas 25-50
Petroleum Industry 5-30
Coal Combustion 7-30
Waste Management System
Landfills 20-70 30
Animal Waste 20-30 25
Domestic Sewage Treatment ? 25
Enteric Fermentation 65-100 80
Biomass Burning 20-80 40
. . . . . . . .Rice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-100
. . . . Paddies . . . . . . . . . . . . . . . . .60
.....
Total Anthropogenic 360
Total Source 515
Sinks
Reaction with OH 330-560 445
Removal in Stratosphere 25-55 40
Removal by Soils 15-45 30
. . . . . . . .Atmospheric . . . . . . . . . . . . . . . . . . . . . . . . .30-40
. . . . . . . . . . . .Increase . . . . . . . . . . . . . . . .37
.....
Total Sink 552
1 1 Tg = 1012 g Source: WMO (1995)
lifetime is given by:
4900 Tg
≈ 8 years (5.51)
600 Tg produced per year
14 C/12 C, and D/H) observed in the atmosphere versus that in organic

matter can be used to estimate the magnitude of these non-biological
sources (see e.g., Quay et al., 1991; 1999; Lassey et al., 1993; Lowe et al.,
1994). Such studies indicate that this latter source must represent about
20 percent of the total production of methane, i.e., between 65 and
170 Tg/yr. Measurement of the isotopic abundance must, however, be
considered an upper limit due to the possibility of sample contamination.
A more detailed analysis of the methane budget is given, for example,
by Crutzen (1983), Crutzen and Gidel (1983), Cicerone and Oremland
(1988), and IPCC (1990). IPCC (2001) reports that 60% of the emissions
are related to human activities including agriculture, fossil fuel use, and
waste disposal. See also the model synthesis of the global methane
cycle by Fung et al. (1991) and the inverse modeling studies of Hein
et al. (1997) and Houweling et al. (1999).
One may conclude that approximately 600 Tg of methane are
produced each year. Since the total atmospheric burden of methane
is about 4900 Tg (corresponding to a mean tropospheric mixing ratio
of about 1.75 ppmv), a global atmospheric lifetime of 8 years can be
derived.
Analyses of air bubbles trapped inside ice cores extracted from
Antarctica and Greenland suggest that the concentration of methane
has more than doubled over the last 200 years (Etheridge et al., 1998),
with a corresponding radiative effect representing a forcing of 0.5 Wm−2
(about 20% of the direct radiative forcing due to greenhouse gases since
the pre-industrial era; see IPCC, 2001). The growth rate has fluctuated
as a function of time: It was as high as 1%/yr in the late 1970s and
seems to have declined to 0.5-0.75%/yr in the late 1980s, and to even
lower values in the 1990s. The cause of the fluctuations (inter-annual
variability) in the methane trends is not entirely clear, but recent changes
in the emissions from natural gas (Dlugokencky et al., 1994; 1998) and
from tropical biomass burning (Lowe et al., 1994) may have contributed.
A positive anomaly in growth rate observed after the eruption of
Mt. Pinatubo in 1991 has been attributed to short-term decreases in
solar UV radiation and related decrease in OH formation rates in the
troposphere (Dlugokencky et al., 1996). Variability in wetland emission
rates and fire recurrence associated with climatic anomalies could be
additional factors.
5.3.2 Methane Oxidation Chemistry
The oxidation of methane produces a methyl CH3 radical, which

reacts rapidly with molecular oxygen:
(c4 ); CH3 + O2 + M → CH3 O2 + M (5.52)

to produce the methyl peroxy radical. This species can be destroyed by
reaction with NO, with itself, with ClO, or with HO2 . The first three
processes lead to the formation of the methoxy radical (CH3 O):
(c5 ); CH3 O2 + NO → CH3 O + NO2 (5.53)
(c14b ); CH3 O2 + CH3 O2 → 2CH3 O + O2 (5.54a)
(c6 ); CH3 O2 + ClO → CH3 O + ClOO (5.55)
The CH3 O2 + CH3 O2 reaction, however, is characterized by several
pathways: The major channel (with 55-60% probability at 298 K) leads
to the formation of formaldehyde and methanol (CH3 OH)
(c14a ); CH3 O2 + CH3 O2 → CH2 O + CH3 OH + O2 . (5.54b)
Methanol is further converted to formaldehyde by
(c16 ); CH3 OH + OH → CH2 O + HO2 + H2 O . (5.56)
If Reaction (5.53) is followed by NO2 photolysis, NO2 + hν → NO
+ O, then the net effect is to produce odd oxygen. This reaction
is an important source of ozone in the lower stratosphere and upper
troposphere (see Crutzen, 1971). Methoxy is in turn destroyed by
reaction with molecular oxygen:
(c15 ); CH3 O + O2 → CH2 O + HO2 (5.57)
to produce formaldehyde.
Reaction
(c7 ); CH3 O2 + HO2 → CH3 OOH + O2 (5.58)
produces methyl peroxide, which either reacts rapidly with OH to reform
the methyl peroxy radical
(c17a ); CH3 OOH + OH → CH3 O2 + H2 O (70%) (5.59a)
or to produce formaldehyde
(c17b ); CH3 OOH + OH → CH2 O + OH + H2 O (30%) (5.59b)
CH3 OOH can also be photolyzed, producing a methoxy and hydroxyl
radical
(JCH3 OOH ); CH3 OOH + hν → CH3 O + OH , (5.60)
and can precipitate in raindrops or aerosols. The primary loss process of
CH3 OOH in the stratosphere is, however, the reaction with OH (5.59).
Regardless of the chemical path taken (with the exception of the
precipitation process in the troposphere), the carbon atom of the
methane molecule must eventually end up as formaldehyde. It should
be noted that the formation of CH2 O also can be accomplished directly

via (5.47b). In the mesosphere and lower thermosphere, reaction with
atomic oxygen can also play a role:
(c3 ); CH3 + O → CH2 O + H (5.61)
Formaldehyde was first observed in the stratosphere using infrared
absorption (see e.g., Barbe et al., 1979). During daytime, the calculated
CH2 O profile results from an equilibrium between the production
mechanisms just described and destruction by OH
(c8 ); CH2 O + OH → CHO + H2 O (5.62)
by atomic oxygen
(c9 ); CH2 O + O → CHO + OH (5.63)
by chlorine atoms
(d10 ); CH2 O + Cl → CHO + HCl (5.64)
and by photolysis. Photodissociation of CH2 O occurs via two distinct
pathways which lead either to the formation of the reactive CHO and H
radicals
(JH−CHO ); CH2 O + hν → CHO + H (5.65a)
or to the longer-lived species, CO and molecular hydrogen:
(JH2 −CO ); CH2 O + hν → CO + H2 (5.65b)
The CHO radical has a very short lifetime because it is rapidly
converted to CO either by photolysis
(JCHO ); CHO + hν → CO + H (5.66)
or by reaction with O2
(c12 ); CHO + O2 → CO + HO2 (5.67)
Figure 5.17 schematically illustrates the presently understood reaction
sequence for carbon compounds in the middle atmosphere. We may
now write the continuity equations for the carbon-containing reactive
intermediates, which are all sufficiently short-lived that transport may
be neglected. Thus,
d(CH3 )
+ [c3 (O) + c4 (O2 )(M)] (CH3 )
dt (5.68)
1
= c1a (O D) + c2 (OH) + d5 (Cl) (CH4 )
d(CH3 O2 )
+ [c5 (NO) + c7 (HO2 ) + 2(c14a + c14b )(CH3 O2 )
dt
+c6 (ClO)] (CH3 O2 ) (5.69)
= c4 (O2 )(M)(CH3 ) + c17a (CH3 OOH)(OH)

Figure 5.17. Schematic diagram of the methane oxidation scheme in the middle
atmosphere.
d(CH3 O)
+ c15 (O2 )(CH3 O) = [c5 (NO) + 2c14b (CH3 O2 )
dt
+c6 (ClO)] (CH3 O2 ) + JCH3 OOH (CH3 OOH) (5.70)
d(CH3 OOH)
+ [JCH3 OOH + (c17a + c17b )(OH)] (CH3 OOH)
dt
= c7 (CH3 O2 )(HO2 ) (5.71)
d(CH3 OH)
+ c16 (OH)(CH3 OH) = c14a (CH3 O2 )2 (5.72)
dt
d(CH2 O)
+ [c8 (OH) + c9 (O) + JH−HCO + JH2 −CO + d10 (Cl)
dt
+c10 (NO3 )] (CH2 O) = c1b (O1 D)(CH4 )
+ c3 (O)(CH3 ) + c15 (O2 )(CH3 O) + c14a (CH3 O2 )2
+ c16 (OH)(CH3 OH) + c17b (OH)(CH3 OOH) (5.73)
d(CHO)
+ [JCHO + c12 (O2 )] (CHO)
dt
= [JH−HCO + c8 (OH) + c9 (O)
+d10 (Cl) + c10 (NO3 )] (CH2 O) (5.74)
Assuming photochemical equilibrium for each species, the daytime

concentration of formaldehyde can be approximately estimated from
% &
c1 (O 1 D) + c2 (OH) + d5 (Cl) (CH4 )
(CH2 O) = (5.75)
JH−HCO + JH2 −CO + c8 (OH) + c9 (O)
It should be noted that the methane oxidation chain also results in the
production and destruction of the hydrogen compounds H, OH and HO2 :
d(H) d(OH) d(HO2 )
+ +
dt dt dt
+ [c2 (CH4 ) + c8 (CH2 O) + c17a (CH3 OOH)] (OH)
+ c7 (CH3 O2 )(HO2 ) = (CH2 O) [(JH−HCO + c9 (O)] (5.76)
+ (HCO) [JHCO + c12 (O2 )] + (CH3 OOH)JCH3 OOH

+ c1a (O1 D)(CH4 ) + c15 (O2 )(CH3 O) + c3 (O)(CH3 )
5.3.3 Some End Products of Methane Oxidation: Carbon

Monoxide and Carbon Dioxide
The relative efficiency of the two paths of formaldehyde photolysis

(Reactions (5.65a) and (5.65b)) is an important parameter because it
determines the fraction of methane which ultimately produces molecular
hydrogen. Reactions (5.62), (5.63) and (5.64), on the other hand,
produce CHO, which in turn yields hydrogen containing radicals through
reactions (5.66) and (5.67). As we will see in Section 5.4, these hydrogen
radicals recombine to produce water vapor. Therefore, the destruction
of methane by this long chain eventually produces water vapor and
molecular hydrogen in the stratosphere and mesosphere, such that the
sum of the hydrogen mixing ratios
4XCH4 + 2XH2 O + 2XH2 O
is approximately constant with altitude (neglecting other much less
abundant hydrogen species such as OH and H). The increase with
height in water vapor abundance observed in the stratosphere (see
below) can be qualitatively explained by this mechanism. Each oxidized
methane molecule also produces carbon monoxide (CO). Noting that the
photolysis of CO2
(JCO2 ); CO2 + hν → CO + O (5.77)
also plays a key role in CO production in the upper mesosphere and
in the thermosphere, one may write the production rate for CO in the
middle atmosphere:

PCO = c1 (O1 D) + c2 (OH) + d5 (Cl) (CH4 ) + JCO2 (CO2 ) (5.78)
On the global scale, taking into account stratospheric as well as
tropospheric processes, the oxidation of methane by the chemical
processes just described leads to an annual production of CO of
about 400-800 Tg. The contribution of the oxidation of nonmethane
hydrocarbons, mostly in the troposphere, is similar in magnitude (200-
600 Tg CO/yr). These sources depend directly on the mean value
of the OH radical density and constitute about 50 percent of the
total CO production. Surface emissions of CO by combustion of
gasoline and natural gas represent an injection of about 300-650 Tg/yr
through combustion (IPCC, 1996; 2001), with a distribution (more
than 90% in the northern hemisphere) that depends on the location
of anthropogenic sources. These emissions are comprised of automobile
and industrial effluents and are localized in industrial areas. Thus, the
direct anthropogenic production of CO is globally of the same order of
magnitude as the production via hydrocarbon oxidation. It is, therefore,
not surprising that the observed abundance of CO varies greatly with
latitude, especially in the troposphere. Observations made at the surface
(see e.g., Novelli et al., 1998) show that the mole fraction of carbon
monoxide in the “remote” locations typically increases from 50±15 ppbv
in the southern polar regions to 140±70 ppbv in the northern hemisphere
at high latitudes. The quoted variations around the mean values reflect
the seasonal variations. Thus, in the marine boundary layer, the mole
fraction of CO reaches a maximum of 200-210 ppbv in the northern
winter and a minimum of 35-45 ppbv in the southern summer. In
regions affected by urban and industrial pollution, concentrations reach
higher values. Tropospheric carbon monoxide has also been observed
from space (e.g., from the Space Shuttle, see Reichle et al., 1990, and
by the MOPITT instrument on board the EOS/Terra satellite).
Other sources of CO which contribute substantially to the global
budget include: domestic combustion of wood for heat, agricultural
operations and savanna burning, principally in tropical zones, as well
as forest fires and plant respiration (see, Seiler and Crutzen, 1980).
CO sources were discussed by Seiler and Conrad (1987), Khalil and
Rasmussen (1990), and Crutzen and Zimmermann (1991). A budget
based on these estimates is shown in Table 5.4. CO sources have
also been derived using model inversion techniques (Manning et al.,
1997; Bergamasschi et al., 2000; Pétron
´ et al., 2002). In spite of
the uncertainties in the budget figures, it seems likely that the total
emissions of CO are between 1250 and 3000 Tg/yr with as much
as 2/3 being produced from anthropogenic activities, mainly through
Table 5.4 Global Budget of Carbon Monoxide

(Tg CO/yr−1 ) (Adapted from WMO, 1999)
Sources Magnitude
Biomass burning 300 – 700

Fossil fuel burning 300 – 500
Methane oxidation 400 – 800
Non-methane oxidation 200 – 600
Vegetation 20 – 200
Oceans 10 – 200
TOTAL (1230 – 3000)
Sinks Magnitude
Chemical loss by OH 1400 – 3000

Uptake by soil 100 – 600
Removal in the stratosphere ∼100
TOTAL 1600 – 3700
combustion processes. This explains why northern hemispheric data

(see, for example, the data recorded at the Jungfraujoch, Switzerland in
1950-1951 and in the late 1980s; Zander et al., 1989) show a positive
trend of 0.9 percent/yr on the average, while no long-term change
has been reported in the southern hemisphere. Novelli et al. (1998)
have deduced from the measurements performed using a distributed air
sampling network that the globally averaged CO mixing ratio at the
surface over the period 1990-1995 decreased by aproximately 2 ppb/yr.
Carbon monoxide released at the surface is mostly destroyed in the
troposphere, but some of the molecules penetrate to the stratosphere.
The only important destruction process is provided by reaction with
OH:
(c20 = a36 ); CO + OH → CO2 + H (5.79)
and the continuity equation for CO in the middle atmosphere may be
written:
∂(CO)
+ ∇ · φCO + c20 (OH)(CO) = PCO (5.80)
∂t
where PCO is given by expression (5.78).
Figure 5.18 presents the vertical profile of the photochemical lifetime
of CO, as well as the time scales for transport. Because of its long
Figure 5.18. Photochemical lifetime of carbon monoxide, and the time constants
for transport processes.
chemical lifetime, the distribution of CO is expected to depend strongly

on meridional and vertical transport throughout the middle atmosphere.
Figure 5.19 shows a number of observations of the vertical distribution of
CO in the stratosphere and troposphere. The stratospheric abundances
are low and a rapid decrease in mixing ratio is typically observed above
the tropopause. This sharp gradient suggests that transport from the
troposphere to the stratosphere is limited at middle and high latitudes,
as discussed previously (see Chapter 3). The observations display an
increase with altitude in the upper stratosphere, as predicted by models.
CO profiles have also been measured in the mesosphere, particularly by
ground-based microwave methods (e.g., Clancy et al., 1982; Forkman
et al., 2003) and from space by the Improved Stratospheric and
Mesospheric Sounder (ISAMS) on board UARS (Figure 5.20) and by
the Swedish satellite ODIN. Its usefulness as a tracer for mesospheric
transport was examined, for example, by Hays and Olivero (1970), and
by Allen et al. (1981).
Carbon dioxide is another end product of the long chemical methane
oxidation chain. The mixing ratio of CO2 is almost constant with
altitude in the homosphere (typically 370 ppmv in year 2000), but
Figure 5.19. Vertical distribution of the CO mixing ratio observed in the

stratosphere. (From Zander et al., 1981.)
decreases with height in the upper mesophere and lower thermosphere

where it is photolyzed by shortwave ultraviolet radiation. This latter
process provides a major source of CO at high altitudes (see e.g., Lopez-
Puertas et al., 2000) Downward transport of CO from the thermosphere
to the upper stratosphere in the wintertime polar night region (where
CO destruction is insignificant) is evident in the ISAMS and the ODIN
measurements (see Figure 5.20). In the troposphere, CO2 exhibits a
weak seasonal variation due mainly to biological processes at the surface.
Other carbon containing species, although present in very small
amounts, can also provide important information regarding the
photochemistry and dynamics of the middle atmosphere. These species
will not be presented here in detail, but we note, for example, that
measurements of lower stratospheric ethane C2 H6 and propane C3 H8
have been used to deduce atomic chlorine densities there (Rudolph
Figure 5.20. Distribution of the zonal mean carbon monoxide mixing ratio
(expressed in ppmv) in the stratosphere and mesosphere represented as a
function of latitude and atmospheric pressure (or approximate altitude), as
measured on 7 August 2001 by the Sub-Millimetre Radiometer on board the
Odin satellite. Note the downward transport of CO at high latitudes in the
winter hemisphere. From Duprey et al., 2004.
et al., 1981). Ehhalt et al., (1991) have reported a secular trend of

0.9 ± 0.3 percent per year in the tropospheric concentration of ethane
over the northern hemisphere, probably in response to anthropogenic
activities and biomass burning. Short-lived hydrocarbons, (e.g., isoprene
and terpenes released in the atmosphere by vegetation), whose local
concentrations may reach 10 ppbv near their sources at the surface, are
destroyed before they can reach the stratosphere.
5.4 Hydrogen Compounds
The presence in the middle atmosphere of fast-reacting radicals such

as H, OH, and HO2 (collectively called odd hydrogen or HOx ) results
from the transport of water vapor (H2 O), methane (CH4 ) and molecular
hydrogen (H2 ) from the troposphere. These source-gases are oxidized
by hydroxyl radicals (OH) or by electronically excited oxygen atoms
(O(1 D)), or are photolyzed by short-wave ultraviolet radiation as they
reach increasingly high altitudes. HOx radicals are efficient catalysts
for destroying odd oxygen (ozone and atomic oxygen), particularly
in the mesosphere, and in the lower stratosphere and troposphere.
Their atmospheric abundance results from an equilibrium between very

rapid photochemical production and destruction mechanisms. The
recombination of OH with HO2 is a loss of HOx in the middle atmosphere
which leads to the formation of water vapor. The reaction of H with HO2
in the mesosphere and lower thermosphere also produces water, but in
addition, can form molecular hydrogen (H2 ) which becomes abundant in
this region of the atmosphere. Above 100 km, atomic hydrogen becomes
a long-lived species whose relative abundance increases with height. The
escape of hydrogen to space is a slow but persistent loss process for
hydrogen from the Earth system, and may have played a role in the
evolution of the atmosphere over geologic timescales.
The following sections provide more quantitative information about
the global budget of hydrogen compounds and about the photochemical
processes affecting these species.
5.4.1 Water Vapor and Molecular Hydrogen
The effects of hydrogen compounds on the behavior of other chemical

species, particularly that of ozone in the mesosphere, were first examined
by Bates and Nicolet (1950), and have been the subject of numerous
subsequent studies (Hampson, 1966; Hunt, 1966; Hesstvedt, 1968;
Crutzen, 1969; Nicolet, 1971). The high reactivity of the hydrogen free
radicals, especially OH, makes these species of particular importance
in atmospheric chemistry. The compounds that initiate the hydrogen
radical chemistry are methane (for which the budget was discussed in
Section 5.3), water vapor, and molecular hydrogen.
Water is quite abundant on planet Earth, but is mostly confined to
the surface and the interior (see Table 5.5). The atmosphere contains
only a very small fraction of the water in the oceans. Water in the oceans
represents, however, a small fraction of the vast number of molecules of
H2 O contained in hydrated minerals which compose the solid part of the
Earth. Clouds contain only a few percent of the water in the atmosphere,
although they cover about half of the surface of the globe, so that most
of the water in the atmosphere occurs in the vapor phase.
A continuous source of atmospheric water vapor is provided by
evaporation at the ocean surface. For this reason, water vapor is quite
abundant at low altitudes. Its ground level mixing ratio is variable,
changing from about 10−2 in the tropics to about 10−4 in the desert.
Vertical soundings made since 1945 show that the water vapor content
decreases with altitude in the troposphere to reach a minimum mixing
ratio of a few ppmv a few kilometers above the tropopause.
Table 5.5 Volume of Water in the Earth System
Location (106 km3 )

Oceans 1350
Glaciers and ice caps 29
Ground water 8
Freshwater lakes 0.1
Salt lakes and inland seas 0.1
Soil moisture and seepage 0.07
Atmosphere 0.013
Rivers and streams 0.001
Brewer (1949) suggested that water vapor reaches the stratosphere in

the ascending branch of the Hadley cell in the tropics (see Chapter 3).
The stratosphere is very dry (typically a few ppmv of water vapor)
because the tropical tropopause is sufficiently cold (190 to 200K) to
“freeze dry” the air and precipitate out much of the water. The relatively
dry air penetrating the stratosphere is then transported upwards by the
Brewer-Dobson circulation (see Chapter 3) and towards polar regions.
The zonally averaged distribution of water vapor measured by LIMS
on board the Nimbus 7 satellite (1978-1979) shown in Figure 5.21
(Gille and Russell, 1984) suggests that the extremely low water vapor
mixing ratios observed in the tropical lower stratosphere are the result
of the freeze-drying mechanism, but the detailed transport processes
are still subject to debate. For example, cumulonimbus clouds which
penetrate the tropopause have been proposed as a primary mechanism
for tropospheric-stratospheric exchanges in the tropics, so that water
vapor may be supplied to the stratosphere via these convective systems
rather than in a slow large-scale upward transport.
The water budget in the tropical troposphere remains poorly
understood, expecially at high altitudes in the vicinity of the tropical
tropopause layer (TTL). This region, which extends approximately from
12 km to the level of temperature minimum at about 16 km, can be
regarded as the transition zone between the tropospheric Hadley and
the stratospheric Brewer-Dobson circulations. In the tropics, the mean
vertical humidity profile is shaped like a “C” with high values in the
boundary layer (moistened by evaporation from the surface) and in the
TTL. The humidity in the TTL is determined by ascent of water in
generally concentrated, intense, and nearly vertical convective towers,
by detrainment from these convective clouds, by large scale subsidience,
by export to the extratropics, and by evaporation of ice crystals or water
droplets (see e.g., Newell et al., 1997; Pierrehumbert, 1998; Jensen et al.,
1999; Gettleman et al., 2000; Zhu et al., 2000; Vömel
¨ et al., 2002). Cirrus
Figure 5.21. Distribution of water vapor observed by the LIMS satellite

experiment. From Gille and Russell (1984).
clouds are often present in this layer of the atmosphere. In the absence of
ice condensation nuclei, however, supersaturation of air parcels relative
to ice is observed (Jensen et al., 1999; Folkins et al., 2002) above the level
of zero diabatic heating, where air detrained from convective clouds is
transported upward and can penetrate into the stratosphere. Potential
long-term changes in the mechanisms (such as convection) that control
the humidity of the tropical tropopause layer could have major impacts
on the climate system. In the extratropical troposphere, moisture is
determined primarily by ascent and descent motions associated with
large scale wave motions, although local convection can play a significant
role.
In the middle and upper stratosphere, where increasing water vapor
mixing ratios are observed, the formation of water vapor results from
methane oxidation via the mechanisms discussed in Section 5.3. In
the mesosphere, the presence of water vapor leads to the formation of
hydrated cluster ions, such as H+ (H2 O)n , which have been commonly
observed in the D-region (see Chapter 7). H2 O photolyzes in
the thermosphere and upper mesosphere by absorption of shortwave
ultraviolet light (see also Section 4.7.2.4):

(JaH2 O ); H2 O + hν (λ < 200 nm) → H + OH (5.81a)
(JbH2 O ); H2 O + hν (λ < 147 nm) → H2 + O(1 D) (5.81b)

reducing the water vapor mixing ratio and increasing those of the
hydrogen free radicals and of molecular hydrogen. Between 175 and
190 nm, the quantum yield for production of OH and H (Reaction 5.81a)
is equal to 1, while at shorter wavelengths H2 (Reaction 5.81b) is also
formed (with a quantum yield of 11% between 105 and 145 nm). In the
stratosphere and the lower mesosphere, water vapor reacts with O(1 D):
(a1∗ ); H2 O + O(1 D) → 2OH (5.82)
Upper stratospheric and mesospheric water vapor has been measured
by microwave techniques, both from the ground (e.g., Gibbins et al.,
1981; Deguchi and Muhleman, 1982; Bevilacqua et al., 1983) and from
aircraft (Waters et al., 1980). Measurements by the HALOE instrument
on board the Upper Atmosphere Research Satellite (UARS; see Russell
et al., 1993; Summers et al., 1997; and Figure 5.22) show that the
water vapor mixing ratio in the mesosphere is of the order of 6 ppmv
up to 70 km, and that it decreases to values less than 4 ppmv near
75 km. The presence of a local maximum of more than 7.5 ppmv in
the tropics detected by the HALOE instrument between 65 and 68 km
(Summers et al., 1997) is not understood. If confirmed by independent
Figure 5.22. Meridional distribution of water vapor (ppmv) measured in the

upper stratosphere and mesosphere during October and November 1994 by the
HALOE instrument on board the UARS satellite (Summers et al., 1997).
observations, it would imply the existence of a mesospheric source of

H2 O, perhaps of extraterrestrial origin. Summers et al. (2001) have
reported the presence of a narrow layer of enhanced water vapor (10-
15 ppmv) in the Arctic summer (82-84 km) coincident with the presence
of noctilucent clouds (ice particles), suggesting that a high degree of
supersaturation is present in this region (see also von Cossart et al.,
1999). In the lower thermosphere, the H2 O density has been inferred
from ion composition measurements (Arnold and Krankowsky, 1977;
Solomon et al., 1982a).
Figure 5.23 shows the photochemical lifetime of water vapor, and the
time constants characterizing atmospheric transport. The production
of water vapor by methane oxidation is essentially complete by about
50 km. Because the photochemical and vertical transport lifetimes
for this gas are comparable above about 50 km, and because there is
no known chemical source of water vapor in this region, it provides
an excellent tracer for mesospheric transport processes (Allen et al.,
1981; Bevilacqua et al., 1983; Le Texier et al., 1988). In addition to
Figure 5.23. Photochemical lifetimes of water vapor and molecular hydrogen,

and the time constants for atmospheric transport processes in the middle
atmosphere.
its chemical properties, water vapor is radiatively active, and plays an

important role in the atmospheric thermal budget through cooling by
infrared emission.
Long-term measurements of stratospheric water vapor using a balloon-
borne frost point hygrometer have been performed in Boulder, CO
(40◦ N) since 1980 (Oltmans and Hofmann, 1995; Oltmans et al., 2000).
A long-term increase in the H2 O concentration, which is larger than
expected from the oxidation of increasing concentrations of stratospheric
methane, has been reported and may be linked to climate variations,
such as a small temperature increase in the troposphere (SPARC, 2000;
Rosenlof et al., 2001). Randel et al. (2004) showed by analyzing HALOE
observations that tropopause temperatures are a primary control for
interannual changes in global stratospheric water vapor, but could not
detect a long-term trend in H2 O during the 1990s.
Because of its long chemical lifetime in the stratosphere, water vapor
should be expected to exhibit variations which are related to atmospheric
transport. Early observations, especially those of Kley et al. (1979, 1980)
suggested that the water vapor profile is highly structured in discrete
layers, with a vertical scale of about 1000 m or less (Figure 5.24a).
These variations imply a stratified structure in the lower part of the
middle atmosphere, and the absence of much vertical mixing in the
region of the measurements. The vertical transport of water vapor in the
tropical lower stratosphere is characterized by the upward propagation
of a succession of wet and dry air masses in response to the seasonal
cycle of convective activity in the equatorial troposphere. This so-called
“tape-recorder” signal (see Figure 5.24b) has been identified by Mote
et al. (1996).
Molecular hydrogen (H2 ) also provides a source of the free radicals H
and OH, because it reacts rapidly with the O(1 D) atom:
(a3∗ ); H2 + O(1 D) → H + OH (5.83)
The destruction of hydrogen also occurs through the following reactions
(a19 ); H2 + OH → H2 O + H (5.84)
and
(a24 ); H2 + O → OH + H (5.85)
which become significant above 40 km. Mathematical models (see e.g.,
Hauglustaine et al., 2002) show that these processes lead to a global H2
loss rate of the order of 15 Tg H2 /yr in the troposphere, and 1 Tg H2 /yr
in the stratosphere. In addition to this chemical loss, possible loss of
hydrogen at ground level must also be considered, and this is likely to
vary with the type of vegetation, temperature, etc. The uptake rate into
Figure 5.24a. Simultaneous observations of ozone and water vapor by Kley

et al. (1980).
various soils (see e.g., Ehhalt, 1973; 1999; Novelli et al., 1999) depends
primarily on microbial activity in soils, soil texture and soil moisture
(Conrad and Seiler, 1985). Estimates of global H2 uptake range from 40
to 90 Tg H2 /yr.
In the troposphere and in the lower stratosphere, the chemical
production of H2 is due to the photolysis of formaldehyde produced by
methane oxidation. About 13 Tg H2 /yr are produced by this mechanism
(Schmidt et al., 1980). Oxidation of isoprene and other terpenes in
the lower troposphere leads to an additional production of about 10-
35 Tg H2 /yr. A precise determination of these quantities requires a
Height-time cross section of the water vapor (ppmv) measured

by HALOE during two consecutive years in the equatorial region highlighting
the “tape recorder” signal in the annual cycle of H2 O.
better understanding of the spatial distribution of formaldehyde. In

the mesosphere, molecular hydrogen can be formed by recombination
of the H and HO2 radicals, which are themselves produced by water
vapor photolysis. This mechanism plays a secondary role in the global
hydrogen budget, producing large H2 densities in the mesosphere and
lower thermosphere (e.g., Liu and Donahue, 1974).
Formation of H2 in soils by anaerobic bacteria is possible, but this
contribution can be neglected for atmospheric purposes because these
hydrogen molecules are rapidly destroyed at ground level by other
microorganisms. The oceans are, however, supersaturated in hydrogen,
and constitute an important source; this source is more important than
that provided by volcanoes. Schmidt et al. (1980) estimate that natural
ground level production is between 2 and 5 Tg/yr.
Several observations have shown that hydrogen is affected by pollution
sources. The hydrogen molecule is an important product of incomplete
combustion in several anthropogenic and natural processes. The
production of H2 by industrial combustion and automobiles is estimated
to vary from 15 Tg H2 /yr (Novelli et al., 1999) to 25 Tg H2 /yr (Schmidt,
1974). According to Crutzen et al. (1979) 9 to 21 Tg H2 /yr are produced
by forest and savanna burning, especially for agricultural purposes in
tropical regions.
Table 5.6 summarizes the components of the global hydrogen budget.
The total amount of hydrogen in the atmosphere is about 200 Tg (170 Tg
Table 5.6 Global Budget of Molecular Hydrogen

Based on the estimates of Novelli et al. (1999)
and Ehhalt (1999)
Production Tg/yr
Industry and automobile 15

Biomass burning 16
Oceans 3
Nitrogen fixation 3
Photochemical Production 40
Total 77
Loss Tg/yr
Oxidation by OH 19
Soil uptake 56
Total 75
Mean mixing ratio (ppbv) 530

Burden (Tg H2 ) 150
Lifetime (yr) 2-3
in the troposphere and 30 Tg above the tropopause). The global cycling

time of hydrogen is of the order of 3±2 years.
Although the atmospheric hydrogen content at ground level was
first measured many years ago (Paneth, 1937; Glueckauf and Kitt,
1957), its detection in the stratosphere occurred much later (Ehhalt
and Heidt, 1973; Ehhalt et al., 1974; 1975a; Fabian et al., 1979). These
observations have mostly been performed by grab sampling, and they
have demonstrated that the H2 mixing ratio is about 0.56 ppmv at the
tropopause and in the lower and middle stratosphere. A small decrease
in the observed mixing ratio with increasing altitude is indicated by
rocket measurements in the upper stratosphere, as shown in Figure 5.25
(Scholz et al., 1970; Ehhalt et al., 1975b). This decrease in mixing
ratio is less pronounced at lower latitudes where vertical motion is
more intense. Abbas et al. (1996a,b) have inferred the vertical profile
of H2 in the upper stratosphere and lower mesosphere based on H2 O
and CH4 mixing ratios observed by the Atmospheric Trace Molecule
Spectroscopy (ATMOS) experiment from four Space Shuttle missions.
They find a minimum mixing ratio of typically 0.15 ± 0.1 ppmv at 50 km
altitude. Model calculations by Le Texier et al. (1988) also suggest the
existence of a broad minimum in the mid- and upper stratosphere (where
Figure 5.25. Observations of the mixing of molecular hydrogen in the

stratosphere, compared to a one-dimensional model. From Ehhalt and
Tonnissen (1980).
H2 oxidation is most rapid) and an increasing mixing ratio at higher

altitudes due to the production of H2 from water vapor photolysis.
5.4.2 Odd Hydrogen Chemistry
The behavior of the hydrogen radicals produced by dissociation of

H2 O, H2 and CH4 varies strongly with altitude. In the mesosphere and
stratosphere, the lifetime of atomic hydrogen is very short because the
three body reaction with oxygen rapidly converts H into a hydroperoxy
radical, HO2
(a1 ); H + O2 + M → HO2 + M (5.86)
This reaction, however, becomes slow at higher altitudes due to the low
air densities found there.
Atomic hydrogen can also react with ozone, and this reaction becomes
important in the upper stratosphere and mesosphere:
(a2 ); H + O3 → O2 + OH†(v ≤ 9) (5.87)
Note that OH formed by this reaction is vibrationally excited (with
a vibrational quantum number v less than or equal to 9) and is de-
activated either by quenching with air molecules or by photon emission
(Meinel bands observed in the airglow; see Chapter 2). At altitudes

above 40 km, the reactions of the OH and HO2 radicals with atomic
oxygen are important:
(a5 ); OH + O → O2 + H (5.88)
and
(a7 ); HO2 + O → O2 + OH (5.89)
On the other hand, in the middle and lower stratosphere and in the
troposphere, the reactions of the hydrogen radicals with ozone must be
considered:
(a6 ); OH + O3 → O2 + HO2 (5.90)
(a6b ); HO2 + O3 → 2O2 + OH (5.91)

These four reactions provide important loss processes for odd oxygen.
Near the tropopause, the OH and HO2 radicals also react with carbon
and nitrogen monoxide:
(a26 ); HO2 + NO → NO2 + OH (5.92)
(c20 ≡ a36 ); OH + CO → CO2 + H (5.79)

In the mesosphere, atomic hydrogen becomes quite abundant and the
following reactions (note that molecular hydrogen can be produced)
become important:
(a23a ); H + HO2 → OH + OH (90 ± 4%) (5.93a)
(a23b ); H + HO2 → H2 + O2 (8 ± 4%) (5.93b)
(a23c ); H + HO2 → H2 O + O (2 ± 2%) (5.93c)
The branching ratios reported here are from the laboratory study of
Keyser (1986).
Hydrogen peroxide also plays an important role in oxygen-hydrogen
chemistry. It is produced by the HO2 “self-reaction”:
(a27 ); HO2 + HO2 → H2 O2 + O2 (5.94)
which has been shown to have a pressure-independent bimolecular
component and a pressure-dependent termolecular component. In the
presence of water vapor (e.g., in the troposphere), the rate constant of
this reaction varies with the abundance of H2 O (Lii et al., 1981; Kircher
and Sander, 1984). H2 O2 is also formed by the three-body reaction
(a83 ); OH + OH + M → H2 O2 + M. (5.95)
It is destroyed during the day through photolysis:
(JH2 O2 ); H2 O2 + hν → 2OH (5.96)
and to a lesser extent, by reaction with OH:

(a30 ); H2 O2 + OH → H2 O + HO2 (5.97)
and with atomic oxygen
(a81 ); H2 O2 + O → OH + HO2 (5.98)
It should be noted that OH is also destroyed by reaction with HO2 :
(a17 ); OH + HO2 → H2 O + O2 (5.99)
which is faster than the reaction
(a16 ); OH + OH → H2 O + O (5.100)
In the lower stratosphere, the presence of large amounts of HNO3 ,
HO2 NO2 , H2 O2 , and HCl lead to an additional destruction of hydroxyl
radical (and odd hydrogen):
(b27 ); OH + HNO3 → H2 O + NO3 (5.101)
(b28 ); OH + HO2 NO2 → H2 O + O2 + NO2 (5.102)
(a30 ); OH + H2 O2 → H2 O + HO2 (5.103)
(d11 ); OH + HCl → H2 O + Cl (5.104)

The effect of these reactions is important below about 25 km.
Figures 5.26a and 5.26b present schematic representations of the most
important reactions in hydrogen chemistry for the lower and upper
portions of the middle atmosphere, respectively. Retaining the most
important of these reactions, we may write the continuity equations for
the long-lived oxygen-hydrogen species:
∂(H2 O)
+ ∇ · φ(H2 O) + JH2 O + a1∗ (O1 D) (H2 O) (5.105)
∂t
= a16 (OH)2 + a17 (OH)(HO2 ) + a19 (OH)(H2 ) + a23c (H)(HO2 )
∂(H2 )
+ ∇ · φ(H2 ) + a3∗ (O1 D) + a19 (OH) + a24 (O) (H2 )
∂t
= a23b (H)(HO2 ) + JbH2 O (H2 O) (5.106)
where JH2 O = JaH2 O + JbH2 O . Neglecting the formaldehyde and methyl
radical reactions discussed above, as well as Reactions (5.101), (5.102),
and (5.104), the equations of H, OH, HO2 , and H2 O2 can be written:
d(H)
+ [a1 (M)(O2 ) + a2 (O3 ) + a23 (HO2 )] (H)
dt
= a5 (O)(OH) + JaH2 O (H2 O) + a∗3 (O1 D)(H2 ) (5.107)
+ a19 (OH)(H2 ) + a24 (O)(H2 ) + a36 (CO)(OH)

Figure 5.26a. Principal chemical reactions affecting hydrogen compounds in the

stratosphere. The numbers in the boxes are concentrations (cm−3 ) calculated
at 25 km altitude (24 h average). The numbers associated with arrows represent
fluxes (cm−3 s−1 ) between different compounds. From Zellner (1999).
d(OH)
+ [a5 (O) + a6 (O3 ) + 2a16 (OH) + a17 (HO2 ) + a19 (H2 )
dt
+ a30 (H2 O2 ) +2a83 (M)(OH) + a36 (CO)] (OH)
= JaH2 O (H2 O) + 2a∗1 (O1 D)(H2 O) + a2 (O3 )(H)
+ a6b (O3 )(HO2 ) + a7 (HO2 )(O) + 2a23a (H)(HO2 )
+ a26 (NO)(HO2 ) + 2JH2 O2 (H2 O2 ) + a81 (O)(H2 O2 )
+ a24 (O)(H2 ) + a24 (O)(H2 ) + a∗3 (O1 D)(H2 )
+ a∗2 (O1 D)(CH4 ) (5.108)
d(HO2 )
+ [a6b (O3 ) + a7 (O) + a23 (H) + a26 (NO) + a17 (OH)
dt
+2a27 (HO2 )] (HO2 ) = a1 (M)(O2 )(H) + a6 (OH)(O3 )
+ a30 (OH)(H2 O2 ) + a81 (O)(H2 O2 ) (5.109)
d(H2 O2 )
+ [JH2 O2 + a30 (OH) + a81 (O)] (H2 O2 )
dt
= a27 (HO2 )2 + a83 (M)(OH)2 (5.110)
Figure 5.26b. Principal aeronomic reactions of hydrogen compounds in the

mesosphere and lower thermosphere.
5.4.3 The Odd Hydrogen Family and Hydrogen

Partitioning
Adding together equations (5.107) through (5.109) and introducing a

transport term, we obtain an approximate equation for an odd hydrogen
family defined by HOx = H + OH + HO2 (neglecting methyl radical
and nitrogen reactions):
∂(HOx )
+ ∇ · φ(HOx ) + 2a23b+23c (H)(HO2 ) + 2a16 (OH)2
∂t
+ 2a17 (OH)(HO2 ) + 2a27 (HO2 )2 + 2a83 (M)(OH)2
= 2JH2 O (H2 O) + 2a3∗ (O1 D)(H2 ) + 2a1∗ (O1 D)(H2 O)
+ 2 [JH2 O2 + a81 (O)] (H2 O2 ) + a2∗ (O1 D)(CH4 ) (5.111)
Figure 5.27 presents calculated altitude profiles of the photochemical
lifetimes of the HOx family and its constituent members, as well as H2 O2 ,
and the time constants for dynamical processes. As in the case of Ox , the
lifetime of the family is several orders of magnitude greater than those of
Figure 5.27. Photochemical lifetime of odd hydrogen radicals, as well as the

time constants for transport by the zonal and meridional winds, and a one-
dimensional diffusive time constant.
some of its members, making the solution of the continuity equation of

the HOx family relatively easy to obtain with simple numerical methods.
Even so, the lifetime of the HOx family is considerably shorter than the
transport time scale up to altitudes near 80 km. Therefore, its density
will not be directly dependent on transport processes except above this
altitude. Throughout the middle atmosphere, the lifetimes of OH and
HO2 are also quite short. Assuming photochemical equilibrium for these
species and keeping only the dominant terms in (5.108) and (5.109), one
may write:
[a5 (O) + a6 (O3 ) + a36 (CO)] (OH)
≈ a2 (O3 )(H) + a7 (HO2 )(O) + a26 (NO)(HO2 ) (5.112)
and
[a6b (O3 ) + a7 (O) + a26 (NO)] (HO2 ) ≈ a1 (M)(O2 )(H) + a6 (OH)(O3 )
(5.113)
The ratio of the concentrations of HO2 and OH can then be written:
(HO2 ) a5 (O) + a36 (CO)
≈
(OH) a7 (O) + a26 (NO) + a6b (O3 )
(5.114)
a1 (M)(O2 ) a6 (O3 )
× +
a1 (M)(O2 ) + a2 (O3 ) a5 (O) + a36 (CO)
This last expression can be further simplified; in the upper part of the
middle atmosphere (z > 40 km), one may assume:
(HO2 ) a5 a1 (M)(O2 )
≈ × (5.115)
(OH) a7 a1 (M)(O2 ) + a2 (O3 )
In the middle stratosphere, we obtain
(HO2 ) a6
≈ (5.116)
(OH) a6b
and near the tropopause, it becomes
(HO2 ) a36 (CO) + a6 (O3 )
≈ (5.117)
(OH) a26 (NO) + a6b (O3 )
In the upper stratosphere and in the mesosphere,
a5 (O)(OH)
(H) ≈ (5.118)
a1 (M)(O2 ) + a2 (O3 )
so that
(H) a5 (O)
≈ (5.119)
(OH) a1 (M)(O2 ) + a2 (O3 )
These ratios establish the approximate partitioning between members of
the HOx family, as was shown previously for Ox :
(OH) 1
= (H)
(5.120)
(HOx ) 1+(OH) + (HO2 )
(OH)
In the upper stratosphere and in the mesosphere, H2 O2 is approximately
in photochemical equilibrium defined by
a27 (HO2 )2
(H2 O2 ) = (5.121)
JH2 O2 + a30 (OH)
In the lower stratosphere the H2 O2 lifetime becomes longer, and the
time dependent equation (5.110), must be used with a transport term
∇ · φ (H2 O2 ) added. Heterogeneous removal in clouds must also be
considered in the troposphere. Figure 5.28 shows calculated vertical
distributions of H, OH, HO2 , and H2 O2 between 10 and 70 km altitude.
In the lower thermosphere, atomic hydrogen becomes the dominant
odd hydrogen species. Adding expressions (5.107) through (5.110),
assuming d(OH)/dt = d(HO2 )/dt = d(H2 O2 )/dt = 0, omitting minor
terms and adding a transport term, we find:
∂(H)
+ ∇ · φ (H) + 2(a23b + a23c )(HO2 )(H) = 2JH2 O (H2 O) (5.122)
∂t
Below 75 km, the concentration of H becomes small relative to OH and
HO2 , and the addition of equations (5.107), (5.108), and (5.109) leads
to the following expression, in which secondary terms are neglected but
Figure 5.28. Calculated distributions of some hydrogen compounds at mid-

latitudes (noon, equinox).
reactions with HNO3 , HO2 , NO2 , and carbon compounds are added:
2a17 (OH)(HO2 ) + 2a30 (OH)(H2 O2 ) + 2b27 (OH)(HNO3 )
+ 2b28 (OH)(HO2 NO2 ) + 2c17 (CH3 OOH)(OH)
= 2JH2 O (H2 O) + 2a1∗ (O 1 D)(H2 O)
+ (O 1 D)(CH4 ) [2c1a (1 + X) + c1b X]
+ 2c2 X(OH)(CH4 ) + 2a3∗ (O 1 D)(H2 )
(5.123)
where X, given by
JH−HCO + c9 (O)
X
JH−HCO + JH2 −CO + c8 (OH) + c9 (O)
1
is approximately equal to 2. This equation can be simplified in the
mesosphere:
a17 (OH)(HO2 ) + (a23b + a23c) (HO2 )(H)
(5.124)
= JH2 O + a1∗ (O 1 D) (H2 O) + a3∗ (O 1 D)(H2 )
and in the stratosphere above 30 km:
a17 (OH)(HO2 ) = a1∗ (O 1 D)(H2 O) + a3∗ (O 1 D)(H2 )
(5.125)
+ (1 + X)c1a (O 1 D)(CH4 ) + c2 X(OH)(CH4 )
while below 30 km,

a17 (OH)(HO2 ) + a30 (OH)(H2 O2 ) + b27 (OH)(HNO3 )
+ b28 (OH)(HO2 NO2 ) = (O 1 D) [a1∗ (H2 O)
+a3∗ (H2 ) + (1 + X)c1a
∗
(CH4 )] + c2 X(OH)(CH4 )
(5.126)
5.4.4 Observations of OH and HO2
OH and HO2 radicals have been measured in the mesosphere

and in the stratosphere by different techniques, including resonance
fluorescence, induced fluorescence, as well as remote sensing methods
(far-infrared and submillimeterwave) (see, e.g., Anderson, 1971, 1976;
Heaps and McGee, 1983; De Zafra et al., 1984; Stimpfle et al., 1990;
Wennberg et al., 1990, 1994; Cohen et al., 1994; Park and Carli, 1991;
Stachnik et al., 1992; Pickett and Peterson, 1993; 1996; Chance et al.,
1996; Jucks et al., 1998). Total column measurements of OH also
provide important information about its stratospheric abundance and
variability (Burnett and Burnett, 1981; 1982). Figure 5.29 shows the
concentration profiles of OH and HO2 measured by Jucks et al. (1998)
during mid-morning at 65◦ N on April 30, 1997, using a balloon-borne
Fourier infra-red spectrometer. In the case of OH, the number density
Figure 5.29. Measured concentration of OH and HO2 on 30 April 1997 at 69◦ N,

149◦ W 9:15 local solar time using a balloon-borne thermal emission far-infrared
transform spectrometer. From Jucks et al. (1998).
increases from 7 × 106 cm−3 at 30 km altitude to a maximum value

of approximately 20 × 106 cm−3 at 46 km. HO2 densities reach a
maximum of about 18 × 106 cm−3 at 39 km. At the stratopause, the
measured concentrations are 17 × 106 cm−3 and 10 × 106 cm−3 for
OH and HO2 , respectively. The observed (HO2 )/(OH) concentration
ratio of 1.7 can be compared with the theoretical values provided by
expression (5.115). In the mesosphere, the observations by the Middle
Atmosphere High Resolution Spectrograph Investigation (MAHRSI) on
board the Space Shuttle (Figure 5.30; Summers et al., 1997) provide
daytime concentrations of 11-15 × 106 cm−3 at 50 km, (3-8) × 106
Figure 5.30. Observations of the vertical distribution of the hydroxyl density

(cm−3 ) in the mesosphere (adapted from Summers et al. 1997). From top to
bottom, the average solar zenith angles for each panel were 62◦ , 44◦ , and 53◦ .
Figure 5.31. Distribution of hydroxl density (cm−3 ) in the mesosphere observed

for different altitudes, latitudes, and local times by the MAHRSI experiment
from the CRISTA-SPAS satellite (R.R. Conway, private communication, 1999).
cm−3 between 55 and 70 km. Figure 5.31 shows measurements of the

OH density by the MAHRSI instrument for different latitudes, altitudes
and local times (or solar zenith angles). The stratospheric abundance
of hydrogen peroxide, reported by Waters et al. (1981), Chance and
Traub (1987), Chance et al. (1991; 1996), and by Park and Carli (1991),
is greater than those of OH and HO2 . Mesospheric atomic hydrogen
has also been inferred from observations of the Lyman α and Lyman β
emissions (Anderson et al., 1980).
5.5. Nitrogen Compounds
The presence in the middle atmosphere of nitrogen oxides (NO and

NO2 generally called, collectively, NOx ) and related nitrogen compounds
such as nitric acid (HNO3 ) results from the oxidation of nitrous oxide
(N2 O) and, to a lesser extent, from the ionization of molecular nitrogen
(N2 ) by solar and galactic high energy particles. Nitrogen oxides can also
be injected in the lower stratosphere by aircraft engines. Nitric oxide
(NO) and nitrogen dioxide (NO2 ) are in photochemical balance during

daytime due to fast conversion mechanisms: NO is converted to NO2
by reaction with ozone, and NO2 is transformed back to NO either by
photolysis or by reaction with atomic oxygen. As these cycles proceed,
large amounts of ozone can be catalytically destroyed. The rate for this
loss mechanism is, however, limited by the transformation of the nitrogen
oxides into less reactive nitrogen compounds. For example, NO2 reacts
with ozone to form nitrogen trioxide (NO3 ) which is most abundant
during nighttime (when photolytic destruction ceases) and reacts with
NO2 to form dinitrogen pentoxide (N2 O5 ). The conversion of nitrogen
oxides to a more stable reservoir, nitric acid (HNO3 ), occurs either
through the reaction between NO2 and OH or through heterogeneous
hydrolysis of N2 O5 on the surface of aerosol particles. The transport
of nitric acid to the troposphere and its removal by precipitation (wet
scavenging) removes nitrogen compounds from the middle atmosphere.
Other nitrogen reservoirs include pernitric acid (HNO4 or HO2 NO2 ) and
chlorine and bromine nitrate (ClONO2 and BrONO2 ).
In the mesosphere and thermosphere, where the abundance of ozone
becomes too low to convert substantial amounts of NO into NO2 and
where the concentration of atomic oxygen is high, reactive nitrogen
is mostly in the form of nitric oxide (NO) and atomic nitrogen (N).
Substantial production of these chemical species results from ionic
processes and N2 photolysis in the thermosphere. NO is transported
downward to the mesosphere where it is destroyed by photolysis and
recombination between N and NO to reform stable N2 . The chemical
coupling between nitrogen oxides in the thermosphere/upper mesosphere
and in the stratosphere is therefore weak, except in the polar night
regions where photochemical destruction does not occur and substantial
quantities of thermospheric nitrogen compounds are transported to the
stratosphere by the general circulation.
Details on the chemical and photochemical processes affecting
nitrogen compounds in the middle atmosphere are given in the following
sections: Sections 5.5.1-5.5.3 deal with stratospheric nitrogen, while
5.5.4 and 5.5.5 focus on nitrogen compounds in the mesosphere and
lower thermosphere.
5.5.1 Sources of Nitrogen Oxides
In the stratosphere, nitric oxide (NO) is produced mostly by

dissociation of N2 O by reaction with an excited oxygen atom in the (1 D)
state (Nicolet, 1971; Crutzen, 1971; McElroy and McConnell, 1971):
(b39 ); N2 O + O(1 D) → 2NO (5.127a)
The following products are also possible:
(b38 ); N2 O + O(1 D) → N2 + O2 (5.127b)
The rate of NO production by reaction (5.127a) is given by:
PNO = 2b39 (N2 O)(O 1 D) (5.128)
and is therefore dependent on the vertical distribution of N2 O. In the

stratosphere, N2 O is, in turn, destroyed primarily by photolysis:
(JN2 O ); N2 O + hν (λ ≤ 200 nm) → N2 + O(1 D) (5.129)
The production reaction
(b0 ) N2 + O(1 D) + M → N2 O + M (5.130)
is extremely slow, so that the in situ production of N2 O in the

stratosphere is insignificant.
The equation of continuity for nitrous oxide can thus be written:
∂(N2 O)
+ ∇ · φN2 O + (N2 O) JN2 O + (b38 + b39 )(O 1 D) = 0 (5.131)
∂t
Nitrous oxide is present throughout the troposphere (and is injected
into the lower stratosphere) with a mixing ratio of about 315 ppbv and a
global burden of 1510 Tg N. N2 O abundances are about 0.8 ppbv greater
in the northern hemisphere, where about 60% of the emissions are
located. This compound is produced principally by bacterial processes
associated with complex nitrification and denitrification mechanisms
in soils (Delwiche, 1978). Thus, as first indicated by Bates and
Witherspoon (1952), the biosphere plays a major role in determining
the source of N2 O and, hence, the abundance of nitrogen oxides in
the stratosphere. In spite of numerous studies, the global budget of
atmospheric N2 O remains insufficiently well quantified, because this
gas is emitted by several small sources which have large uncertainties.
Table 5.7 provides an estimate of the global N2 O sources and sinks.
Budget estimates by Mosier et al. (1998) and by Kroeze et al. (1999)
suggest that the total source of N2 O is in the range 6.7-36.6 Tg N/yr
with the most likely value close to 17.7 Tg N/yr (9.6 for the natural
contribution and 8.1 for the anthropogenic contribution). The global
lifetime of nitrous oxide is estimated to be approximately 120 years.
Table 5.7 Estimated Sources and Sinks of Nitrous Oxide (Tg/yr)

(From WMO, 1995)
Sources
A. Natural
Oceans 1.4-5.2
Tropical Soils
Wet Forests 2.2-3.7
Dry Savannas 0.5-2.0
Temperate Soils
Forests 0.05-2.0
Grassland ??
B. Anthropogenic
Cultivated Soils 1-3
Animal Waste 0.2-0.5
Biomass Burning 0.2-1.0
Stationary Combustion 0.1-0.3
Mobile Sources 0.1-0.6
Adific Acid Production 0.4-0.6
Nitric Acid Production 0.1-0.3
Sinks
Removal by Soils ?
Photolysis in the Stratosphere 9-17
Atmospheric Increase 3.1-4.7
Figure 5.32 shows the photochemical time constant of N2 O and those

appropriate to transport processes. Like CH4 , N2 O is an excellent
tracer for transport in the middle stratosphere, where its lifetime is
comparable to those for advection by the mean meridional circulation.
At higher altitudes in the upper stratosphere and lower mesosphere, the
N2 O lifetime remains close to the mean meridional transport lifetimes,
making it a more sensitive tracer than CH4 in this region.
The distribution of N2 O has been measured by cryogenic sampling
and gas phase chromatography (Scholz et al., 1970; Ehhalt et al., 1975b;
Schmeltekopf et al., 1977) as well as by infrared absorption (Harries,
1973; Farmer, 1974; Farmer et al., 1974, 1980). Figure 5.33a shows
the early observations by Schmeltekopf et al. at different latitudes. In
each case, a decrease in mixing ratio is observed above the tropopause.
As was shown previously for CH4 , this decrease is less rapid at the
Figure 5.32. Photochemical lifetime of N2 O, as well as the time constants for

transport by the zonal and meridional winds, and a one-dimensional diffusive
time constant.
equator than at higher latitudes, presumably because of the rapid rate

of upward transport associated with the rising branch of the Hadley cell
at these latitudes. Space observations were also performed in the late
1970s by the SAMS instrument aboard the Nimbus 7 satellite, and in the
1990s by the CLAES and ISAMS sounders aboard UARS (Figure 5.33b)
and by the Cryogenic Infrared Spectrometers and Telescopes for
the Atmosphere (CRISTA) experiment on board the Space Shuttle
(Offermann et al., 1999 and Plate 3). Like CH4 (Figure 5.16), large
values are obtained in the tropical lower stratosphere. Pre-industrial
mole fractions of N2 O were approximately 270 ppbv (Flückinger
¨ et al.,
1999). The concentration increase during the industrial period has
produced a current radiative forcing of 0.15 Wm−2 (IPCC, 2001). An
analysis of the infrared spectra recorded in the Swiss Alps in 1950-1951
and from 1984-1992 suggests that the column abundance of N2 O has
increased by 0.23±0.04%/yr from 1951 to 1984 and by 0.36±0.06%/yr
between 1984 and 1992 (Zander et al., 1994). Analyses of data collected
during oceanic expeditions (Weiss, 1994) and from a global network of
stations (Swanson et al., 1993) indicate that, on the average, the growth
rate in the surface mixing ratio of N2 O was 0.5-0.6 ppbv/yr during 1976-
1982, increased to a maximum of 0.8-1.0 ppbv/yr in 1998-1989, and
declined to 0.5-0.6 ppbv in the early 1990s.
Figure 5.34 shows the distribution of the nitric oxide production by
oxidation of N2 O. This production is directly related to atmospheric
dynamics that carries N2 O to its loss region; it reaches a maximum
in the mid-stratosphere where its value is about 100 molec cm−3 s−1 .
Crutzen and Schmailzl (1983) suggest a global value between 1.1 and
1.9 ×108 molec NO cm−2 s−1 , using the N2 O distribution measured by
the SAMS satellite.
Figure 5.33a. Observations of the mixing ratio of nitrous oxide by Schmeltekopf

et al. (1977).
Figure 5.33b. Zonally averaged distribution of the N2 O mixing ratio (in pptv)
measured by the CLAES instrument on UARS during 1-20 September, 1992.
Courtesy of W. Randel, NCAR.
Figure 5.34. Diurnally and annually averaged rate of NO production in the

stratosphere through reaction of N2 O with O1 D. From Crutzen and Schmailzl
(1983).
Figure 5.35. Distribution of the rate of production of NO by cosmic rays at

solar minimum. Adapted from Nicolet (1975b).
Stratospheric nitrogen atoms can be formed by dissociative ionization

and dissociation of molecular nitrogen by galactic cosmic rays. These
processes generally occur through secondary electrons ejected by heavy
cosmic particles. One ion pair formed by cosmic radiation leads to the
production of 1-1.6 atoms of nitrogen (see Chapter 7). Figure 5.35
shows the spatial distribution of PN for solar minimum (from Nicolet,
1975b) as a function of geomagnetic latitude. In the stratosphere,
atomic nitrogen reacts rapidly with molecular oxygen to form nitric
oxide. Stratospheric production of nitrogen oxides by cosmic radiation
should not be neglected, especially in polar regions where the oxidation
of N2 O is slow. Nicolet (1975b) indicated that cosmic radiation leads
to an integrated NO production of about (5 ± 1) × 107 cm−2 s−1 in polar
regions and 3 × 107 cm−2 s−1 in the tropics.
Crutzen et al. (1975) noted that penetration of large amounts of
protons into the middle atmosphere during solar proton events would
also lead to an intense production of atomic nitrogen. These episodic
events can produce enough nitrogen oxide to influence the ozone content,
particularly at high latitudes (see e.g., Jackman, 1991; Jackman et al.,

1995). The chemistry of the ionization and odd nitrogen production
processes is described in more detail in Chapter 7.
Several other gases released at the Earth’s surface carry additional
nitrogen into the middle atmosphere. This is the case of hydrogen
cyanide (HCN) and methyl cyanide (CH3 CN). In both cases, only limited
observations are available and the global budget is not well quantified.
The numbers quoted below represent, therefore, orders of magnitude
only.
Hydrogen cyanide is believed to be produced primarily as a result of
biomass burning (Lobert et al., 1990) with a global release corresponding
to 0.5-3.0 Tg N/yr (see also Li et al., 2000). Emissions by fossil fuel
consumption appears to be very small (0.04 Tg N/yr) in comparison.
The major atmospheric loss of HCN is probably due to ocean uptake
(1-2.5 Tg N/yr; see Li et al., 2000). Reactions with OH (0.3 Tg N/yr)
and O(1 D) (3 × 10−4 Tg N/yr), as well as photolysis (2 × 10−3 Tg N/yr)
are small on the global scale (Li et al., 2000), but should be the major
destruction mechanisms in the middle atmosphere. The presence of HCN
in the atmosphere has been detected by infra-red spectroscopy (Coffey
et al., 1981a; Mahieu et al., 1995; 1997; Rinsland et al., 1999; Zhao et al.,
1999). The observed seasonal variation (factor 2-3 at northern mid-
latitudes and a factor 5 enhancement in the tropical upper troposphere
during the dry season) suggests that the global lifetime of HCN could
be as short as 2-4 months (Li et al., 2000). The mixing ratio of HCN
appears to be relatively constant with height (100-200 pptv) between
the surface and the upper stratosphere with a total atmospheric burden
of approximately 0.5 Tg N/yr.
Methyl cyanide (also called acetonitrile) is also produced primarily by
biomass burning (0.3-0.4 Tg N/yr; see Lobert et al., 1990; Holzinger
et al., 1999). Emissions from automobile exhausts and industrial
activities are believed to be small (0.01 Tg N/yr). The ocean is probably
a significant sink for atmospheric CH3 CN, but the magnitude of this
loss is not yet well quantified. The largest chemical loss is provided by
reactions with OH, O(1 D) and Cl, and by photolysis (Arijs and Brasseur,
1985; Hamm and Warneck, 1990). Rainout is another loss process in
the troposphere. The high variability in the measured tropospheric
concentrations of methyl cyanide suggests that the atmospheric lifetime
of this molecule is significantly shorter than predicted when only the
photochemical loss processes are taken into account. Since methyl
cyanide has a large proton affinity, it plays a major role in the formation
of positive ion clusters (called non-proton hydrates) in the stratosphere
(see Chapter 7). CH3 CN has been detected in the stratosphere by the
MLS instrument on board the Upper Atmosphere Research Satellite
(Livesey et al., 2001). Typical mole fractions of 40-60 ppbv are reported
in the tropics between 20 and 30 km altitude. Values are decreasing at
higher altitudes (about 10 pptv at 45 km near the equator) and higher
latitudes (20 pptv in summer and 10 pptv in winter at 60◦ latitude and
30 km altitude).
5.5.2 Odd Nitrogen Chemistry in the Stratosphere
In the stratosphere nitric oxide is converted to nitrogen dioxide by

reaction with ozone:
(b4 ); NO + O3 → NO2 + O2 (5.132)
by reaction with peroxy radicals (RO2 = HO2 or CH3 O2 ):
(a26 ); NO + HO2 → NO2 + OH (5.92)
(c5 ); NO + CH3 O2 → NO2 + CH3 O (5.53)
and by reactions with ClO and BrO (see Sections 5.6 and 5.7):
(d4 ); NO + ClO → NO2 + Cl (5.133)
(e4 ); NO + BrO → NO2 + Br (5.134)
The conversion mechanism by the three-body reaction
(b82 ); NO + O + M → NO2 + M (5.135)
can generally be neglected.
The NO2 molecule reacts rapidly during the day to reform NO, either
by reaction with atomic oxygen:
(b3 ); NO2 + O → NO + O2 (5.136)
or by photolysis:
(JNO2 ); NO2 + hν(λ < 405 nm) → NO + O(3 P) (5.137)
Reaction of NO with O3 (5.132), followed by reaction of NO2 with O
represents a catalytic cycle which provides the major loss process for
odd oxygen in the stratosphere.
In the upper stratosphere and above, nitric oxide is photolyzed in the
δ-bands (see Chapter 4)

(JNO ); NO + hν → N 4
S +O (5.138)
but ground-state atomic oxygen reacts rapidly with O2 to reform NO

(b7 ); N 4
S + O2 → NO + O . (5.139)
However, the reaction of N(4 S) and NO

(b6 ); N 4
S + NO → N2 + O (5.140)
represents a loss of odd nitrogen, which plays a major role in the

mesosphere during the daytime.
Additional reactions should also be considered in the stratosphere.
These lead to the formation of other nitrogen containing species
such as NO3 , N2 O5 , HNO3 , and HO2 NO2 , which represent nitrogen
“reservoirs”.
The formation of nitrogen trioxide (NO3 ) results from the reaction of
NO2 with ozone
(b9 ); NO2 + O3 → NO3 + O2 (5.141)
and leads to high NO3 concentrations during nighttime. Other minor
sources of NO3 are provided by Reactions (5.101), (5.156b), (5.159a),
and (5.160a), and by reaction
(b81 ); NO2 + O + M → NO3 + M (5.142)
During the day the NO3 radical is rapidly photolyzed
(JaNO3 ); NO3 + hν (λ ≤ 620 nm) → NO2 + O (5.143a)
(JbNO3 ); → NO + O2 (5.143b)
According to the laboratory measurements of Orlando et al. (1993),
the second channel of the NO3 photolysis (Reaction 5.143b) contributes
less than 10% of the total photodissociation. Additional destruction
mechanisms for NO3 include
(b71 ); NO3 + O → NO2 + O2 (5.144)
(b72 ); NO3 + OH → NO2 + HO2 (5.145)
(b73a ); NO3 + HO2 → NO2 + OH + O2 (5.146a)
(b73b ); → HNO3 + O2 (5.146b)

(b84 ); NO3 + NO → 2NO2 (5.147)
Dinitrogen pentoxide (N2 O5 ), formed by the reaction of NO2 and
NO3 , is produced almost entirely at night, when NO3 is abundant:
(b12 ); NO3 + NO2 + M → N2 O5 + M (5.148)
It is destroyed during the day through photolysis
(JaN2 O5 ); N2 O5 + hν (λ ≤ 380 nm) → NO3 + NO2 (5.149a)
(JbN2 O5 ); → NO3 + NO + O (5.149b)
and by collisional decomposition (the rate of which is highly temperature
dependent)
(b32 ); N2 O5 + M → NO3 + NO2 + M (5.150)
The quantum yield for process (5.149b) decreases with wavelength, and
is reported to be 0.72, 0.38, 0.21, and 0.15 at 248, 266, 287, and 289 nm,
respectively (Ravishankara et al., 1986).
Another important stratospheric nitrogen containing species is HNO3 .
This molecule, which introduces an interaction between nitrogen and
hydrogen compounds, is formed by the three body process:
(b22 ); NO2 + OH + M → HNO3 + M (5.151)
A heterogeneous reaction converting N2 O5 into HNO3
(h1 ) ; N2 O5 (g) + H2 O(, s) → 2HNO3 (5.152)
occurs on the surface of ice particles [H2 O (solid)], water droplets
[H2 O (liquid)], nitric acid ice [HNO3 · 3H2 O (solid)], sulfate aerosol
[H2 SO4 /H2 O] and ternary solution [H2 SO4 /HNO3 /H2 O] particles in
the lower stratosphere. In the case of sulfate aerosols, the reaction
probability γ is close to 0.1 (see e.g., Mozurkiewicz and Calvert, 1988;
Hanson and Ravishankara, 1992; Fried et al., 1994; Lovejoy and Hanson,
1995; Hanson, 1997) with only a modest dependence on the H2 SO4
content and on temperature. For a water ice particle, however, the
reaction probability is much less (of the order of 0.01, see Hanson and
Ravishankara, 1993). If the ice particle becomes coated with HNO3 ,
the value of γ drops even further (approximately 0.001). This reaction
provides an efficient mechanism to convert nitrogen oxides into nitric
acid, especially at high latitudes where the gas-phase conversion of NO2
to HNO3 by the OH radical is slow, particularly during wintertime.
When particles are present in the stratosphere, Reaction (5.152) is rapid,
and the rate of heterogeneous conversion of nitrogen oxides (NOx ) to
nitric acid (HNO3 ) is limited by the rate at which the gas-phase reaction
of NO2 with ozone (Reaction 5.141) proceeds. Note also that, as the
concentration of NO2 is reduced by Reaction (5.152), the presence of
sulfate aerosol particles in the stratosphere, and occasionally of polar
stratospheric clouds (PSCs), tends to enhance the levels of ClO and
OH in the lower stratosphere. The concentration of ClO is increased
mainly because the rate at which this radical recombines with NO2
(Reaction 5.157) is reduced. The level of OH is enhanced mainly because
its gas-phase reaction with NO2 to form HNO3 (Reaction 5.151) becomes
slower.
The two primary chemical loss processes for HNO3 are:
(JHNO3 ); HNO3 + hν (λ ≤ 310 nm) → OH + NO2 (5.153)
(b27 ); HNO3 + OH → NO3 + H2 O (5.101)

HNO3 molecules are eventually transported from the stratosphere to
the troposphere where they are readily incorporated into rain droplets.
Gravitational sedimentation of HNO3 -containing solid or liquid particles

can also lead to denitrification of the lower polar stratosphere during
winter.
The HO2 NO2 molecule is formed by the reaction:
(b23a ); NO2 + HO2 + M → HO2 NO2 + M (5.154)
and destroyed by unimolecular decomposition at low altitudes:
(b23b ); HO2 NO2 + M → HO2 + NO2 + M (5.155)
as well as by photolysis
(JaHO2 NO2 ); HO2 NO2 + hν (λ < 325 nm) → NO2 + HO2
(5.156a)
(JbHO2 NO2 ); → NO3 + OH
(5.156b)
and reaction with OH:
(b28 ); HO2 NO2 + OH → H2 O + O2 + NO2 (5.102)
At 248 nm, photolysis (Reaction 5.156) yields about one-third OH and
NO3 and two-thirds HO2 + NO2 (McLeod et al., 1988). The chemistry
of HO2 NO2 is important in both nitrogen and hydrogen chemistries.
Other nitrogen-containing reservoirs ClONO2 and BrONO2 are
formed by reactions with chlorine and bromine monoxide, respectively
(d31 ); ClO + NO2 + M → ClONO2 + M (5.157)
(e13 ); BrO + NO2 + M → BrONO2 + M (5.158)
Chlorine and bromine nitrate are photolyzed with different possible
channels
(JaClONO2 ); ClONO2 + hν (λ ≤ 320 nm) → Cl + NO3
(5.159a)
(JbClONO2 ); → ClO + NO2
(5.159b)
(JaBrONO2 ); BrONO2 + hν (λ ≤ 450 nm) → Br + NO3
(5.160a)
(JbBrONO2 ); → BrO + NO2
(5.160b)
Reactions involving halogen atoms will be further discussed in Sec-
tions 5.6 and 5.7. Note that these reactions introduce a strong chemical
coupling between nitrogen and halogen species in the stratosphere.
The primary reactions of these species are depicted in Figure 5.36.
Considering the most important chemical and photochemical processes,
Figure 5.36. Principal chemical reactions affecting nitrogen species in the

stratosphere. Concentrations ((cm−3 ) and reaction fluxes (cm
( −3 s−1 ) between
different compounds calculated at 25 km altitude (24-hour average) are
indicated. After Zellner (1999).
the continuity equations for NO, NO2 , NO3 , N2 O5 , HNO3 , and HO2 NO2
can be written:
d (N)
+ [b6 (NO) + b7 (O2 )] (N) = JNO (NO) + PN (5.161)
dt
d (NO)
+ [b4 (O3 ) + b6 (N) + b84 (NO3 ) + a26 (HO2 )
dt
+c5 (CH3 O2 ) + d4 (ClO) + e4 (BrO) + JNO ] (NO)
= [b3 (O) + JNO2 ] (NO2 ) + JbNO3 (NO3 )
+ JbN2 O5 (N2 O5 ) + b7 (O2 ) (N) + P (NO) (5.162)
d (NO2 )
+ [JNO2 + b3 (O) + b12 (M) (NO3 ) + b22 (M) (OH)
dt
+ b23a (M) (HO2 ) + b81 (M) (O) + d31 (M) (ClO)
+ e13 (M) (BrO)] (NO2 ) = [b4 (O3 ) + a26 (HO2 )
+ c5 (CH3 O2 ) + d4 (ClO) + e4 (ClO)] (NO)
%
+ JHNO3 (HNO3 ) + JaNO3 + b71 (O) + b72 (OH)
+b73a (HNO2 ) + 2b84 (NO)] (NO3 )

% &
+ JaN2 O5 + b32 (M) (N2 O5 )
% &
+ JaHO2 NO2 + b23b (M) + b28 (OH) (HO2 NO2 )
+ JbClONO2 (ClONO2 ) + JbBrONO2 (BrONO2 ) (5.163)
d(NO3 )
+ [b12 (M)(NO2 ) + JNO3 + b71 (O) + b72 (OH)
dt
+ (b73a + b73b ) (HO2 ) + b84 (NO)] (NO3 ) = b9 (O3 )(NO2 )
+ [JN2 O5 + b32 (M)] (N2 O5 ) + b27 (OH)(HNO3 )
+ b81 (M) (O) (NO2 ) + JaClONO2 (ClONO2 )
+ JaBrONO2 (BrONO2 ) (5.164)
d(N2 O5 )
+ (N2 O5 ) [JN2 O5 + b32 (M) + h1 ]
dt
= b12 (M)(NO2 )(NO3 ) (5.165)
d (HNO3 )
+ [JHNO3 + b27 (OH)] (HNO3 )
dt
= b22 (M) (OH) (NO2 ) + 2h1 (N2 O5 )
+ b73b (HO2 ) (NO3 ) (5.166)
d(HO2 NO2 )
+ (HO2 NO2 ) [JHO2 NO2 + b28 (OH) + b23b (M)]
dt
= b23a (NO2 )(HO2 )(M) (5.167)
where PN and PNO account for all ionic production of N and NO, JNO3 =
JaNO3 + JbNO3 , JHO2 NO2 = JaHO2 NO2 + JbHO2 NO2 , and JN2 O5 = JaN2 O5 +
JbN2 O5 .
5.5.3 The Odd Nitrogen Family in the Stratosphere and

Nitrogen Partitioning
The odd nitrogen family (NOy ) is defined as the sum of all nitrogen-
containing species except the major atmospheric constituent N2 and the
source gases such as nitrous oxide (N2 O), hydrogen cyanide (HCN) and
methyl cyanide (CH3 CN). Thus, in the middle atmosphere, if we include
halogenated nitrates,
NOy = N + NO + NO2 + NO3 + 2 N2 O5
+ HNO3 + HO2 NO2 + ClONO2 + BrONO2
it is easy to show by adding (5.161) to (5.167) and accounting for the
contributions of ClONO2 and BrONO2 , that
∂ (NOy )
+ ∇φNOy + 2b6 (N) (NO) + βNOy (NOy ) = PN + PNO (5.168)
∂t
where βNOy represents an effective loss coefficient accounting for wet
scavenging of HNO3 .
In many cases, the NOx family is formed as the sum of NO and
NO2 , and accounts for the most reactive nitrogen species. The NOx /
NOy concentration ratio, which is often reported from field observations,
is an indicator of the reactivity of odd nitrogen and its ability to
destroy stratospheric ozone (or to affect other chemical families including
chlorine and bromine compounds). The value of this ratio increases
with altitude above 30 km to reach a value of nearly one in the
upper stratosphere and mesosphere. It decreases substantially when the
stratospheric aerosol load is enhanced, for example, after large volcanic
eruptions (Fahey et al., 1993), and substantial amounts of nitrogen
oxides are converted to nitric acid by heterogeneous reaction (5.152).
It is also low in the polar regions, especially in air masses processed by
polar stratospheric clouds.
Note that NOz is sometimes defined as NOy – NOx (and refers
essentially to odd nitrogen reservoirs), and that some authors define NOx
as the sum of N + NO + NO2 + NO3 + 2 N2 O5 + HO2 NO2 (all NOy
compounds except HNO3 ), considering thereby all nitrogen compounds
which are rapidly converted to NO2 and NO. This latter definition is
adopted in Figure 5.37. In this case, the continuity equation for NOx
can be written
d (NOx )
+ b22 (M) (OH) (NO2 ) + 2b6 (N) (NO) + 2h1 (N2 O5 )
dt
= 2b39 (N2 O) O1 D + PN + JHNO3 (HNO3 ) (5.169)
+ b27 (OH) (HNO3 )

Figure 5.37 displays the altitude profiles of the photochemical lifetimes
of the NOy family and its constituent members (gas phase chemistry
only), as well as the time constants for transport at middle latitudes.
Note that the photochemical lifetimes for NOx and HNO3 are both
comparable to the transport lifetimes in the lower stratosphere, so that
the distribution of these species will be quite dependent on atmospheric
transport. It is important to treat them separately in numerical model
Figure 5.37. Photochemical time constants for NOy , N, NO, NO2 , NO3 , N2 O5 ,
HO2 NO2 , and HNO3 , as well as the time constants for transports by the zonal
and meridional winds and a one-dimensional eddy diffusion coefficient.
studies. In high latitude winter, it also becomes necessary to treat N2 O5

separately from the other nitrogen species; this will be discussed in more
detail below.
The partitioning of the NOy family is somewhat more complicated
than for the families presented previously. The ratio between N and NO
(below 80 km) can be derived from Eq. (5.161) in which b6 (NO) << b7
(O2 ), PN is neglected, and photochemical conditions (d(N)/dt = 0) are
assumed:
(N) JNO
= (5.170)
(NO) b7 (O2 )
NO2 is also in equilibrium with NO throughout the sunlit middle
atmosphere. Because atmospheric absorption in the 400 nm region is
negligibly small, JNO2 is given by its optically thin value of about 1×10−2
s−1 even at altitudes as low as 15 km, which corresponds to a lifetime of
about 2 minutes. NO2 can therefore be considered to be in immediate
photochemical equilibrium with NO during the day, and by conserving
the largest terms, we may deduce from (5.163) the following expression
for the ratio of NO to NO2 :

(NO) JNO2 + b3 (O)
= (5.171)
(NO2 ) b4 (O3 ) + a26 (HO2 ) + c5 (CH3 O2 ) + d4 (ClO)
The value of this ratio is close to one over most of the stratosphere
during daytime, but increases rapidly above 40 km because of increasing
atomic oxygen densities. At night, NO is quickly converted to NO2 up to
altitudes of about 60 km. Figure 5.38 shows calculated one-dimensional
model distributions of the various nitrogen species. The curves represent
a steady state solution for average solar insolation.
Setting NO3 in photochemical equilibrium in the sunlit stratosphere
yields:
(NO3 ) b9 (O3 )
≈ (5.172)
(NO2 ) JNO3
while at night
b9 (O3 )
(NO3 )night = (5.173)
b12 (M)
Calculated vertical distributions of nitrogen species at mid-

latitudes (noon, equinox).
For HO2 NO2 , we obtain

(HO2 NO2 ) b23a (HO2 )(M)
= (5.174)
(NO2 ) JHO2 NO2 + b26 (OH)
The major difficulty in partitioning the NOz family is that in the
middle atmosphere a simple instantaneous ratio cannot describe the
relationship between N2 O5 and the other family members. This is
because N2 O5 is produced almost entirely at night, when concentrations
of NO3 are large. About 30-50% of the atmospheric NOx is in the form
of N2 O5 at the end of a typical night in the lower stratosphere. This
accumulated N2 O5 is destroyed primarily during the day by photolysis,
which produces a diurnal variation in NO and NO2 . Therefore NO, NO2 ,
and N2 O5 are never really in an instantaneous stationary state; rather
they achieve a diurnally changing equilibrium over the 24-hour period.
A good approximation for the diurnal average steady state ratio can be
derived as follows (see also, Crutzen, 1971): First, write the appropriate
continuity equation for N2 O5 averaged over 24 hours:
d(N2 O5 )
≈ b12 (M)(NO2 )night (NO3 )night Fd − JN2 O5 (N2 O5 )Fs (5.175)
dt
where Fd and Fs represent the fractions of the day which are dark and
sunlit, respectively. We next assume that
(NO2 )night = (NO2 )day + (NO)day
(5.176)
JNO2 + b3 (O)
≈ 1+ (NO2 )day
b4 (O3 )
since, as mentioned above, all the nitric oxide in the stratosphere is
almost immediately converted to NO2 at sunset (and using 5.171).
Substituting for (NO3 )night and (NO2 )night with the expressions given
above, and assuming steady state for N2 O5 , we obtain

J +b (O)
(N2 O5 )avg b9 (O3 ) 1 + NOb24 (O33) Fd
= (5.177)
(NO2 )day JN2 O5 Fs
An alternate approach is simply to integrate N2 O5 separately from
the other NOy species. Indeed, this becomes necessary near the polar
night region. In the upper stratosphere, where the lifetime of nitric
acid becomes relatively short, the HNO3 /NO2 concentration ratio can
be expressed as follows, if the most important reactions are retained and
heterogeneous chemical processes are neglected:
(HNO3 ) b22 (M)(OH)
= (5.178)
(NO2 ) JHNO3 + b27 (OH)
In the lower stratosphere, nitric acid is strongly influenced by transport
and photochemical equilibrium conditions do not apply.
The NOy family can be partitioned using the ratios defined above:
(NO2 )
=
(NOy )
1
(N) (NO) (NO) (NO3 ) (HO2 NO2 ) 2×(N2 O5 ) (HNO3 )
1+ (NO) × (NO2 ) + (NO2 ) + (NO2 ) + (NO2 ) + (NO2 ) + (NO2 )
(5.179a)
and
(NO) (NO) (NO2 )
= × etc. (5.179b)
(NOy ) (NO2 ) (NOy )
Figure 5.39 presents a model calculation of the diurnal variation of
NO, NO2 , NO3 , N2 O5 , and HNO3 at 20 and 40 km.
5.5.4 Observations of Nitrogen Species in the Stratosphere
Stratospheric nitrogen species have been measured by balloon-borne

instruments since about 1970, and more recently from satellites. Most
measurements have been performed by chemiluminescence techniques
Figure 5.39. Diurnal variation of nitrogen species calculated for May 10 and
65◦ N at 20 km (left) and 40 km (right).
(in the case of NO, see Ridley et al., 1987; Fahey et al., 1989; Kondo
et al., 1992), mass spectroscopy (in the case of HNO3 , see Spreng and
Arnold, 1994), UV-visible absorption (in the case of NO2 , see Noxon,
1979; Pommereau and Piquard, 1994; Pfeilsticker and Platt, 1994) and
infrared absorption and emission (in the case of NO2 , N2 O5 , HNO3
and ClONO2 , see Oelhaf et al., 1994; Rinsland et al., 1996). Figure
5.40 shows the vertical profile of NO, NO2 , N2 O5 , HNO3 , and ClONO2
observed by a balloon-borne Fourier Transformed Infrared instrument
at 35◦ N in early fall (Sen et al., 1998).
In the case of NO, the mixing ratio increases by a factor of about 100
from 20 to 35 km, and is about 10 ppbv near 40 km. A large degree of
variability is also observed, associated with atmospheric dynamics. The
rapid increase of the NO mixing ratio after sunrise, as predicted from
model calculations (see Figure 5.39), is seen in the measurements made
by Ridley et al. (1977) (see Figure 5.41).
The observations of nitrogen dioxide reveal an increase in the mixing
ratio with increasing altitude up to about 35 km, while above that
altitude a progressive decrease is seen, reflecting the changing ratio of
NO2 /NO, e.g., Eq. (5.171). More recent satellite observations of NO2 by
LIMS, SAGE and UARS have provided further information, confirming
Figure 5.40. Volume mixing ratio (ppbv) of serveral nitrogen-containing species

(NO, NO2 , N2 O5 , HNO3 , ClONO2 ) measured at sunrise by a balloon-borne
infrared instrument at 35◦ N in September 1993. Adapted from Sen et al. (1998).
Figure 5.41. Time variation of the NO mixing ratio at about 26.5 km by Ridley
et al. (1977).
many of the features inferred from earlier studies. Figure 5.42 shows
a monthly averaged distribution of nighttime NO2 measured by LIMS.
Note the small abundances observed at low altitudes in the tropics, and
the high values found at mid-latitudes in summer; the low densities seen
in high latitude winter are related to the formation of N2 O5 as discussed
above. Note the tendency for the altitude of maximum abundance
to decrease at higher latitudes, presumably as a result of downward
transport by the mean meridional circulation.
The earliest measurements of the NO2 total column as a function of
latitude (Noxon, 1979; Noxon et al., 1979; and Coffey et al., 1981b)
showed that the abundance of NO2 increases with increasing latitude
in summer, while in winter the abundance decreases poleward of about
40 or 50 N. As noted by Noxon (1979), very sharp gradients are often
observed in the winter hemisphere (about a factor of four in total
column in only 5-10 degrees of latitude), particularly at the edge of the
winter polar vortex (this is sometimes referred to as the Noxon “cliff”).
Plate 10 shows a satellite view of the NO2 column in the polar region in
November.
The behavior of nitrogen compounds in the polar regions and
specifically the observed decrease in the abundance of NOy observed
in the vicinity of the winter vortex are the result of complex chemical
and physical processes. In the polar night, NO is converted into NO2
Figure 5.42. Zonally averaged distribution of nighttime NO2 from the LIMS
instrument (Nimbus 7) on 15 January 1979 (Gille, personal communication).
and further into NO3 by reactions with ozone (see equations (5.132) and
(5.141)). NO3 then combines with NO2 to form N2 O5 which provides
an important reservoir for nitrogen compounds in the winter hemisphere
(see e.g., Evans et al., 1982; Solomon and Garcia, 1983). N2 O5 , however,
can be rapidly converted into nitric acid by heterogeneous reaction
(5.152) on the surface of sulfate aerosol particles and, if the temperature
is sufficiently low, on the surface of polar stratospheric cloud particles.
The gravitational sedimentation of these latter particles provides a
mechanism to “denitrify” the lower polar stratosphere (e.g., Fahey et al.,
1990). Inside the Antarctic vortex during winter and early spring, for
example, the abundance of NOy (and hence of NO, NO2 and HNO3 ) is
very low. Denitrification processes are also observed occasionally in the
Arctic, but are less frequent than in Antarctica since the temperature is
generally somewhat higher during wintertime.
In the sunlit high latitude wintertime lower stratosphere, the
photochemical lifetime of N2 O5 is as large as a few days. This is
comparable to the time scale associated with transport. In the presence
of planetary wave structure, winds often flow across latitude lines,
rapidly moving N2 O5 from regions where its lifetime is long (e.g., polar
night) to sunlit regions where its lifetime is less than a day. Under
these circumstances, the N2 O5 content of a given air parcel is far from
local equilibrium, and carries a “memory” of where it has been traveling

(Solomon and Garcia, 1983).
The nighttime distribution of NO3 has been measured, for example,
by Naudet et al. (1981) using stellar occultation, and the distribution is
close to that predicted by theory. Its total column abundance has also
been reported by Noxon et al. (1978), Solomon et al. (1989a) and by
Weaver et al. (1996).
Nitric acid constitutes an important reservoir for odd nitrogen,
particularly in the lower stratosphere where its density exceeds those of
NO and NO2 . Its concentration has been measured by optical methods
(for example, Murcray et al., 1975) and by in-situ filter sampling (Lazrus
and Gandrud, 1974). Recent observations by the Cryogenic Limb Array
Etalon Spectrometer (CLAES) on board the UARS satellite (Roche
et al., 1993; 1994) are shown in Figure 5.43. The global distribution of
HNO3 has also been measured from the Space Shuttle by the CRISTA
instrument (Offermann et al., 1999). Most of the nitric acid is contained
in the lower stratosphere, where its photochemical lifetime is relatively
long (Figure 5.37), and its distribution is quite sensitive to atmospheric
transport. As in the case of ozone, the nitric acid column abundance
is greater at high latitudes, where its variability also increases. Note
2
4
6
4
8
10
0 6
2
Figure 5.43. Zonally averaged ((climatological)) distribution of the nitric acid

mixing ratio (ppbv) derived from measurements by the CLAES instrument on
board UARS for July. (Courtesy of W. Randel, NCAR).
the low concentration of HNO3 in the winter polar region of the lower
stratosphere, where this compound has been removed by sedimentation
of particles.
The general tendency for NOy to increase with latitude can be
qualitatively understood in terms of atmospheric transport. Like ozone,
NOy is a long-lived tracer in the lower stratosphere whose source lies in
the tropical middle stratosphere. Downward and poleward transport by
the meridional circulation (see Chapter 3) leads to increased total NOy
abundances towards higher latitudes, just as is found for O3 . Space
observations indicate the importance of planetary waves in establishing
odd nitrogen distributions in the middle atmosphere. They also suggest
that odd nitrogen transport from the thermosphere to the mesosphere,
and even to the stratosphere, occurs during the high latitude polar
night period. This subject will be discussed in somewhat more detail in
Section 5.5.7.
Figure 5.44 shows the vertical distribution of NOy at 24◦ N between
20 and 55 km deduced from space observations. The maximum mixing
ratio of approximately 18 ppbv is reached between 35 and 40 km altitude
where the production by the oxidation of nitrous oxide is large. The
decrease in the mixing ratio near and above the stratopause is associated
with the mesosphere destruction through the N + NO reaction.
Simultaneous measurements of NOy and N2 O provide a useful tool to
quantify the budget of odd nitrogen in the stratosphere. Observations
(Loewenstein et al., 1993; Kondo et al., 1996) show that there is
often a tight linear anticorrelation between these two compounds (see
Figure 5.45), confirming that N2 O is the dominant source of NOy and
that there is no efficient removal process for NOy in most parts of the
stratosphere. Deviations from this relationship found, for example,
in the lower polar stratosphere suggest that NOy is removed from
this region of the atmosphere by sedimentation of nitrate-containing
particles (Fahey et al., 1990). The departure from the linear NOy -N2 O
relationship for low N2 O concentrations (upper stratosphere) is another
manifestation of the N + NO reaction (Kondo et al., 1996; Chang et al.,
1996).
5.5.5 Odd Nitrogen Chemistry in the Lower Thermosphere

and Mesosphere
In the thermosphere, atomic nitrogen can be formed by dissociation

of N2 , either through the effects of solar radiation or of energetic
particles. The determination of the magnitude of this production (and
Figure 5.44. Mixing ratio of odd nitrogen (NOy ) as a function of altitude

measured in November by ATMOS and UARS instruments at 24◦ N. The figure
illustrates the role of NO photolysis parameterizations (see Nevison et al., 1997).
Figure 5.45. Correlations of NOy with N2 O concentrations measured from

ATMOS aboard the Space Shuttle and instruments aboard the ER-2 high
altitude aircraft. Adapted from Chang et al. (1996).
its variability with solar activity) are presented in Chapter 7. We note

here, however, that atomic nitrogen can be produced either in the ground
4 S or the excited 2 D state by these processes. N(2 D) reacts rapidly with
molecular oxygen:
(b7∗ ); N(2 D) + O2 → NO + O (5.180a)
or is de-activated by quenching with atomic oxygen
(bQ ); N(2 D) + O → N(4 S) + O (5.180b)
while the ground N(4 S) state reacts much more slowly, and the rate of
its reaction depends strongly on temperature:
(b7 ); N(4 S) + O2 → NO + O (5.139)
A reaction with OH also converts N(4 S) to NO:
(b7a ); N( S) + OH → NO + H
4
(5.181)
Photodissociation of nitric oxide mostly in the δ-bands (see Chapter 4),
initiates the primary loss process for odd nitrogen in the middle
atmosphere:
(JNO ); NO + hν → N(4 S) + O (5.138)
because it leads to the formation of an N(4 S)
atom, which can then react
with NO in a so-called “cannibalistic reaction”
(b6 ); N(4 S) + NO → N2 + O (5.140)
wherein two odd nitrogen particles are destroyed. The following kinetic
equations can be written for mesospheric and thermospheric conditions:
d(N4 S)
+ [b6 (NO) + b7 (O2 ) + b7a (OH)] (N4 S)
dt
= JNO (NO) + bQ (O)(N2 D) + P(N4 S) (5.182)
d(N2 D)
+ b7∗ (O2 ) + bQ (O)(N2 D) = P(N2 D) (5.183)
dt
d(NO)
+ b4 (O3 ) + a26 (HO2 ) + c5 (CH3 O2 ) + d4 (ClO) + b6 (N4 S)
dt
+ JNO ] (NO) = [JNO2 + b3 (O)] (NO2 ) + b7∗ (O2 )(N2 D)
+ [b7 (O2 ) + b7a (OH)] (N4 S) + P(NO)
(5.184)
or, considering Eq. (5.171),
d(NO)
+ b6 (N4 S) + JNO (NO)
dt
= b7∗ (O2 )(N2 D) + [b7 (O2 ) + b7a (OH)] (N4 S) + P(NO)
(5.185)
The rate of production, P(N4 S) and P(N2 D) is discussed in Chapter 7

and involves ionization and dissociation of molecular nitrogen by high
energy particles and by shortwave solar radiation. P(NO) is given by
Equation (5.128).
5.5.6 The Odd Nitrogen Family in the Lower

Thermo sphere and Mesosphere
The lifetime of atomic nitrogen is short throughout the middle

atmosphere, and steady-state conditions may therefore be assumed
for N(2 D) and N(4 S). Thus, for example, the concentration of atomic
nitrogen in the ground state (which far exceeds the concentration of N
in the electronically excited 2 D state) is given by:
JNO (NO) + P (N4 S) + bQ (O)(N2 D)
(N4 S) = (5.186)
b6 (NO) + b7 (O2 ) + b7a (OH)
This reduces approximately to
JNO (NO) + P(N4 S)
(N4 S) ≈ (5.187)
b6 (NO) + b7 (O2 )
If, again, we define the odd nitrogen (NOy ) family as
(NOy ) = (N4 S) + (N2 D) + (NO) + (NO2 ) + (NO3 )
+ 2(N2 O5 ) + (HNO3 ) + (HO2 NO2 )
which reduces essentially to
(NOy ) = (NO) + (NO2 )
in the upper stratosphere, and to
(NOy ) = (NO) + (NO2 ) + (N)
in the mesosphere and lower thermosphere, we find by adding the
equations of continuity for all of the reactive nitrogen species
d(NOy )
+ 2b6 (N4 S)(NO) = P(N2 D) + P(N4 S) + P(NO) (5.188)
dt
In the lower thermosphere and mesosphere, we may define R as follows:
(NOy ) (NO2 ) (N4 S)
R= ≈1+ +
(NO) (NO) (NO)
The value of this factor can be obtained by examination of the continuity
equations for each species. Considering transport, and substituting for
N(4 S) with Eq. (5.187), we find

∂(NOy ) 2b6 JNO (NOy )2
+ ∇ · φ(NOy ) +
∂t R(b6 (NOy ) + Rb7 (O2 ))

2 Rb7 (O2 ) − b6 (NOy )
= P(NO) + P(N D) + P(N4 S)
Rb7 (O2 ) + b6 (NOy )
(5.189)
5.5.7 Observations of Nitric Oxide in the Thermosphere

and Mesosphere
Figure 5.46 displays early observations of NO in the thermosphere and

mesosphere. The observed variability indicates the strong sensitivity
of NO to dynamic processes in the atmosphere, as one might expect
based on the comparison between its photochemical lifetime and the
time constant for dynamics in this region (see Figure 5.37). The
observed minimum near the mesopause is due to the photodissociation
of NO and its subsequent recombination with N(4 S). The depth of the
minimum depends on the competition between downward transport and
this photochemical loss. Its variability as depicted in the figure is
probably related to both seasonal and short term temporal variations
in mesospheric transport parameters.
Figure 5.46. Observations of nitric oxide from 40 to 110 km altitude. In

the lower part, the values of Horvath and Mason are shown, while at higher
altitudes, the observations include those of Meira (1971), Tisone (1973),
Tohmatsu and Iwagami (1975, 1976), Witt et al. (1976), and Baker et al. (1977).
Figure 5.47a. Meridional distribution of the NO mixing ratio (ppbv) derived

from the UARS/HALOE instrument in the mesosphere.
In the mesosphere, NO provides the primary source of electrons during

geomagnetically quiet conditions, and is responsible for the formation of
the ionospheric D-region. This will be discussed in detail in Chapter 7.
Figure 5.47a shows a distribution of NO in the mesosphere and
upper stratosphere obtained by HALOE on UARS. The calculated
variations with respect to latitude reflect the influence of solar elevation
on odd nitrogen destruction, and of spatial variability in transport
processes. Note the large NO densities in the high latitude winter
stratosphere, particularly near the polar night, where there is virtually
no photochemical loss of odd nitrogen (because JNO is zero there).
Models show that these high NO concentrations result from the
production of odd nitrogen in the thermosphere and its transport down
to the mesosphere and stratosphere at these latitudes (see, e.g., Frederick
and Orsini, 1982; Solomon et al., 1982; Brasseur, 1982; Siskind et al.,
1997; Marsh and Roble, 2002).
Figures 5.47b and c show the concentration of NO observed by the
Student Nitric Oxide Explorer (SNOE) in the lower thermosphere on 2
specific days corresponding to low and high auroral activity, respectively
(Barth et al., 1999; Solomon et al., 1999). In both cases, the NO density
reaches a maximum near 110 km altitude. The value of this maximum in
the tropics is correlated with the intensity of solar soft X-rays (2-10 nm)
during low geomagnetic activity (Barth et al., 1999). At high latitudes,
the NO density varies considerably (typically a factor 3-5) with auroral

activity. Auroral NO is generated by direct particle impact (dissociation
of N2 into excited nitrogen atoms) and by heating resulting from ion-
neutral collisions (that accelerates the conversion of N atoms into NO)
(Solomon et al., 1999).
Figure 5.47b. Nitric oxide daily averaged density ((107 cm−3 ) measured by the
SNOE satellite (Solomon et al., 1999) in the lower thermosphere. Low auroral
activity.
Figure 5.47c. Same as Figure 5.47b but for high auroral activity.
The HALOE and SNOE observations also show the dynamic behavior
of the nitrogen oxides at different heights (Plate 11). For example, above
100 km altitude, the NO concentration is characterized by a strong
seasonal component associated with solar energy input. At 97 km, the
evolution of the NO concentration is primarily influenced by the intensity
of the atmospheric tides. The vertical wind produced by the migrating
diurnal tides can induce factor 2 changes over 12 hours in the mixing
ratio of NO at the equator (Marsh and Russell, 2000).
5.6 Halogen Compounds
The presence of halogens (chlorine, fluorine, bromine and perhaps

iodine) in the middle atmosphere results from the upward transport from
the troposphere of halocarbons which are released at the Earth’s surface
as a results of natural or anthropogenic processes. These compounds
break up in the stratosphere and release halogen atoms.
Once released in the middle atmosphere, fluorine, chlorine and
bromine atoms react rapidly with ozone to form FO, ClO and BrO.
Further reactions, which will be presented in the following sections, lead
to efficient catalytic destruction of odd oxygen via ClO and BrO.
It is useful to stress that chemical partitioning processes play a
major role in halogen-catalyzed ozone destruction that is dramatically
illustrated by the contrasts between the gas-phase chemistry of different
halogen species shown in simplified form in Figure 5.56. Briefly, the
halogen atoms released in the stratosphere from chloro- bromo- and
fluorocarbon source gases can form acids (through abstraction of a
hydrogen atom) and nitrates (through reaction with NO2 ). In the
case of fluorine, the acid HF is quickly formed and so tightly bound
that essentially all fluorine released from fluorine source gases in the
stratosphere is irreversibly and rapidly ‘neutralized’ as HF (Rowland
and Molina, 1975; Stolarski and Rundel, 1975; Sze, 1978). The amount
of atomic fluorine and FO available to participate in ozone-destroying
catalytic cycles (or, in chemical terms, the catalytic chain length, see
Johnston and Podolske, 1978) is therefore extremely low and fluorine
has a negligible impact on ozone. Chlorine forms both HCl (Stolarski
and Cicerone, 1974) and ClONO2 reservoirs (Rowland et al., 1976).
Unlike HF, these gases can be reconverted to chlorine atoms by gas-
phase stratospheric chemistry (i.e., by reaction with OH and photolysis,
respectively). The amount of Cl and ClO available to participate in
ozone-destroying catalytic chemistry therefore is critically dependent
on the partitioning of chlorine between these ‘active’ chlorine radicals
and the non-ozone-destroying ‘reservoirs’, HCl and ClONO2 . The rates

of chemical formation and destruction of the reservoirs control this
partitioning. Bromine is less tightly bound than chlorine, so that
relatively little of the bromine released from bromocarbons is tied up
in HBr and BrONO2 , rendering this atom quite effective for ozone loss
in the lower stratosphere (Wofsy et al., 1975; Yung et al., 1980; Lary,
1996; Daniel et al., 1999), especially in combination with chlorine (Tung
et al., 1986; McElroy et al., 1986). Although each bromine atom released
in the stratosphere is about 50 times more effective than chlorine, the
contemporary abundances of total man-made stratospheric bromine are
about 300 times smaller than those of chlorine (e.g., Schauffler et al.,
1993; Wamsley et al., 1998). Iodine may also participate in ozone-
destroying catalytic cycles with bromine and chlorine (Solomon et al.,
1994), but its stratospheric abundance is believed to be much smaller
than either those of bromine and chlorine, and its primary sources are
believed to be natural rather than largely or partly man-made as in the
case of fluorine, chlorine, and bromine.
The effect of the polar stratospheric clouds, which form during winter
when the temperature decreases below a threshold of approximately
190 K, is to convert chlorine reservoirs (HCl and ClONO2 ) back to
reactive chlorine through heterogeneous processes that occur on the
surface of the cloud particles. This activation of chlorine (and analogous
reactions for bromine) leads to rapid ozone depletion at high latitudes
after the return of solar light in late winter/early spring. These processes
explain the formation of the Antarctic ozone hole in September/October
and the large depletion of ozone observed in the Arctic during cold
winters. Heterogeneous reactions occurring on the surface of cold sulfate
aerosol particles in the stratosphere also lead to chlorine activation
and ozone destruction. This effect is considerably amplified after large
volcanic eruptions (such as the eruption of Mount Pinatubo in 1991)
when the aerosol load in the stratosphere can be enhanced by several
orders of magnitude.
The next sections present more details about the photochemical
processes affecting halogens in the middle atmosphere, and highlight
their relations with other chemical families. Chapter 6 contains a
detailed review of ozone depletion by these gases.
5.6.1 Sources of Inorganic Halogens
Numerous chlorine, fluorine, and bromine-containing compounds

(halocarbons) are produced at the Earth’s surface, both by natural and
industrial processes. Fully halogenated molecules are especially stable,

and nearly insoluble in water so they do not rain out to any significant
degree. The halocarbons are transported to the stratosphere where they
are photo-decomposed (by short-wave solar radiation) and constitute
the main source of inorganic halogen species. Methyl chlorine (CH3 Cl)
is a compound with natural sources, including the oceans, but it is also
produced by combustion of biomass, especially in tropical agriculture.
Its destruction results mainly from reaction with the hydroxyl radical
in the troposphere, but due to its relatively long lifetime (1.5 years),
significant amounts of this gas penetrate into the stratosphere, where
photolysis becomes another important loss mechanism. Because the
anthropogenic contribution to CH3 Cl emissions is small, methyl chloride
should be regarded as the largest natural source of inorganic chlorine
in the middle atmosphere. Its mixing ratio at the surface is about
600 pptv, while those of the industrial compounds represented about
3 ppbv (3000 pptv) as of 1992 (see Schauffler et al., 1993).
Methyl bromide (CH3 Br), which is also released by the ocean (Singh
et al., 1983) and produced by biomass burning (Manõ˜ and Andreae,
1994), is believed to be the largest bromine reservoir in the atmosphere,
with surface mixing ratios of typically 9 pptv in the southern hemisphere
and 11 pptv in the northern hemisphere. Its lifetime against destruction
by OH in the troposphere and uptake by the ocean is of the order
of 8 months, so that significant amounts of CH3 Br are transported
to the stratosphere where photodecomposition leads to the formation
of inorganic bromine. A significant amount of methyl bromide is also
produced for agricultural purposes, specifically as a pre-planting and
post-harvesting fumigant.
Methyl iodide (CH3 I) is produced by marine algae and phytoplankton
in the ocean and probably also by biomass burning. Its lifetime in the
atmosphere is only a few days, so that its transport to the stratosphere
should occur only during strong convective events (Solomon et al., 1994).
The surface mixing ratio of CH3 I is typically a few pptv in the boundary
layer with values reaching 10-20 pptv over highly productive regions
of the ocean. An estimated budget of natural halocarbons is given in
Table 5.8.
While many natural processes produce chlorine at ground level
(including for example, sea salt and volcanic emissions of HCl), these
compounds are efficiently removed in precipitation (rain and snow)
due to high solubility. The removal of HCl emitted, for example, in
volcanic plumes (which contain a great deal of water and hence form
rain) is extremely efficient (see, e.g., Tabazadeh and Turco, 1993).
This renders even the most explosive volcanic plumes ineffective at
providing significant inputs of chlorine to the stratosphere. Observations
Table 5.8 Estimated Budgets of CH3 Cl, CH3 Br, and CH3 I
(After Brasseur et al., 1999)
CH3 Cl CH3 Br CH3 I

(Tg yr−1 ) (Tg yr−1 ) (Tg yr−1 )
Typical surface mixing 650 10 2
ratio (pptv)
Sources
Biomass burning source 0.7 0.01-0.05 -
Ocean source 0.4 ∼0.03 1-2
Terrestrial ecosystems 0.1 - -
Anthropogenic source 0.05 0.02-0.06 -
TOTAL SOURCES ∼1.2 ∼0.05-0.15 ∼1.5
Loss to soils 0.1 0.01-0.07 -

Atmospheric loss 3 0.06-0.11 5
Atmospheric lifetime 1.5 years 1.3 years 5 days
of limited or even unobservable changes in HCl after the eruptions of the

El Chichón
´ and Mt. Pinatubo volcanoes in 1981 and 1991 demonstrates
this effect (Mankin et al., 1992 and Wallace and Livingston, 1992).
Over recent decades, the global budget of halogens has been
considerably modified by the increasingly large industrial production
of chlorofluorocarbons (CFC’s). These chlorine and fluorine-containing
compounds, which are generally very stable in the atmosphere (10-
100 years lifetime), have been used extensively as refrigerants (CFC-
11, -12, -114), solvents (CFC-113), inflating agents (CFC-11, -12) in
the manufacturing of foams, and as aerosol propellants. Other organic
agents such as methyl chloroform (CH3 CCl3 ) are also used as solvents,
while bromine-containing halons (e.g., CBrClF2 , CBrF3 ) are used as fire
retardants. It is estimated that, in the mid-1990s, approximately 75-
85% of stratospheric chlorine was of anthropogenic origin with a total
CFC emission of approximately 1 Tg yr−1 . Because of the threat to
the ozone layer, CFC’s, methyl chloroform, carbon tetrachloride, and
the halons have been gradually phased out as a result of international
agreements (Montreal Protocol in 1987 and subsequent amendments;
see Chapter 6), and replaced by hydrochlorofluorocarbons (HCFC’s) and
hydrofluorocarbons (HFC’s). The contribution of the HCFC-substitute
compounds to ozone depletion is limited by the fact that they are partly
destroyed by the OH radical in the troposphere. Observations have
confirmed that the temporal trends in global surface-level abundances of

chlorofluorocarbons are consistent with the known industrial emissions
(e.g., Montzka et al., 1996a; Prinn et al., 1995; WMO/UNEP, 1985;
1991; 1995; 1999), both in terms of the buildup of these gases in past
decades and the slower accumulation in the 1990s following reductions
in global use. The HFC’s do not contain chlorine atoms at all.
However, all organic halogen compounds absorb infrared radiation,
and hence contribute to greenhouse warming. Taken together, their
current radiative forcing is estimated at 0.34 Wm−2 (IPCC, 2001).
Table 5.9 provides a list of the major halocarbons with estimated
surface concentrations in the 1990s and calculated atmospheric lifetimes,
while Figure 5.48 shows a zonally averaged distribution of CFC-12
derived from the observations by CLAES on board UARS (Nightingale
et al. 1996). Global distributions of CFC-11 observed by the spaceborne
CRISTA instrument have been reported by Riese et al. (1999). Note
that, as a result of measures taken to reduce or phase-out the production
of the most harmful halocarbons, the concentration of inorganic chlorine
and bromine is expected to decrease in the future. This downward trend,
however, should be slow due to the long residence time of these chemical
species in the atmosphere. Models suggest that the level of chlorine in
the stratosphere should fall below the levels at which ozone depletion
first became observable (about 2 ppbv) around the year 2050.
Figure 5.48. Meridional distribution of CFC-12 ((CF2 Cl2 ) provided by CLAES

(on board the UARS satellite). Mixing ratio values (in ppbv) representative of
July (Courtesy of W. Randel, NCAR).
Table 5.9 Abundances and Lifetimes of Halogenated Organic Source

Gases (Adapted from Brasseur et al., 1999)
Molecule Concentration (ppt) Lifetime(2)

1990(1) 1992(2) 1995(3) (years)
CF2 Cl2 (CFC 12) 474-479 503 532 102
CFCl3 (CFC 11) 254-263 268 272 50±5
CF2 ClCFCl2 (CFC 113) 71-75 82 84 85
CF2 ClCF2 Cl (CFC 114) 15-20 20 300
CF3 CFCl2 (CFC 114a) 5 400
CF3 Cl (CFC 13) 5 640
CCl4 (CFC 10 or
carbon tetrachloride) 107 106(3) 103 42
CH3 CCl3 (methylchloroform) 140(3) 135(3) 109 5.4±0.4
CHF2 Cl (HCFC 22) 90(5) 102(5) 117 13.3
CH3 Cl (methyl chloride) 600 600 ∼550 1.5
CHCl3 (chloroform) 15 0.6
CH2 Cl2 (dichloromethane) 30 0.4
CCl2 CCl2 (tetrachloroethane) 10 0.3
CH2 ClCH2 Cl (dichloroethane) 35
CHClCCl2 (trichloroethane) 2-3 0.01
CH3 Br (methyl bromide) 10-15 1.3
CH2 Br2 (dibromomethane) 0.5-3 0.5
CHBr3 (bromoform) 0.2-0.3 ∼0.06
CH2 ClBr 1-2
CHClBr2 1
CHCl2 Br 1
C2 H4 Br2 (ethylene dibromide) <1
C2 H5 Br (ethyl bromide) 2-3
CF2 ClBr (H 1211) 1.2-2.2 20
CF3 Br (H 1301) 1.7-2.5 65
CH3 I (methyl iodide) ≈1-10 0.01
CH3 CFCl2 (HCFC-141b) 0.3 3.5 9.4
CH3 CF2 Cl (HCFC-142b) 3.5 7.0 19.5
CF3 CH2 F (HFC-134a) 0.3(6) 1.73(6) 14
Total Cl/Br 3800/30
(1) Kaye et al., 1994
(2) WMO, 1995
(3) Montzka et al., 1996a
(4) WMO, 1992
(5) Montzka et al., 1993
(6) Montzka et al., 1996b
Sulfur hexafluoride (SF6 ), a fluorine-containing gas released by

high voltage electrical equipment (e.g., powerplants) and emitted by
metallurgy facilities, is not a signifcant source of inorganic fluorine in
the atmosphere. Its abundance is small (a few pptv), but has been
growing rapidly (i.e., 7%/yr) in recent decades. This gas, which is
photochemically destroyed only above 70 km altitude and hence has
a lifetime of 800 to 3200 years (see Maiss et al., 1996), is of particular
interest because it can be used to estimate the mean age of air after it
penetrates into the stratosphere (see Chapter 3).
5.6.2 Halogen Chemistry
5.6.2.1 Chlorine
The natural production of chlorine atoms in the stratosphere is
provided by the destruction of methyl chloride, either by photolysis
(JCH3 Cl ) ; CH3 Cl + hν → CH3 + Cl (5.190)
or by reaction with the hydroxyl radical
(d0 ) ; CH3 Cl + OH → CH2 Cl + H2 O (5.191)
A major source of anthropogenic chlorine currently results from
the photolysis of chlorofluorocarbons and other chlorocarbons in the
stratosphere. In the case of CFCl3 (also called chlorofluorocarbon-11 or
CFC-11) the most likely degradation scheme is the following:
CFCl3 + hν (λ < 226 nm) → CFCl2 + Cl (5.192a)
CFCl2 + O2 + M → CFCl2 O2 + M (5.192b)
CFCl2 O2 + NO → CFCl2 O + NO2 (5.192c)
CFCl2 O + M → COFCl + Cl + M (5.192d)
COFCl + hν → FCO + Cl (5.192e)
Following this chain, three chlorine atoms are released. Similar reaction
paths apply to other chlorofluorocarbons including CFC-12. In this case,
the primary destruction mechanism is
CF2 Cl2 + hν (λ < 215 nm) → CF2 Cl + Cl (5.193)
A secondary destruction path is provided by the reaction with the
electronically excited oxygen atom O(1 D). For example, in the case of
CFC-12,
CF2 Cl2 + O(1 D) → CF2 Cl + ClO (5.194)
Other channels are possible and could produce Cl or Cl2 . The
importance of the degradation of industrially manufactured chloro-
fluorocarbons as a source of stratospheric chlorine (with potential impact

on ozone) was first noted by Molina and Rowland (1974a) and Rowland
and Molina (1975). Other significant sources of anthropogenic chlorine
include the photolysis of carbon tetrachloride
(JaCCl4 ) ; CCl4 + hν → CCl3 + Cl (5.195a)
(JbCCl4 ) ; CCl4 + hν → CCl2 + Cl2 (5.195b)

and of methyl chloroform
(JCH3 CCl3 ) ; CH3 CCl3 + hν → CH3 CCl2 + Cl (5.196)
This latter compound can also react with the hydroxyl radical
(d1 ) ; CH3 CCl3 + OH → CH2 CCl3 + H2 O (5.197)
Further processes lead to the release of chlorine atoms.
The increasing use of hydrochlorofluorocarbons could significantly
affect the stratospheric budget of chlorine. For example, reactive
chlorine is released following the photolysis of HCFC-22
(JCF2 HCl ); CF2 HCl + hν → products (5.198)
and its oxidation by OH
(d1a ); CF2 HCl + OH → CF2 Cl + H2 O (5.199)
Assuming that the intermediate photolysis products are eventually
completely dissociated in the stratosphere and neglecting destruction
reactions with O(1 D), we can write to a good approximation:
PCl = 4JCCl4 (CCl4 ) + 3JCFCl3 (CFCl3 ) + 2JCF2 Cl2 (CF2 Cl2 )
+ [JCH3 Cl + d0 (OH)] (CH3 Cl)

(5.200)
+ 3 [JCH3 CCl3 + d1 (OH)] (CH3 CCl3 )
+ [JCF2 HCl + d1a (OH)] (CF2 HCl)
When a chlorine atom is produced in the stratosphere, it is converted
to chlorine monoxide mainly by ozone:
(d2 ); Cl + O3 → ClO + O2 (5.201)
and, to a lesser extent, by HO2
(d33 ); Cl + HO2 → ClO + OH (5.202)
Chlorine monoxide undergoes two principal reactions during the
daytime, one of them with atomic oxygen (mostly in the upper
stratosphere and mesosphere)
(d3 ); ClO + O → Cl + O2 (5.203)
and a second with NO (mostly in the lower stratosphere)
(d4 ); ClO + NO → Cl + NO2 (5.133)
These reactions reform atomic chlorine. The reaction between ClO and
NO represents an important coupling between the chlorine and nitrogen
cycles, and the pair of reactions (5.201) and (5.203) constitutes an
important catalytic cycle which destroys odd oxygen (primarily in the
upper stratosphere).
The photodissociation of ClO
(JClO ); ClO + hν → Cl + O (5.204)
and conversion of ClO to Cl by OH
(d8 ); ClO + OH → Cl + HO2 (5.205)
play a secondary role. However, this latter reaction has a small efficiency
to produce hydrogen chloride HCl (5-7%)
(d46 ); ClO + OH → HCl + O2 (5.206)
This channel reduces significantly the calculated ClO/HCl ratio in the
upper stratosphere and, hence, the sensitivity of ozone to chlorine.
There are also other slow processes that yield hydrogen chloride. Like
HNO3 , HCl provides a relatively inert reservoir which sequesters a
photochemically active species and hence reduces the rate of its catalytic
reaction with odd oxygen.
(d5 ); Cl + CH4 → CH3 + HCl (5.48)
(d6 ); Cl + H2 → H + HCl (5.207)

(d7 ); Cl + HO2 → O2 + HCl (5.208)
(d8 ); Cl + H2 O2 → HO2 + HCl (5.209)
(d10 ); Cl + CH2 O → HCO + HCl (5.64)

Reaction (5.48) is generally the most important of these, although
(5.208) plays a substantial role in the upper stratosphere. HCl can be
destroyed by reaction with OH, O(1 D), and O(3 P) to yield a chlorine
atom:
(d11 ); HCl + OH → H2 O + Cl (5.210)
(d25 ); HCl + O(1 D) → OH + Cl (5.211)
(d12 ); HCl + O(3 P) → OH + Cl (5.212)
Laboratory studies (Wine et al., 1986) suggest that the reaction between
HCl and O(1 D) leads mainly to the formation of OH and Cl, but that
a channel producing H and ClO is also significant. Photolysis must also
be considered:
(JHCl ); HCl + hν (λ < 220nm) → H + Cl (5.213)
Finally, HCl is soluble in water, and can therefore be removed from the
atmosphere in cloud processes.
To end this brief discussion on the reactions of chlorine species, we
must consider the formation of other constituents such as ClOO, OClO,
ClONO2 , HOCl, etc. Chlorine nitrate ClONO2 , another important
“chemical reservoir” for chlorine, is formed by the reaction of ClO with
NO2 :
(d31 ); ClO + NO2 + M → ClONO2 + M (5.157)
and dissociated in the ultraviolet (Chang et al., 1979)
(JaClONO2 ); ClONO2 + hν → Cl + NO3 (5.159a)
with a quantum yield of approximately 0.8. The efficiency of a second
channel
(JbClONO2 ); ClONO2 + hν → ClO + NO2 (5.159b)
is of the order of 0.15-0.20 while the yield of a third channel
(JcClONO2 ); ClONO2 + hν → ClONO + O (5.159c)
is less than 0.05. A loss with atomic oxygen, atomic chlorine and the
hydroxyl radical should also be mentioned:
(d32 ); ClONO2 + O → products (5.214)
(d82 ); ClONO2 + Cl → Cl2 + NO3 (5.215)
(d87 ); ClONO2 + OH → HOCl + NO3 (5.216)
This latter reaction leads to the formation of HOCl. The primary
production of HOCl, however, is provided by the reaction
(d33 ); HO2 + ClO → HOCl + O2 (5.217)
HOCl is destroyed mostly by photolysis
(JHOCl ); HOCl + hν (λ < 503 nm) → Cl + OH (5.218)
and by reaction with OH:
(d34 ); OH + HOCl → H2 O + ClO (5.219)
It can also react with atomic oxygen
(d35 ); HOCl + O → ClO + OH (5.220)
and with atomic chlorine
(d37 ); HOCl + Cl → Cl2 + OH (5.221a)
with a possible channel
(d37a ); HOCl + Cl → HCl + ClO (5.221b)
The concentration of inorganic chlorine compounds in the strato-
sphere can also be affected by heterogeneous reactions on the surface
of solid and liquid particles. For example, hydrolysis of chlorine nitrate

(h2 ); ClONO2 (g) + H2 O(, s) → HOCl(g) + HNO3 (5.222)
on the surface of water ice particles, nitric acid ice, supercooled
ternary solutions, and liquid sulfuric acid particles (aerosols) leads to
the formation of HOCl and of nitric acid. Because HOCl is rapidly
photolyzed into Cl and OH, heterogeneous reaction (5.222) provides an
efficient mechanism to activate chlorine (i.e., convert a relatively inert
chlorine reservoir into a very reactive form of chlorine) in the lower
stratosphere. The efficiency of the reaction, and hence the value of the
reaction probability, depends strongly on the concentration of sulfuric
acid in the solution and hence on the temperature (e.g., Rossi et al.,
1987; Tolbert et al., 1988; Hanson and Ravishankara, 1994; Manion
et al., 1994; Williams et al., 1994; Robinson et al., 1997; Hanson, 1998).
It decreases typically from 0.1 for a H2 SO4 content of 40% by weight
(corresponding to a stratospheric temperature T of 190 K) to 10−3 for
60 wt% H2 SO4 (or T = 205 K), assuming a water partial pressure
representative of the lower stratosphere (Figure 2.5 in Chapter 2). The
reactivity of ClONO2 is increased when significant quantities of HCl are
dissolved in the H2 SO4 /H2 O solution (Hanson, 1998; see Figure 2.6).
As the stratospheric temperature decreases and hence the solubility of
HCl in a liquid aerosol increases, reaction
ClONO2 (g) + HCl(s, ) → Cl2 (g) + HNO3 (5.223)
becomes competitive with and even faster than the reaction of chlorine
nitrate with water molecules. When the hydrolysis of ClONO2 takes
place on water ice, the reaction probability is of the order of 0.3 (Hanson
and Ravishankara, 1991); it drops to less than 0.01 when the ice surface
contains HNO3 . Note that the nitric acid that is produced by the
hydrolysis of N2 O5 and ClONO2 is partitioned between the gas and
the condensed phases according to thermodynamics. For example, at
relatively high temperature (T > 190 K), when the particles present
in the stratosphere are liquid sulfuric acid aerosols, HNO3 is released
mostly in the gas phase. A similar reaction with dissolved HBr also
occurs
(h3 ); ClONO2 (g) + HBr(s, ) → BrCl(g) + HNO3 (5.224)
Heterogeneous reactions
(h4 ); N2 O5 (g) + HCl(s,) → ClNO2 (g) + HNO3 (5.225)
and
(h5 ); HOCl(g) + HCl(s,) → Cl2 (g) + H2 O (5.226)
on the surface of particles are responsible for additional chlorine
activation. Note that Reaction (5.225) produces ClNO2 . This compound
can also be formed by the three-body reaction

(d36 ); Cl + NO2 + M → ClNO2 + M (5.227)
But the major product (90%) of this reaction is believed to be ClONO:
(d36a ); Cl + NO2 + M → ClONO + M (5.228)
ClONO is either rapidly photolyzed or converted by isomerization into
ClNO2 .
In the sunlit atmosphere, the photolysis of Cl2 , HOCl, and ClNO2
(JCl2 ); Cl2 + hν → Cl + Cl (5.229)
(JHOCl ); HOCl + hν (λ < 370 nm) → Cl + OH (5.230)

(JClNO2 ); ClNO2 + hν → Cl + NO2 (5.231)
produces reactive chlorine atoms.
When high ClO concentrations (0.5-2 ppbv) are present in the
stratosphere (e.g., in polar regions during winter when air masses are
processed by polar stratospheric clouds), the three-body reaction
(d60 ); ClO + ClO + M → Cl2 O2 + M (5.232)
leads to the formation of significant quantities of the ClO dimer (Cl2 O2 ).
During nightime, in the springtime polar region, Cl2 O2 is often the
most abundent inorganic chlorine-containing compound in the lower
sratosphere. During daytime, Cl2 O2 is photolyzed
(JCl2 O2 ); Cl2 O2 + hν → Cl + ClOO (5.233)
producing a chlorine atom and the ClOO radical, which is rapidly
photolyzed or is decomposed according to
(d62 ) ClOO + M → Cl + O2 + M. (5.234)
In the PSC processed airmasses of the springtime polar lower
stratosphere, ClO becomes the most abundant inorganic chlorine species
during daytime. The Cl2 O2 dimer itself can also be decomposed upon
collision with an atmospheric molecule M
(d61 ) Cl2 O2 + M → ClO + ClO (5.235)
at a rate which increases substantially with increasing temperature.
Thermal decomposition (5.235) maintains significant levels of ClO in
the polar region during the night, if the temperature is sufficiently high
(Lefèvre
f et al., 1994).
The analysis of the termolecular reaction (d60 ) is complicated by the
existence of three bimolecular channels
(d47 ); ClO + ClO → OClO + Cl (5.236a)
(d48 ); → Cl2 + O2 (5.236b)
(d49 ); → ClOO + Cl . (5.236c)

The reactions between chlorine and bromine monoxide also play an
important role in the lower stratosphere, particularly at high latitudes
in winter. Three channels must be considered
(e5a ); ClO + BrO → OClO + Br (5.237a)
(e5b ); → ClOO + Br (5.237b)

(e5c ); → BrCl + O2 (5.237c)
Pathway (a), with an efficiency of approximately 60% of the overall
reaction at 200 K, is the major production mechanism for OClO.
This latter compound, however, is rapidly photolyzed in the sunlit
atmosphere
(JOClO ); OClO + hν → ClO + O (5.238)
so that its abundance is greatest during nighttime. BrCl, produced by
pathway (c), with a yield of approximately 8%, is photodissociated in
sunlit regions
BrCl + hν (λ < 540 nm) → Br + Cl . (5.239)

Retaining the major reactions affecting inorganic chlorine compounds
(see also Figure 5.49), the continuity equations for Cl, ClO, HCl, HOCl,
ClONO2 , Cl2 O2 , OClO, ClOO, BrCl, Cl2 , and ClNO2 can be easily
derived:
d(Cl)
+ (Cl) [d2 (O3 ) + d5 (CH4 ) + d6 (H2 ) + d7 (HO2 ) + d8 (H2 O2 )
dt
+d10 (CH2 O)] = PCl + [d3 (O) + d4 (NO) + JClO + d8 (OH)] (ClO)
+ [d12 (O) + d11 (OH) + JHCl ] (HCl) + JaClONO2 (ClONO2 )

+ JHOCl (HOCl) + JCl2 O2 (Cl2 O2 ) + d62 (M)(ClOO)
+ (d47 + d49 )(ClO)2 + 2JCl2 (Cl2 ) + JClNO2 (ClNO2 )

+ JBrCl (BrCl)
(5.240)
where PCl represents production of chlorine from organo-chlorine
molecules.
d(ClO)
+ (ClO) [d3 (O) + d4 (NO) + JClO + (d8 + d46 ) (OH)
dt
+ d31 (M)(NO2 ) + 2d60 (M)(ClO) + 2(d47 + d48 + d49 )(ClO)

+(e5a + e5b + e5c )(BrO)]

= d2 (O3 )(Cl) + JbClONO2 + d32 (O) (ClONO2 )
+ 2d61 (M)(Cl2 O2 ) + JOClO (OClO) (5.241)
d(HCl)
+ (HCl) [d12 (O) + d11 (OH) + JHCl + h3 (ClONO2 ) + h4 (N2 O5 )
dt
+h5 (HOCl) + βp ] = (Cl) [d5 (CH4 ) + d6 (H2 ) + d7 (HO2 )
+d8 (H2 O2 ) + d10 (CH2 O)] + d46 (OH)(ClO) (5.242)

where βp represents the pseudo first-order rate (s−1 ) for wet scavenging
of HCl (in the troposphere)
d(HOCl)
+ [JHOCl + d34 (OH)
dt
+h5 (HCl)] (HOCl) = d33 (HO2 )(ClO) + h2 (H2 O)(ClONO2 )
(5.243)
Principal chemical reactions affecting chlorine compounds in

the stratosphere. Concentrations (cm−3 ) and reaction fluxes (cm−3 s−1 )
between different compounds calculated at 25 km altitude (24-hour average)
are indicated. After Zellner (1999).
d(ClONO2 )
+ (ClONO2 ) [JClONO2 + d32 (O) + h2 (H2 O)
dt (5.244)
&
+h3 (HCl) + h3 (HBr) = d31 (M)(ClO)(NO2 )
d(Cl2 O2 )
+ (Cl2 O2 ) [JCl2 O2 + d61 (M)] = d60 (M)(ClO)2 (5.245)
dt
d(OClO)
+ JOClO (OClO) = d47 (ClO)2 (5.246)
dt
d(ClOO)
+ d62 (M)(ClOO) = d49 (ClO)2 + JCl2 O2 (Cl2 O2 ) (5.247)
dt
d(BrCl)
+ JBrCl (BrCl) = e5c (ClO)(BrO) (5.248)
dt
d(Cl2 )
+ JCl2 (Cl2 ) = h3 (ClONO2 )(HCl) + h5 (HOCl)(HCl)
dt
+ d48 (ClO)2 (5.249)
d(ClNO2 )
+ JClNO2 (ClNO2 ) = h4 (HCl)(N2 O5 ) (5.250)
dt
where JClONO2 represents the total photolysis coefficient of ClONO2 .
One-dimensional model predictions of the theoretical distributions of
chlorine compounds are shown in Figure 5.50. The diurnal variations of
these species at two specific altitudes are also displayed in Figure 5.51.
5.6.2.2 Fluorine
The dissociation of chlorofluorocarbons such as CFCl3 and CF2 Cl2 ,
as well as other species such as CF4 , represents a source of atmospheric
fluorine. Fluorine atoms react rapidly with molecular oxygen
F + O2 + M → FO2 + M (5.251)
to form FO2 , which in turn is rapidly converted back to fluorine atoms:
FO2 + NO → FNO + O2 (5.252)
FNO + hν → F + NO . (5.253)
The F atom also reacts with ozone:
(ff2 ); F + O3 → FO + O2 (5.254)
to produce fluorine monoxide, which, like chlorine monoxide, can react
either with atomic oxygen
(ff3 ); FO + O → F + O2 (5.255)
Figure 5.50. Calculated vertical distribution of reactive chlorine species at mid-

latitudes equinox conditions, noon values. Total amount of chlorine is assumed
to be 3 ppbv.
Figure 5.51. Diurnal variation of chlorine species calculated for May 10 and
65◦ N at 20 km (a) and 40 km (b).
or with nitric oxide:

(ff4 ); FO + NO → F + NO2 (5.256)
Like chlorine, the fluorine atom is also converted into an acid, HF:
(ff5 ); F + CH4 → HF + CH3 (5.257)
(ff6 ); F + H2 → HF + H (5.258)
(ff7 ); F + H2 O → HF + OH (5.259)
Unlike HCl however, hydrogen fluoride does not react appreciably with
OH, so that the fluorine atoms, which react with methane, hydrogen,
or water vapor, are effectively stabilized as HF. Since the sources of HF
and HCl are similar but their loss rates are quite different as a result
of the reaction of OH with HCl at low altitudes, the HF/HCl ratio can
provide an indication of the abundance of the OH radical (see, e.g., Sze
and Ko, 1980).
Because of the stability of HF, the atmospheric densities of F and FO
are very small and the effect of fluorine on odd oxygen is insignificant.
The reaction of HF with O(1 D) is chemically possible, but is negligible
due to the low abundance of this excited atom. The distribution of HF
is therefore largely determined by the rates of the surface emission of
fluorine containing gases, of photochemical destruction of these gases,
and atmospheric dynamics.
5.6.2.3 Bromine
Molecules of the type Cx Hy Brz produce bromine atoms upon
photodissociation and oxidation, like the other halogen atoms just
discussed. Unlike fluorine, the chemistry of bromine is broadly similar
to that of chlorine (see Yung et al. (1980); Poulet et al. (1992); Lary
(1996) and Figure 5.52) and is therefore capable of affecting the ozone
budget. The production of bromine atoms is followed by the reaction
(e2 ); Br + O3 → BrO + O2 (5.260)
which produces BrO, the most abundant inorganic bromine compound
in the stratosphere below 40 km altitude. In the lower stratosphere, Br
is also converted into BrO by
(e71 ); Br + OClO → BrO + ClO (5.261)
The main loss of BrO in the upper stratosphere is due to its reaction
with atomic oxygen
(e3 ); BrO + O → Br + O2 , (5.262)
Figure 5.52. Principal chemical reactions affecting bromine species in the

stratosphere. Concentrations (cm−3 ) and reaction fluxes (cm−3 s−1 ) between
different compounds calculated at 25 km altitude (24-hour average) are
indicated. After Zellner (1999).
while, in the lower stratosphere, the major destruction mechanisms are

provided by photolysis
(JBrO ); BrO + hν (λ < 380 nm) → Br + O (5.263)
and the reaction with NO
(e4 ); BrO + NO → Br + NO2 (5.134)
Conversion of BrO to Br also occurs via
BrO + OH → Br + HO2 (5.264)
Other reactions such as
(e5a ); BrO + ClO → Br + OClO (5.237a)
(e5b ); → Br + ClOO (5.237b)
(e5c ); → BrCl + O2 (5.237c)
(e6 ); BrO + BrO → 2Br + O2 (5.265)

contribute only marginally to the destruction of BrO. Bromine thus plays
a role not only through its own catalytic ozone destruction cycle, but can
also influence the chlorine cycle through reaction e5 (see Wofsy et al.,
1975). Indeed, the effect of these halogens on the ozone molecule can be
significantly influenced by this “synergy” between the catalytic cycles,

wherein both become more effective. Hydrogen bromide (HBr) is formed
almost entirely by reactions
(e7 ); Br + HO2 → HBr + O2 (5.266)
and
Br + H2 CO → HBr + HCO (5.267)
because Br does not react efficiently with methane or molecular
hydrogen, and because the reaction
(e9 ); Br + H2 O2 → HBr + HO2 (5.268)
proceeds slowly at middle atmospheric temperatures. The photolysis of
HBr in the stratosphere is slow because it involves shortwave ultraviolet
light. HBr reacts, however, rapidly with OH, and with O(1 D)
(e11 ); HBr + OH → Br + H2 O . (5.269)
(e∗12 ); HBr + O(1 D) → Br + OH (5.270)
The reaction with atomic oxygen O(3 P)
(e12 ); HBr + O(3 P) → Br + OH (5.271)
becomes important only above 50 km altitude. Finally, BrO can, like
ClO, associate with NO2 :
(e13 ); BrO + NO2 + M → BrONO2 + M (5.158)
and react with HO2 :
(e15 ); BrO + HO2 → HOBr + O2 (5.272)
to form BrONO2 and HOBr, both of which rapidly photodissociate.
By analogy with ClONO2 , the photolysis of BrONO2 (at λ < 450
nm) is believed to produce mainly Br + NO3 which leads to catalytic
ozone destruction (see Burkholder et al., 1995, and cycle (5.337)). The
photolysis of HOBr (λ < 425 nm), which leads to the formation of Br
and OH, is the major destruction mechanism for this species in the lower
stratosphere, and is also an odd oxygen destruction mechanism. Above
approximately 25 km, the main loss of HOBr is provided by the reaction
with atomic oxygen
(e16 ); HOBr + O → BrO + OH . (5.273)
It is important to note that gas-phase BrONO2 can be converted into
HOBr by heterogenous reaction on the surface of sulfate aerosol and
PSC particles.
(h10 ); BrONO2 (g) + H2 O(s, ) → HOBr(g) + HNO3 (5.274)
(see Hanson and Ravishankara, 1994). This reaction provides a
significant source of HOx due to the subsequent photolysis of HOBr. As
a result of the enhanced OH concentration, additional HCl is converted

to reactive chlorine (ClO) which leads to increased ozone loss at all
latitudes and for all seasons, particularly when high loadings of sulfate
aerosols are present in the stratosphere (e.g., Lary et al., 1996).
Finally, as shown by reaction (5.237c), the reaction between BrO and
ClO leads to the formation of BrCl. Other production mechanisms are
provided by heteorogeneous reactions on the surface of aerosol and of
PSC particles:
(h13 ); BrONO2 (g) + HCl(s,) → BrCl(g) + HNO3 (5.275)
(h11 ); HOBr(g) + HCl(s,) → BrCl(g) + H2 O (5.276)
(h12 ); HOCl(g) + HBr(s,) → BrCl(g) + H2 O (5.277)
The loss of BrCl below 30 km is due almost entirely to its photolysis
(JBrCl ); BrCl + hν (λ < 540 nm) → Br + Cl , (5.278)
while, above this altitude, the major destruction mechanism is the
reaction with atomic oxygen
(e17 ); BrCl + O → BrO + Cl . (5.279)
Figure 5.53 presents calculated vertical distributions of a number of
bromine compounds, and Figure 5.54 shows the diurnal variation of
these species at two selected altitudes. The most abundant bromine
species during daytime below 35-40 km is BrO, despite its short chemical
lifetime (approximately 10 seconds in the sunlit stratosphere below
35 km). The second most abundant compound in the sunlit atmosphere
is HOBr (which represents about 20-30% of odd bromine in the lower
stratosphere). Its lifetime during daytime varies from several minutes
at 15 km to a few seconds at 50 km. BrONO2 is the most abundant
inorganic bromine compound during nighttime in the stratosphere up
to 40 km, and is approximately as abundant as HOCl during daytime
near 20 km altitude. The lifetime of BrONO2 is of the order of several
minutes throughout the sunlit stratosphere, while that of HBr is close to
1 day at 15 km and 1 hour at 50 km. Reactive bromine compounds are
thus short-lived, and hence subject to diurnal variations (Figure 5.54).
5.6.2.4 Iodine
Ozone in the lower stratosphere may in principle be affected by iodine
chemistry (Solomon et al., 1994). The abundance of total iodine in the
troposphere is believed to be of the order of pptv, but the fractional
partitioning of iodine free radicals (I and IO) is much higher than in the
case of other halogens (chlorine and even of bromine; see Section 5.6.3).
Figure 5.53. Vertical distribution of bromine compounds in the stratosphere

(24-hour-average at 30◦ N equinox). Adapted from McElroy (2002).
Figure 5.54. Diurnal variation of bromine species calculated for May 10 and
65◦ N at 20 km (a) and 40 km (b).
Thus, even if limited amounts of iodine penetrated into the stratosphere,

they could be very efficient in depleting ozone.
A likely source of active iodine is provided by the photolysis of methyl
iodide (CH3 I) and perhaps of other iodocarbons. As the atmospheric
lifetime of CH3 I is relatively short (a few days), the tropospheric
abundance of this compound is generally lower than 10 pptv (Moyers
and Duce, 1972; Singh et al., 1983; Atlas et al., 1993) although local
maxima are found over the productive regions of the ocean (Oram and
Penkett, 1994). If iodine atoms are released above the tropopause, they
react with ozone to form the iodine monoxide radical
(g2 ); I + O3 → IO + O2 . (5.280)
IO undergoes rapid photolysis
(JIO ); IO + hν → I + O (5.281)
and reacts with NO
(g4 ); IO + NO → I + NO2 (5.282)
or with oxygen atoms
(g3 ); IO + O → I + O2 . (5.283)
Reservoir species (HOI, IONO2 and HI) are produced and destroyed by
the following reactions:
(g15 ); IO + HO2 → HOI + O2 (5.284)
(JHOI ); HOI + hν → I + OH (5.285)
(g13 ); IO + NO2 + M → IONO2 + M (5.286)
(JIONO2 ); IONO2 + hν → products (5.287)

(g7 ); I + HO2 → HI + O2 (5.288)
(g11 ); HI + OH → I + H2 O (5.289)
Finally, inter-halogen reactions such as
(g5 ); IO + ClO → Cl + I + O2 (5.290)
(g6 ); IO + BrO → Br + I + O2 (5.291)
which represent rate-limiting steps for ozone destruction, could be
significant in the lowermost stratosphere, even if the abundance of total
iodine is as low as a few tenths of a pptv (Solomon et al., 1994).
Observations by Wennberg et al. (1997), and Pundt et al. (1998) suggest
upper limits close to this value, suggesting a minor or negligible role.
However, measurements by Wittrock et al. (2000) provide evidence for
a greater contribution at high latitudes. Further, the formation and
photolysis of HOI is particularly effective for ozone destruction near

the tropopause, even for iodine abundances less than 0.5 pptv (Gilles
et al., 1997). On the other hand, formation of iodine reservoirs has been
highlighted as a factor in iodine’s impact on the atmosphere (Cox et al.,
1999). The reaction
(g16 ); IO + ClO → OIO + Cl (5.292)
is of particular importance, since photolysis of OIO to re-form odd

oxygen as the products IO + O could short-circuit ozone loss via this
reaction. However, some studies suggest that OIO photolysis produces
I + O2 (see Ingham et al., 2000) and the photolysis products of this
key reaction are not well understood. Tropospheric observations of IO
(Friess et al., 2001) and OIO (Allan et al., 2001) suggest that iodine
species may play a role in particle formation (Hoffmann et al., 2001).
Such processes could reduce iodine’s gas-phase role in ozone chemistry,
but could be relevant for formation of the stratospheric aerosol layer,
particularly if iodine can promote growth of sulfate clusters (Kulmala
et al., 2000).
5.6.3 The Halogen Families and Halogen Partitioning
Halogen families are variously defined to account for all chlorine,

bromine, iodine, and fluorine atoms attached to inorganic halogen
compounds. For example, the odd chlorine family Cly may be defined
as the sum of Cl + ClO + 2 Cl2 O2 + OClO + ClOO + 2 Cl2 + HCl
+ HOCl + ClONO2 + ClNO2 + BrCl. Similarly, odd bromine Bry is
provided by the sum of Br + BrO + HBr + HOBr + BrONO2 + BrCl
plus any other inorganic form of bromine-containing molecule or radical.
Note that a compound like BrCl (whose contribution is generally small
except during nighttime in activated air masses of the polar regions)
belongs to the odd chlorine as well as the odd bromine family. A similar
definition can be provided for odd iodine Iy and odd fluorine Fy .
A general continuity equation for the halogen families Xy (where X
can be Cl, Br, I or F) can be written as
∂ (Xy )
+ ∇ · φX + βx (Xy ) = PX (5.293)
∂t
where βx represents the loss coefficient (expressed in s−1 ) accounting for
physical removal (i.e., wet scavenging) of Xy , and PX is the production
rate (cm−3 s−1 ) associated with the photodecomposition of source gases.
An expression for this term is given by expression (5.200) in the case
of chlorine (Xy = Cly ). The flux divergence ∇·φX accounts for the
transport of all inorganic compounds of the Xy family.
In some cases, sub-families of the chemical family Xy (representing,
for example, the relatively reactive species) can be defined. In the case
of chlorine, for example, the most reactive compounds can be grouped
into the Clx family, defined as the entire odd chlorine family minus the
slow-reacting chlorine-containing species HCl and ClONO2 :
∂(Clx )
+ ∇ · φ(Clx ) + [d5 (CH4 ) + d6 (H2 ) + d7 (HO2 )
∂t
+d8 (H2 O2 ) + d9 (HNO3 ) + d10 (CH2 O)] (Cl) + [d31 (M)(NO2 )
+d46 (OH)] (ClO) = [d11 (OH) + d12 (O) + JHCl + h3 (ClONO2 )

+h4 (N2 O5 ) + h5 (HOCl)] (HCl) + [JClONO2 + d32 (O)
&
+h2 (H2 O) + h3 (HCl) + h3 (HBr) (ClONO2 )
(5.294)
The terms associated with heterogeneous reactions (h2 , h3 , h3 , h4 , h5 )
highlight the activation of chlorine when particles are present in the
stratosphere.
It is important to underline here the differences between the different
halogen families. In the case of bromine (and iodine), all inorganic
compounds are fast-reacting species and the concentration of the
individual members of the Bry (and Iy ) families can be derived through
a simple partitioning procedure assuming photoequilibrium conditions.
The abundances of Bry (and Iy ) are obtained by solving an equation of
type (5.293). In the case of fluorine, almost all inorganic fluorine is in
the form of HF, whose concentration can be obtained by solving (5.293)
with Xy = Fy ≈ HF. In the case of inorganic chlorine compounds, it
is generally necessary to solve for HCl and ClONO2 separately from
the other members of the odd-chlorine family, because of the relatively
long lifetime of hydrogen chloride and chlorine nitrate, especially in the
lower stratosphere (see Figure 5.55). Individual species belonging to
the Clx compounds can be partitioned assuming equilibrium conditions.
Retaining the most important terms in (5.240), (5.243), (5.245), and
(5.246) and adding terms involving Br and BrO,
(ClO) d2 (O3 )
= (5.295)
(Cl) d3 (O) + d4 (NO) + JClO
(HOCl) d33 (HO2 )
= (5.296)
(ClO) JHOCl + d34 (OH)
Figure 5.55. Photochemical lifetimes of Clx , Cl, ClO, HOCl, ClONO2 , HCl,
CF2 Cl2 , and CFCl3 , as well as the time constants appropriate to atmospheric
transport.
(Cl2 O2 ) d60 (M)(ClO)

= (5.297)
(ClO) JCl2 O2 + d61 (M)
(OClO) d47 (ClO) + e5a (BrO)
= (5.298)
(ClO) JOClO + e71 (Br)
In the upper stratosphere, where the photochemical lifetimes of
ClONO2 and HCl become shorter, we may evaluate the equilibrium
ratios between these chlorine containing compounds to serve as a rough
estimate of their relative abundances. In the absence of heterogeneous
processes, using Eq. (5.244),
(ClONO2 ) d31 (M)(NO2 )
= (5.299)
(ClO) JClONO2 + d32 (O)
and from Eq. (5.242),
(HCl) d5 (CH4 ) + d6 (H2 ) + d7 (HO2 ) + d8 (H2 O2 )
= (5.300)
(Cl) d10 (O) + d11 (OH) + JHCl
or, approximately
(HCl) d5 (CH4 ) + d7 (HO2 )

≈ (5.301)
(Cl) d11 (OH)
These ratios illustrate the importance of the interactions of chlorine

chemistry with oxygen, hydrogen and nitrogen species. Figure 5.56
provides an estimate of the partitioning associated with different halogen
families.
The above partitioning between halogen species can be dramatically
modified when particles (e.g., polar stratospheric clouds, sulfate
aerosols) are present in the atmosphere. For example, in air masses
processed by polar stratospheric clouds, ClONO2 and HCl are rapidly
converted by heterogeneous reactions to more reactive Clx compounds,
and specifically to Cl2 , HOCl, ClNO2 , BrCl. In the presence of solar
radiation, these species are photolyzed and high levels of ClO are
observed during daytime. ClO recombines with itself and is converted
into the Cl2 O2 dimer, which at night becomes the dominant inorganic
Model calculation of the partitioning of free radicals and total

halogen content for chlorine, bromine, and iodine at 8◦ N and 50◦ N in December.
From Solomon et al. (1994).
chlorine-containing compound. Thus, in PSC processed airmasses

during the springtime, the total abundance of inorganic chlorine is
approximately Cly ClO + 2 Cl2 O2 . ClO can participate in ozone-
destroying catalytic cycles, but high concentrations of ClO can only
be maintained if the level of nitrogen oxides is very low (to avoid the
formation of ClONO2 ). As indicated above, heterogeneous processes
tend to reduce substantially the NOx /NOy concentration ratio. With
ClO mixing ratios of typically 0.5-2.0 ppbv, in the Antarctic in late
winter/early spring, substantial amounts of ozone are depleted over a
time period of only a few weeks (see Chapter 6).
5.6.4 Observations of Halogen Compounds
The ClO radical has been directly measured since 1976 using in-
situ resonance fluorescence techniques and by Menzies (1979), Anderson
et al. (1980), Waters et al. (1981), De Zafra et al. (1989), and
Stachnik et al. (1992) using remote sensing microwave methods.
Global observations such as those provided by the MLS instrument
aboard UARS (Waters et al., 1993; Figure 5.57) have improved the
understanding of the behavior of ClO and the amplitude of its variations.
This constitutes an important element in the study of the chlorine-ozone
interaction, discussed in more detail in the next chapter.
Infrared spectra of ClONO2 provided the first estimates of the
abundance of this key compound (Murcray et al., 1977). Global
distributions have also been measured by the CLAES instrument on
board UARS (Roche et al., 1993; 1994; Figure 5.58) and by the CRISTA
and ATMOS instruments (Zander et al., 1996; Offermann et al., 1999;
Riese et al., 1999) on board the Space Shuttle. The largest mixing
ratios are found at high latitudes near 25-30 km altitude. The low
concentrations observed in the winter polar region below 20 km altitude
are due to heterogeneous conversion of ClONO2 .
HOCl has been observed by infrared thermal emission (Chance et al.,
1996) and by infrared solar absorption (Toon et al., 1992a). The column
abundance of OClO has been measured in Antarctica using visible
absorption methods, for example, by Solomon et al. (1987), and in the
Arctic by Wahner and Schiller (1992).
Hydrogen chloride (HCl) has been detected by spectroscopic absorp-
tion (Ackerman et al., 1976; Farmer et al., 1976; Raper et al., 1977;
Williams et al., 1976; Zander et al., 1996), by absorption radiometry
(Eyre and Roscoe, 1977), by in-situ filter sampling and chemical analysis
(Lazrus et al., 1975, 1976), by submillimeterwave heterodyne techniques
Figure 5.57. Meridional distribution of the ClO mixing ratio (ppbv) in July
based on the measurements by the UARS/MLS / instrument. Elevated levels in
the lower stratosphere in winter at high latitudes ((here south of 60◦ S between
15 and 28 km)) are associated with heterogeneous reactions on PSC particles.
(Courtesy of W. Randel, NCAR).
Figure 5.58. Meridional distribution of the ClONO2 mixing ratio (ppbv) in

July based on the measurements by the UARS/CLAES instrument (Courtesy
of W. Randel, NCAR).
(Stacknik et al., 1992), and by in-situ tunable diode laser spectrometry

(May and Webster, 1989). In the 1990s, global distributions of this
gas were obtained by the HALOE infrared instrument aboard UARS
(Russell et al., 1993; Figure 5.59). The distribution and variability of
the total column abundances of HCl have been reported, for example,
by Mankin and Coffey (1983).
Observations of inorganic bromine, and specifically BrO, have also
been reported (see e.g., Brune et al., 1989; Carroll et al., 1989; Toohey
et al., 1990; Avallone et al., 1995; Fish et al., 1997). Figure 5.60 shows
some measured mixing ratios of BrO at mid-latitudes and in the polar
regions. Values range from typically a few pptv at the tropopause to
10-15 pptv in the middle and upper troposphere. The stratospheric
column abundance of BrO has been measured at mid-latitudes by Arpag
et al. (1994), in Antarctica by Solomon et al. (1989b), and in the Arctic
by Wahner and Schiller (1992).
Observations of bromine-bearing gases show large variations in the
abundance of lower stratospheric Bry , perhaps for both physical reasons
(i.e., true variability) and instrumental uncertainty. BrO observations
using optical absorption methods (Pfeilsticker et al., 2000) provide a
new approach that supports the view that the lower stratospheric Bry
may often be considerably larger than suggested by consideration only
of halon and CH3 Br sources. Short-lived molecules such as CH2 Br2 and
even CHBr3 are possible additional sources (see Schauffler et al., 1993)
whose origins are largely natural but which may transport bromine to
the upper troposphere or even to the stratosphere in convective towers
(Ko et al., 1997; Dvortsov et al., 1999).
The impact of fluorine to the stratosphere is reflected in observed
abundances of HF, its sole significant stratospheric reservoir. HF is
observed mainly by infrared absorption and long records are available
from ground-based sites (e.g., Zander, 1981) which confirm its primary
origin in anthropogenic sources. Satellite observations from UARS (see
Figure 3.26) map its global distribution (Russell et al., 1993), and its
long lifetime makes HF a useful tracer for atmospheric motions (Toon
et al., 1992b).
5.7 Stratospheric Aerosols and Clouds
The existence of aerosol surfaces in the stratosphere has been known

for many decades, but it was not until the discovery of the Antarctic
ozone hole that their role in surface chemistry was recognized. Here the
stratospheric sulfate layer and polar stratospheric clouds will be
Figure 5.59. Meridional distribution of the HCl mixing ratio (ppbv) in January
based on the measurements by the UARS/HALOE instrument (Courtesy of W.
Randel, NCAR).
Figure 5.60. Observed stratospheric BrO mixing ratios (pptv) at mid- and high
latitudes (Harder et al., 1998).
described. Their critical roles in both mid-latitude and polar ozone

depletion are discussed in Chapter 6.
5.7.1 The Sulfate Layer
Atmospheric sulfur is released at the Earth’s surface by a number

of processes. Some are natural: hydrogen sulfide, H2 S, for example,
is a by-product of decomposition of organic matter; dimethyl sulfide
(DMS or (CH3 )2 S), carbonyl sulfide (COS), and CS2 are produced
by the oceans. DMS is produced from dimethyl sulfide propriothetin
(DMSP) by certain species of phytoplankton. COS results from the
action of sunlight on dissolved organic compounds. Other sources are
related to anthropogenic activities. Industry produces large amounts of
sulfur dioxide, particularly through combustion of coal (especially coal
fired electric power plants), as well as smaller quantities of H2 S, CS2
and COS. Sulfur oxides and sulfuric acid vapors are also produced in
combustion engines. Finally, volcanoes are responsible for the injection
of appreciable amounts of sulfur into the atmosphere (SO2 , H2 S), either
continuously through venting or episodically during volcanic eruptions.
Although sulfur is always present in the atmosphere, its density
can vary considerably as a result of geophysical or anthropogenic
perturbations. Urban regions exhibit large densities of sulfur-containing
compounds, while most of the sulfur present above the ocean is of
biogenic origin. The density of SO2 in polluted air is quite variable,
and depends on meteorological conditions as well as the source of the
air masses, because the tropospheric lifetime of SO2 is several days.
In clean air, the mole fraction of SO2 at low altitudes ranges from
typically 10 to 500 pptv (Friend, 1973; Breeding et al., 1973; Graedel
et al., 1974; Maroulis et al., 1980; Nguyen et al., 1983; Herman and
Jaeschke, 1984). Values measured over the Pacific Ocean are close
to 90±60 pptv in the northern hemisphere and 60±20 pptv in the
southern hemisphere (Maroulis et al., 1980). Measurements made in
the remote marine atmosphere of the western Atlantic Ocean (Thornton
et al., 1987) resulted in median SO2 mole fractions of 65 pptv in the
free troposphere, compared with a value of approximately 1 ppbv in
the boundary layer. Measurements by Bandy et al. (1993a) show that
the mole fraction of SO2 at 4 km over the coast of Brazil ranges from
15 to 60 pptv. Values close to 700 pptv have, for example, been
reported by Boatman et al. (1989) in the central United States. Near
the tropopause, abundances of about 50 pptv are characteristically
observed over continental areas (Jaeschke et al., 1976; Meixner, 1984),
indicating that less than 10% of the molecules typically reach the lower
stratosphere. Other sulfur compounds such as H2 S and (CH3 )2 S are
short-lived and are therefore rapidly destroyed in the troposphere. On
the other hand, COS is very stable and its mole fraction is of the order
of 500 pptv throughout the troposphere (Hanst et al., 1975; Sandalls
and Penkett, 1977; Maroulis et al., 1977; Friedl et al., 1991; Weiss
et al., 1995), even at tropopause levels (Inn et al., 1979; Mankin et al.,
1979). The lifetime of tropospheric COS is long, of the order of a year,
and only above about 15 km does the mole fraction begin to decrease,
reaching about 15 pptv at 30 km. CS2 possesses an intermediate
lifetime, as evidenced by an observed tropospheric variability of about
50% (Sandalls and Penkett, 1977). Its surface mixing ratio is about
100 pptv over polluted areas, but can be as low as 1-5 pptv in remote
regions (Bandy et al., 1993b). Therefore, middle atmospheric sulfur
is linked in part to transport of COS across the tropopause as first
suggested by Crutzen in 1976. COS is primarily produced by biologic,
volcanic and anthropogenic sources. Thus, a potential increase in COS,
particularly as a result of the use of large quantities of coal, could
alter the composition of the atmosphere, increasing the concentration
of sulfur aerosols and modifying the terrestrial radiative balance (see
Crutzen, 1976; Turco et al., 1980). No trend in COS has, however, been
observed (Bandy et al., 1992; Rinsland et al., 1992). Large hemispheric
gradients (15-25%), with larger concentrations measured in the Northern
hemisphere, have been reported over the North Atlantic (Bingemer et al.,
1990; Johnson et al., 1993). Table 5.10 provides an estimate of the
global budget of sulfur in the atmosphere and shows that the largest
gaseous emissions (approximately 90 Tg/yr) are of anthropogenic origin.
However, only a fraction of this reaches the stratosphere. This budget
can, however, be perturbed substantially by large volcanic eruptions.
It is estimated that after the eruption of El Chichon ´ (Mexico) in 1982
nearly 10 Tg of sulfur were injected into the stratosphere. The eruption
of Mt. Pinatubo (the Philippines) in 1991 injected approximately 20 Tg S
above the tropopause.
When COS reaches the stratosphere, photolysis occurs:
(JCOS ); COS + hν → CO + S (5.302)
producing carbon monoxide and atomic sulfur. Sulfur reacts immedi-

ately with molecular oxygen:
(s1 ); S + O2 → SO + O (5.303)
Table 5.10 Global Budget of Sulfur (in Tg S/yr)
Sources
Compounds Flux
Volcanoes SO2 , H2 S, COS 7-10

Vegetation and Soils H2 S, DMS, COS, CS2 , DMDS 0.4-1.2
Biomass Burning SO2 , COS, H2 S 2-4
Oceans DMS, COS, CS2 , H2 S 10-40
Anthropogenic SO2 , sulfates 88-92
Sinks
Dry Deposition SO2 , sulfates 50-75

Wet Deposition SO2 , sulfates 50-75
to produce sulfur monoxide. This species can also be formed by reaction

of COS and atomic oxygen:
COS + O → CO + SO (5.304)
which, like the slow reaction
COS + OH → products (5.305)
represents an additional destruction mechanism for COS, but these are
considerably less important than the direct photolysis process. Sulfur
monoxide is in rapid equilibrium with SO2 . Its destruction is dominated
by
(s2 ); SO + O2 → SO2 + O (5.306)
which is more important (Black et al., 1982a,b) than
(s3 ); SO + O3 → SO2 + O2 (5.307)
and
(s4 ); SO + NO2 → SO2 + NO (5.308)
Sulfur dioxide is relatively stable and is thus present in fairly large
abundance in the stratosphere. It can be photodissociated by radiation
in the 200-230 nm region, where its absorption cross section is of the
order of 10−18 to 10−17 cm2 .
(JSO2 ); SO2 + hν → SO + O (5.309)
or converted to HSO3 by the following process:
(s5 ); SO2 + OH + M → HSO3 + M (5.310)
The HSO3 intermediate is probably destroyed by O2 (Calvert and

Stockwell, 1983; McKeen et al., 1984):
(s6 ); HSO3 + O2 → HO2 + SO3 (5.311)
Direct conversion of SO2 to SO3 by reactions such as
(s7 ); SO2 + HO2 → SO3 + OH (5.312)
(s8 ); SO2 + O + M → SO3 + M (5.313)

can be neglected.
Sulfur trioxide can combine with water vapor to form sulfuric acid:
(s9 ); SO3 + H2 O → H2 SO4 (5.314)
The acid can be photodissociated
(JH2 SO4 ); H2 SO4 + hν → HSO3 + OH (5.315)
or can precipitate in water droplets, (as can SO2 and HSO3 ) yielding
sulfate aerosols. A complete analysis of sulfur compounds must therefore
include heterogeneous processes.
It should be noted that the reaction sequences which convert SO2
to sulfate do not modify the HOx budget (McKeen et al., 1984). The
intrusion of large amounts of SO2 in the stratosphere during volcanic
eruptions could enhance the short-term amount of ozone in the middle
atmosphere by direct chemical effects (Crutzen and Schmailzl, 1983;
Bekki et al., 1993) through the following mechanism:
SO2 + hν → SO + O
SO + O2 → SO2 + O
2(O + O2 + M) → 2(O3 + M)
Net: 3 O2 → 2 O3
(5.316)
In the longer-term, and with contemporary loads of chlorine in the
stratosphere, the presence of volcanic aerosols causes a reduction in
the ozone abundance resulting from the large increase in aerosol surface
area available for heterogeneous reactions (Hofmann and Solomon, 1989;
Brasseur and Granier, 1992; Michelangeli et al., 1992; Pitari and Rizzi,
1993; Tie et al., 1994). These effects will be further discussed in
Chapter 6.
Figure 5.61 shows a schematic diagram of the chemistry of sulfur
and Figure 5.62 presents calculated distributions of the gaseous sulfur
compounds from the model by Whitten et al. (1980). Some observations
are also shown for comparison. This model shows that below 30 km
Figure 5.61. Primary sulfur reactions in the middle atmosphere.
Vertical distribution of the primary sulfur compounds, as

calculated by Turco et al. (1979).
during volcanically quiescent periods, sulfur is mostly in the form of

COS, while at higher altitude it is mostly sulfuric acid.
Analyses of observations of negative ion densities can be used to derive
the abundance of H2 SO4 , as shown, for example, by Arnold and Fabian
(1980), Arijs et al. (1981), and Ingels et al. (1987). These observations
confirm that the gaseous stratospheric H2 SO4 concentration is largely
determined by the equilibrium between the aerosol and vapor phases.
The gas and aerosol chemistry of sulfur are discussed in detail in the
excellent review by Turco et al., (1982).
The H2 SO4 molecule plays a major role in the microphysics of
nucleation and condensation, and is therefore essential to the formation
and growth of aerosols. These sub-micron liquid particles also
coagulate, evaporate, or sediment out under the force of gravity. The
exact mechanisms of nucleation, especially the initial processes, are
poorly known, but it is presently thought that heterogeneous and
heteromolecular nucleation of the H2 SO4 -H2 O system around fine
Aitken particles (condensation nuclei of radius less than 0.1 µm)
is predominant. These Aitken particles are probably transported
from the troposphere and perhaps also produced by aircraft engines.
Other processes may also be important in aerosol formation, such as
homogeneous nucleation (nucleation of polar molecules such as H2 O,
HNO3 , H2 SO4 around ions). Evidence for such a process is implied
by the formation of sulfate radicals and complex neutral species during
ion-ion recombination.
After nucleation, the droplets grow through heteromolecular conden-
sation, provided that the partial pressure of H2 SO4 in the surrounding
air is greater than the vapor pressure above the droplet. Otherwise, the
H2 SO4 will evaporate and the particles will shrink. The H2 SO4 /H2 O
concentration ratio inside an aerosol particle varies significantly with
temperature. Under stratospheric conditions, the H2 SO4 concentration
inside an aerosol particle is typically 80% wt at 240 K and 40% wt
at 190 K. The concentration of stratospheric sulfate aerosols (SSA) is
typically 10-20 particles cm−3 . Figure 5.63 presents a schematic diagram
of the physical and chemical properties pertaining to stratospheric
aerosol. It should be noted that although sulfur enters the stratosphere
in gaseous form (essentially COS, CS2 , and SO2 ), it leaves the middle
atmosphere in the form of particles (sulfate) transported downward by
the stratospheric circulation and by gravitational settling, followed by
precipitation in tropospheric clouds (see Turco et al., 1982).
An important aerosol property is the size distribution, which is usually
expressed in an empirical form (power law, exponential, or log normal
function) whose parameters are derived from observations. This size
distribution depends on the nature of the particles and varies with
Figure 5.63. Schematic diagram of the physical and chemical processes related
to the formation, growth, transport, and destruction of atmospheric aerosols: C
= coagulation, Ch = chemistry, D = diffusion, E = evaporation, Em = emission,
G = condensation and growth, I = injection, N = nucleation, P = photolysis,
S = sedimentation, W = washout and rainout. From Turco et al. (1979).
altitude. As shown in Figure 5.64, for example, the concentration

of condensation nuclei with radii smaller than 0.01 µm decreases
monotoneously with increasing height above the tropopause, while
the concentration of the larger particles (radius greater than 0.15µm)
exhibits a broad peak in the lower stratosphere. The existence of this
aerosol layer (called the Junge layer)
r approximately 5-7 km above the
tropopause indicates that these particles grow in-situ in the stratosphere
by sulfuric acid formation and by coalescence. Stratospheric sulfate
aerosols remain generally liquid, although nonspherical (solid) sulfuric
acid hydrate particles have occasionally been observed. They serve as
condensation nuclei for the formation of polar stratospheric clouds in
the Arctic and Antarctic (see Section 5.8).
Global climatologies of stratospheric aerosols have been produced
from observations by satellite instruments. Hitchman et al. (1994), for
example, have provided meridional distributions of measured extinction
ratios based on 9 years of data from the Stratospheric Aerosol and
Gas Experiment (SAGE I and II) and the Stratospheric Aerosol
Figure 5.64. Observed particle size distribution with altitude. From Hofmann
and Rosen (1981).
Measurement (SAM II) instruments. This climatology has provided

insight into the general circulation of the stratosphere, and specifically
on the circulation in the tropics associated with the quasi-biennial
oscillation. Under background (non-volcanic) situations, the surface
area density provided by aerosol particles is typically 0.5-1.0 µm2 cm−3
in the lower stratosphere and less than 0.2 µm2 cm−3 above 25 km
altitude. After large volcanic eruptions such as those of El Chichón
(Mexico) in April, 1982, and of Mt. Pinatubo (the Philippines) in June,
1991, during which large quantities (1-20 Tg) of SO2 were injected into
the stratosphere, the aerosol surface area increases dramatically (e.g.,
1-2 orders of magnitude after the eruption of Mt. Pinatubo). The decay
time for such perturbation is typically 2-3 years. Heterogeneous reaction
rates are enhanced proportionally, with substantial perturbations in
the chemistry of the lower stratosphere (see e.g., Hoffmann and
Solomon, 1989; Brasseur and Granier, 1992). Global climatologies
of stratospheric aerosol surface area density have been established by
Hitchman et al. (1994) as well as by Thomason et al. (1997) (in this
case on the basis of multiwavelength aerosol extinction measurements
from space, see Plate 12). The meridional distribution of the aerosol
surface area density (annual mean) derived by Thomason et al. (1997)
is shown in Figure 5.65. Plate 8 shows the evolution of the surface area
density at 20.5 km altitude from 1979 to 1995. The spread with latitude
Figure 5.65. Median meridional distribution of the sulfate aerosol surface area
density (µm2 cm−3 ) for the period 1985-1994, based on observations by the
SAGE II satellite. From Thomason et al., 1997.
of the aerosol perturbation following the eruption of Mt. Pinatubo is

visible on this figure. Additional studies of aerosol changes following
the Mt. Pinatubo eruption have also been performed by Langford
et al. (1995), Russell et al. (1996), and others.
5.7.2 Polar Stratospheric Clouds
For more than a century, the presence of thin clouds has been
reported at high altitudes in the polar regions. Because of their colorful
appearances, especially at low solar elevation, these clouds have been
named “mother-of-pearl” or nacreous clouds, or–more recently in the
context of polar ozone research–polar stratospheric clouds (PSCs).
While PSCs form in the winter lower polar stratosphere, there is a
second class of high altitude clouds that are formed during summertime
at mesopause levels, when the temperature observed at these heights
drops below approximately 150 K (see Plate 13). The chemical role
of these mesospheric clouds is not yet well understood, but it has
been suggested that the frequency of appearance of such clouds could
increase in the future in response to human-induced cooling of the
middle atmosphere (associated with enhanced levels of CO2 and other
“greenhouse gases” in the atmosphere).
PSCs (see Plate 14) are observed during winter in both the Antarctic
and Arctic polar stratosphere (Stanford and Davis, 1974; McCormick
et al., 1982), typically from 12 to 26 km altitude (see e.g., Poole and
McCormick, 1988; Browell et al., 1990; Gobbi and Adriani, 1993; Godin
et al., 1994). The presence of these clouds is associated with cold
airmasses (T < 200 K). It was originally assumed that these PSCs
were composed mainly of water ice (Steele et al., 1983). Such particles
are often observed in the Antarctic (e.g., Poole et al., 1988; Fahey
et al., 1990) and only occasionally in the Arctic (e.g., Carslaw et al.,
1998a,b). Their diameter is typically 1-30 µm and their concentration
can vary by several orders of magnitude. Because of their relatively large
size, ice particles sediment in the atmosphere by typically 100 m per
hour (assuming a diameter of 20 µm), leading to regular dehydration
of the lower Antarctic stratosphere and very occasionally also in the
Arctic. Such particles, which have been observed, for example, by
lidar, can be identified by their optical properties (high depolarization
corresponding to nonspherical and therefore solid particles). Early
lidar observations made in Antarctica by Poole and McCormick (1988)
have shown that, in addition to ice particles (called type II PSCs),
optically thinner PSCs were also present at temperatures up to 5-6 K
above the ice frost point. Crutzen and Arnold (1986) as well as Toon
et al. (1986) suggested that these small particles could be formed by
uptake of nitric acid and water onto pre-existing frozen sulfate aerosols
(sulfuric acid tetrahydrate or SAT). The particles, first called Type I
PSCs, have been frequently observed in the Antarctic (Fahey et al.,
1989; Pueschel et al., 1989; Goodman et al., 1989; 1997) and in the
Arctic (Larsen et al., 1996; 1997; Rosen et al., 1997; Enell et al., 1999).
Fahey et al. (1989) observed that these particles contained considerable
amounts of reactive nitrogen, which could possibly be identified as
crystals of nitric acid trihydrates (HNO3 · 3 H2 O or NAT), which are
the most thermodynamically stable solid hydrates under stratospheric
conditions (Hanson and Mauersberger, 1988). Final evidence for the
existence of NAT in the winter polar stratosphere has been provided
by balloon-borne mass spectrometric measurements (Voigt et al., 2000).
The presence of nitric acid monohydrate (NAM) (McElroy et al., 1986)
and dihydrate (NAD) (Tolbert et al., 1992; Worsnop et al., 1993) has
also been considered as a possibility. NAT particles are characterized
by low backscatter ratios and high depolarization, and by atmospheric
concentrations of generally less than 0.1 particles cm−3 . The diameter
of these NAT particles ranges typically from 1 to 5 µm, and the
volume densities of these particles (typically 1 µm3 cm−3 ) inferred
from observations is considerably smaller than for ice particles. Fahey
et al. (2001) have observed in the Arctic very large HNO3 -containing
particles (probably NAT or NAD). With diameters of 10-20 µm, such

particles sediment rapidly (fall speed of 1.5 km day−1 for 14 µm spherical
particles at 18-19 km), providing an efficient mechanism to denitrify the
lower stratosphere.
Recent laboratory studies and field observations have led to a new
conceptual picture of the composition and chemistry of PSCs (see
Table 5.11). For example, the mechanisms advanced earlier to describe
the formation of solid NAT particles appears unlikely to occur in the
atmosphere. Lidar measurements by Browell et al. (1990) revealed very
low depolarization of the backscatter signal, suggesting the presence of
liquid particles (Toon et al., 1990). Observations by Dye et al. (1992) and
their analysis by Carslaw et al. (1994), Drdla et al. (1994) and Tabazadeh
et al. (1994) have shown that, in addition to NAT particles (in Lidar
terminology also called type Ia particles), liquid supercooled ternary
solutions H2 SO4 /HNO3 /H2 O (STS, also called type Ib particles) are
often present in the winter stratosphere. These spherical particles have
been observed in both polar regions (see e.g., Adriani et al., 1995; Hamill
Table 5.11 Particles in the Stratosphere
Particle type Composition/ Phase Approximate Observational

name temperature evidence
threshold
Stratospheric H2 SO4 /H2 O small liquid T < 261 K strong

sulfate droplets
aerosol
Sulfuric acid H2 SO4 ·4 H2 O small solid T < 213 K weak
tetrahydrate (SAT) crystals
Nitric acid HNO3 · 3 H2 O solid T < 196 K confirmed
trihydrate (NAT) crystals
Nitric acid HNO3 · 2 H2 O solid T < 193 K vague
dihydrate (NAD) crystals
Supercooled H2 SO4 /HNO3 /H2 O liquid T < 192 K strong
ternary (STS) droplets
solution
Ice H2 O large solid T < 189 K strong
crystals
Amorphous H2 SO4 /HNO3 /H2 O solid T < 170 K not
solution particles confirmed
et al., 1996; Beyer et al., 1997; David et al., 1997; Rosen et al., 1997;
Larsen et al., 1997; Carslaw et al., 1998a,b; Steele et al., 1999). They
seem to form a few degrees below the stability temperature threshold
for NAT. In well developed clouds, STS are submicron particles at
concentrations similar to ambient sulfate aerosols, whereby a significant
increase in particle volume is observed 3-4 K below the NAT stability
temperature.
The formation of PSC particles, as the temperature decreases, can
no longer be represented by a “step function” as the stability point
for NAT is reached, but should be regarded as a gradual transition
from aqueous sulfuric acid aerosols to liquid PSC particles. Carslaw
et al. (1997) have provided a detailed review of the thermodynamics of
ternary solutions. As shown by Figure 5.66, in cold environments (T
< 195 K), significant quantities of nitric acid are absorbed by sulfate
aerosol particles, forming the H2 SO4 /HNO3 /H2 O ternary solution. As
the temperature decreases even further (T < 191 K), more HNO3 and
H2 O are absorbed from the gas phase, leading to the formation of a
Figure 5.66. Relative mass concentration of H2 SO4 , HNO3 and HCl dissolved
within a liquid aerosol particle as a function of temperature. Note the
change from linear to logarithmic scales below 2 wt%. Adapted from Carslaw
et al. (1994).
Figure 5.67. Total particle volume density observed by Dye et al. (1992) from
the high altitude ER-2 aircraft on 24 January 1989, north of Stavanger, Norway,
at 19 km altitude. As the temperature decreases, the aerosol particles initially
grow as a result of H2 O uptake by the H2 SO4 /H2 O binary solution. At a
temperature below 193 K, the volume increase is much more rapid, and suggests
combined uptake of H2 O and HNO3 by a H2 SO4 /HNO3 /H2 O ternary solution.
The data points are consistent with mixing ratios of HNO3 ranging between
5 and 15 ppbv. The light solid line corresponding to the H2 SO4 /H2 O binary
aerosol is calculated for a water vapor mixing ratio of 5 ppmv. The dash line
corresponds to NAT particle growth without nucleation barrier. From Carslaw
et al., (1994).
quasi-binary HNO3 /H2 O solution. At very low temperatures (T < 188

K), HCl becomes increasingly soluble in the liquid particle. Shown in
Figure 5.67 is the growth of different types of particles (background
sulfate aerosols, supercooled ternary solution, NAT) resulting from the
uptake of water and nitric acid as the temperature decreases from 210
to 185 K. The interpretation of cloud observations is often complicated
due to the fact that different types of particles can co-exist in the same
cloud. It seems that type Ib particles (liquid STS) are often observed in
air parcels that have encountered temperatures below the SAT melting
point during less than a day, while type Ia particles (NAT) are found in
air parcels exposed to temperatures below the NAT stability point for
several days (see e.g., Tabazadeh et al., 1995; 1996; Toon et al., 2000).
Although the thermodynamics of liquid solutions is rather well
understood, the circumstances under which solid PSCs are formed
remain unresolved (Peter, 1997; Tolbert and Toon, 2001). It seems
unlikely that NAT forms through direct nucleation on SAT because SAT
particles become unstable and melt in the presence of gas-phase HNO3
(Koop and Carslaw, 1996). NAT cannot easily or quickly be produced

by direct freezing of the aqueous ternary solutions except, perhaps,
if the temperature drops below the ice frost point (189 K) and water
precipitates from the solution (Koop et al., 1995). Rapid temperature
fluctuations (associated, for example, with lee waves downstream from
mountain ridges) may lead to non-equilibrium conditions under which
the ternary solution could freeze (Meilinger et al., 1995; Tsias et al.,
1997). Tabazadeh et al. (1995) have suggested that, under stratospheric
conditions, amorphous solid binary or ternary solutions might form and
be precursors to the formation of NAT-type crystals. As stressed,
however, by Toon et al. (2000), it is difficult to establish if such
particles have a chemical composition distinct from NAT or if they are
just formed with a different nucleation rate. One study suggests that
freezing may occur if particles remain cold for more than a day (Salcedo
et al., 2001). MacKenzie et al. (1995) and Drdla (1994) have suggested
that trace impurities in particles may facilitate the freezing process.
Further details on the microphysics and heterogeneous chemistry of polar
stratospheric clouds can be found in the reviews by Peter (1997) and
Carslaw et al. (1997). Further discussion on the role of the clouds in
the observed ozone depletion in the Arctic and Antarctic is presented in
Chapter 6.
5.8 Generalized Ozone Balance
The analysis presented above permits us to establish a generalized

ozone and atomic oxygen balance, and to correct the equations derived
previously by considering only a pure oxygen atmosphere. Detailed
reviews of the ozone balance are provided by Johnston and Podolske
(1978), and Crutzen and Schmailzl (1983). A general algorithm for the
determination of all significant pathways involved in the ozone budget
is provided by Lehmann (2004).
Above 40 km, the catalytic cycles involving hydrogen free radicals
represent a very rapid loss process for odd oxygen (Nicolet, 1970):
OH + O → H + O2
H + O2 + M → HO2 + M
HO2 + O → OH + O2
Net: O + O → O2
(5.317)
or
OH + O → H + O2
H + O3 → OH + O2
Net: O + O3 → 2 O2
(5.318)
With some algebraic manipulation, the chemistry of these hydrogen
catalyzed reactions can be included in the equation for the photochem-
ical equilibrium ozone abundance between about 50 and 80 km to yield
the following simple formulas at various altitudes (Allen et al., 1984):
2 1 2 1 4
(JO2 ) 3 (k4a a17 ) 3 [k2 (M)] 3 (N2 ) 3 (O2 ) 3
(O3 )50 km ≈ 2 1 1 (5.319)
(JO3 ) 3 (J3∗ a1∗ a5 a7 ) 3 (H2 O) 3
1
JO2 (a17 ) 2 k2 (M)(O2 )2
(O3 )70 km ≈ 1 1 (5.320)
JO3 (JH2 O a5 a7 ) 2 (H2 O) 2
(JO2 )2 a23 k2 (M)(O2 )2

(O3 )80 km ≈ (5.321)
a2 JO2
JH2 O a7 a1 (H2 O)(M) JO3 + a1 (M)
The important role of odd hydrogen catalyzed destruction of odd
oxygen is illustrated by comparing these expressions to Eq. (5.43),
which presents the analytic expression for the equilibrium ozone density
assuming only pure oxygen chemistry.
The hydrogen radicals also participate in the following cycle:
OH + O3 → HO2 + O2
HO2 + O3 → OH + 2O2
Net: 2O3 → 3 O2
(5.322)
This process dominates the odd oxygen destruction near the tropopause
because it is the most effective catalytic cycle involving only ozone as the
reactive odd oxygen species. Most of the other HOx , NOx , and ClOx
cycles (see below) also require reaction with atomic oxygen, which is
present only in very small amounts at low altitudes. For example, the
following cycles are important in the middle and upper stratosphere,
and in the mesosphere
OH + O3 → HO2 + O2
HO2 + O → OH + O2
Net: O + O3 → 2 O2
(5.323)
or
OH + O → H + O2
H + O2 + M → HO2 + M
HO2 + O3 → OH + 2O2
Net: O + O3 → 2O2
(5.324a)
or
OH + O → H + O2
H + O3 → OH + O2
Net: O + O3 → 2O2
(5.324b)
The nitrogen compounds can also catalyze the destruction of odd
oxygen. This cycle is most efficient near 35-45 km (Crutzen, 1970;
Johnston, 1971):
NO + O3 → NO2 + O2
NO2 + O → NO + O2
Net: O + O3 → 2O2
(5.325)
This cycle must be compared with another cycle that has no effect on
odd oxygen:
NO + O3 → NO2 + O2
NO2 + hν → NO + O
O + O2 + M → O3 + M
Net: Null
(5.326)
Thus, the efficiency of the NOx catalyzed destruction of odd oxygen
depends on the competition between photolysis and reaction with atomic
oxygen for the NO2 radical; the rate-limiting step for the destruction of
odd oxygen is therefore provided by the NO2 + O reaction. Other
secondary cycles should also be noted:
NO + O3 → NO2 + O2
NO2 + O3 → NO3 + O2
NO3 + hν → NO + O2
Net: 2 O3 → 3 O2
(5.327)
and
NO + O3 → NO2 + O2
NO2 + O3 → NO3 + O2
NO3 + NO2 + M → N2 O5 + M
N2 O5 + hν → NO3 + NO2
NO3 + hν → NO + O2
Net: 2 O3 → 3 O2
(5.328)
Note that these cycles result in ozone destruction only when the
photolysis of NO3 leads to NO + O2 rather than NO2 + O. In the
latter case, the effect of these cycles is also null.
The catalytic destruction of odd oxygen by chlorine species must also
be considered, particularly in the upper stratosphere where this cycle
is quite effective (Stolarski and Cicerone, 1974; Molina and Rowland,
1974a):
Cl + O3 → ClO + O2
ClO + O → Cl + O2
Net: O + O3 → 2 O2
(5.329)
A similar cycle involves bromine compounds:
Br + O3 → BrO + O2
BrO + O → Br + O2
Net: O + O3 → 2 O2
(5.330)
Below 30 km except in regions where heterogeneous reactions are
proceeding rapidly, the densities of Cl and ClO are generally much
reduced due to the formation of slow-reacting chlorine reservoirs
including HCl and ClONO2 . Formation of these reservoirs limits the
ozone loss in the lower stratosphere, and the effect of the direct chlorine
catalyzed destruction on odd oxygen is greatly reduced. However, the

following cycle is generally important near 20 km:
Cl + O3 → ClO + O2
ClO + HO2 → HOCl + O2
HOCl + hν → Cl + OH
OH + O3 → HO2 + O2
Net: 2O3 → 3O2

(5.331)
Here also similar cycles with bromine must be considered; e.g.,
Br + O3 → BrO + O2
BrO + HO2 → HOBr + O2
HOBr + hν → Br + OH
OH + O3 → HO2 + O2
Net: 2 O3 → 3 O2
(5.332)
When high concentrations (∼ 1 ppbv) of ClO are present in the lower
stratosphere (e.g., in air masses wherein the reservoir’s HCl and ClONO2
have been processed by polar stratospheric clouds), additional cycles
must be considered (see Figure 5.68); one of them involves the formation
of the Cl2 O2 dimer which dominates the destruction of polar ozone (e.g.,
Antarctic ozone hole):
ClO + ClO + M → Cl2 O2 + M
Cl2 O2 + hν → Cl + ClOO
ClOO + M → Cl + O2 + M
2 (Cl + O3 → ClO + O2 )
Net: 2O3 → 3O2

(5.333)
The formation rate of the ClO dimer and the photolysis rate of Cl2 O2
are critical pieces of information to calculate the ozone loss in the polar
regions (Avallone and Toohey, 2001). A second cycle, which introduces
an interaction between the chlorine and bromine families also contributes
to the destruction of ozone in polar regions, as well as at middle latitudes:
Figure 5.68. Reaction scheme responsible for rapid ozone destruction in polar
regions when large amounts of chlorine and bromine are activated by polar
stratospheric clouds. Concentrations (cm−3 ) and fluxes (cm−3 s−1 ) associated
with chemical reactions are calculated for Antarctic springtime conditions
(lower stratosphere). After Zellner (1999).
ClO + BrO → Br + Cl + O2
Br + O3 → BrO + O2
Cl + O3 → ClO + O2
Net: 2O3 → 3O2

(5.334)
where the BrO+ClO → Br+Cl+O2 reaction, as written here, represents
one of the 2 possible processes
ClO + BrO → Br + ClOO
ClOO + M → Cl + O2 + M
or
ClO + BrO → BrCl + O2
BrCl + hν → Br + Cl
A similar possible cycle involves an interaction between the bromine and
iodine families:
BrO + IO → Br + I + O2
I + O3 → IO + O2
Br + O3 → ClO + O2
Net: 2O3 → 3O2

(5.335)
The importance of chlorine activation in the Antarctic region during
winter and spring and the associated ozone destruction by cycles as
(5.333) and (5.334) can be seen in Figure 5.69. This diagram shows
the chemical changes affecting air masses in the southern polar region,
when the temperature becomes so low that polar stratospheric clouds
are formed (see Section 5.7.2). Heterogeneous reactions (e.g., (5.222)
and (5.223)) occurring on the surface of cloud particles tend to convert
chlorine reservoirs into more reactive compounds. These are photolyzed
as the Sun returns in the polar region, and the levels of ClO that are
produced are sufficiently high to destroy large amounts of ozone in a few
weeks. The concentrations of ClO remain high because nitrogen oxides
are converted into nitric acid by heterogenous process (see Reaction
(5.152)) and therefore cannot react with ClO to form ClONO2 . Note
also the denitrification and dehydration of the lower polar stratosphere
through sedimentation of the largest cloud particles. As the temperature
gradually increases duirng springtime, and the polar stratospheric clouds
disappear, nitrogen oxides produced by photolysis of nitric acid survive
a sufficiently long time to rapidly convert ClO into ClONO2 . Chlorine
atoms react also with methane to form HCl. At that time, the fast
catalytic cycles (5.333) and (5.334) cease to be efficient mechanisms to
deplete ozone. A more detailed description of the processes affecting
polar ozone is given in Chapter 6.
Other reactions affecting ozone in the lower stratosphere must be
considered. The following cycle is significant when the photolysis
products of ClONO2 are Cl and NO3 ,
ClO + NO2 → ClONO2

ClONO2 + hν → Cl + NO3
Cl + O3 → ClO + O2
NO3 + hν → NO + O2
NO + O3 → NO2 + O2
Net: 2 O3 → 3 O2
(5.336)
Figure 5.69. Evolution with time of temperature ((upper left panel) and chemical
composition (other panels) of an airmass situated at 82◦ S and 20 km altitude,
highlighting the activation of halogen compounds in the presence of polar
stratospheric clouds. As temperature drops below the threshold for PSC I
formation, initial signs of denitrification are seen (June). Dehydration occurs
about a month later as the temperature reaches the threshold for water ice
formation. The reservoir species (HCl and ClONO2 ) are converted to reactive
chlorine species by heterogeneous reactions on the surface of PSC particles.
ClOx is enhanced throughout the polar night and ozone is destroyed as solar
radiation becomes available in the spring. The partitioning between the ClOx
species (here defined as Cl + ClO + 2Cl2 O2 + OClO + HOCl + 2Cl2 ) is
shown in the lower right panel. Ozone recovery takes place in late spring as
PSC disappear and reactive chlorine is converted back to ClONO2 and HCl.
Values based on the MOZART-3 global chemical/transport model. Courtesy of
D. Kinnison (NCAR).
Similarly, BrONO2 photolysis can destroy ozone efficiently in the

lower stratosphere (Burkholder, 1995):
BrO + NO2 → BrONO2

BrONO2 + hν → Br + NO3
Br + O3 → BrO + O2
NO3 + hν → NO + O2
NO + O3 → NO2 + O2
Net: 2 O3 → 3 O2
(5.337)
For typical levels of BrONO2 , this latter cycle is believed to be more
effective at destroying ozone than the similar cycle involving chlorine
compounds (Toumi et al., 1993; Lary, 1996). These cycles illustrate the
importance of the interactions between various chemical families.
In the ozone destruction cycle presented above, the conversion of NO
to NO2 by direct reactions with ozone can be replaced by other pathways,
including the reactions involving the peroxy radical HO2 ,
NO + HO2 → OH + NO2
OH + O3 → HO2 + O2
or the chlorine oxide radical ClO
NO + ClO → Cl + NO2
Cl + O3 → ClO + O2
It is important to note that certain cycles can lead to ozone formation;
this phenomenon is readily observed in the troposphere (see Box 5.4),
which is rich in anthropogenic hydrocarbons and nitrogen oxides. In
the free troposphere and lower stratosphere, the conversion of NO to
NO2 by peroxy radicals (HO2 and CH3 O2 ) produced by the oxidation
of methane and carbon monoxide, followed by the photodissociation of
NO2 leads to the formation of O3 . The complete chains are the following
(Crutzen, 1974):
CH4 + OH → CH3 + H2 O
CH3 + O2 + M → CH3 O2 + M
CH3 O2 + NO → CH3 O + NO2
O + O2 + M → O3 + M
CH3 O + O2 → CH2 O + HO2
HO2 + NO → NO2 + OH
O + O2 + M → O3 + M
CH2 O + hν → CO + H2
CH4 + 4 O2 + hν → H2 O + CO + H2 + 2 O3
(5.338)
Box 5.4. The Oxidizing Power of the Atmosphere

The influx into the stratosphere of chemical compounds emitted at
the surface (e.g., biogenic gases, primarly pollutants) or photochemically
produced in the lower atmosphere (e.g., secondary pollutants) depends
on the rate at which these compounds are destroyed in the troposphere.
The region below the tropopause often acts as a “filter” against species
that are transported upwards: soluble molecules are wet-scavenged
by precipitation; other species are chemically converted by oxidation
processes.
The oxidation potential of the atmosphere determines the rate at which
most reactive species are destroyed in the atmosphere. Oxidation by
molecular oxygen (O2 ) is generally slow. The most efficient oxidant
is the hydroxyl (OH) radical, although other compounds such as NO3 ,
H2 O2 and O3 play a significant role in certain cases. As in the
middle atmosphere, the OH radical in the troposphere is produced
primarily during the daytime by the reaction between water vapor (H2 O)
and the electronically-excited oxygen atom O(1 D) (see Reaction (5.82)
and Engleman, 1965; Levy, 1971) In the upper troposphere, however,
especially in the tropics, where the water vapor abundance is relatively
low, and molecules released at the surface can be rapidly transported by
convective uplift, an additional primary source of hydrogen and radicals
(HOx ) is provided by the photolysis of organic molecules, including
acetone, aldehydes, and organic peroxides (e.g., Singh et al., 1995; Prather
and Jacob, 1997; Muller
¨ and Brasseur, 1999).
OH is converted to HO2 by reactions with carbon monoxide (Reaction
(5.79)) and ozone (Reaction (5.90)), and HO2 is converted back to OH
by reactions with nitric oxide (Reaction (5.92)) and ozone (Reaction
(5.91)). The loss of HOx results from the conversion of OH and HO2
into soluble species like H2 O2 and HNO3 , followed by wet scavenging of
these molecules through precipitation.
Since the presence (in very small quantities) of O(1 D) in the
troposphere results from the photolysis of ozone (see Reaction (5.16)),
an assessment of the oxidizing power of the atmosphere requires that the
global budget of tropospheric ozone be accurately quantified.
Approximately 10% of atmospheric ozone resides in the troposphere.
Since solar radiation of wavelengths less than 242 nm that photolyzes
molecular oxygen (see Reaction (5.11)) is entirely absorbed above the
tropopause, and hence cannot produce ozone in the lower atmosphere,
it was believed for a long time that the presence of O3 in the free
troposphere was due to downward transport of ozone-rich air masses from
the stratosphere. Dry deposition of O3 on vegetation was believed to be
the only significant loss process.
In the early 1950s, Hagen-Smit (1952) suggested that the high ozone
levels frequently observed in urban areas, like Los Angeles during summer-
Box 5.4 (Continued)

time, were associated with the emission of primary pollutants (NOx ,
CO, hydrocarbons) by automobiles and power plants. About 20 years
later, Crutzen (1973, 1974) and Chameides and Walker (1973) showed
that similar processes could provide a substantial photochemical source of
ozone in the free troposphere at the global scale. Rather than involving
the photolysis of O2 as in the middle atmosphere, the formation of
atomic oxygen (which recombines with O2 to form ozone) results from
the photolysis of NO2 into O and NO. The process that converts NO back
to NO2 without consuming ozone, and hence “limits” the rate of ozone
formation is provided by
NO + RO2 → NO2 + RO (5.339)
where the peroxy radicals RO2 (which can be a hydroperoxy radical

HO2 or organic peroxy radical such as CH3 O2 , C2 H5 O2 , etc.) are
produced by the oxidation of the so-called ozone precursors (CO, CH4 ,
non-methane hydrocarbons). Oxidation of these precursors leads to a net
ozone production only for sufficiently high atmospheric levels of nitrogen
oxides (see Reactions (5.338) and (5.340)).
The major photochemical losses of tropospheric ozone are provided by
reactions of O3 with OH (Reaction (5.90)) and HO2 (Reaction (5.91)),
as well as by ozone photolysis (Reaction (5.16)), if it is followed by the
irreversible loss of O(1 D), for example, by the reaction of this tome with
water vapor.
Figure 5.70 shows the number density of OH and HO2 , as well as the net
ozone photochemical production as a function of the NOx mixing ratio,
near the surface and in the upper troposphere. These daytime quantities
are calculated by a zero-dimensional “box” model for typical values of
H2 O, CH4 CO and nonmethane hydrocarbon concentrations. This figure
suggests that the oxidizing properties of the troposphere are strongly
affected by the level of tropospheric NOx . Over the oceans, for example,
where the abundance of NOx is generally very low, the atmosphere acts
as a loss for tropospheric ozone and the OH/HO2 concentration ratio
is low. In moderately polluted areas, where NOx levels are higher, the
net ozone production and the OH concentration increase with the NOx
abundance. In highly polluted areas (e.g., urban zones) with very high
NOx concentrations, HOx radicals are efficiently converted into nitric acid,
and the net production of ozone decreases dramatically.
The nitrate radical (NO3 ) which is present in the troposphere primarily
during nighttime is also a powerful oxidant, and reacts efficiently
with many organic compounds (nonmethane hydrocarbons, DMS, etc.).
Hydrogen peroxide (H2 O2 ) is a major oxidant for SO2 inside water
droplets, and contributes to the formation of sulfate aerosols.
Box 5.4 (Continued)
Figure 5.70. Upper panel: Concentration (cm−3 ) of OH and HO2 as a

function of the NOx (NO + NO2 ) mole fraction (pptv) for conditions
representative of the upper troposphere (UT-solid line) and planetary
boundary layer (PBL-dashed line). Lower panel: Same, but for the net
photochemical rate (ppb/hr) of ozone. The values corresponding to the
upper troposphere are multiplied by 10. From Brune (2000).
The oxidizing properties of the troposphere have a strong influence

on the lifetime of chemical compounds in the atmosphere, and hence on
the probability for a molecule to reach the middle atmosphere. Most
hydrocarbons, for example, including hydrogenated halocarbons, are
efficiently destroyed by the OH radical in the troposphere before they
can penetrate into the stratosphere. Compounds that are not oxidized
in the troposphere (e.g., chlorofluorocarbons) or weakly oxidized (e.g.,
methane) reach the stratosphere more easily.
and
CO + OH → CO2 + H
H + O2 + M → HO2 + M
HO2 + NO → OH + NO2
O + O2 + M → O3 + M
Net: CO + 2 O2 + hν → CO2 + O3
(5.340)
If, however, NO is reduced by efficient heterogeneous reactions (e.g.,
on polar stratospheric clouds), the oxidation of methane becomes
catalyzed by ClO, and, rather than producing two ozone molecules per
cycle, destroys two ozone molecules for each methane molecule oxidized
(Crutzen et al., 1992). The chain is the following (Lary, 1996):
CH4 + OH → CH3 + H2 O
CH3 + O2 + M → CH3 O2 + M
CH3 O2 + ClO → CH3 O + Cl + O2
Cl + O3 → ClO + O2
CH3 O + O2 → CH2 O + HO2
HO2 + ClO → HOCl + O2
HOCl + hν → OH + Cl
Cl + O3 → ClO + O2
CH2 O + hν → CO + H2
Net: CH4 + 2 O3 + hν → H2 O + CO + H2 + 2O2

(5.341)
The complete ozone equation is complex as a result of all these
processes. Considering only the most important processes, the general
behavior of ozone can be described by the following equations:
d(O3 )
+ JO3 (O3 ) + k3 (O)(O3 ) + a2 (H)(O3 ) + a6 (OH)(O3 )
dt
+ a6b (HO2 )(O3 ) + b4 (NO)(O3 ) + b9 (NO2 )(O3 ) (5.342)
+ d2 (Cl)(O3 ) + e2 (Br)(O3 ) = k2 (M)(O2 )(O)

d(O)
+ 2k1 (M)(O)2 + k2 (M)(O2 )(O) + k3 (O3 )(O) + a5 (OH)(O)
dt
+ a7 (HO2 )(O) + b3 (NO2 )(O) + d3 (ClO)(O) + e3 (BrO)(O)
= 2JO2 (O2 ) + JO3 (O3 ) + JNO2 (NO2 ) + JBrO (BrO) + JaNO3 (NO3 )
(5.343)
where equilibrium has been assumed for O(1 D). Adding Eqs. (5.342)
and (5.343), the rate of change for odd oxygen (Ox ) becomes
d(Ox )
+ 2k1 (M)(O)2 + 2k3 (O)(O3 ) + a2 (H)(O3 )
dt
+ [a5 (OH) + a7 (HO2 )] (O) + [a6 (OH) + a6b (HO2 )] (O3 )
+ b3 (NO2 )(O) + b4 (NO)(O3 ) + b9 (NO2 )(O3 )
+ d3 (ClO)(O) + d2 (Cl)(O3 ) + e3 (BrO)(O) + e2 (Br)(O3 )

= 2JO2 (O2 ) + JNO2 (NO2 ) + JaNO3 (NO3 )
(5.344)
This equation still accounts for gross production and destruction rates
associated with rapid chemical cycling. To eliminate large, and often
compensating terms, it is therefore more appropriate to define the “odd
oxygen” as the sum of all chemical compounds that include one or more
“odd oxygen” attached to them. If we define Oy as the sum
Oy =O(3 P) + O(1 D) + O3 + NO2 + 2NO3 + HNO3
+ HO2 NO2 + 2N2 O5 + ClO + 2Cl2 O2 + 2OClO

+ 2ClONO2 + BrO + 2BrONO2
and if we add the continuity equations for all the corresponding species,
we obtain, after eliminating the small terms and adding a transport term
∂(Oy )
+ ∇ · φ (Oy ) + 2k1 (M)(O)2 + 2k3 (O)(O3 ) + a2 (H)(O3 )
∂t
+ [a5 (OH) + a7 (HO2 )] (O) + [a6 (OH) + a6b (HO2 )] (O3 )
+ 2b3 (NO2 )(O) + 2d3 (ClO)(O) + 2e3 (BrO)(O)
+ d33 (ClO)(HO2 ) + e15 (BrO)(HO2 ) + 2JCl2 O2 (Cl2 O2 )
+ 2 [d48 + d49 ] (ClO)(ClO) + 2e6 (BrO)(BrO)
+ 2 [e5b + e5c ] (BrO)(ClO) + 2JbNO3 (NO3 )
= 2JO2 (O2 ) + a26 (HO2 )(NO) + c5 (CH3 O2 )(NO)
(5.345)
Under this convention, the terms that appear in this general equation
for Oy represent the net production and destruction rates of odd oxygen,
and correspond to the “limiting rates” in most of the cycles presented
above. At all altitudes (Oy ) ≈ (Ox ), so that Eq. (5.345) can be regarded
as a form of the odd oxygen equation. In the stratosphere below 50 km,
(Oy ) ≈ (Ox ) = (O3 ) + (O) ≈ (O3 )
while, above 70 km
(Oy ) ≈ (Ox ) ≈ (O).
Figure 5.71 shows the relative contribution of the Ox (Chapman cycle),

NOx , HOx , ClOx , and BrOx families to the destruction of ozone in
the stratosphere for different latitudes and for two seasons. Note the
important role played by NOx in the mid-stratosphere and by HOx in
the mesosphere and lower stratosphere. Halogen species play a major
role at high latitudes (20 km and 45 km). The role of the Chapman
reactions is generally small except in the upper stratosphere. In the polar
regions during periods of activated chlorine, the major loss in the lower
stratosphere results from cycle (5.333) as well as cycle (5.334). This
last loss process, which is limited by the rate of reactions between ClO
and BrO, plays also a significant role in the lower stratosphere at mid-
latitudes, where chlorine and bromine are activated on sulfate aerosol
particles. A diagnostic of ozone loss in the lower stratosphere, based on
observational data, is given by Lee et al. (2002) and Rex et al. (2003).
In closing, we remark that although the complete chemistry of the
middle atmosphere involves several hundred reactions, most of the
important processes can be identified by examining the magnitudes of
the photochemical terms in the continuity equations for each species,
and the resulting lifetimes. When this information is used to define
chemical families, a simple framework is obtained in which the species
subject to the influence of transport can be readily identified, and the net
versus gross production and loss processes are clearly distinguished from
one another. Indeed, the situation is somewhat similar to the difficulties
associated with understanding the significance of the net and gross mean
meridional and eddy transports in the stratosphere in the classical and
residual Eulerian representations, as was described in Chapter 3. Thus,
in the photochemistry as well as the dynamics of the middle atmosphere,
the occurrence of many simultaneous processes, some of which merely
cancel with one another to a great degree, requires careful study to
recognize the roles played by each of them. The particular frameworks
chosen to accomplish that task in this volume represent one possible
choice which may facilitate that study.
Figure 5.71. Relative contribution (percent)

( ) of various chemical families (NO
( x,
HOx , ClOx , BrOx , and Ox ) to the destruction of odd oxygen in the stratosphere
in June (based on calculations by the NASA/Goddard 2-D model). Courtesy
of C.H. Jackman, NASA.
References
Abbas et al. (1996) p. 44 Abbas, M.M., M.R. Gunson, M.J. Newchurch, H.A.
Michelsen, R.J. Salawitch, M. Allen, M.C. Abrams, A.Y. Chang, A. Goldman, F.W.
Irion, E.J. Moyer, R. Nagaraju, C.P. Rinsland, G.P. Stiller, and R. Zander, The
hydrogen budget of the stratosphere inferred from ATMOS measurements of H2 O
and CH4 . Geophys Res Lett: 23, 2405, 1996a.
Abbas, M.M., H.A. Michelsen, M.R. Gunson, M.C. Abrams, M.J. Newchurch, R.J.
Salawitch, A.Y. Chang, A. Goldman, F.W. Irion, G.L. Manney, E.J. Moyer, R.
Nagaraju, C.P. Rinsland, G.P. Stiller, and R. Zander, Seasonal variations of water
vapor in the lower stratosphere inferred from ATMOS/ATLAS-3 measurements of
H2 O and CH4 . Geophys Res Lett: 23, 2401, 1996b.
Ackerman, M., D. Frimout, A. Girard, M. Gottignies, and C. Muller, Stratospheric
HCl from infrared spectra. Geophys Res Lett: 3, 81, 1976.
Ackerman, M., D. Frimout, and C. Muller, Stratospheric CH4 , HCl and ClO and the
chlorine ozone cycle. Nature: 269, 226, 1977.
Adriani, A., T. Deshler, G. Di Donfranceso, and G.P. Gobbi, Polar stratospheric
clouds and volcanic aerosol during spring 1992 over McMurdo Station Actarctica:
Lidar and particle counter comparison. J Geophys Res: 100, 25,877, 1995.
Allan, P.J., J.M.C. Plane, and G. McFiggans, Observations of OIO in the remote
marine boundary layer. Geophys Res Lett: 28, 1945, 2001.
Allen, M., Y.L. Yung, and J.W. Waters, Vertical transport and photochemistry in
the terrestrial mesosphere and lower thermosphere (50-170 km). J Geophys Res: 86,
3617, 1981.
Allen, M., J.I. Lunine, and Y.L. Yung, The vertical distribution of ozone in the
mesosphere and lower thermosphere. J Geophys Res: 89, 4841, 1984.
Anderson, J.G., Rocket measurement of OH in the mesosphere. J Geophys Res: 76,
7820, 1971.
Anderson, J.G., The absolute concentration of OH in the earth’s stratosphere.
Anderson, J.G., Free radicals in the earth’s stratosphere: A review of recent results,
Anderson, J.G., H.J. Grassl, R.E. Shetter, and J.J. Margitan, Stratospheric free
chlorine measured by balloon-borne in situ resonance fluorescence. J Geophys Res:
85, 2869, 1980.
Angell, J.K., and J. Korshover, Quasi-Biennal variations in temperature, total ozone
and tropopause height. J Atmos Sci: 21, 5, 479, 1964.
Angell, J.K., and J. Korshover, Quasi-Biennal and long-term fluctuation in total
ozone. Mon Wea Rev: 101, 426, 1973.
Arijs, E., and G. Brasseur, Acetonitrile in the stratosphere and implications for
positive ion composition. J Geophys Res: 91, 4003, 1985.
Arijs, E., D. Nevejans, P. Frederick, and J. Ingels, Negative ion composition
measurements in the stratosphere. Geophys Res Lett: 8, 121, 1981.
Arnold, F., and D. Krankowsky, Water vapour concentrations at the mesopause.
Nature: 268, 218, 1977.
Arnold, F., and R. Fabian, First measurements of gas phase sulfuric acid in the
stratosphere. Nature: 283, 55, 1980.
Arpag, K.A., P.V. Johnston, H.L. Miller, R.W. Sanders, and S. Solomon,
Observations of the stratospheric BrO column over Colorado, 40◦ N. J Geophys Res:
99, 8175, 1994.
Atlas, E., W. Pollock, J. Greenberg, L. Heidt, and A.M. Thompson, Alkyl nitrates,
nonmethane hydrocarbons, and halocarbon gases over the equatorial Pacific Ocean
during SAGA 3. J Geophys Res: 98, 16,933, 1993.
Avallone, L.M., and D.W. Toohey, Tests of halogen photochemistry using in situ
measurements of ClO and BrO in the lower polar stratosphere. J Geophys Res: 106,
10,411, 2001.
Avallone, L.M., S.M. Schauffler, W.H. Pollock, L.E. Heidt, E.L. Atlas, and K.R.
Chan, In situ measurements of BrO during AASE II. Geophys Res Lett: 22, 831,
1995.
Baker, K.D., A.F. Nagy, R.O. Olsen, E.S. Oran, J. Randhawa, D.F. Strobel, and T.
Tohmatsu, Measurement of the nitric oxide altitude distribution in the mid-latitude
mesosphere. J Geophys Res: 82, 3281, 1977.
Bandy, A.R., D.C. Thornton, D.L. Scott, M. Lalevic, E.E. Lewin, and A.R.
Driedger III, A time series of carbonyl sulfide in the northern hemisphere. J Atmos
Chem: 14, 527, 1992.
Bandy, A.R., D.C. Thornton, and A. Driedger III, Airborne measurements of sulfur
dioxide, dimethyl sulfide, carbon disulfide, and carbon sulfide by isotope dilution gas
chromatography/mass spectrometry. J Geophys Res: 98, 23,423, 1993a.
Bandy, A.R., D.C. Thornton, and J.E. Johnson, Carbon disulfide measurements in
the atmosphere of the western north Atlantic and the northwestern south Atlantic
oceans. J Geophys Res: 98, 23,449, 1993b.
Barbe, A., P. Mache, C. Secroun, and P. Jouve, Measurements of tropospheric and
stratospheric H2 CO by an infrared high resolution technique. Geophys Res Lett: 6,
463, 1979.
Barth, C.A., D.W. Rusch, R.J. Thomas, G.H. Mount, G.J. Rottman, G.E. Thomas,
R.W. Sanders, and G.M. Lawrence, Solar mesosphere explorer: Scientific objectives
and results. Geophys Res Lett: 10, 237, 1983.
Barth, C.A., S.M. Bailey, and S.C. Solomon, Solar-terrestrial coupling: Solar osft
X-rays and thermospheric nitric oxide. Geophys Res Lett: 26, 1251, 1999.
Bates, D.R., and M. Nicolet, Atmospheric hydrogen. Publ Astron Soc Pacific: 62,
106, 1950.
Bates, D.R., and A.E. Witherspoon, The photochemisry of some minor constituents
of the earth’s atmosphere (CO2 , CO, CH4 , N2 O). Mon Notic Roy Astron Soc: 112,
101, 1952.
Bekki, S., R. Toumi, and J.A. Pyle, Role of sulphur photochemistry in tropical ozone
changes after the eruption of Mount Pinatubo. Nature: 362, 331, 1993.
Bergamaschi, P.R., M. Hein, M. Heinmann, and P.J. Crutzen, Inverse modeling of
the global CO cycle, 1. Inversion of CO mixing rations. J Geophys Res: 105, 1909,
2000.
Bevilacqua, R.M., J.J. Olivero, P.R. Schwartz, C.J. Gibbins, J.M. Bologna, and D.J.
Thacker, An observational study of water vapor in the mid-latitude mesosphere using
ground-based microwave technique. J Geophys Res: 88, 8523, 1983.
Beyer, K.D., B. Luo, R. Neuber, and T. Peter, Temperature dependence of ternary
solution particle volumes as observed by lidar in the Arctic stratosphere during winter
1992/1993. J Geophys Res: 102, 3603, 1997.
Bingemer, H.G., S. Bürgermeister,

¨ R.L. Zimmermanm, and H.-W. Georgii,
Atmospheric OCS: Evidence for a contribution of anthropogenic sources? J Geophys
Res: 95, 20 617, 1990.
Black, G., R.L. Sharpless, and T.G. Slanger, Rate coefficients for SO reactions with
O2 and O3 over the temperature range of 230 to 420 K. Chem Phys Lett: 90, 598,
1982a.
Black, G., R.L. Sharpless, and T.G. Slanger, Rate coefficients at 198 K for SO
reactions with O2 , O3 , and NO2 . Chem Phys Lett: 90, 55, 1982b.
Boatman, J.F., D.L. Wellman, C.C. Van Valin, R.L. Gunter, J.D. Ray, H. Sievering,
Y. Kim, S.W. Wilkinson, and M. Luria, Airborne sampling of selected trace chemicals
above the central United States. J Geophys Res: 94, 5081, 1989.
Brasseur, G.P., The response of the middle atmopshere to a long-term and short-term
solar variability: A two dimensional model. J Geophys Res: 98, 23,079, 1993.
Brasseur, G., and C. Granier, Mount Pinatubo aerosols, chlorofluorocarbons, and
ozone depletion. Science: 257,
7 1239, 1992.
Brasseur, G.P., and S. Madronich, Chemistry-transport models, in Climate System
Modeling. K.E. Trenberth, ed., Cambridge University Press, 1992.
Brasseur, G., A. De Rudder, G.M. Keating, and M.C. Pitts, Response of the middle
atmosphere to short-term solar ultraviolet variations, 2. Theory. J Geophys Res: 92,
903, 1987.
Brasseur, G.P., J.J. Orlando, and G.S. Tyndall (eds), Atmospheric Chemistry and
Global Change, Oxford University Press, 1999.
Breeding, R.J., J.P. Lodge, Jr., J.B. Pate, D.C. Sheesley, H.B. Klonis, B. Fogle, J.A.
Anderson, T.R. Englert, P. Haagenson, R.B. McBeth, A.L. Morris, R. Pogue, and
A.F. Wartburg, Background trace gas concentrations in the central United States. J
Geophys Res: 78, 7057, 1973.
Brewer, A.W., Evidence for a world circulation provided by the measurements of
helium and water vapour distributions in the stratosphere. Quart J Roy Met Soc: 75,
351, 1949.
Browell, E.V., C.F. Butler, S. Ismail, P.A. Robinette, A.F. Carter, N.S. Higdon, O.B.
Toon, M.R. Schoeberl, and A.F. Tuck, Airborne lidar observations in the wintertime
Arctic stratosphere: Polar stratospheric clouds. Geophys Res Lett: 17,7 385, 1990.
Brownsword, R.A., M. Hillenkamp, T. Laurent, R.K. Vatsa, H.-R. Volpp, and J.
Wolfrum, Quantum yield for H atom formation in the methane dissociation after
photoexcitation at the Lyman-α (121.6 nm) wavelength. Chem Phys Lett: 266, 259,
1997.
Brune, W., OH and HO2 : Sources, interactions with nitrogen oxides, and ozone
production. IGAC Newsletter: 21, 2000.
Brune, W.H., J.G. Anderson, and K.R. Chan, In situ observations of BrO over
Antarctica: ER-2 aircraft results from 54◦ S to 72◦ S latitude. J Geophys Res: 94,
16,639, 1989.
Burkholder, J.B., A.R. Ravishankara, and S. Solomon, UV/visible and IR absorption
cross sections of BrONO2 . J Geophys Res: 100, 16,793, 1995.
Burnett, C.R., and E.B. Burnett, Spectroscopic measurement of the vertical column
abundance of hydroxyl (OH) in the earth’s atmosphere. J Geophys Res: 86, 5185,
1981.
Burnett, C.R., and E.B. Burnett, Vertical column abundance of atmospheric OH at
solar maximum from Fritz Peak, Colorado. Geophys Res Lett: 9, 708, 1982.
Bush, J.A., A.L. Schmeltekopf, F.C. Fehsenfeld, D.L. Albritton, J.R. McAffe, P.D.
Goldan, and E.E. Ferguson, Stratospheric measurements of methane at several
latitudes. Geophys Res Lett: 5, 1027, 1978.
Byrne, G.D., and A.C. Hindmarsh, A polyalgorithm for the numerical solution of
ordinary differential equations. ACM Trans. Math Software: 1, 71, 1975.
Calvert, J.G., and W.R. Stockwell, Acid generation in the troposphere by gas phase
chemistry. Environ Sci Technol: 17,
7 428A, 1983.
Carroll, M.A., R.W. Sanders, S. Solomon, and A.L. Schmeltekopf, Visible and near-
ultraviolet spectroscopy at McMurdo station, Antarctica, 6: Observations of BrO. J
Geophys Res: 94, 16,633, 1989.
Carslaw, K.S., T. Peter, and S.L. Clegg, Modeling the composition of liquid
stratospheric aerosols. Rev Geophys: 35, 125, 1997.
Carslaw, K.S., M. Wirth, A. Tsias, B.P. Luo, A. Dörnbrack,
¨ M. Leutbecher, H.
Volkert, W. Renger, J.T. Bacmeister, and T. Peter, Particle microphysics and
chemistry in remotely observed mountain polar stratospheric clouds. J Geophys Res:
103, 5785, 1998a.
Volkert, W. Renger, J.T. Bacmeister, and T. Peter, Increased stratospheric ozone
depletion due to mountain-induced atmospheric waves. Nature: 391, 675, 1998b.
Chameides, W.L., and J.C.G. Walker, A photochemical theory of tropospheric ozone.
J Geophys Res: 78, 8751, 1973.
Chance, K.V., and W.A. Traub, Evidence for stratospheric hydrogen peroxide. J
Geophys Res: 92, 3061, 1987.
Chance, K.V., D.G. Johnson, W.A. Traub, and K.W. Jucks, Measurement of the
stratospheric hydrogen peroxide concentration profile using far infrared thermal
emission spectroscopy. Geophys Res Lett: 18, 1003, 1991.
Chance, K.V., W.A. Traub, D.G. Johnson, K.W. Jucks, P. Ciarfallini, R.A. Stachnik,
R.J. Salawitch, and H.A. Michelsen, Simultaneous measurements of stratospheric
HOx , NOx , and Clx : Comparison with a photochemical model. J Geophys Res: 101,
9031, 1996.
Chandra, S., The solar UV related changes in total ozone from a solar rotation to a
solar cycle. Geophys Res Lett: 18, 837, 1991.
Zander, C.P. Rinsland, M. Loewenstein, J.R. Podolske, M.H. Proffitt, J.J. Margitan,
D.W. Fahey, R.-S. Gao, K.K. Kelly, J.W. Elkins, C.R. Webster, R.D. May, K.R.
Chan, M.M. Abbas, A. Goldman, F.W. Irion, G.L. Manney, M.J. Newchurch, and
G.P. Stiller, A comparison of measurements from ATMOS and instruments aboard
the ER-2 aircraft: Tracers of atmospheric transport. Geophys Res Lett: 23, 2389,
1996.
Chang, J.S., R.R. Barker, J.E. Davenport, and D.M. Goldan, Chlorine nitrate
photolysis by a new technique: Very low pressure photolysis. Chem Phys Lett: 60,
385, 1979.
Chapman, S., On ozone and atomic oxygen in the upper atmosphere. Phil Mag: 10,
369, 1930.
Chubachi, S., A special ozone observation at Syowa station Antarctica from February
1982 to January 1983, in Atmospheric Ozone. S. Zerefos and A. Ghazi (eds.), Reidel,
Dordrecht, pp. 285, 1985.
Cicerone, R.J., and R.S. Oremland, Biogeochemical aspects of atmospheric methane.
Global Biogeochem Cycles: 2, 299, 1988.
Clancy, R.T., D.O. Muhlemon, and G.L. Berge, Microwave spectra of terrestrial
mesosphere CO. J Geophys Res: 87,
7 5009, 1982.
Coffey, M.T., W.G. Mankin, and R.J. Cicerone, Spectrascopic detection of
stratospheric hydrogen cyanide. Science: 214, 333, 1981a.
Coffey, M.T., W.G. Mankin, and A. Goldman, Simultaneous spectroscopic
determination of the latitudinal, seasonal and diurnal variability of stratosphere N2 O,
NO, NO2 and HNO3 . J Geophys Res: 86, 7731, 1981b.
Cohen, R.C., P.O. Wennberg, R.M. Stimpfle, J. Koplow, J.G. Anderson, D.W. Fahey,
E.L. Woodbridge, E.R. Keim, R. Gao, M.H. Proffitt, M. Loewenstein, and K.R. Chan,
Are models of catalytic removal of O3 by HOx accurate? Constraints from in-situ
measurements of the OH to HO2 ratio. Geophys Res Lett: 21, 2539, 1994.
Conrad, R., and W. Seiler, Influence of temperature, moisture, and organic carbon
on the flux of H2 and CO between soil and atmosphere: Field studies in subtropical
regions. J Geophys Res: 90, 5699, 1985.
Cox, R.A., W.J. Bloss, R.L. Jones, and D.M. Rowley, OIO and the atmospheric cycle
of iodine. Geophys Res Lett: 26, 1857, 1999.
Crutzen, P.J., Determination of parameters appearing in the “dry” and the “wet”
photochemical theories for ozone in the stratosphere. Tellus: 21, 368, 1969.
Crutzen, P.J., The influence of nitrogen oxide on the atmospheric ozone content.
Quart J Roy Met Soc: 96, 320, 1970.
Crutzen, P.J., Ozone production rates in oxygen-hydrogen-nitrogen oxide atmosphere.
J Geophys Res: 76, 7311, 1971.
Crutzen, P.J., A discussion of the chemistry of some minor constituents in the
stratosphere and troposphere. Pure Appl Geophys: 106, 1385, 1973.
Crutzen, P.J., Photochemical reactions initiated by the influencing ozone in
unpolluted tropospheric air. Tellus: 26, 47, 1974.
Crutzen, P.J., The possible importance of CSO for the sulfate layer of the
stratosphere. Geophys Res Lett: 3, 73, 1976.
Crutzen, P.J., Atmospheric interactions: Homogenous gas reactions of C, N and S
containing compounds, in The Major Biochemical Cycles and Their Interactions. B.
Bolin and R.B. Cook (eds.), SCOPE 21, 67-114, Wiley, Chichester, 1983.
Crutzen, P.J., I.S.A. Isaksen, and G.C. Reid, Solar Protons events: Stratospheric
sources of nitric oxide. Science: 189, 457, 1975.
Crutzen, P.J., and L.T. Gidel, The tropospheric budgets of anthropogenic
chlorocarbons, CO, CO4 , CH3 Cl and the effect of various NOx sources on tropospheric
ozone. J Geophys Res: 88, 6641, 1983.
Crutzen, P.J., and U. Schmaizl, Chemical budgets of the stratosphere. Planet Space
Sci: 31, 1009, 1983.
Crutzen, P.J., and F. Arnold, Nitric acid cloud formation in the cold Antarctic
stratosphere: A major cause for the springtime ‘ozone hole.’ Nature: 324, 651, 1986.
Crutzen, P.J., and P.H. Zimmermann, The changing photochemistry of the
troposphere. Tellus: 43A-B, 136, 1991.
Crutzen, P.J., L.E. Heidt, J.P. Krasnec, W.H. Pollock, and W. Seiler, Biomass burning
as a source of the atmosphere gases CO, H2 , N2 O, NO, CH3 Cl and COS, Nature:
282, 253, 1979.
Crutzen, P.J., R. Müller,
¨ C. Bruhl,
¨ and T. Peter, On the potential importance of the
gas phase reaction CH3 O2 + ClO → ClOO + CH3 O and the heterogenous reaction
HOCl + HCl → H2 O + CL2 in “ozone hole” chemistry. Geophys Res Lett: 19, 1113,
1992.
Daniel, J.S., S. Solomon, R.W. Portmann, and R.R. Garcia, Stratospheric ozone
destruction: The importance of bromine relative to chlorine. J Geophys Res: 104,
23,871, 1999.
David, D., S. Godin, G. Megie,
´ Y. Emery, and C. Flesia, Physical state and
composition of polar stratospheric clouds from airborne lidar measurements during
sesame. J Atmos Chem: 27, 7 1, 1997.
Deguchi, S., and O.O. Muhleman, Mesopheric water vapor. J Geophys Res: 87,
7 1343,
1982.
Delwiche, C.C., Biological production and utilization of N2 O. Pageoph: 116, 414,
1978.
De Zafra, R.L., A. Parrish, P.M., Solomon and J.W. Barrett, A measurement of
stratospheric HO2 by ground-based nm-wave spectroscopy. J Geophys Res: 89, 1321,
1984.
De Zafra, R.L., M. Jaramillo, J. Barrett, L.K. Emmons, P.M. Solomon, and A.
Parrish, New observations of large concentration of ClO in the springtime lower
stratosphere over Antarctica and its implications for ozone-depleting chemistry. J
Geophys Res: 94, 11,423, 1989.
Dlugokencky, E.J., L.P. Steel, P.M. Lang, and K.A. Masarie, The growth rate and
distribution of atmospheric methane. J Geophys Res: 99, 17,021, 1994.
Dlugokencky, E.J., E.G. Dutton, P.C. Novelli, P.P. Tans, K.A. Masarie, K.O. Lantz,
and S. Madronich, Changes in CH4 and CO growth rates after the eruption of Mt.
Pinatubo and their link with changes in tropical tropospheric UV flux. Geophys Res
Lett: 23, 2761, 1996.
Dlugokencky, E.J., K.A. Masarie, P.M. Lang, and P.P. Tans, Continuing decline in
growth rate of the atmospheric methane burden. Nature: 393, 447, 1998.
Dobson, G.M.B., Observations of the amount of ozone in the earth’s atmosphere and
its relation to other geophysical conditions. Proc Roy Soc London: A 129, 411, 1930.
Dobson, G.M.B., Exploring the Atmosphere, Clarendon Press (Oxford), 1963.
Drdla, K., A. Tabazadeh, R.P. Turco, M.Z. Jacobsen, J.E. Dye, C. Twohy, and D.
Baumgardner, Analysis of the physical state of one arctic polar stratospheric cloud
based on observations. Geophys Res Lett: 21, 2475, 1994.
Dupuy, É., J. Urban, P. Ricaud, É. Le Flochmoen,
¨ N. Lauti´ é, D. Murtagh, J. De
La Noe,
¨ L. El Amraoui, P. Eriksson, P. Forkman, U. Frisk, F. J´ égou, C. Jiménez,
and M. Olberg, Strato-mesospheric measurements of carbon monoxide with the Odin
Sub-Millimetre Radiometer: Retrieval and first results. Geophys Res Lett, in press.
Dutsch,
¨ H.U., Atmospheric ozone: A short review. J Geophys Res: 75, 1707, 1970.
Dutsch,
¨ H.U., Vertical ozone distribution and tropospheric ozone, Proc of the
NATO Advanced Institute on Atmospheric Ozone, A.C. Aikin, ed., U.S. Dept. of
Dvortsov, V.L., M.A. Geller, S. Solomon, S.M. Schauffler, E.L. Atlas, and D.R. Blake,
Rethinking reactive halogen budgets in the midlatitude lower stratosphere. Geophys
Res Lett: 26, 1699, 1999.
Dye, J.E., D. Baumgardner, B.W. Gandrud, S.R. Kawa, K.K. Kelly, M. Loewenstein,
G.V. Ferry, K.R. Chan, and B.L. Gary, Particle size distribution in Arctic polar
stratospheric clouds, growth and freezing of sulfuric acid droplets, and implications
for cloud formation. J Geophys Res: 97,
7 8015, 1992.
Eckman, R.S., The response of ozone to short-term variations in the solar ultraviolet
irradiance, 1. A theoretical model. J Geophys Res: 91, 6695, 1986.
Ehhalt, D.H., On the uptake of tritium by soil water and groundwater. Water Resour
Res: 9, 1073, 1973.
Ehhalt, D.H., The atmospheric cycle of methane. Tellus: 26, 58, 1974.
Ehhalt, D.H., Gas phase chemistry in the troposphere, in Global Aspects of
Atmospheric Chemistry, R. Zellner, ed., Springer Verlag, Topics in Physical
Chemistry: 6, 21, 1999.
Ehhalt, D.H., and L.E. Heidt, The concentration of molecular H2 and CH3 in the
stratosphere. Pageophys: 106, 1352, 1973.
Ehhalt, D.H., and A. Tonnissen, Hydrogen and carbon compounds in the stratosphere,
Ehhalt, D.H., L.E. Heidt, R.H. Lueb, and N. Roper, Vertical profiles of CH4 , H2 ,
CO, N2 O and CO2 in the stratosphere, in Proc of the Third Conference on CIAP.
U.S. Dept of Transportation, Washington, D.C., USA, 1974.
Ehhalt, D.H., L.E. Heidt, R.H. Lueb, and W. Pollock, The vertical distribution of
trace gases in the stratosphere. Pageoph: 113, 389, 1975a.
Ehhalt, D.H., L.E. Heidt, R.H. Lueb, and E.A. Martell, Concentrations of CH4 , CO,
CO2 , H2 , H2 O and N2 O in the upper stratosphere. J Atmos Sci: 32, 163, 1975b.
Ehhalt, D.H., U Schmidt, R. Zander, P. Demoulin, and C.P. Rinsland, Seasonal
cycle and secular trend of the total and tropospheric abundance of ethane above the
Jungfraujoch. J Geophys Res: 96, 4985, 1991.
Enell, C.-F., A. Steen, T. Wagner, U. Friess, K. Pfeilsticker, and U. Platt, Occurrence
of polar stratospheric clouds at kiruna. Ann Geophys: 17, 7 1457, 1999.
Engleman, R., The vibrational state of hydroxyl radicals produced by flash photolysis
of water-ozone-argone mixture. J Am Chem Soc: 87, 7 4193, 1965.
Etheridge, D.M., L.P. Steele, R.J. Francey, and R.L. Langenfelds, Atmospheric
methane between 1000 AD and present: Evidence of anthropogenic emissions and
climate variability. J Geophys Res: 103, 15 979, 1998.
Evans, W.J.J., O.M. Hunter, E.J. Lewellyn, and A. Vallance-Jones, Altitude profile
of the infrared system of oxygen in the dayglow. J Geophys Res: 73, 2885, 1968.
Evans, W.F.J., J.B. Kerr, C.T. McElroy, R.S. O’Brien, and J.C. McConnell,
Measurement of NO, NO2 and HNO3 during a stratospheric warming at 54◦ N in
February 1979. Geophys Res Lett: 9, 493, 1982.
Eyre, J.R., and H.K. Roscoe, Radiometric measurements of stratospheric HCl.
Nature: 266, 243, 1977.
Fabian, P., R. Borchers, K.H. Weiler, U. Schmidt, L.A. Volz, P.H. Ehhalt, W. Seiler,
and F. Muller, Simultaneously measured vertical profiles of H2 , CH4 , CO, N2 O,
CFCl3 , and CF2 Cl2 in the mid-latitude stratosphere and troposphere. J Geophys
Res: 84, 3149, 1979.
Fahey, D.W., K.K. Kelly, G.V. Ferry, L.R. Poole, J.C. Wilson, D.M. Murphy, M.
Loewenstein, and K.R. Chan, In situ measurements of total nitrogen, total water,
and aerosol in a polar stratospheric cloud in the Antarctic. J Geophys Res: 94,
11,299, 1989.
Fahey, D.W., K.K. Kelly, S.R. Kawa, A.F. Tuck, M. Loewenstein, K.R. Chan,
and L.E. Heidt, Observations of denitrification and dehydration in the winter polar
stratospheres. Nature: 344, 321, 1990.
Fahey, D.W., S.R. Kawa, E.L. Woodbridge, P. Tin, J.C. Wilson, H.H. Jonsson, J.E.
Dye, D. Baumgardner, S. Borrmann, D.W. Toohey, L.M. Avallone, M.H. Proffitt, J.
Margitan, M. Loewenstein, J.R. Podolske, R.J. Salawitch, S.C. Wofsy, M.K.W. Ko,
D.E. Anderson, M.R. Schoeberl, and K.R. Chan, In situ measurements constraining
the role of sulfate aerosols in mid-latitude ozone depletion. Nature: 363, 509, 1993.
Fahey, D.W., R.S. Gao, K.S. Carslaw, J. Kettleborough, P.J. Popp, M.J. Northway,
J.C. Holecek, S.C. Ciciora, R.J. McLaughlin, T.L. Thompson, R.H. Winkler, D.G.
Baumgardner, B. Gandrud, P.O. Wennberg, S. Dhaniyala, K. McKinney, Th. Peter,
R.J. Salawitch, T.P. Bui, J.W. Elkins, C.R. Webster, E.L. Atlas, H. Jost, J.C.
Wilson, R.L. Herman, A. Kleinböhl,
¨ and M. von Konig,
¨ The detection of large HNO3 -
containing particles in the winter arctic stratosphere. Science: 291, 1026, 2001.
Farman, J.C., B.G. Gardiner, and J.D. Shanklin, Large losses of total ozone in
Antarctic reveal seasonal ClOx /NOx interaction. Nature: 315, 207, 1985.
Farmer, C.B., Infrared measurements of stratospheric composition. Canad J Chem:
52, 1544, 1974.
Farmer, C.B., O.F. Raper, R.A. Toth, and R.A. Schindler, Recent results of aircraft
infrared observations of the stratosphere. Proc Third Conf on the Climatic Impact
Assessment Program, Rep. No. DOT-TSC-OST-74-15, U.S. Dept. of Transportation,
Washington, D.C., USA, 234, 1974.
Farmer, C.B., O.F. Raper, and R.H. Norton, Spectroscopic detection and vertical
distribution of HCl in the troposphere and stratosphere. Geophys Res Lett: 3, 13,
1976.
Farmer, C.B., O.F. Raper, B.D. Robbins, R.A. Toth, and C. Muller, Simultaneous
spectroscopic measurements of stratospheric species: O3 , CH4 , CO, CO2 , N2 O, H2 O,
HCl and HF at northern and southern mid-latitudes. J Geophys Res: 85, 1621, 1980.
Fish, D.J., S.R. Aliwell, and R.L. Jones, Midlatitude observations of the seasonal
variation of BrO: 2, Interpretation and modelling study. Geophys Res Lett: 24, 1199,
1997.
Fluckiger,
¨ J.A., T. D¨
allenbach,
¨ T. Blunier, B. Stauffer, T.F. Stocker, D. Raynaud, and
J.-M. Barnola, Variations in atmospheric N2 O concentration during abrupt climate
changes. Science: 285, 227, 1999.
Forkman, P., P. Eriksson, A. Winnberg, R.R. Garcia, and D. Kinnison, Longest
continuous ground-based measurements of mesospheric CO. Geophys Res Lett: 30,
1532, doi: 10.1029/2003GL016931, 2003.
Frederick, J.E., and N. Orsini, The distribution and variability of mesospheric odd
nitrogen: A theoretical investigation. J Atmos Terr Phys: 44, 4798, 1982.
Fried, A., B.E. Henry, J.G. Calvert, and M. Mozukewich, The reaction probability of
N2 O5 with sulfuric acid aerosols at stratospheric temperatures and compositions. J
Geophys Res: 99, 3517, 1994.
Friedl, A., J.R. Drummond, B. Henry, and J. Fox, Versatile integrated tunable diode
laser system for ambient measurements of OCS. Appl Optics: 30, 1916, 1991.
Friend, J.P., The global sulfur cycle, in Chemistry of the Lower Atmosphere. Rasool,
I., ed., Plenum Press, 177, 1973.
Friess, U., T. Wagner, I. Pundt, K. Pfeilsticker, and U. Platt, Spectroscopic
measurements of troposphere iodine oxide at Neumayer Station, Antarctica. Geophys
Res Lett: 28, 1941, 2001.
Fung, I., J. John, J. Lerner, E. Matthews, M. Prather, L. Steele, and P. Fraser, Three-
dimensional model synthesis of the global methane cycle. J Geophys Res: 96, 13,033,
1991.
Garcia, R.R., and S. Solomon, A numerical model of the zonally averaged dynamical
Garcia, R.R., and S. Solomon, The effect of breaking gravity waves on the dynamics
and chemical composition of the mesosphere and lower thermosphere. J Geophys Res:
90, 3850, 1985.
Garcia, R.R., and S. Solomon, A new numerical model of the middle atmosphere: 2,
Ozone and related species. J Geophys Res: 99, 12,937, 1994.
Gear, C.W., Numerical Initial Value Problems in Ordinary Differential Equations.
Prentice Hall, 1971.
Gettelman, A., J.R. Holton, and A.R. Douglass, Simulations of water vapor in the
lower stratosphere and upper troposphere, J Geophys Res: 105, 9003, 2000.
Gibbins, C.J., P.R. Schmitz, D.L. Thacher, and R.M. Bevilacqua, The variability of
mesospheric water vapor. Geophys Res Lett: 8, 1059, 1981.
Gille, J.C., and J.M. Russell, The limb infrared monitor of the stratosphere (LIMS):
Experiment description, performance and results. J Geophys Res: 89, 5125, 1984.
Gille, J.C., C.M. Smyth, and D.F. Heath, Observed ozone response to variations in
solar ultraviolet radiation. Science: 225, 315, 1984.
Gilles, M.K., A.A. Turnipseed, J.B. Burkholder, A.R. Ravishankara, and S. Solomon,
Kinetics of the IO radical: 2, Reaction of IO with BrO. J Phys Chem: 101, 5526,
1997.
Glueckauf, E., and G.P. Kitt, The hydrogen content of air at ground level. Quart J
Roy Met Soc: 83, 522, 1957.
Gobbi, G.P., and A. Adriani, Mechanisms of formation of stratospheric clouds
observed during the Antarctic late winter of 1992. Geophys Res Lett: 20, 1427,
1993.
Godin, S., G. Megie,
´ C. David, D. Haner, C. Flesia, and Y. Emery, Airborne lidar
observations of mountain-wave-induced polar stratospheric clouds during EASOE.
Goodmann, J., O.B. Toon, R.F. Pueschel, K. Snetsinger, and S. Verma, Antarctic
stratospheric ice crystals. J Geophys Res: 94, 16,449, 1989.
Goodmann, J., S. Verma, R.F. Pueschel, P. Hamill, G.V. Ferry, and D. Webster, New
evidence of size and composition of PSC particles. Geophys Res Lett: 24, 615, 1997.
Graedel, T.E., B. Kleiner, and G.C. Patterson, Measurements of extreme
concentrations of tropospheric hydrogen sulfide. J Geophys Res: 79, 4467, 1974.
Gunson, M.R., C.B. Farmer, R.H. Norton, R. Zander, C.P. Rinsland, J.H. Shaw, and
B.-C. Gao, Measurements of CH4 , N2 O, CO, H2 O, and O3 in the middle atmosphere
by the Atmospheric Trace Molecule Spectroscopy experiment on Spacelab 3. J
Geoophys Res: 95, 13,867, 1990.
Hagen-Smit, A.J., Chemistry and physiology of Los Angeles Smog. Ind Eng Chem:
44, 1342, 1952.
Hammill, P., A. Tabazadeh, S. Kinne, O.B. Toon, and R.P. Turco, On the growth
of ternary system HNO3 /H2 SO4 aerosol particles in the stratosphere. Geophys Res
Lett: 23, 753, 1996.
Hamm, S., and P. Warneck, The interhemispheric distribution and the budget of
acetonitrile in the troposphere. J Geophys Res: 95, 20,593, 1990.
Hampson, J., Chemiluminescent emissions observed in the stratosphere and
mesosphere, in Les Problemes Meteorologiques de la Stratosphere et de la Mesosphere.
M. Nicolet, ed., Presses Universitaries de France, Paris, 393, 1966.
Hanson, D.R., Reaction of N2 O5 with H2 O on bulk liquids and on particles and the
effect of dissolved HNO3 . Geophys Res Lett: 24, 1087, 1997.
Hanson, D.R., Reaction of ClONO2 with H2 O and HCl in sulfuric acid and
HNO3 /H2 SO4 /H2 O mixtures. J Phys Chem: 102, 4794, 1998.
Hanson, D.R., and K. Mauersberger, The vapor pressures of HNO3 -H2 O solutions. J
Phys Chem: 92, 6167, 1988.
Hanson, D.R., and A.R. Ravishankara, The reaction probabilities of ClONO2 and
N2 O5 on 40 to 75% sulfuric acid solutions. J Geophys Res: 96, 17307, 1991.
Hanson, D.R., and A.R. Ravishankara, Investigation of the reactive and nonreactive
processes involving ClONO2 and HCl on water and nitric acid doped ice. J Phys
Chem: 96, 2682, 1992.
Hanson, D.R., and A.R. Ravishankara, Uptake of HCl and HOCl onto sulfuric acid:
Solubilities, diffusivities and reaction. J Phys Chem: 97,
7 12309, 1993.
Hanson, D.R., and A.R. Ravishankara, Reactive uptake of ClONO2 onto sulfuric acid
due to reaction with HCl and H2 O. J Phys Chem: 98, 5728, 1994.
Hanst, P.L., L.L. Speller, D.M. Watts, J.W. Spence, and F.M. Miller, Infrared
measurements of fluorocarbons, carbon tetrachloride, carbonyl sulfide, and other
atmospheric trace gases. Air Pollution Control Assn J: 24, 12, 1220, 1975.
Harder, H., C. Camy-Peyret, F. Ferlemann, R. Fitzenberger, T. Hawat, H.
Osterkamp, D. Perner, U. Platt, M. Schneider, P. Vradelis, and K. Pfeilsticker,
Stratospheric BrO profiles measured at different latitudes and seasons: Atmospheric
observations. Geophys Res Lett: 25, 3843, 1998.
Harries, J.E., Measurements of stratospheric water vapor using infrared techniques,
J. Atmos. Sci., 30, 1691, 1973.
Hauglustaine, D.A., and D.H. Ehhalt, A three-dimensional model of molecular hy-
drogen in the troposphere. J Geophys Res: 107,
7 D17, 4330, doi:1029/2001JD001156,
2002.
Hays, P.B., and J.J. Olivero, Carbon dioxide and monoxide above the troposphere.
Heaps, W.S., and J.J. McGee, Balloon borne lidar measurements of stratospheric
hydroxyl radical. J Geophys Res: 86, 5281, 1983.
Hein, R., P.J. Crutzen, and M. Heinmann, An inverse modeling approach to
investigate the global atmospheric methane cycle. Global Biogeochem Cycles: 11,
43, 1997.
Herman, J., and W. Jaeschke, Measurements of H2 S and SO2 over the Atlantic ocean.
J Atmos Chem: 1, 111, 1984.
Hesstvedt, E., On the photochemistry of ozone in the ozone layer. Geophys Publ: 27,
7
4, 1968.
Hesstvedt, E., O. Høv, and I.S.A. Isaksen, Photochemistry of mixtures of
hydrocarbons and nitrogen oxides in air. Geophys Norv: 31, 27, 1978.
Hitchman, M.H., M. McKay, and C.R. Trepte, A climatology of stratospheric aerosol.

J Geophys Res: 99, 20,689, 1994.
Hoffmann, D.J., and J.M. Rosen, Balloon-borne Observations of Stratospheric Aerosol
and Condensation Nuclei During the Year Following the Mt. St. Helens Eruption.
University of Wyoming Report AP-63, 1981.
Hoffmann, D.J., and S. Solomon, Ozone destruction through heterogeneous chemistry
Hoffman, T., C.D. O’Dowd, and J.H. Seinfeld, Iodine oxide homogeneous nucleation:
An explanation for coastal new particle formation. Geophys Res Lett: 28, 1949, 2001.
Holton, J.R., P.H. Haynes, M.E. McIntyre, A.R. Douglass, R.B. Rood, and C. Pfister,
Holzinger, R., C. Warneke, A. Hansel, A. Jordan, W. Lindinger, D.H. Scharffe, G.
Schade, and P.J. Crutzen, Biomass burning as a source of formaldehyde, acetaldehyde,
methanol, acetone, acetonitrile, and hydrogen cyanide. Geophys Res Lett: 26, 1161,
1999.
Hood, L.L., Solar ultraviolet radiation induced variations in the stratosphere and
mesophere. J Geophys Res: 92, 876, 1987.
Hood, L.L., and J.L. Jirikowic, Stratospheric dynamical effects on solar ultraviolet
variations: Evidence from zonal mean ozone and temperature data. J Geophys Res:
96, 7565, 1991.
Hood, L.L., and J.P. McCormack, Components of interannual ozone change based on
NIMBUS 7 TOMS data. Geophys Res Lett: 19, 2309, 1992.
Hood, L.L., Z. Huang, and S.W. Bougher, Mesospheric effects of solar ultraviolet
variations: Further analysis of SME IR ozone and NIMBUS 7 SAMS temperature
data. J Geophys Res: 96, 12,989, 1991.
Hood, L.L., J.L. Jirikowic, and J.P. McCormack, Quasi-decadal variability of the
stratosphere: Influence of long-term solar ultraviolet variations. J Atmos Sci: 50,
3941, 1993.
Houweling, S., T. Kaminski, F. Dentener, J. Lelieveld, and M. Heinmann, Inverse
modeling of methane sources and sinks using the adjoint of a global transport model.
J Geophys Res: 104, 26,137, 1999.
Huang, T.Y.W., and G.P. Brasseur, The effect of long-term solar variability in a two-
dimensional interactive model of the middle atmosphere. J Geophys Res: 98, 20,410,
1993.
Hunt, B.G., Photochemistry of ozone in a moist atmosphere. J Geophys Res: 71,
1385, 1966.
Ingels, J.D. Nevejans, P. Frederick, and E. Arijs, Acetonitrile and sulfuric acid con-
centrations derived from ion composition measurements during the MAP/GLOBUS
1983 Campaign. Planet Space Sci: 35, 685, 1987.
Ingham, T., M. Cameron, and J.N. Crowley, Photodissociation of IO (355 nm) and
OIO (532 nm): Quantum yields for O3 P and I production. J Phys Chem: 104, 8001,
2000.
Inn, E.C.Y., J.F. Vedder, B.J. Tyson, and D. O’Hara, COS in the stratosphere.
IPCC, Climate Change: The IPCC Scientific Assessment, J.T. Houghton, C.J.
Jenkins, and J.J. Ephraums, eds., Intergovernmental Panel on Climate Change,
Cambridge University Press, Cambridge, UK, 1990.
IPCC, Climate Change 1995: The Science of Climate Change. Contribution of

Working Group I to the Second Assessment Report of the Intergovernmental Panel
on Climate Change, J.T.Houghton, L.G. Meira Filho, B.A. Callander, N. Harris,
A. Kattenberg, and K. Maskell, eds., Cambridge University Press, Cambridge, UK,
572 pp., 1996.
IPCC, Climate Change 2001: The Scientific Basis. Contribution of Working Group I
to the Third Assessment Report of the Intergovernmental Panel on Climate Change,
J.T. Houghton, Y. Ding, D.J. Griggs, M. Noguer, P.J. van der Linden, X. Dai,
K. Maskell, and C.A. Johnson, eds. Cambridge University Press, Cambridge, UK,
881 pp., 2001.
Jaeschke, W., R. Schmitt, and H.W. Georgi, Preliminary results of stratospheric SO2
Jackman, C.H., Effects of energetic particles on minor constituents in the middle
atmosphere. J Geomagn Geoelec: 43, 637, 1991.
Jackman, C.H., M.C. Cerniglia, J.E. Nielsen, D.J. Allen, J.M. Zawodny, R.D.
McPeeters, A.R. Douglass, J.E. Rosenfield, and R.B. Rood, 2-Dimensional and 3-
Dimensional model simulations, measurements, and interpretation of the influence of
the October 1989 solar proton events on the middle atmosphere. J Geophys Res: 100,
11,641, 1995.
Jacobson, M.Z., Computation of global photochemistry with SMVGEAR II. Atmos
Environ: 29, 2541, 1995.
Jacobson, M.Z., Improvement of the SMVGEAR II on vector and scalar machines
through absolute tolerance control. Atmos Environ: 32, 791, 1998.
Jacobson, M.Z., and R.P. Turco, SMVGEAR: A sparse-matrix vectorized gear code
for atmospheric models. Atmos Environ: 28A, 273, 1994.
Jensen, E.J., et al., High humidities and subvisible cirrus near the tropical tropopause,
Johnson, J.E., A.R. Bandy, D.C. Thornton, and T.S. Bates, Measurements of
atmospheric carbonyl sulfide during the NASA Chemical Instrumentation Test and
Evaluation Project: Implication for the global COS budget. J Geophys Res: 98,
23,443, 1993.
Johnston, H., Reduction of stratospheric ozone by nitrogen oxide catalysts from
supersonic transport exhaust. Science: 173, 517, 1971.
Johnston, H.S., and J. Podolske, Interpretation of stratospheric photochemistry. Rev
Jucks, K.W., D.G. Johnson, K.V. Chance, W.A. Traub, J.J. Margitan, G.B.
Osterman, R.J. Salawitch, and Y. Sasano, Observations of OH, HO2 , H2 O and O3 in
the upper stratosphere, implications for HOx chemistry. Geophys Res Lett: 25, 3935,
1998.
Kaye, J.A., S. Penkett, and F.M. Ormond, (eds.), Report on Concentrations, Lifetimes
and Trends of CFCs, Halons, and Related Species, NASA Reference Publication 1339,
NASA Office of Mission to Planet Earth, 247pp., National Aeronautics and Space
Administration, Washington, D.C., USA, 1994.
Keating, G.M., M.C. Pitts, G. Brasseur, and A. De Rudder, Response of middle
atmosphere to short-term solar ultraviolet variations: 1, Observations. J Geophys
Res: 92, 889, 1987.
Keyser, L.F., Absolute rate constant and branching fractions for the H + HO2 reaction
from 245 to 300 K. J Phys Chem: 90, 2994, 1986.
Khalil, M.A.K., and R.A. Rasmussen, Carbon monoxide in the Earth’s atmosphere:
Increasing trend. Science: 224, 54, 1990.
Kircher, C.C., and S.P. Sander, Kinetics of HO2 and DO2 disproportionations. J
Phys Chem: 88, 2082, 1984.
Kley, D., E.J. Stone, W.R. Henderson, J.W. Drummond, W.J. Harrop, A.L.
Schmeltekopf, T.L. Thompson, and R.H. Winkler, In situ measurements of the mixing
ratio of water vapor in the stratosphere. J Atmos Sci: 36, 2513, 1979.
Kley, D., J.W. Drummond, and A.L. Schmeltekopf, On the structure and
microstructure of stratospheric water vapor, in Atmospheric Water Vapor. A. Deepak,
T.D. Wilkerson, and L.H. Ruhnke, eds., Academic Press, 315, 1980.
Ko, M.K.W., N.D. Sze, and D.K. Weisenstein, Roles of dynamical and chemical
processes in determining the stratospheric concentration of ozone in one-dimensional
and two-dimensional models. J Geophys Res: 94, 9889, 1989.
Ko, M.K.W., N.D. Sze, C.J. Scott, and D.K. Weisenstein, On the relation between
stratospheric chlorine/bromine loading and short-lived tropospheric source gases. J
Geophys Res: 102, 25,507, 1997.
Kondo, Y., P. Aimedieu, M. Koike, Y. Iwasaka, P.A. Newman, U. Schmidt, W.A.
Mathews, M. Hayashi, and W.R. Sheldon, Reactive nitrogen, ozone, and nitrate
aerosols observed in the Arctic stratosphere in January 1990. J Geophys Res: 97,
7
13,025, 1992.
Kondo, Y., U. Schmidt, T. Sugita, A. Engel, M. Koike, P. Aimedieu, M.R.
Gunson, and J. Rodriguez, NOy correlation with N2 O and CH4 in the midlatitude
stratosphere. Geophys Res Lett: 23, 2369, 1996.
Koop, T., and K.S. Carslaw, Melting of H2 SO4 ·4H2 O particles upon cooling:
Implications for polar stratospheric clouds. Science: 272, 1638, 1996.
Koop, T., U.M. Biermann, W. Raber, B.P. Luo, P.J. Crutzen, and Th. Peter, Do
stratospheric aerosols freeze above the ice frost point? Geophys Res Lett: 22, 917,
1995.
Kroeze, C., A. Mosier, and L. Bouwman, Closing the N2 O budget: A restrospective
analysis. Global Biogeochem Cycles: 13, 1, 1999.
Kulmala, M., L. Pirjola, and J.M. Mäkel
¨ ä, Stable sulfate clusters as a source of new
atmospheric particles. Nature: 404, 66, 2000.
Langford, A.O., T.J. O’Leary, M.H. Proffit, and M.H. Hitchman, Transport of the
Pinatubo aerosol to a northern midlatitute site. J Geophys Res: 100, 9007, 1995.
Larsen, N., B.M. Knudsen, J.M. Rosen, N.T. Kjome, and E. Kyrö,¨ Balloon-borne
backscatter observations of type 1 PSC formation: Inference about physical state
from trajectory analysis. Geophys Res Lett: 23, 1091, 1996.
Larsen, N., B.M. Knudsen, J.M. Rosen, N.T. Kjome, R. Neuber, and E. Kyrö,
Temperature histories in liquid and solid polar stratospheric cloud formation. J
Geophys Res: 102, 23,505, 1997.
Lary, D.J., Gas phase bromine photochemistry. J Geophys Res: 101, 1505, 1996.
Lary, D.J., M.P. Chipperfield, R. Toumi, and T. Lenton, Heterogeneous atmospheric
bromine chemistry. J. Geophys Res: 101, 1489, 1996.
Lassey, K.R., D.C. Lowe, C.A.M. Brenninkmeijer, and A.J. Gomez, Atmospheric
methane and its carbon isotopes in the southern hemisphere: Their time series and
an instructive model. Chemosphere–Global Change Science: 26, 95, 1993.
Lazrus, A.L., and B.W. Gandrud, Distribution of stratospheric nitric acid vapor. J
Atmos Sci: 31, 1102, 1974.
Lazrus, A.L., B.W. Gandrud, R.N. Woodard, and W.A. Sedlacek, Stratospheric
halogen measurements. Geophys Res Lett: 2, 439, 1975.
Lazrus, A.L., B.W. Gandrud, R.N. Woodard, and W.A. Sedlacek, Direct
measurements of stratospheric chlorine and bromine. J Geophys Res: 81, 1067, 1976.
Lee, A.M., R.L. Jones, I. Kilbane-Dawe, and J.A. Pyle, Diagnosing ozone loss in the
extratropical lower stratosphere. J Geophys Res: 107,
7 10.1029/2001JD000538, 2002.
Lefevre,
f
f` F., G.P. Brasseur, I. Folkins, A.K. Smith, and P. Simon, Chemistry of the
1991-1992 stratospheric winter: Three-dimensional model simulations. J Geophys
Res: 99, 8183, 1994.
Lehmann, R., An algorithm for the determination of all significant pathways in
chemical reaction systems. J Atm Chem: 47,
7 45, 2004.
Le Texier, H., S. Solomon, and R.R. Garcia, The role of molecular hydrogen and
methane oxidation in the water vapor budget of the stratosphere. Quart J Roy Met
Soc: 114, 281, 1988.
Levy II, H., Normal atmosphere: Large radical and formaldehyde concentrations
predicted. Science: 173, 141, 1971.
Li, Q., D.J. Jacob, I. Bey, R.M. Yantosca, Y. Zhao, Y. Kindo, and J. Notholt,
Atmospheric hydrogen cyanide (HCN): Biomass burning source, ocean sink? Geophys
Res Lett: 27,
7 357, 2000.
Lii, R.R., M.C. Sauer, Jr., and S. Gordon, Temperature dependence of the gas-phase
self reaction of HO2 in the presence of H2 O. J Chem Phys: 85, 2833, 1981.
Liu, S.C., and T.M. Donahue, The aeronomy of hydrogen in the atmosphere of the
earth. J Atmos Sci: 31, 1118, 1974.
Livesey, N.J., J.W. Waters, R. Khosravi, G.P. Brasseur, G.S. Tyndall, and W.G.
Read, Stratospheric CH3 CN from the UARS Microwave Limb Sounder, Geophys Res
Lett: 28, 779, 2001.
Lobert, J.M., D.H. Scharffe, W.M. Hao, and P.J. Crutzen, Importance of biomass
burning in the atmospheric budgets of nitrogen-containing gases. Nature: 346, 552,
1990.
Loewenstein, M., J.R. Podolske, D.W. Fahey, E.L. Woodbridge, P. Tin, A. Weaver,
P.A. Newman, S.E. Strahan, S.R. Kawa, M.R. Schoeberl, and L.R. Lait. New
observations of the NOy /N2 O correlation in the lower stratosphere. Geophys Res
Lett: 20, 2531, 1993.
London, J., Radiative energy sources and sinks in the stratosphere and mesophere,
London, J., and G.M.W. Haurwitz, Ozone and sunspots. J Geophys Res: 68, 795,
1963.
Lopez-Puertas, M., M.A. Lopez-Valverde, R.R. Garcia, and R.G. Roble, A review of
CO2 and CO abundancies in the middle atmosphere, in Atmospheric Science Across
the Stratosphere, Geophysical Monograph 123, American Geophysical Union, 2000.
Lowe, D.C., C.A.M. Brenninkmeijer, G.W. Brailsford, K.R. Lassey, and A.J.
Gomez, Concentration and 13 C records of atmospheric methane in New Zealand and
Antarctica: Evidence for changes in methane sources. J Geophys Res: 99, 16,913,
1994.
Lovejoy, E.R., and D.R. Hanson, Measurement of the kinetics of reactive uptake by
submicron sulfuric acid particles. J Phys Chem: 99, 2080, 1995.
MacKenzie, A.R., M. Kulmala, A. Laaksonen, and T. Vesala, On the theories of

type 1 polar stratospheric cloud formation. J Geophys Res: 100, 11,275, 1995.
Mahieu, E., C.P. Rinsland, R. Zander, P. Demoulin, L. Delbouille, and G. Roland,
Vertical column abundances of HCN deduced from ground-based infrared solar
spectra: Long-term trends and variability. J Atmos Chem: 20, 299, 1995.
Mahieu, E., R. Zander, L. Delbouille, P. Demoulin, G. Roland, and C. Servais,
Observed trends in total vertical column abundances of atmospheric gases from IR
solar spectra recorded at the Jungfraujoch. J Atmos Chem: 28, 227, 1997.
Maiss, M., L.P. Steele, R.J. Francey, P.J. Fraser, R.L. Langenfelds, N.B.A. Trivelt,
and I. Levin, Sulfur hexafluoride, a powerful new atmospheric tracer. Atmos Environ:
30, 1621, 1996.
Manion, J.A., C. M. Fittschen, D.M. Golden, L.R. Williams, and M.A. Tolbert, Israel
J Chem: 34, 355, 1994.
Mankin, W.G., and M.T. Coffey, Latitudinal distributions and temporal changes of
stratosphere HCl and HF. J Geophys Res: 88, 10,776, 1983.
Mankin, W.G., M.T. Coffey, D.W.T. Griffith, and S.R. Drayson, Spectroscopic
measurement of carbonyl sulfide (OCS) in the stratosphere. Geophys Res Lett: 6,
853, 1979.
Mankin, W.G., M.T. Coffey, and A. Goldman, Airborne observations of SO2 , HCl,
and O3 in the stratospheric plume of the Pinatubo volcano in July, 1991. Geophys
Res Lett: 19, 179, 1992.
Manning, M.R., E.A.M. Brenningkmeijer, and W. Allan, The atmospheric carbon
monoxide budget of the southern hemisphere. J Geophys Res: 102, 10,673, 1997.
Mano,
˜ S., and M.O. Andreae, Emission of methyl bromide from biomass burning.
Science: 263, 1255, 1994.
Maroulis, P.J., A.L. Torres, and A.R. Bandy, Atmospheric concentrations of carbonyl
sulfide in the south-western and eastern United States. Geophys Res Lett: 4, 510,
1977.
Maroulis, P.J., A.L. Torres, A.B. Goldberg, and A.R. Bandy, Atmospheric SO2
measurements on project Gametag. J Geophys Res: 85, 7345, 1980.
Marsh, D.R., and J.M. Russell III, A tidal explanation of the sunrise/sunset anomaly
in HALOE low-latitude nitric oxide observations. Geophys Res Lett: 27, 7 3197, 2000.
Marsh, D.R., and R. Roble, TIME-GCM simulations of lower-thermospheric nitric
oxide seen by the halogen occultation experiment. J Atmos Terr Phys: 64, 889,
2002.
May, R.D., and C.R. Webster, In situ stratospheric measurements of HNO3 and
HCl near 30 km using the balloon-borne laser in situ sensor tunable diode laser
spectrometer. J Geophys Res: 94, 16,343, 1989.
McCormick, M.P., H.M. Steele, P. Hamill, W.P. Chu, and T.J. Swissler, Polar
stratospheric cloud sightings by SAM II. J Atmos Sci: 39, 1387, 1982.
McElroy, M.B., The Atmospheric Environment: Effects of Human Activities.
Princeton University Press, 2002.
McElroy, M.B., and J.C. McConnell, Nitrous oxide: A natural source of stratospheric
NO. J Atmos Sci: 28, 1095, 1971.
McElroy, M.B., R.J. Salawitch, S.C. Wofsy, and J.A. Logan, Reductions of Antarctic
ozone due to synergistic interactions of chlorine and bromine. Nature: 321, 759, 1986.
McKeen, S.A., S.C. Liu, and C.S. Kiang, On the chemistry of stratospheric SO2 from
volcanic eruptions. J Geophys Res: 89, 4873, 1984.
McLeod, H., G.P. Smith, and D.M. Golden, Photodissociation of pernitric acid
(HO2 NO2 ) at 248 nm. J Geophys Res: 93, 3813, 1988.
Meilinger, S.K., T. Koop, B.P. Luo, T. Huthwelker, K.S. Carslaw, U. Krieger, P.J.
Crutzen and Th. Peter, Size-dependent stratospheric droplet composition in leewave
temperature fluctuations and their potential role in PSC freezing. Geophys Res Lett:
22, 3031, 1995.
Meira, L.G., Jr., Rocket measurements of upper atmospheric nitric oxide and their
consequences to the lower ionosphere. J Geophys Res: 76, 202, 1971.
Meixner, F.X., The vertical sulfur dioxide distribution at the tropopause level. J
Atmos Chem: 2, 175, 1984.
Menzies, R.T., Remote measurement of ClO in the stratosphere. Geophys Res Lett:
6, 151, 1979.
Michelangeli, D.V., M. Allen, Y.L. Yung, R.L. Shia, D. Crisp, and J. Eluszkiewicz,
Enhancement of atmospheric radiation by an aerosol layer, J Geophys Res: 97,
7 865,
1992.
Migeotte, M.V., Methane in the earth’s atmosphere, Astrophys J: 107,
7 400, 1948.
Mlynzack, M.G., A new perspective on the molecular oxygen and hydroxyl airglow
emissions. J Geophys Res: 104, 27,535, 1999.
Mlynzack, M.G., and B.T. Marshall, A reexamination of the role of solar heating
in the O2 atmospheric and infrared atmospheric bands. Geophys Res Lett: 23, 657,
1996.
Molina, J.S., and F.S. Rowland, Stratospheric sink for chlorofluoromethanes: Chlorine
atom-catalyzed destruction of ozone. Nature: 249, 810, 1974a.
Molina, J.S., and F.S. Rowland, Predicted present stratospheric abundances of
chlorine species from photodissociation of carbon tetrachloride. Geophys Res Lett:
1, 309, 1974b.
Molina, L.T., and M.J. Molina, Production of Cl2 O2 from the self-reaction of ClO
radical. J Phys Chem: 91, 433, 1987.
Montzka, S.A., R.C. Myers, J.H. Butler, J.W. Elkins, and S.O. Cummings, Global
tropospheric distribution and calibration scale of HCFC-22, Geophys Res Lett: 20,
703, 1993.
Montzka, S.A., J.H. Butler, R.C. Myers, T.M. Thompson, T.H. Swanson, A.D. Clarke,
L.T. Lock, and J.W. Elkins, Decline in the tropospheric abundance of halogen from
halocarbons: Implications for stratospheric ozone depletion. Science: 272, 1318,
1996.
Mosier, A., C. Kroeze, C. Nevison, O. Oenema, S. Seitzinger, and O. Van Cleemput,
Closing the N2 O budget: Nitrous oxide emissions through the agricultural nitrogen
cycle — OECD/IPCC/IEA phase II development of IPCC guidelines for national gas
inventory methodology. Nutrient Cycling in Agroecosystems: 52, 225, 1998.
Mote, P.W., K.H. Rosenlof, M.E. McIntyre, E.S. Carr, J.C. Gille, J.R. Holton, J.S.
Kinnersley, H.C. Pumphrey, J.M. Russell III, and J.W. Waters, An atmospheric tape-
recorder: The imprint of tropical tropopause temperatures on stratospheric water
vapor. J Geophys Res: 101, 3989, 1996.
Moyers, J.L., and R.A. Duce, Gaseous and particulate iodine in the marine
atmosphere. J Geophys Res: 77,
7 5529, 1972.
Mozurkewich, M., and J.G. Calvert, Reaction probability of N2 O5 on aqueous
Muller,
¨ J.F., and G. Brasseur, Sources of upper tropospheric HOx : A three-
dimensional study. J Geophys Res: 104, 1705, 1999.
Murcray, D.G., D.D. Barker, J.N. Brooks, A.Goldman, and W.J. Williams, Seasonal
and latitudinal variations of the stratospheric concentration of HN O3 . Geophys Res
Lett: 2, 223, 1975.
Murcray, D.G., A. Goldman, W.J. Williams, F.H. Murcray, F.S. Bonono, G.M.
Bradford, G.R. Cole, P.L. Hanst, and M.J. Molina, Upper limit for stratospheric
ClONO2 from balloon-borne infrared measurements. Geophys Res Lett: 4, 227, 1977.
Naudet, J.P., D. Hugenin, P. Rigaud, and D. Cariolle, Stratospheric observations of
NO3 and its experimental and theoretical distribution between 20 and 40 km. Planet
Space Sci: 29, 707, 1981.
Nevison, C.D., S. Solomon, and R.R. Garcia, Model overestimates of NOy in the
upper stratosphere. Geophys Res Lett: 24, 803, 1997.
Newell, R. E., The circulation of the upper atmosphere. Sci Amer: 210, 62, 1964.
Newell, R.E., Y. Zhu, W.G. Read, and J.W. Waters, Relationship between tropical
upper tropospheric moisture and eastern tropical Pacific sea surface temperature at
seasonal and interannual time scales. Geophys Res Lett: 24, 25, 1997.
Nguyen, B.C., B. Bonsang, and A. Gaudry, The role of the ocean in the global
atmospheric sulfur cycle. J Geophys Res: 88, 10,903, 1983.
Nicolet, M., Ozone and hydrogen reactions. Ann Geophys: 26, 531, 1970.
Nicolet, M., Aeronomic reactions of hydrogen and ozone, in Mesospheric Models and
Related Experiments. G. Fiocco, ed., D. Reidel Publ., The Netherlands, 1971.
Nicolet, M., Stratospheric ozone: An introduction to its study. Rev Geophys Space
Phys: 13, 593, 1975a.
Nicolet, M., On the production of nitric oxide by cosmic rays in the mesosphere and
stratosphere. Planet Space Sci: 23, 637, 1975b.
Nightingale, R.W., A.E. Roche, J.B. Kumer, J.L. Mergenthaler, J.C. Gille, S.T.
Massie, P.L. Bailey, D.P. Edwards, M.R. Gunson, G.C. Toon, B. Sen, J.-F. Blavier,
and P.S. Connell, Global CF2 Cl measurements by UARS cryogenic limb array etalon
spectrometer: Validation by correlative data and a model. J Geophys Res: 101, 9711,
1996.
Novelli, P.C., K.A. Masarie, and P.M. Lang, Distributions and recent changes of
carbon monoxide in the lower troposphere. J Geophys Res: 103, 19,015, 1998.
Novelli, P.C., P.M. Lang, K.A. Masarie, D.F. Hurst, R. Myers, and J.W. Elkins,
Molecular hydrogen in the troposphere: Global distribution and budget. J Geophys
Res: 104, 30,427, 1999.
Noxon, J.F., Stratospheric N O2 , 2. Global behavior. J Geophys Res: 84, 5067, 1979.
Noxon, J.F., R.B. Norton, and W.R. Henderson, Observations of atmospheric N O3 .
Noxon, J.F., E.C. Whipple, Jr., and R.S. Hyde, Stratospheric NO2 , 1. Observational
method and behavior at midlatitude. J Geophys Res: 84, 5047, 1979.
Oelhaf, H., T.V. Clarmann, H. Fischer, F. Friedl-Vallon, C. Fritzsche, A. Linden, C.
Piesch, M. Seefeldner, and W. Volker, Stratospheric ClONO2 and HNO3 profiles
inside the Arctic vortex from MIPAS-B limb emission spectra obtained during
EASOE. Geophys Res Lett: 21, 1263, 1994.
Offermann, D., K.-U. Grossman, P. Varthol, P. Knieling, M. Riese, and R. Tank,
Cryogenic Infrared Spectrometers and Telescopes for the Atmosphere (CRISTA)
experiment and middle atmosphere variability. J Geophys Res: 104, 16,311, 1999.
Oltmans, S.J., and D.J. Hofmann, Increase in lower stratospheric water vapor at a
midlatitude Northern Hemisphere site from 1981 to 1994. Nature: 374, 146, 1995.
Oltmans, S.J., H. Volmel, D.J. Hofmann, K. Rosenlof, and D. Kley, The increase in
stratospheric water vapor from balloon-borne frost point hygrometer measurements
at Washington, DC and Boulder, Colorado. Geophys Res Lett: 27, 7 3453, 2000.
Oram, D.E., and S.A. Penkett, Observations in eastern England of elevated methyl
iodide concentrations in air of Atlantic origin. Atmos Environ: 28, 1159, 1994.
Orlando, J.J., G.S. Tyndall, G.K. Moortgat, and J.G. Calvert, Quantum yields for
NO3 photolysis between 570 and 635 nm. J Phys Chem: 97,7 10,996, 1993.
Paetzold, H.K., F. Piscalar, and H. Zschorner, Secular variation of the stratospheric
ozone layer over middle Europe during the solar cycles from 1951 to 1972. Nature:
240, 106, 1972.
Paneth, F.A., The chemical composition of the atmosphere. Quart J Roy Meteorol
Soc: 63, 433, 1937.
Park, J.H., and B. Carli, Spectroscopic measurements of HO2 , H2 O2 , and OH in the
stratosphere. J Geophys Res: 96, 22,535, 1991.
Perliski, L., S. Solomon, and J. London, On the interpretation of seasonal variations
in stratospheric ozone. Planet Space Sci: 37,
7 1527, 1989.
Peter, T., Microphysics and chemistry of polar stratospheric cloud particles. Ann Rev
Phys Chem: 48, 785, 1995.
Petron,
´ G., C. Granier, B. Khattatov, J.-F. Lamarque, V. Yudin, J.-F. M¨
uller,
¨ and J.
Gille, Inverse modeling of carbon monoxide surface emissions using Climate Modeling
and Diagnostics Laboratory network observations, J Geophys Res: 107, 7 D24, 4761,
doi:10.1029/2001JD001305, 2002.
Pfeilsticker, K., and U. Platt, Airborne measurements during the Arctic stratospheric
experiment: Observations of O3 and NO2 . Geophys Res Lett: 21, 1375, 1994.
Pfeilsticker, K., W.T. Sturges, H. Bösch,
¨ C. Camy-Peyret, M.P. Chipperfield,
A. Engel, R. Fitzenberger, M. Müller,
¨ S. Payan, and B.-M. Sinnhuber, Lower
stratospheric organic and inorganic bromine budget for the Arctic winter 1998/99.
Geophys Res Lett: 27,
7 3305, 2000.
Pierrehumbert, R.T., Lateral mixing as a source of subtropical water vapor, Geophys
Res Lett: 25, 151, 1998.
Pickett, H.M., and D.B. Peterson, Stratospheric OH measurements with a far-infrared
limb observing spectrometer. J Geophys Res: 98, 20,507, 1993.
Pickett, H.M., and D.B. Peterson, Comparison of measured stratospheric OH with
prediction. J Geophys Res: 101, 16,789, 1996.
Pitari, G., and V. Rizzi, An estimate of the chemical and radiative perturbation of
stratospheric ozone following the eruption of Mount Pinatubo. J Atmos Sci: 50,
3260, 1993.
Pommerau, J.P., and J. Piquard, Ozone and nitrogen-dioxide vertical distributions
by UV-visible solar occulation from balloons. Geophys Res Lett: 21, 1231, 1994.
Poole, L.R., and M.P. McCormick, Airborne lidar observations of Arctic polar
stratospheric clouds: Indications of two distinct growth stages. Geophys Res Lett:
15, 21, 1988.
Poulet, G., M. Pirre, F. Maguin, M. Ramarosson, and G. Le Bras, Role of BrO +
HO2 reaction in the stratospheric chemistry of bromine. Geophys Res Lett: 19, 2305,
1992.
Prather, M.J., and D.J. Jacob, A persistent inbalance between HOx photochemistry
of the upper troposphere driven by deep tropical convection. Geophys Res Lett: 24,
3189, 1997.
Prinn, R. G., R.F. Weiss, B.R. Miller, J. Huang, F.N. Alyea, D.M. Cunnold,
P.J. Fraser, D.E. Hartler, and P.G. Simmonds, Atmospheric trends and lifetime of
CH3 CCl3 and global OH concentrations. Science: 269, 187, 1995.
Pueschel, R.F., K.G. Snetsinger, J.K. Goodmann, O.B. Toon, G.V. Ferry, V.R.
Oberbeck, J.M. Livingston, S. Verma, W. Fong, W.L. Starr, and K.R. Chan,
Condensed nitrate, sulfate, and chloride in Antarctic stratospheric clouds. J Geophys
Res: 94, 11,271, 1989.
Pundt, I., J.P. Pommereau, C. Phillips, and E. Lateltin, Upper limit of iodine oxide
in the lower stratosphere. J Atmos Chem: 30, 173, 1998.
Quay, P.D., S.L. King, J. Stutsman, D.O. Wilbur, L.P. Steele, I. Fung, R.H. Gammon,
T.A. Brown, G.W. Farwell, P.M. Grootes, and F.H. Schmidt, Carbon isotopic
composition of atmospheric CH4 : Fossil and biomass burning source strengths. Global
Biogeochem Cycles: 5, 25, 1991.
Quay, P., J. Stutsman, D. Wilbur, A. Stover, E. Duglokencky, and T. Brown, The
isotopic composition of atmospheric methane. Global Biogeochem Cycles: 13, 445,
1999.
Ramanathan, K.R., Bi-annual variation of atmospheric ozone over the tropics, Quart
J Roy Met Soc: 89, 540, 1963.
Randel, W.J., F. Wu, S.J. Oltmans, K. Rosenlof, and G.E. Nedoluka, Interannual
changes of stratospheric water vapor and correlations with tropical tropospause
temperatures. J Atm Sci: 61, 2133, 2004.
Raper, O.F., C.B. Farmer, R.A. Toth, and B.D. Robbins, The vertical distribution
of HCl in the stratosphere. Geophys Res Lett: 4, 531, 1977.
Ravishankara, A.R., P.H. Wine, C.A. Smith, P.E. Barbone, and A. Torabi, N2 O5
photolysis: Quantum yields for NO3 and O(3 P ). J Geophys Res: 91, 5355, 1986.
Reichle, H.G., V.S. Connors, J.A. Holland, R.T. Sherrill, H.A. Wallio, J.C. Casas,
E.P. Condon, B.B. Gormsen, W. Seiler, The distribution of middle atmosphere carbon
monoxide during early October 1984. J Geophys Res: 95, 9845, 1990.
Rex, M., R.J. Salawitch, M.L. Santee, J.W. Waters, K. Hoppel, and R. Bevilacqua,
On the unexplained stratospheric ozone losses during cold Arctic Januaries. Geophys
Res Lett: 30, doi: 10.1029/2002GL016008, 2003.
Ridley, B.A., M. McFarland, J.T. Bruin, H.I. Schiff, and J.C. McConnell, Sunrise
measurements of stratospheric nitric oxide. Can J Phys: 55, 212, 1977.
Ridley, B.A., M. McFarland, A.L. Schmeltekopf, M.H. Proffitt, D.L. Albritton, R.H.
Winkler, and T.L. Thompson, Seasonal differences in the vertical distributions of NO,
NO2 , and O3 in the stratosphere near 50◦ N. J Geophys Res: 92, 11,919, 1987.
Riese, M., R. Spang, P. Preusse, M. Ern, M. Jarisch, D. Offermann, and K.-U.
Grossmann, Cryogenic Infrared Spectrometers and Telescope for the Atmosphere
(CRISTA) data processing and atmospheric temperature and trace gas retrieval. J
Geophys Res: 104, 16,349, 1999.
Rinsland, C.P., R. Zander, E. Mahieu, P. Demoulin, A. Goldman, D. Ehhalt, and
J. Rudolph, Ground-based infrared measurements of carbonyl sulfide total column
abundances: Long-term trends and variability. J Geophys Res: 97,
7 5995, 1992.
Rinsland, C.P., M.R. Gunson, R.J. Salawitch, H.A. Michelsen, R. Zander, M.J.
Newchurch, M.M. Abbas, M.C. Abrams, G.L. Manney, A.Y. Chang, F.W. Irion, A.
Goldman, and E. Mahieu, ATMOS/ATLAS-3 measurements of stratospheric chlorine
and reactive nitrogen partitioning inside and outside the November 1994 Antarctic
vortex. Geophys Res Lett: 23, 2365, 1996.
Rinsland, C.P., A. Goldman, F.J. Murcray, T.M. Stephen, M.S. Pougatchev, J.
Fischman, S.J. David, R.D. Blatherwick, P.C. Novelli, N.B. Jones, and B.J. Connor,
Infrared solar spectroscopic measurements of free tropospheric CO, C2 H5 and HCN
above Mauna Loa, Hawaii: Seasonal variations and evidence for enhanced emissions
from the southeast Asian tropical fires of 1997-1998. J Geophys Res: 104, 18,667,
1999.
Robinson, G.N., D.R. Worsnop, J.T. Jayne, and C.E. Kolb, Heterogeneous uptake of
ClONO2 and N2 O5 by sulfuric acid solutions. J Geophys Res: 102, 3583, 1997.
Roche, A.E., J.B. Kumer, J.L. Mergenthaler, G.A. Ely, W.G. Uplinger, J.F. Potter,
T.C. James, and L.W. Sterritt, The Cryogenic Limb Array Etalon Spectrometer,
CLAES on UARS: Experiment description and performance. J Geophys Res: 98,
10,763, 1993.
Roche, A.E., J.B. Kumer, J.L. Mergenthaler, R.W. Nightingale, W.G. Uplinger, G.A.
Ely, J.F. Potter, D.J. Wuebbles, P.S. Connell, and D.E. Kinnison, Observations of
lower stratospheric ClONO2 , HNO3 , and aerosol by the UARS CLAES experiment
between January 1992 and April 1993. J Atmos Sci: 51, 2877, 1994.
Rosen, J.M., N.T. Kjome, N. Larsen, B.M. Knudsen, E. Kyrö, ¨ R. Kivi, J. Karhu,
R. Neuber, and I. Benigna, Polar stratospheric cloud threshold temperatures in the
1995-1996 Arctic vortex. J Geophys Res: 102, 28,195, 1997.
Rosenlof, K.H., S.J. Oltmans, D. Kley, J.M. Russell III, E.W. Chiou, W.P. Chu,
D.G. Johnson, K.K. Kelly, H.A. Michelsen, G.E. Nedoluha, E.E. Remsberg, G.C.
Toon, and M.P. McCormick, Stratospheric water vapor increases over the past half
century. Geophys Res Lett: 28, 1195, 2001.
Rossi, M.J., R. Malhotra, and D.M. Golden, Heterogeneous chemical reaction of
chlorine nitrate and water on sulfuric-acid surfaces at room temperatures. Geophys
Res Lett: 14, 127, 1987.
Rowland, F.S., J.E. Spencer, and M.J. Molina, Stratospheric formation and photolysis
of chlorine nitrate. J Phys Chem: 80, 2711, 1976.
Rudolph, J., D.H. Ehhalt, and A. Tonissen, Vertical profiles of ethane and propane
in the stratosphere. J Geophys Res: 86, 7767, 1981.
Russell, J.M., III, L.L. Gordley, J.H. Park, S.R. Drayson, D.H. Hesketh, R.J. Cicerone,
A.F. Tuck, J.E. Frederick, J.E. Harries, and P.J. Crutzen, The Halogen Occultation
Experiment. J Geophys Res: 98, 10,777, 1993.
Russell, P.B., J.M. Livingston, R.F. Pueschel, J.B. Pollack, S. Brooks, P. Hamill, J.
Hughes, L. Thomason, L. Stowe, T. Deshler, and E. Dutton, Global to microscale
evolution of the Pinatubo volcanic aerosol, derived from diverse measurements and
analyses. J Geophys Res: 101, 18,745, 1996.
Salcedo, D., L.T. Molina, and M.J. Molina, Homogeneous freezing of concentrated
aqueous nitric acid solutions at polar stratospheric temperatures. J Phys Chem: 105,
1433, 2001.
Sandalls, F.J., and S.A. Penkett, Measurements of carbonyl sulfide and carbon
disulfide in the atmosphere. Atmos Environ: 11, 197, 1977.
Schauffler, S.M., L.E. Heidt, W.H. Pollack, T.M. Gilpin, J.F. Vedder, S. Solomon,
R.A. Lueb, and E.L. Atlas, Measurements of halogenated organic compounds near
the tropical tropopause. Geophys Res Lett: 20, 2567, 1993.
Schmeltekopf, A.L., D.L. Albritton, P.J. Crutzen, P.D. Goldan, W.J. Harrop, W.R.
Henderson, J.R. McAfee, M. McFarland, H.I. Schiff, T.L. Thompson, D.L. Hofmann,
and N.T. Kjome, Stratospheric nitrous oxide altitude profiles at various latitudes. J
Atmos Sci: 34, 729, 1977.
Schmidt, U., Molecular hydrogen in the atmosphere. Tellus: 26, 78, 1974.
Schmidt, U., G. Kulessa, and E. P. Roth, The Atmospheric H2 cycle, Proc of the
NATO Advanced Institute on Atmospheric Ozone, A.C. Aikin, ed., U.S. Dept. of
Scholz, T.G., D.H. Ehhalt, L.E. Heidt, and E.A. Martell, Water vapor molecular
hydrogen, methane, and tritium concentrations near the stratopause. J Geophys Res:
75, 3049, 1970.
Seiler, W., and P.J. Crutzen, Estimates of gross and net fluxes of carbon between the
biosphere and the atmosphere from biomass burning. Clim Change: 2, 207, 1980.
Seiler, W., and R. Conrad, Contribution of tropical ecosystems to the global budget
of trace gases, especially CH4 , H2 , CO and N2 O, in The Geophysiology of Amazonia,
R. Dickinson, ed., John Wiley, pp. 133-160, 1987.
Sen, B., G.C. Toon, G.B. Osterman, J.-F. Blavier, J.J. Margitan, R.J. Salawitch, and
G.K. Yue, Measurements of reactive nitrogen in the stratosphere. J Geophys Res:
103, 3571, 1998.
Shah, G.M., Quasi-biennial oscillation in ozone. J Atmos Sci: 24, 396, 1967.
Singh, H.B., L.J. Salas, and R.E. Stiles, Methyl halides in and over the Eastern
Pacific. J Geophys Res: 88, 3684, 1983.
Singh, H.B., M. Kanakidou, P.J. Crutzen, and D.J. Jacob, High concentrations and
photochemical fate of oxygenated hydrocarbons in the global troposphere. Nature:
378, 50, 1995.
Siskind, D.E., J.T. Bacmeister, M.E. Summers, and J.M. Russell III, Two-dimensional
calculations of nitric oxide transport in the middle atmosphere and comparison with
halogen occultation experiment data. J Geophys Res: 102, 3527, 1997.
Solomon, S., and R.R. Garcia, Simulation of NOx partitioning along isobaric parcel
trajectories. J Geophys Res: 88, 5497, 1983.
Solomon, S., P.J. Crutzen, and R.G. Roble, Photochemical coupling between the
thermosphere and the lower atmosphere, I. Odd nitrogen from 50 to 120 km. J
Geophys Res: 87,
7 7206, 1982.
Solomon, S., R.R. Garcia, F.S. Rowland, and D.J. Wuebbles, On the depletion of
Antarctic ozone. Nature: 321, 755, 1986.
Solomon, S., G.H. Mount, R.W. Sanders, and A.L. Schmeltekopf, Visible spectroscopy
at McMurdo Station, Antarctica, 2. Observation of OClO. J Geophys Res: 92, 8329,
1987.
Solomon, S., H.L. Miller, J.P. Smith, R.W. Sanders, G.H. Mount, A.L. Schmeltekopf,
and J.F. Noxon, Atmospheric NO3 : 1. Measurement technique and the annual cycle
at 40◦ N. J Geophys Res: 94, 11,041, 1989a.
Solomon, S., R.W. Sanders, M.A. Carroll, and A.L. Schmeltekopf, Visible and near-
ultraviolet spectroscopy at McMurdo Station, Antarctica: 5. Observations of the
diurnal variations of BrO and OClO. J Geophys Res: 94, 11,393, 1989b.
Solomon, S., R.R. Garcia, and A.R. Ravishankara, On the role of iodine in ozone
depletion. J Geophys Res: 99, 20,491, 1994.
Solomon, S.C., C.A. Barth, and S.M. Bailey, Auroral production of nitric oxide
measured by the SNOE satellite. Geophys Res Lett: 26, 1259, 1999.
Spreng, S., and F. Arnold, Balloon-borne mass spectrometer measurements of

HNO3 and HCN in the winter Arctic stratosphere–Evidence for HNO3 -processing
by aerosols. Geophys Res Lett: 21, 1251, 1994.
Stachnik, R.A., J.C. Hardy, J.A. Tarsala, J.W. Waters, and N.R. Erickson,
Submillimeter heterodyne measurements of stratospheric ClO, HCl, and HO2 : First
results. Geophys Res Lett: 19, 1931, 1992.
Stanford, J.L., and J.S. Davis, A century of stratospheric cloud reports: 1870-1972.
Bull Amer Meteorol Soc: 55, 213, 1974.
Steele, H.M., P. Hamill, M.P. McCormick, and T.J. Swissler, The formation of polar
stratospheric clouds. J Atmos Sci: 40, 2055, 1983.
Steele, H.M., K. Drdla, R.P. Turco, J.D. Lumpe, R.M. Bevilacqua, Tracking polar
stratospheric cloud development with POAM II and a microphysical model. Geophys
Res Lett: 26, 287, 1999.
Stimpfle, R.M., P.O. Wennberg, L.B. Lafson, and J.G. Anderson, Simultaneous in
situ measurements of OH and HO2 in the stratosphere. Geophys Res Lett: 17,
7 1905,
1990.
Canad J Chem: 52, 1610, 1974.
Stolarski, R.R., and R.D. Rundel, Fluorine chemistry in the stratosphere. Geophys
Res Lett: 2, 443, 1975.
SPARC (Stratospheric Processes and their Role in Climate), Assessment of Upper
Tropospheric and Stratospheric Water Vapor, WMO/TD 1043, SPARC Office,
Verrieres le Buisson, France, 2000.
Summers, M.E., D.F. Strobel, R.M. Bevilacqua, X. Zhu, M.T. DeLand, M. Allen, and
G.M. Keating, A model study of the response of mesopheric ozone to the short-term
solar ultraviolet flux variations. J Geophys Res: 95, 22,523, 1990.
Summers, M.E., R.R. Conway, D.E. Siskind, M.H. Stevens, D. Offermann, M.
Riese, P. Preusse, D.F. Strobel, and J.M. Russell III, Implications of satellite OH
observations for middle atmospheric H2 O and ozone. Science: 227,
7 1967, 1997.
Summers, M.E., R.R. Conway, C.R. Englert, D.E. Siskind, M.H. Stevens, J.M.
Russell III, L.L. Gordley, and M.J. McHugh, Discovery of a water vapor layer in
the Arctic summer mesosphere: Implications for polar mesospheric clouds. Geophys
Res Lett: 28, 3601, 2001.
Swanson, T.H., J.W. Eltkins, J.H. Butler, S.A. Montzka, R.C. Myers, T.M.
Thompson, T.J. Baring, S.O. Cummins, G.S. Dutton, A.H. Hayden, J.M. Lobert,
G.A. Holcomb, W.T. Sturges, and T.M. Gilpin, 5. Nitrous oxide and halocarbons
division, in Climate Monitoring and Diagnostics Laboratory Summary Report No.
21, 1992. J.T. Petersen and R.M. Rosson (eds.), 59-75, NOAA Environ. Res. Labs.,
Boulder, Colo, USA, 1993.
Sze, N.D., Stratospheric fluorine: A comparison between theory and measurements.
Sze, N.D., and M.K.W. Ko, Photochemistry of COS, CS2 , CH3 SCH3 and H2 S:
Implications for the atmospheric sulfur cycle. Atmos Environ: 14, 1223, 1980.
Tabazadeh, A., and R.P. Turco, Stratospheric chlorine injection by volcanic eruptions:
HCl scavenging and implications for ozone. Science: 260, 1082, 1993.
Tabazadeh, A., and O.B. Toon, Freezing behavior of stratospheric sulfate aerosols
inferred from trajectory studies. Geophys Res Lett: 22, 1725, 1995.
Tabazadeh, A., and O.B. Toon, The presence of metastable HNO3 /H2 O solid phases
in the stratosphere inferred from ER-2 data. J Geophys Res: 101, 9071, 1996.
Tabazadeh, A., R.P. Turco, and M.Z. Jacobson, A model for studying the composition
and chemical effects of stratospheric aerosols. J Geophys Res: 99, 12,897, 1994.
Tabazadeh, A., O.B. Toon, and P. Hamill, Freezing behavior of stratospheric sulfate
aerosols inferred from trajectory studies. Geophys Res Lett: 22, 1725, 1995.
Tabazadeh, A., O.B. Toon, B.L. Gary, J.T. Bacmeister, and M.R. Schoeberl,
Observational constraints on the formation of type Ia polar stratospheric clouds.
Thomas, R.J., C.A. Barth, G.J. Rottman, D.W. Rusch, G.H. Mount, G.M. Lawrence,
R.W. Sanders, G.E. Thomas, and L.E. Clemens, Ozone density in the mesosphere (50-
90 km) measured by the SME near infrared spectrometer. Geophys Res Lett: 10, 245,
1983.
Thomason, L.W., L.R. Poole, and T. Deshler, A global climatology of stratospheric
aerosol surface area density deduced from Stratospheric Aerosol and Gas Experiment
II measurements: 1984-1994. J Geophys Res: 102, 8967, 1997.
Thornton, D.C., A.R. Bandy, and A.R. Driedger III, Sulfur dioxide over the western
Atlantic Ocean. Glob Biogeochem Cycles: 1, 317, 1987.
Tie, X., G.P. Brasseur, B. Briegleb, and C. Granier, Two-dimensional simulation of
Pinatubo aerosol and its effect on stratospheric ozone. J Geophys Res: 99, 20,545,
1994.
Tisone, G.C., Measurements of NO densities during sunrise at Kauai. J Geophys Res:
78, 746, 1973.
Tohmatsu, T., and N. Iwagami, Measurement of nitric oxide distribution in the upper
atmosphere. Space Res: 15, 241, 1975.
Tohmatsu, T., and N. Iwagami, Measurement of nitric oxide abundance in equatorial
upper atmosphere. J Geomagn Geoelec: 28, 343, 1976.
Tolbert, M.A., and O.B. Toon, Solving the PSC mystery. Science: 292, 61, 2001.
Tolbert, M.A., M.J. Rossi, and D.M. Golden, Heterogeneous interactions of chlorine
nitrate, hydrogen chloride and nitric acid with sulfuric acid surfaces at stratospheric
temperatures. Geophys Res Lett: 15, 847, 1988.
Tolbert, M.A., B.G. Koehler, and A.M. Middlebrook, Spectroscopic studies of model
polar stratospheric cloud films. Spectrochimica Acta: 48A, 1303, 1992.
Toohey, D.W., J.G. Anderson, W.H. Brune, and K.R. Chan, In situ measurements of
BrO in the arctic stratosphere. Geopohys Res Lett: 17,
7 513, 1990.
Toon, G.C., C.B. Farmer, P.W. Schafer, L.L. Lowes, and R.H. Norton, Composition
measurements of the 1989 Arctic stratosphere by airborne infrared solar absorption
spectroscopy. J Geophys Res: 97,
7 7939, 1992a.
Toon, G.C., C.B. Farmer, P.W. Schafer, L.L. Lowes, R.H. Norton, M.R. Schoeberl,
L.R. Lait, and P.A. Newman, Evidence for subsidience in the 1989 Arctic winter
stratosphere from airborne infrared composition measurements. J Geophys Res: 97,
7
7963, 1992b.
Toon, O.B., P. Hamill, R.F. Turco, and J. Pinto, Condensation of HNO3 and HCl in
the polar stratospheres. Geophys Res Lett: 13, 1284, 1986.
Toon, O.B., E.V. Browell, S. Kinne, and J. Jordan, An analysis of Lidar observations
of polar stratospheric clouds. Geophys Res Lett: 17,
7 393, 1990.
Toon, O.B., A. Tabazadeh, E.V. Browell, and J. Jordan, Analysis of lidar observations
of arctic polar stratospheric clouds. J Geophys Res: 105, 20,598, 2000.
Toumi, R., R.L. Jones, and J.A. Pyle, Stratospheric ozone depletion by ClONO2
photolysis. Nature: 365, 37, 1993.
Tsias, A., A.J. Prenni, K.S. Carslaw, T.P. Onasch, B.P. Luo, M.A. Tolbert, and T.
Peter, Freezing of polar stratospheric clouds in orographically induced strong warming
events. Geophys Res Lett: 24, 2303, 1997.
Tung, K.K., On the two-dimensional transport of stratospheric trace gases in
isentropic coordinates. J Atmos Sci: 39, 2330, 1982.
Tung, K.K., M.K.W. Ko, J.M. Rodriguez, and N.D. Sze, Are Antarctic ozone
variations a manifestation of dynamics or chemistry? Nature: 333, 811, 1986.
Turco, R.P., P. Hamill, O.B. Toon, R.C. Whitten, and C.S. Kiang, A one-dimensional
model describing aerosol formation and evolution in the stratosphere, I. Physical
processes and mathematical analogs, J Atmos Sci: 36, 399, 1979.
Turco, R.P., R.C. Whitten, O.B. Toon, J.B. Pollack, and P. Hamill, OCS,
stratospheric aerosols and climate. Nature: 283, 283, 1980.
Turco, R.P., R.C. Whitten, and O.B. Toon, Stratospheric aerosols: Observations and
theory. Rev Geophys Space Sci: 20, 233, 1982.
Voigt, C., J. Schreiner, A. Kohlmann, P. Zink, K. Mauersberger, N. Larsen, T.
Deshler, C. Kroger,
¨ J. Rosen, A. Adriani, F. Cairo, G. Di Donfrancesco, M. Viterbini,
J. Ovarlez, H. Overlez, C. David, and A. Dornbrack,
¨ Nitric acid trihydrate (NAT) in
polar stratospheric cloud particles. Science: 290, 1756, 2000.
Vömel
¨ H. et al., Balloon-borne observations of water vapor and ozone in the
tropical upper troposphere and lower stratosphere. J Geophys Res: 107,
7 4210, doi:
10.1029/2001JD000707, 2002.
von Cossart, G., J. Fielder, and V. von Zahn, Size distributions of NLC particles as
determined from 3-color observations of NLC by ground-based lidar. Geophys Res
Lett: 26, 1513, 1999.
Wallace, L., and W. Livingston, The effect of the Pinatubo cloud on hydrogen chloride
and hydrogen fluoride. Geophys Res Lett: 19, 1209, 1992.
Wamsley, P.R., J.W. Elkins, D.W. Fahey, G.S. Dutton, C.M. Volk, R.C. Myers, S.A.
Montzka, J.H. Butler, A.D. Clarke, P.J. Fraser, L.P. Steele, M.P. Lucarelli, E.L. Atlas,
S.M. Schauffler, D.R. Blake, F.S. Rowland, W.T. Sturges, J.M. Lee, S.A. Penkett, A.
Engel, R.M. Stimpfle, K.R. Chan, D.K. Weisenstein, M.K.W. Ko, and R.J. Salawitch,
Distribution of halon-1211 in the upper troposphere and lower stratosphere and the
1994 total bromine budget. J Geophys Res: 103, 1513, 1998.
Wahner, A., and C. Schiller, Twilight variation of veritcal column abundances of
OClO and BrO in the North Polar region. J Geophys Res: 97, 7 8047, 1992.
Waters, J.W., J.J. Gustincic, and P.N. Swanson, Atmospheric Water Vapor. A.
Deepak, ed., Academic Press, 1980.
Waters, J.W., J.C. Hardy, R.F. Jarnot, and H.M. Pickett, Chlorine monoxide
radical, ozone and hydrogen peroxide: Stratospheric measurements by microwave
limb sounding. Science: 214, 61, 1981.
Waters, J.W., L. Froidevaux, W.G. Read, G.L. Manney, L.S. Elson, D.A. Flower,
R.F. Janot, and R.S. Harwood, Stratospheric ClO and ozone from the Microwave
Limb Sounder on the Upper Atmosphere Research Satellite. Nature: 362 , 597, 1993.
Weaver, A., S. Solomon, R.W. Sanders, K. Arpag, and H.L. Miller, Jr., Atmospheric
NO3 , 5. Off-axis measurements at sunrise: Estimates of tropospheric NO3 at 40◦ N.
J Geophys Res: 101, 18,605, 1996.
Weiss, R.F., Changing global concentration of atmospheric nitrous oxide, Proc. Int.
Sym. on Global Cycles of Atmospheric Greenhouse Gases, Sendai, Japan, 78, 1994.
Weiss, P.S., J.E. Johnson, R.H. Gammon, and T.S. Bates, Re-evaluation of the open
ocean source of carbonyl sulfide to the atmosphere. J Geophys Res: 100, 23,083,
1995.
Wennberg, P.O., R.M. Stimple, E.M. Weinstock, A.E. Dessler, S.A. Lloyd, L.B.
Lapson, J.J. Schwab, and J.G. Anderson. Simultaneous in situ measurements of
OH, HO2 , O3 , and H2 O: A test of modeled stratospheric HOx chemistry. Geophys
Res Lett: 17,
7 1909, 1990.
Wennberg, P.O., R.C. Cohen, R.M. Stimpfle, J.P. Koplow, J.G. Anderson, R.J.
Salawitch, D.W. Fahey, E.L. Woodbridge, E.R. Keim, R.S. Gao, C.R. Webster, R.D.
May, D.W. Toohey, L.M. Avallone, M.H. Proffitt, M. Loewenstein, J.R. Podolske,
K.R. Chan, and S.C. Wofsy, Removal of stratospheric O3 by radicals: In situ
measurements of OH, HO2 , NO, NO2 , ClO, and BrO. Science: 266, 398, 1994.
Wennberg, P.O., J.W. Brault, T.F. Hanisco, R.J. Salawitch, and G.H. Mount, The
atmospheric column abundance of IO: Implications for stratospheric ozone. J Geophys
Res: 102, 8887, 1997.
Whitten, R.C., O.B. Toon, and R.P. Turco, The stratospheric sulfate aerosol layer:
Processes, models observations and simulations. Pageoph: 118, 86, 1980.
Willett, H.C., The relationship of total atmospheric ozone to the sunspot cycle. J
Geophys Res: 67,
7 661, 1962.
Willett, H.C., and J. Prohaska, Further evidence of sunspot-ozone relationships. J
Atmos Sci: 22, 493, 1965.
Williams, L.R., J.A. Manion, D.M. Golden, and M.A. Tolbert, Laboratory
measurements of heterogeneous reactions on sulfuric acid surfaces. J Appl Meteorol:
33, 785, 1994.
Williams, W.J., J.J. Kostus, A. Goldman, and D.G. Murcray, Measurements of the
stratospheric mixing ratio of HCl using an infrared absorption technique. Geophys
Res Lett: 3, 383, 1976.
Wine, P.H., J.R. Wells, and A.R. Ravishankara, Channel specific rate constants for
reaction of O (1 D) with HCl and HBr. J Chem Phys: 84, 1349, 1986.
Witt, G., J.E. Dye, and N. Wilhelm, Rocket-borne measurements of scattered sunlight
in the mesosphere. J Atmos Terr Phys: 38, 223, 1976.
Wittrock, F., R. Mueller, A. Richter, H. Bovensmann, and J.P. Burrows,
Measurements of iodine monoxide (IO) above Spitsbergen. Geophys Res Lett: 27,
7
1471, 2000.
Wofsy, S.C, M.B. McElroy, and Y.L. Yung, The chemistry of atmospheric bromine.
World Meteorological Organization (WMO), Atmospheric Ozone, Report No. 16,
Geneva, 1985.
World Meteorological Organization/United Nations Environment Programme
(WMO/UNEP), Scientific assessment of ozone depletion: 1991, Report No. 25,
Geneva, 1991.
(WMO/UNEP), Scientific assessment of ozone depletion: 1994, Report No. 37,
Geneva, 1995.

(WMO/UNEP), Scientific Assessment of Ozone Depletion: 1998, Report No. 44,
Geneva, 1999.
Worsnop, D.R., L.E. Fox, M.S. Zahniser, and S.C. Wofsy, Vapor-pressures of solid
hydrates of nitric-acid. Implications for polar stratospheric clouds. Science: 259, 71,
1993.
Yang, H., and K.K. Tung, On the phase propagation of extratropical ozone quasi-
biennial oscillation in observational data. J Geophys Res: 100, 9091, 1995.
Yung, Y.L., J.P. Pinto, R.T. Watson, and S.P. Sander, Atmospheric bromine and
ozone perturbations in the lower stratosphere. J Atmos Sci: 37,
7 339, 1980.
Zander, R., Recent observations of HF and HCl in the upper stratosphere. Geophys
Res Lett: 8, 413, 1981.
Zander, R., H. Leclercq, and L.D. Kaplan, Concentrations of carbon monoxide in the
upper troposphere. Geophys Res Lett: 8, 365, 1981.
Zander, R., M.R. Gunson, and C.B. Farmer, Remote sensing of the earth’s atmosphere
by infrared absorption spectroscopy — An update of the ATMOS program, in:
Advanced Optical Instrumentation for Remote Sensing of the Earth’s Surface from
Space, Proc. SPIE: 1129, 10, 1989.
Zander, R., D.H. Ehhalt, C.P. Rinsland, U. Schmidt, E. Mahieu, J. Rudolph,
P. Demoulin, G. Roland, L. Delbouille, and A.J. Sauval, Secular trend and
seasonal variability of the column abundance of N2 O above the Jungfraujoch station
determined from IR solar spectra. J. Geophys Res: 99, 16,745, 1994.
Zander, R. E. Mahieu, M.R. Gunson, M.C. Abrams, A.Y. Chang, M. Abbas, C.
Aellig, A. Engel, A. Goldman, F.W. Irion, N. Kämpfer,
¨ H.A. Michelsen, M.H.
Newchurt, C.P. Rinsland, R.J. Salawitch, G.P. Stiller, and G.C. Toon, The 1994
northern midlatitude budget of stratospheric chlorine derived from ATMOS/ATLAS-
3 observations. Geophys Res Lett: 23, 2357, 1996.
Zellner, R., Chemistry of the stratosphere, in Global Aspects of Atmospheric
Chemistry. R. Zellner, ed., Springer Verlag, 1999.
Zhao, Y., Y. Kondo, F.J. Murcray, X. Liu, M. Koike, H. Irie, K. Strong, K. Suzuki,
M. Sera, and Y. Ikegami, Seasonal variations of HCN over northern Japan measured
by ground-based infrared spectroscopy. Geophys Res Lett: 27,
7 2085, 1999.
Zhu, Y., R.E. Newell, and W.G. Read, Factors controlling upper-troposphere water
vapor. J Clim: 13, 836, 2000.
Chapter 6
OZONE PERTURBATIONS
6.1 Introduction
The previous chapters describe the array of physical and chemical

processes that can influence the odd oxygen family (and thereby its
most prominent member, ozone). The photochemical lifetime of odd
oxygen is less than a day throughout the region from about 35 to 70 km
in the sunlit atmosphere (see Figure 5.3). This is much faster than those
of the corresponding zonally-averaged transport processes, so that this
region provides a testbed for photochemical theory.
Section 6.2 discusses how local temperature, HOx , NOx , and
ClOx chemistries affect ozone at chemically-controlled altitudes in the
upper stratosphere and mesosphere, providing important constraints
on scientific understanding of ozone perturbations. The observation of
trends in upper stratospheric ozone has now demonstrated the long-term
impact of chlorofluorocarbons on ozone in the upper stratosphere, as was
first predicted in the 1970s. The vertical and meridional structure of the
observed ozone trends in this region also yield important insights into
chlorine-catalyzed ozone destruction. These will be a particular focus
of Section 6.2, which discusses how ozone in the chemically controlled
regime can be perturbed by changes of temperature, solar radiation, and
the chemical families described in Chapter 5.
Below about 25 km, the chemical time scale for odd oxygen increases
to months or years in most locations, so that photochemistry is not
the only component of the zonally-averaged ozone budget of the lower
stratosphere. This dramatic change in lifetime is due largely to the
attenuation of the solar radiation field available to drive photochemistry
at deeper and deeper altitudes of penetration (see Chapter 4). Where
the chemical lifetime of odd oxygen exceeds a month (below about
25 km), it becomes comparable to the time scale characterizing the
zonally-averaged transport by the mean meridional circulation. The
bulk of the ozone column resides in this dynamically-sensitive region
as discussed in Section 5.2.3, so that dynamical processes can play a
critical role in modulating the total ozone column not only from day
to day, but also on a seasonal and interannual basis. Any chemically-
443
driven change in ozone at these altitudes depends upon how quickly

chemistry can remove the ozone as it flows, i.e., on the potential of
chemistry to destroy ozone ‘in transit’. If chemistry is negligibly slow
compared to dynamics, then ozone will pass through without significant
alteration. If chemical processes can remove a portion of the moving
ozone flowing through this region, then the column abundance will
decrease. Even at altitudes as low as 20 km, chemical processes can
sometimes become remarkably fast, for example, when surface chemistry
(e.g., on polar stratospheric clouds in Antarctic spring) comes into play.
Section 6.3 discusses lower stratospheric ozone depletion. Section 6.3.1
describes the Antarctic ozone hole, focusing on the observations that
have shown that chlorine (and, to a lesser extent, bromine) compounds
are the primary cause of this spectacular change in ozone’s behavior since
about 1980. Sections 6.3.2 and 6.3.3 focus on the complex interplays
between dynamics and chemistry that affect Arctic and mid-latitude
ozone depletion, respectively. Section 6.4 is a summary of the chapter
and presents a look ahead to the likely future of the ozone layer to be
expected in the 21st century.
6.2 The Photochemically-Controlled Upper

Stratosphere and Mesosphere: 25-75 km
Chapter 5 presented the components of the continuity equation for
odd oxygen over a range of altitudes. Figure 6.1 shows how each of the
rate-limiting steps in the catalytic cycles that affect ozone varies with
altitude. The figure illustrates the competing roles of NOx and ClOx -
driven catalytic loss processes from about 30 to 45 km, the emerging
importance of purely Ox -driven loss near 40-55 km, and the domination
of ozone chemistry by HOx -driven catalytic cycles in the mesosphere.
Observations have provided important support for much of the key
chemistry controlling upper stratospheric and mesospheric ozone, not
only in terms of measurements of, for example, the abundance and
distribution of OH, ClO, NO2 , and other trace gases as discussed in
Chapter 5 (e.g., Conway et al., 2000; Waters et al., 1983; Anderson
et al., 1977; Russell et al., 1996), but also through some direct and
dramatic responses to perturbations, detailed in following sections.
6.2.1 Temperature Variability and Ozone Response
It is well known that the primary source of heat in the stratosphere

and mesosphere is the absorption of solar ultraviolet radiation by ozone
OZONE PERTURBATIONS 445
Figure 6.1. Daytime averaged rates of destruction of ozone at 40◦ N in winter

by a range of chemical processes for 1998 levels of reactive chemicals, from the
numerical model of Garcia and Solomon (see e.g., Solomon et al., 1998, and
references therein).
(see Chapter 3); indeed, if it were not for the presence of ozone, the
Earth would not have a stratosphere at all. If the concentration of ozone
changes due to some perturbation (such as NOx increases due to solar
proton events, see next section), then the temperature will change in
response. A simple, linearized description of this linkage was suggested
by Dickinson (1973):
PH ∆ (O3 )
∆T = · (6.1)
α (O3 )
where PH is the total gross heating rate, α is the relaxation rate (see
Eq. (4.89)) and ∆ (O3 )/(O3 ) is the relative ozone change. At 50 km, for
example, PH ≈ 10K/day and α ≈ 0.2/day, so that for an ozone reduction
of 20%, a temperature decrease of about 10◦ K is expected based on this
relationship.
Such a description can serve as a useful guide, but its applicability
is limited by several factors. Ozone plays some role in the emission of
infrared radiation as well as in absorbing ultraviolet radiation, so that
the relaxation rate in the above equation is also linked to ozone change.
Further, a change in temperature will modify all temperature-dependent

photochemical rates and hence will feed back to the ozone chemistry.
This renders the temperature-ozone relation into a tightly coupled and
non-linear system that can be better understood with the aid of some
simple chemical arguments. If a pure oxygen atmosphere is assumed
(see Chapter 5), the interaction between ozone density and temperature
dependent rates is expressed by:
∆(O3 ) 1 ∆(k2 /k3 )
= (6.2)
(O3 ) 2 (k2 /k3 )
Based upon the known temperature dependencies of reactions k2 and
k3 , we can write:
∆(O3 ) −1400
≈ ∆T (6.3)
(O3 ) T2
or
d ln(O3 )
≈ −1400 K (6.4)
dT−1
This expression suggests that in a region dominated by pure oxygen
chemistry, a temperature decrease of 10◦ K (imposed for example through
a change in circulation) would produce about a 20% increase in ozone.
In the real atmosphere, however, both ozone and temperature change
in concert, and there is hence a stabilizing feedback between the two
at upper stratospheric altitudes (Luther et al., 1977). As a chemical
perturbation occurs, the associated change in temperature assuming a
pure oxygen atmosphere and including chemical coupling changes to:
PH /α ∆ (O3 )
∆T ≈ PH
(6.5)
1 + α (1400/T ) (O3 )
2
Let us assume that a perturbation is applied which would produce a 10%

ozone decrease in the absence of coupling. As the ozone concentration
begins to decrease, the temperature decreases. This decreases the rates
of ozone-destroying chemical reactions that have a positive energy of
activation (such as the reaction between O + O3 cited above for pure Ox
chemistry). The net effect is an equilibrium ozone change that is roughly
half as large as that obtained in the uncoupled system. Similarly, the
temperature response is also halved.
These considerations lead to an anti-correlation of ozone and
temperature above 30 km, first noted by Barnett et al. (1975); later
studies include those of Krueger et al. (1980) and Finger et al. (1995).
Rood and Douglass (1985), Douglass et al. (1985), and Douglass and
Rood (1985) noted that there can be substantial phase differences
between ozone and temperature changes when dynamical processes
are very rapid, due to differences in the local response timescales for
temperature and ozone. However, Finger et al. (1996) showed good

agreement of observed and calculated anti-correlations between ozone
and temperature over much of the upper stratosphere in many cases.
Thus, observations and related theoretical modelling of the anticorre-
lation between temperature and ozone in the upper stratosphere provide
a graphic illustration of the role of temperature-dependent chemistry in
modulating ozone, as shown in Figure 6.2.
Eq. (6.5) was derived for a pure-oxygen atmosphere and is therefore
only an idealized case. A more realistic analysis includes the temperature
Figure 6.2. Observed variations in ozone and temperature based upon satellite
data from 30◦ N to 50◦ N (top) and from 30◦ N to 30◦ S (bottom), from Finger
et al. (1995).
effects of hydrogen, nitrogen, and chlorine chemistry. Haigh and Pyle

(1982) suggested the following more general expression:
d ln(O3 ) a + b f1 + c f2 + d f3
−1
= (6.6)
dT 2 + f1 + f2 + f3
where f1 , f2 , and f3 are the ratios of the rates of odd oxygen destruction
due to HOx , NOx , and ClOx , respectively, compared to the rate of
destruction by the pure oxygen reactions. These factors depend on the
chemical state of the atmosphere, and, in particular, on the amplitude of
the perturbation. The other constants depend only on chemical kinetics
and were given by Haigh and Pyle (1982) as a = 2810 K, b = 510 K,
c= 1200 K. The value of d is less than 275 K, showing that increasing
chlorine should reduce the stabilizing positive feedback between ozone
and temperature. Recent kinetic studies suggest that the ClOx and
NOx catalytic cycles are even less temperature sensitive than earlier
calculations, underscoring this aspect of the chemistry.
It should be noted that other chemical species besides odd oxygen
also depend upon temperature-dependent chemical reactions, further
complicating this coupled picture. For example, the formation of HCl
through the reaction of Cl + CH4 also displays moderately strong
temperature sensitivity, so that ozone’s full sensitivity to temperature is
linked not only to the temperature dependence of those reactions directly
involved in ozone destruction, but also those of others that are indirectly
linked through other chemicals. In general, these effects are relatively
small compared to the direct roles of the competing and temperature-
sensitive rate-limiting steps in the odd oxygen chemistry, so that useful
information on the competition between chemical mechanisms can be
obtained from analyses of ozone-temperature relations in the upper
stratosphere.
6.2.2 Particle Precipitation
Perhaps the most striking evidence for the prominent role of the
NOx cycle in influencing ozone was first provided by observations of the
changes in ozone following major charged particle precipitation events
(see Chapter 7). Protons, electrons, and alpha particles released from
the sun are channelled by our planet’s magnetic field lines and produce
ionization as they enter the Earth’s atmosphere. Ionization in turn
results in production of NOx (about 1.5 NO molecules are produced
per ionization) and HOx (about 2 HOx molecules are produced below
75 km per ionization). Solar disturbances can be associated with input
of large quantities of charged particles to the Earth’s atmosphere, which
result in the northern and southern lights (the aurorae) due to emissions
by excited atoms and molecules. They can also lead to relativistic
electron precipitation (REPs) and solar proton events (SPE, also referred
to as polar cap absorption or PCA events). While REPs are related
mainly to mesospheric ionization and can occur at latitudes as low as
about 50◦ , SPEs and PCA events are characterized by particles with
higher energies, some of which can penetrate to the stratosphere, but
are generally restricted to high latitudes near the geomagnetic poles.
The possibility that odd nitrogen would be produced by energetic
particle precipitation was first advanced by Dalgarno (1971), but the
role of odd nitrogen in ozone chemistry was not understood at that
time. In the early 1970s Crutzen et al. (1975) predicted that particle
events could be linked to transient ozone loss. A notable depletion of
upper stratospheric ozone was observed following the major solar proton
event of August 1972. Ozone dropped by as much as about 20% near
35 km at 75◦ N on the day after the primary event, and remained low
for weeks afterward (see e.g., Heath et al., 1977). A rapid drop in ozone
supports the short photochemical lifetime of ozone in this height range,
while the continued perturbation is consistent with the long lifetime of
the enhanced NOy in this region (order of months to a year). While
dramatic, such events are rare. Major solar proton events are relatively
infrequent, and typically occur only near the maximum of the 11-year
cycle of solar activity.
More recent work on later solar proton events has benefited from
concurrent observations, not only of ozone but also of NOx compounds
(e.g., Jackman et al., 2001), allowing a better link to be drawn between
ozone and the enhanced catalytic NOx concentrations. Figure 6.3
presents a striking illustration of the ability of current models to simulate
the general features of observed NOx and O3 changes during a solar
proton event observed in July 2000 (from Jackman et al., 2001). Note
that HOx also contributes to the observed decreases, especially in the
mesosphere. Further, it has been shown that the sensitivity of ozone
to increasing NOx depends upon stratospheric chlorine loading: for the
low total chlorine amounts of the early 1970s, upper stratospheric ozone
depletion was predicted and observed, but a SPE that occurred in the
1990s actually showed evidence for ozone increases in some regions, due
to the much higher chlorine loading present then (Jackman et al., 2000).
Thus solar proton events have provided a useful probe not only of the
NOx -catalyzed destruction of ozone, but also of its coupling to chlorine
over past decades of change in the ozone layer.
Production of mesospheric and thermospheric NOx by particle
precipitation does provide a possible source of stratospheric NOy , as
illustrated by the data in Figure 6.3. However, its magnitude is
Figure 6.3. Satellite observations from the HALOE instrument of ozone and
NOx changes following the major solar proton event of July 2000, compared to
model calculations. From Jackman et al. (2001).
critically dependent upon transport from the source region to lower

altitudes. Destruction of this potential pool of NOy is governed mainly
by the reaction of NO with atomic nitrogen (which in turn depends

on photochemical processes, particularly the photolysis of NO, see
Section 5.5). Mesospheric NOy , therefore, has a longer lifetime in winter
than in summer, and a particularly long lifetime in the dark polar
night region. Numerous studies have examined the possible downward
transport of this ionization-induced NOx to the stratosphere, and have
shown that there is some contribution but that it is largely confined to
high latitudes in winter (see e.g., Solomon et al., 1982; Brasseur et al.,
1983; Russell et al., 1984; Siskind et al., 2000; Jackman et al., 2001).
In contrast to NOx , the depletion of ozone due to HOx production
during particle precipitation events was first observed, and only later
understood in terms of theory. The first observations of mesospheric
ozone changes related to HOx production during solar proton events
were reported by Weeks et al. (1972). Swider and Keneshea (1973)
proposed that these unusual though limited rocket measurements
might be explained in terms of production of odd hydrogen due to
ion chemistry (see Chapter 7). Thomas (1983) presented extensive
satellite measurements showing major perturbations (in excess of 60%
depletion) to mesospheric ozone during solar proton events, and Solomon
et al. (1983) demonstrated that these agreed well with theory. Recent
studies of solar proton and relativistic electron precipitation (REP)
events include those of Jackman and McPeters (1987) and Jackman
(2001). While the impact of HOx production on mesospheric ozone can
be quite large and hence sometimes spectacular, it is very short-lived
since the lifetime of HOx is on the order of hours or less everywhere
below about 80 km. As might be expected based upon the vertical
profile of the HOx cycle’s role in ozone depletion, the impacts of
proton events through HOx chemistry are most pronounced above
about 50 km. Nevertheless, particle precipitation events and their
associated observable perturbations have provided key tests of scientific
understanding of some of the HOx and NOx chemistry controlling ozone
in the stratosphere and mesosphere.
6.2.3 The Effect of Changes in the Solar Irradiance
Changes in solar irradiance can modify the physical structure,

dynamics, and chemical composition of the middle atmosphere. The
question of effects of solar activity on the troposphere, particularly the
surface climate, has been a subject of scientific and public fascination
for many decades. However, it has been stressed that “despite a massive
literature on the subject, there is at present little or no evidence of
statistically significant or practically useful correlations between sunspot

cycles and the weather or climate” (Pittock, 1978). More recent
assessments show that many of the concerns raised by this quotation
still apply (see e.g., IPCC, 1994; 2001).
The goal of this section is not to address the Sun’s influence on
surface climate, but rather to consider its role in variability of ozone
in the mesosphere and stratosphere, again with a view towards this
perturbation as a test of some components of ozone photochemistry.
The following solar activity variations are known to influence the
radiant energy emitted by the Sun:
* Short-term variations (seconds to hours) related to solar flares and
disruptive prominences.
* Intermediate-term variations (days to weeks) due to active regions
and the solar rotation period (27 days).
* Long-term variations, particularly those over the 11-year cycle of
sunspot activity and perhaps the 22-year magnetic field cycle.
Much of the Sun’s energy is emitted in the visible, a wavelength range
where high-quality observations over the 27-day and 11-year periodicities
show very limited variability (less than a percent near 300 nm, see
IPCC, 2001). However, at shorter wavelengths in the ultraviolet (see
Figure 6.4), the variability is much larger (order of 15-20% over the
11-year solar cycle near 190-200 nm), so that solar activity variations
can be expected to alter some of the wavelengths of light that play
key roles in stratospheric and mesospheric photochemistry. Changes in
the heating rates due to absorption of the same radiation by molecular
oxygen and ozone should also be considered. While the major heating is
provided by ozone absorption at Hartley band wavelengths (250-300 nm)
where the solar variability is small, any change in ozone itself (induced,
for example, by solar-driven variability at other wavelengths) will feed
back to the radiative budget and temperatures through the mechanisms
delineated in Section 6.2.1 above. Finally, we note an indirect effect
that can be significant: changes in photolysis rates can modify the ozone
column at a given altitude, thereby affecting the penetration of radiation
to lower heights and producing a non-local perturbation.
The largest variability in solar output is generally observed at the
shortest wavelengths. Extreme ultraviolet radiation varies by a factor
of two or more over an 11-year solar cycle, and produces nitric oxide
in the thermosphere. As already mentioned, NO can be transported
to the mesosphere and, to some extent, in winter even to the upper
stratosphere where it can perturb upper stratospheric ozone. This is an
indirect mechanism linking solar activity at short wavelengths (which
do not penetrate below the lower thermosphere) to chemistry at lower
altitudes.
Figure 6.4. Spectral distribution of the variability in the solar flux associated
with the 11-year solar cycle. Based on the observations by UARS/SOLSTICE
/
(see also Figure 4.10). Courtesy of G. Rottman, Univ. of Colorado.
The singular role played by the solar Lyman-alpha line at 121.6 nm

was discussed in Chapters 4 and 5. This solar emission feature varies by
about 30% over the 27-day rotation and by about a factor of 2 over the
11-year solar cycle, and it has a particularly important role in affecting
both water vapor and oxygen photolysis in the mesosphere. Thus, solar
activity over a solar rotation or an 11-year cycle affects both the loss and
production rates of odd oxygen, and the net effect on ozone depends on
the balance between competing effects. Furthermore, fast photolysis
of water vapor at Lyman-alpha can influence its vertical profile, so
that transport cannot be ignored when considering mesospheric solar
impacts. However, Lyman-alpha penetrates only to about 65 km.
Longer wavelengths are likely to control the solar interactions at lower
altitudes. For the lower mesosphere and upper stratosphere, changes in
the wavelength range from about 170 to 200 nm can affect photolysis in
the Schumann-Runge bands that control production of ozone through
the dominant oxygen photolysis process at these altitudes. This acts to
increase ozone during periods of high solar activity, although competing
effects such as increased photolysis of chlorofluorocarbons in the same
wavelength range can provide some compensating effects.
Changes in ozone through any of the above mechanisms can influence

temperatures as already noted, and changes in temperature gradients
can in turn affect zonal winds, planetary wave propagation, and perhaps
even the mean meridional circulation of the stratosphere (see e.g., Geller
and Alpert, 1980; Haigh, 1996; Shindell et al., 1999). This possible
feedback provides an important mechanism whereby the relatively large
changes in solar output at short wavelengths can affect dynamical
processes, at least at stratospheric altitudes. Their potential for
propagation downward to lower heights (through, for example, reflection
of upward-propagating planetary waves) has been a focus of considerable
scientific interest (e.g., Haigh, 1996) and provides a possible but as yet
unproven mechanism for solar coupling of surface climate through the
stratosphere. Further, any change in the mean circulation can affect
transport of ozone and related species such as N2 O, CH4 , H2 O, CFCs,
etc. Recent studies have focussed on understanding the vertical profile
of solar-related coupled changes in both ozone and temperature, and
probing the time lags observed between the two (e.g., Chen et al., 1997).
Figure 6.5 shows observed and modelled responses of ozone based upon
observations of solar variability on the 27-day solar rotation timescale.
The figure illustrates good agreement in the upper stratosphere, where
local chemistry and chemistry/temperature coupling are both important
Figure 6.5. Comparison of the observed and calculated ozone sensitivities over
27-day solar rotation timescales, from Chen et al. (1997). The dashed line
shows a one-dimensional model calculation, while the solid and dot-dashed lines
display two-dimensional model estimates using two different prescriptions for
the absolute irradiance changes. Observations from Nimbus-7 SBUV are also
presented for comparison.
(Chen et al., 1997). Perhaps most important is the ability of models

to simulate the overall shapes of the vertical profiles of the responses
obtained in both ozone and temperature, since these in turn support
the calculation of the penetration of radiation into the atmosphere and
hence provide a key test of theory.
6.2.4 Long-term Trends and Halogen Chemistry in the

Upper Stratosphere
Section 5.6.3 discussed the chemistries of the halogens chlorine and

bromine, and outlined their interactions with one another and with
ozone. More than two decades after the pioneering prediction of possible
ozone destruction through humankind’s use of halogenated chemicals,
upper stratospheric ozone observations began to reveal systematic
depletion indicative of a changing chemical state (see SPARC, 1998,
and references therein).
Chapter 5 described the important role of chemical partitioning
processes in modulating ozone destruction by halogen compounds.
Observations of HCl, ClO, ClONO2 , and other chlorine-bearing gases by
infrared spectroscopy onboard the space shuttle (Michelsen et al., 1996;
Zander et al., 1996) or from satellites (Dessler et al., 1995; 1998) allow
study of how chlorine is chemically partitioned in the middle and upper
stratosphere in some detail, as shown in the left-hand side of Figure 6.6.
In the uppermost stratosphere above about 45 km, nearly all of the
chlorine released from source molecules such as CFCs is sequestered in
the HCl reservoir, due largely to the efficacy of the reaction of Cl + CH4
at warm temperatures and high Cl/ClO ratios there. Models (Michelsen
et al., 1996), stratospheric observations (Stachnik et al., 1992; Chandra
et al., 1993) and laboratory measurements (Lipson et al., 1997; see JPL,
1997; 2000) suggest that a small yield of HCl in the reaction of ClO with
OH also affects the HCl/ClO partitioning in this region.
Because nearly all of the chlorine and fluorine released from
chlorofluorocarbons resides as HCl and HF in the stratosphere near
50 km, observations of those two gases in this region provide an
important verification of their attribution to CFC sources. Global data
by Russell et al. (1996) display abundances and trends in both HCl
and HF near 50 km that are quantitatively consistent with observations
of the chlorofluorocarbons at ground level; these observations therefore
confirm that CFCs are the key sources for stratospheric chlorine
and fluorine. Such observations also suggest that the abundance of
reactive chlorine reached its maximum in the late 1990s, while fluorine
Figure 6.6. Observations of chlorine partitioning as a function of altitude from

an instrument on board the space shuttle (Zander et al., 1996) are shown at
left. The observed vertical profile of the ozone trend at northern mid-latitudes
is shown at the right (SPARC, 1998) together with a model estimate (from
Solomon et al., 1997).
continued to increase (Anderson et al., 2000). These observations

are broadly consistent with expectations based on global decreases in
chlorofluorocarbon emissions, and corresponding increases in substitute
compounds such as hydrofluorocarbons that bear fluorine but no
chlorine. Taken together, the global measurements of HCl, HF, other
chlorine compounds, and the CFC source gases both at the surface and
in the tropopause region (e.g., Russell et al., 1993; Zander et al., 1996;
Montzka et al., 1996; Schauffler et al., 1993) provide direct evidence that
the chlorine content of the stratosphere is greatly perturbed, with about
85% of the 1992 stratospheric chlorine burden attributable to human
activities.
Crutzen (1974) and Crutzen et al. (1978) carried out some of the first
detailed chemical models of chlorine-catalyzed ozone depletion, building
upon the studies of Stolarski and Cicerone (1974), Molina and Rowland
(1974) and Rowland and Molina (1975) and including the general
chemical understanding outlined above. Crutzen (1974) predicted a
relative maximum in ClO near 40 km, which was broadly confirmed a
few years later by observations of ClO by Anderson et al. (1977), Parrish

et al. (1981), and Waters et al. (1981). The left hand side of Figure 6.6
shows that direct observations of the ClO/Cly profile in the middle and
upper stratosphere agree with those early predictions and observations.
Largely because of this relative maximum in ClO, Crutzen (1974)
predicted a maximum in ozone depletion in the same region (although
other factors such as the availability of atomic oxygen also make some
contribution to the vertical profile of ozone depletion). While there were
indications of ozone depletion in the upper stratosphere beginning in the
1980s (e.g., Ozone Trends Panel, 1988), it is only within the 1990s that
a clear trend could be identified, distinct from natural fluctuations (e.g.,
Miller et al., 1995; SPARC, 1998). Sample observations of the northern
mid-latitude ozone profile trends deduced from satellite and in-situ data
as shown in the right-hand side of Figure 6.6 display a maximum near
40 km. The close agreement between the vertical shapes of the observed
and predicted ozone changes in the upper stratosphere provides strong
evidence for gas-phase chlorine-catalyzed ozone depletion chemistry.
Further, the correspondence between the shapes of the vertical profiles
of the observed ozone depletion and of the ClO/Cly ratio attests to the
role of partitioning processes in modulating this chemistry. Indeed, this
is the first of several ‘fingerprints’ that support the role of chlorine in
ozone depletion.
Figure 6.7 presents another fingerprint illustrating the role of gas-
phase chemistry and chlorine partitioning in ozone depletion, namely
the latitudinal gradients observed in upper stratospheric ClO and ozone
depletion. In the early 1980s, global measurements of methane by
satellite illustrated that the strong upwelling of the Brewer-Dobson
circulation in the tropical upper stratosphere (see Chapter 3) gives
rise to a maximum in methane there (Jones and Pyle, 1984). The
enhanced methane in turn was predicted to lead to a tropical minimum
in ClO through its dominant role in ClO/HCl partitioning and hence
in ozone depletion (Solomon and Garcia, 1984). Later global satellite
observations of ClO (Waters et al., 1993; Waters et al., 1999), as shown in
the lower panel of Figure 6.7, indeed display such a latitudinal gradient,
with a pronounced minimum in the tropics. Satellite observations of
the latitudinal variation of the ozone trends over the period from the
early 1980s to the mid-1990s (top panel of Figure 6.7) reveal a similar
spatial pattern in the upper stratosphere — again as predicted. Thus,
not only the vertical profile but also the latitudinal structure of the
ozone depletion in the upper stratosphere parallel the patterns observed
in ClO. These spatial variations in ozone depletion point towards gas-
phase chlorine chemistry and highlight the role of chemical partitioning
in modulating ozone depletion.
Figure 6.7. Observations of the global distribution of zonally averaged daytime

ClO abundance from satellite-based microwave remote sensing (Waters
( et al.,
1999)) together with the zonally averaged ozone trend deduced from satellite-
based visible spectroscopy (SPARC, 1998; see McCormick et al., 1992). Trends
are statistically significant outside the shaded region.
Other processes that could contribute to upper stratospheric ozone

changes include trends in methane, nitrous oxide, and water vapor.
These source gases can, for example, lead to changes in HOx and
NOx , which can in turn affect ozone loss rates and the competition
between different catalytic cycles. However, the effect of these changes
is considerably smaller than the dramatic impact of the five-fold
enhancement in chlorine caused by human activities. By the turn of the
21st century, observations and modelling studies showed that chlorine
chemistry dominated the trends found in upper stratospheric ozone (see
Muller
¨ et al., in WMO/UNEP, 1999).
The upper stratosphere was not only depleted in ozone, but was also
colder by the 1990s than it had been 20 years earlier (Ramaswamy et al.,
1996). The cooling of the upper stratosphere is due in part to the
chlorine-driven ozone changes and is hence subject to ozone/temperature
feedback as discussed in Section 6.2.1. But much of the upper
stratospheric cooling is driven not by ozone itself but by increases in
greenhouse gases (mainly carbon dioxide and water vapor). While
greenhouse gases warm the planet surface, they cool the stratosphere.
Thus, changes in the upper stratospheric thermal structure driven by
trends in these gases were believed to be responsible for an average
cooling of about 1.5◦ K near 40 km in 1997 as compared to 1979, and
an associated damping of the ozone trend that would otherwise have
occurred (see e.g., Forster and Shine, 1999).
6.3 Lower Stratospheric Ozone Depletion:

Observations and Explanations
The chemical lifetime of ozone in the sunlit atmosphere increases from

the order of days near 35 km to about a year near 20 km. This dramatic
change largely reflects slower ozone photolysis rates and increasing air
densities that shift the balance between O and O3 strongly in favor of
ozone, thereby reducing the effectiveness of many of the odd oxygen
catalytic destruction cycles whose rate-limiting steps involve reactions
with O (e.g., the reactions of O with NO2 , ClO, and HO2 ). While some
chemical cycles involve only ozone, these tend to be slower than those
linked to the highly reactive oxygen atom. As chemical processes slow
down, dynamical mechanisms play larger relative roles in the odd oxygen
budget. Particularly in winter when the mean meridional circulation is
relatively fast, chemistry and dynamics both influence ozone abundances
(see e.g., Perliski et al., 1989). In summer, reduced wave activity
results in a slower circulation, and a detailed analysis of chemical
rates is required to determine the balance of competing processes.
The hallmark of studies of lower stratospheric ozone changes has been
ongoing examination of chemical processes occurring on relatively slow
time scales, and the need to understand changes in dynamical processes
— both those driven by the ozone changes themselves (e.g., feedbacks
from ozone depletion to dynamics) and those perhaps driven by other
mechanisms (such as changing greenhouse gas concentrations that can
change dynamics or changes in tropospheric meteorology).
Throughout the 1970s and until about the mid-1980s, modelling
studies predicted that modest changes in column ozone might someday
occur due to human activities (see Figure 6.8). By the early

1980s, gas-phase photochemical theories predicted that human uses
of chlorofluorocarbons would likely produce less than a 5% integrated
change in column ozone at steady state as shown in the figure. That
steady state would not be achieved until about the middle of the 21st
century. The bulk of those changes were predicted to occur in the upper
stratosphere, following the arguments regarding chemical partitioning
discussed above.
Prior to the discovery of heterogeneous chemistry, models suggested a
strong anti-correlation between estimates of the calculated future mid-
latitude ozone changes to be expected at steady state due to CFC
emissions and that predicted for NOx emissions from supersonic aircraft,
as illustrated in Figure 6.8. For a period of more than a decade (roughly
1974 to the mid-1980s), when the estimated future NOx impact on ozone
increased, the calculated CFC-related ozone depletion decreased. Of
particular importance is the reaction:
HO2 + NO → NO2 + OH
which has the effect of increasing the net production of odd oxygen
Figure 6.8. Calculated one-dimensional model ozone column change resulting

from specific fixed perturbations (increased NOx emissions at 20 km, simulating
possible high altitude aircraft releases, and the steady-state response to
1974 emissions of chlorofluorocarbons), as function of the year in which
the calculation was performed. Changes over time reflect changes in the
understanding of stratospheric chemistry, not real atmospheric changes. Figure
courtesy of D.J. Wuebbles (personal communication).
through methane oxidation chemistry in the lowermost stratosphere.

Changes in the early estimates of the rate of this single reaction
from laboratory studies (see e.g., Crutzen and Howard, 1978) had
large impacts on the estimation of ozone depletion due to emisions
of NOx from supersonic aircraft — note in particular the transition
from predictions of ozone depletions due to NOx input at 20 km to
projected increases which occurred around 1979. Other improvements
in laboratory kinetics relating to HOx chemistry (e.g., the rate of OH
+ NO2 → HNO3 ) also had significant effects on calculations of the
estimated effect of NOx on stratospheric ozone, for similar reasons.
Further, the coupling of the ClOx and NOx cycles through, for example,
the reactions ClO + NO2 → ClONO2 and ClO + NO → Cl + NO2
are key elements in the competition between NOx and ClOx . This
balance between chlorine and nitrogen chemistry changed dramatically,
however, with the discovery of surface chemistry’s large impact on
chlorine, nitrogen, and hydrogen radicals in the lower stratosphere (see
e.g., Rodriguez et al., 1991). Indeed, Figure 6.8 ends with the last
systematic calculation of mid-latitude ozone trends employing gas phase
chemistry alone, in 1988. When the first statistically significant ozone
changes were observed, they were far larger than model estimates, were
found in the lower rather than upper stratosphere, and were located in
a region where no one had predicted a special sensitivity: Antarctica.
6.3.1 Antarctic Ozone
Measurable total column ozone depletion was first documented in

the Antarctic spring at the British Antarctic Survey station at Halley
(Farman et al., 1985). Figure 6.9 presents an updated set of these
pioneering measurements, showing a remarkable decline in October
average ozone that began near the late 1970s. These findings were
quickly confirmed by space-based measurements (Stolarski et al., 1986)
and by observations at other Antarctic sites (e.g., Komhyr et al.,
1986). Observations of total column ozone using infrared and visible
spectroscopy (Farmer et al., 1987; Mount et al., 1987) provided further
support for the seasonal depletion of springtime ozone using independent
methods. As the satellite measurements confirmed that the depletion
extended over roughly the entire continent, the phenomenon became
known as the Antarctic ozone ‘hole’. Farman et al. showed that the
ozone hole is confined to particular seasons (i.e., spring) and to south
polar latitudes.
6.3.1.1 What is the Ozone Hole?

While not a true ‘hole’ in the sense that some column ozone remains
even in the most extreme depletions observed in the mid 1990s (when
October ozone minima were near 100 Dobson Units over the South Pole,
or depletion of about two-thirds of the historical levels, see Hofmann
et al., 1997), the descriptor captures the fact that the peak depletion
is sharply limited to Antarctic latitudes. Dobson (1968 and references
therein) noted that there is less ozone naturally present over Antarctica
than over the Arctic in winter and much of the spring, but this
climatological difference between the natural ozone levels over the poles
of the two hemispheres should not be confused with the abrupt decline
that began near the mid-1970s as depicted in Figure 6.9. Newman (1994)
discusses these and other historical measurements of total ozone and
shows that the Antarctic ozone hole began only in the last few decades.
Measurements of the vertical profile of the depletion within the ozone
hole were first presented by Chubachi (1984), and were rapidly followed
by other data such as that of Hofmann et al. (1987), Gardiner (1988),
Iwasaka and Kondoh (1987), and satellite studies such as Bevilacqua
et al. (1997). Hofmann et al. (1997) presented a detailed analysis of
many years of ozonesonde measurements at the South Pole. Figure 6.10
summarizes a key result of that study, showing that the depletion of
the Antarctic ozone column is largely confined to altitudes from about
12-25 km, far below the altitudes where gas-phase chlorine chemistry
would predict significant changes. Figure 6.10 also illustrates the shape
of the unperturbed ozone ‘layer’ observed in historical ozone data, and
the near-total removal of ozone in the heart of the layer in a typical
contemporary sonding for early October. Finally, the figure shows the
close correspondence between the region where most of the ozone is
depleted and a vertical profile of a typical polar stratospheric cloud
(PSC) observed at the South Pole (Collins et al., 1993).
The latitudinal gradients in Antarctic ozone depletion are related
to the dynamical structure of the polar winter stratosphere, whose
circulation can be viewed as a vortex (see e.g., Schoeberl et al., 1992;
Holton et al., 1995). Briefly, the absence of solar illumination in
high latitude winter leads to cooling over the poles and hence a large
temperature gradient near the polar terminator. This thermal gradient
implies rapid zonal (west-east) flow characterizing the ‘jet’ at the edge
of the vortex, while the air within the vortex is relatively isolated
compared to surrounding mid-latitude regions, allowing deep depletion
to develop. Differences in the pre-1970s ozone abundances in the two
polar vortices first noted by Dobson (1968) are related to differences in
Figure 6.9. Observations of total ozone at various locations. The Antarctic

data are from Halley (Farman et al., 1985; Jones and Shanklin, 1995) and
updated courtesy of J. Shanklin. The Arctic data are from satellite observations
described in Newman et al. (1997), updated courtesy of P. Newman. The
Arosa, Switzerland, dataset is the longest running in the world (Staehelin
et al., 1998a,b). Satellite observations from a slightly higher mid-latitude region
are shown for comparison (Hollandsworth et al., 1995), updated courtesy of
R. Nagatani. The satellite data are zonally and monthly averaged, while the
ground-based data at each site have also been averaged over time as indicated
in each case. Adapted from Solomon (1999).
Figure 6.10. Observations of the vertical profile of ozone observed at the South
Pole in the Octobers of the late 1960s and early 1970s, contrasted with those of
1986 and 1997. Total ozone ((DU)) is indicated for each profile, from Hofmann
et al. (1997). The right hand panel shows a typical polar stratospheric cloud
observed at the South Pole from the observations of Collins et al. (1993). From
Solomon (1999).
atmospheric waves and circulation patterns, which are in turn driven by

factors relating to surface topography (e.g., distribution of mountains,
oceans, and continents). The north polar vortex is generally more
disturbed by atmospheric waves forced from beneath by flow over a
more variable surface topography. These lead to greater mixing and
faster downward motion which both increases the natural wintertime
Arctic ozone abundances (by bringing down ozone-rich air from above)
and warms the lower stratosphere (through adiabatic compression).
Temperatures in the Antarctic vortex are both colder and less variable
than those of the Arctic, which strongly influences the polar ozone
depletion in the two hemispheres (see Figure 6.17 below).
Figure 6.11 shows measurements of the seasonal cycle of ozone at
Halley in historical and recent data, which show that the depletion
occurs only over a limited portion of the year. These observations
demonstrate that contemporary observations of ozone at Halley in late
August (end of austral winter) are near historical levels, while the bulk
of the ozone loss there occurs rapidly during the month of September.
In recent years, measurable ozone depletion is also observed in Antarctic
summer — in part the result of dilution of the extreme losses in spring.
Figure 6.11. Observations of the full seasonal cycle of daily ozone at Halley,
Antarctica, in the years before the ozone hole (1976/7
( / and 1977/8),
/ ) and in
1987/88
/ and 1996/7 / ((courtesy of J. Shanklin).
) Note the rapid September drop
in total ozone in the years displaying the ozone hole. From Solomon (1999).
6.3.1.2 Early Explanations

By the late 1980s, the chemical foundation of the ozone hole in
heterogeneous chlorine chemistry was firmly established (WMO, 1989,
and references therein). However, the suite of proposals initially
advanced to explain its origins prior to that time is of interest
both historically and as an illustration of how differing proposals
constructively add to and refine scientific debate. Here we briefly
summarize the primary explanations advanced to explain the ozone hole
following its surprising discovery.
Dynamics. Tung et al. (1986) pointed out that the extreme isolation of
the Antarctic winter circumpolar vortex leads to very low temperatures
approaching radiative equilibrium, in contrast to the Arctic winter
vortex, which is much warmer. When radiative equilibrium is achieved,
the rate of radiative cooling becomes zero, so the net heating equals
the radiative heating alone. Tung et al. suggested that the return of
sunlight to the Antarctic spring (and the associated radiative heating)
could therefore lead to net upward motion, assuming

Q=H−C
and
dT
Q= + Γ w∗
dt
where Q is the net radiative heating, H is the radiative heating rate,
Γ is the static stability and w∗ is the mean vertical velocity. C is the
radiative cooling rate, which becomes small as the stratosphere cools in
winter. When sunlight returns in spring, C is small, but H may be large,
implying that Q could be positive. If the net radiative heating is positive,
net upward motion may occur in the mean meridional circulation system.
Since the ozone mixing ratio decreases with decreasing altitude below
the ozone maximum, upward motion would transport air which is
relatively ozone-poor into the lower stratosphere, thereby depleting the
ozone column. This type of transport can explain qualitatively the
observed decline in spring total ozone, but has a number of quantitative
shortcomings. The decadal trend in total ozone can be explained in the
context of this theory only if a long-term change in the stratospheric
circulation system has occurred. Tung et al. (1986) suggested that a
long-term trend in Antarctic spring total ozone might be due to changes
in the strength of this upward mean circulation over time, possibly
because of increased aerosol heating associated with the volcanic aerosol
emitted by El Chichon;´ see also an elegant later paper by Tung (1986).
However, observations of conservative tracers that illustrate the direction
of dynamical flow, such as aerosols (Hofmann et al., 1987), nitrous oxide
(Parrish et al., 1988; Loewenstein et al., 1989) and other long-lived gases
(Jaramillo et al., 1989; Toon et al., 1989) quickly demonstrated that
the ozone hole could not be caused by such upward motion. Indeed,
much as Brewer (1949) deduced the nature of global transport from
observations of water vapor as a tracer of atmospheric dynamics, so have
observations of a wide variety of chemical tracers shown that transport
is directed downwards within the Antarctic stratosphere in spring rather
than upwards, although there is still some debate about the strength of
this fluid flow and the degree of exchange of air between lower latitudes
and polar regions (e.g., Hartmann et al., 1989; Tuck, 1989; Tuck et al.,
1997; Schoeberl et al., 1990; 1992; 1995; Manney et al., 1995a,b).
Mahlman and Fels (1986) also considered changes in atmospheric
dynamics that might explain the ozone trend through dynamical
phenomena. They suggested a reduction in wintertime planetary-scale
wave activity in the southern hemisphere troposphere, which would lead
to reduced downward transport of ozone into the Antarctic during the
winter and spring. However, an explanation relying upon an extended
winter season reduction in ozone transport would imply reduced ozone

not only in October, but also earlier in the southern winter (e.g., in
August and September). Observations (i.e., earlier versions of the type
of data shown in Figure 6.11) demonstrated that the ozone depletion is
not a slow process occurring all winter long, but rather a rapid removal
that begins around late August as shown above.
Other authors noted that Antarctic stratospheric temperatures
differed by 18 K in October at about the 30-mbar level during the years
1979 versus 1985, and pointed out that the colder temperatures appeared
well correlated with the total ozone changes observed by TOMS —
perhaps indicating a dynamical cause (Newman and Schoeberl, 1986).
However, it is important to note that ozone provides the primary source
of heat to the stratosphere through its absorption of UV radiation. Shine
(1986) suggested that the ozone hole should be expected to lead to a
stratospheric cooling (implying that a cooling trend in October could be
an effect of the ozone depletion rather than its cause). Observations have
since demonstrated that the Antarctic temperature trends indeed occur
later in the spring season than the onset of ozone depletion, and thus
cannot be the fundamental cause of the ozone hole but rather illustrate
one of its effects (e.g., Newman and Randel, 1988). Indeed, the removal
of this source of heat has become large enough to influence the warming
of the Antarctic stratosphere at the end of the winter-spring season.
Jones and Shanklin (1995) demonstrated progressively later warmings
over Halley and the ozone observations in Figure 6.11 also illustrate this
effect through the delayed austral spring ozone increases (now generally
seen in December rather than November).
Solar Activity. Enhanced nitrogen oxides from high solar activity that
occurred in the early 1980s were also proposed as a possible cause of the
ozone hole (Callis and Natarajan, 1986), drawing upon the well-known
catalytic chemistry of NOx and its enhancement by processes such as
solar proton events as discussed in Sections 5.5.2 and 6.2.2. However,
the solar theory proved to be in conflict with observations of nitrogen
compounds. Measurements of the nitrogen dioxide column by Noxon
(1978) and McKenzie and Johnston (1984) displayed reduced rather
than enhanced nitrogen oxides over the south polar regions. After the
discovery of the ozone hole, similar measurements confirmed those early
data using both infrared and visible spectroscopy methods (Coffey et al.,
1989; Farmer et al., 1987; Mount et al., 1987). Airborne measurements
of the latitudinal gradient of nitric oxide at 20 km (in the heart of the
south polar ozone destruction region by Fahey et al., 1989a) using a
chemiluminescence method are compared to the data of Noxon (1978) in
Figure 6.12. The two datasets both show that the southern high latitude
winter/spring stratosphere contains a minimum in nitrogen oxides rather
Figure 6.12. Observations of the ‘cliff’ in NO2 reported by Noxon (1978).

The solid and dashed lines represent Northern Hemisphere measurements while
the filled circles and crosses show Southern Hemisphere evening and morning
twilight data, respectively. The NO measurements of Fahey et al. (1989a) are
shown for comparison. The two molecules interchange rapidly with one another
in the sunlit atmosphere and hence provide a measure of NOx . Both datasets
show very low NOx in the high latitude stratosphere. From Solomon (1999).
than a maximum as would be required for the solar theory of Antarctic

ozone depletion to operate in the key region displaying the maximum
ozone loss. The differences in the location of the steep gradient between
the two datasets likely reflect differences in season and local motion of
the polar vortex. The observed shape of the profile of ozone depletion
is also in conflict with the solar theory, since downward transport of
ozone-depleting nitrogen radicals from above would yield greater ozone
losses at higher altitudes. This mechanism is therefore inconsistent with
the observed ozone removal in a confined range of altitude from about
12-24 km as shown in Figure 6.10.
It is interesting to note that the observations of Noxon (1978; 1979)

revealed strong evidence for a ‘cliff’ in NO2 in polar regions several
years before the Antarctic ozone hole was discovered. The chemistry
of this anomaly was not understood and it was called out as one of
the key challenges to scientific understanding of stratospheric chemistry
of the time (see e.g., WMO, 1985). We now know that the chemistry
that produced Noxon’s ‘cliff’ is indeed tied to that of the ozone hole;
the relationship between NO2 and ClO will be discussed further below.
Arguably, the ozone hole might have been predicted before it was
observed had the Noxon ‘cliff’ been better understood in the early
1980s. For example, a model calculation with an imposed deficit in
polar nitrogen based on the data would have displayed enhanced polar
ozone depletion.
Halogen Chemistry on PSCs. The cold temperatures that occur in
polar winter can lead to formation of clouds within the stratosphere, and
there are visual sightings of such Arctic clouds dating back hundreds
of years. In the unpopulated Antarctic, the earliest explorers noted
unusually colorful high clouds in winter. The term polar stratospheric
clouds (or PSCs) was coined by McCormick et al. (1982), who first
presented satellite observations of high-altitude clouds in the Antarctic
and Arctic stratospheres, but the clouds were considered little more than
a scientific curiosity until the ozone hole was discovered.
Solomon et al. (1986) suggested that HCl and ClONO2 might react
on the surfaces of polar stratospheric clouds, perturbing gas-phase
chlorine partitioning in a manner that could greatly accelerate ozone
loss in the Antarctic lower stratosphere (HCl + ClONO2 → HNO3 +
Cl2 ). The Cl2 formed would photolyze quickly in sunlit air and rapidly
form ClO. Solomon et al. stressed that this and related heterogeneous
reactions would suppress the concentration of NO2 by forming HNO3 ,
so that the ClO thereby released could not readily reform the ClONO2
reservoir. They also showed that such changes in NOx partitioning had
the potential to better explain the observed gradients in HNO3 and NO2
in polar regions — i.e., providing insight into the then-eight-year-old
mystery of the Noxon cliff. Thus it was recognized that rapid ozone loss
via chlorine chemistry would require (i) the heterogeneous ‘activation’
of chlorine from both the HCl and ClONO2 reservoirs and (ii) the
suppression of NO2 — an essential element in keeping the chlorine active.
The production of Cl2 in this process implies that sunlight would be
required to release Cl, so that the ozone depletion would occur when
air was not only cold but also sunlit (i.e., largely in spring as observed,
rather than in winter when the polar cap is continuously dark or in
summer when it is warm). Some ozone loss can take place even in polar
winter, however, due to atmospheric waves that move polar air out to
the sunlit atmosphere for brief periods (Tuck, 1989; Sanders et al., 1993;
Roscoe et al., 1997).
Observations of PSCs, low NO2 amounts in polar regions (Figure
6.12), enhanced polar HNO3 (Murcray et al., 1975; Williams et al., 1982)
and the vertical profile of the ozone depletion based upon the Japanese
measurements (Chubachi, 1984) were cited in support of heterogeneous
chemistry as the primary process initiating Antarctic ozone depletion.
Such a mechanism would be most effective in the Antarctic due to
colder temperatures and greater PSC frequencies there than in the
corresponding seasons in the Arctic (McCormick et al., 1982), a point
discussed further below.
As in the discussion of gas-phase chemistry, a complete understanding
of ozone depletion requires consideration not only of how much ClO is
present (i.e., ClO/Cly ), but also of the catalytic cycles in which ClO
may engage. Solomon et al. (1986) emphasized the catalytic ozone
destruction initiated by the reaction between HO2 and ClO. However,
this process cannot destroy enough ozone early enough in the spring
season to be consistent with the seasonality of the ozone loss process as
shown above in Figure 6.11.
Molina and Molina (1987) showed that very rapid ozone depletion
can occur through a previously-unrecognized catalytic cycle involving
formation and photolysis of a ClO dimer, Cl2 O2 . Following a period
of some uncertainty regarding the kinetics and photochemistry of the
dimer, laboratory studies confirmed its importance (e.g., Sander et al.,
1989; see Rodriguez et al., 1990 for model calculations and JPL, 2000,
for a summary of laboratory data). This cycle is now well-recognized
as the primary catalytic process responsible for about 75% of the ozone
removal in the ozone hole. Note that the reaction of ClO with itself
and a third body to form Cl2 O2 is the rate limiting step in this critical
catalytic cycle, implying that measurements of ClO provide a direct
measure of the rate of this odd-oxygen loss process.
McElroy et al. (1986) and Tung et al. (1986) emphasized the role of
bromine chemistry in ozone hole formation (in particular, its coupling to
chlorine through the reaction between ClO and BrO); this cycle is now
known to contribute about 20% to the annual formation of the Antarctic
ozone hole (e.g., Anderson et al., 1989). Both McElroy et al. (1986)
and Tung et al. (1986) also emphasized the need for reduced NO2 in
order for ClO to remain active (noting the links to the Noxon ‘cliff’),
and McElroy et al. (1986) also emphasized the Japanese ozonesonde
observations, particularly the observation of ozone loss at low altitudes,
where bromine can be very effective for ozone destruction.
6.3.1.3 Observations of Chemical Species Involved in Antarctic

Ozone Destruction
Foremost among the data that established the cause of the ozone hole
are observations of active chlorine species, particularly the ClO that
determines ozone loss through Cl2 O2 formation. De Zafra et al. (1987;
1989) presented ground-based microwave emission measurements at
McMurdo Station, Antarctica, showing evidence for greatly enhanced
ClO in the lower stratosphere. Near 20 km, the observations
suggested mixing ratios of about 1 ppbv in September, about 100 times
greater than the 10 pptv predicted by gas-phase photochemical theory.
Anderson et al. (1989) presented in-situ airborne measurements of ClO
using a resonance fluorescence method. The latitudinal coverage of the
airborne data taken flying south from Chile near 20 km showed a very
steep gradient in ClO as the airplane crossed into cold regions within
the Antarctic vortex, increasing to about 1 ppbv in mid-September as
shown in Figure 6.13. Hence, within a few years after the discovery of the
Figure 6.13. Observations of the latitude gradients in ClO, NOy , and H2 O

on a flight of the ER-2 aircraft in September 1987, showing evidence for
extremely high ClO in the Antarctic together with substantial denitrification
and dehydration (removal of NOy and H2 O) associated with PSCs. From
Solomon (1999).
ozone hole, two independent methods confirmed remarkably elevated

ClO abundances in the ozone hole region, which are only possible if
chlorine is released from both of the reservoir gases, HCl and ClONO2 .
Figure 6.14 shows both ground-based and airborne measurements of
ClO in September over Antarctica from 1987, and compares them to gas-
phase and heterogeneous photochemical theory. More recently, satellite
observations as depicted in Plate 6 have allowed study of the full global
distribution of ClO (Waters et al., 1993; 1999) using microwave emission
methods which further tie the ozone destruction region with the spatial
distribution of enhancements in ClO (Manney et al., 1995b; MacKenzie
et al., 1996). The data show that the region of reduced ozone extends
over an area larger than the continent beneath, and illustrates the close
spatial correspondence between the regions of depleted ozone and those
of enhanced ClO (as first emphasized in Anderson et al., 1989). Plate 6
provides a dramatic illustration of the spatial structure of enhanced ClO
and depleted ozone over Antarctica.
Observation of chlorine dioxide (OClO) via visible spectroscopy is
another independent method of probing the chlorine chemistry that also
Figure 6.14. Observations of the vertical profiles of ClO in the Antarctic

stratosphere in September 1987 from both ground-based microwave remote
sensing (de Zafra et al., 1989) and aircraft resonance fluorescence techniques
(Anderson et al., 1989). These data are compared to a gas-phase
photochemistry model and to the heterogeneous chemistry model of Jones
et al. (1989), which accounts for air parcel trajectories. From Solomon (1999).
revealed hundred-fold enhancements of active chlorine in the Antarctic

vortex (Solomon et al., 1987; Wahner et al., 1989; Kreher et al., 1996).
This technique also allowed study of the seasonal changes in Antarctic
chlorine activation and its links to PSC chemistry. These showed
a seasonal decline of OClO between late August and early October,
associated with increasing temperatures in the spring and hence the
cessation of PSC chemistry (Solomon et al., 1987).
Observations of the HCl column and, in particular, its ratio to the
HF column, strongly suggested that HCl had indeed been converted to
active chlorine in the Antarctic spring (Farmer et al., 1987; Toon et al.,
1989; Coffey et al., 1989). The recent global satellite data by Russell
et al. (1993) further demonstrate this behavior on larger spatial scales.
The first in-situ measurements of HCl showing evidence for conversion to
active chlorine were obtained in Arctic studies by Webster et al. (1993);
see Section 6.3.2.1. Concurrent global HCl and ClONO2 data from the
UARS satellite illustrated the simultaneous chemical conversion of both
species where polar stratospheric cloud surfaces were also present (Geller
et al., 1995; Yudin et al., 1997). A detailed view of the temperature
dependence of chlorine activation is provided from Antarctic in-situ
observations of both ClO and HCl, which dramatically illustrate rapid
activation at temperatures below about 195 K (Kawa et al., 1997; see
earlier studies by Toohey et al., 1993 and Schoeberl et al., 1993a,b), and
provide a key demonstration of rapid heterogeneous chemistry under
cold conditions.
The combination of simultaneous observations of ClO, HCl, OClO,
NO, NO2 , and other gases by a variety of independent chemical
methods demonstrates that the springtime Antarctic stratosphere is
indeed heavily perturbed compared to expectations from gas-phase
chemistry, and in a manner consistent with heterogeneous reactions on
PSC particles. Through the resulting enhancements in the ClO/Cly
ratio, chlorine’s effectiveness for ozone destruction is greatly enhanced.
For ClO abundances near 1-1.3 ppbv as observed since 1986/7 (and
BrO abundances near 7-10 pptv, see Brune et al., 1989), Antarctic
ozone is destroyed near 20 km in September at a rate of about 0.06-
0.1 ppmv/day, so that within about 40-60 days virtually all of the
ozone at this level can be depleted unless rapid dynamical resupply
occurs, broadly consistent with Figure 6.10 (see Anderson et al., 1989;
Murphy, 1991; MacKenzie et al., 1996). The cold temperatures observed
in the Antarctic during September in most years suggest that net
downward motion (which would tend to warm the air through adiabatic
compression) and horizontal mixing is relatively limited at this time of
year. This general picture of relative dynamical isolation in the heart
of the ozone depletion region is supported by a number of dynamical
studies (e.g., Hartmann et al., 1989; Manney et al., 1994b; 1995a,b;

Schoeberl et al., 1995; 1996).
The observations offer several different spatial and temporal finger-
prints that strongly support the identification of chlorine chemistry and
its perturbations by heterogeneous processes as the principal cause of
the ozone hole. The measurements shown in Figure 6.14 reveal that
the enhanced ClO occurs over about the 12-24 km range, the region
where PSCs are observed and where the ozone is depleted as shown in
Figure 6.10. The airborne data of Anderson and colleagues, as depicted
for example in Figure 6.13, demonstrate the steep latitudinal gradient in
ClO, consistent with the connection of the ozone hole to cold Antarctic
latitudes. A fully three-dimensional view of the same behavior is based
upon satellite data (Waters et al., 1993; 1999) is illustrated in Plate 15.
Seasonal observations of OClO, ClO, HCl, NO2 , and other gases have
been used to show that the large temporal changes in the abundances
of these chemical species are consistent with the time evolution of the
ozone hole and with heterogeneous chemistry. In short, the vertical,
latitudinal, and seasonal behavior observed in active chlorine and a host
of related species all provide independent evidence confirming the basic
processes that control the occurrence of the ozone hole.
6.3.1.4 The Composition and Chemistry of Polar Stratospheric

Clouds
Partly because of limited analysis of Antarctic stratospheric temper-

atures, early studies that followed the discovery of the ozone hole were
not specific about the type of particles of which the observed PSCs were
composed. It was generally assumed that the particles were mainly water
ice (Steele et al., 1983). Stratospheric ice clouds are frequently optically
thick and brilliant in color. Such clouds form when temperatures drop
below the temperature at which ice forms (frost point), and are now
referred to as Type 2 PSCs. However, sensitive satellite measurements
(McCormick et al., 1982) suggested that optically thinner PSCs were
also present at warmer temperatures.
As noted in Chapter 5, Toon et al. (1986) and Crutzen and Arnold
(1986) pointed out that the PSC particles might be composed not
only of water ice but also of solid nitric acid trihydrate (NAT). Both
studies noted that such composition could affect the impact on ozone
in two ways: (i) by reducing the amount of nitrogen oxide that could
be present (i.e., not only by forming nitric acid but by removing it
from the gas phase) and (ii) by raising the temperature at which clouds
could form, since thermodynamic analyses suggested that NAT could
condense at temperatures well above the frost point. These clouds

came to be known as Type 1 PSCs. In addition, Toon et al. (1986)
suggested that sedimentation of large particles could serve to ‘denitrify’
the stratosphere. The removal of nitric acid not only from the gas
phase but from the stratosphere altogether would have a potential to
further reduce NO2 concentrations, and hence enhance ClO/ClONO2
ratios and attendant chlorine-catalyzed ozone loss. McElroy et al. (1986)
also considered the possibility of nitric acid/water particles, suggesting
that nitric acid monohydrate (NAM) was likely to form.
Figure 6.13 shows evidence for NOy and H2 O removal in the same
region displaying enhanced ClO in the Antarctic stratosphere (poleward
of about 64◦ S in that particular transect) from airborne studies. Later
measurements documenting the ‘denitrification’ and ‘dehydration’ of
the Antarctic lower stratosphere include those of Santee et al. (1998,
1999), Voemel et al. (1995), and Pierce et al.(1994), but see below
for a discussion of the roles of denitrification and dehydration in ozone
depletion chemistry.
As discussed in Section 5.7.2, observatons have revealed that
polar stratospheric clouds are sometimes composed of supercooled
liquid solutions, raising important questions about the mechanisms of
formation and freezing of various types of PSCs. What is the effect
of different surfaces on ozone depletion chemistry? It is extremely
important to consider whether the phase and microphysical mechanisms
underlying PSC formation are important for ozone depletion. The
uptake of condensable vapors enhance surface areas when PSCs are
present. This increases the gas-particle collision frequency and hence can
enhance the rates of heterogeneous reactions. Background stratospheric
aerosols grow into Type 1a (solid) and/or Type 1b (liquid) particles as
they cool, then further grow into Type 2 water ice PSCs if temperatures
fall below the freezing point. The impact of enhanced surface areas
for chemistry depends not only on the frequency with which gases
strike these surfaces but also on the reactivity of the surfaces and the
availability of those gases.
Let us first consider how reactive the different surfaces are, and where
they are to be found in the stratosphere. Laboratory studies have shown
that water ice, NAT, and liquid ternary solutions are all effective for
activating chlorine heterogeneously, but with differing efficiencies and
with different dependencies on temperature, water vapor abundance,
and pressure (e.g., Carslaw et al., 1997a; JPL, 1997, and references
therein). These dependencies are related to the thermodynamics of the
different surfaces, which not only control their surface areas, but also
their composition (especially the uptake of HCl onto/into the particles).
Figure 6.15. The left panel illustrates laboratory data on the efficiency of the
reaction between HCl and ClONO2 for ice, nitric acid/water
/ solid surfaces,
and liquid sulfuric acid/water solutions. The right panel depicts the altitude
variation of the temperature at which the efficiency of this reaction on liquid
sulfuric acid/water solutions becomes greater than 0.3 for water vapor mixing
ratios that can be observed in the lower stratosphere.
Figure 6.15 summarizes a number of laboratory measurements of the

efficiency of the key heterogeneous surface reaction HCl + ClONO2 →
Cl2 + HNO3 (where 1 indicates reaction on every collision of ClONO2
with a surface, 0.1 indicates one reaction in 10 collisions, etc.). Water
ice is believed to be highly reactive wherever it can form, but the
thermodynamics of ice condensation imply that rather cold temperatures
are required to form it in the stratosphere (e.g., below about 188 K near
20 km). There is currently debate about the efficiency of NAT for this
reaction (Carslaw and Peter, 1997; JPL, 1997) as well as the conditions
under which NAT can form, but from a thermodynamic viewpoint NAT
may form near 20 km when temperatures drop below about 195 K,
thus allowing reaction on a solid surface at temperatures above the frost
point. In the case of liquid solutions, the efficiency of reaction depends
strongly on the fraction of water in the particle (e.g., Tolbert et al.,
1988; Hanson et al., 1994). The HCl + ClONO2 reaction becomes
more efficient for lower percentages of sulfuric acid (and higher water
in the liquid particles, which greatly increases the solubility of HCl;
see Robinson et al., 1998 for a detailed recent analysis). This reaction
takes place in liquid solutions with an efficiency greater than 1 in one
hundred (0.01 as shown in Figure 6.15) for temperatures colder than
about 197 K at 20 km, and an efficiency of 1 in ten for temperatures
below about 195 K. Ravishankara and Hanson (1996) have emphasized
that liquid PSCs can be comparable to or more effective than solid PSCs
for many surface reactions at temperatures below about 195 K at 20 km,
a point also illustrated in Cox et al. (1994), Borrmann et al. (1997a),
and Del Negro et al. (1997).
Because liquid aerosols are present throughout the global stratosphere
and because the water vapor pressures available to condense into them
increase with increasing total pressure, the temperatures at which
effective reactions may occur in liquid particles are higher for lower
altitudes (Hofmann and Oltmans, 1992), as is also shown in Figure 6.15.
This is an intriguing issue for both the polar and mid-latitude lower
stratospheres. Some studies suggest that both liquid and solid surfaces
can activate chlorine efficiently near the tropopause (Borrmann et al.,
1996; 1997b; Bregman et al., 1997; Solomon et al., 1997). Observations
of enhanced ClO and reduced NO close to the tropopause for relatively
wet (15 ppmv of H2 O) conditions provide some evidence for such liquid
surface chemistry at mid-latitudes (Keim et al., 1996; see also Brune
et al., 1990), although there is also evidence for some suppression of
such chemistry at the tropopause itself (Smith et al., 2001). Recent
work has emphasized that ammonium may play a role along with
sulfate in forming aerosols in the upper troposphere and perhaps at the
tropopause (see e.g., Tabazadeh and Toon, 1998; Kaercher and Solomon,
1999; Prenni et al., 2001), and the effect of ammonium ions or other
impurities such as organics on heterogeneous chemistry is not currently
known. Note that Figure 6.15 is based upon the thermodynamic model
of Carslaw et al. (1997a); its extrapolation to very high water vapor
pressures (e.g., greater than 5 ppmv at 200 mbar) is uncertain at present
and requires laboratory studies for those conditions (see Robinson et al.,
1998); impurities such as ammonium have also not been considered here.
It is useful to note that the HCl + ClONO2 reaction competes with
the H2 O + ClONO2 reaction for the available ClONO2 on liquid aerosol
surfaces. Thus, if HCl has been depleted, the rate of the latter reaction
increases, so that effective heterogeneous activation of chlorine is not
dependent upon both HCl and ClONO2 being present. Further, the
reactions of HCl with HOCl and HOBr are also quite efficient on liquid
aerosol under moderately cold and/or wet conditions (Ravishankara and
Hanson, 1996; JPL, 1997), providing additional pathways for chlorine
activation.
While there are differences and uncertainties in the reactivity of

various surfaces, laboratory data suggest that rather effective chlorine-
activating reactions can occur irrespective of particle phase below about
198 K at 20 km and below 200-210 K near 12-14 km. As an air
parcel cools and particle reactivities increase, liquid chemistry will occur
first. This may be followed by reactions on NAT and ice, depending
on factors including microphysics, the minimum temperature reached,
and whether or not all of the chlorine activation has already occurred
(Turco and Hamill, 1992). This latter point is critical. For example,
if effective chemical processing on liquid surfaces has depleted all of
the available HCl and/or ClONO2 within an air parcel, then further
lowering of temperature and formation of, for example NAT, may have a
limited effect on ozone depletion. Moreover, an increased rate of reaction
and/or an increased surface area (through, for example, formation of
NAT Type 1s or Type 2s) may not enhance ozone depletion in a time
averaged sense. If, for example, reactions on sulfate aerosols are sufficient
to activate all of the available chlorine within a day, ozone depletion will
not be increased if instead reactions on ice activate all of the chlorine in
an hour, since the ozone depletion is a process that occurs over a much
longer period (weeks) following the activation. Hence, the details of
the reactivities and the microphysics that control particle surface areas,
while playing a role to some degree, are not critical to formation of the
ozone hole (e.g., Portmann et al., 1996; Carslaw et al., 1997b). They are
likely to be more important at the margins, particularly regions where
temperatures are cold but not extremely cold.
6.3.1.5 The Depth and Area of the Ozone Hole

Current understanding of the key roles of liquid sulfate aerosols,
solid PSCs, temperature, and sunlight provides a framework for
understanding variations in the depth of the ozone hole (e.g., the
minimum possible vertical column) and its spatial area (i.e., the number
of square kilometers of depleted ozone).
Primary sources of sulfur to the stratosphere are carbonyl sulfide
(Crutzen, 1976; Chin and Davis, 1995) and explosive volcanic eruptions
that inject SO2 gas directly into the stratosphere (e.g., McCormick
et al., 1995) which subsequently forms liquid sulfate aerosols (see
Section 5.7.1). Observations of PSC extinction show that the major
eruptions of El Chichón´ in 1981 and Pinatubo in 1991 led to large
increases in particle surface areas in polar regions (e.g., McCormick
et al., 1995; Deshler et al., 1992; Thomason et al., 1997; Hofmann
et al., 1992; 1997). Hofmann and Oltmans (1993) showed that enhanced
aerosol surface areas due to Pinatubo and Hudson (a South American
volcano) expanded the altitude range of significant Antarctic ozone

depletion down to lower, warmer altitudes (i.e., 10-14 km) where
solid PSCs do not form. Observations of OClO in the Antarctic
fall season (March-April) at temperatures above 200 K in the year
immediately following Pinatubo also indicate significant activation of
chlorine through sulfate aerosol processing (Solomon et al., 1993).
Hence, both ozone and trace gas observations provide support for
the role of temperature-dependent heterogeneous chemistry on liquid
aerosols and its modulation by volcanic eruptions. Portmann et al.,
(1996) showed that volcanically enhanced PSC surface areas were likely
responsible for the sharp onset of the ozone hole in the early 1980s (due
to the El Chichon
´ eruption), and for the very deep ozone holes observed
in the early 1990s following the Pinatubo eruption.
The strong sensitivity of heterogeneous chemistry to temperature
suggests that warm Antarctic springs, or years of early final warmings,
could display less ozone than colder ones. Indeed, the relatively
small Antarctic ozone depletions in 1986, 1988, and 2002 were
partly due to temperature effects in these warm years — and are
qualitatively simulated by numerical models of stratospheric chemistry
when constrained by observed temperatures. Since the fundamental
cause of warmer and colder years is the propagation of waves into the
Antarctic stratosphere, dynamical impacts (e.g., resupply of ozone from
other regions) also play a role in dampening the ozone loss in warm years.
In 2002, a massive Antarctic wave led to stratospheric temperatures in
September that were higher than observed in the previous 50 years,
affecting both the area and depth of the ozone hole (as shown in Plate
9; see also Allen et al., 2003). Current models can simulate many key
aspects of the onset, growth, and interannual variability in the depth
of the Antarctic ozone hole based upon the growth of stratospheric
chlorine in past decades along with the modulating effects of known
temperature fluctuations and volcanic input of particulates (Portmann
et al., 1996; WMO/UNEP, 2003). However, the possibility of changes in
dynamics remain incompletely understood, and it is not known whether
the unprecedented behavior of the ozone hole of 2002 was a random
event related to the chaos of atmospheric flow or the beginning of a
more systematic shift (see 6.3.2.3).
The Antarctic ozone hole cannot get very much deeper than it
presently is — since virtually all the ozone is now removed each year
in those altitudes where PSCs occur. Ozone column minima near
100 DU were observed after the Pinatubo eruption, when the efficiency of
heterogeneous chemistry was especially enhanced by very large aerosol
amounts over a broad altitude range as noted above. This suggests
that the ozone hole could only become deeper in the absence of major
volcanic eruptions if PSCs were to extend to significantly higher or lower

altitudes, implying a major change in south polar thermal structure.
While substantial changes of this type appear unlikely, the possibility of
increases in the horizontal extent of the hole cannot be ruled out. The
ozone hole is largely confined to the Antarctic polar vortex, and year-
to-year variability in its size can occur. These fluctuations are mainly
linked to variability in temperature (e.g., slightly colder or warmer years
than average). But changes in exposure to sunlight also play a role,
especially in August and early September (e.g., through atmospheric
waves that photolyze Cl2 particularly early in the depletion season).
Figure 6.16 shows the growth in size of the ozone hole during the 1980s,
and its interannual ups and downs. In general, years of deeper ozone
holes are also years of large spatial extent. Several decades will be
required before changes in the size or depth of the ozone hole will signal
a recovery or a worsening that lies outside of the modulations introduced
by temperature, volcanic aerosol content, and wave-driven excursions in
sunlight exposure. Some studies suggest that cooling (e.g., that expected
from greenhouse gases) may enhance ozone losses in the ‘edge’ region
around the vortex in the future (Lee et al., 2001), perhaps increasing
Figure 6.16. Observations of the year-to-year variations in the size of the

Antarctic ozone hole based on satellite observations of the region where total
ozone drops below 220 Dobson Units. Figure courtesy of P. Newman (personal
communication).
the size of the hole. However, as discussed below in the context of the
Arctic ozone depletion, the interactions of waves with the mean flow
is a complex process that makes a definitive prediction of the future
temperatures of the polar and sub-polar stratospheres a challenge.
6.3.1.6 Changes in Understanding

Here we pause for a brief summary of the conceptual picture of
Antarctic ozone depletion that emerged in the late 1980s and describe
how it has changed. Initial observations of Antarctic chemistry
as discussed above showed evidence for both heterogeneous chlorine
activation on PSCs and for denitrification. An understanding emerged
that chlorine-activating reactions took place on solid PSCs in Antarctic
winter, accompanied by denitrification that allowed the depletion to
persist in spring, even in the absence of further PSC formation. This
picture was simple and very easy to explain in chemical terms. However,
the current understanding suggests that denitrification can increase
ozone destruction somewhat, but is not required and probably not
sufficient on its own for massive polar ozone losses. Even in the
Antarctic, denitrification is not 100% effective — some reactive nitrogen
remains present (as HNO3 ). That HNO3 will photolyze in spring,
reforming ClONO2 . Since NOy abundances are roughly 10 times
greater than Cly abundances in the lower stratosphere, photolysis of
the leftover HNO3 can neutralize reactive chlorine in the absence of
further heterogeneous processing even if denitrification is 80% effective at
removing NOy . A more effective way to drive ozone depletion processes
is continued chlorine activation on solid or liquid aerosols in spring. Such
reprocessing is sufficient to keep the chlorine active in cold sunlit air
whether denitrified or not (Portmann et al., 1996; Chipperfield and Pyle,
1998). This heterogeneous chemistry must occur faster than photolysis
of HNO3 to maintain high ClO levels, but HNO3 photolysis is not a very
rapid chemical reaction and it has been shown that PSC reprocessing
can support rapid springtime ozone loss for typical Antarctic conditions.
Observations following the eruption of Mt. Pinatubo further support the
view that liquid aerosol reprocessing has been a key factor in determining
the depth and duration of the ozone hole after major eruptions. More
generally, the expansion of heterogeneous chemistry from ice, to NAT, to
liquid sulfate aerosols has lessened the expected dependence of the ozone
loss on extreme cold to one of relative cold, thereby expanding the height,
time and latitude ranges in which ozone depletion may be expected to
be enhanced by heterogeneous processes that affect chlorine partitioning.
Further, Arctic observations suggest that ozone depletion there is limited
mainly by the seasonal cycle of north polar temperatures (i.e., whether
or not temperatures are cold in the spring season when sunlight is also
available), and not by denitrification. Gao et al. (2001) showed that
extensive denitrification, as documented by Fahey et al. (2001), did
enhance O3 loss (by about 30%) in a narrow layer in 2000, but Tabazadeh
(2001) showed that the more important question may be the depth of
such layers. Years of deepest ozone depletion observed in the 1990s were
often associated with little or no denitrification (Santee et al., 1998;
1999), and the years of greatest denitrification (e.g., 1999/2000) were
not those of greatest Arctic ozone loss. This is the subject of the next
section.
6.3.2 Arctic Ozone Depletion
Perhaps ironically in view of the extremely remote nature of the

Antarctic, ozone depletion was more readily observed there than in the
Arctic. This was due in part to the fact that no correponding ‘hole’
developed in the Arctic stratosphere in the early 1980s, but also to
the paucity of ground-based long-term measurements in the high Arctic
and to the greater local variability of Arctic ozone associated with
atmospheric waves. As the mechanism for Antarctic ozone depletion
began to be elucidated in the latter half of the 1980s, it was understood
in general terms that Arctic ozone depletion would likely be smaller due
to warmer temperatures (hence, fewer PSCs as already documented in
McCormick et al., 1982) and the associated dynamical differences (i.e.,
a less isolated vortex). The top panel of Figure 6.17 illustrates the
climatological differences in the seasonal cycles of temperature for 65◦ N
and 65◦ S, edge regions of the Arctic and Antarctic. Colder temperatures
are typically found at higher latitudes, but this region is shown here
in order to illustrate accompanying satellite total ozone data (which is
only available in the sunlit atmosphere). Perhaps most importantly,
the typical springtime increase in stratospheric temperatures occurs in
association with much earlier stratospheric warmings in the north than
in the south (e.g., Andrews et al., 1987), suggesting that the overlap
between cold temperatures and sunlight must be limited, and the Arctic
ozone depletion hence less severe (see e.g., the review by Pyle et al.,
1992). However, not all years are typical, and there is some evidence
that the Arctic stratosphere may be systematically changing — a point
discussed below.
Figure 6.17. Observations of the average temperatures at 65◦ S and 65◦ N from
the Fleming et al. (1990) Cospar International Reference Atmosphere (CIRA)
climatology (top). The middle and bottom panels show satellite measurements
of the annual cycles of total ozone at 65◦ N and 65◦ S in the late 1970s and in
1996 and 1997 (from SBUV/SBUV2, courtesy of R. Nagatani). From Solomon
(1999).
6.3.2.1 Chemical Processes in the Arctic
As in the Antarctic, direct observations of a broad range of chemical

species have shown that heterogeneous chemistry greatly perturbs the
composition of the Arctic vortex. Evidence for effective winter activation
of Arctic chlorine was first provided by measurements of enhanced OClO
(Solomon et al., 1988) and ClO (Brune et al., 1990). Both gases have
since been observed in many other studies (e.g., Schiller et al., 1990;
Pommereau and Piquard, 1994a; Perner et al., 1994; Pfeilsticker and
Platt, 1994; Brune et al., 1990; 1991; Toohey et al., 1993; Waters et al.,
1993; Crewell et al., 1994; Bell et al., 1994; de Zafra et al., 1994; Shindell
et al., 1994; Donovan et al., 1997). Decreased NO and NO2 were also
observed with several independent methods (Fahey et al., 1990a; Noxon
et al., 1978; Toon et al., 1994; Mankin et al., 1990; Wahner et al., 1990;
Pommereau and Piquard, 1994b; Goutail et al., 1994; Pfeilsticker and
Platt, 1994; Van Roozendael et al., 1994). The column abundances of
HCl and HF further supported the view that chlorine activation on PSCs
must be effective in the Arctic (Toon et al., 1994; Mankin et al., 1990;
Traub et al., 1994).
In-situ measurements of HCl were particularly important in sharp-
ening the link between enhanced ClO and conversion from HCl, as
shown for example in Figure 6.18 (see Webster et al., 1993; Michelsen
et al., 1999). Measurements of ClONO2 also displayed evidence for
heterogeneous processing on PSCs (e.g., von Clarmann et al., 1993;
Roche et al., 1994; Oelhaf et al., 1994; Adrian et al., 1994; Geller
et al., 1995; Yudin et al., 1997). Concurrent in-situ and space-based
observations of ClO and PSCs together with trajectory studies further
linked the activated chlorine to heterogeneous chemistry (Jones et al.,
1990b; Yudin et al., 1997; Dessler et al., 1998 and references therein).
Thus, the same general fingerprints of heterogeneous chemistry that
were first observed in the Antarctic were not only apparent but further
strengthened by Arctic data.
Observations of NOy and water vapor displayed signs of sporadic
and limited denitrification and dehydration (compare Figures 6.13 and
6.18), in marked contrast with the pervasive and extensive characteristics
of these chemical conditions in the Antarctic (Kawa et al., 1990;
Fahey et al., 1990b; see later work by Kondo et al., 1994; Oelhaf
et al., 1994; Khattatov et al., 1994; Rinsland et al., 1996; Santee
et al., 1998a,b). Denitrification was also observed to occur without
accompanying dehydration, suggesting that NAT particles sometimes
can become large enough to precipitate without necessarily growing to
Type 2 (water ice) proportions (see e.g., Toon et al., 1990; Gandrud
Figure 6.18. Observations of the chemical composition of the Arctic

stratosphere from the ER-2 aircraft in February 1992. The data show high
ClO abundances associated with reduced HCl abundances as would be expected
from heterogeneous conversion. H2 O does not display evidence for dehydration
on this Arctic transect. From Solomon (1999).
et al., 1990; Salawitch et al., 1989; Koop et al., 1995; Fahey et al.,
2001).
Based on ClO observations (Brune et al., 1990) and related model
calculations, observed and calculated rates of ozone loss in February
1989, were shown to be of the order of 20 ppbv/day near 20 km
(Schoeberl et al., 1990; Salawitch et al., 1990; McKenna et al., 1990).
Further, the BrO observations of Toohey et al. (1990) revealed that
the ClO-BrO catalytic cycle was probably of particular importance for
the Arctic, since ClO enhancements were generally smaller than in the
Antarctic and hence the efficiency of the ClO dimer cycle was reduced;
note that the rate of the latter depends on the square of ClO density,
e.g., Salawitch et al., 1990; 1993). However, the early warming observed,
for example, in February 1989, as illustrated in Figure 6.19, prevented
extensive total ozone loss in that year. Some early studies suggested that
the less extensive denitrification of the Arctic would limit ozone losses
Figure 6.19. Observations of the zonally averaged temperatures in the Arctic

for 82◦ N near 18 km, from the National Center for Environmental Prediction
(NCEP) meteorological database for two illustrative recent years, together with
Antarctic observations at 82◦ S in 1997 (shifted by six months for comparison).
there (Brune et al., 1991; Salawitch et al., 1993) through less effective
NOx reduction and hence an early cutoff of the depletion process in
spring; the previous section illustrates how new understanding of liquid
aerosol chemistry has affected this picture. Thermal decomposition of
the Cl2 O2 dimer (which cuts off the ClO dimer ozone loss cycle) also
affects the degree of possible ozone loss as air warms in spring even if
denitrified (McKenna et al., 1990; MacKenzie et al., 1996).
6.3.2.2 Quantifying Arctic Ozone Depletion

The more complex dynamics of the Arctic vortex as compared to
the Antarctic demand the application of sophisticated tools for analysis
of ozone destruction. The greater wave activity of the Northern
Hemisphere can enhance ozone losses even in winter by increasing the
exposure of polar air to sunlight in the distortions caused by atmospheric
waves as compared to the Southern Hemisphere (see e.g., Jones et al.,
1990a). However, the same wave activity can warm the air and perhaps
even distort it sufficiently to mix with its surroundings, thus reducing

ozone depletion. Detailed methods have been used to evaluate the
air parcel trajectories along which ozone and other trace gases are
transported (Schoeberl et al., 1990; Manney et al., 1994b; 1995a,b;
1996). These help to quantify the amount of ozone chemically destroyed
by revealing that while the time evolution of inert tracers such as N2 O
can be well simulated in the Arctic using such approaches, the evolution
of ozone shows large departures from conservation that likely reflect
chemical loss (Manney et al., 1994a; 1995a,b,c; 1996; 1997). Further, the
regions of apparent ozone depletion identified in this manner occur in
regions of enhanced ClO as measured by concurrent satellite observations
(e.g., Waters et al., 1993; Manney et al., 1994a; 1995c; Lutman et al.,
1994a,b; MacKenzie et al., 1996).
Tracer-ozone correlations are another method used to provide insights
into polar ozone loss. Briefly, changes in the amount of ozone
observed for a given amount of conserved tracer such as N2 O or CH4
provide a useful (albeit imperfect) diagnostic for ozone loss based
upon understanding of ozone-inert tracer relationships and their spatial
distributions (Proffitt et al., 1993; Müller
¨ et al., 1996; 1997a,b; Plumb
and Ko, 1992). Using satellite data for CH4 , O3 , and HCl, for
example, Müller
¨ et al. (1996) suggest that 60 DU of total Arctic
ozone was depleted on constant CH4 surfaces in a manner inconsistent
with transport from any other region of the stratosphere in the Arctic
winter of 1991/2. The reduced ozone was associated with pronounced
HCl depletion observed in the same air, as expected based upon
heterogeneous chemistry on PSCs.
Another method of quantifying ozone destruction involves the use of
trajectory analyses of airflow together with multiple ozonesondes to find
‘matches’ wherein the air observed at one site is observed again some
days later. Changes in the observed ozone then provide a measure of
ozone loss (von der Gaathen et al., 1995). This approach has yielded
strong evidence for extensive Arctic ozone depletion that is closely tied
to cold temperatures near 195 K (Rex et al., 1997; 1998). Understanding
can be further tested by comparing the observed depletion derived from
such ‘matches’ with chemistry calculations along the same trajectories.
These studies have shown good agreement in February and March, but
some evidence for mid-winter ozone loss that exceeds photochemical
theory has also been suggested (Rex et al., 1998; Becker et al., 1998).
This intriguing result may suggest dark or twilight chemistry that is not
fully understood, or contributions from dynamical phenomena not fully
captured in the trajectories.
Fully three-dimensional models driven in some cases by the
meteorological data for specific years have also been used to probe the
Arctic ozone losses and test photochemical understanding. These models

have succeeded in explaining much of the observed ozone depletion,
documenting its connections to chemical processes, and even reproducing
much of the observed variability seen from one year to another (see e.g.,
Chipperfield et al., 1994; 1996; Chipperfield and Pyle, 1998; Deniel et al.,
1998; Douglass et al., 1995).
Taken together, these combined approaches to transport analyses
using tracers, ‘matches’, chemical transport models, or Lagrangian
calculations along with ozone and trace constituent observations provide
strong evidence for a chemically-driven Arctic ozone loss (order of 60-
120 DU) in several recent years. Each approach is subject to different
sources of quantitative error and uncertainty, such as inaccuracies
in temperature data used as input in observationally-based transport
studies, incomplete understanding of the factors influencing tracer-tracer
correlations, and small scale dynamical processes that are not well
represented in modelling studies (e.g., mountain waves). In spite of these
shortcomings and in contrast with the Antarctic, there is substantial
evidence for a dynamical contribution to recent trends as well — these
are discussed in the next section.
6.3.2.3 Variability of Arctic Temperatures

Since Antarctic ozone depletion occurs mainly in September under
cold conditions, it is natural to consider whether comparable conditions
are ever attained in the Arctic in the analogous month of March.
Nagatani et al. (1990) pointed out that while Arctic temperatures below
about 195 K appeared to be quite rare based on the available record
at that time (which extended back to about the 1950s), they are not
unknown. For example, the Arctic winter of 1975/6 was a year in which
March Arctic temperatures were close to those typically seen in the
Antarctic in September, but chlorine loading was small in 1975 and no
discernible Antarctic ozone hole was observed at that point (see Figure
6.9). Nagatani et al. (1990) noted that extensive Arctic ozone loss might
be expected if such meteorological conditions were to be realized in an
atmosphere with current chlorine loadings.
There were several unusually cold Arctic winters in the 1990s, with
correspondingly large Arctic ozone losses (see Newman et al., 1997;
Coy et al., 1997) illustrated in Figure 6.9. Not all the years since
1990 have been cold, as reflected for example in the high spring ozone
observed in 1998, for example. Enhanced volcanic aerosol from the
Mt. Pinatubo eruption probably contributed to the very low ozone
observed in 1992 and 1993 through enhanced PSC chemistry, but
the continuing depletions in, for example, 1996 and 1997 suggest a
strong effect of temperature. Figure 6.19 illustrates the full seasonal

behavior of temperatures observed in some recent cold years, and
contrasts their behavior with the Antarctic. As has already been
emphasized, the warmer temperatures generally observed in the winter
Arctic stratosphere as compared to the Antarctic reflect adiabatic
heating associated with faster downward motion, which also leads to
a rapid wintertime increase in Arctic total ozone (from values of around
300 DU in September to as much as 450 DU at the spring maximum in
March in 1979 and 1980, for example, as shown in Figure 6.17). In the
much colder Antarctic, pre-ozone hole total ozone did not show such a
winter increase (remaining instead near 250-300 DU from March through
September, see Figures 6.11 and 6.19), suggestive of a much less dynamic
vortex. In today’s Antarctic atmosphere, an abrupt drop in ozone occurs
in September deep in the vortex as shown in Figure 6.11, reflecting
rapid chemical removal in sunlit air with limited dynamical resupply as
discussed earlier. In the relatively stagnant Antarctic vortex, the total
ozone therefore actually decreases in spring to form a ‘hole’ compared to
the surrounding mid-latitude air. In the more dynamic Arctic, transport
replaces a substantial portion of the ozone lost, even in the very cold
years of the 1990s (see e.g., Manney et al., 1997). Indeed, Figure 6.17
shows that even in 1996 and 1997, Arctic ozone continued to increase at
65◦ N during spring — it simply did not do so as rapidly as it did in 1979
and 1980. The fact that Antarctic ozone drops rapidly in spring while
Arctic ozone does not is one key reason why the depletion in the Arctic
is not generally referred to as an ozone hole. Hence, large chemical
ozone losses of the order of 60-120 DU do occur in the Arctic, but it is
difficult to say that a ‘hole’ formed since the ozone was apparently being
simultaneously resupplied to a major extent (see the detailed analyses
by Manney et al., 1997; Müller
¨ et al., 1997a,b). The formation of a true
Arctic ozone hole according to this definition would require not only
cold March temperatures but also very cold temperatures throughout
the winter, in order to cause both activation of chlorine in sunlit air
and to inhibit the winter buildup of ozone through downward transport.
Such conditions did not occur in the 1990s (see Coy et al., 1997; Zurek
et al., 1996).
It is important to note that denitrification was observed but was
rather limited in degree in the Arctic springs of 1993, 1996, and 1997
(Santee et al., 1998; 1999), so that the observations of ozone depletion
of the order of 60-120 DU in each of these years are not associated with
extensive denitrification. Rather, as in the Antarctic and consistent
with current understanding of liquid aerosol chemistry, the evidence
suggests that heterogeneous reactions in the sunlit atmosphere are
mainly responsible for maintaining the high ClO (Santee et al., 1997)
that depleted the Arctic ozone in those years (Manney et al., 1997).
Figure 6.11 shows that the Antarctic ozone hole fills in to a substantial
degree around December. This increase in ozone is associated with
the final stratospheric warming that breaks down the polar vortex
and brings in ozone-rich air from lower latitudes. However, it is also
important to note that the seasonal increase in Antarctic ozone that
marks the warming now occurs systematically later than it did prior
to the discovery of the ozone hole — note the ozone seasonal increases
observed not in December but in November in the 1970s (see Jones
and Shanklin, 1995). Theory suggests that the massive destruction of
Antarctic ozone has greatly altered the thermal budget of the Antarctic
stratosphere, and is likely to be responsible for this considerable delay
in the timing of the Antarctic stratospheric spring warming (e.g.,
Randel and Wu, 1999). If there were to be a similar delay in the
breakdown of the Arctic polar vortex caused by ozone depletion or by
another mechanism (such as greenhouse gas changes), then it could have
profound implications for Arctic ozone loss in the future. There is some
evidence that Arctic stratospheric warmings are tending to occur about
2 weeks later than they did in the past (Zhou et al., 2000). If this trend
were to continue, Arctic ozone depletion may deepen in the future, since
the spring seasonal overlap between cold temperatures and the amount
of sunlight is critical in driving ozone depletion.
In summary, there is abundant evidence for some chemical perturba-
tions and ozone destruction in the Arctic even in relatively warm years,
but the degree of ozone depletion depends upon cold temperatures in
sunlit conditions, just as in the Antarctic. An unprecedented number
of cold years occurred in the Arctic in the 1990s. Each of these was
reflected in low ozone in the Arctic record as shown in Figure 6.9. Five
of the years from 1991-1998 have been significantly colder than average
(Coy et al., 1997; Zurek et al., 1996), and there is evidence that Arctic
warmings may be occurring later than before. This series of unusually
cold years raises the key question of cause. Randel and Wu (1999) argue
that the cooling observed in both the Arctic and the Antarctic is due
to the ozone depletion itself. They propose a feedback mechanism as
in Shine (1986) wherein ozone losses lead to colder temperatures and
hence even greater depletion. The studies by Thompson and Wallace
(2000) and Thompson et al. (2000) suggest that changes in the dynamics
of the north polar vortex may be linked to the underlying tropospheric
wave field, particularly the North Atlantic and Arctic Oscillations (NAO,
AO) in wave patterns (i.e., the spatial distribution of tropospheric
disturbances). These authors suggest a wave-driven systematic linkage
between tropospheric waves and stratospheric temperature, which could
reflect ozone changes. Hartley et al. (1998) argue for a similar linkage
involving the modification of stratospheric dynamics due to the ozone
changes, with tropospheric propagation as a key element. Other
evidence for downward propagation of stratospheric waves is presented
by Baldwin and Dunkerton (2001). However, other studies argue for
smaller effects, and suggest mechanisms including greenhouse gases and
surface temperatures could account for these changes (see references in
Thompson et al., 2000).
As has already been emphasized, the ‘greenhouse effect’ due to
increases in CO2 and other gases warm the planet surface but cool
the stratosphere (e.g., Fels et al., 1980). This effect is predicted to
be small (only a few tenths of a degree in today’s atmosphere, far
less than the recent coolings observed in the Arctic), but dynamical
amplification of such changes is also possible. A number of studies
have suggested that increased CO2 and other greenhouse gases could
substantially affect stratospheric dynamics, thereby cooling the Arctic
stratosphere far more than the radiative effect alone would predict and
leading to greater Arctic ozone depletion (e.g., Austin et al., 1992;
Shindell et al., 1998). The work of Shindell et al. argues for a key
role for such a dynamical feedback both in the 1990s and perhaps in
future years, with the peak Arctic ozone losses being predicted to occur
near 2010, well after the expected peak of chlorine loading (see also
Dameris et al., 1998). However, some other three-dimensional models
display the opposite effect — those studies suggest that the Arctic
stratosphere could warm rather than cool in the future if emissions
of CO2 and other greenhouse gases continue. The different signs of
the temperature effects predicted by various models demonstrate great
sensitivity to how planetary waves propagate, dissipate, and couple to
the mean circulation and temperature fields. At present the possibility
that the recent colder Arctic temperatures are simply part of a natural
low-frequency cycle that could, for example, induce a series of colder
years every fifty years or so cannot be ruled out given the short record
of existing global stratospheric temperature data. Hence, while it is clear
that there has been significant chemical ozone depletion associated with
the cold Arctic winter/spring seasons of recent years, the fundamental
reasons for those cold temperatures and their future persistence remain
a topic of research.
6.3.3 Mid-Latitude Ozone Depletion
The discovery of the Antarctic ozone hole naturally raised the question
of whether mid-latitudes might also display greater ozone depletion than
expected. Within a few years after the ozone hole was discovered,
statistically significant trends in ozone were documented at northern
mid-latitudes (Ozone Trends Report, 1988, and references therein). By
the 1990s, significant trends had been established for both northern
and southern mid-latitudes, not only in winter and spring but also in
summer (WMO/UNEP, 1991; 2003; Stolarski et al., 1991; McPeters
et al., 1996a,b; Harris et al., 1997). Mid-latitude ozone column trends
as of the late-1990s were of the order of 5-10%, much smaller than those
of the Antarctic (Figure 6.20) but far greater than gas-phase model
predictions. Analyses of vertical profiles measured by ozonesondes and
satellites suggested that the bulk (at least 2/3) of the northern mid-
latitude ozone decline is occurring in the lower stratosphere near 12-
20 km (see McCormick et al., 1992; Miller et al., 1995; WMO/UNEP,
1994; Bojkov and Fioletov, 1997; SPARC, 1998). Figure 6.20 depicts the
ozone record from Arosa, Switzerland, the sole station in the world where
ozone measurements date back to the 1920s. Systematic records were
begun at many other stations in the northern hemisphere in the 1950s or
60s (see Ozone Trends Panel Report, 1988), but the length of the record
at Switzerland is unique. The figure shows the large seasonal cycle of
ozone at mid-latitudes and illustrates that variability in that cycle is
one factor that makes observation of a trend difficult. However, when
attention is restricted to a single month, such as January for example
(shown by the filled circles in the figure), the data do suggest a trend
after about 1980. The trend becomes far clearer when the data are
averaged over several years, in order to reduce sensitivity to processes
such as the quasi-biennial oscillation or the El Niño,
˜ which can cause
short-term fluctuations of a few years due to dynamical perturbations.
Here we present a five-year running mean. The trends in this long-term
averaged record are evident; indeed the figure shows evidence for about
a 6% maximum trend in mid-latitude ozone and displays considerable
similarity to the Antarctic depletion (albeit much smaller in magnitude).
We next probe the causes for this striking change — this Antarctica in
miniature — observed for mid-latitude ozone.
6.3.3.1 Heterogeneous Chemistry and Mid-Latitude Ozone

De pletion
One mechanism that could affect mid-latitude ozone depletion in
the lower stratosphere is heterogeneous chemistry. It had long been
suspected that some heterogeneous process involving N2 O5 might be
responsible for the Noxon ‘cliff’ and for anomalously high HNO3
abundances in middle to high latitudes of the Northern Hemisphere
(e.g., Wofsy, 1978; Noxon, 1979; Austin et al., 1986), but it was not
Figure 6.20. Observations of total ozone at Arosa, Switzerland since the 1920s,
the world’s longest-running total ozone record. The top panel shows monthly
mean data for all months and years, illustrating both the large seasonal cycle
and its variability. Januaries are indicated by the filled circles, and this month
does display a long-term downward trend. The bottom panel presents the 5-
year running mean of the same data. Periods when modelling studies suggest
volcanic influences due to the El Chichón ´ and Mt. Pinatubo eruptions are
indicated.
until the late 1980s that laboratory studies showed that N2 O5 can
hydrolyze rapidly (reaction efficiency of about 0.1) on sulfuric acid/water
films (Tolbert et al., 1988) and particles (Mozurkiewicz and Calvert,
1988). Hence the possibility of heterogeneous chemistry on the liquid
sulfate layer that is pervasive throughout the stratosphere began to be
considered in earnest (but see also Cadle et al., 1975, for an early and
interesting exploratory paper).
The hydrolysis of N2 O5 reduces NOx and its impact on ozone in the
lower stratosphere, and indirectly enhances the effect of ClO through
its control of the ClONO2 /ClO ratio, as discussed earlier. Recent
studies have examined the dependence of the N2 O5 hydrolysis reaction
on temperature and pressure (Robinson et al., 1997) and have probed
reaction conditions in extensive detail (JPL, 1997; 2000). This key
reaction and the related hydrolysis of bromine nitrate (Hanson et al.,
1996) both take place rapidly at virtually all stratospheric conditions,
making their influence extremely widespread (and, as shown below,
extremely important). In addition to these indirect effects, however,
there is evidence for some direct activation of chlorine on liquid sulfate
aerosols as well. As noted in connection with polar chemistry, Tolbert
et al. (1988) suggested that ClONO2 could react with water and with
HCl on sulfuric acid/water surfaces, but with a strong dependence on
the particle composition, specifically the water content of the particle
(hence on the temperature and water vapor pressure).
The laboratory investigations of Tolbert (1988) and Mozurkiewicz
and Calvert (1988) prompted Hofmann and Solomon (1989) to study
the role of N2 O5 hydrolysis and chlorine activation on sulfuric acid
aerosols at mid-latitudes, particularly under volcanic conditions when
such processes would be enhanced. They suggested that this chemistry
could be significant for both background and volcanically enhanced
conditions, and that the ozone reductions noted by several authors
(e.g., Adriani et al., 1987) following the eruption of El Chichón
´ in 1981
might be linked to heterogeneous reactions similar to those occurring
in Antarctica, albeit with reduced rates. Observations of marked
reductions in NO2 over New Zealand after the El Chichón ´ eruption
(Johnston and McKenzie, 1989) provided some of the first chemical
evidence that such processes could be important at mid-latitudes.
Several modelling studies further probed the role of this chemistry
in determining global ozone trends and related questions of chemical
partitioning and odd oxygen destruction cycles studies (e.g., Rodriguez
et al., 1991; 1994; Brasseur and Granier, 1992; Prather et al., 1992;
McElroy et al., 1992; Pitari et al., 1993; Toumi et al., 1993). Prather
et al. (1992) pointed out that the hydrolysis of N2 O5 saturates beyond
a certain aerosol load at which N2 O5 is converted to HNO3 as fast as
it can be formed, so that further increases in aerosol do not affect NOx

abundances through this process.
Rodriguez et al. (1991) and McElroy et al.(1992) noted that
hydrolysis of N2 O5 would have the effect of dramatically altering
the competition between the various catalytic cycles in the lower
stratosphere, enhancing the roles of the odd hydrogen and odd
chlorine/bromine destruction mechanisms, even for background aerosol
conditions. Direct observations by Wennberg et al. (1994) later
confirmed this view by providing simultaneous measurements of a suite
of key radicals including OH, HO2 , NO, and ClO near 20 km; these
data can be related in hindsight to the anomalously low NO mixing
ratios reported in the mid-latitude lowermost stratosphere by Ridley
et al. (1987). Cohen et al. (1994) present a detailed chemical argument
demonstrating the dominance of HOx chemistry in the natural lower
stratospheric ozone balance based on observations. Taken together, this
improved understanding of the balance of terms among chemical cycles
of ozone destruction is a key building block for attempts to evaluate
ozone loss, which is tied to the competition of chlorine and bromine-
catalyzed destruction compared to other chemical processes and to
transport. In short, slower rates of ozone loss through other processes
(especially NOx chemistry) result in a larger relative role for human-
induced perturbations at mid-latitude due to chlorine and/or bromine
increases.
From about 1988 to the early 1990s, the scientific understanding
of mid-latitude ozone depletion evolved from a gas-phase picture into
the expectation of enhanced ozone depletion at least via the N2 O5
hydrolysis process, not only for volcanically perturbed conditions but
also for background aerosol loading. The eruption of Mt. Pinatubo in
June, 1991, was the largest in the 20th century, and occurred near the
peak loading of atmospheric chlorine. This geophysical event provided
numerous lines of evidence suggesting that heterogeneous chemical
reactions on sulfate aerosols play a role in ozone chemistry and its
depletion (see the review by Toohey, 1995). Gleason et al. (1993) were
the first to report record low northern mid-latitude ozone abundances in
the following year. Hofmann et al. (1994) and McGee et al. (1994)
demonstrated that substantial ozone losses occurred in the lower
stratosphere following Pinatubo, particularly in winter and spring, with
peak local depletions near 20 km at 40-50◦ N as large as about 25%.
Randel and Cobb (1994) showed that changes in temperatures relating
to aerosol heating can provide an important means of distinguishing
ozone losses due to volcanic eruptions from those relating to the QBO,
El Nino,
˜ or other perturbations in statistical analyses of ozone data
(see also Jaeger and Wege, 1990; Bojkov et al., 1993; Zerefos et al.,
1994). The high aerosol load present just after the eruption in mid-
1991 changed stratospheric heating and hence reduced tropical ozone
through dynamical effects (Brasseur and Granier, 1992), but this lasted
only a few months and was largely confined to the tropics (see Schoeberl
et al., 1992; Tie et al., 1994). For reviews of the many studies
establishing the large and persistent mid-latitude ozone changes after
Pinatubo, see Toohey (1995) and WMO/UNEP (1994; 1998). Because
the stratospheric Brewer-Dobson circulation transports material upward
and poleward, major volcanic eruptions that inject material into the
tropical stratosphere can have the greatest and longest impacts on
global ozone, while volcanic injections at higher latitudes are removed
by downward motion. Both El Chichón ´ and Pinatubo are tropical
volcanoes.
Chemical measurements after Pinatubo have identified many signa-
tures of heterogeneous reactions on sulfate aerosols at mid-latitudes that
are akin to those occurring in the ozone hole region. Observations
from New Zealand showed both reduced NO2 and enhanced HNO3
column abundances (Johnston et al., 1992; Koike et al., 1994). Aircraft,
ground-based, balloon- and shuttle-borne experiments revealed similar
large changes in NOx /NOy partitioning associated with the roughly
thirty-fold increases in aerosol surface observed (e.g., Rinsland et al.,
1994; Webster et al., 1994; Fahey et al., 1993; Mills et al., 1993;
Coffey and Mankin, 1993; Sen et al., 1998). Figure 6.21 shows direct
observations of perturbations in NOx /NOy and ClO/Cly at 20 km from
Fahey et al. (1993) associated with the buildup of Pinatubo aerosols
at mid-latitudes. NOx /NOy decreases follow the behavior broadly
predicted by Prather (1992) and expected from the dominance of N2 O5
hydrolysis. It is important to note that these observations confirm that
some of the ozone depletion observed following Pinatubo must have
been due to chlorine chemistry, although they do not in themselves
establish its magnitude. Randinaya et al. (1997) and Slusser et al. (1997)
used summer polar observations of NO2 to show evidence for BrONO2
hydrolysis on sulfate aerosols as well. Observations of enhanced OH
at sunrise further suggest that the latter process is significant not only
in reducing NOx via heterogeneous chemistry but also as a source of
OH (Hanson and Ravishankara, 1995, 1996; Salawitch et al., 1994; Lary
et al., 1996).
Turning to the key chlorine-related species, Avallone et al. (1993a,b)
and Wilson et al. (1993) showed post-Pinatubo ClO observations
suggesting heterogeneous perturbations in mid-latitude air. HCl
observations by Webster et al. (1998) provide evidence that not only
is ClO enhanced by high volcanic loading at mid-latitudes as an indirect
effect through shifts in NO2 , but also is directly affected to some degree
by chlorine activation (as reflected in reduced HCl). Debate on the

magnitude of the latter effect in some regions has focussed on both the
rates of chlorine activation in liquid aerosols for mid-latitude conditions
and on mass-balance among Cly species (see Dessler et al., 1996; 1997;
1998; Stimpfle et al., 1994; Smith et al., 2001). While chlorine-activating
reactions on liquid sulfate aerosols are thought to be relatively slow
at 20 km for the average temperatures that prevail at mid-latitudes,
the strongly non-linear dependence of these reactions on temperatures
must be considered. Murphy and Ravishankara (1994) showed that
the reaction rate averaged over the actual temperatures including cold
fluctuations associated with wave motions will substantially exceed the
rate computed for the average temperature. In other words, brief
exposure to cold temperatures may alter ClO/Cly partitioning and
hence enhance ozone depletion at latitudes outside of the polar vortex,
especially under high aerosol loads (Webster et al., 1998; Solomon et al.,
1998).
Figure 6.21 shows that even rather modest changes in aerosol
abundances can substantially affect the ClO/Cly partitioning near
20 km. Indeed, Figure 6.21 suggests that aerosol surface area increases
of a factor of only about 5 (as observed, for example, in northern mid-
latitudes following the relatively minor Mount St. Helens eruption;
see Thomason et al., 1997) could increase ClO/Cly by 50%, thus
greatly enhancing the chlorine-driven local ozone destruction reactions.
In addition to local depletion, enhanced ozone destruction at higher
latitudes may also mix out to mid-latitudes.
As has been emphasized throughout this chapter, processes that
enhance ClO partitioning relative to Cly are at the heart of ozone
depletion. A point of useful comparison may be drawn by noting
that if ClO/Cly had been constant from 1980 to 1990, then ClO
would increase by about 50% over this decade (due to the roughly
50% increase in Cly due to chlorofluorocarbon emissions). However,
Figure 6.21 demonstrates that far larger changes in ClO can be rapidly
induced by volcanic aerosol increases via their effects on chemical
partitioning. Solomon et al. (1996; 1998) showed that both the
long-term ozone trend at northern mid-latitudes and its year-to-year
variations in the 1980s and 1990s are highly likely to be closely tied to
volcanic aerosol-driven changes in ClO/Cly partitioning (see Figure 6.23
below). Jackman et al. (1996) and Zerefos et al. (1997) reached similar
conclusions with their models, and showed that solar cycle contributions
to interannual ozone depletion are much smaller. Several authors have
shown that the ozone response to volcanic aerosols before humans
perturbed stratospheric chlorine (e.g., in the 19th century after the
eruption of Krakatoa) would likely be a slight column increase as a result
Figure 6.21. Observations of the changes in chemical partitioning as a function

of aerosol load after the eruption of Mt. Pinatubo at 20 km at mid-latitudes
((Fahey et al., 1993).
) The primary chemistry responsible for the observed
behavior is depicted by the schematic diagram. From Solomon (1999).
of suppression of NOx -catalyzed destruction as depicted in Figure 6.21

rather than the observed decreases obtained for current chlorine loads
(Solomon et al., 1996; Tie and Brasseur, 1995). It is also useful to
note that observations such as those at Arosa in Figure 6.20 show no
noticeable depletion after a series of large eruptions in the 1960s, most
notably the major tropical eruption of Agung in 1963. Observations from
many other ground-based sites confirm that the enhancements in aerosol
of the 1960s had little effect on ozone (e.g., Bojkov et al., 1995). It is
only in the period since about 1980 that chlorine levels have become
sufficiently elevated that volcanic perturbations to ClO/Cly such as
shown in Figure 6.21 result in signficant ozone loss. Hence, the evidence
suggests that volcanic particles at mid-latitudes exacerbate halogen-
induced ozone depletion in the contemporary stratosphere (much as
PSCs do for polar regions, and with some similar chemistry), but cannot
on their own significantly destroy stratospheric ozone.
Lary et al. (1997) were the first to suggest that soot may also
affect northern hemispheric mid-latitude ozone, mainly through possible
reactions involving reactive nitrogen species (see Rogaski et al., 1997).
A study by Bekki (1997) further probed this chemistry in some detail,

and argued for significant impacts on ozone trends in the vicinity of
the tropopause. There is currently debate about the surface area of
soot available at stratospheric altitudes, the extent to which it can
remain active for chemistry or be quickly ‘poisoned’, and whether or
not chemical data support such perturbations (Gao et al., 1998).
The observation of large mid-latitude ozone depletion following
Pinatubo and El Chichón, ´ substantial related changes in chemical
species, and a wide range of modelling studies (e.g., Hofmann and
Solomon, 1989; Brasseur and Granier, 1992; Michelangeli et al., 1989;
Pitari et al., 1993; Bekki and Pyle, 1994; Tie et al., 1997; Solomon et al.,
1996; Jackman et al., 1996) provide strong evidence that heterogeneous
sulfate aerosol chemistry plays a major role together with manmade
chlorine in the processes controlling mid-latitude ozone trends. The
observed ClO/Cly and NOx /NOy dependencies upon volcanic aerosol
amounts as shown, for example, in Figure 6.21 may be considered a
chemical fingerprint underlying these effects, like the observations of
greatly enhanced ClO in the ozone hole region. Another parallel with
Antarctic ozone depletion is the observation of a close correspondence in
altitude between the region of enhanced Pinatubo aerosol abundances
and ozone depletion (as in McGee et al., 1994 and Hofmann et al., 1994,
for example). A third fingerprint is the onset and slow relaxation of
the ozone depletion after Pinatubo observed at mid-latitudes (see e.g.,
Solomon et al., 1996; 1998; Jackman et al., 1996) over a period of a few
years, mirroring in a slower manner the seasonal depletion of the ozone
hole.
6.3.3.2 Dynamical Processes and Mid-latitude Ozone Trends

Several studies examined the extent to which dynamical processes
might spread the influence of the ozone hole, either through a one-time
‘dilution’ at the end of the winter when stratospheric warmings break
up the polar vortex or through vortex ‘processing’ wherein flow of air
through the vortex (and hence chemical activation of chlorine) might
be transported to lower latitudes (e.g., Tuck, 1989; Tuck et al., 1992;
Waugh et al., 1994; 1997; Wauben et al., 1997; Tuck and Proffitt, 1997).
The amount of ozone depletion observed at both northern and
southern mid-latitudes is considerably greater than that implied by a
one-time end-of-winter dilution process (see e.g., Sze et al., 1989; Prather
et al., 1990; Pitari et al., 1992). For the southern hemisphere such one-
time dilution likely provides a contribution to the mid-latitude column
ozone depletion of about 1-2%. Locally larger but transient dilution
effects following the breakup of the Antarctic ozone hole in late spring
have been documented in observations over New Zealand, Australia,

Brazil, and Chile (Atkinson et al., 1989; Lehmann et al., 1992; Kirchhoff
et al., 1996; 1997a). The city of Punta Arenas, Chile at 53◦ S occasionally
lies just beneath the tip of the Antarctic ozone hole itself for brief periods
in October when wave disturbances push the vortex overhead (Kirchhoff
et al., 1997a,b). Due to greater dynamical activity, the northern vortex
is likely to be subject to a greater degree of processing, and many studies
conclude that there is ample evidence for the spread of polar ‘filaments’
to mid-latitudes at times (Tuck et al., 1992; Gerber and Kampfer, 1994;
Pyle et al., 1995; Lutman et al., 1997). However, dynamical analyses
and tracer studies suggest that the transport from polar regions alone
cannot fully account for the observed ozone losses in mid-latitudes (e.g.,
Schoeberl et al., 1992; Waugh et al., 1994; Manney et al., 1994; Jones
and MacKenzie, 1995; Chipperfield et al., 1996; Wauben et al., 1997;
Grewe et al., 1998).
In addition to vortex processing as described above, the notion of
PSC processing has also been suggested (wherein PSCs forming outside
the vortex provide the sites for heterogeneous reactions), particularly
in association with locally cold temperatures that may be related to
mountain lee waves and hence of quite small spatial and temporal scale
(e.g., Godin et al., 1994; Carslaw et al., 1998). Although the mountain
wave clouds themselves are small, air flows through these quasi-
stationary clouds, allowing for processing of large quantities of chlorine.
All processing mechanisms depend upon heterogeneous chlorine-related
chemistry in some fashion and hence connect mid-latitude ozone
depletion to chlorine trends, but with important differences in the degree
of non-local (i.e., transport-related) linkages. Thus such a mechanism
remains tied to chemistry and does not constitute evidence for a purely
dynamically-driven effect. Chipperfield (1999) suggests that transport
of ozone-depleted air from the Arctic may play a key role in Northern
Hemisphere mid-latitude ozone loss. As in the Antarctic, dynamical
conditions that modify heterogeneous chemistry should be interpreted
as key aspects of setting the stage for ozone loss. However, they can
only contribute independently to ozone trends if there is evidence for a
trend in those dynamical processes.
Some authors have argued that changes in stratospheric dynamics
could have contributed to the observed mid-latitude ozone trends. A
review is provided in Ravishankara et al. (1999; see references therein).
In brief, some studies (e.g., Hood and Zaff, 1995; McCormack and Hood,
1997; Hood et al., 1997; Fusco and Salby, 1999) have argued for a
component of purely dynamical change in mid-latitude ozone relating,
for example, to changes in the meridional transport of ozone. It is
well known that dynamical processes strongly influence ozone variability
from year-to-year, particularly in winter (see for example, Fusco and

Salby, 1999). However, evaluation of trends requires long records and
analysis of low-frequency changes (i.e., time scales of the order of a
decade) rather than a focus on higher frequency (e.g., year-to-year)
variations. Some authors suggest that planetary wave activity may
indeed have changed in the Northern Hemisphere in particular (e.g.,
Fusco and Salby, 1999), perhaps in association with the changes in the
Arctic Oscillation (AO) noted earlier. Thompson et al. (2000) suggest
that the systematic changes in the AO could have contributed about
30% to the observed northern mid-latitude ozone depletion. However,
the cause of the changes observed in the AO are not well established.
They may reflect natural climatic variability, or greenhouse gas coupling
to dynamics, or they may be driven at least in part by the ozone
change itself through linkage with stratospheric temperature, winds,
and/or wave propagation. In the latter case, the dynamical perturbation
should be considered a feedback and not a cause. Indeed, changes in
ozone can play a major role in temperature gradients, zonal wind and,
hence, in stratospheric dynamics, and the feedbacks of ozone loss to
dynamics are difficult to distinguish from changes due to other causes
(see WMO/UNEP, 2003).
Observations show that about 25% of the observed mid-latitude ozone
column depletion occurs above about 25 km, in the altitude range where
gas-phase photochemistry is rapid and it is difficult for dynamics to
compete (see e.g., SPARC, 1998). A further contribution due to PSC
processing and to vortex breakdown is highly likely. Thus, a substantial
chemical contribution of at least half of the trend in the column seems
difficult to dispute even if locally-driven chemical ozone depletion in the
mid-latitude lower stratosphere were to be substantially smaller than
suggested by modelling studies and by the post-Pinatubo measurements
described above. Thus, the evidence suggests that chlorine chemistry
has played an important and very likely dominant role in the observed
trends in mid-latitude ozone over the past two decades (for a recent
review, see WMO/UNEP, 2003).
6.4 Summary and Outlook
This chapter has outlined the history and conceptual understanding

of the processes responsible for ozone depletion by chlorofluorocarbons
in the stratosphere. In brief, the long lifetimes of chlorofluorocarbons are
reflected in their observed worldwide accumulation in the atmosphere.
Their role in stratospheric ozone depletion depends critically on
partitioning processes that follow release of halogen atoms; indeed,

the marked contrasts between fluorine (which does not deplete
stratospheric ozone), chlorine, and bromine illustrate the central role
of partitioning chemistry. Table 6.1 summarizes a series of spatial and
temporal fingerprints that connect chlorine chemistry to ozone depletion.
Observational evidence for gas-phase chlorine chemistry impacts on
ozone is provided, for example, by observations of the ClO/Cly and
ozone trend profiles above about 25 km at mid latitudes and by the
similarities in their observed latitudinal distributions.
The cold conditions of the Antarctic winter and spring stratosphere
lead to formation of polar stratospheric clouds. Heterogeneous
chemistry involving manmade chlorine takes place on these surfaces and
results in the dramatic and unanticipated Antarctic ozone hole. The
heterogeneous activation of chlorine from both its HCl and ClONO2
reservoirs and the suppression of the NO2 (that would otherwise
reform ClONO2 ) alters chlorine partitioning and allows effective ozone
loss in cold sunlit air. The close correspondence between observed
enhancements in ClO and depleted Antarctic ozone through independent
observational methods as functions of altitude, latitude and longitude
illustrates the key role of these chemical partitioning processes in
producing the ozone hole. A broad range of chemical observations of
HCl, HNO3 , NO2 , OClO, and other species support and extend this
picture. Quantitative numerical modelling studies that include detailed
analyses of transport and chemistry further connect the enhanced ClO
produced by heterogeneous chemistry to the formation of the Antarctic
ozone hole.
Scientific understanding of PSCs and heterogeneous chemistry has
evolved considerably in recent years. The detailed microphysical mecha-
nisms responsible for freezing of PSC particles and for denitrification are
subject to debate at present, but these processes appear to be less critical
to ozone depletion than once thought. There is evidence from field,
laboratory, and modelling studies that PSCs can be composed not only
of solid water ice and nitric acid hydrates but also of liquid solutions of
water, sulfuric acid, and nitric acid. The chemistry associated with these
varying surfaces displays important differences in detail, but has the
common feature that all can suppress NO2 and activate chlorine from the
reservoir species, making the ozone depletion process more continuous in
temperature and less dependent upon the abrupt temperature thresholds
that are associated with formation of solids than previously thought.
Observations of enhanced Antarctic and mid-latitude ozone depletion
following the eruption of Pinatubo confirm the impact of liquid aerosol
surfaces on chlorine and nitrogen partitioning chemistry. Observations
and laboratory studies have demonstrated the efficacy of heterogeneous
processes on such surfaces. As in the Antarctic, concurrent observations

of a broad range of chemical species show evidence for surface reactions
associated with particles, which work to enhance ClO/Cly partitioning
at mid-latitudes. Dilution and processing of the polar ozone losses also
contribute to mid-latitude ozone depletion. While some studies suggest
a role for a purely dynamical trend in mid-latitude ozone depletion,
these have not yet succeeded in quantifying a significant independent
contribution.
There is abundant evidence for heterogeneous perturbations to Arctic
chemistry through observations of ClO, OClO, HCl, and many other
key gases. Arctic ozone reached record low values in many years in
the 1990s, linked not only with heterogeneous chemistry on Pinatubo
aerosols but also with unusually cold spring temperatures. A chemical
contribution to these low values has been documented with a variety of
methods including trajectory ‘matches’, chemistry transport modelling,
and tracer correlation studies. The fundamental question of the cause or
causes of record-low temperatures in many of the Arctic winter/spring
seasons of the 1990s remains a topic of debate and a key issue.
In closing, Figures 6.22 and 6.23 are presented to show the impact
of changes in global emissions of chlorofluorocarbons and the likely
future of the ozone layer based on the conceptual picture developed
in this book. Figure 6.22 displays surface observations of CFC-11 and
methyl chloroform. The latter gas is the only short-lived industrial
chlorofluorocarbon produced in large amounts in the 1970s and 1980s.
Because of its five year lifetime, the abundances of methyl chloroform
have already begun to decline, due to reduced global emissons. Those of
CFC-11 are just passing their peak and are projected to decline slowly
in coming decades, reflecting its 50-year lifetime.
Figure 6.23 (top) shows the past and future projections of the total
tropospheric chlorine content (which leads the stratosphere by 3-5 years).
It was anticipated that the combined effect of all CFCs would lead
to a peak stratospheric chlorine loading in the late 1990s. Indeed,
observations of stratospheric HCl displayed a pronounced slowing in the
late 1990s, and suggest that the stratospheric Cly then began to level
off (Anderson et al., 2000).
By about 2050, the chlorine will return to levels close to those of
the mid- to late 1970s when ozone depletion was first apparent. All
other things being equal, the Antarctic ozone hole and mid-latitude
ozone depletion will likely disappear around this time. However, the key
role of temperature and aerosols in modulating ozone depletion must
also be considered. The unusually cold Arctic winter/springs of recent
years stand at the time of this writing as a critical challenge to our
understanding that could affect the future of polar ozone depletion in
Figure 6.22. Observations of CH3 CCl3 and CFC-11 from the Atmospheric
Lifetime Experiment/ / Global Atmospheric Gases Experiment ((ALE/GAGE)/ )
and Climate Monitoring and Diagnostics Laboratory (CMDL) databases,
respectively. The projections from the baseline emission scenario of
WMO/UNEP
/ ((1998)) are shown for comparison. The scenario includes
estimated industrial production and emission for each year ((including the effects
of delayed release in some applications such as refrigeration, see WMO/UNEP,
/
1998).
) The methyl chloroform data show a rapid decline observed in recent
years due to reduced emissions and the 5-year lifetime of this gas ((Prinn et al.,
1995; WMO/UNEP,
/ 1998), while the CFC-11 abundances have just passed their
peak ((Elkins et al., 1993; Montzka et al., 1996; updated courtesy of J. Elkins
and S. Montzka) and are projected to decline slowly in the future due to the
50-year lifetime of this gas. From Solomon (1999).
Figure 6.23. The top panel shows the total tropospheric chlorine content
estimated from the baseline scenario in WMO/UNEP (1998); this is based
on a gas-by-gas analysis like those shown in Figure 6.22. The bottom panel
shows the changes in the 5-year running mean ozone observed over Switzerland
(Staehelin et al., 1998a,b) compared to a model calculation for 45◦ N applying
the same time averaging, with and without considering the effects of volcanic
enhancements in aerosol chemistry (from the model of Solomon et al., 1996;
1998). The major eruptions since 1980 were those of El Chichón´ in 1982 and
Mt. Pinatubo in 1991. Updated from Solomon (1999).
both hemispheres. For example, if the majority of future Arctic winters

were to be colder than average, then the Arctic ozone depletion would
likely be prolonged. The bottom panel in Figure 6.23 shows the long-
running Arosa, Switzerland, ozone record illustrating the onset of mid-
latitude ozone depletion, its likely links to heterogeneous chemistry, and
its simulation with a current stratospheric chemistry model including
the processes described here. The changes in ozone observed over Arosa
are in good agreement with the zonally averaged global satellite data
discussed earlier, and the time-averaged trends obtained there appear
to be generally representative of northern mid-latitudes. The figure
illustrates that the future of mid-latitude ozone depletion is likely to
be linked not only to chlorine but also in part to volcanoes for at
least several decades. If there were to be an extremely large volcanic
eruption such as that of Tamboura (whose 1815 eruption is estimated
at about three times the stratospheric impact of Mt. Pinatubo) in
coming decades, it is likely that mid-latitude ozone depletion would
be increased even though the chlorine content of the stratosphere is
expected to be lower than it is today. This illustrates the connection
between the accumulation of chlorine in today’s atmosphere due to
human activities of the industrial era and the unpredictable time scales
of geologic phenomena that couple into this altered chemical state.
This book has emphasized many spatial and temporal ‘fingerprints’
that illustrate the role of chlorine in depleting ozone in the contemporary
stratosphere (Table 6.1). Indeed, it is the structure of the ozone loss
in space (e.g., in the 40 km region) and time (e.g., in the Antarctic
spring and in mid-latitudes in the years following Mt. Pinatubo) that
tests and supports scientific understanding, illustrating how gas-phase
and heterogeneous modulation of ClO/Cly partitioning affects ozone
depletion. Through the impacts of this chemistry, the stratospheric
ozone layer in the 21st century will continue to reflect the impact of
the changes in chlorine enacted in the 20th .
Table 6.1 Summary of Key Fingerprints of Ozone Depletion
Observation Method Latitude (◦ ) Altitude (km) Primary Chemistry

Linkage
Profile shapes of upper Satellite and ground-based Northern Hemisphere 30-50 Gas-phase chlorine
stratospheric ozone depletion mid-latitude chemistry, particularly
and ClO/Cly partitioning processes
Latitudinal structure of upper Satellite Polar, mid-latitude, 30-50 Gas-phase chlorine
stratospheric ozone depletion and tropical chemistry, particularly
and ClO HCl/ClO
OZONE PERTURBATIONS
Latitudinal structure of ClO, Airborne and satellite ≈50-85◦ S; 50-85◦ N 20 Heterogeneous chlorine
HCl, NO, NO2 , and ozone loss activation and NO2
in polar regions suppression
Vertical structure of seasonal Balloon-borne, lidar, and 90◦ S, 50-85◦ S, 12-24 km Heterogeneous chlorine
ozone loss and PSCs in polar satellite 50-85◦ N activation
regions
Seasonal changes in PSCs, Balloon-borne, lidar, ≈50-90◦ in both ≈12-24 km and Heterogeneous
ozone depletion, OClO, ClO, ground-based, and satellite polar regions column chemistry
HCl, and ClONO2
Post-Pinatubo (≈1992-1995) Ground-based and some Mid-latitudes and near 20 km and Heterogeneous chemistry
ozone depletion and recovery; satellite polar regions column
contrast with post-Agung
(≈1964-1968)
Post-Pinatubo changes in Airborne, ground-based, Mid-latitudes and near 20 km, near Heterogeneous
stratospheric chemical state balloon-borne, and satellite some polar tropopause, and chemistry, particularly
and aerosol content (NOx , column N2 O5 hydrolysis and
ClO, HNO3 , OClO) some chlorine activation
507
References
Adrian, G.P., M. Baumann, T. Blumenstock, H. Fischer, A. Friedle, L. Gerhardt, G.

Maucher, H. Oelhof, W. Scheuerpflug, P. Thomas, O. Trieschmann, and A. Wegner,
First results of ground-based FTIR measurements of atmospheric trace gases in north
Sweden and Greenland during EASOE. Geophys Res Lett: 21, 1343, 1994.
Adriani, A., G. Fiocco, G.P. Gobbi, and F. Congeduti, Correlated behavior of aerosol
and ozone contents of the stratosphere after the El Chichón
´ eruption. J Geophys Res:
92, 8365, 1987.
Adriani, A., T. Deshler, G. Di Donfrancesco, and G.P. Gobbi, Polar stratospheric
clouds and volcanic aerosol during spring 1992 over McMurdo Station, Antarctica:
Lidar and particle counter comparisons. J Geophys Res: 100, 25,877, 1995.
Allen, D.R., R.M. Bevilacqua, G.E. Nedoluha, C.E. Randall, and G.L. Manney,
Unusual stratospheric transport and mixing during the 2002 Antarctic winter,
Anderson, J.G., J.J. Margitan, and D.H. Stedman, Atomic chlorine and the chlorine
monoxide radical in the stratosphere: Three in-situ observations. Science: 198, 501,
1977.
Anderson, J.G., W.H. Brune, and M.H. Proffitt, Ozone destruction by chlorine
radicals within the Antarctic vortex: The spatial and temporal evolution of ClO-
O3 anticorrelation based on in situ ER-2 Data. J Geophys Res: 94, 11,479, 1989.
Anderson, J., J.M. Russell III, S. Solomon, and L.E. Deaver, Halogen occultation
experiment confirmation of stratospheric chlorine decreases in accordance with the
Montreal Protocol. J Geophys Res: 105, 4483, 2000.
Academic Press, Orlando, Fla, USA, 1987.
Atkinson, R.J., W.A. Matthews, P.A. Newman, and R.A. Plumb, Evidence of mid-
latitude impact of Antarctic ozone depletion. Nature: 340, 290, 1989.
Austin, J., R.R. Garcia, J.M. Russell III, S. Solomon, and A.F. Tuck, On the
atmospheric photochemistry of nitric acid. J Geophys Res: 91, 5477, 1986.
Austin, J., N. Butchart, and K.P. Shine, Possibility of an Arctic ozone hole in a
doubled-CO2 climate. Nature: 360, 221, 1992.
Avallone, L.M., D.W. Toohey, W.H. Brune, R.J. Salawitch, A.E. Dessler, and J.G.
Anderson, Balloon-borne in situ measurements of ClO and ozone: Implications for
heterogeneous chemistry and mid-latitude ozone loss. Geophys Res Lett: 20, 1795,
1993a.
Avallone, L.M., D.W. Toohey, M.H. Proffitt, J.J. Margitan, K.R. Chan, and J.G.
Anderson, In situ measurements of ClO at mid-latitudes: Is there an effect from Mt.
Pinatubo? Geophys Res Lett: 20, 2519, 1993b.
Baldwin, M.P., and T.J. Dunkerton, Stratospheric harbingers of anomalous weather
regimes. Science: 29, 581, 2001.
Barnett, J.J., J.T. Houghton, and J.A. Pyle, Temperature dependence of the ozone
concentration near the stratopause. Quart J Roy Meteorol Soc: 101, 245, 1975.
Becker, G., R. Müller,
¨ D.S. McKenna, M. Rex, and K.S. Carslaw, Ozone loss rates
in the Arctic stratosphere in the winter 1991/92: Model calculations compared with
match results. Geophys Res Lett: 23, 4325, 1998.
Bekki, S., On the possible role of aircraft-generated soot in the middle latitude ozone
depletion. J Geophys Res: 102, 10,751, 1997.
Bekki, S., and J.A. Pyle, A two-dimensional modeling study of the volcanic eruption
of Mount Pinatubo. J Geophys Res: 99, 18,861, 1994.
Bell, W., N.A. Martin, T.D. Gardiner, N.R. Swann, P.T. Woods, P.F. Fogal, and
J.W. Waters, Column measurements of stratospheric trace species over Åre, Sweden
in the winter of 1991-1992. Geophys Res Lett: 21, 1347, 1994.
Bevilacqua, R.M., C.P. Aellig, D.J. Debrestian, M.D. Fromm, K. Hoppel, J.D. Lumpe,
E.P. Shettle, J.S. Hornstein, C.E. Randall, D.W. Rusch, and J.E. Rosenfield, POAM
II ozone observations in the Anarctic ozone hole in 1994, 1995, and 1996. J Geophys
Res: 102, 23,643, 1997.
Bojkov, R.D., and V.E. Fioletov, Changes of the lower stratospheric ozone over
Europe and Canada. J Geophys Res: 102, 1337, 1997.
Bojkov, R.D., C.S. Zerefos, D.S. Balis, I.C. Ziomas, and Z.F. Bais, Record low total
ozone during northern winters of 1992-1993. Geophys Res Lett: 20, 1351, 1993.
Bojkov, R.D., L. Bishop, and V.E. Fioletov, Total ozone trends from quality controlled
ground-based data (1964-1994). J Geophys Res: 100, 25,867, 1995.
Borrmann, S., S. Solomon, J.E. Dye, and B. Luo, The potential of cirrus clouds for
heterogenous chlorine activation. Geophys Res Lett: 23, 2133, 1996.
Borrmann, S., S. Solomon, J.E. Dye, D. Baumgardner, K.K. Kelly, and K.R. Chan,
Heterogeneous reactions on stratospheric background aerosols, volcanic sulfuric acid
droplets, and Type I polar stratospheric clouds: Effects of temperature fluctuations
and differences in particle phase. J Geophys Res: 102, 3639, 1997a.
Borrmann, S., S. Solomon, L. Avallone, D. Toohey, and D. Baumgardner, On the
occurrence of ClO in cirrus clouds and volcanic aerosol in the tropopause region.
Geophys Res Lett: 24, 2011, 1997b.
Brasseur, G., The response of the middle atmosphere to long-term and short-term
solar variability: A two-dimensional model. J Geophys Res: 98, 23,079, 1993.
Brasseur, G., and S. Solomon, Aeronomy of the Middle Atmosphere. D. Reidel,
Hingham, Mass., 1984.
Brasseur, G., and M.H. Hitchman, Stratospheric response to trace gas perturbations:
Changes in ozone and temperature distributions. Science: 240, 634, 1988.
Brasseur, G., and C. Granier, Mount Pinatubo aerosols, chlorofluorocarbons, and
ozone depletion. Science: 257,
7 1239, 1992.
Brasseur, G., P. DeBaets, and A. de Rudder, Solar variability and minor constituents
in the lower thermosphere and in the mesosphere. Space Sci Rev: 34, 377, 1983.
Brasseur, G.P., X.X. Tie, P.J. Rasch, and F. Lefevre, A three dimensional simulation
of the Antarctic ozone hole: Impact of anthropogenic chlorine on the lower
stratosphere and upper troposphere. J Geophys Res: 102, 8909, 1997.
Bregman, A., M. van den Broek, K.S. Carslaw, R. Müller,
¨ T. Peter, M.P. Scheele, and
J. Leieveld, Ozone depletion in the late winter lower Arctic stratosphere: Observations
and model results. J Geophys Res: 102, 10,815, 1997.
Brewer, A.W., Evidence for a world circulation provided by the measurements of
helium and water vapour distribution in the stratosphere. Quart J Roy Meteorol Soc:
75, 351, 1949.
Brune, W.H., J.G. Anderson, and K.R. Chan, In situ observations of BrO over
Antarctica: ER-2 Aircraft results from 54◦ S to 72◦ S Latitude. J Geophys Res: 94,
16,639, 1989.
Brune, W.H., D.W. Toohey, J.G. Anderson, and K.R. Chan, In situ observations of
ClO in the Arctic stratosphere: ER-2 aircraft results from 59◦ N to 80◦ N latitude.
Geophys Res Lett: 17,
7 505, 1990.
Brune, W.H., J.G. Anderson, D.W. Toohey, D.W. Fahey, S.R. Kawa, R.L. Jones,
D.S. McKenna, and L.R. Poole, The potential for ozone depletion in the arctic polar
stratosphere. Science: 252, 1260, 1991.
Cadle, R.D., P. Crutzen, and D. Ehhalt, Heterogeneous chemical reactions in the
stratosphere. J Geophys Res: 80, 3381, 1975.
Callis, L.B., and M. Natarajan, The Antarctic ozone minimum: Relationship to odd
nitrogen, odd chlorine, the final warming, and the 11-year solar cycle. J Geophys Res:
91, 10,771, 1986.
Carslaw, K.S., and T. Peter, Uncertainties in reactive uptake coefficients for solid
stratospheric particles-1. Surface chemistry. Geophys Res Lett: 24, 1743, 1997.
Carslaw, K.S., T. Peter, and S.L. Clegg, Modeling the composition of liquid
stratospheric aerosols. Rev Geophys: 35, 125, 1997a.
Carslaw, K.S., T. Peter, and R. Muller,
¨ Uncertainties in reactive uptake coefficients
for solid stratospheric particles-2. Effect on ozone depletion. Geophys Res Lett: 24,
1747, 1997b.
Volkert, W. Renger, J.T. Bacmeister, E. Reimer, and T. Peter, Increased stratospheric
ozone depletion due to mountain-induced atmospheric waves. Nature: 391, 675, 1998.
Chandra, S., C.H. Jackman, A.R. Douglass, E.L. Fleming, and D.B. Considine,
Chlorine catalyzed destruction of ozone: Implications for ozone variability in the
upper stratosphere. Geophys Res Lett: 20, 351, 1993.
Chen, L., J. London, and G. Brasseur, Middle atmospheric ozone and temperature
responses to solar irradiance variations over 27-day periods. J Geophys Res: 102,
29,957, 1997.
Chin, M., and D.D. Davis, A reanalysis of carbonyl sulfide as a source of stratospheric
background sulfur aerosols. J Geophys Res: 100, 8993, 1995.
Chipperfield, M.P., Multi-annual simulations with a three-dimensional chemical
transport model. J Geophys Res: 104, 1781, 1999.
Chipperfield, M.P., and J.A. Pyle, Model sensitivity studies of Arctic ozone depletion.
J Geophys Res: 103, 28,389, 1998.
Chipperfield, M.P., D. Cariolle, and P. Simon, A 3D transport model study of chlorine
activation during EASOE. Geophys Res Lett: 21, 1467, 1994.
Chipperfield, M.P., A.M. Lee, and J.A. Pyle, Model calculations of ozone depletion
in the Arctic polar vortex for 1991/92 to 1994/95. Geophys Res Lett: 23, 559, 1996.
Chubachi, S., Preliminary result of ozone observations at Syowa Station from
February, 1982, to January, 1983. Mem Natl Inst Polar Res Jap, Spec Issue 34,
13, 1984.
Coffey, M.T., and W.G. Mankin, Observations of the loss of stratospheric NO2
following volcanic eruptions. Geophys Res Lett: 20, 2873, 1993.
Coffey, M.T., W.G. Mankin, and A. Goldman, Airborne measurements of
stratospheric constituents over Antarctica in the austral spring, 1987: 2, Halogen
and nitrogen trace gases. J Geophys Res: 94, 16,597, 1989.
Cohen, R.C., P.O. Wennberg, R.M. Stimpfle, J. Koplow, J.G. Anderson, D.W. Fahey,
E.L. Woodbridge, E.R. Keim, R. Gao, M.H. Proffitt, M. Loewenstein, and K.R. Chan,
Are models of catalytic removal of O3 by HOx accurate? Constraints from in situ
measurements of the OH to HO2 ratio. Geophys Res Lett: 21, 2539, 1994.
Collins, R.L., K.P. Bownan, and C.S. Gardner, Polar stratospheric clouds at the
South Pole in 1990: Lidar observations and analysis. J Geophys Res: 98, 1001, 1993.
Conway, R.R., M.E. Summers, M.H. Stevens, J.G. Cardon, P. Preusse, and D.
Offermann, Satellite observations of upper stratospheric and mesospheric OH: the
HOx dilemma. Geophys Res Lett: 27,7 2613, 2000.
Cox, R.A., A.R. MacKenzie, R.H. Müller,
¨ T. Peter, and P.J. Crutzen, Activation of
stratospheric chlorine by reactions in liquid sulphuric acid. Geophys Res Lett: 21,
1439, 1994.
Coy, L., E.R. Nash, and P.A. Newman, Meteorology of the polar vortex: Spring 1997.
¨
Crewell, S., K. Kunzi, H. Nett, T. Wehr, and P. Hartogh, Aircraft measurements of
ClO and HCl during EASOE 1991/92. Geophys Res Lett: 21, 1267, 1994.
Crutzen, P.J., The influence of nitrogen oxide on the atmospheric ozone content.
Quart J Roy Meteorol Soc: 96, 320, 1970.
Crutzen, P.J., Ozone production rates in an oxygen-hydrogen-nitrogen oxide
Crutzen, P.J., Estimates of possible future ozone reductions from continued use of
fluoro-chloro-methanes (CF2 Cl2 , CFCl3 ). Geophys Res Lett: 1, 205, 1974.
Crutzen, P.J., The possible importance of OCS for the sulfate layer of the atmosphere.
Crutzen, P.J., and C.T. Howard, Effect of the HO2 + NO reaction rate constant on
one-dimensional model calculations of stratospheric ozone perturbations. Pure Appl
Geophys: 116, 497, 1978.
Crutzen, P.J., and F. Arnold, Nitric acid cloud formation in the cold Antarctic
stratosphere: A major cause for the springtime ‘ozone hole’. Nature: 324, 651, 1986.
Crutzen, P.J., I.S.A. Isaksen, and G.C. Reid, Solar proton events: Stratospheric
Crutzen, P.J., I.S.A. Isaksen, and J.R. McAfee, Impact of chlorocarbon industry on
the ozone layer. J Geophys Res: 83, 345, 1978.
Dalgarno, A., Atmospheric reactions with energetic particles. Space Res: 7,
7 849,
1971.
Dameris, M., V. Grewe, R. Hein, C. Schnadt, C. Brühl,
¨ and B. Steil, Assessment of
the future development of the ozone layer. Geophys Res Lett: 25, 3579, 1998.
Del Negro, L.A., D.W. Fahey, S.G. Donnelly, R.S. Gao, E.R. Keim, R.C. Wamsley,
E.L. Woodbridge, J.E. Dye, D. Baumgardner, B.W. Gandrud, J.C. Wilson, H.H.
Jonsson, M. Loewenstein, J.R. Podolske, C.R. Webster, R.D. May, D.R. Worsnop, A.
Tabazadeh, M.A. Tolbert, K.K. Kelly, and K.R. Chan, Evaluating the role of NAT,
NAD, and liquid H2 SO4 /H2 O/HNO3 solutions in Antarctic polar stratospheric cloud
aerosol: Observations and implications. J Geophys Res: 102, 13,255, 1997.
de Zafra, R.L., M. Jaramillo, A. Parrish, P. Solomon, B. Connor, and J. Barnett,
High concentrations of chlorine monoxide at low altitudes in the Antarctic spring
stratosphere: Diurnal variation. Nature: 328, 408, 1987.
de Zafra, R.L., M. Jaramillo, J. Barrett, L.K. Emmons, P.M. Solomon, and A. Parrish,
New observations of a large concentration of ClO in the springtime lower stratosphere
over Antarctica and its implications for ozone-depleting chemistry. J Geophys Res:
94, 11,423, 1989.
de Zafra, R.L., L.K. Emmons, J.M. Reeves, and D.T. Shindell, An overview
of millimeter-wave spectroscopic measurements of chlorine monoxide at Thule,
Greenland, February-March, 1992: Vertical profiles, diurnal variation, and longer-
term trends. Geophys Res Lett: 21, 1271, 1994.
Deniel, C., R.M. Bevilacqua, J.P. Pommereau, and F. Lefevre, Arctic chemical ozone
depletion during the 1994-1995 winter deduced from POAM II satellite observations
and the REPROBUS three-dimensional model. J Geophys Res: 103, 19,231, 1998.
Deshler, T., A. Adriani, G.P. Gobbi, D.J. Hofmann, G. Di Donfrancesco, and B.J.
Johnson, Volcanic aerosol and ozone depletion within the Antarctic polar vortex
during the austral spring of 1991. Geophys Res Lett: 19, 1819, 1992.
Dessler, A.E., D.B. Considine, G.A. Morris, M.R. Schoeberl, J.M. Russell III,
A.E. Roche, J.B. Kumer, J.L. Mergenthaler, J.W. Waters, J.C. Gille, and G.K.
Yue, Correlated observations of HCl and ClONO2 from UARS and implications for
stratospheric chlorine partitioning. Geophys Res Lett: 22, 1721, 1995.
Dessler, A.E., S.R. Kawa, A.R. Douglass, D.B. Considine, J.B. Kumer, A.E. Roche,
J.L. Mergenthaler, J.W. Waters, J.M. Russell III, and J.C. Gille, A test of the
partitioning between ClO and ClONO2 using simultaneous UARS measurements of
ClO, NO2 and ClONO2 . J Geophys Res: 101, 12,515, 1996.
Dessler, A.E., D.B. Considine, J.E. Rosenfield, S.R. Kawa, A.R. Douglass, and
J.M. Russell, III, Lower stratospheric chlorine partitioning during the decay of the
Mt. Pinatubo aerosol cloud. Geophys Res Lett: 24, 1623, 1997.
Dessler, A.E., M.D. Burrage, J.-U. Grooß, J.R. Holton, J.L. Lean, S.T. Massie, M.R.
Schoeberl, A.R. Bouglass, and C.H. Jackman, Selected science highlights from the
first 5 years of the Upper Atmosphere Research Satellite (UARS) Program. Rev
Geophys: 36, 183, 1998.
Dickinson, R.E., Method of parameterization for infrared cooling between the
altitudes of 30 and 70 km. J Geophys Res: 78, 4451, 1973.
Dobson, G.M.B., Observations of the amount of ozone in the Earth’s atmosphere and
its relation to other geophysical conditions. Proc R Soc London, Ser A.: 129, 411,
1930.
Dobson, G.M.B., Forty years’ research on atmospheric ozone at Oxford: A history.
App Opt: 7,
7 387, 1968.
Donovan, D.P., H. Fast, Y. Makino, J.C. Bird, A.I. Carswell, J. Davies, T.J. Duck,
J.W. Kaminski, C.T. Mc Elroy, R.L. Mittermeier, S.R. Pal, V. Savastiouk, D. Velkov,
and J.A. Whiteway, Ozone, column ClO, and PSC measurements made at the NDSC
Eureka observatory (80◦ N, 86◦ W) during the spring of 1997. Geophys Res Lett: 24,
2709, 1997.
Douglass, A.R., and R.B. Rood, Derivation of photochemical information near 1 mbar
from ozone and temperature data. J Geophys Res: 91, 13,153, 1986.
Douglass, A.R., R.B. Rood, and R.S. Stolarski, Interpretation of ozone temperature
correlations. 2, Analysis of SBUV ozone data. J Geophys Res: 90, 10,693, 1985.
Douglass, A.R., M.R. Schoeberl, R.S. Stolarski, J.W. Waters, J.M. Russell III, A.E.
Roche, and S.T. Massie., Interhemispheric differences in springtime production of
HCl and ClONO2 in the polar vortices. J Geophys Res: 100, 13,967, 1995.
Elkins, J.W., T.M. Thompson, T.H. Swanson, J.H. Butler, B.D. Hall, S.O.E
Cummings, D.A. Fisher, and A.G. Raffo, Decrease in the growth rates of atmospheric
chlorofluorocarbons 11 and 12. Nature: 364, 780, 1993.
Fahey, D.W., D.M. Murphy, K.K. Kelly, M.K.W. Ko, M.H. Proffitt, C.S.E Eubank,
G.V. Ferry, M. Loewenstein, and K.R. Chan. Measurements of nitric oxide and
total reactive nitrogen in the Antarctic stratosphere: Observations and chemical
implications. J Geophys Res: 94, 16,665, 1989a.
Fahey, D.W., K.K. Kelly, G.V. Ferry, L.R. Poole, J.C. Wilson, D.M. Murphy,
M. Loewensten, and K.R. Chan, In situ measurements of total reactive nitrogen,
total water and aerosol in a polar stratospheric cloud in the Antarctic. J Geophys
Res: 94, 11,299, 1989b.
Fahey, D.W., S.R. Kawa, and K.R. Chan, Nitric oxide measurements in the arctic
winter stratosphere. Geophys Res Lett: 17,
7 489, 1990a.
Fahey, D.W., K.K. Kelly, S.R. Kawa, A.F. Tuck, M. Loewenstein, K.R. Chan,
and L.E. Heidt, Observations of denitrification and dehydration in the winter polar
stratospheres. Nature: 344, 321, 1990b.
Fahey, D.W., S.R. Kawa, E.L. Woodbridge, P. Tin, J.C. Wilson, H.H. Jonsson, J.E.
Dye, D. Baumgardner, S. Borrmann, D.W. Toohey, L.M. Avallone, M.H. Proffitt, J.
Margitan, M. Loewenstein, J.R. Podolske, R.J. Salawitch, S.C. Wofsy, M.K.W. Ko,
D.E. Anderson, M.R. Schoeberl, and K.R. Chan, In situ measurements constraining
the role of sulfate aerosols in mid-latitude ozone depletion. Nature: 363, 509, 1993.
Fahey, D.W., R.S. Gao, K.S. Carslaw, J. Kettleborough, P.J. Popp, M.J. Northway,
J.C. Holecek, S.C. Ciciora, R.J. McLaughlin, T.L. Thompson, R.H. Winkler, D.G.
Baumgardner, B. Gandrud, P.O. Wennberg, S. Dhaniyala, K. McKinney, Th. Peter,
R.J. Salawitch, T.P. Bui, J.W. Elkins, C.R. Webster, E.L. Atlas, H. Jost, J.C. Wilson,
R.L. Herman, A. Kleinböhl,
¨ M. von Konig,
¨ The detection of large HNO3 -containing
particles in the winter Arctic stratosphere. Science: 291, 1026, 2001.
Farman, J.C., B.G. Gardiner, and J.D. Shanklin, Large losses of total ozone in
Antarctica reveal seasonal ClOx /NOx interaction. Nature: 315, 207, 1985.
Farmer, C.B., G.C. Toon, P.W. Shaper, J.F. Blavier, and L.L. Lowes, Stratospheric
trace gases in the spring 1986 Antarctic atmosphere. Nature: 329, 126, 1987.
Fels, S.B., J.D. Mahlman, M.D. Schwarzkopf, and R.W. Sinclair, Stratospheric
sensitivity to perturbations in ozone and carbon dioxide radiative and dynamical
response. J Atmos Sci: 37,7 2265, 1980.
Finger, F.G., R.M. Nagatani, M.E. Gelman, C.S. Long, and A.J. Miller, Consistency
between variations of ozone and temperature in the stratosphere. Geophys Res Lett:
22, 3477, 1995.
Fleming, E.L., S. Chandra, J.J. Barnett, and M. Corney, Zonal mean temperatures,
pressures, zonal winds, and geopotenial heights as a function of latitude. Adv Space
Res: 10, 1211, 1990.
Forster, P.M., and K.P. Shine, Stratospheric water vapor changes as a possible
contributor to observed stratospheric cooling. Geophys Res Lett: 26, 3309, 1999.
Fusco, A.C., and M.L. Salby, Interannual variations of total ozone and their
relationship to variations of planetary wave activity. J Clim: 12, 1619, 1999.
Gandrud, B.W., J.E. Dye, D. Baumgardner, G.V. Ferry, M. Loewenstein, K.R. Chan,
L. Sanford, B. Gary, and K. Kelly, The January 30, 1989 Arctic polar stratospheric
clouds (PSC) event: Evidence for a mechanism of dehydration. Geophys Res Lett:
17,
7 457, 1990.
Gao, R.S., B. Kaercher, E.R. Keim, and D.W. Fahey, Constraining the heterogne-
neous loss of O3 on soot particles with observations in jet engine exhaust plumes.
Gao, R.S., E.C. Richard, P.J. Popp, G.C. Toon, D.F. Hurst, P.A. Newman, J.C.
Holecek, M.J. Northway, D.W. Fahey, M.Y. Danilin, B. Senh, K. Aikin, P.A.
Romashkin, J.W. Elkins, C.R. Webster, S.M. Schauffler, J.B. Greenblatt, C.T.
McElroy, L.R. Lait, T.P. Bui, and D. Baumgardner, Observational evidence for the
role of denitrification in Arctic stratospheric ozone loss. Geophys Res Lett: 28, 2879,
2001.
Gardiner, B.C., Comparative morphology of the vertical ozone profile in the Antarctic
spring. Geophys Res Lett: 15, 901, 1988.
Geller, M.A., and J.C. Alpert, Planetary wave coupling between the troposphere and
the middle atmosphere as a possible sun-weather mechanism. J Atmos Sci: 37,7 1197,
1980.
Geller, M.A., V. Yudin, A.R. Douglass, J.W. Waters, L.S. Elson, A.E. Roche, and
J.M. Russell III, UARS PSC, ClONO2 , HCl and ClO measurements in early winter:
Additional verification of the paradigm for chlorine activation. Geophys Res Lett: 22,
2937, 1995.
Gerber, L., and N. Kampfer,
¨ Millimeter-wave measurements of chlorine monoxide at
the Jungfraujoch Alpine Station. Geophys Res Lett: 21, 1279, 1994.
Gleason, J.F., P.K. Bhartia, J.R. Herman, R. Mc Peters, P. Newman, R.S. Stolarski,
L. Flynn, G. Labow, D. Larko, C. Seftor, C. Wellemeyer, W.D. Komhyr, A.J. Miller,
and W. Planet, Record low global ozone in 1992. Science: 260, 523, 1993.
Godin, S., G. Megie,
´ C. David, D. Haner, C. Flesia, and Y. Emery, Airborne lidar
observations of mountain-wave-induced polar stratospheric clouds during EASOE.
Goutail, F., J.P. Pommereau, A. Sarkissian, E. Kyro, and V. Dorokhov, Total nitrogen
dioxide at the Arctic polar circle since 1990. Geophys Res Lett: 21, 1371, 1994.
Grewe, V., M. Dameris, and R. Sausen, Impact of stratospheric dynamics and
chemistry on northern hemisphere midlatitude ozone loss. J Geophys Res: 103,
25,417, 1998.
Haigh, J.D., The impact of solar variability of climate. Science: 272, 981, 1996.
Haigh, J.D., and J.A. Pyle, Ozone perturbation experiments in a two-dimensional
circulation model. Quart J Roy Meteor Soc: 108, 551, 1982.
Haigh, J.D., and J.A. Pyle, A two-dimensional calculation including atmospheric
carbon dioxide and stratospheric ozone. Nature: 279, 222, 1979.
Hampson, J., Photochemical Behavior of the Ozone Layer. Tech Note, Can Armament
Res and Dev Estab, TN 1627/64, Valcartier, Canada, 1964.
Hanson, D. R., and K. Mauersberger, The vapor pressures of HNO3 -H2 O solutions.
J Phys Chem: 92, 6167, 1988.
Hanson, D.R., and A.R. Ravishankara, Heterogeneous chemistry of bromine species
in sulfuric acid under stratospheric conditions. Geophys Res Lett: 22, 385, 1995.
Hanson, D.H., A.R. Ravishankara, and S. Solomon, Heterogeneous reactions in
sulfuric acid aerosols: A framework for model calculations. J Geophys Res: 99,
3615, 1994.
Hanson, D.R., A.R. Ravishankara, and E.R. Lovejoy, Reactions of BrONO2 and H2 O
on submicron sulfuric acid aerosol and the implications for the lower stratosphere.
J Geophys Res: 101, 9063, 1996.
Harris, N.R.P., G. Ancellet, L. Bishop, D.J. Hofmann, J.B. Kerr, R.D. Mc Peters, M.
Prendez, W.J. Randel, J. Staehelin, B.H. Subbaraya, A. Volz-Thomas, J. Zawodny,
and C.S. Zerefos, Trends in stratospheric and free tropospheric ozone. J Geophys
Res: 102, 1571, 1997.
Hartley, D.E., J.T. Villarin, R.X. Black, and C.A. Davis, A new perspective on the
dynamical link between the stratosphere and troposphere. Nature: 391, 471, 1998.
Hartley, W.N., On the probable absorption of solar radiation by atmospheric ozone.
Chem News: 42, 268, 1880.
Hartmann, D.L., L.E. Heidt, M. Loewenstein, J.R. Podolske, J. Vedder, W.L. Starr,
and S.E. Strahan, Transport into the south polar vortex in early spring. J Geophys
Res: 94, 16,779, 1989.
Heath, D.F., A.J. Krueger, and P.J. Crutzen, Solar proton event: Influence on
stratospheric ozone. Science: 197,
7 886, 1977.
Hofmann, D.J., and S. Solomon, Ozone destruction through heterogeneous chemistry
Hofmann, D.J., and S.J. Oltmans, The effect of stratospheric water vapor on the
heterogeneous reaction rate for ClONO2 and H2 O for sulfuric acid aerosol. Geophys
Res Lett: 19, 2211, 1992.
Hofmann, D.J., and S.J. Oltmans, Anomalous Antarctic ozone during 1992: Evidence
for Pinatubo volcanic aerosol effects. J Geophys Res: 98, 18,555, 1993.
Hofmann, D.J., J.W. Harder, S.R. Rolf, and J.M. Rosen, Balloon-borne observations
of the development and vertical structure of the Antarctic ozone hole in 1986. Nature:
326, 59, 1987.
Hofmann, D.J., S.J. Oltmans, J.M. Harris, S. Solomon, T. Deschler, and B.J. Johnson,
Observations and possible causes of new ozone depletion in Antarctica in 1991.
Nature: 359, 283, 1992.
Hofmann, D.J., S.J. Oltmans, W.D. Komhyr, J.M. Harris, J.A. Lathrop, A.O.
Langford, T. Deshler, B.J. Johnson, A. Torres, and W.A. Matthews, Ozone loss in the
lower stratosphere over the United States in 1992-1993: Evidence for heterogeneous
chemistry on the Pinatubo aerosol. Geophys Res Lett: 21, 65, 1994.
Hofmann, D.J., S.J. Oltmans, J.M. Harris, B.J. Johnson, and J.A. Lathrop, Ten
years of ozonesonde measurements at the south pole: Implications for recovery of
springtime Antarctic ozone. J Geophys Res: 102, 8931, 1997.
Hollandsworth, S.M., R.D. McPeters, L.E. Flynn, W. Planet, A.J. Miller, and S.
Chandra, Ozone trends deduced from combined Nimbus 7 SBUV and NOAA 11
SBUV/2 data. Geophys Res Lett: 22, 905, 1995.
Holton, J.R., P.H. Haynes, M.E. McIntyre, A.R. Douglass, R.B. Rood, and L. Pfister,
Hood, L.L., and D.A. Zaff, Lower stratospheric stationary waves and the longitude
dependence of ozone trends in winter. J Geophys Res: 100, 25,791, 1995.
Hood, L.L., J.P. McCormack, and K. Labitzke, An investigation of dynamical
contributions to midlatitude ozone trends in winter. J Geophys Res: 102, 13,079,
1997.
Intergovernmental Panel on Climate Change (IPCC), Radiative Forcing of Climate
Change: The Scientific Assessment, Cambridge University Press, 1994.
Intergovernmental Panel on Climate Change (IPCC), Climate Change: The Scientific
Assessment, Cambridge University Press, 2001.
Iwasaka, Y., and K. Kondoh, Depletion of Antarctic ozone: Height of ozone loss
region and its temporal changes. Geophys Res Lett: 14, 87, 1987.
Jackman, C.H., and R.D. McPeters, Solar proton events as tests for the fidelity of
middle atmosphere models. Physica Scripta T18, 309, 1987.
Jackman, C.H., E.L. Fleming, S. Chandra, D.B. Considine, and J.E. Rosenfield,
Past, present, and future modeled ozone trends with comparisons to observed trends.
J Geophys Res: 101, 28,753, 1996.
Jackman, C.H., E.L. Fleming, and F.M. Vitt, Influence of extremely large solar proton
events in a changing stratosphere. J Geophys Res: 105, 11,659, 2000.
Jackman, C.H., R.D. McPeters, G.J. Labow, E.L. Fleming, C.J. Praderas, and J.M.
Russell, Northern Hemisphere atmospheric effects due to the July 2000 solar proton
event. Geophys Res Lett: 28, 2883, 2001.
Jaeger, H., and K. Wege, Stratospheric ozone depletion at northern midlatitudes after
major volcanic eruptions. J Atmos Chem: 10, 273, 1990.
Jaramillo, M., R.L. de Zafra, J. Barrett, L.K. Emmons, P.M. Solomon, and A. Parrish,
Measurements of stratospheric hydrogen cyanide at McMurdo Station, Antarctica:
Further evidence of winter stratospheric susidence? J Geophys Res: 94, 16,773, 1989.
Johnston, P.V., and R.L. McKenzie, NO2 observations at 45◦ S during the decreasing
phase of solar cycle 21, from 1980 to 1987. J Geophys Res: 94, 3473, 1989.
Johnston, P.V., R.L. McKenzie, J.G. Keys, and W.A. Matthews, Observations of
depleted stratospheric NO2 following the Pinatubo volcanic eruption. Geophys Res.
Lett: 19, 211-213, 1992.
Jones, R.L., and A.R. MacKenzie, Observational studies on the role of polar regions
in midlatitude ozone loss. Geophys Res Lett: 22, 3485, 1995.
Jones, A.E., and J.D. Shanklin, Continued decline of total ozone over Halley,
Antarctica, since 1985. Nature: 376, 409, 1995.
Jones, R.L., and J.A. Pyle, Observations of CH4 and N2 O by the NIMBUS 7 SAMS:
A comparison with in-situ data and two dimensional numerical model calculations.
J Geophys Res: 89, 5263, 1984.
Jones, R.L., J. Austin, D.S. McKenna, J.G. Anderson, D.W. Fahey, C.B. Farmer,
L.E. Heidt, K.K. Kelly, D.M. Murphy, M.H. Proffitt, A.F. Tuck, and J.F. Vedder,
Lagrangian photochemical modeling studies of the 1987 Antarctic spring vortex, 1.
Comparison with AAOE observations. J Geophys Res: 94, 11,529, 1989.
Jones, R.L., S. Solomon, D.S. McKenna, L.R. Poole, W.H. Brune, D.W. Toohey,
J.G. Anderson, and D.W. Fahey, The polar stratospheric cloud event of January 24,
2. Photochemistry. Geophys Res Lett: 17,
7 541, 1990a.
Jones, R.L., D.S. McKenna, L.R. Poole, and S. Solomon, On the influence of polar
stratospheric cloud formation on chemical composition during the 1988/89 arctic
winter. Geophys Res Lett: 17,
7 545, 1990b.
Use in Stratospheric Modelling, Evaluation number 12 of the NASA panel for data
evaluation, JPL publication 97-4, 1997.
Use in Stratospheric Modelling, Evaluation number 13 of the NASA panel for data
evaluation, JPL publication 00-3, 2000.
Kaercher, B., and S. Solomon, On the composition and optical extinction of particles
in the tropopause region. J Geophys Res: 104, 27441, 1999.
Kawa, S.R., D.W. Fahey, L.C. Anderson, M. Loewenstein, and K.R. Chan,
Measurements of total reactive nitrogen during the airborne Arctic stratospheric
expedition. Geophys Res Lett: 17,
7 485, 1990.
Kawa, S.R., P.A. Newman, L.R. Lait, M.R. Schoeberl, R.M. Stimpfle, D.W. Kohn,
C.R. Webster, R.D. May, D. Baumgardner, J.E. Dye, J.C. Wilson, K.R. Chan, and
M. Loewenstein, Activation of chlorine in sulfate aerosol as inferred from aircraft
observations. J Geophys Res: 102, 3921, 1997.
Keim, E.R., D.W. Fahey, L.A. Del Negro, E.L. Woodbridge, R.S. Gao, P.O.
Wennberg, R.C. Cohen, R.M. Stimpfle, K.K. Kelly, E.J. Hintsa, J.C. Wilson, H.H.
Jonsson, J.E. Dye, D. Baumgardner, S.R. Kawa, R.J. Salawitch, M.H. Proffitt, M.
Loewenstein, J.R. Podolske, and K.R. Chan, Observations of large reductions in
the NO/NOy ratio near the mid-latitude tropopause and the role of heterogeneous
chemistry. Geophys Res Lett: 23, 3223, 1996.
Kelly, K.K., A.F. Tuck, D.M. Murphy, M.H. Proffitt, D.W. Fahey, R.L. Jones, D.S.
McKenna, M. Loewenstein, J.R. Podolske, S.E. Strahan, G.V. Ferry, K.R. Chan, J.F.
Vedder, G.L. Gregory, W.D. Hypes, M.P. McCormick, E.V. Browell, and L.E. Heidt,
Dehydration in the lower Antarctic statosphere during late winter and early spring,
1987. J Geophys Res: 94, 11,317, 1989.
Khattatov, V., V. Yushkov, M. Khaplanov, I. Zaitzev, J. Rosen, and N. Kjome, Some
results of water vapor, ozone and aerosol balloon-borne measurements during EASOE.
Kirchhoff, V.W.J.H., N.J. Schuch, K.K. Pinheiro, and J.M. Harris, Evidence for an
ozone hole perturbation at 30◦ South. Atmos Environ: 30, 1481, 1996.
Kirchhoff, V.W.J.H., C.A.R. Casiccia S., Zamorano B., and F. Zamorano B., The
ozone hole over Punta Arenas, Chile. J Geophys Res: 102, 8945, 1997a.
Kirchhoff, V.W.J.H., Y. Sahai, C.A.R. Casiccia S., F. Zamorano B., and
V. Valderrama V., Observations of the 1995 ozone hole over Punta Arenas, Chile.
J Geophys Res: 102, 16,109, 1997b.
Knight, J.R., J. Austin, R.G. Grainger, and A. Lambert, A three-dimensional model
simulation of the impact of Mt. Pinatubo aerosol on the Antarctic ozone hole. Quart
J Roy Meteor Soc: 124, 1527, 1998.
Koike, M., N.B. Jones, W.A. Matthews, P.V. Johnston, R.L. McKenzie, D. Kinnison,
and J. Rodriguez, Impact of Pinatubo aerosols on the partitioning between NO2 and
HNO3 . Geophys Res Lett: 21, 597, 1994.
Komhyr, W.D., R.D. Grass, and R.K. Leonard, Total ozone decrease at South Pole
Antarctica, 1964-1985. Geophys Res Lett: 13, 1248, 1986.
Kondo, Y., U. Schmidt, T. Sugita, P. Aimendieu, M. Koike, H. Ziereis, and
Y. Iwasaka, Total reactive nitrogen, N2 O and ozone in the winter Arctic stratosphere.
Koop, T., U.M. Biermann, W. Raber, B.P. Luo, P.J. Crutzen, and T. Peter, Do
stratospheric aerosols freeze above the ice frost point? Geophys Res Lett: 22, 917,
1995.
Kreher, K., J.G. Keys, P.V. Jonston, U. Platt, and X. Liu, Ground-based
measurements of OClO and HCl in austral spring 1993 at Arrival Heights, Antarctica.
Krueger, A.J., B.W. Buether, A.J. Fleig, D.F. Heath, E. Hilsenrath, R. Mc Peters,
and C. Prabakhara, Satellite ozone measurements. Phil Trans Roy Soc London: A,
191, 1980.
Lary, D.J., Gas phase bromine photochemistry. J Geophys Res: 101, 1505, 1996.
Lary, D.J., M.P. Chipperfield, R. Toumi, and T. Lenton, Heterogeneous atmospheric
bromine chemistry. J Geophys Res: 101, 1489, 1996.
Lary, D.J., A.M. Lee, R. Toumi, M.J. Newchurch, M. Pirre, and J.B. Renard, Carbon
aerosols and atmospheric photochemistry. J Geophys Res: 102, 3671, 1997.
Lee, A.M., H.K. Roscoe, A.E. Jones, P.H. Haynes, E.F. Shuckburgh, M.W. Morrey,
and H.C. Pumphrey, The impact of the mixing properties within the Antarctic
stratospheric vortex on ozone loss in spring. J Geophys Res: 106, 3203, 2001.
Lehmann, P., D.J. Karoly, P.A. Newman, T.S. Clarkson, and W.A. Matthews, An
investigation into the causes of stratospheric ozone loss in the southern australasian
region. Geophys Res Lett: 19, 1463, 1992.
Lipson, J.B., M.J. Elrod, T.W. Beiderhase, L.T. Molina, and M.J. Molina,
Temperature dependence of the rate constant and branching ratio for the OH+ClO
reaction. J Chem Soc: 93, Faraday Trans, 2665, 1997.
Loewenstein, M., J.R. Podolske, K.R. Chan, and S.E. Strahan, Nitrous oxide as a
dynamical tracer in the 1987 airborne Antarctic ozone experiment. J Geophys Res:
94, 11,589, 1989.
Luther, F.M., D.J. Wuebbles, and J.S. Chang, Temperature feedback in a
stratospheric model, J Geophys Res: 82, 4935, 1977.
Lutman, E.R., R. Toumi, R.L. Jones, D.J. Lary, and J.A. Pyle, Box model studies
of ClOx deactivation and ozone loss during the 1991/92 northern hemisphere winter.
Geophys Res Lett: 21, 1415, 1994a.
Lutman, E.R., J.A. Pyle, R.L. Jones, D.J. Lary, A.R. MacKenzie, I. Kilbane-Dawe,
N. Larsen, and B. Knudsen, Trajectory model studies of ClOx activation during the
1991/92 northern hemispheric winter. Geophys Res Lett: 21, 1419, 1994b.
Lutman, E.R., J.A. Pyle, M.P. Chipperfield, D.J. Lary, I. Kilbane-Dawe, J.W. Waters,
and N. Larsen, Three-dimensional studies of the 1991/1992 northern hemisphere
winter using domain-filling trajectories with chemistry. J Geophys Res: 102, 1479,
1997.
MacKenzie, A.R., M. Kulmala, A. Laaksonen, and T. Vesala, On the theories of type I
polar stratospheric cloud formation. J Geophys Res: 100, 11,275, 1995.
MacKenzie, I.A., and R.S. Harwood, Arctic ozone destruction and chemical-radiative
interaction. J Geophys Res: 105, 9033, 2000.
MacKenzie, I.A., R.S. Harwood, L. Froidevaux, W.G. Read, and J.W. Waters,
Chemical loss of polar vortex ozone inferred from UARS MLS measurements of ClO
during the Arctic and Antarctic late winters of 1993. J Geophys Res: 101, 14,505,
1996.
Mahlman, J.D., and S.B. Fels, Antarctic ozone decreases: A dynamical cause?
Mankin, W.G., M.T. Coffey, A. Goldman, M.R. Schoeberl, L.R Lait, and P.A.
Newman, Airborne measurements of stratospheric constituents over the Arctic in
the winter of 1989. Geophys Res Lett: 17,
7 473, 1990.
Mankin, W.G., M.T. Coffey, and A. Goldman, Airborne observations of SO2 , HCl,
and O3 in the stratospheric plume of the Pinatubo volcano in July, 1991, Geophys
Res Lett: 19, 179, 1992.
Manney, G.L., L. Froidevaux, J.W. Waters, R.W. Zurek, W.G. Read, L.S. Elson, J.B.
Kumer, J.L. Mergenthaler, A.E. Roche, A. O’Neill, R.S. Harwood, I. MacKenzie, and
R. Swinbank, Chemical depletion of ozone in the Arctic lower stratosphere during
winter 1992-93. Nature: 370, 429, 1994a.
Manney, G.L, R.W. Zurek, A. O’Neill, and R. Swinbank, On the motion of air through
the stratospheric polar vortex. J Atmos Sci: 51, 2973, 1994b.
Manney, G.L., R.W. Zurek, W.A. Lahoz, R.S. Harwood, J.C. Gille, J.B. Kumer, J.L.
Mergenthaler, A.E. Roche, A. O’Neill, R. Swinbank, and J.W. Waters, Langrangian
transport calculations using UARS data. Part I: Passive tracers. J Atmos Sci: 52,
3049, 1995a.
Manney, G.L., R.W. Zurek, L. Froidevaux, J.W. Waters, A. O’Neill, and R. Swinbank,
Lagrangian transport calculations using UARS data. Part II: Ozone. J Atmos Sci:
52, 3069, 1995b.
Manney, G.L., R.W. Zurek, L. Foidevaux, and J.W. Waters, Evidence for arctic ozone
depletion in late February and early March 1994. Geophys Res Lett: 22, 2941, 1995c.
Manney, G.L., M.L. Santee, L. Froidevaux, J.W. Waters, and R.W. Zurek, Polar
vortex conditions during the 1995-96 Arctic winter: Meteorology and MLS ozone.
Manney, G.L., L. Froidevaux, M.L. Santee, R.W. Zurek, and J.W. Waters, MLS
observations of Arctic ozone loss in 1996-97. Geophys Res Lett: 24, 2697, 1997.
McCormack, J.P., and L.L. Hood, Modeling the spatial distribution of total ozone in
northern hemisphere winter: 1979-1991. J Geophys Res: 102, 13,711, 1997.
McCormick, M.P., H.M. Steele, P. Hamill, W.P. Chu, and T.J. Swissler, Polar
stratospheric cloud sightings by SAM II. J Atmos Sci: 39, 1387, 1982.
McCormick, M.P., R.E. Veiga, and W.P. Chu, Stratospheric ozone profile and total
ozone trends derived from the SAGE I and SAGE II data. Geophys Res Lett: 19,
269, 1992.
McCormick, M.P., L.W. Thomason, and C.R. Trepte, Atmospheric effects of the
Mt. Pinatubo eruption. Nature: 373, 399, 1995.
McElroy, M.B., R.J. Salawitch, S.C. Wofsy, and J.A. Logan, Reductions of Antarctic
ozone due to synergistic interactions of chlorine and bromine. Nature: 321, 759, 1986.
McElroy, M.B., R.J. Salawitch, and K. Minschwaner, The changing stratosphere.
McGee, T.J., P. Newman, M. Gross, U. Singh, S. Godin, A-M. Lacoste, and G. Mégie,
´
Correlation of ozone loss with the presence of volcanic aerosols. Geophys Res Lett:
21, 2801, 1994.
McKenna, D.S., R.L. Jones, L.R. Poole, S. Solomon, D.W. Fahey, K.K. Kelly,
M.H. Proffitt, W.H. Brune, M. Loewenstein, and K.R. Chan, Calculations of ozone
destruction during the 1988/89 Arctic winter. Geophys Res Lett: 17,
7 553, 1990.
McKenzie, R.L., and P.V. Johnston, Springtime stratospheric NO2 in Antarctica.
McPeters, R.D., and G.J. Labow, An assessment of the accuracy of 14.5 years of
Nimbus 7 TOMS Version 7 ozone data by comparison with the Dobson network.
Geophys Res Lett: 23, 3695, 1996a.
McPeters, R.D., S.M. Hollandsworth, L.E. Flynn, J.R. Herman, and C.J. Seftor,
Long-term ozone trends derived from the 16-year combined Nimbus 7/Meteor 3
TOMS Version 7 record. Geophys Res Lett: 23, 3699, 1996b.
Michelangeli, D.V., M. Allen, and Y.L. Yung, El Chichón
´ volcanic aerosols: Impact
of radiative, thermal, and chemical perturbations. J Geophys Res: 94, 18,429, 1989.
Michelsen, H.A., R.J. Salawitch, M.R. Gunson, C. Aellig, N. Kämpfer,
¨ M.M. Abbas,
M.C. Abrams, T.L. Brown, A.Y. Chang, A. Goldman, F.W. Irion, M.J. Newchurch,
C.P. Rinsland, G.P. Stiller, and R. Zander, Stratospheric chlorine partitioning:
Constraints from shuttle-borne measurements of [HCl], [ClNO3 ] and [ClO]. Geophys
Res Lett: 23, 2361, 1996.
Michelsen, H.A., C.M. Spivakovsky, and S.C. Wofsy, Aerosol-mediated partitioning

of stratospheric Cly and NOy at temperatures below 200. Geophys Res Lett: 26, 299,
1999.
Miller, A.J., G.C. Tiao, G.C. Reinsel, D. Wuebbles, L. Bishop, J. Kerr, R.M.
Nagatani, J.J., DeLuisi, and C.L., Mateer, Comparisons of observed ozone trends
in the stratosphere through examinations of Umkehr and balloon ozonesonde data. J
Geophys Res: 100, 11,209, 1995.
Mills, M.J., A.O. Langford, T.J. O’Leary, K. Arpag, H.L. Miller, M.H. Proffitt, R.W.
Sanders, and S. Solomon, On the relationship between stratospheric aerosols and
nitrogen dioxide. Geophys Res Lett: 20, 1187, 1993.
Molina, M.J., and F.S. Rowland, Stratospheric sink for chlorofluoromethanes:
Chlorine atom catalyzed destruction of ozone. Nature: 249, 820, 1974.
Molina, L.T., and M.J. Molina, Production of Cl2 O2 from the self-reaction of the
ClO radical. J Phys Chem: 91, 433, 1987.
Molina, M. J., T-L. Tso, L.T. Molina, and F.C.-Y. Wang, Antarctic stratospheric
chemistry of chlorine nitrate, hydrogen chloride, and ice: Release of active chlorine,
Science: 238, 1253, 1987.
Montzka, S.A., J.H. Butler, R.C. Myers, T.M. Thompson, T.H. Swanson, A.D. Clarke,
L.T. Lock, and J.W. Elkins, Decline in the tropospheric abundance of halogen from
halocarbons: implications for stratospheric ozone depletion. Science: 272, 1318, 1996.
Mount, G.H., R.W. Sanders, A.L. Schmeltekopf, and S. Solomon, Visible spectroscopy
at McMurdo Station, Antarctica, 1, Overview and daily variations of NO2 and O3
during austral spring, 1986. J Geophys Res: 92, 8320, 1987.
Mozurkewich, M., and J.G. Calvert, Reaction probability of N2 O5 on aqueous
Muller,
¨ R., P.J. Crutzen, J-U. Grooß, C. Br¨ uhl,
¨ J.M. Russell III, and A.F. Tuck,
Chlorine activation and ozone depletion in the Arctic vortex: Observations by the
halogen occultation experiment on the upper atmosphere research satellite. J Geophys
Res: 101, 12,531, 1996.
Muller,
¨ R., P.J. Crutzen, J-U. Grooß, C. Br¨
uhl,
¨ J.M. Russell III, H. Gernandt, D.S.
McKenna, and A.F. Tuck, Severe chemical ozone loss in the Arctic during the winter
of 1995-96. Nature: 389, 709, 1997a.
Muller,
¨ R., J-U. Grooß, D.S. McKenna, P.J. Crutzen, C. Br¨ uhl,
¨ J.M. Russell III, and
A.F. Tuck, HALOE observations of the vertical structure of chemical ozone depletion
in the Arctic vortex during winter and early spring 1996-1997. Geophys Res Lett: 24,
2717, 1997b.
Muller,
¨ R, R.J. Salawitch, P.J. Crutzen, W.A. Lahoz, G.L. Manney, and R.
Tuomi, Upper stratospheric processes, in Scientific Assessment of Ozone Depletion:
1998, Rep. 44, pp. 6.1-6.4, World Meteorological Organization/United Nations
Environment Programme (WMO/UNEP), Geneva, 1999.
Murcray, D.G., D.B. Barker, J.N. Brooks, A. Goldman, and W.J. Williams, Seasonal
and latitudinal variation of the stratospheric concentration of HNO3 . Geophys Res
Lett: 2, 223, 1975.
Murphy, D.M., Ozone loss rates calculated along ER-2 flight tracks. J Geophys Res:
96, 5045, 1991.
Murphy, D.M., and A.R. Ravishankara, Temperature averages and rates of
stratospheric reactions. Geophys Res Lett: 21, 2471, 1994.
Nagatani, R.M., A.J. Miller, M.E. Gelman, and P.A. Newman, A comparison of
Arctic lower stratospheric winter temperatures for 1988-89 with temperatures since
1964. Geophys Res Lett: 17,7 333, 1990.
Newman, P.A., Antarctic total ozone in 1958. Science: 264, 543, 1994.
Newman, P.A., and M.R. Schoeberl, October Antartic temperature and total ozone
trends from 1979 to 1985. Geophys Res Lett: 13, 1207, 1986.
Newman, P.A., and W.J. Randel, Coherent ozone-dynamical changes during the
southern hemisphere spring, 1979-1986. J Geophys Res: 93, 12,585, 1988.
Newman, P.A., Antarctic total ozone in 1958. Science: 264, 543, 1994.
Newman, P.A., J.G. Gleason, R.D. McPeters, and R.S. Stolarski, Anomalously low
ozone over the Arctic. Geophys Res Lett: 24, 2689, 1997.
Noxon, J.F., Stratospheric NO2 in the Antarctic winter. Geophys Res Lett: 5, 1021,
1978.
Noxon, J.F., Stratospheric NO2 , 2. Global behavior. J Geophys Res: 84, 5067, 1979.
Oelhaf, H., T. von Clarmann, H. Fischer, F. Friedl-Vallon, C. Fritzsche, A. Linden,
C. Piesch, M. Seefeldner, and W. Volker,
¨ Stratospheric ClONO2 and HNO3 profiles
inside the Arctic vortex from MIPAS-B limb emission spectra obtained during
EASOE. Geophys Res Lett: 21, 1263, 1994.
Ozone Trends Panel (OTP), Report of the International Ozone Trends Panel, WMO
Report Number 18, 1988.
Parrish, A., R.L. deZafra, P.M. Solomon, J.W. Barrett, and E.R. Carlson, Chlorine
oxide in the stratospheric ozone layer: Ground-based detection and measurement.
Science: 211, 1158, 1981.
Parrish, A., R.L. de Zafra, M. Jaramillo, B. Connor, P.M. Solomon, and J.W. Barrett,
Extremely low N2 O concentrations in the springtime stratosphere at McMurdo
Station, Antarctica. Nature: 332, 53, 1988.
Perliski, L., S. Solomon, and J. London, On the interpretation of seasonal variations
in stratospheric ozone. Planet Space Sci: 37,
7 1527, 1989.
Perner, D., A. Roth, and T. Klüpfel,
¨ Ground based measurements of stratospheric
OClO, NO2 and O3 at Søndre Strømfjord in winter 1991/92. Geophys Res Lett: 21,
1367, 1994.
Pfeilsticker, K., and U. Platt, Airborne measurements during the Arctic Stratospheric
Experiment: Observations of O3 and NO2 . Geophys Res Lett: 21, 1375, 1994.
Pierce, R.B., W.L. Grose, J.M. Russell III, A.F. Tuck, R. Swinbank, and A. O’Neill,
Spring dehydration in the Antarctic stratospheric vortex observed by HALOE.
J Atmos Sci: 51, 2931, 1994.
Pitari, G., and V. Rizi, An estimate of the chemical and radiative perturbation of
stratospheric ozone following the eruption of Mt. Pinatubo. J Atmos Sci: 50, 3260,
1993.
Pitari, G., G. Visconti, and M. Verdecchia, Global ozone depletion and the Antarctic
ozone hole. J Geophys Res: 97,7 8075, 1992.
Pitari, G., and V. Rizi, An estimate of the chemical and radiative perturbation of
stratospheric ozone following the eruption of Mt. Pinatubo. J Atmos Sci: 50, 3260,
1993.
Pittock, A.B., A critical look at long-term sun-weather relationships. Rev Geophys
Space Phys: 16, 400, 1978.
Plumb, R.A., and M.K.W. Ko, Interrelationships between mixing raitos of long-lived
stratospheric constituents. J Geophys Res: 97,
7 10,145, 1992.
Pommereau, J-P., and J. Piquard, Observations of the vertical distribution of
stratospheric OClO. Geophys Res Lett: 21, 1231, 1994a.
Pommereau, J-P., and J. Piquard, Ozone and nitrogen dioxide vertical distributions
by UV-visible solar occultation from balloons. Geophys Res Lett: 21, 1227, 1994b.
Portmann, R.W., S. Solomon, R.R. Garcia, L.W. Thomason, L.R. Poole, and M.P.
McCormick, Role of aerosol variations in anthropogenic ozone depletion in polar
regions. J Geophys Res: 101, 22,991, 1996.
Prather, M., Catastrophic loss of stratospheric ozone in dense volcanic clouds.
J Geophys Res: 97,
7 10,187, 1992.
Prather, M., M.M. Garcia, R. Suozzo, and D. Rind, Global impact of the Antarctic
ozone hole: Dynamical dilution with a three dimensional chemical transport model.
J Geophys Res: 95, 3449, 1990.
Prenni, A.J., M.E. Wise, S.D. Brooks, and M.A. Tolbert, Ice nucleation in sulfuric
acid and ammonium sulfate particles. J Geophys Res: 106, 3037, 2001.
Prinn, R.G., R.F. Weiss, B.R. Miller, J. Huang, F.N. Alyea, D.M. Cunnold, P.J.
Fraser, D.E. Hartley, and P.G. Simmonds, Atmospheric trends and lifetime of
CH3 CCl3 and global OH concentrations. Science: 269, 187, 1995.
Proffitt, M.H., K. Aikin, J.J. Margitan, M. Loewenstein, J.R. Podolske, A. Weaver,
K.R. Chan, H. Fast, and J.W. Elkins, Ozone loss inside the northern polar vortex
during the 1991-1992 winter. Science: 261, 1150, 1993.
Pyle, J.A., G. Carver, J.L. Grenfell, J.A. Kettleborough, and D.J. Lary, Ozone loss
in Antarctica: The implications for global change. Phil Trans R Soc London B: 338,
219, 1992.
Pyle, J.A., M.P. Chipperfield, I. Kilbane-Dawe, A.M. Lee, R.M. Stimpfle, D. Kohn,
W. Renger, and J.W. Waters, Early modelling results from the SESAME and ASHOE
Campaigns. Faraday Discuss: 100, 371, 1995.
Ramaswamy, V., M.D. Schwarzkopf, and W.J. Randel, Fingerprint of ozone depletion
in the spatial and temporal pattern of recent lower-stratospheric cooling. Nature: 382,
616, 1996.
Randel, W.J., and J.B. Cobb, Coherent variations of monthly mean total ozone and
lower stratospheric temperature. J Geophys Res: 99, 5433, 1994.
Randel, W.J., and F. Wu, Cooling of the Arctic and Antarctic polar stratospheres
due to ozone depletion. J Climate: 12, 1467, 1999.
Randeniya, L.K., P.F. Vohralik, I.C. Plumb, and K.R. Ryan, Heterogeneous BrONO2
hydrolysis: Effect on NO2 columns and oxone at high latitudes in summer. J Geophys
Res: 102, 23,543, 1997.
Ravishankara, A.R., and D.R. Hanson, Differences in the reactivity of Type 1 PSC’s
depending on their phase. J Geophys Res: 101, 3885, 1996.
Ravishankara, A.R., and T.G. Shepherd (Lead Authors), M.P. Chipperfield, P.H.
Haynes, S.R. Kawa, T. Peter, R.A. Plumb, R.W. Randel, D.W. Waugh, and D.R.
Worsnop, Lower Stratospheric Processes, Chapter 7 in Scientific Assessment of
Ozone Depletion: 1998, Global Ozone Research and Monitoring Project–Report
No. 44, World Meteorological Organization/United Nations Environment Programme
(WMO/UNEP), Geneva, 1999.
Rex, M., N.R.P. Harris, P. von der Gathen, R. Lehmann, G.O. Braathen, E. Reimer,
A. Beck, M.P. Chipperfield, R. Alfier, M. Allaart, F. O’Connor, H. Dier, V. Dorokhov,
H. Fast, M. Gil, E. Kyrö,

¨ Z. Litynska, I.S. Mikkelsen, M.G. Molyneux, H. Nakane, J.
Notholt, M. Rummukainen, P. Viatte, and J. Wenger, Prolonged stratospheric ozone
loss in the 1995-96 Arctic winter. Nature: 389, 835, 1997.
Rex, M., P. von der Gathen, N.R.P. Harris, D. Lucic, B.M. Knudsen, G.O. Braathen,
S.J. Reid, H. DeBacker, H. Claude, R. Fabian, H. Fast, M. Gil, E. Kyrö, ¨ I.S.
Mikkelsen, M. Rummukainen, H.G. Smit, J. Stähelin,
¨ C. Varotsos, and I. Zaitcev,
In situ measurements of stratospheric ozone depletion rates in the Arctic winter
1991/1992: A Lagrangian approach. J Geophys Res: 103, 5843, 1998.
Ridley, B.A., M. McFarland, A.L. Schmeltekopf, M.H. Proffitt, D.L. Albritton, R.H.
Winkler, and T.L. Thompson, Seasonal differences in the vertical distributions of NO,
NO2 , and O3 in the stratosphere near 50◦ N. J Geophys Res: 92, 11,919, 1987.
Rinsland, C.P., M.R. Gunson, M.C. Abrams, L.L. Lowes, R. Zander, E. Mahieu, A.
Goldman, M.K.W. Ko, J.M. Rodriguez, and N.D. Sze, Heterogeneous conversion of
N2 O5 to HNO3 in the post-Mt. Pinatubo eruption stratosphere. J Geophys Res: 99,
8213, 1994.
Rinsland, C.P., M.R. Gunson, R.J. Salawitch, M.J. Newchurch, R. Zander, M.M.
Abbas, M.C. Abrams, G.L. Manney, H.A. Michelsen, A.Y. Chang, and A. Goldman,
ATMOS measurements of H2 O+2CH4 and total reactive nitrogen in the November
1994 Antarctic stratosphere: Dehydration and denitrification in the vortex. Geophys
Res Lett: 23, 2397, 1996.
Robinson, G.N., D.R. Worsnop, J.T. Jayne, C.E. Kolb, and P. Davidovits,
Heterogeneous uptake of ClONO2 and N2 O5 by sulfuric acid solutions. J Geophys
Res: 102, 3583, 1997.
Robinson, G.N., D.R. Worsnop, J.T. Jayne, C.E. Kolb, E. Swartz, and P. Davidovits,
Heterogeneous uptake of HCl by sulfuric acid solutions. J Geophys Res: 103, 25,371,
1998.
Roche, A.E., J.B. Kumer, J.L. Mergenthaler, R.W. Nightingale, W.G. Uplinger, G.A.
Ely, J.F. Potter, D.J. Wuebbles, P.S. Connell, and D.E. Kinnison, Observations of
lower-stratospheric ClONO2 , HNO3 , and aerosol by the UARS, CLAES Experiment
between January 1992 and April 1993. J Atmos Sci: 51, 2877, 1994.
Rodriguez, J.M., M.K.W. Ko, and N.D. Sze, The role of chlorine chemistry in
Antarctic ozone loss: Implications of new kinetic data. Geophys Res Lett: 17,
7 255,
1990.
Rodriguez, J.M., M.K.W. Ko, and N.D. Sze, Role of heterogeneous conversion of
N2 O5 on sulphate aerosols in global ozone losses. Nature: 352, 134, 1991.
Rodriguez, J.M., M.K.W. Ko, N.D. Sze, C.W. Helsey, G.K. Yue, and M.P.
McCormick, Ozone response to enhanced heterogeneous processing after the eruption
of Mt. Pinatubo. Geophys Res Lett: 21, 209, 1994.
Rogaski, C.A., D.M. Golden, and L.R. Williams, Reactive uptake and hydration
experiments on amorphous carbon treated with NO2 , SO2 , O3 , HNO3 , and H2 SO4 .
Rood, R.B., and A.R. Douglass, Interpretation of ozone temperature correlations, 1.
Theory. J Geophys Res: 90, 5733, 1985.
Roscoe, H.K., A.E. Jones, and A.M. Lee, Midwinter start to Antarctic ozone
depletion: Evidence from observations and models. Science: 278, 93, 1997.
Geophys: 13, 1, 1975.
Russell III, J.M., S. Solomon, L.L. Gordley, E.E. Remsberg, and L.B. Callis, The
variability of stratospheric and mesospheric NOx in the polar winter night observed
by LIMS. J Geophys Res: 89, 7267, 1984.
Russell III, J.M., A.F. Tuck, L.L. Gordley, J.H. Park, S.R. Drayson, J.E. Harries,
R.J. Cicerone, and P.J. Crutzen, HALOE Antarctic observations in the spring of
1991. Geophys Res Lett: 20, 719, 1993.
Russell III, J.M., M. Luo, R.J. Cicerone, and L.E. Deaver, Satellite confirmation
of the dominance of chlorofluorocarbons in the global stratospheric chlorine budget.
Nature: 379, 526, 1996.
Salawitch, R.J., G.P. Gobbi, S.C. Wofsy, and M.B. McElroy, Denitrification in the
Antarctic stratosphere. Nature: 339, 525, 1989.
Salawitch, R.J., M.B. Mc Elroy, J.H. Yatteau, S.C. Wofsy, M.R. Schoeberl, L.R. Lait,
P.A. Newman, K.R. Chan, M. Loewenstein, J.R. Podolske, S.E. Strahan, and M.H.
Proffitt, Loss of ozone in the Arctic vortex for the winter of 1989. Geophys Res Lett:
17,
7 561, 1990.
Salawitch, R.J., S.C. Wofsy, E.W. Gottlieb, L.R. Lait, P.A. Newman, M.R. Schoeberl,
M. Loewenstein, J.R. Podolske, S.E. Strahan, M.H. Proffitt, C.R. Webster, R.D. May,
D.W. Fahey, D. Baumgardner, J.E. Dye, J.C. Wilson, K.K. Kelly, J.W. Elkins, K.R.
Chan, and J.G. Anderson, Chemical loss of ozone in the Arctic polar vortex in the
winter of 1991-1992. Science: 261, 1146, 1993.
Salawitch, R.J., S.C. Wofsy, P.O. Wennberg, R.C. Cohen, J.G. Anderson, D.W.
Fahey, R.S. Gao, E.R. Keim, E.L. Woodbridge, R.M. Stimpfle, J.P. Koplow, D.W.
Kohn, C.R. Webster, R.D. May, L. Pfister, E.W. Gottlieb, H.A. Michelsen, G.K. Yue,
M.J. Prather, J.C. Wilson, C.A. Brock, H.H. Jonsson, J.E. Dye, D. Baumgardner,
M.H. Proffitt, M. Loewenstein, J.R. Podolske, J.W. Elkins, G.S. Dutton, E.J. Hintsa,
A.E. Dessler, E.M. Weinstock, K.K. Kelly, K.A. Boering, B.C. Daube, K.R. Chan,
and S.W. Bowen, The diurnal variation of hydrogen, nitrogen, and chlorine radicals:
Implications for the heterogeneous production of HNO2 . Geophys Res Lett: 21, 2551,
1994.
Sander, S.P., R.R. Friedl, and Y.L. Yung, Rate of formation of the ClO dimer in the
polar stratosphere: Implications for ozone loss. Science: 245, 1095, 1989.
Sanders, R.W., S. Solomon, J.P. Smith, L. Perliski, H.L. Miller, G.H. Mount, J.G.
Keys, and A.L. Schmeltekopf, Visible and near-ultraviolet spectroscopy at McMurdo
Station, Antarctica, 9. Observations of OClO from April to October, 1991. J Geophys
Res: 98, 7219, 1993.
Santee, M.L., W.G. Read, J.W. Waters, L. Froidevaux, G.L. Manney, D.A. Flower,
R.F. Jarnot, R.S. Harwood, and G.E. Peckham, Interhemispheric differences in polar
stratospheric HNO3 , H2 O, ClO, and O3 . Science: 267,
7 849, 1995.
Santee, M.L., G.L. Manney, W.G. Read, L. Foidevaux, and J.W. Waters, Polar vortex
conditions during the 1995-96 Arctic winter: MLS ClO and HNO3 . Geophys Res Lett:
23, 3207, 1996.
Santee, M.L., G.L. Manney, L. Froidevaux, R.W. Zurek, and J.W. Waters, MLS
observations of ClO and HNO3 in the 1996-97 Arctic polar vortex. Geophys Res Lett:
24, 2713, 1997.
Santee, M.L., A. Tabazadeh, G.L. Manney, R.J. Salawitch, L. Froidevaux, W.G. Read,
and J.W. Waters, UARS Microwave limb sounder HNO3 observations: Implications
for Antarctic polar stratospheric clouds. J Geophys Res: 103, 13,285, 1998.
Santee, M.L., G.L. Manney, L. Froidevaux, W.G. Read, and J.W. Waters, Six years of
UARS microwave limb sounder HNO3 observations: Seasonal, interhemispheric, and
interannual variations in the lower stratosphere. J Geophys Res: 104, 8225, 1999.
Schauffler, S.M., L.E. Heidt, W.H. Pollock, T.M. Gilpin, J.F. Vedder, S. Solomon,
R.A. Lueb, and E.L. Atlas, Measurements of halogenated organic compounds near
the tropical tropopause. Geophys Res Lett: 20, 2567, 1993.
Schiller, C., A. Wahner, U. Platt, H-P. Dorn, J. Callies, and D.H. Ehhalt, Near UV
atmospheric absorption measurements of column abundances during airborne arctic
stratospheric expedition, January-February 1989, 2. OClO Observations. Geophys
Res Lett: 17,
7 501, 1990.
Schmidt, U., and A. Khedim, In situ measurements of carbon dioxide in the winter
Arctic vortex and at mid-latitudes: An indicator of the age of stratospheric air.
Schoeberl, M.R., and D.L. Hartmann, The dynamics of the polar vortex. Science:
251, 46, 1991.
Schoeberl, M.R., A.R. Douglass, R.S. Stolarski, P.A. Newman, L.R. Lait, D. Toohey,
L. Avallone, J.G. Anderson, W. Brune, D.W. Fahey, and K. Kelly, Stratospheric
constituent trends from ER-2 profile data. Geophys Res Lett: 17,
7 469, 1990.
Schoeberl, M.R., and D.L. Hartmann, The dynamics of the polar vortex. Science:
251, 46, 1991.
Schoeberl, M.R., L.R. Lait, P.A. Newman, and J.E. Rosenfield, The structure of the
polar vortex. J Geophys Res: 97,
7 7859, 1992.
Schoeberl, M.R., P.K. Bhartia, and J.R. Herman, Tropical ozone loss following the
eruption of Mt. Pinatubo. Geophys Res Lett: 20, 29, 1992.
Schoeberl, M.R., A.R. Douglass, R.S. Stolarski, P.A. Newman, L.R. Lair, D. Toohey,
L. Avallone, J.G. Anderson, W. Brune, D.W. Fahey, and K. Kelly, The evolution of
ClO and NO along air parcel trajectories. Geophys Res Lett: 20, 2511, 1993a.
Schoeberl, M.R., R.S. Stolarski, A.R. Douglass, P.A. Newman, L.R. Lait, J.W.
Waters, L. Froidevaux, and W.G. Ready, MLS ClO observations and the Arctic polar
vortex temperatures. Geophys Res Lett: 20, 2861, 1993b.
Schoeberl, M.R., M. Luo, and J.E. Rosenfield, An analysis of the Antarctic halogen
occultation experiment trace gas observations. J Geophys Res: 100, 5159, 1995.
Schoeberl, M.R., A.R. Douglass, S.R. Kawa, A.E. Dessler, P.A. Newman, R.S.
Stolarski, A.E. Roche, J.W. Waters, and J.M. Russell III, Development of the
Antarctic ozone hole. J Geophys Res: 101, 20,909, 1996.
Sen, B., G.C. Toon, G.B. Osterman, J.-F Blavier, J.J. Margitan, and R.J. Salawitch,
and G.K. Yue, Measurements of reactive nitrogen in the stratosphere. J Geophys
Res: 103, 3571, 1998.
Shindell, D.T., J.M. Reeves, L.K. Emmons, and R.L. de Zafra, Arctic chlorine
monoxide observations during spring 1993 over Thule, Greenland, and implications
for ozone depletion. J Geophys Res: 99, 25,697, 1994.
Shindell, D.T., D. Rind, and P. Lonergan, Increased polar stratospheric ozone losses
and delayed eventual recovery owing to increasing greenhouse-gas concentrations.
Nature: 392, 589, 1998.
Shindell, D., D. Rind, N. Balachandran, J. Lean, and P. Lonergan, Solar cycle
variability, ozone, and climate. Science: 284, 305, 1999.
Shine, K.P., On the modelled thermal response of the Antarctic stratosphere to a
depletion of ozone. Geophys Res Lett: 13, 1331, 1986.
Siskind, D.E., G.E. Nedoluha, C.E. Randall, M. Fromm, and J.M. Russell III, An
assessment of Southern Hemisphere stratospheric NOx enhancements due to transport
from the upper atmosphere. Geophys Res Lett: 27,7 329, 2000.
Slusser, J.R., D.J. Fish, E.K. Strong, and R.L. Jones, Five years of NO2 vertical
column measurements at Faraday (65◦ S): Evidence for the hydrolysis of BrONO2 on
Pinatubo aerosols. J Geophys Res: 102, 12,987, 1997.
Smith, J.B., E.J. Hintsa, N.T. Allen, R.M. Stimpfle, and J.G. Anderson, Mechanisms
for mid-latitude ozone loss: Heterogeneous chemistry in the lowermost stratosphere.
J Geophys Res: 106, 1297, 2001.
Solomon, S., The mystery of the Antarctic ozone hole. Rev Geophys: 26, 131, 1988.
Solomon, S., Stratospheric ozone depletion: A review of concepts and history. Rev
Geophys: 37,
7 275, 1999.
Solomon, S., and R.R. Garcia, On the distributions of long-lived tracers and chlorine
species in the middle atmosphere. J Geophys Res: 89, 11,633, 1984.
thermosphere and the lower atmosphere, I. Odd nitrogen from 50 to 120 km. J
Geophys Res: 87,
7 7206, 1982.
Solomon, S., G.C. Reid, D.W. Rusch, and R.J. Thomas, Mesospheric ozone depletion
during the solar proton event of July 13, 1982, Part II. Comparison between theory
and measurements. Geophys Res Lett: 10, 257, 1983.
Solomon, S., R.R. Garcia, F.S. Rowland, and D.J. Wuebbles, On the depletion of
Antarctic ozone. Nature: 321, 755, 1986.
Solomon, S., G.H. Mount, R.W. Sanders, and A.L. Schmeltekopf, Visible spectroscopy
at McMurdo Station, Antarctica, 2. Observations of OClO. J Geophys Res: 92, 8329,
1987.
Solomon, S., G.H. Mount, R.W. Sanders, R.O. Jakoubek, and A.L. Schmeltekopf,
Observations of the nighttime abundance of OClO in the winter stratosphere above
Thule, Greenland. Science: 242, 550, 1988.
Solomon, S., R.W. Sanders, R.R. Garcia, and J.G. Keys, Increased chlorine dioxide
over Antarctica caused by volcanic aerosols from Mount Pinatubo. Nature: 363, 245,
1993.
Solomon, S., R.W. Portmann, R.R. Garcia, L.W. Thomason, L.R. Poole, and M.P.
McCormick, The role of aerosol variations in anthropogenic ozone depletion at
northern midlatitudes. J Geophys Res: 101, 6713, 1996.
Solomon, S., S. Borrmann, R.R. Garcia, R. Portmann, L. Thomason, L.R. Poole, D.
Winker, and M.P. McCormick, Heterogeneous chlorine chemistry in the tropopause
region. J Geophys Res: 102, 21,411, 1997.
Solomon, S., R.W. Portmann, R.R. Garcia, W. Randel, F. Wu, R. Nagatani, J.
Gleason, L. Thomason, L.R. Poole, and M.P. McCormick, Ozone depletion at mid-
latitudes: Coupling of volcanic aerosols and temperature variability to anthropogenic
chlorine. Geophys Res Lett: 25, 1871, 1998.
SPARC (Stratospheric Processes and their Role in Climate), Assessment of Trends in
the vertical distribution of ozone. WMO Report No. 43, Geneva, Switzerland, 1998.
Stachnik, R.A., J.C. Hardy, J.A. Tarsala, J.W. Waters, and N.R. Erickson,
Submillimeter wave heterogyne measurements of stratospheric ClO, HCl, O3 , and
HO2 : First results. Geophys Res Lett: 19, 1931, 1992.
Staehelin, J., A. Renaud, J. Bader, R. McPeters, P. Viatte, B. Hoegger, V. Bugnion,
M. Giroud, and H. Schill, Total ozone series at Arosa (Switzerland): Homogenization
and data comparison. J Geophys Res: 103, 5827, 1998a.
Staehelin, J., R. Kegel, and N.R.P. Harris, Trend analysis of the homogenized total
ozone series of Aroso (Switzerland), 1926-1996. J Geophys Res: 103, 8389-8399,
1998b.
Steele, H.M., P. Hamill, M.P. McCormick, and T.J. Swissler, The formation of polar
stratospheric clouds. J Atmos Sci: 40, 2055, 1983.
Stimpfle, R.M., J.P. Koplow, R.C. Cohen, D.W. Kohn, P.O. Wennberg, D.M. Judah,
D.W. Toohey, L.M. Avallone, J.G. Anderson, R.J. Salawitch, E.L. Woodgridge, C.R.
Webster, R.D. May, M.H. Proffitt, K. Aiken, J. Margitan, M. Loewenstein, J.R.
Podolske, L. Pfister, and K.R. Chan, The response of ClO radical concentrations to
variations in NO2 radical concentrations in the lower stratosphere. Geophys Res Lett:
21, 2543, 1994.
Can J Chem: 52, 1610, 1974.
Stolarski, R.S., A.J. Krueger, M.R. Schoeberl, R.D. McPeters, P.A. Newman, and J.C.
Alpert, Nimbus 7 satellite measurements of the springtime Antarctic ozone decrease.
Nature: 322, 808, 1986.
Stolarski, R.S., P. Bloomfield, R.D. McPeters, and J.R. Herman, Total ozone trends
deduced from Nimbus 7 TOMS data. Geophys Res Lett: 18, 1015, 1991.
Swider, W., and T. Keneshea, Decrease of ozone and atomic oxygen in the lower
mesosphere during a PCA event. Planet Space Sci: 21, 1969, 1973.
Sze, N.D., M.K.W. Ko, D.K. Weisenstein, J.M. Rodriguez, R.S. Stolarski, and M.R.
Schoeberl, Antarctic ozone hole: Possible implications for ozone trends in the southern
hemisphere. J Geophys Res: 94, 11,521, 1989.
Tabazadeh, A., and R.P. Turco, Stratospheric chlorine injection by volcanic eruptions:
HCl scavenging and implications for ozone. Science: 260, 1082, 1993.
Tabazadeh, A., and O.B. Toon, The role of ammoniated aerosols in cirrus cloud
nucleation. Geophys Res Lett: 25, 1379, 1998.
Tabazadeh, A., E.J. Jensen, O.B. Toon, K. Drdla, and M.R. Schoeberl, Role of the
stratospheric polar freezing belt in denitrification. Science: 291, 2591, 2001.
Thomas, R.J., Mesospheric ozone depletion during the solar proton event of July 13,
1982, Part I. Measurement. Geophys Res Lett: 10, 253, 1983.
Thomason, L.W., L.R. Poole, and T. Deshler, A global climatology of stratospheric
aerosol surface area density deduced from Stratospheric Aerosol and Gas Experi-
ment II measurements: 1984-1994. J Geophys Res: 102, 8967, 1997.
Thompson, D.J., and J.M. Wallace, Annular modes in the extratropical circulation,
Part I: Month to month variability. J Clim: 13, 1000, 2000.
Thompson, D.J., J.M. Wallace, and G.C. Hegerl, Annular modes in the extratropical
circulation, Part II: Trends. J Clim: 13, 1018, 2000.
Tie, X.X., and G. Brasseur, The response of stratospheric ozone to volcanic eruptions:
Sensitivity to atmospheric chlorine loading. Geophys Res Lett: 22, 3035, 1995.
Tie, X.X., G.P. Brasseur, B. Briegleb, and C. Granier, Two-dimensional simulation
of Pinatubo aerosol and its effect on stratospheric ozone. J Geophys Res: 99, 20,545,
1994.
Tie, X.X., C. Granier, W. Randel, and G.P. Brasseur, Effects of interannual variation
of temperature on heterogeneous reactions and stratospheric ozone. J Geophys Res:
102, 23,519, 1997.
Tolbert, M.A., M.J. Rossi, and D.M. Golden, Heterogeneous interactions of chlorine
nitrate hydrogen chloride and nitric acid with sulfuric acid surfaces at stratospheric
temperatures. Geophys Res Lett: 15, 847, 1988.
Toohey, D.W., A critical review of stratospheric chemistry research in the U.S.: 1991-
1994. Rev Geophys Suppl, U.S. National Report to International Union of Geodesey
and Geophysics 1991-1994, 759, 1995.
Toohey, D.W., J.G. Anderson, W.H. Brune, and K.R. Chan., In situ measurements
of BrO in the Arctic stratosphere. Geophys Res Lett: 17,
7 513, 1990.
Toohey, D.W., L.M. Avallone, L.R. Lait, P.A. Newman, M.R. Schoenberl, D.W.
Fahey, E.L. Woodbridge, and J.G. Anderson, The seasonal evolution of reactive
chlorine in the northern hemisphere stratosphere. Science: 261, 1134, 1993.
Toon, G.C., C.B. Farmer, L.L. Lowes, P.W. Schaper, J.-F. Blavier, and R.H. Norton,
Infrared aircraft measurements of stratospheric composition over Antarctica during
September 1987. J Geophys Res: 94, 16,571, 1989.
Toon, G.C., J.-F. Blavier, and J.T. Szeto, Latitude variations of stratospheric trace
gases. Geophys Res Lett: 21, 2599, 1994.
Toon, O.B., P. Hamill, R.P. Turco, and J. Pinto, Condensation of HNO3 and HCl in
the winter polar stratosphere. Geophys Res Lett: 13, 1284, 1986.
Toon, O.B., E.V. Browell, S. Kinne, and J. Jordan, An analysis of lidar observations
of polar stratopheric clouds. Geophys Res Lett: 17,
7 393, 1990.
Toumi, R., S. Bekki, and R. Cox, A model study of ATMOS observations and the
heterogeneous loss of N2 O5 by the sulphate aerosol layer. J Atmos Chem: 16, 135,
1993.
Traub, W.A., K.W. Jucks, D.G. Johnson, and K.V. Chance, Chemical change in the
Arctic vortex during AASE II. Geophys Res Lett: 21, 2595, 1994.
Tuck, A.F., Synoptic and chemical evolution of the Antarctic vortex in late winter
and early spring, 1987. J Geophys Res: 94, 11,687, 1989.
Tuck, A.F., and M.H. Proffitt, Comment on ”On the magnitude of transport out of
the Antarctic polar vortex” by W.M.F. Wauben et al., J Geophys Res: 102, 28,215,
1997.
Tuck, A.F, T. Davies, S.J. Hovde, M. Noguer-Alba, D.W. Fahey, S.R. Kawa, K.K.
Kelly, D.M. Murphy, M.H. Proffitt, J.J. Margitan, M. Loewenstein, J.R. Podolske,
S.E. Strahan, and K.R. Chan, Polar stratospheric cloud processed air and potential
vorticity in the northern hemisphere lower stratosphere at mid-latitudes during
winter. J Geophys Res: 97,
7 7883, 1992.
Tuck, A.F., D. Baumgardner, K.R. Chan, J.E. Dye, J.W. Elkins, S.J. Hovde, K.K.
Kelly, M. Loewenstein, J.J. Margitan, R.D. May, J.R. Podolske, M.H. Proffitt,
K.H. Rosenlof, W.L. Smith, C.R. Webster, and J.C. Wilson, The Brewer-Dobson
circulation in the light of high altitude in-situ aircraft observations. Quart J Roy
Meteorol Soc: 123, 1, 1997.
Tung, K.K., M.K.W. Ko, J.M. Rodriguez, and N.D. Sze, Are Antarctic ozone
variations a manifestation of dynamics or chemistry? Nature: 333, 811, 1986.
Turco, R.P., and P. Hamill, Supercooled sulfuric acid droplets: Perturbed
stratospheric chemistry in early winter. Ber Bunsenges Phys Chem: 96, 323, 1992.
Van Roozendael, M., C. Fayt, D. Bolse, P.C. Simon, M. Gil., M. Yela, and J. Cacho,
Ground-based stratospheric NO2 monitoring at Keflavik (Iceland) during EASOE.
Von Clarmann, T., H. Fischer, F. Friedl-Vallon, A. Linden, H. Oelhof, C. Piesch, M.

Seefeldner, and W. Volker,
¨ Retrieval of stratospheric O3 , HNO3 and ClONO2 profiles
from 1992 MIPAS-B limb emission spectra. J Geophys Res: 98, 20,495, 1993.
Voemel, H., S.J. Oltmans, D.J. Hofmann, T. Deshler, and J.M. Rosen, The evolution
of the dehydration in the Antarctic stratospheric vortex. J Geophys Res: 100, 13,919,
1995.
Von der Gathen, P., M. Rex, N.R.P. Harris, D. Lucic, B.M. Knudsen, G.O. Braathen,
H. de Backer, R. Fabian, H. Fast, M. Gil, E. Kyrö,
¨ I.S. Mikkelsen, M. Rummukainen,
J. Stahelin,
¨ and C. Varotsos, Observational evidence for chemical ozone depletion
over the Arctic in winter 1991-2. Nature: 375, 131, 1995.
Wahner, A., R.O. Jakoubek, G.H. Mount, A.R. Ravishankara, and A.L. Schmeltekopf,
Remote sensing observations of nighttime OClO column during the airborne Antarctic
ozone experiment, September 8, 1987. J Geophys Res: 94, 11,405, 1989.
Wahner, A., J. Callies, H-P Dorn, U. Platt, and C. Schiller, Near UV atomospheric
absorption measurements of column abundances during airborne Arctic stratospheric
expeditions, January-February 1989: 1. Technique and NO2 observations. Geophys
Res Lett: 17,
7 497, 1990.
Waters, J.W., J.C. Hardy, R.F. Jarnot, and H.M. Pickett, Chlorine monoxide radical,
ozone, and hydrogen peroxide: Stratospheric measurements by microwave limb
sounding. Science: 214, 61, 1981.
Waters, J.W., L. Froidevaux, W.G. Read, G.L. Manney, L.S. Elson, D.A. Flower,
R.F. Jarnot, and R.S. Harwood, Stratospheric ClO and ozone from the microwave
limb sounder on the upper atmosphere research satellite. Nature: 362, 597, 1993.
Waters, J.W., W.G. Read, L. Froidevaux, R.F. Jarnot, R.E. Cofield, D.A. Flower,
G.K. Lau, H.M. Pickett, M.L. Santee, D.L. Wu, M.A. Boyles, J.R. Burke, R.R. Lay,
M.S. Loo, N.J. Livesey, T.A. Lungu, G.L. Manney, L.L. Nakamura, V.S. Perun, B.P.
Ridenoure, Z. Shippony, P.H. Siegel, R.P. Thurstans, R.S. Harwood, H.C. Pumphrey,
and M.J. Filipiak, The UARS and EOS Microwave Limb Sounder (MLS) experiments.
J Atmos Sci: 56, 194, 1999.
Wauben, W.M.F., R. Bintanja, P.F.J. van Velthoven, and H. Kelder, On the
magnitude of transport out of the Antarctic polar vortex. J Geophys Res: 102,
1229, 1997.
Waugh, D.W., R.A. Plumb, R.J. Atkinson, M.R. Schoeberl, L.R. Lait, P.A. Newman,
M. Loewenstein, D.W. Toohey, L.M. Avallone, C.R. Webster, and R.D. May,
Transport out of the lower stratospheric Arctic vortex by Rossby wave breaking.
J Geophys Res: 99, 1071, 1994.
Waugh, D.W., R.A. Plumb, J.W. Elkins, D.W. Fahey, K.A. Boering, G.S. Dutton,
C.M. Volk, E. Keim, R.S. Gao, B.C. Daube, S.C. Wofsy, M. Loewenstein, J.R.
Podolske, K.R. Chan, M.H. Proffitt, K.K. Kelly, P.A. Newman, and L.R. Lait, Mixing
of polar vortex air into middle latitudes as revealed by tracer-tracer scatterplots.
J Geophys Res: 102, 13,119, 1997.
Webster, C.R., R.D. May, D.W. Toohey, L.M. Avallone, J.G. Anderson, P. Newman,
L. Lait, M.R. Schoeberl, J.W. Elkins,and K.R. Chan, Chlorine chemistry on polar
stratospheric cloud particles in the Arctic winter. Science: 261, 1130, 1993.
Webster, C.R., R.D. May, M. Allen, L. Jaegle, and M. P. McCormick, Balloon profiles
of stratospheric NO2 and HNO3 for testing the heterogeneous hydrolysis of N2 O5 on
sulfate aerosols. Geophys Res Lett: 21, 53, 1994.
Webster, C.R., R.D. May, H.A. Michelsen, D.C. Scott, J.C. Wilson, H.H. Jonsson,
C.A. Brock, J.E. Dye, D. Baumgardner, R.M. Stimpfle, J.P. Koplow, J.J. Margitan,
M.H. Proffitt, L. Jaeglé,
´ R.L. Herman, H. Hu, G.J. Flesch, M. Loewenstein, Evolution
of HCl concentrations in the lower stratosphere from 1991 to 1996 following the
eruption of Mt. Pinatubo. Geophys Res Lett: 25, 995, 1998.
Weeks, C.H., R.S. Cuikay, and J.R. Corbin, Ozone measurements in the mesosphere
during the solar proton event of 2 November 1969. J Atmos Sci: 29, 1138, 1972.
Wennberg, P.O., R.C. Cohen, R.M. Stimpfle, J.P. Koplow, J.G. Anderson, R.J.
Salawitch, D.W. Fahey, E.L. Woodbridge, E.R. Keim, R.S. Gao, C.R. Webster, R.D.
May, D.W. Toohey, L.M. Avallone, M.H. Proffitt, M. Loewenstein, J.R. Podolske,
K.R. Chan,and S.C. Wofsy, Removal of stratospheric O3 by radicals: In situ
measurements of OH, HO2 , NO, NO2 , ClO, and BrO. Science: 266, 398, 1994.
Williams, W.J., J.J. Kosters, and D.G. Murcray, Nitric acid column densities over
Antarctica. J Geophys Res: 87,
7 8976, 1982.
Wilson, J.C., H.H. Jonsson, C.A. Brock, D.W. Toohey, L.M. Avallone, D.
Baumgardner, J.E. Dye, L.R. Poole, D.C. Woods, R.J. DeCoursey, M. Osborn, M.C.
Pitts, K.K. Kelly, K.R. Chan, G.V. Ferry, M. Loewenstein, J.R. Podolske, and A.
Weaver, In situ observations of aerosol and chlorine monoxide after the 1991 eruption
of Mount Pinatubo: Effect of reactions on sulfate aerosol. Science: 261, 1140, 1993.
WMO (World Meteorological Organization), Atmospheric Ozone. Report No. 16,
1985.
1989.
1991.
1994.
Geneva, 1999.
Geneva, 2003.
Wuebbles, D.J., F.M. Luther, and J.E. Penner, Effect of coupled anthropogenic
perturbations on stratospheric ozone. J Geophys Res: 88, 1444, 1983.
Yudin, V.A., M.A. Geller, B.V. Khattatov, A.R. Douglass, M.C. Cerniglia, J.W.
Waters, L.S. Elson, A.E. Roche, and J.M. Russell III, A UARS study of lower
stratospheric polar processing in the early stages of northern and southern winters.
J Geophys Res: 102, 19,137, 1997.
Zander, R., E. Mahieu, M.R. Gunson, M.C. Abrams, A.Y. Chang, M. Abbas, C.
Aellig, A. Engel, A. Goldman, F.W. Irion, N. Kämpfer,¨ H.A. Michelsen, M.J.
Newchurch, C.P. Rinsland, R.J. Salawitch, G.P. Stiller, and G.C. Toon, The 1994
northern mid-latitude budget of stratospheric chlorine derived from ATMOS/ATLAS-
3 observations. Geophys Res Lett: 23, 2357, 1996.
Zerefos, C.S., K. Tourpali, and A.F. Bais, Further studies on possible volcanic signal
to the ozone layer. J Geophys Res: 99, 25,741, 1994.
Zerefos, D.S., K. Tourpali, B.R. Bojkov, D.S. Balis, B. Rognerund, and I.S.A. Isaksen,
Solar activity-total column ozone relationships: Observations and model studies with
heterogeneous chemistry. J Geophys Res: 102, 1561, 1997.
Zhou, S., M.E. Gelman, A.J. Miller, and J.P. McCormack, An inter-hemisphere
comparison of the persistent stratospheric Polar vortex. Geophys Res Lett: 27,
7 1123,
2000.
Zurek, R.W., G.L. Manney, A.J. Miller, M.E. Gelman, and R.M. Nagatani,
Interannual variability of the north polar vortex in the lower stratosphere during
the UARS mission. Geophys Res Lett: 23, 289, 1996.
Chapter 7
THE IONS
7.1 Introduction
At the end of the 19th century, it was suggested that the presence of
an electrically conducting atmospheric layer could explain the observed
diurnal variation in the terrestrial magnetic field. In 1901, Marconi
established the first radiowave transmission between Europe and North
America, and Kennelly and Heaviside independently suggested that
this communication was possible only because of the reflection of radio
signals by a conducting layer near 80 km altitude. Radiowave methods
later led to the first quantitative studies of this layer, through analysis of
emitted signals reflected to the surface. Systematic sounding of this type
showed that the base of the electrically conducting layer was located at
about 50-60 km. Present understanding of chemical processes affecting
ions is due primarily to laboratory investigations and to in situ mass
spectrometric measurements made from balloon and rocket platforms.
Such studies have revealed, for example, that electrons are present in
very small abundances below the stratopause, and that positive and
negative ions dominate the ion composition of the stratosphere while
positive ions and electrons dominate the mesosphere and thermosphere.
Atmospheric gaseous ions not only affect the electrical properties of the
atmospheric medium; they also play an interesting role in the chemistry
of the atmosphere.
The ionosphere is a weakly ionized fluid of net neutral charge.
Atmospheric atoms and molecules can be ionized either by short
wavelength solar radiation (UV and X-rays), or by precipitating
energetic particles, e.g.,
X + hν → X+ + e (7.1)
X + e∗ → X+ + 2e (7.2)
where e∗ represents an energetic electron. These processes are the
starting points for a series of reactions which determine the structure
of the ionosphere. Several types of general reactions of importance
in ion chemistry should be outlined in order to study the ionosphere.
533
For example, different ions may be produced from the initial (also
called primary) particles through charge exchange reactions with neutral
molecules:
X+ + Y → Y+ + X (7.3)
Negative ions can also be formed by electron attachment on neutral
species:
e + Z + M → Z− + M (7.4)
Electrons can be released from negative ions, either by photodetachment
or by collisional detachment:
Z− + hν → Z + e (7.5)
Z− + M → Z + M + e (7.6)
Positively and negatively charged particles can recombine:
X+ + e → neutral products (7.7)
X+ + Y− → neutral products (7.8)
Understanding the formation of the ions requires knowledge of
the spectral distribution of solar radiation at short wavelengths, the
structure of solar and galactic cosmic rays, as well as the chemical
composition of the atmosphere and its physical characteristics such as
pressure, temperature, and transport. The variations of solar activity
must also be considered.
It is customary to divide the ionosphere into a number of characteristic
layers based on the mean vertical profile of electron density.
The D-region is located between about 60 km and 90-95 km. In this
layer, ionization results mostly from photoionization of NO by Lyman
α radiation (121.6 nm). High energy cosmic rays also contribute to the
ionization of O2 and N2 below 70 km. Radio waves are readily absorbed
in the D-region, as will be discussed in Section 7.6.
The E-region extends from about 90-95 km to 130 km. It is produced
by ionization of molecular and atomic oxygen, and molecular nitrogen,
by X-rays and Lyman β radiation (102.6 nm). Although the predomi-
nant ions in this layer are O+ +
2 and NO , relatively large concentrations
of Fe+ , Mg+ , Si+ , and Ca+ are also observed. These latter ions are
produced mainly by meteor ablation between 85 and 130 km.
The F-region begins above 130 km and is sometimes subdivided
into two layers, F1 and F2 . It is primarily produced by ionization of
atomic oxygen and molecular nitrogen by extreme ultraviolet radiation
(9-91 nm). The atomic oxygen ion, O+ , dominates. The electron density
attains its maximum value of about 106 cm−3 in this layer. The F-region
plays an important role in the transmission of certain radio waves, which
THE IONS 535
can be reflected or refracted if their frequency is less than 5-10 MHz, or

transmitted if their frequency is above this limit (television waves).
In the D, E, and F1 layers, the electron concentration is greatest at
maximum solar elevation (local noon). At night, the electrons disappear
almost entirely in the D-region, and their density is reduced by a factor
of 100 in the E- and F1 -regions. In these layers, the lifetime of the ions
is short compared to the transport time scale, and the charged particle
concentrations are therefore determined by a photochemical equilibrium
between production (by ionization processes) and loss (by recombination
of positive ions with electrons or with negative ions). In the F2 -region,
on the other hand, the electron density is no longer due to a simple
equilibrium between ionization and recombination and the effects of
transport by molecular diffusion must be considered, particularly above
300 km. The structure of this part of the ionosphere is complex, and
varies considerably with geomagnetic latitude.
Above the F2 layer, the electron density rapidly decreases. Helium
He+ and then hydrogen H+ ions dominate, and the effect of the magnetic
field on these charged particles becomes more important. As a result,
the particles no longer occur in horizontal layers; rather they tend to be
aligned by the Earth’s magnetic field.
A detailed description of the ionospheric layers is given by Banks
and Kockarts (1973) and, more recently, by Hargreaves (1979; 1992).
Physical and chemical processes occurring in the upper atmosphere are
discussed by Rees (1989) and Kelly (1989).
Electrically charged particles play a substantial role in the middle and
upper atmosphere. First, they are directly involved in the occurrence
of aurora and airglow, since the formation of energetically-excited
molecules or atoms that emit visible radiation are often formed by the
recombination of positive ions and electrons. The formation of polar
mesospheric clouds near 85 km altitude during summer seems to be
facilitated by the presence of condensation nuclei in the form of large
proton hydrates. Ions also have an impact on neutral species. As
already mentioned in Chapter 5, ionic reactions lead to the formation
of odd nitrogen and odd hydrogen compounds with potential impact
on mesospheric and stratospheric ozone, especially during solar events.
Finally, as the fraction of ionized compounds increases with altitude, air
parcels become increasingly affected by electromagnetic forces. The so-
called “ion-drag” process (see Chapter 3) becomes a significant force in
the thermosphere and needs to be taken into account in the calculation
of the winds.
Figure 7.1 shows the shape of the mean vertical distribution of
electrons below 1000 km and indicates the dominant ions in each
layer. In this chapter, we will discuss only the ions in the middle
Figure 7.1. Definition of the ionospheric layers based on the vertical distribution
of electron density. From Banks and Kockarts (1973).
atmosphere, i.e., those observed in the D-and lower E-regions, and in

the stratosphere.
To study the behavior of the ions and electrons in a simple fashion, we
first write the following equations of continuity, assuming only a single
ion of each type:
dn+
= nI − αD n+ ne − αi n+ n− (7.9)
dt
dn−
= ae nna ne− n− (dp + fd n+αi n+ ) (7.10)
dt
dne
= nI − αD n+ ne − ae n2 ne + n− (dp + fd n) (7.11)
dt
where
n = concentration of the neutral species which ionize, attach, and
detach electrons (e.g., NO, OH, etc.)
THE IONS 537
n+ = concentration of the positive ion

n− = concentration of the negative ion
ne = electron concentration
na = concentration of main atmospheric molecules (N2 , O2 )
αD = electron-ion recombination rate
αi = ion-ion recombination rate
ae = electron-neutral attachment rate
dp = negative ion photodetachment rate
fd = negative ion collisional detachment rate
I = ionization frequency (formation of an electron and a positive ion).
Further, electrical neutrality is assumed:
n+ = n− + ne = (1+λ) ne (7.12)
where
λ = n− /ne (7.13)
is the ratio of negative ions to electrons. Expressions (7.9) and (7.10)
can now be transformed (assuming na = n):
dne nI ne dλ
= − (αD + λαi )n2e − (7.14)
dt 1+λ 1 + λ dt
dne ae n2 ne ne dλ
= − ne [dp + fd n+(1 + λ)αi ne ] − (7.15)
dt λ λ dt
eliminating dne /dt, we obtain the ratio, λ:

1 dλ ae n2 nI
= + (dp + fd n) − + (αi − αD ) ne
λ(1 + λ) dt λ (1 + λ)ne
(7.16)
which is a Riccati differential equation. Analysis of approximate
magnitude of the respective terms in Eq. (7.16) shows that it is possible
to neglect the contribution of the ionization and recombination terms,
so that the equation reduces approximately to
dλ
= (1 + λ) ae nna − (1 + λ) λ (dp + fd n) (7.17)
dt
Since the time constant for negative ion formation by attachment (ae ) is
at most about an hour at the top of the middle atmosphere (and is much
less at lower altitudes), photochemical steady state can be assumed:
ae nna
λ= (7.18)
dp + fd n
This equation shows that the ratio of the negative ions to electrons
depends principally on the rate of attachment of electrons on neutral
particles, and on the rate of detachment by solar radiation and collisions.
The continuity equation for electrons in the D-region can be written:
dne nI
= − (αD + λαi ) n2e (7.19a)
dt 1+λ
but, in the E-region, where λ << 1,
dne
= nI − αD n2e (7.19b)
dt
A more realistic analysis considers formation of several ions (j), and
involves more complex expressions of the type
dne nj Ij ne
= − (αD, j + λαi, j )n+
j (7.20)
dt j
1 + λ 1 + λ j
with

nj + = (1 + λ) ne (7.21)
j
An effective recombination rate can then be defined:

(αD, j + λαi, j ) n+
j
j
αeff = + (7.22)
n j
j
such that the equation for electrons becomes
dne nj Ij
= − αeff n2e (7.23)
dt j
1 + λ
Because the lifetime of electrons in this part of the atmosphere is

quite short, we can assume photochemical equilibrium, and the electron
concentration is given by
⎛ ⎞1/2
nj Ij
⎜ j ⎟
ne = ⎝ ⎠ (7.24)
(1 + λ) αeff
Figure 7.2.a/7.2.b presents the shape of the vertical distributions of

αeff and λ , and Figure 7.3 shows some representative observed electron
density profiles in the D- and lower E-regions.
THE IONS 539
Figure 7.2.a. Vertical distribution of the effective recombination coefficient αeff

obtained with a mathematical model. From Brasseur (1982).
Figure 7.2.b. Ratio λ of negative ions to electrons as a function of altitude as

calculated with a mathematical model for low solar activity. From Brasseur
(1982).
7.2 Formation of Ions in the Middle Atmosphere
7.2.1 Effect of Solar Radiation
Examination of the photoionization thresholds and absorption cross

sections for atmospheric constituents shows that the solar radiation
capable of producing ionization is primarily absorbed by O2 , N2 , NO,
and O. Table 7.2 presents the photoionization thresholds for various
Figure 7.3. Vertical distributions of electron density observed during daytime

at Wallops Island, Virginia, during different seasons, for a zenith angle of
60 degrees. Observations in April 1964, June 1965, September 1965, and
December 1965. From Mechtly and Smith (1968).
atmospheric molecules. The photoionization limit of molecular oxygen

is 102.8 nm, implying that oxygen is ionized by the Lyman β and C III
lines and by solar X-rays. On the other hand, molecular nitrogen can
only be ionized by wavelengths less than 79.58 nm, and is therefore
unaffected by the principal ultraviolet lines which are available below
100 km. Richards et al. (1994) present a detailed solar EUV model (5-
105 nm) and the sensitivity of the flux to solar activity. Another detailed
model of the solar flux is provided by Tobiska et al. (2000). Table 7.1
gives some typical values of the incident solar fluxes for different levels
of solar activity.
Given the ionization (σ) and absorption (K) cross sections of N2 , O2 ,
and O for different spectral intervals, the photoionization coefficient can
be calculated using the following expression:

I (X) = σi (X)q∞,i e−τ (X = N2 , O2 or O) (7.25a)
where
τi = Ki,O2 (O2 ) + Ki,N2 (N2 ) + Ki,O (O) (7.25b)
and where the other symbols have been defined in Chapter 4.
Photoionization and photoabsorption cross sections needed to calculate
the ion production using 37 wavelength bins are given by Richards
et al. (1994).
Stewart (1970) suggested analytic expressions which can be used to
obtain the photoionization coefficients of N2 , O2 , and O to within 15%
THE IONS 541
Table 7.1 Flux of ionizing solar radiation (cm−2 s−1 ) for different
levels of solar activity.
Wavelength Very Quiet Sun Moderate Solar High solar

Interval (nm) Quiet Sun Activity Activity
102.6 (L−β) 3.5 (9) 5.0 (9) 8.0 (9) 1.2 (10)
97.7 (C III) 4.4 (9) 5.0 (9) 6.0 (9) 1.0 (10)
91-79.6 7.5 (9) 1.0 (10) 1.3 (10) 1.5 (10)
79.6-73.2 1.0 (9) 1.3 (9) 1.5 (9) 2.0 (9)
73.2-66.5 5.0 (8) 6.0 (8) 8.0 (8) 1.0 (9)
66.5-37.5 4.0 (9) 6.0 (9) 8.0 (9) 1.2 (10)
37.5-27.5 7.5 (9) 1.0 (10) 1.5 (10) 2.0 (10)
27.5-15.0 7.5 (9) 1.0 (10) 1.5 (10) 2.0 (10)
15.0-8.0 5.0 (8) 7.5 (8) 1.0 (9) 2.0 (9)
8.0-6.0 2.5 (7) 5.0 (7) 1.0 (8) 1.5 (8)
6.0-4.1 2.5 (7) 5.0 (7) 1.0 (8) 1.5 (8)
4.1-3.1 7.5 (6) 1.5 (7) 3.0 (7) 4.5 (7)
0.8-0.5 2.9 (2) 2.9 (3) 2.9 (4) 2.9 (5)
0.33-0.5 2.0 (1) 2.0 (2) 2.0 (3) 2.0 (4)
0.15-0.33 1.0 (0) 1.0 (1) 1.0 (2) 1.0 (3)
Table 7.2 Ionization potentials of atoms and molecules (wavelength

in nm).
Al 207.1 Fe 157.1 O3 96.9 CO2 89.9

Na 241.2 NO 134.0 H2 O 98.5 O 91.0
Ca 202.8 NH3 122.1 N2 O 96.1 CO 88.5
Mg 162.2 O2 (1 ∆g) 111.8 CH4 95.4 N 85.2
Si 152.1 O2 102.8 OH 94.0 N2 79.6
C 110.0 SO2 100.8 H 91.1 Ar 78.7
accuracy. Defining the effective column density as:

N∗ (z, χ) = N (N2 , z, χ) + N (O2 , z, χ) + 0.8 × N (O, z, χ) (7.26)
where N (N2 , z, χ) represents the total column abundance of N2 (molec
cm−2 ) along the solar ray path, and N (O2 , z, χ) and N (O, z, χ) are
defined similarly. Then
I−1 ∗
X+ ≈ A0 + A1 (N (z, χ) × 10
−17
) + A2 (N∗ (z, χ) × 10−17 )α (7.27)
Using this expression, the photoionization frequencies of N+ +

2 , O2 ,
+
and O can be estimated for mean solar activity by substituting the
appropriate constants given in Table 7.3 below:
Table 7.3 Constants for Photoionization Frequencies
Constituent Ao A1 A2 α
IN+2 2.2 (6) 2.2 (6) 6.4 (6) 2.0

IO+2 1.1 (6) 1.0 (6) 1.3 (6) 2.5
IO+ 2.3 (6) 3.2 (6) 1.0 (6) 2.4
Figure 7.4 shows the ionization rate for four primary ions (O+ +
2 , N2 ,
O+ , +
and N ) between 90 and 120 km altitude. Note that the N ion is+
produced primarily by dissociative ionization of N2 .

In the D-region, under typical conditions, most of the ionization is due
to the effect of the solar Lyman α ray on nitric oxide (Nicolet, 1945).
The ionization potential of the NO molecule is only 9.25 eV, which
corresponds to a wavelength of 134 nm. In the spectral region of the
Lyman α line an atmospheric window exists due to the low absorption
cross section of O2 in this interval, and thus the ionizing radiation can
Figure 7.4. Production rate by EUV radiation of primary ions in the E-region
of the ionosphere calculated for 50◦ N at local noon during summertime (from
Fritzenwallner, 1997).
THE IONS 543
penetrate relatively far into the mesosphere. The rate of production of

the NO+ ion is given by
P (NO+ ) = INO (NO) (7.28)
and can be estimated by

INO = 6 (1 + R) × 10−7 exp −1 × 10−20 N (O2 , z, χ) (s−1 ) (7.29)
where R is a factor that accounts for solar variability and which ranges
from 0 (solar minimum) to approximately 0.8 (solar maximum). This
expression assumes that the incident flux (q∞ ) of Lyman α is 3(1 + R) ×
1011 cm−2 s−1 , σNO = 2.0 × 10−18 cm−2 , and σO2 = 1.0 × 10−20 cm−2 .
N (O2 , z, χ) represents the integrated oxygen abundance along the line
of sight, as above. It should be noted that even at night, when the
direct incident flux of Lyman α is zero, there is still a considerable
flux of scattered Lyman α reflected by the hydrogen geocorona in the
uppermost levels of the atmosphere. Its intensity is about 100 to
1000 times weaker than the direct daytime flux. It is, however, a major
ion source of the D-region during nighttime. A similar effect occurs for
Lyman β, and is important in the E-region, where it is a major source
of O+2 ions.
The morphology of the D-region is therefore closely related to the
distribution of nitric oxide, which is strongly tied to dynamics (Brasseur
and Nicolet, 1973; Solomon et al., 1982a,b; Brasseur and De Baets,
1986). Nitric oxide is produced in large amounts in the stratosphere
and thermosphere, while in the mesosphere its distribution is heavily
dependent on the competition between transport from these neighboring
regions and destruction by photochemical processes (see Chapter 5).
The absorption of solar radiation in the spectral region from 102.7
[photoionization limit of O2 (1 Σ−g )] to 111.8 nm [photoionization limit
1
of O2 ( ∆g )] represents an additional source of D-region ionization as
suggested by Hunten and McElroy (1968). In this region, however, the
incident radiation is absorbed both by O2 and CO2 , so that the rate of
formation of the O+ 2 ion
P (O+ 1
2 ) = IO2 (1 ∆g ) (O2 ( ∆g )) (7.30)
is much smaller than the rate of ionization of NO. The rate of production
has been determined by Paulsen et al. (1972) using the solar fluxes
reported by Hall and Hinteregger (1970). The following approximate
expression can be applied from 70 to 90 km.

IO2 (1 ∆g ) = 0.549 × 10−9 exp −2.406 × 10−20 N (O2 , z, χ)
(7.31)
−9 −20 −1
+ 2.614 × 10 exp −8.508 × 10 N (O2 , z, χ) (s )
The effect of solar variability is not accounted for in this expression

and needs to be added for detailed calculations.
Hard X-rays (λ < 1 nm) also penetrate into the D-region. The
production of ionization by X-rays is somewhat complex, since it involves
the absorption of an X-ray photon by a neutral particle, leading to
the production of photoelectrons which are sufficiently energetic to
ionize other particles. It thus becomes necessary to distinguish between
primary and secondary ionization. The mean number of secondary ion
pairs produced depends on the energy of the incident photon. For
example, this number has been estimated at about 45 for λ = 0.6 nm,
75 for λ = 0.4 nm, and 165 for λ = 0.2 nm (Nicolet and Aikin, 1960).
For low levels of solar activity, this source of ions is small compared to
that produced by the Lyman α line. On the other hand, its contribution
becomes more important during disturbed conditions: from 0.1 to
0.8 nm, for example, the solar irradiance is about 1000 times more
intense for active conditions, and may increase by an additional factor
of 100 during solar flares.
The ionization rate due to X-rays in the D-regions is thus quite
variable. During solar flares, the X-ray contribution is reflected in
observed changes in ion densities. Because of the relative stability of
the dominant ion sources in this region at other times, the variability
observed in the electron density (especially in the upper part of the
D-region) is believed to be associated with meteorological fluctuations,
including planetary wave propagation.
Soft X-rays (1-10 nm) are only important in the upper part of the
middle atmosphere (z > 85km). Their flux also varies with solar activity.
7.2.2 The Effect of Energetic Particles
7.2.2.1 General
Energetic charged particles can penetrate into the Earth’s atmosphere
and ionize atmospheric species. They propagate along helical
trajectories following the Earth’s geomagnetic field. In polar regions
(magnetic latitude Λ > 75◦ ) the magnetic field lines are open and
cosmic particles (e.g., galactic) can easily enter the atmosphere. The
depth of penetration of charged particles depends on their mass and
energy. When energetic particles collide with neutral gas particles, a
small amount of their energy is converted to X-rays as they are rapidly
decelerated. This process is called the bremstrahlung process. The X-
rays can penetrate deeper in the atmosphere than the original particles.
THE IONS 545
Figure 7.5. Depth of atmospheric penetration for different charged particles

and X-rays as a function of energy. Adapted from Potemra (1974) and Thorne
(1977a).
Figure 7.5 depicts the approximate depth of atmospheric penetration for

different types of charged particles as a function of their energy. This
diagram also shows the altitude of bremsstrahlung X-ray penetration
produced by precipitation of high energy particles. In the auroral
zone (70◦ < Λ < 75◦ ), low energy particles (mostly electrons from
1-10 keV) precipitate into the atmosphere from the magnetospheric
plasmasheet. These particles provide the optical displays known as the
aurora, but do not penetrate much below 100 km. In the subauroral
zone, (Λ ≤ 70◦ ) particles can be accelerated in the radiation belt,
attaining energies as high as a few MeV, and can thus penetrate into the
upper stratosphere. Figure 7.6 shows the configuration of the Earth’s
geomagnetic field and the zones where charged particle precipitation
occurs. It should be noted that high energy particles produce substantial
fluxes of secondary electrons (energies from 10 to 100 eV) that are
responsible for a considerable fraction of the energy transfer from the
primary particles to the atmosphere. Knowledge of both the primary and
secondary electron fluxes as a function of energy (also called the particle
spectrum) and their cross sections for interaction with the principal
atmospheric species is thus an important part of evaluating the impact
of particle precipitation in aeronomic processes.
Figure 7.6. Schematic representation of the structure of the magnetosphere.

From Kelley (1989) and courtesy of J. Roederer.
7.2.2.2 Cosmic Rays

The ionization of atmospheric constituents by galactic cosmic
rays provides the dominant source of ions in the lower mesosphere,
stratosphere, and troposphere. The rate of ion pair production increases
exponentially with penetration into the atmosphere, i.e., in proportion
to the atmospheric density (see Figure 7.7) to reach a maximum near 10
to 15 km (Figure 7.8). The production is larger at high latitude than in
the tropics, and is modulated by solar activity (see Eqs. (7.32)-(7.35),
below). Galactic cosmic rays are produced outside the solar system,
and consist mostly of protons (about 83%) and α particles (about 12%)
with energy in the MeV to hundreds of MeV range. They tend to follow
the magnetic field lines as they approach the Earth, and to penetrate
near the magnetic poles. This is the primary reason for the observed
meridional gradient in cosmic ray fluxes, particularly for less energetic
particles which are more strongly influenced by the magnetic field.
During periods of high solar activity, galactic cosmic rays tend to be
removed from the solar system by the intense solar wind. The amplitude
of cosmic ray precipitation thus decreases, as does the associated ion pair
production. The reverse is true at solar minimum.
Heaps (1978a) derived a convenient parameterization of the rate of
ion pair production by cosmic rays QCR (cm−3 s−1 ) as a function of
THE IONS 547
Figure 7.7. Rate of ionization provided by cosmic rays at different geomagnetic

latitudes (0, 50, and 70◦ ) and for minimum and maximum levels of solar activity.
These values are compared to the ion pair production producted by direct and
diffuse Lyman α. From Rosenberg and Lanzerotti (1979).
Figure 7.8. Rate of ionization by cosmic rays in the lower stratosphere and
troposphere. From Brasseur and Nicolet (1973).
latitude, altitude, and solar activity level. For latitudes less than 53◦ ,
the following expressions can be used:
If
X1 = 1.74 × 10−18 (7.32a)
X2 = 1.93 × 10−17 (solar max), 2.84 × 10−17 (solar min) (7.32b)
X3 = 0.6 + 0.8| cos φ| (7.32c)
then, for total number density (M) greater than 3 × 1017 molec cm−3 ,
(1−X3 )
QCR = X1 + X2 | sin φ|4 3 × 1017 (M)X3 (7.33a)
while for (M) less than 3 × 1017 cm−3 ,

QCR = X1 + X2 | sin φ|4 (M) (7.33b)
where φ is the latitude. For latitudes greater than 53◦ (polar caps)
QCR = X1 (M) (solar mean) (7.34a)
QCR = (X1 + X2 )(M) (solar min) (7.34b)
QCR = (X1 − X2 )(M) (solar max) (7.34c)
with
X1 = 1.44 × 10−17 ; X2 = 4.92 × 10−18 (7.35)
The impact of cosmic rays on the chemical composition of the
atmosphere has been assessed, for example, by Jackman et al. (1993).
7.2.2.3 Magnetospheric Electrons and X-rays Produced by

Bremsstrahlung
Electron precipitation from the Earth’s radiation belt is sporadically
observed to produce ionization at high latitudes. As we have already
mentioned, the penetration depth of these electrons depends directly
on their energy and knowledge of their corresponding spectrum is thus
necessary. The height of maximum ionization is typically 120 km for
electron energy of 10 keV, 80 km for 100 keV and 70 km for 500 keV.
Significant fluxes at energies exceeding 100 keV are often observed, and
their precipitation is generally associated with geomagnetic storms in the
subauroral radiation belt. The induced ionization events are generally
referred to as relativistic electron precipitation (REP). The lifetime of
these electrons is short (a few hours) and the frequency of REP events
is not well known. Thorne (1977b) estimated that this phenomenon
probably occurs between 1 and 10 percent of the time. Figure 7.9
shows the ionization rate above 60 km altitude due to monoenergetic
electrons at various initial energies for an electron flux of 1 cm−2 s−1 .
THE IONS 549
Figure 7.9. Ion production rate (cm−3 s−1 ) as a function of altitude due to
monoenergetic particles at various initial energies. Production rate due to a
normalized flux of 1 electron cm−2 s−1 at energies ranging from 1 to 300 keV
(adapted from Rees, 1963).
Realistic vertical distributions of ion production rates for a given (e.g.,

observed) energy spectrum of electron precipitation is obtained from
the convolution of the monoenergetic ionization rate by this energy
spectrum. Figure 7.10 shows an estimate of the average number of ions
formed per year by relativistic electron and cosmic ray deposition in the
middle atmosphere. The effect of associated bremsstrahlung (X-rays) is
also indicated. In any case, in the subauroral zone, this mechanism of
ion pair production is of considerable importance. It should be noted,
however, that the belt only extends over about 7 percent of the globe.
The effect of this process on ionospheric structure and trace neutral
species distributions is not yet well established (see e.g., Callis et al.,
1991).
7.2.2.4 Solar Proton Events

During large solar flares, heavy particles (mostly protons of energies
from 10-500 MeV) can be emitted from the surface of the Sun in sub-
stantial amounts. These particles can produce intense ionization in the
Earth’s D-region, primarily at high latitudes (polar cap). High energy
particles also penetrate below the stratopause (E > 10 MeV) and can
even reach the tropopause (E ≥ 500 MeV). Although this phenomenon
Rate of ion pair production by magnetospheric electrons (REP)

in the subauroral zone. These values correspond to an annual average, and
are compared to the effect of cosmic rays and the total ion pair production
associated with the solar proton event (SPE) of August 1972. From Thorne
(1977a).
is not frequent, it often lasts for several days. These events can change
the ionization rate from its normal value of about 10 cm−3 s−1 to 104
or 105 cm−3 s−1 . As we shall see below, these events can also alter the
composition of the neutral atmosphere (see e.g., Jackman et al., 1990;
1995; Reid et al., 1991). Figure 7.11 presents the vertical distribution of
the ion pair production rates associated with some of the solar proton
events of recent years. The events of August 1972, July 1982, October
1989, and October 2003 were particularly intense.
Figure 7.11. Ionization rates associated with some solar proton events,
compared to that due to cosmic rays. Adapted from Solomon et al. (1983).
THE IONS 551
7.2.2.5 Ionization Rate Calculation

The calculation of the rate of ion pair production due to corpuscular
radiation requires knowledge of the particle energy spectrum and the
energy degradation rate in the atmosphere. If j(z,E) dE represents the
flux of ionizing particles at altitude z of energies from E to E + dE (the
differential flux j(E) is expressed in cm−2 s−1 eV−1 sr−1 ) and if dE/ds is
the energy loss per particle in an inelastic collision, then the ionization
rate at altitude z is given by (Dubach and Barker, 1971):

ρ(z) dE
Q(z) = j (z, E) dE dΩ (7.36)
W ds
E Ω
where W is the mean energy required per ion pair formation (W ≈
35 eV), ρ (z) is the air density (g cm−3 ), and Ω is the solid angle over
which Eq. (7.36) is integrated.
The ionization rates of the major species, molecular nitrogen and
oxygen, can be determined from the rate of total ion pair production,
Q, if it is assumed that the fraction of ionization is proportional to the
mass of the target particle (Rusch et al., 1981):
.88 (N2 )
QN2 = Q (7.37a)
.88 (N2 ) + (O2 )
(O2 )
QO2 = Q (7.37b)
.88 (N2 ) + (O2 )
The ion chemistry is initiated by the impact of energetic electrons,
e∗ , on the nitrogen molecule:
e∗ + N2 → N+
2 + 2e (7.38a)
e∗ + N2 → N + N+ + 2e (7.38b)
e∗ + N2 → N + N + e (7.38c)
and on the oxygen molecule:
e∗ + O2 → O+
2 + 2e (7.39a)
e∗ + O2 → O + O+ + 2e (7.39b)
The ratio of the peak cross section for simple ionization versus that for
dissociative ionization is 0.76:0.24 for N2 (Eqs. (7.38a,b)), and 0.67:0.33
for O2 (Eqs. (7.39a,b)) (Rapp et al., 1965). Assuming that the rate of
production is proportional to these cross section ratios, we obtain the
following approximate expressions (Rusch et al., 1981):
PN+2 = 0.76 × 0.77Q = 0.585Q (7.40a)
PN+ = 0.24 × 0.77Q = 0.185Q (7.40b)

PO+2 = 0.67 × 0.23Q = 0.154Q (7.40c)
PO+ = 0.33 × 0.23Q = 0.076Q (7.40d)

It should also be noted that the reactions discussed above can also lead
to the formation of neutral oxygen and nitrogen atoms, thus initiating
processes which can affect the neutral chemistry. These will be discussed
in Section 7.5.
7.2.3 Comparison of Different Ionization Processes
Summarizing some of the statements made previously, we note that

the ionization above about 90 km is due primarily to the effect of
solar ultraviolet radiation (Lyman β and extreme UV) and X-rays on
the major species, particularly molecular oxygen. From 60 to 90 km,
ionization is produced mainly through the effect of solar Lyman α
radiation on nitric oxide, with photoionization of O2 (1 ∆g ) and hard
X-rays (λ < 1 nm) playing a secondary role under normal conditions.
The effect of cosmic rays dominates below about 65 km. Figure 7.12
compares the contributions of each of these terms as a function of
altitude during the day (upper panel) and the night (lower panel), for
mean solar conditions.
In the absence of solar radiation, the ionization is due primarily to
the precipitation of high energy particles (e.g., see Vampola and Gorney,
1983), diffuse Lyman α radiation and cosmic rays. The effect of X-rays
emitted by certain stars (Sco XR1, GX333-2.5) plays only a minor role.
As we have already mentioned, sudden perturbations on the Sun
(increases in X-ray production, precipitation of energetic protons or
electrons) can greatly modify the ionization rate and the morphology
of the ionosphere. Table 7.4 presents order of magnitude estimates of
the energy penetrating the middle atmosphere, during both quiet and
disturbed periods.
7.3 Positive Ion Chemistry
The chemistry of positive ions in the middle atmosphere is relatively

well understood. Ions such as O+ +
2 and NO , which dominate in
the lower thermosphere, are rapidly lost below 80 km in a set of
clustering reactions ending with stable proton hydrates of the type
H+ (H2 O)n . The hydration order n depends on temperature and water
vapor concentration. In the stratosphere, water ligands are partly or
totally replaced by other molecules such as methyl cyanide (CH3 CN),
whose proton affinity is larger than that of water molecules.
THE IONS 553
Figure 7.12. Ionization rates produced by various sources in the region from
50 to 100 km at noon (50◦ ) (upper panel) and during the night (50◦ ) (lower
panel) in January with mean solar activity. From Fritzenwallner (1997).
Table 7.4 Sources of ionization in the middle atmosphere. (From

Rosenberg and Lanzerotti, 1979).
Permanent sources Flux (ergs cm−2 s−1 )
Galactic cosmic rays 1 (–3) to 1 (–2)

Cosmic X-rays: 0.1 to 1 nm 4 (–9)
Solar X-rays: weak activity, λ 1 nm 1 (–3) to 1 (–1)
Solar X-rays: weak activity, λ 1-10 nm 1 (–1) to 1
Solar H Lyman α: direct 6
Solar H Lyman α: scattered by the
geocorona 6 (–3) to 6 (–2)
Magnetospheric electrons: auroral zones 1 (–1) to 1
Magnetospheric electrons: mid-latitudes 1 (–4) to 1 (–3)
Sporadic sources
Solar cosmic rays (SPE) 1 (–3) to 1 (50 on 8/2/72)

Solar X-rays: solar flares <1 nm <3
Solar X-rays: solar flares 1-10 nm <35
Cosmic X-rays: SCO X-1 0.1-1 nm 4 (–7)
Magnetospheric electrons: auroral zones 1 to 1 (3)
Magnetospheric electrons: mid-latitudes 1 (–3) to 1 (–2)
7.3.1 Positive Ions in the E-Region
The primary O+ 2 ion is present in large abundances because its

destruction rate by dissociative recombination
(αO+2 ); O+
2 +e→O+O (7.41)
and by charge exchange with NO
(γ5 ); O+
2 + NO → NO + O2
+
(7.42)
is relatively slow. In contrast, the N+
2 ion is very short lived because of
rapid charge exchange reaction with oxygen:
(γ9 ); N+ +
2 + O2 → N2 + O2 (7.43)
which provides an additional source of O+ +
2 . N2 also reacts with atomic
oxygen, which is present in relatively large amounts in the E-region:
(γ3 ); N+
2 + O → NO + N
+
(7.44)
THE IONS 555
to produce NO+ . The rates of reactions (7.43) and (7.44) are

comparable, implying that the relative importance of each process is
dependent on the abundances of O2 and O. Below about 95-100 km,
Reaction (7.43) dominates. The dissociative recombination of NO+ must
also be considered:
(αNO+ ); NO+ + e → N (4 S, 25%;2 D, 75%) + O (7.45)
The equations for the ionic constituents of the E-region can now be
written. Neglecting minor reactions and eliminating the density of N+
2,
we can write
IN+2 (N2 ) + IO+2 (O2 )
(O+
2)= (7.46)
αO+2 (ne ) + γ5 (NO)
and
γ3 (O)
IN+2 (N2 ) γ3 (O)+γ 9 (O2 )
+ γ5 (NO)(O+
2)
(NO+ ) = (7.47)
αNO+ (ne )
The equation of electrical neutrality must also be considered in
ionospheric models. In the E-region, we may write:
(ne ) ≈ (O+ +
2 ) + (NO ) (7.48)
Figure 7.13 presents a vertical distribution of O+ +
2 and NO as observed
by Keneshea et al. (1970). Figure 7.14 presents a schematic diagram
of E-region ion chemistry. For completeness, it should be noted that
the O+ produced by photoionization is destroyed rapidly by reaction
with O2 , N2 , and CO2 , and thus plays a negligible role in the middle
atmosphere. Its maximum density reaches only about 10 percent of that
of O+2 . All of these polyatomic ions are produced directly by ionization
or charge transfer with neutral species, and are referred to as molecular
ions.
Gas phase metal ions have been observed at mid-latitudes by several
investigators and appear to be related to the layers of metallic atoms
present in the upper mesosphere and lower thermosphere. Meteoric
ablation is believed to be the dominant source of these metals (Plane,
1991). The presence of atomic sodium (Na) was first detected from its
noctural emission by Slipher (1929) and is discussed by Chapman (1939).
In recent decades several other metallic atoms (e.g., K, Li, Fe, Ca, etc.)
and ions (e.g., Mg+ , Fe+ , Al+ , Na+ , Ca+ ) have been identified. These
species are present in permanent layers from 85 to 110 km, which have
been studied in detail using resonance fluorescence lidar techniques (see
e.g., Granier et al., 1989; Bills and Gardner, 1990; Qian and Gardner,
1995).
Figure 7.13. Observations of the vertical distributions of O+ +

2 and NO . From
Keneshea et al. (1970).
Figure 7.14. Schematic diagram of E-region ion chemistry.
In addition, at altitudes near 100-105 km, thin layers (1-3 km thick) of

greatly enhanced metal ion densities are found to extend over horizontal
areas of a few hundred kilometers. This phenomenon is called a sporadic
E layer, and arises primarily from the effects of wind shears in the
presence of a magnetic field, although other mechanisms have also been
THE IONS 557
Figure 7.15. Evolution between 21:30 and 23:00 of the Ca+ density (expressed
in number of ions per cubic centimeter) measured by resonant lidar technique
between 92 and 106 km on October 27, 1983. The time integration for the lidar
observation is 8 minutes. From Granier et al.(1989).
proposed. The electron density in these layers is considerably larger

than in the nearby regions, leading to enhancement of recombination
of O+ +
2 and NO , reducing their densities relative to those of the metal
ions, which have extremely low recombination coefficients. Figure 7.15
shows the evolution of the ionic calcium density (Ca+ ) measured in
southern France by Granier et al. (1989) over a period of 90 minutes.
The permanent layer (93-96 km) and a sporadic layer (101-104 km), as
detected by the resonant lidar method, are visible.
Part of the ionization of the metallic neutral species X may occur
during ablation, but most of the ions are probably produced by
photoionization
X + hν → X+ + e− (m1 )
and by charge transfer with O+

2
X + O+
2 → X + O2
+
(m2 )
and with NO+
X + NO+ → X+ + NO (m3 )
Similar reactions can, in principle, apply to neutral oxides such as XO

or XO2 , leading to the XO+ and XO+ 2 ions. The major chemical paths
leading to the formation of these ions are, however, provided by the
following reactions
X+ + O3 → XO+ + O2 (7.49)
X+ + O2 + M → XO+
2 +M (7.50)
Conversion of XO+ to X+ occurs through reactions with O, O3 , CO and

NO
XO+ + O → X+ + O2
XO+ + O3 → X+ + 2 O2
(7.51)
XO+ + CO → X+ + CO2
XO+ + NO → X+ + NO2
In addition
XO+
2 + M → X + O2 + M
+
(7.52)
Finally, these ions recombine with electrons
X+ + e− → X + hν
XO+ + e → X + O (7.53)
XO+
2 + e → X + O2
Figure 7.16 provides a schematic representation of the most important

reactions affecting metal ions in the E-region of the ionosphere. Note,
however, that the scheme differs slightly from one metal to another.
For example, Si+ reacts rapidly with water vapor to form the SiOH+
ion. The reaction of this latter ion with ozone leads to the formation of
SiO2 H+ . In the case of sodium (see e.g., Plane et al., 1999), the reaction
of Na+ with ozone is slow, but the following reactions involving water
vapor and carbon dioxide are important
Na+ + N2 + M → Na+ N2 + M
Na+ N2 + CO2 → Na+ CO2 + N2 (7.54)
Na+ CO2 + H2 O → Na+ H2 O + CO2
Additional association reactions could also play a significant role (see

Sze et al., 1982; Jégou et al., 1985a,b; Plane, 1991).
Figure 7.17 shows estimates of the vertical distribution of the ion
densities for several metals, as calculated by the model of Fritzenwallner
(1997; see also Fritzenwallner and Kopp, 1998a).
THE IONS 559
Figure 7.16. Schematic representation of the major chemical processes affecting

metal ions (Fe, Mg, Si) in the lower thermosphere and upper mesosphere.
7.3.2 Positive Ions in the D-Region
In 1945 Nicolet suggested that D-region ionization must be due to

photoionization of nitric oxide by Lyman α (see Section 7.2). The first
theoretical study of the chemistry of this layer by Nicolet and Aikin
(1960) led to the expectation that the D-region should be composed
primarily of O+ 2 and NO .
+ However, the first rocket-borne mass
spectrometric measurements by Narcisi and Bailey (1965) revealed
that in addition to O+ +
2 and NO , other species of mass 19 and 37,
corresponding to the H3 O and H5 O+
+
2 ions (hydrated ions of oxonium,
commonly named proton hydrates), were present in larger abundances
than the primary ions below about 82 km. The presence of relatively
large amounts of heavy ions (mass numbers greater than 45) was also
established by these measurements. Since the early 1970s, many other
“in-situ” measurements of positive ions have confirmed that the transfer
of molecular ions (O+ + +
2 and NO ) to proton hydrates (H (H2 O)n ) is a
characteristic feature of the lower ionosphere. The altitude profiles of D-
region ions observed by the pioneering experiments by Narcisi and Bailey
(1965) are shown in Figure 7.18. The formation and stability of these
proton hydrates (PH) reflects the large proton affinity of H2 O (170 kcal
mole−1 ) and the strong bonding of H2 O to positive ions in general.
Figure 7.17. Vertical profile of the ion concentration ((cm−3 ) calculated for
different metallic species at 50◦ N for the month of January ((noon). ) The
concentration of a corresponding (dominant) neutral species is also shown (from
Fritzenwallner, 1997).
Figure 7.18. First mass spectrometric observations of positive ions from 65 to

90 km (Narcisi and Bailey, 1965).
THE IONS 561
Other cluster ions such as NO+ (H2 O), NO+ (CO2 ), NO+ (H2 O)2 , O+ 4,
O+2 (H2 O), etc., are also observed. The altitude (known as the transition
height) at which the cross over from proton hydrates to molecular ions
(NO+ and O+ 2 ) occurs varies from about 70 to 90 km according to
season and latitude. The hydration order of the ions (n) is dependent
on geophysical conditions (particularly temperature) as well as the
atmospheric water vapor content.
Figure 7.19 shows the density profile of the main ions and ion classes
measured by Kopp and Herrmann (1984) above Red Lake (Ontario,
Canada) on 24 February 1979. As indicated above, NO+ and O+ 2 are
the dominant ions above approximately 80 km, while proton hydrates are
the most abundant positive ions below this altitude. The density profiles
of each individual proton hydrate (hydration orders 1 to 4), measured
during the same experiment, are shown in Figure 7.20.
Figure 7.19. Vertical profile of the main ion densities (cm−3 ) between 60 and
130 km measured by magnetic mass spectrometry above Red Lake (Ontario)
on 24 February 1979. From Kopp and Herrmann (1984).
Vertical profile of the main proton hydrate densities (cm−3 ) in the

D-region of the ionosphere measured above Red Lake (Ontario) on 24 February
1979. From Kopp and Herrmann (1984).
The most abundant ions generally display hydration orders from 2 to

4, but at cold temperatures near the mesopause hydration orders of 8 or
9 are not uncommon. Bjorn ¨ and Arnold (1981) observed ions as large as
+
H (H2 O)20 near a very cold summer mesopause at high latitude. These
authors also suggest that such ions may be important in the formation
of noctilucent clouds.
The transition from a region dominated by NO+ and O+ 2 ions (z >
80 km) to a region dominated by proton hydrates H+ (H2 O)n results
from the fact that three-body reactions affecting O+ +
2 and NO become
faster at lower altitude than two-body reactions or dissociative electron
recombination. The first processes to be considered are the reactions of
primary ions (N+ + + +
2 , O2 , N , O ) with major gases (N2 , O2 ) leading to
+
the formation of O4 and NO+ :
N+ +
2 + O2 → O2 + N2
N+
2 + O → O + N2
+
N+
2 + O → NO + N
+
(7.55)
O+ + O2 → O+
2 +O
O + N2 → NO+ + N
+
O+ +
2 + O2 + M → O4 + M
THE IONS 563
O+ +
4 reacts with atomic oxygen to reform O2 and O3 :
O+ +
4 + O → O2 + O3 (7.56)
and with water vapor to form the first hydrated species:
O+ +
4 + H2 O → O2 · H2 O + O2 (7.57)
The O+
2· H2 O ion is very short lived because of the fast hydration
reactions
O+
2 · H2 O + H2 O → H3 O · OH + O2
+
(7.58)
H3 O+ · OH + H2 O → H+ (H2 O)2 + OH (7.59)
As a result, a proton hydrate of the form H+ ·
(H2 O)n and an OH
radical are produced. Successive hydration can occur through the
following equilibrium reactions:
H+ · (H2 O)n + H2 O + M ⇐⇒ H+ · (H2 O)n+1 + M (7.60)
It should be noted that under high atomic oxygen concentrations, the
efficiency of O+ 2 hydration is reduced as a result of reaction (7.56).
Thus, the rapid decrease in hydrated ions which is observed above
80 km is due at least in part to the rapid increase in the atomic oxygen
abundance which occurs in the lower thermosphere. The reaction scheme
just described does not completely explain the observed D-region ion
distributions, since the primary ion at these altitudes is predominantly
NO+ rather than O+ 2.
Another chain of reactions was found to resolve this discrepancy
(Ferguson, 1971; Reid, 1977), explaining how proton hydrate formation
occurred starting from NO+ as the primary ion. Although the direct
hydration processes
NO+ + H2 O + M ⇐⇒ NO+ · H2 O + M (7.61)
NO+ · (H2 O)n + H2 O + M ⇐⇒ NO+ · (H2 O)n+1 + M (7.62)
were found to be too slow to explain the observations, it was shown that
the equivalent reaction can occur via two indirect sequences involving
CO2 and N2 :
NO+ + CO2 + M ⇐⇒ NO+ · CO2 + M (7.63)
NO+ · CO2 + H2 O → NO+ · H2 O + CO2 (7.64)
and
NO+ + N2 + M ⇐⇒ NO+ · N2 + M (7.65)
NO+ · N2 + CO2 → NO+ · CO2 + N2 (7.66)
followed by
NO+ · CO2 + H2 O → NO+ · H2 O + CO2
Similar processes occur for higher hydration (e.g., involving NO+ · H2 O

· CO2 , see Figure 7.14) until NO+ is converted to NO+ · (H2 O)3 . This
species reacts rapidly with water vapor to produce H3 O+ · (H2 O)2 :
NO+ · (H2 O)3 + H2 O → H+ · (H2 O)3 + HNO2 (7.67)
Higher order clusters can then be formed via reaction (7.60). Most
of the equilibria (e.g., Reactions (7.61), (7.62), (7.63) and (7.65)) are
strongly temperature sensitive, implying that the composition of the
D-region should be quite variable with season and latitude, and that
sporadic changes associated with local temperature variations should be
observed.
Figure 7.21 presents a schematic diagram of the aeronomy of positive
ions in the D-region. Recombination of these ions with negative ions
and electrons must also be considered. The recombination coefficient
of positive ions with electrons should be between about 10−7 and
10−5 cm3 s−1 ; the coefficient is faster for clusters than for molecular ions
such as NO+ and O+ 2 (Leu et al., 1973). Models of positive ions in the
D-region have been presented by Reid (1977), Thomas (1983), Brasseur
and De Baets (1986), Koshelev (1987), Burns et al.(1991), Fritzenwallner
(1997), Kull et al. (1997), and Kazil et al.(2003). Figure 7.22 presents
model distributions of the principal ions calculated at noon between 50
and 120 km for winter conditions (50◦ N) (Fritzenwallner, 1997).
7.3.3 Positive Ions in the Stratosphere
In the stratosphere, ionization is produced by cosmic rays. Many of

the reaction processes are similar to those occurring in the D-region. The
rate of formation of NO+ , however, becomes negligible; the precursor
ions are O+ + +
2 and N2 (which immediately forms O2 by charge exchange
with O2 ). The reaction chain leading to water vapor clusters is, however,
modified by the presence of certain stratospheric neutral species. The
reaction of O+4 with O3 must also be mentioned:
O+ +
4 + O3 → O5 + O2 (7.68)
and this process competes with reaction with H2 O (Reaction 7.57). The
reaction with atomic oxygen is negligible in the stratosphere due to the
reduced abundances of this atom there. O+ 5 is almost as abundant as
O+4 near 25 to 30 km, where ozone densities are greatest. O+
5 reacts
rapidly with H2 O:
O+ +
5 + H2 O → O2 · H2 O + O3 (7.69)
THE IONS 565
Figure 7.21. Schematic diagram of D-region positive ion chemistry. From

Ferguson (1979).
and the ozone molecule destroyed by reaction (7.68) is thus reproduced.

The formation chain leading to proton hydrates (PH) can then continue
as in the D-region (see Figure 7.23).
The presence of stratospheric species such as HNO3 and N2 O5 can
alter the chain through reactions of the form:
Figure 7.22. Vertical distributions of selected positive ion densities between 50

and 120 km altitude at 50◦ N calculated for noon conditions in winter. From
Fritzenwallner (1997)
H3 O+ + N2 O5 → NO+
2 · H2 O + HNO3 (7.70)
H3 O+ + HNO3 → NO+
2 · H2 O + H2 O (7.71)
followed by
NO+ +
2 · H2 O + H2 O + M → NO2 · (H2 O)2 + M (7.72)
NO+
2 · (H2 O)2 + H2 O → H · (H2 O)2 + HNO3
+
(7.73)
These processes accelerate the conversion of the first hydrated water
cluster ion to the second. It should be noted that Reaction (7.70)
converts N2 O5 to HNO3 . Bohringer
¨ et al. (1983) have shown that this
process is probably not fast enough to be important in stratospheric
neutral chemistry, however.
THE IONS 567
Figure 7.23. Schematic of the aeronomic reactions of positive ions in the

stratosphere. From Ferguson (1979).
The first detection of stratospheric positive ions by rocket borne mass

spectrometer (Arnold et al., 1977) showed that above 45 km the most
abundant ions were proton hydrates (H+ · (H2 O)n ) while below that
altitude non-proton hydrates (NPH) of masses 29 ± 2, 42 ± 2, 60 ± 2,
and 80 ± 2 were dominant. Arijs et al. (1978) also observed ions with a
mass number of 96 ± 2.
Arnold et al. (1977) proposed that the non-proton hydrates are formed
by reactions of proton hydrates with formaldehyde, which possesses a
large proton affinity, but this hypothesis was rejected when the reaction
of formaldehyde with higher order proton hydrates was found to be
endothermic (Fehsenfeld et al., 1978). Ferguson (1978) suggested that
proton hydrates might react rapidly with NaOH. If sodium (injected
near 90 km by meteor ablation) could reach the stratosphere, it should
be present there as NaOH, a species with a large proton affinity.
Other observations by Arnold et al. (1978) and Arijs et al. (1978; 1980)
confirmed the presence of clusters of the form H+ Xn (H2 O)m , where X
was an unidentified molecule. Arijs et al. (1980) unambiguously showed
that the mass of X was equal to 41, implying that the identity of X
could not be NaOH (mass 40). MgOH (mass 41) cannot account for X
because the isotopic ratio which would be expected to result from the
25 Mg and 26 Mg isotopes was not observed.
Theoretical and experimental observations, as well as laboratory

studies by Bohringer
¨ and Arnold (1981) indicated that the identity
of X is almost certainly methyl cyanide CH3 CN (mass 41 and proton
affinity of 186 kcal mole−1 ). This species has been detected in the
lower atmosphere (where it is produced by forest fires, automobiles and
industrial activities), but the complete budget of its sources and sinks
remains poorly quantified (e.g., Becker and Ionescu, 1982; Snider and
Dawson, 1984; Arnold and Hauck, 1985; Knop and Arnold, 1987a,b;
Hamm et al., 1989; Schneider et al., 1997) (see Chapter 5). Studies
of the fractional ion abundances of proton hydrates and non-proton
hydrates have been used to infer the profile of CH3 CN from about
15 to 45 km (Arnold et al., 1981; Henschen and Arnold, 1981; Arijs
et al., 1982a; Arijs et al., 1983a,b). These profiles have been fitted to
model calculations of the neutral chemistry and transport of CH3 CN
(Brasseur et al., 1983; Beig et al., 1993a). Figure 7.24 presents a
schematic diagram of the key chemical processes believed to affect ion
formation and destruction in the stratosphere, while Figure 7.25 shows
the relative amount of proton hydrates and non-proton hydrates present
in this region of the stratosphere. Using a two-dimensional model,
Beig et al. (1993b) have shown that the level where the NPH/PH
THE IONS 569
Figure 7.24. Schematic diagram of stratospheric positive ion reactions proposed

by Arnold (1980), with X representing a molecule of methyl cyanide (CH3 CN).
concentration ratio is unity varies from approximately 48 km at the

equator to about 30 km at the poles.
Especially in the lower stratosphere, other processes can occur,
including the reaction with ammonia:
H+ · (H2 O)n + NH3 → NH+

4 · (H2 O)m + (n − m) H2 O (7.74)
(where m = 0, 1, 2,...n). This process is quite rapid, particularly

for n = 2, 3, and 4 (Fehsenfeld and Ferguson, 1973). However, the
stratospheric ammonia abundance is small due to its large solubility and
removal in clouds. Reaction (7.74) may be important in the troposphere.
The reaction with methanol (proton affinity = 182.2 kcal mole−1 )
H+ · (H2 O)n + CH3 OH → H+ CH3 OH(H2 O)m + (n − m) H2 O (7.75)
could play a role, but the vertical distribution of methanol is not well
known.
Figure 7.25. Percentage abundance of proton hydrates (PH) and nonproton

hydrates ((NPH)) as a function of altitude at 45◦ N. The full line indicates model
values and dots and triangles are experimental data. From Beig et al. (1993a).
Table 7.5 shows the mass of positive ions measured in the stratosphere
by mass spectrometry and the identification of the corresponding
ions. Comprehensive reviews of the chemistry and observations of
stratospheric positive ions are provided by Arnold (1980), Arijs (1983),
and Arijs (1992). Models of positive ions in the stratosphere have been
developed by Brasseur and Chatel (1983), Arijs and Brasseur (1986),
Beig and Chakrabarty (1988), Beig et al. (1993a), etc.
7.4 Negative Ion Chemistry
Negative ions are observed below typically 70–80 km altitude, where

the air density is sufficiently high to enable the attachment of electrons
to oxygen molecules. A complex set of reactions leads to the formation
THE IONS 571
Table 7.5 Stratospheric positive ion species detected by balloon-

borne mass spectrometers (Arnold, 1986).
Mass Ion Mass Ion

19 H3 O+ 104 ±1 H+ CH3 CN(H2 O)2
23 ± 1 Na+ 109 H+ (H2 O)6
29 ± 2 H+ HCN 110 ±1 H+ CH3 OH(H2 O)2
37 H+ (H2 O)2 114 ±1 H+ CH3 CN(H2 O)4
42 H+ CH3 CN 117 ±1
45 ± 1 H+ HCN·H2 O 119 H+ (CH3 CN)2 (H2 O)2
49 ± 1 H+ CH3 OH·H2 O 121 ± 1
55 H+ (H2 O)3 125 ± 1
58 ± 1 Na+ (H2 O)2 128 ± 1
60 H+ CH3 CH·H2 O 134 ± 2
63 ± 1 H+ HCN(H2 O)2 139 ± 2
67 ± 1 H+ CH3 OH(H2 O)2 142 ± 1 H+ H2 O(CH3 CN)3
73 H+ (H2 O)4 151 ± 2
78 H+ CH3 CN(H2 O)2 158 ± 3
81 ± 1 H+ HCN(H2 O)3 168 ± 3
83 H+ (CH3 CN)2 179 ± 3
86 ± 1 H+ CH3 OH(H2 O)3 182 ± 3
91 H+ (H2 O)5 186 ± 3
96 H+ CH3 CN(H2 O)3 190 ± 3
101 H+ (CH3 CN)2 H2 O 202 ± 3
of ions such as NO− −

3 and CO3 , which can cluster with water molecules.
In the stratosphere, NO3 and HSO−
−
4 core ions with H2 O, HNO3 , and
H2 SO4 as ligands become dominant.
7.4.1 Negative Ions in the D-Region
The first studies of the ionosphere using radiowave techniques

demonstrated that electrons are nearly absent below 65 or 70 km during
the day and 75 or 80 km at night. However, electrical neutrality requires
a balance between charged particles. The observation of positive ions
implies that equal quantities of a negatively charged particle must be
present.
Johnson et al. (1958) performed the first observations of negative ions,
and suggested that the most abundant negative ion in the D-region is
NO− 2 at mass 46. Since that time, only a very few measurements have
been performed. Narcisi et al. (1971) and Arnold et al. (1971) observed
the D-region at night. The presence of a transition zone above which
the concentration of negative ions rapidly decreases was established by
observation but the altitude of this layer seemed to vary considerably
with time (Arnold and Krankowsky, 1977; Arnold, 1980).
This transition zone between electrons and negative ions appears to be
located near 70 km during the day, and 5 to 10 km higher at night. But
the few available observations are not in agreement as to the composition
of the D-region negative ions. Narcisi et al. suggested that the dominant
ions were hydrated clusters, probably of the type NO− 3 · (H2 O)n , but
Arnold et al. (1971) found that over Northern Europe (69◦ N) the most
abundant species were CO− − −
3 , HCO3 and Cl . Figure 7.26 shows some
vertical profiles derived by Arnold et al. (1971). Further observations
are badly needed.
The chemistry of negative ions has been intensively studied in the
laboratory since the 1960s by Ferguson and coworkers. A schematic
diagram of the suggested chemistry resulting from their experiments is
Figure 7.26. Negative ion concentration from 70 to 85 km at Andoya, Norway

(March, 1970). (Arnold et al., 1971).
THE IONS 573
Figure 7.27. Schematic diagram of the negative ion chemistry of the D-region.
From Ferguson (1979).
presented in Figure 7.27. The chain of reactions is initiated by electron

attachment on an oxygen molecule:
e + O2 + M → O−
2 +M (7.76)
The rate of this reaction varies with the atmospheric density (M), leading
to a nearly complete disappearance of free electrons below a certain
altitude. The attached electron can be liberated, however, by several
reactions:
O−
2 + O → O3 + e (7.77)
O−
2 + O2 ( ∆g ) → 2O2 + e
1
(7.78)
O−
2 + hν → O2 + e
These reactions lead to a balance between negative ions and electrons.
The increasing densities of atomic oxygen in the upper mesosphere
result in increased electron abundances there through Reaction (7.77).
The chain leading to stable negative ions occurs through the following
sequence:
O− −
2 + O3 → O3 + O2 (7.79)
O− −
3 + CO2 → CO3 + O2 (7.80)
The CO−
3 ion can also be formed by a second sequence:
O− −
2 + O2 + M → O4 + M (7.81)
O− −
4 + CO2 → CO4 + O2 (7.82)
CO− −
4 + O → CO3 + O2 (7.83)
This last reaction competes with the following processes:
CO− −
4 + O3 → O3 + CO2 + O2 (7.84)
O− −
3 + CO2 → CO3 + O2 (7.85)
Another source of CO−
3 is provided by the binary formation of O−
e− + O3 → O− + O2 (7.86)
followed by a reaction with CO2
O− + CO2 + M → CO−
3 +M (7.87)
The CO−
3 ion is rapidly destroyed by reaction with atomic oxygen
CO− −
3 + O → O2 + CO2 (7.88)
which interrupts the formation of negative ions. However, the sequence
above can also be followed by:
CO− −
3 + NO → NO2 + CO2 (7.89)
NO− −
2 + O3 → NO3 + O2 (7.90)
which leads to the formation of a terminal ion NO−
3. This species is
particularly stable because of its very high electron affinity. If, however,
THE IONS 575
photodetachment of electrons from NO−

3
NO−
3 + hν → NO3 + e (7.91)
takes place at the rate suggested by the measurements of Smith
et al. (1979), the lifetime of NO−3 could be as small as 20 s, and the
concentration of this ion could be considerably smaller than that of
major negative ions such as CO− − −
3 , HCO3 and Cl . Reaction (7.89)
is relatively slow and controls the rate of formation of the NO− 3 ion.
Reaction (7.90) is fast, and thus the NO− 2 ion can only be present in
small quantities.
The OH− ion is produced by the following reactions
O− + CH4 → OH− + CH3
(7.92)
NO−2 + H → OH−
+ NO
and leads to the eventual formation of HCO−
3 , another terminal ion,
through
OH− + CO2 + M → HCO−
3 +M (7.93)
if it is not destroyed by these reactions:
OH− + O → HO2 + e (7.94)
OH− + H → H2 O + e (7.95)
In the presence of hydrogen chloride, Cl− ions are formed as follows:
O− −
2 + HCl → Cl + HO2 (7.96)
O− + HCl → Cl− + OH (7.97)
NO− −
2 + HCl → Cl + HNO2 (7.98)
Cl− is destroyed by the chemical detachment process:
Cl− + H → HCl + e (7.99)
In most cases, Figure 7.27 does not indicate the likely formation of
ion hydrates X− · (H2 O)n , where X− represents, for example, NO− 3,
CHO− 3 , Cl− , OH− , NO− , O− , and O− . The formation of such ions
2 3 2
by reaction with water vapor has some limited implications for the
schematic picture just described (Fehsenfeld and Ferguson, 1974). For
example, as suggested by Figure 7.27, the reaction of CO− 3 · H2 O with
NO tends to accelerate the conversion of CO−3 to NO −
3 . The partitioning
between ions and their hydrates is a strong function of the temperature
and the water vapor concentration. Figure 7.28 shows the relative
amount of the four dominant negative ions between 50 and 100 km
altitude together with the contribution of the first and second hydrates
(n = 1 and 2). This figure shows that, during summer, when the
mesospheric temperature is low (see Chapter 3), a substantial fraction
Figure 7.28. Vertical distribution of the 4 dominant negative ions (0) and of
their single ((1)) and doubled ((2)) hydrated ions in the D-region as modeled by
Fritzenwallner (1997) and Fritzenwallner and Kopp (1998b) at 50◦ (summer).
of the ions are converted to their hydrates near 80 km altitude (67 km

in the case of Cl− ). It should be noted that, as in the case of NO+ ,
the formation of clusters ions with O2 , N2 , CO2 , and H2 O could play
an important part in producing ions of the form X− · (H2 O)n , but these
reactions have not yet been studied in the laboratory.
The eventual destruction of negative ions, particularly the terminal
ions, occurs by recombination with positive ions. The rate of this
reaction has not been measured for all of the species that are actually
present in the atmosphere, but a mean rate for reactions of this type,
i.e.,
(α); X− + Y+ → neutrals
of 6 × 10−8 cm3 s−1 can be adopted (Smith et al., 1976). Note, however,
that this reaction is also believed to exhibit a pressure dependence
which may be important in the lower stratosphere (Smith et al., 1981).
It should be emphasized that several issues related to the vertical
distribution of negative ions in the D-region remain to be solved.
Perhaps the main problem is to explain the strong variability in ion and
electron concentrations as well as in the λ parameter (see Eq. 7.13)
when ion sources are relatively stable (quiet solar conditions). The
variability in ionic composition is probably related to fluctuations in the
THE IONS 577
concentration of neutrals, and hence suggests a strong meteorological

control of the D-region.
Models of negative ions in the D-region have been presented by
Wisemberg and Kockarts (1980), Thomas and Bowman (1985), Brasseur
and De Baets (1986), Burns et al. (1991), Kull et al. (1997), Fritzenwall-
ner (1997), and Fritzenwallner and Kopp (1998b). Figure 7.29 shows
an example of the vertical distributions of negative ions calculated by
Fritzenwallner (1997). As in the case of positive ions, the vertical profile
of individual negative ion species is strongly related to temperature and
minor species concentrations in the mesosphere.
7.4.2 Negative Ions in the Stratosphere
In order to describe the behavior of negative ions in the stratosphere,

it is useful to adopt the D-region scheme just described and consider the
changes which occur as a result of differences in pressure, temperature,
and composition. The chain of reactions leading to NO− 3 from O2
−
can still occur, and is even more efficient because the low densities of
Figure 7.29. Vertical distributions of selected negative ions at 50◦ N calculated

for noon conditions in winter. From Fritzenwallner (1997).
atomic oxygen there decrease the frequency of the interrupting reactions.

Further, additional reactions facilitate the formation of NO−
3:
CO− −
3 + NO2 → NO3 + CO2 (7.100)
This reaction is 20 times faster than (7.89) and becomes dominant in

the stratosphere, where the densities of NO and NO2 are comparable.
This mechanism also avoids the chain breaking reactions presented in
Section 7.4.1. Further, the rapid reactions of O− , O− − −
2 , CO3 , NO2 , and
−
Cl with HNO3 , and the reaction of the latter three species with N2 O5 ,
lead to additional production of NO− 3 . The formation of hydrated ions,
for example by reaction with O− 3 , can reduce the formation rate of NO−3,
however. Figure 7.30 presents a schematic diagram of these processes.
The possibility of destruction of NO− 3 must also be considered. The
hydration of this ion must proceed as in the D-region. Kinetic studies
(Fehsenfeld et al., 1975) showed that HNO3 can substitute for H2 O
in a cluster ion. Ions of the form NO− 3 · (H2 O)n (HNO3 )m should
thus be formed. NO− 3 · (HNO )
3 n reacts rapidly, however, with H2 SO4
(Viggiano et al., 1980), at least for n = 0, 1, or 2. Thus HSO− 4 or
−
a cluster of the form HSO4 · (H2 SO4 )m · (HNO3 )n · (H2 O)p should be
the terminal ion in the stratosphere. Figure 7.31 presents a schematic
diagram of the corresponding reactions proposed by Arijs et al. (1981).
Figure 7.30. Schematic diagram of the negative ion chemistry in the

stratosphere. From Ferguson (1979).
THE IONS 579
Figure 7.31. Schematic diagram of the negative ion chemistry of the

stratosphere suggested by Arijs et al. (1981) based on in-situ observations.
The first observations by Arnold and Henschen (1978), and later by Arijs
et al. (1981) confirmed the presence of some of these cluster ions; the
most abundant species observed at 35 km were HSO− −
4 · (H2 SO4 )3 , HSO4 ·
− −
(HNO3 ), NO3 · (HNO3 )2 , and HSO4 · (H2 SO4 )2 . Figure 7.32 shows the
fractional abundance between 23 and 38 km of major negative cluster ion
families containing NO− −
3 and HSO4 cores (Viggiano and Arnold, 1981).
Studies of the fractional abundance of negative ions can be used to infer
the vertical profile of neutral H2 SO4 in the stratosphere, as discussed
briefly in Chapter 5.
Figure 7.32. Relative abundance of NO− −

3 and HSO4 core ions observed at mid-
latitudes in the stratosphere (September 1980) by Viggiano and Arnold (1981).
Note that the NO− 3 (HNO3 )n ion family dominates below 30 km, while the
HSO− 4 (H 2 SO4 m (HNO3 )n family is dominant above this altitude.
)
Other polar molecules are also likely to be important. Investigations

of minor mass peaks in negative ion spectra (McCrumb and Arnold,
1981; Arijs et al., 1982a,b) have revealed the presence of ions which
may contain other ligands in addition to HNO3 and H2 SO4 , such as
HOCl and HCl. Table 7.6 provides the mass (and the corresponding
identification) of negative ions measured by mass spectroscopy in the
stratosphere.
Models of the positive and negative ion distributions in the
stratosphere were presented by Brasseur and Chatel (1983), Kawamoto
and Ogawa (1984, 1985), and by Beig et al. (1993a,b). Theoretical
studies of both positive and negative ions are hampered by the lack of
laboratory data regarding some of the relevant rate constants.
In conclusion, the chemistry of negative ions in both the stratosphere
and mesosphere remains at present in a primitive state. Systematic
observations as a function of altitude are needed, as well as laboratory
kinetic studies, before a deeper understanding can be acquired.
THE IONS 581
Table 7.6 Negative ion species detected by balloon-borne mass

spectrometers in the stratosphere (Arnold, 1986)
Mass Ion Mass Ion
26 ± 2 CN− 223 ± 1 HSO−4 (HNO3 )2

43 ± 2 CN− H2 O 241 ± 1 HSO−4 (HNO3 )2 H2 O
61 ± 1 CO−
3 , NO3
−
251 ± 1 NO−
3 (HNO3 )3
80 ± 2 CO3 H2 O, NO−
−
3 H2 O 258 ± 1 HSO−4 ·H2 SO4 ·HNO3
97 ± 1 HSO−4 274 ± 1 HSO−4 (HNO3 )2 HClO
109 ± 2 NO−
3 HNO2 HSO−4 ·H2 SO4 ·HSO3
125 NO−
3 HNO3 286 ± 1 HSO−4 (HNO3 )3
133 ± 1 HSO−4 (H2 O)2 293 ± 1 HSO−4 (H2 SO4 )2
HSO−4 HCl 374 ± 1 HSO−4 (H2 SO4 )2 HSO3
143 ±1 NO−
3 HNO3 H2 O 391 ± 1 HSO−4 (H2 SO4 )3
148 ±2 HSO−4 HOCl 409 ± 1 HSO−4 (H2 SO4 )3 H2 O
160 HSO−4 HNO3 427 ± 1 HSO−4 (H2 SO4 )3 (H2 O)2
174 ±1 NO−
3 HNO3 HNO2 454 ± 1 HSO−4 (H2 SO4 )3 HNO3
188 NO−
3 (HNO3 )2 472 ± 1 HSO−4 (H2 SO4 )3 HSO3
195 HSO−4 H2 SO4 489 ± 1 HSO−4 (H2 SO4 )3 HSO3 H2 O
206 ±1 NO−
3 (HNO3 )2 H2 O
7.5 Effect of Ionic Processes on Neutral Constituents
Reactions between ions and neutral species, can, in principle, affect

the budgets of neutral species as mentioned in Chapter 5.
In the stratosphere, the present understanding of the ion composition
and chemistry is not adequate to completely describe the possible
interactions. The ion production by cosmic rays appears, however, to be
too small to modify the neutral composition substantially, except near
the tropopause, where the ionization and decomposition of N2 leads
to a small production of nitrogen oxides (Warneck, 1972; Nicolet and
Peetermans, 1972; Brasseur and Nicolet, 1973).
During intense solar proton events, the number of ion pairs produced
can increase greatly even at relatively low altitudes, perturbing the
neutral atmosphere and leading to a considerable increase in the rate
of production of nitrogen and hydrogen oxides (see e.g., Swider and
Keneshea, 1973; Crutzen et al., 1975; Jackman et al., 1980; 1990; 1995;
Reid et al., 1991).
In principle, ions can provide a loss mechanism for species formed at
ground level which are long lived in the lower atmosphere. For example,
ion reactions with chlorofluorocarbons which would lead to a lifetime less

than a hundred years could be important. However, the most abundant
ions, H+ · (H2 O)n and NH+ 4 · (H2 O)n do not react with CF2 Cl2 and
CFCl3 .
Only the reaction
H3 O+ + CFCl3 → CCl+
3 + HF + H2 O (7.101)
proceeds rapidly, but H3 O+
is not present in large enough amounts to
be important.
In the stratosphere, positive ions could contribute to the conversion
of nitrogen oxides into nitric acid by the following chain:
NO2 + O3 → NO3 + O2
NO3 + NO2 + M → N2 O5 + M
N2 O5 + H3 O+ → HNO3 + NO+
2 · H2 O
The importance of this mechanism is also limited by the abundance of
the H3 O+ ion, and should be compared with that of other homogeneous
and heterogeneous NO2 to HNO3 conversion processes. Another
possible mechanism involved in the conversion of NOx to HNO3 is the
recombination of proton hydrates with NO−3
H+ (H2 O)n + NO−
3 → HNO3 + n(H2 O)
In the thermosphere, the effect of the ions in producing neutral
nitrogen oxides is very important. In addition to photolysis,
(JN2 ); N2 + hν (80 < λ < 125 nm) → 2N (2 D or 4 S) [F0 ]
(7.102)
(where F0 represents the fraction of nitrogen atoms produced in the 2 D
state) ionic reactions can lead to the production of nitrogen atoms. The
reaction of N+2 with atomic oxygen is one such source
(γ3 ); N+
2 + O → NO + N ( D or S)
+ 2 4
[F1 ] (7.103)
as well as recombination:
(αN+2 ); N+
2 + e → 2N ( D or S)
2 4
[F2 ] (7.104)
The recombination of NO+
(αNO+ ); NO+ + e → N (2 D; 4 S) + O [F3 ] (7.105)
and the following reactions must also be considered:
(γ1 ); O+ + N2 → NO+ + N (4 S) (7.106)
(γ12 ); N+ + O2 → O+ 2 4
2 + N ( D; S) [F4 ] (7.107)
+
(γ12 ); N+ + O2 → NO+ + O (7.108)
THE IONS 583
Fast secondary electrons produced by cosmic and solar particles can

dissociate or dissociatively ionize the nitrogen molecule:
N2 + e∗ → 2N (2 D; 4 S) + e [F5 ] (7.109)
N2 + e∗ → N (2 D; 4 S) + N+ + e [F6 ] (7.110)
The fractional production of the nitrogen atoms in each electronic state
is not yet known with accuracy. This parameter represents an extremely
important factor in determining the density of nitric oxide in the E- and
D-regions, as discussed below (see also, Norton and Barth, 1970; Rees
and Roble, 1979; Solomon, 1983).
Almost all of the atoms which are produced in the 2 D state react with
O2 in the middle atmosphere, producing nitric oxide:
(b∗7 ); N (2 D) + O2 → NO + O (7.111)
Deactivation by atomic oxygen and electrons play only a minor role
below 100 km:
N (2 D) + O → N (4 S) + O (7.112)
N(2 D) + e → N (4 S) + e (7.113)
On the other hand, the reaction of the ground state atom with molecular
oxygen is relatively slow:
(b7 ); N (4 S) + O2 → NO + O (7.114)
so that almost all of these nitrogen atoms are destroyed in the lower
thermosphere by the rapid reaction with NO:
(b6 ); N (4 S) + NO → N2 + O (7.115)
Therefore, when dissociation of molecular nitrogen yields one N (4 S)
and one N (2 D) atom in the lower thermosphere, the net odd nitrogen
production is extremely small: almost every N (2 D) atom produces
one NO molecule, but almost every N (4 S) atom immediately destroys
one at these altitudes. Net production is provided only by the very
small fraction of N (4 S) atoms which react with oxygen rather than
with NO. Even very small deviations from a 50%-50% branching ratio
can have large effects on lower thermospheric NO densities, making this
parameter an extremely important one in NOx chemistry. Laboratory
studies by Kley et al. (1997) concerning the state of the product atoms
produced in these reactions show that F3 = 0.76, while studies by Zipf
and McLaughlin (1978) suggest that F5 = 0.5. To model the effects
of these processes, indirect estimates based on airglow studies are often
adopted. Rusch et al. (1981) adopted, for example, F1 = 1, F2 = 0.5,
and F4 = 0. Massie (1979) assumed F1 = 1, F2 = 1, F4 = 1, and
F5 = 0.75. Figure 7.33 shows the chemistry of nitrogen compounds in
Figure 7.33. Chemistry of nitrogen compounds in the thermosphere and the

interactions between neutrals and ions. From Massie (1979).
the thermosphere. We can now evaluate the rates of production PNO ,

PN (4 S) , and PN (2 D) which were referred to in Eqs. (5.161) and (5.162).
PN(2 D) = 2F0 JN2 (N2 ) + F1 γ3 (N+
2 )(O) + 2αN+
2
(N+
2 ) ne F2
(7.116)
+ F3 αNO+ (NO+ ) ne + 2PD F5 + PI F6
PN (4 S) = (1 − F0 )2JN2 (N2 ) + (1 − F1 ) γ3 (N+

2 )(O)
+ 2αN+2 (1 − F2 )(N+
2 ) ne + (1 − F3 )αNO+ (NO ) ne
+
+ γ12 (N+ )(O2 ) + γ1 (O+ )(N2 ) + 2PD (1 − F5 ) + PI (1 − F6 )

(7.117)
PNO = 2b39 (N2 O)(O1 D) + PM (NO) + γ1 (O+ )(N2 ) (7.118)
where PD is the rate of production of nitrogen atoms due to dissociation
of N2 by energetic electrons (7.109), PI is the production due to
THE IONS 585
dissociative ionization (7.110), and PM is the rate of odd nitrogen

production associated with meteors. The rate of production of the
nitrogen atom by dissociative ionization is equal to the rate of production
of N+ by the same process (see Eq. (7.40b)):
PI = PN+ = 0.185Q (7.119)
where Q is the rate of ion pair production. Note that the N+ and NO+
ions formed by these reactions will ultimately yield reactive nitrogen,
either by Reaction (7.107) or by Reaction (7.108) followed by (7.104).
The relative values of the cross section for dissociation (Winters, 1966;
Niehaus, 1967) and dissociative ionization (Rapp et al., 1965) imply a
rate for PD of about 0.85Q (Rusch et al., 1981).
The O+ ion can also produce nitrogen atoms by reaction (7.106) with
nitrogen and dissociative recombination of NO+ . This contribution is
only about 0.03Q in the middle atmosphere, however. Below 90 km
altitude, each ion pair produced during particle precipitation in the
atmosphere (cosmic rays, magnetospheric electrons, and solar proton
events) results in a total gross odd nitrogen production
PNOx = PN(2 D) + PN(4 S) + PNO
of about 1.25Q (see e.g., Jackman et al., 1979; Rusch et al., 1981). The
net production depends on the fractional production of N4 S and N2 D
atoms.
Below the mesopause, where water cluster ions are present, the
ions can contribute to the formation of hydrogen free radicals (see
Figure 7.34). Referring to the reactions in Section 7.3.2, we note that
the O+2 · H2 O ion can lead to the following chain, producing one H and
one OH radical :
O+
2 · H2 O + H2 O → H3 O · OH + O2
+
H3 O+ · OH + e → H + OH + H2 O
Net: H2 O → H + OH
or
O+
2 · H2 O + H2 O → H3 O · OH + O2
+
H3 O+ · OH + H2 O → H3 O+ · H2 O + OH
H3 O+ · H2 O + nH2 O → H3 O+ · (H2 O)n+1
H3 O+ · (H2 O)n+1 + e → H + (n + 2) H2 O
Net: H2 O → H + OH
Figure 7.34. Schematic diagram of the ion chemistry leading to odd hydrogen
production. From Solomon et al. (1981).
Other secondary chains must be considered. The processes and rates

of formation of hydrogen radical are described by Frederick (1976),
Heaps (1978b), Crutzen and Solomon (1980), and Solomon et al. (1981).
Below about 75 km, nearly two odd hydrogen particles are formed
per O+ 2 ionization; above that altitude, odd hydrogen production
depends on the ionization rate and the atomic oxygen and water vapor
densities. Figure 7.34 presents a schematic diagram of the positive ion
chemistry responsible for odd hydrogen production in the D-region of the
ionosphere. Note that odd hydrogen production is much more efficient
if the primary ion is O+ +
2 than if the primary ion is NO (i.e., during
particle precipitation as opposed to quiet times). Figure 7.35 presents
a calculated vertical distribution of the rates of production of HOx and
NOx per ion pair for two values of Q.
7.6 Radio Waves in the Lower Ionosphere
The study of the ionosphere has largely been performed by observation

of the properties of radio wave propagation through this medium. In
this work we will not attempt to present an exhaustive study of the
interaction of electromagnetic waves and a weakly ionized plasma, nor
THE IONS 587
Figure 7.35. Rate of production of nitrogen atoms and odd hydrogen radicals
per ion pair as a function of altitude. From Fritzenwallner (1997).
will we review the many techniques used for such studies. The object of
this section is to indicate the basic principles which underlie studies
of this sort and to present an elementary description of magneto-
ionic theory (Appleton, 1937; Hartree, 1931), and finally to show how
this approach yields information about the properties of the D-region,
particularly in winter.
The conditions of propagation of an electromagnetic wave in the
ionospheric medium are determined by the Maxwell and Lorentz
equations. These equations determine the properties of the electric
(E) and magnetic (H) fields, as well as the displacement (D) and the
induction (B) as a function of the electric charge and current densities
(J) (see, for example, Budden, 1961; Davies, 1965):
∇·D=ρ (7.120)
∇·B=0 (7.121)
∂D
∇×H=J+ (7.122)
∂t
∂B
∇×E=− (7.123)
∂t
with
D = 0 E + P (7.124)
B = µ0 H (7.125)
where P is the polarization vector, 0 is the permitivity of free space, ρ
is the space charge density, and µ0 is the magnetic permeability. The
Lorentz equation describes the flow of charge as a function of the electric
and induction fields. For electrons, one may write
dV
me = −e (E + V × B) − me νc V (7.126)
dt
where me is the mass of the electron, V is its speed, -e is the electrical
charge and νc is the collision frequency between electrons and neutrals.
If one assumes that a plane transverse wave
E = E0 exp i(ωr t − kx) (7.127)
of amplitude Eo and frequency ωr propagates as a function of time t
in the x direction with a propagation factor k = 2π/λ, λ being the
actual wavelength in the medium, then it can be shown (after some
mathematical manipulation), that the index of refraction of the ionized
medium is given by the Appleton formula:
X
n2 = 1 − 4
1/2 (7.128)
YT
1 − iZ − 2 (1−X−iZ)
YT
± 4(1−X−iZ) 2
2 + YL
where X = ne e2 /0 me ωr2 , (a dimensionless parameter proportional to

the electron concentration ne ), YL = eBL /me ωr , and YT = eBT /me ωr
represent the effects of the longitudinal (L) and transverse (T) magnetic
field components and Z = νc /ωr is a dimensionless parameter describing
the effect of electron-neutral collisions. This formula shows that the
refractive index is a complex number when the collision frequency cannot
be neglected compared to the frequency of the wave. The real part of the
refractive index describes the refracting and dispersive properties of the
medium and the imaginary part its absorption properties. The Appleton
formula also shows that if YL or YT are different from zero, two values
can be assigned to n2 , indicating that the presence of a magnetic field
introduces anisotropy into the medium, so that two propagation modes
exist for each frequency (ordinary and extraordinary wave); this is a
phenomenon similar to the optical birefringence of certain crystals.
The most simple form of the Appleton equation is obtained by
neglecting the effects of the magnetic field and collisions. This equation
is adequate to describe the propagation of high frequency waves (above
30 MHz) in the ionosphere, and can be used to obtain an approximate
understanding of the propagation characteristics of lower frequency
waves. In this case, one obtains
Kne
n2 = 1 − 2 (7.129)
f
THE IONS 589
where K is a constant which is equal to 80.5 if the wave frequency

f is expressed in Hertz. Coupling this expression with Snell’s law
(which describes the propagation of the electromagnetic radiation), the
level of reflection of the wave can be derived. Further, by tuning the
wave frequency of the transmitter, information on the vertical profile of
electron density can be deduced.
In the D-region, collisions cannot be neglected except for very high
frequency propagation. Neglecting the effect of the magnetic field, the
refractive index then becomes:
X X iXZ
n2 = 1 − =1− − (7.130)
1 − iZ 1 + Z2 1 + Z2
It follows that, to this order of approximation, a wave traveling through
an ionized medium with negligible refraction is progressively attenuated
with an absorption coefficient given by
e2 ne νc
K= (7.131)
2o me c ωr + νc2
2
which is often expressed in units of nepers per meter (a neper

corresponds to 8.69 dB). The Appleton-Hartree theory used to obtain
this expression employs an invariant collision frequency for all electrons.
However, as pointed out by Phelps and Pack (1959), the cold plasma
approximation is not valid in the D-region and the collision frequency
should vary with the velocity of the electrons. Sen and Wyller (1960)
generalized the Appleton-Hartree theory by assuming a realistic electron
energy distribution. In practical applications, however, the above
expressions can still be used, provided that νc is replaced by an effective
collision frequency νeff , such that the ratio νeff /νc , varies with the radio
wave frequency.
Study of the amplitude of waves in the MHz range, reflected in the
E- and F-regions, gives a measure of D-region absorption. Techniques
involving vertical reflection of pulses (called the A1 method) or oblique
reflection of continuous waves (A3 method) provide useful indications
of some properties of the D-region, particularly the change in electron
concentration with time at a given point for a layer where νc ≈ ωr .
Such systematic observations of absorption have established the nature
of variations which are related to ionization processes. For example, a
variation of the absorption L with solar zenith angle is found, of the
form:
L = C cosa χ (7.132)
where a is typically between 0.6 and 1.0 at mid-latitudes, and C is a
constant which varies with season, geomagnetic activity, etc. A variation
with solar activity has also been noted (see e.g., Appleton and Piggott,
1954). It has been found that while the absorption varies systematically
with the mean zenith angle in summer, abnormally high absorption
is observed in winter. In the absence of changes in the atmospheric
composition, one would expect to obtain lower absorption in winter
than in summer because of the larger wintertime solar zenith angle.
However, observations suggest the opposite (winter mean values of the
electron concentration are larger than the summertime values, producing
the so-called “smooth winter anomaly”). During winter, the day-to-day
variability in the electron concentration is also higher (typically a factor
of 3-4). In addition, unexpected intense absorption events lasting a few
days are often observed in winter even under quiet solar conditions (see
Figure 7.36). This effect was observed by Appleton as early as 1937,
and is often referred to as “sporadic winter anomaly”. It has been the
subject of considerable research. The phenomenon is clearly related to
an increase in electron concentration (Mechtly and Shirke, 1968) and
NO+ ion density (Zbinden et al., 1975). Under these conditions, the
level at which the water cluster ion densities equal that of the molecular
ions (NO+ and O+ 2 ) is typically 5 km below normal.
The observations obtained during a measurement campaign at El
Arenosillo (37◦ N) (European winter anomaly campaign 1975/6, see
Offermann, 1979) as well as model studies (e.g., Koshelev, 1979; Solomon
et al., 1982b) have indicated that the origin of the winter anomaly
probably lies in sporadic injection of large amounts of nitric oxide from
the thermosphere to the mesosphere in association with changes in
atmospheric dynamics. D-region models which include the effects of
both chemistry and dynamics (Solomon et al., 1982a,b; Brasseur and
De Baets, 1986) show that ionization is more intense in winter than in
summer, even under normal conditions, as a result of the greater NO
densities in the winter season. The observed changes are consistent
with the observations of a “smooth winter anomaly” underlying the
sporadic enhancements. Measurements of NO densities during sporadic
winter anomaly events (Beran and Bangert, 1979; Beynon et al., 1976;
Arnold and Krankowsky, 1979; Aikin et al., 1977) indicate that the
NO concentration is greatly elevated during these conditions. The
sporadic winter anomaly may be related to atmospheric heating, which
decreases the positive ion-electron recombination coefficient because the
concentrations of the cluster ions decrease while those of the molecular
ions increase (Danilov and Taubenheim, 1983). It is now believed,
however, that the NO enhancement and winter anomalous absorption of
radio signals is related to transport processes associated with planetary
waves (Labitzke et al., 1979; Kawahira, 1982, Garcia et al., 1987).
THE IONS 591
Figure 7.36. Time series of D-region absorption during summer and winter at
mid-latitude. From Schwentek (1971).
References
Aikin, A.C., R.A. Goldberg, W. Jones, and J.A. Kane, Observations of the mid-
latitude lower ionosphere in winter. J Geophys Res: 82, 1869, 1977.
Appleton, E.V., Regularities and irregularities in the ionosphere, 1. Proc Roy Soc A:
162, 451, 1937.
Appleton, E., and W.R. Piggott, Ionospheric absorption measurements during a
sunspot cycle. J Atmos Terr Phys: 8, 141, 1954.
Arijs, E., Positive and negative ions in the stratosphere. Ann Geophys: 1, 149, 1983.
Arijs, E., Stratospheric ion chemistry: Present understanding and outstanding
problems. Planet Space Sci: 40, 255, 1992.
Arijs, E., and G. Brasseur, Acetonitrile in the stratosphere and implication for positive
ion composition. J Geophys Res: 91, 4003, 1986.
Arijs, E., J. Ingels, and D. Nevejans, Mass spectrometric measurement of positive ion
composition in the stratosphere. Nature: 271, 642, 1978.
Arijs, E., D. Nevejans, and J. Ingels, Unambiguous mass determination of major
stratospheric positive ions. Nature: 288, 684, 1980.
Arijs, E., D. Nevejans, P. Frederick, and J. Ingels, Negative ion composition
measurements in the stratosphere. Geophys Res Lett: 8, 121, 1981.
Arijs, E., D. Nevejans, and J. Ingels, Stratospheric positive ion composition

measurements, ion abundances and related trace gas detection. J Atmos Terr Phys:
44, 43, 1982a.
Arijs, E., D. Nevejans, P. Frederick, and J. Ingels, Stratospheric negative ion
composition measurements, ion abundances and related trace gas detection. J Atmos
Terr Phys: 44, 681, 1982b.
Arijs, E., D. Nevejans, and J. Ingels, Positive ion composition measurements and
acetonitrile in the upper stratosphere. Nature: 303, 314, 1983a.
Arijs, E., D. Nevejans, J. Ingels, and P. Frederick, Positive ion composition
measurements between 33 and 20 km altitude. Ann Geophys: 2, 161, 1983b.
Arnold, F., The middle atmosphere ionized component, in Proc of the ESA-
Symposium on Rocket- and Balloon-Programmes. Bournemouth, 1980.
Arnold, F., Atmospheric ions. Studies in Env Sci: 26, 103, 1986.
Arnold, F., and D. Krankowsky, Ion composition and electron- and ion-loss processes
in the earth’s atmosphere, pp. 93-127, in Dynamical and Chemical Coupling Between
the Neutral and Ionized Atmosphere. Grandal, B., and Holtet, J.A., (eds.), D. Reidel
Publishing Company, Dordrecht, Holland, 1977.
Arnold, F., and G. Henschen, First mass analysis of stratospheric negative ions,
Nature: 275, 521, 1978.
Arnold, F., and D. Krankowsky, Mid-latitude lower ionosphere structure and
composition measurements during winter. J Atmos Terr Phys: 41, 1127, 1979.
Arnold, F., and G. Hauck, Lower stratosphere trace gas detection using aircraft-borne
active chemical ionization mass spectrometry. Nature: 315, 307, 1985.
Arnold, F., J. Kissel, D. Krankowsky, H. Wieder, and J. Zähringer,
¨ Negative ions in
the lower ionosphere: A mass spectrometric measurement. J Atmos Terr Phys: 33,
1169, 1971.
Arnold, F., D. Krankowsky, and K.H. Marien, First mass spectrometric measurements
of positive ions in the stratosphere. Nature: 267,
7 30, 1977.
Arnold, F., H. Böhringer,
¨ and G. Henschen, Composition measurements of
stratospheric positive ions. Geophys Res Lett: 5, 653, 1978.
Arnold, F., G. Henschen, and E.E. Ferguson, Mass spectrometric measurements of
fractional ion abundances in the stratosphere — Positive ions. Planet Space Sci: 29,
185, 1981.
Banks, P., and G. Kockarts, Aeronomy. Academic Press, New York, 1973.
Becker, K.H., and A. Ionescu, Acetonitrile in the lower troposphere. Geophys Res
Lett: 9, 1349, 1982.
Beig, G., and D.K. Chakrabarty, On modeling stratospheric positive ions, J Atmos
Chem: 6, 175, 1988.
Beig, G., S. Walters, and G. Brasseur, A two-dimensional model of ion composition
in the stratosphere, 1. Positive ions. J Geophys Res: 98, 12,767, 1993a.
Beig, G., S. Walters, and G. Brasseur, A two-dimensional model of ion composition
in the stratosphere, 2. Negative ions. J Geophys Res: 98, 12,775, 1993b.
Beran, D., and W. Bangert, Trace constituents in the mesosphere and lower
thermosphere during winter anomaly events, J Atmos Terr Phys: 41, 1091, 1979.
Beynon, W.J.G., E.R. Williams, F. Arnold, D. Krankowsky, W.C. Bain, and P.H.G.
Dickinson, D-region rocket measurements in winter anomaly absorption conditions.
Nature: 261, 118, 1976.
THE IONS 593
Bills, R.E., and C.S. Gardner, Lidar observations of mesospheric Fe and sporadic Fe
layers at Urbana, Illinois. Geophys Res Lett: 17,
7 143, 1990.
Björn,
¨ L.G., and F. Arnold, Mass spectrometric detection of precondensation nuclei
at the arctic summer mesopause. Geophys Res Lett: 8, 1167, 1981.
Bohringer,
¨ H., and F. Arnold, Acetonitrile in the stratosphere — Implications from
laboratory studies. Nature: 290, 321, 1981.
Bohringer,
¨ H., D.W. Fahey, F.C. Fehsenfeld, and E.E. Ferguson, The role of ion-
molecule reactions in the conversion of N2 O5 to HNO3 in the stratosphere. Planet
Space Sci: 31, 185, 1983.
Brasseur, G., Physique et Chimie de l’Atmosphere Moyenne. Masson, Paris, 1982.
Brasseur, G., and M. Nicolet, Chemospheric processes of nitric oxide in the
mesosphere and stratosphere. Planet Space Sci: 21, 939, 1973.
Brasseur, G., and A. Chatel, Modelling of stratospheric ions: A first attempt. Ann
Geophys: 1, 173, 1983.
Brasseur, G., and P. De Baets, Minor constituents in the mesosphere and lower
thermosphere, J Geophys Res: 91, 4025, 1986.
Brasseur, G., E. Arijs, A. De Rudder, D. Nevejans, and J. Ingels, Acetonitrile in the
atmosphere. Geophys Res Lett: 10, 725, 1983.
Budden, K.G., Radio Waves in the Ionosphere. Cambridge Univ. Press, 1961.
Burns, C.J., E. Turunen, M. Matveinen, H. Ranta, and J.K. Hargreaves, Chemical
modeling of the quiet D- and E-Region using EISCAT electron density profiles. J
Atmos Terr Phys: 53, 115, 1991.
Callis, L.B., R.E. Boughner, N. Natarajan, J.D. Lambeth, D.N. Baker, and J.B.
Blake, Ozone depletion in the high altitude lower stratosphere, 1979-1990. J Geophys
Res: 96, 2921, 1991.
Chapman, S., Notes on atmospheric sodium. Astrophys J: 90, 309, 1939.
Crutzen, P.J., and S. Solomon, Response of mesospheric ozone to particle
precipitation. Planet Space Sci: 28, 1147, 1980.
Crutzen, P.J., I.S.A. Isaksen, and G.C. Reid, Solar proton events: Stratospheric
Danilov, A.D., and J. Taubenheim, NO and temperature control of the D-Region.
Space Sci Rev: 34, 413, 1983.
Davies, K., Ionospheric radio propagation, National Bureau of Standard, Monograph
80, 1965 and Dover Publications, New York, 1966.
Dubach, J., and W.A. Barker, Charged particle induced ionization rates in planetary
atmospheres. J Atmos Terr Phys: 33, 1287, 1971.
Fehsenfeld, F.C., and E.E. Ferguson, Thermal energy positive ion reactions in a wet
atmosphere containing ammonia. J Chem Phys: 59, 6272, 1973.
Fehsenfeld, F.C., and E.E. Ferguson, Laboratory studies of negative ion reactions
with atmospheric trace constituent. J Chem Phys: 61, 3181, 1974.
Fehsenfeld, F.C., C.J. Howard, and A.L. Schmeltekopf, Gas phase ion chemistry of
HNO3 . J Chem Phys: 63, 2835, 1975.
Fehsenfeld, F.C., I. Dotan, D.L. Albritton, C.J. Howard, and E.E. Ferguson,
Stratospheric positive ion chemistry of formaldehyde and methanol, J Geophys Res:
83, 1333, 1978.
Ferguson, E.E., D-Region ion chemistry. Rev Geophys Space Phys: 9, 997, 1971.
Ferguson, E., Sodium hydroxide in the stratosphere. Geophys Res Lett: 5, 1035, 1978.
Ferguson, E.E., Ion-molecule reactions in the atmosphere, in Kinetics of Ion-Molecule
Reactions. Ausloos, P. (ed.), Plenum Press, New York, 1979.
Frederick, J.E., Solar corpuscular emission and neutral chemistry in the Earth’s
middle atmosphere. J Geophys Res: 81, 3179, 1976.
Fritzenwallner, J., Globales Modell zur Verteilung der Positiven und Negativen Ionen
in der Unteren Ionosphären
¨ der Erde. Ph.D. thesis, University of Bern, Switzerland,
1997.
Fritzenwallner, J., and E. Kopp, Model calculations of the silicon and magnesium
chemistry in the mesosphere and lower thermosphere. Adv Space Res: 21, 859, 1998a.
Fritzenwallner, J., and E. Kopp, Model calculations of the negative ion chemistry in
the mesosphere with special emphasis on the chlorine species and the formation of
cluster ions. Adv Space Res: 21, 891, 1998b.
Garcia, R.R., S. Solomon, G.C. Reid, S.K. Avery, Transport of nitric oxide and the
D region winter anomaly. J Geophys Res: 92, 977, 1987.
Granier, C., J.P. Jegou,
´ and G. M´
égie, Atomic and ionic calcium in the earth’s upper
Hall, L.A., and H.E. Hinteregger, Solar radiation in the extreme ultraviolet and its
variation with solar rotation. J Geophys Res: 75, 6959, 1970.
Hamm, S., G. Helas, and P. Warneck, Acetonitrile in the air over Europe. Geophys
Res Lett: 16, 483, 1989.
Hargreaves, J.K., The Upper Atmosphere and Solar-Terrestrial Relations. Van
Nostrand Reinhold, 1979.
Hargreaves, J.K., The Solar-Terrestrial Environment. Cambridge Univ. Press, 1992.
Hartree, D.R., The propagation of electro-magnetic waves in a refracting medium in
a magnetic field, in Proc Cambridge Phil Soc: 27,
7 143, 1931.
Heaps, M.G., Parameterization of the cosmic ray ion-pair production rate above
18 km. Planet Space Sci: 26, 513, 1978a.
Heaps, M.G., The Effect of a Solar Proton Event on the Minor Neutral Constituents
of the Summer Polar Mesosphere. US Army Atmospheric Sci Lab Report ASL-TR-
0012, 1978b.
Henschen, G., and F. Arnold, Extended positive ion composition measurements in
the stratosphere — Implication for neutral trace gases. Geophys Res Lett: 8, 999,
1981.
Hunten, D.M., and M.B. McElroy, Metastable O2 (1 ∆) as a major source of ions in
the D-region. J Geophys Res: 73, 2421, 1968.
Jackman, C.H., Effect of energetic particles on minor constituents in the middle
atmosphere. J Geomagn Geoelect: 43, 637, 1991.
Jackman, C.H., H.S. Porter, and J.E. Frederick, Upper limits on production rate of
NO per ion pair. Nature: 280, 170, 1979.
Jackman, C.H., J.E. Frederick, and R.S. Stolarski, Production of odd nitrogen in the
stratosphere and mesosphere: An intercomparison of source strengths. J Geophys
Res: 85, 7495, 1980.
Jackman, C.H., A.P. Douglass, R.B. Rood, R.D. McPeters, and P.E. Maede, Effect of
solar events on the middle atmosphere during the past two solar cycles as computed
using a two-dimensional model. J Geophys Res: 95, 7417, 1990.
THE IONS 595
Jackman, C.H., J.E. Nielsen, D.J. Allen, M.C. Cerniglia, R.D. McPeeters, A.R.
Douglass, and R.B. Rood, The effects of the October 1989 solar proton events on
the stratosphere as computed using a 3-Dimensional model. Geophys Res Lett: 20,
459, 1993.
Jackman, C.H., M.C. Cerniglia, J.E. Nielsen, D.J. Allen, J.M. Zawodny, R.D.
McPeters, A.R. Douglass, J.E. Rosenfield, and R.B. Rood, Two-dimensional and
three-dimensional model simulations, measurements, and interpretation of the
influence of the October 1989 solar proton events on the middle atmosphere. J
Geophys Res: 100, 11,641, 1995.
Jegou,
´ J.P., C. Granier, M.L. Chanin, and G. M´ égie, General theory of the alkali
metals present in the earth’s upper atmosphere, 1. Flux model: Chemical and
dynamical processes. Ann Geophys: 3, 163, 1985a.
Jegou,
´ J.P., C. Granier, M.L. Chanin, and G. M´ égie, General theory of the alkali
metals present in the earth’s upper atmosphere, 2. Seasonal and meridional variations.
Ann Geophys: 3, 299, 1985b.
Johnson, C.Y., E.B. Meadows, and J.C. Holmes, Ion composition of the arctic
ionosphere. J Geophys Res: 63, 443, 1958.
Kawahira, K., An observational study of the D-region winter anomaly and sudden
stratospheric warmings. J Atmos-Terr Phys: 44, 947, 1982.
Kawamoto, H., and T. Ogawa, A steady-state model of negative ions in the lower
stratosphere. Planet Space Sci: 32, 1223, 1984.
Kawamoto, H., and T. Ogawa, Minor species of negative ions in the lower
stratosphere. Res Lett Atmos Electr: 5, 1, 1985.
Kazil, J., E. Kopp, S. Chabrillat, and J. Bishop, The University of Bern atmosphere
ion model: Time–dependent modeling of the ions in the mesophere and lower
thermosphere. J Geophys Res: 108, d14, 4432, doi 10.1029/2002JD003024, 2003.
Kelley, M.C., The Earth’s Ionosphere: Plasma Physics and Electrodynamics.
Academic Pres, 1989.
Keneshea, T.J., R.S. Narcisi, and W. Swider, Jr., Diurnal model of the E region. J
Geophys Res: 75, 845, 1970.
Kley, D., G.M. Lawrence, and E.J. Stone, The yield of N (2 D) atoms in the
dissociative recombination of NO+ . J Chem Phys: 66, 4157, 1977.
Knop, G., and F. Arnold, Stratospheric trace gas detection using a new balloon-borne
ACIMS method: Acetonitrile, acetone and nitric acid. Geophys Res Lett: 14, 1262,
1987a.
Knop, G., and F. Arnold, Atmospheric acetonitrile measurements in the tropopause
region using aircraft-borne active chemical ionization mass spectrometry. Planet
Space Sci: 35, 259, 1987b.
Kopp, E., and U. Herrmann, Ion composition in the lower ionosphere. Ann Geophys:
2, 83, 1984.
Koshelev, V.V., Variations of transport conditions and winter anomaly in the D-
ionospheric region. J Atmos Terr Phys: 41, 431, 1979.
Koshelev, V.V., The winter anomaly in the ionospheric D-region – Some numerical
calculations. J Atmos Terr Phys: 49, 81, 1987.
Kull, A., E. Kopp, C. Granier, and G. Brasseur, Ions and electrons of lower-latitude
D region. J Geophys Res: 102, 9705, 1997.
Labitzke, K., K. Paetzoldt, and H. Schwentek, Planetary waves in the strato- and
mesosphere during the western European winter anomaly campaign 1975/76 and
their relation to ionospheric absorption. J Atmos Terr Phys: 41, 1149, 1979.
Leu, M.T., M.A. Biondi, and R. Johnsen, Measurements of the recombination of
elections with H3 O+ (H2 O)n series ions. Phys Rev: A7,
7 292, 1973.
Massie, S.T., Nitric Oxide Delta Band Absorption Measurements in the Lower
Thermosphere. PhD thesis, University of Colorado, Boulder, Colo, 1979.
McCrumb, J.L., and F. Arnold, High-sensitivity detection of negative ions in the
stratosphere. Nature: 294, 136. 1981.
Mechtly, E.A., and J.S. Shirke, Rocket electron concentration measurements on winter
days of normal and anomalous absorption. J Geophys Res: 73, 6243, 1968.
Mechtly, E.A., and L.G. Smith, Seasonal variation of the lower ionosphere at Wallops
Island during the IQSY. J Atmos Terr Phys: 30, 1555, 1968.
Narcisi, R.S., and A.D. Bailey, Mass spectrometric measurements of positive ions at
altitudes from 64 to 112 kilometers. J Geophys Res: 70, 3687, 1965.
Narcisi, R.S., A.D. Bailey, L. Della Lucca, C. Sherman, and D.M. Thomas, Mass
spectrometric measurements of negative ions in the D- and lower E-regions. J Atmos
Terr Phys: 33, 1147, 1971.
Nicolet, M., Contribution à` l’étude de la structure de l’ionosphère. Mem Inst Mét´
´ or
Belgique: 19, 83, 1945.
Nicolet, M., and A.C. Aikin, The formation of the D region of the ionosphere. J
Geophys Res: 65, 1469, 1960.
Nicolet, M., and W. Peetermans, The production of nitric oxide in the stratosphere
by oxidation of nitrous oxide. Ann Geophys: 28, 751, 1972.
Niehaus, A., Excitation and dissociation of molecules by electron bombardment.
Measurement of the formation probability for neutral fragments as a function of
electron energy (in German). Z Naturf: 22a, 690, 1967.
Norton, R.B., and C.A. Barth, Theory of nitric oxide in the earth’s atmosphere. J
Geophys Res: 75, 3903, 1970.
Offermann, D., An integrated GBR campaign for the study of the D region winter
anomaly in western Europe 1975/76. J Atm Terr Phys: 41, 1047, 1979.
Paulsen, D.E., R.E. Huffman, and J.C. Larrabee, Improved photoionization of
O2 (1 ∆g ) in the D region. Radio Sci: 7
7, 51, 1972.
Phelps, A.V., and J.L. Pack, Electron collision frequencies in nitrogen and in the
lower ionosphere. Phys Rev Lett: 3, 340, 1959.
Plane, J.M.C., The chemistry of meteoric metals in the earth’s upper atmosphere.
Int Rev Phys Chem: 10, 55, 1991.
Plane, J.M.C., C.S. Gardner, J. Yu, C.Y. She, R.R. Garcia, and H.C. Pumphrey,
Mesopheric Na layer at 40◦ N, Modeling and observations. J Geophys Res: 104, 3773,
1999.
Potemra, T.A., Ionizing radiation affecting the lower ionosphere, in ELF-VLF
Radio Wave Propagation. Holtet, J.A. (ed.), Reidel Publishing, Dordrecht, The
Netherlands, 21, 1974.
Qian, J., and C.S. Gardner, Simultaneous lidar measurements of mesospheric Ca, Na,
and temperature profile at Urbana, Illinois. J Geophys Res: 100, 7453, 1995.
Rapp, D., P. Englander-Golden, and D.D. Briglia, Cross sections for dissociative
ionization of molecules by electron impact. J Chem Phys: 42, 4081. 1965.
THE IONS 597
Rees, M.H., Aurnal ionization and excitation by incident energetic electrons. Planet
Space Sci: 11, 1209, 1963.
Rees, M.H., and R.G. Roble, The morphology of N and NO in auroral substorms.
Planet Space Sci: 27,
7 453, 1979.
Rees, M.H., Physics and Chemistry of the Upper Atmosphere. Cambridge Univ.
Press, 1989.
Reid, G.C., The production of water cluster positive ions in the quiet daytime D-
region, J Geophys Res: 25, 275, 1977.
Reid, G.C., S. Solomon, and R.R. Garcia, Response of the middle atmosphere to the
solar proton events of August-December 1989. Geophys Res Lett: 18, 1019, 1991.
Richards, P.G., J.A. Fennelly, and D.G. Torr, EUVAC: A solar EUV flux model for
aeronomic calculations. J Geophys Res: 99, 8981, 1994.
Rosenberg, T.J., and L.J. Lanzerotti, Direct energy inputs to the middle atmosphere,
in Middle Atmosphere Electrodynamics, N.C. Maynard (ed.), NASA CP-2090, 43,
1979.
Rusch, D.W., J.-C. Gérard,
´ S. Solomon, P.J. Crutzen, and G.C. Reid, The effect
of particle precipitation events on the neutral and ion chemistry of the middle
atmosphere, 1. Odd nitrogen. Planet Space Sci: 29, 767, 1981.
Schneider, J., V. Burger,
¨ and F. Arnold, Methyl cyanide and hydrogen cyanide
measurements in the lower stratosphere: Implications for methyl cyanide sources
and sinks. J Geophys Res: 102, 25,501, 1997.
Schwentek, H., Regular and irregular behavior of the winter anomaly in ionospheric
absorption. J Atmos Terr Phys: 33, 1647, 1971.
Sen, H.K., and A.A. Wyller, On the generalizations of the Appleton-Hartree
magnetoionic formulas. J Geophys Res: 65, 3931, 1960.
Slipher, V.M., Emissions in the spectrum of the light of the night sky. Publ Astron
Soc Pac: 41, 262, 1929.
Smith, D., N.G. Adams, and M.J. Church, Mutual neutralization rates of
ionospherically important ions. Planet Space Sci: 24, 697, 1976.
Smith, D., N.G. Adams, and E. Alge, Ion-ion mutual neutralization and ion-neutral
switching reactions of some stratospheric ions. Planet Space Sci: 29, 449, 1981.
Smith, G.P., L.C. Lee, and P.C. Cosby, Photodissociation and photodetachment of
molecular negative ions, VIII. Nitrogen oxides and hydrates. J Chem Phys: 71, 4464,
1979.
Snider, J.R., and G.A. Dawson, Surface acetonitrile near Tucson, Arizona. Geophys
Res Lett 11, 241, 1984.
Solomon, S., The possible effects of translationally excited nitrogen atoms on lower
thermospheric odd nitrogen. Planet Space Sci: 31, 135, 1983.
Solomon, S., D.W. Rusch, J.-C. Gérard,
´ G.C. Reid, and P.J. Crutzen, The effect
of particle precipitation events on the neutral and ion chemistry of the middle
atmosphere, 2. Odd hydrogen. Planet Space Sci: 29, 885, 1981.
thermosphere and the lower atmosphere, 1. Odd nitrogen from 50 to 120 km. J
Geophys Res: 87,
7 7206, 1982a.
Solomon, S., G.C. Reid, R.G. Roble, and P.J. Crutzen, Photochemical coupling
between the thermosphere and the lower atmosphere, 2. D region ion chemistry and
the winter anomaly. J Geophys Res: 87,
7 7221, 1982b.
Solomon, S., G.C. Reid, D.W. Rusch, and R.J. Thomas, Mesospheric ozone depletion
during the solar proton cvent of July 13, 1982. 2. Comparison between theory and
Stewart, A.I., Photoionization coefficients and photoelectron impact excitation
efficiences in the daytime ionosphere. J Geophys Res: 75, 6333, 1970.
Swider, W., and T.J. Keneshea, Decrease of ozone and atomic oxygen in the lower
mesosphere during a PCA event. Planet Space Sci: 21, 1969, 1973.
Sze, N.D., M.K.W. Ko, W. Swider, and E. Murad, Atmospheric sodium chemistry, I.
The altitude region 70-100 km. Geophys Res Lett: 9, 1187, 1982.
Thomas, L., Modeling the ion composition of the middle atmosphere. Ann Geophys:
1, 61, 1983.
Thomas, L., and M.R. Bowman, Model studies of the D-region negative-ion
composition during day-time and night-time. J Atmos Terr Phys: 35, 397, 1985.
Thorne, R.M., Influence of relativistic electron precipitation on the lower ionosphere
and stratosphere, in Dynamical and Chemical Coupling Between the Neutral and
Ionized Atmosphere. Grandal, B., and J.A. Holtet (eds.), 161, Reidel Publishing,
Dordrecht, The Netherlands, 1977a.
Thorne, R.M., Energetic radiation belt electron precipitation: A natural depletion
mechanism for stratospheric ozone. Science: 195, 287, 1977b.
Tobiska, W.K., T. Woods, F. Eparvier, R. Viereck, L. Floyd, D. Bouwer, G. Rottman,
and O.R. White, The SOLAR2000 empirical solar irradiance model and forecast tool.
J Atmos Solar-Terr Phys: 62, 1233, 2000.
Vampola, A.L., and D.J. Gorney, Electron energy deposition in the middle
Viggiano, A.A., and F. Arnold, The first height measurements of the negative ion
composition of the stratosphere. Planet Space Sci: 29, 895, 1981.
Viggiano, A.A., R.A. Perry, D.L. Albritton, E.E. Ferguson, and F.C. Fehsenfeld, The
role of H2 SO4 in stratospheric negative-ion chemistry. J Geophys Res: 85, 4551,
1980.
Warneck, P., Cosmic radiation as a source of odd nitrogen in the stratosphere. J
Geophys Res: 77,
7 6589, 1972.
Winters, H.F., Ionic absorption and dissociation cross section for nitrogen. J Chem
Phys: 44, 1472, 1966.
Wisemberg, J., and G. Kockarts, Negative ion chemistry in the terrestrial D-region
and signal flow graph theory. J Geophys Res: 84, 4642, 1980.
Zbinden, P.A., M.A. Hidalgo, P. Eberhardt, and J. Geiss, Mass spectrometer
measurements of the positive ion composition in the D- and E-regions of the
ionosphere. Planet Space Sci: 23, 1621, 1975.
Zipf, E.C., and R.W. McLaughlin, On the dissociation of nitrogen by electron impact
and EUV photo-absorption. Planet Space Sci: 26, 449, 1978.
Appendix 1
PHYSICAL CONSTANTS
AND OTHER DATA
Table A.1.1
General and Universal Constants
Base of natural logarithms (e) 2.7182818285

π 3.14159265
Molar gas constant (R = NA k) 8.3143 J K−1 mol−1
Boltzmann’s constant (k) 1.38066 × 10−23 J K−1
Stefan-Boltzmann constant (σ) 5.67032 × 10−8 W m−2 K−4
Planck’s constant (h) 6.626176 × 10−34 J s
Speed of light (in vacuo) (c) 2.9979246 × 108 m s−1
Gravitational constant (G) 6.67259 × 10−11 m3 s−2 kg−1
Permittivity of free space (µ0 ) 8.85 × 10−12 F m−1
Mass of electrons (me ) 9.1096 × 10−31 kg
Charge of electron (e) 1.6022 × 10−19 C
Atomic mass unit (amu) 1.6605402 × 10−27 kg
Avogadro’s number (NA ) 6.022137 × 1023 mol−1
Triple-point temperature
of water (TT ) 273.16 K
Energy per unit wavenumber (hc) 1.986445 × 10−25 J m
First radiation constant
(c1 = 2πhc2 ) 3.741832 × 10−16 W m2
Second radiation constant
(c2 = hc/k) 1.438775 × 10−2 m K
Wien displacement constant
(b = λmax T) 2.897769 × 10−3 m K
599
Sun
Flux at Sun surface 6.312 × 107 W m−2

Luminosity 3.85 × 1026 W
Mass 1.99 × 1030 kg
Equatorial radius 6.9598 × 108 m
Surface area 6.087 × 1018 m2
Mean angle subtended by 31.988 arc min
photosphere at Earth
Emission temperature 5783 K
Earth
Average radius 6.371 × 106 m

Equatorial radius 6.378 × 106 m
Polar radius 6.357 × 106 m
Average height of land 840 m
Average depth of oceans 3730 m
Area of Earth’s surface 5.10 × 1014 m2
Area of Earth’s continents 1.49 × 1014 m2
Northern Hemisphere land surface area 1.03 × 1014 m2
Southern Hemisphere land surface area 0.46 × 1014 m2
Area of Earth’s oceans 3.61 × 1014 m2
Standard surface gravity 9.80665 m s−2
Mass of Earth 5.973 × 1024 kg
Mass of ocean 1.4 × 1021 kg
Mass of atmosphere 5.136 × 1018 kg
Mass of the biosphere 1.15 × 1016 kg
Eccentricity of orbit 0.016709
Equatorial escape velocity 1.18 × 104 m s−1
Inclination of rotation axis 23.45 deg or 0.409 rad
Mean angular rotation rate 7.292 × 10−5 rad s−1
Earth orbital period 365.25463 days
Length of sidereal day 23.94 hours
Lunar orbital period 27.32 days
Solar constant 1367 ± 2 W m−2
Mean distance from sun (1AU) 1.496 × 1011 m
Furthest Sun-Earth distance (4 July) 1.5196 × 1011 m
Nearest Sun-Earth distance (3 January) 1.4696 × 1011 m
Human population (1987) 5.0 × 109
Earth’s mean albedo 0.29
APPENDIX 1 601
Dry Air
Average molecular mass 28.97 g mol−1

Specific gas constant 287 J K−1 kg−1
Standard surface pressure 1.01325×105 Pa
Standard temperature 273.15 K
Mass density at 0◦ C and 101,325 Pa 1.2928 kg m−3
Molecular density at 0◦ C and 2.68678 × 1025 m−3
101,325 Pa
Molar volume at 0◦ C and 101,325 Pa 22.414 × 10−3 m3 mol−1
Specific heat at constant pressure (cp ) 1005 J K−1 kg−1
Specific heat at constant volume (cv ) 717.6 J K−1 kg−1
Index of refraction for air 1.000277
Dry adiabatic lapse rate –9.75 K km−1
Speed of sound in standard, calm air 343.15 m s−1
Water
Molecular weight 18.016 g mol−1

Gas constant for vapor 461.5 J K−1 kg−1
Density of pure water at 0◦ C 1000 kg m−3
Density of ice at 0◦ C 917 kg m−3
Density of water vapor at STP 0.803 kg m−3
Specific heat of vapor at constant pressure 1952 J K−1 kg−1
Specific heat of vapor at constant volume 1463 J K−1 kg−1
Specific heat of liquid water at 0◦ C 4218 J K−1 kg−1
Specific heat of ice at 0◦ C 2106 J K−1 kg−1
Latent heat of vaporization at 0◦ C 2.501 × 106 J kg−1
Latent heat of vaporization at 100◦ C 2.25 × 106 J kg−1
Latent heat of fusion at 0◦ C 3.34 × 105 J kg−1
Index of refraction for liquid water 1.336
Index of refraction for ice 1.312
Appendix 2
CONVERSION FACTORS AND

MULTIPLYING PREFIXES
Table A.2.1
International System of Units

Quantity Name of Unit Symbol Definition
Length meter m
Mass kilogram kg
Time second s
Electrical current ampere A
Temperature kelvin K
Derived Units
Force newton N kg m s−2
Pressure pascal Pa N m−2
Energy joule J kg m2 s−2
Power watt W J s−1
Electric potential
difference volt V W A−1
Electrical charge coulomb C As
Electrical resistance ohm Ω V A−1
Electrical capacitance farad F A s V−1
Frequency hertz Hz s−1
603
Conversion Factors
Area 1 ha = 104 m2
Volume 1 liter = 10−3 m3
Force 1 N = 105 dyn
Pressure 1 bar = 105 Pa
1 atm = 1.01325 × 105 Pa = 760.0 Torr
Energy 1 cal = 4.1855 J
1 eV = 1.6021 × 10−19 J
Power 1 W = 14.3353 cal min−1
Temperature T(◦ C) = T(K) - 273.15
T(◦ F) = 1.8 T(◦ C) + 32
Mixing ratios 1 ppb = 10−3 ppm
1 ppt = 10−3 ppb = 10−6 ppm
Logarithms ln x = 2.3026 log10 x
Table A.2.2
Multiplying Prefixes
Multiple Prefix Symbol Multiple Prefix Symbol
10−1 deci d 10 deca da

10−2 centi c 102 hecto h
10−3 milli m 103 kilo k
10−6 micro µ 106 mega M
10−9 nano n 109 giga G
10−12 pico p 1012 tera T
10−15 femto f 1015 peta P
10−18 atto a 1018 exa E
The International System of units is used except where usage suggests otherwise. The
unit of pressure is the pascal (Pa), but meteorologists commonly use the millibar (mb) or
equivalently the hectopascal (hPa). Number densities are often expressed in molecules
per cubic centimeter (cm−3 ). Volume mixing ratios are given in percent, parts per
million (ppmv), parts per billion (ppbv), or parts per trillion (pptv). Mass mixing ratios
are given in kilograms per kilograms or grams per kilograms. Wavelengths are often
expressed in micrometers (µm) or nanometers (nm). Wavenumbers are expressed in
inverse centimeters (cm−1 ).
Table A.2.3
Pressure Pa dyn cm2 atm torr hPa
1 Pa 1 10 9.86923 × 10−6 7.50062 × 10−3 10−2

APPENDIX 2
1 dyn cm−2 10−1 1 9.86923 × 10−7 7.50062 × 10−4 10−3

1 atm 1.01325 × 105 1.01325 × 106 1 760 1.01325 × 103
1 torr 1.33322 × 102 1.33322 × 103 1.31579 × 10−3 1 1.33322
1 hPa 102 103 9.86923 × 10−4 7.50062 × 10−1 1
Energy J eV erg cm−1 * K*
1J 1 6.24151 × 1018 107 5.03412 × 1022 7.24296 × 1022

1 eV 1.60218 × 10−19 1 1.60218 × 10−12 8.06554 × 103 1.16045 × 104
1 erg 10−7 6.24151 × 1011 1 5.03412 × 1015 7.24296 × 1015
1 cm−1 1.98645 × 10−23 1.23984 × 10−4 1.98645 × 10−16 1 1.43878
1K 1.38065 × 10−23 8.61734 × 10−5 1.38065 × 10−16 6.95036 × 10−1 1
∗ Derived from E = hcν̃ = kT.

605
Appendix 3
RATE COEFFICIENTS FOR SECOND-

ORDER GAS-PHASE REACTIONSa,b
Table A.3.1
Reaction A Factor Ea /R k298
O + O3 → O2 + O2 8.0 × 10−12 2060 8.0 × 10−15

O(1 D) + O2 → O + O2 3.2 × 10−11 –70 4.0 × 10−11
O(1 D) + O3 → O2 + O2 1.2 × 10−10 0 1.2 × 10−10
→ O2 + O + O 1.2 × 10−10 0 1.2 × 10−10
O(1 D) + H2 → OH + H 1.1 × 10−10 0 1.1 × 10−10
O(1 D) + H2 O → OH + OH 2.2 × 10−10 0 2.2 × 10−10
O(1 D) + N2 → O + N2 1.8 × 10−11 –110 2.6 × 10−11
O(1 D) + N2 O → N2 + O2 4.9 × 10−11 0 4.9 × 10−11
→ NO + NO 6.7 × 10−11 0 6.7 × 10−11
O(1 D) + CH4 → products 1.5 × 10−10 0 1.5 × 10−10
O(1 D) + CCl4 → products 3.3 × 10−10 0 3.3 × 10−10
O(1 D) + CH3 Br → products 1.8 × 10−10 0 1.8 × 10−10
O(1 D) + CH2 Br2 → products 2.7 × 10−10 0 2.7 × 10−10
O(1 D) + CHBr3 → products 6.6 × 10−10 0 6.6 × 10−10
O(1 D) + CHClF2 → products 1.0 × 10−10 0 1.0 × 10−10
O(1 D) + CCl3 F → products 2.3 × 10−10 0 2.3 × 10−10
O(1 D) + CCl2 F2 → products 1.4 × 10−10 0 1.4 × 10−10
O + OH → O2 + H 2.2 × 10−11 –120 3.3 × 10−11
O + HO2 → OH + O2 3.0 × 10−11 –200 5.9 × 10−11
H + O3 → OH + O2 1.4 × 10−10 470 2.9 × 10−11
H + HO2 → products 8.1 × 10−11 0 8.1 × 10−11
a The rate constant (cm3 molecule−1 s−1 ) is expressed by k = A exp(−Ea /RT), where
A, the pre-exponential factor, is given in cm3 molecule−1 s−1 , and Ea /R (activation
energy of the reaction divided by the gas constant) is given in Kelvin. The value of the
rate constant k298 at 298 K is given in cm3 molecule−1 s−1 .
b Sander, S.P., R.R. Friedl, D.M. Golden, M.J. Kurylo, R.E. Huie, V.L. Orkin,
G.K. Moortgat, A.R. Ravishankara, C.E. Kolb, M.J. Molina, and B.J. Finlayson-Pitts,
Chemical Kinetics and Photochemical Data for Use in Atmospheric Studies, Evaluation
No. 14, JPL Publication No. 02-25, Pasadena, Calif, USA, 2003.
607
OH + O3 → HO2 + O2 1.7 × 10−12 940 7.3 × 10−14

OH + H2 → H2 O + H 5.5 × 10−12 2000 6.7 × 10−15
OH + OH → H2 O + O 4.2 × 10−12 240 1.9 × 10−12
OH + HO2 → H2 O + O2 4.8 × 10−11 –250 1.1 × 10−10
OH + H2 O2 → H2 O + HO2 2.9 × 10−12 160 1.7 × 10−12
HO2 + O3 → OH + 2O2 1.0 × 10−14 490 1.9 × 10−15
HO2 + HO2 → H2 O2 + O2 2.3 × 10−13 –600 1.7 × 10−12
M
→ H2 O2 + O2 1.7 × 10−33 [M] –1000 4.9 × 10−32 [M]
O + NO2 → NO + O2 6.5 × 10−12 –120 9.7 × 10−12
OH + HNO3 → H2 O + NO3 (see Note)
OH + HO2 NO2 → products 1.3 × 10−12 –380 4.6 × 10−12
HO2 + NO → NO2 + OH 3.5 × 10−12 –250 8.1 × 10−12
HO2 + NO3 → products 3.5 × 10−12
N + O2 → NO + O 1.5 × 10−11 3600 8.5 × 10−17
N + NO → N2 + O 2.1 × 10−11 –100 3.0 × 10−11
N + NO2 → N2 O + O 5.8 × 10−12 –220 1.2 × 10−11
NO + O3 → NO2 + O2 2.0 × 10−12 1400 1.8 × 10−14
NO + NO3 → 2NO2 1.5 × 10−11 –170 2.6 × 10−11
NO2 + O3 → NO3 + O2 1.2 × 10−13 2450 3.2 × 10−17
OH + CO → products 1.5 × 10−13 × 0 1.5 × 10−13 ×
(1 + 0.6 Patm ) (1 + 0.6 Patm )
OH + CH4 → CH3 + H2 O 2.45 × 10−12 1775 6.3 × 10−15
OH + CH2 O → H2 O + HCO 1.0 × 10−11 0 1.0 × 10−11
OH + CH3 OH → products 6.7 × 10−12 600 8.9 × 10−13
OH + CH3 OOH → products 3.8 × 10−12 –200 7.4 × 10−12
OH + HCN → products 1.2 × 10−13 400 3.1 × 10−14
OH + CH3 CN → products 7.8 × 10−13 1050 2.3 × 10−14
HO2 + CH3 O2 → CH3 OOH + O2 3.8 × 10−13 –800 5.6 × 10−12
NO3 + CH2 O → HCO + HNO3 6.8 × 10−16
HCO + O2 → CO + HO2 5.2 × 10−12 0 5.2 × 10−12
CH2 OH + O2 → CH2 O + HO2 9.1 × 10−12
CH3 O + O2 → CH2 O + HO2 3.9 × 10−14 900 1.9 × 10−15
CH3 O2 + CH3 O2 → products 2.5 × 10−13 –190 4.7 × 10−13
CH3 O2 + NO → CH3 O + NO2 2.8 × 10−12 –300 7.7 × 10−12
O + FO → F + O2 2.7 × 10−11 0 2.7 × 10−11
F + O3 → FO + O2 2.2 × 10−11 230 1.0 × 10−11
F + H2 → HF + H 1.4 × 10−10 500 2.6 × 10−11
F + H2 O → HF + OH 1.4 × 10−11 0 1.4 × 10−11
F + CH4 → HF + CH3 1.6 × 10−10 260 6.7 × 10−11
FO + NO → NO2 + F 8.2 × 10−12 –300 2.2 × 10−11
FO2 + NO → FNO + O2 7.5 × 10−12 690 7.5 × 10−13
CF3 O + O3 → CF3 O2 + O2 2 × 10−12 1400 1.8 × 10−14
APPENDIX 3 609
CF3 O + NO → CF2 O + FNO 3.7 × 10−11 –110 5.4 × 10−11

CF3 O + CH4 → CF3 OH + CH3 2.6 × 10−12 1420 2.2 × 10−14
CF3 O2 + O3 → CF3 O + 2O2 < 3 × 10−15
CF3 O2 + NO → CF3 O + NO2 5.4 × 10−12 –320 1.6 × 10−11
OH + CF3 CH2 F (HFC-134a) 1.05 × 10−12 1630 4.4 × 10−15
→ H2 O + CF3 CHF
O + ClO → Cl + O2 3.0 × 10−11 –70 3.8 × 10−11
O + OClO → ClO + O2 2.4 × 10−12 960 1.0 × 10−13
O + HCl → OH + Cl 1.0 × 10−11 3300 1.5 × 10−16
O + HOCl → OH + ClO 1.7 × 10−13 0 1.7 × 10−13
O + ClONO2 → products 2.9 × 10−12 800 2.0 × 10−13
OH + ClO → Cl + HO2 7.4 × 10−12 –270 1.8 × 10−11
→ HCl + O2 6.0 × 10−13 –230 1.3 × 10−12
OH + OClO → HOCl + O2 4.5 × 10−13 –800 6.8 × 10−12
OH + HCl → H2 O + Cl 2.6 × 10−12 350 8.0 × 10−13
OH + HOCl → H2 O + ClO 3.0 × 10−12 500 5.0 × 10−13
OH + ClONO2 → products 1.2 × 10−12 330 3.9 × 10−13
OH + CH3 Cl → CH2 Cl + H2 O 2.4 × 10−12 1250 3.6 × 10−14
HO2 + Cl → HCl + O2 1.8 × 10−11 –170 3.2 × 10−11
→ OH + ClO 4.1 × 10−11 450 9.1 × 10−12
HO2 + ClO → HOCl + O2 2.7 × 10−12 –220 5.6 × 10−12
Cl + O3 → ClO + O2 2.9 × 10−11 260 1.2 × 10−11
Cl + H2 → HCl + H 3.7 × 10−11 2300 1.6 × 10−14
Cl + H2 O2 → HCl + HO2 1.1 × 10−11 980 4.1 × 10−13
Cl + CH4 → HCl + CH3 1.1 × 10−11 1400 1.0 × 10−13
Cl + CH2 O → HCl + HCO 8.1 × 10−11 30 7.3 × 10−11
Cl + OClO → ClO + ClO 3.4 × 10−11 –160 5.8 × 10−11
ClO + NO → NO2 + Cl 6.4 × 10−12 –290 1.7 × 10−11
ClO + ClO → Cl2 + O2 1.0 × 10−12 1590 4.8 × 10−15
→ ClOO + Cl 3.0 × 10−11 2450 8.0 × 10−15
→ OClO + Cl 3.5 × 10−13 1370 3.5 × 10−15
OH + CHF2 Cl (HCFC-22) 1.0 × 10−12 1600 4.7 × 10−15
→ CF2 Cl + H2 O
OH + CH3 CFCl2 (HCFC-141b) 1.25 × 10−12 1600 5.8 × 10−15
→ CH2 CFCl2 + H2 O
OH + CH3 CF2 Cl (HCFC-142b) 1.3 × 10−12 1770 3.4 × 10−15
→ CH2 CF2 Cl + H2 O
O + BrO → Br + O2 1.9 × 10−11 –230 4.1 × 10−11
OH + HBr → H2 O + Br 1.1 × 10−11 0 1.1 × 10−11
OH + CH3 Br → CH2 Br + H2 O 2.35 × 10−12 1300 3.0 × 10−14
HO2 + Br → HBr + O2 1.5 × 10−11 600 2.0 × 10−12
HO2 + BrO → products 3.4 × 10−12 –540 2.1 × 10−11
Br + O3 → BrO + O2 1.7 × 10−11 800 1.2 × 10−12

Br + H2 O2 → HBr + HO2 1.0 × 10−11 >3000 < 5 × 10−16
Br + CH2 O → HBr + HCO 1.7 × 10−11 800 1.1 × 10−12
BrO + O3 → Br + 2O2 ∼ 1.0 × 10−12 >3200 < 2 × 10−17
BrO + NO → NO2 + Br 8.8 × 10−12 –260 2.1 × 10−11
BrO + ClO → Br + OClO 9.5 × 10−13 –550 6.0 × 10−12
→ Br + ClOO 2.3 × 10−12 –260 5.5 × 10−12
→ BrCl + O2 4.1 × 10−13 –290 1.1 × 10−12
BrO + BrO → 2Br + O2 2.4 × 10−12 –40 2.7 × 10−12
→ Br2 + O2 2.8 × 10−14 –860 5.0 × 10−13
OH + OCS → products 1.1 × 10−13 1200 1.9 × 10−15
S + O2 → SO + O 2.3 × 10−12 0 2.3 × 10−12
SO + O2 → SO2 + O 2.6 × 10−13 2400 8.4 × 10−17
SO + O3 → SO2 + O2 3.6 × 10−12 1100 9.0 × 10−14
Note: The rate coefficient for this reaction is a complex function of
temperature and pressure and can be obtained as follows:
k2 k3 [M]
k(M, T) = k0 +
k2 + k3 [M]
where k0 = 7.2 × 10−15 exp (785/T)
k2 = 4.1 × 10−16 exp (1440/T)
k3 = 1.9 × 10−33 exp (725/T)
Appendix 4
RATE COEFFICIENTS FOR GAS-PHASE

ASSOCIATING REACTIONS a,b
Table A.4.1
Reaction Low-Pressure Limit High-Pressure Limit

k0300 n k0300 m
M
O + O2 → O3 6.0 × 10−34 2.4
M
H + O2 → HO2 5.7 × 10−32 1.6 7.5 × 10−11 0
M
OH + NO2 → HNO3 2.0 × 10−30 3.0 2.5 × 10−11 0
M
HO2 + NO2 → HO2 NO2 1.8 × 10−31 3.2 4.7 × 10−12 1.4
M
NO2 + NO3 → N2 O5 2.0 × 10−30 4.4 1.4 × 10−12 0.7
M
CH3 + O2 → CH3 O2 4.5 × 10−31 3.0 1.8 × 10−12 1.7
M
FO + NO2 → FONO2 2.6 × 10−31 1.3 2.0 × 10−11 1.5
M
ClO + NO2 → ClONO2 1.8 × 10−31 3.4 1.5 × 10−11 1.9
M
ClO + ClO → Cl2 O2 1.6 × 10−32 4.5 2.0 × 10−12 2.4
M
BrO + NO2 → BrONO2 5.2 × 10−31 3.2 6.9 × 10−12 2.9
M
OH + SO2 → HOSO2 3.0 × 10−31 3.3 1.5 × 10−12 0
a The rate constant (expressed in cm3 molecule−1 s−1 ) is given by

k0 (T)[M] 2 −1
k(M, T) = 0.6{1+[log10 (k0 (T)[M]/k∞ (T))] }
1 + k0 (T)[M]/k∞ (T)
where [M] is the air density (molecules cm−3 ), and T is the temperature (kelvin).
k0 is expressed in cm6 molecule−2 s−1 , and k∞ in cm3 molecule−1 s−1 . k0 (T) =
k300 −n , k (T) = k300 (T/300)−m .
0 (T/300) ∞ ∞
b Sander, S.P., R.R. Friedl, D.M. Golden, M.J. Kurylo, R.E. Huie, V.L. Orkin,
G.K. Moortgat, A.R. Ravishankara, C.E. Kolb, M.J. Molina, and B.J. Finlayson-Pitts,
Chemical Kinetics and Photochemical Data for Use in Atmospheric Studies, Evaluation
No. 14, JPL Publication No. 02-25, Pasadena, Calif, USA, 2003
611
Appendix 5
SURFACE REACTION PROBABILITY
Selected surface reaction probabilities (γ) for irreversible reactive uptake of

trace gases on condensed surfaces.a
Table A.5.1
Reaction Surface Type Temp. (K) γ Un-

certainty
Factor
N2 O5 (g) + H2 O
→ 2HNO3 H2 O (s) 188-195 0.02 2
H2 O (l) 260-295 ∼0.05 2
HNO3 · 3H2 O (s) 200 4 × 10−4 3
H2 SO4 · 4H2 O (s) 195-207 0.006 2
N2 O5 (g) + HCl (s)
→ ClNO2 + HNO3 H2 O/HCl (s) 190-220 0.03
HNO3 · 3H2 O · HCl (s) 200 0.003 2
HOCl (g) + HCl (s)
→ Cl2 + H2 O H2 O · HCl (s) 195-200 0.2 2
HNO3 · 3H2 O · HCl (s) 195-200 0.1 2
ClONO2 (g) + H2 O (s)
→ HOCl + HNO3 H2 O (s) 180-200 0.3 3
HNO3 · 3H2 O (s) 200 0.004 3
H2 SO4 · nH2 O (l) 200-265 see Note 1
ClONO2 (g) + HCl (s)
→ Cl2 + HNO3 H2 O (s) 180-200 0.3 3
HNO3 · 3H2 O · HCl (s) 185-210 0.2 2
ClONO2 (g) + HBr (s)
→ BrCl + HNO3 H2 O/HBr (s) 200 >0.3
HNO3 · 3H2 O · HBr (s) 200 >0.3
a Adapted from Sander, S.P., R.R. Friedl, D.M. Golden, M.J. Kurylo, R.E. Huie, V.L.
Orkin, G.K. Moortgat, A.R. Ravishankara, C.E. Kolb, M.J. Molina, and B.J. Finlayson-
Pitts, Chemical Kinetics and Photochemical Data for Use in Atmospheric Studies, Evaluation
No. 14, JPL Publication No. 02-25, Pasadena, Calif, USA, 2003.
Note 1: Values of γ are strongly dependent on the H2 SO4 concentration of the droplet,
increasing with decreasing [H2 SO4 ].
613
Table A.5.2
Reaction Surface Type Temp. (K) γ Uncertainty

Factor
HOBr (g) + HCl (s)

→ BrCl + H2 O H2 O/HBr (s) 180-228 0.3 3
HOBr (g) + HBr (s)
→ Br2 + H2 O H2 O/HBr (s) 228 0.1 3
BrONO2 (g) + H2 O
→ HOBr + HNO3 H2 O (s) 190-200 >0.3
Appendix 6
ATMOSPHERIC PROFILES
Appendix 6 presents numerical values for several atmospheric

quantities as a function of the log-pressure altitude (km). All the
variables are represented by zonally, latitudinally and annually averaged
values derived from the SOCRATES model (Brasseur et al., J Geophys
Res: 95, 5639, 1990). The values listed in the Tables should be
considered as approximate, and should be used only for crude estimates
of physical and chemical parameters of the atmosphere.
The first Table provides an estimate of the geometric height z (km),
the atmospheric pressure p (hPa), the absolute temperature T (K), the
atmospheric scale height H (km) and the air number density n (cm−3 ).
The subsequent Tables provide indicative values of the volume mixing
ratio (left column) and number density (cm−3 , right column) for several
chemical constituents of the atmosphere.
615
Table A.6.1.
log-palt. z Pressure Temperature H Air Dens

(km) (km) (hPa) (K) (km) (cm−3 )
0 0.0 1.013E+03 290.2 8.6 2.5E+19

5 6.0 4.959E+02 250.5 7.4 1.4E+19
10 11.2 2.428E+02 215.6 6.3 8.2E+18
15 15.7 1.188E+02 198.0 5.8 4.4E+18
20 20.3 5.818E+01 208.0 6.1 2.0E+18
25 25.1 2.848E+01 216.1 6.4 9.6E+17
30 30.1 1.394E+01 221.5 6.5 4.6E+17
35 35.2 6.826E+00 228.1 6.8 2.2E+17
40 40.5 3.341E+00 240.5 7.1 1.0E+17
45 46.2 1.636E+00 251.9 7.5 4.7E+16
50 51.9 8.008E−01 253.7 7.5 2.3E+16
55 57.7 3.920E−01 247.2 7.4 1.2E+16
60 63.2 1.919E−01 235.4 7.0 5.9E+15
65 68.4 9.395E−02 220.6 6.6 3.1E+15
70 73.2 4.599E−02 207.3 6.2 1.6E+15
75 77.8 2.251E−02 198.0 5.9 8.3E+14
80 82.3 1.102E−02 195.3 5.9 4.1E+14
85 86.8 5.396E−03 195.4 5.9 2.0E+14
90 91.3 2.641E−03 192.5 5.8 1.0E+14
95 95.7 1.293E−03 185.2 5.6 5.1E+13
100 99.8 6.330E−04 176.3 5.4 2.6E+13
105 104.0 3.099E−04 177.6 5.5 1.3E+13
110 108.4 1.517E−04 192.5 6.0 5.7E+12
Table A.6.2.a
log−palt. CH
H4 H2 O CO HCl HBr H2
0 1.7E−06 4.3E+13 1.7E−02 4.4E+17 7.5E−08 1.9E+12 1.3E−12 3.2E+07 3.7E−14 9.4E+05 2.7E−07 6.9E+12
5 1.7E−06 2.4E+13 1.2E−03 1.7E+16 5.8E−08 8.4E+11 1.5E−12 2.1E+07 1.9E−14 2.7E+05 2.7E−07 3.9E+12
APPENDIX 6
10 1.7E−06 1.4E+13 3.6E−05 3.0E+14 5.5E−08 4.5E+11 3.2E−12 2.6E+07 2.0E−14 1.7E+05 2.7E−07 2.2E+12
15 1.7E−06 7.3E+12 3.4E−06 1.5E+13 4.6E−08 2.0E+11 5.9E−11 2.6E+08 1.0E−13 4.4E+05 2.7E−07 1.2E+12
20 1.5E−06 3.1E+12 2.1E−06 4.2E+12 2.3E−08 4.7E+10 9.7E−10 2.0E+09 3.1E−13 6.2E+05 3.0E−07 6.0E+11
25 1.4E−06 1.3E+12 2.6E−06 2.5E+12 1.7E−08 1.7E+10 1.3E−09 1.3E+09 3.0E−13 2.8E+05 3.3E−07 3.1E+11
30 1.1E−06 5.0E+11 3.4E−06 1.5E+12 2.2E−08 9.9E+09 1.7E−09 7.9E+08 2.4E−13 1.1E+05 3.6E−07 1.6E+11
35 7.6E−07 1.7E+11 4.1E−06 8.9E+11 2.2E−08 4.9E+09 2.4E−09 5.1E+08 2.4E−13 5.2E+04 3.7E−07 8.0E+10
40 4.4E−07 4.4E+10 4.8E−06 4.8E+11 2.6E−08 2.6E+09 2.6E−09 2.7E+08 2.4E−13 2.4E+04 3.6E−07 3.6E+10
45 2.2E−07 1.0E+10 5.2E−06 2.5E+11 4.8E−08 2.3E+09 2.9E−09 1.4E+08 3.1E−13 1.5E+04 3.3E−07 1.5E+10
50 1.2E−07 2.8E+09 5.4E−06 1.2E+11 8.7E−08 2.0E+09 3.3E−09 7.5E+07 3.5E−13 8.1E+03 3.4E−07 7.8E+09
55 1.0E−07 1.1E+09 5.3E−06 6.1E+10 1.7E−07 1.9E+09 3.4E−09 3.9E+07 3.0E−13 3.4E+03 4.1E−07 4.7E+09
60 9.3E−08 5.5E+08 5.2E−06 3.1E+10 3.1E−07 1.9E+09 3.5E−09 2.1E+07 2.6E−13 1.5E+03 5.1E−07 3.0E+09
65 8.5E−08 2.6E+08 5.0E−06 1.6E+10 5.4E−07 1.7E+09 3.5E−09 1.1E+07 2.5E−13 7.7E+02 6.9E−07 2.1E+09
70 5.9E−08 9.6E+07 4.6E−06 7.4E+09 1.1E−06 1.7E+09 3.5E−09 5.6E+06 2.5E−13 4.0E+02 1.1E−06 1.8E+09
75 2.8E−08 2.3E+07 3.6E−06 3.0E+09 2.2E−06 1.8E+09 3.5E−09 2.9E+06 2.4E−13 2.0E+02 2.0E−06 1.6E+09
80 1.1E−08 4.3E+06 2.3E−06 9.5E+08 4.3E−06 1.8E+09 3.5E−09 1.4E+06 2.0E−13 8.1E+01 2.9E−06 1.2E+09
85 4.0E−09 8.0E+05 1.1E−06 2.2E+08 7.2E−06 1.4E+09 3.0E−09 6.1E+05 2.8E−14 5.6E+00 3.3E−06 6.6E+08
90 1.7E−09 1.7E+05 5.0E−07 5.0E+07 9.9E−06 9.9E+08 2.3E−09 2.3E+05 4.9E−16 4.9E−02 3.3E−06 3.3E+08
95 9.2E−10 4.6E+04 2.7E−07 1.4E+07 1.2E−05 6.1E+08 1.8E−09 9.2E+04 2.0E−17 1.0E−03 3.3E−06 1.7E+08
100 5.7E−10 1.5E+04 1.7E−07 4.4E+06 1.4E−05 3.7E+08 1.5E−09 4.0E+04 1.9E−18 4.9E−05 3.3E−06 8.5E+07
105 2.6E−10 3.3E+03 7.6E−08 9.6E+05 2.0E−05 2.5E+08 1.1E−09 1.4E+04 0.0E+00 0.0E+00 3.1E−06 3.9E+07
110 7.6E−11 4.3E+02 1.8E−08 1.0E+05 4.2E−05 2.4E+08 3.8E−10 2.2E+03 0.0E+00 0.0E+00 2.8E−06 1.6E+07
617
Table A.6.2.b 618
log-palt. CH
H2 O CCl4 CFC-11 CFC-12 CH
H3 CCl3 CH
H3 Cl
0 1.2E−10 3.1E+09 1.1E−10 2.7E+09 2.7E−10 6.8E+09 4.6E−10 1.2E+10 1.5E−10 3.8E+09 6.0E−10 1.5E+10
5 3.2E−11 .6E+08 1.1E−10 1.5E+09 2.7E−10 3.9E+09 4.6E−10 6.7E+09 1.5E−10 2.1E+09 5.8E−10 8.3E+09
10 1.0E−11 8.1E+07 1.1E−10 8.8E+08 2.7E−10 2.2E+09 4.6E−10 3.8E+09 1.5E−10 1.2E+09 5.8E−10 4.7E+09
15 4.8E−12 2.1E+07 1.0E−10 4.5E+08 2.6E−10 1.1E+09 4.5E−10 2.0E+09 1.4E−10 6.2E+08 5.6E−10 2.4E+09
20 9.9E−12 2.0E+07 6.3E−11 1.3E+08 1.7E−10 3.4E+08 3.9E−10 7.8E+08 8.6E−11 1.7E+08 4.6E−10 9.3E+08
25 2.4E−11 2.3E+07 1.3E−11 1.3E+07 5.3E−11 5.1E+07 2.8E−10 2.7E+08 1.8E−11 1.7E+07 3.2E−10 3.1E+08
30 5.6E−11 2.5E+07 1.1E−13 5.2E+04 1.8E−12 8.3E+05 1.4E−10 6.4E+07 2.1E−13 9.6E+04 1.7E−10 7.9E+07
35 8.7E−11 1.9E+07 6.2E−18 1.3E+00 1.2E−15 2.6E+02 2.9E−11 6.3E+06 2.7E−17 5.8E+00 5.0E−11 1.1E+07
40 9.1E−11 9.2E+06 1.8E−12 1.8E+05 6.0E−12 6.0E+05
45 6.2E−11 2.9E+06 5.2E−14 2.4E+03 3.3E−13 1.6E+04
50 3.6E−11 8.3E+05 1.5E−15 3.5E+01 2.2E−14 5.0E+02
55 2.7E−11 3.1E+05 8.5E−17 9.8E−01 4.0E−15 4.6E+01
60 2.1E−11 1.2E+05 7.8E−18 4.6E−02 1.4E−15 8.2E+00
65 1.2E−11 3.7E+04 1.1E−18 3.4E−03 6.4E−16 2.0E+00
70 3.9E−12 6.2E+03 3.3E−16 5.2E−01
75 1.1E−12 9.4E+02 1.7E−16 1.4E−01
80 3.5E−13 1.4E+02 9.3E−17 3.8E−02
85 7.6E−14 1.5E+01 5.4E−17 1.1E−02
90 1.8E−14 1.8E+00 3.5E−17 3.5E−03
95 8.8E−15 4.4E−01 2.5E−17 1.3E−03
100 8.6E−15 2.2E−01 2.0E−17 5.2E−04
105 5.8E−15 7.3E−02 1.3E−17 1.7E−04
110 7.4E−16 4.2E−03 6.7E−18 3.8E−05
AERONOMY OF THE MIDDLE ATMOSPHERE
Table A.6.2.c
log-palt. HF CH
H3 Br O3 O((3 P) O(1 D) H
0 6.9E−13 1.8E+07 1.0E−11 2.6E+08 3.8E−08 9.5E+11 1.3E−16 3.2E+03

5 9.8E−13 1.4E+07 9.9E−12 1.4E+08 4.8E−08 6.9E+11 2.9E−16 4.2E+03
APPENDIX 6
10 1.5E−12 1.2E+07 9.8E−12 8.0E+07 6.5E−08 5.3E+11 9.4E−16 7.7E+03

15 8.0E−12 3.5E+07 9.3E−12 4.1E+07 2.1E−07 9.0E+11 8.4E−15 3.7E+04
20 6.8E−11 1.4E+08 5.3E−12 1.1E+07 1.8E−06 3.6E+12 3.9E−13 7.9E+05
25 1.6E−10 1.6E+08 7.9E−13 7.6E+05 5.2E−06 4.9E+12 6.1E−12 5.9E+06 3.0E−18 2.9E+00
30 2.7E−10 1.3E+08 2.8E−15 1.3E+03 8.2E−06 3.8E+12 5.5E−11 2.5E+07 2.9E−17 1.3E+01 4.1E−18 1.9E+00
35 3.5E−10 7.6E+07 8.1E−06 1.8E+12 3.5E−10 7.6E+07 1.8E−16 4.0E+01 1.6E−16 3.5E+01
40 3.8E−10 3.9E+07 6.8E−06 6.9E+11 2.7E−09 2.7E+08 1.0E−15 1.0E+02 9.1E−15 9.2E+02
45 4.0E−10 1.9E+07 4.2E−06 2.0E+11 1.8E−08 8.6E+08 3.8E−15 1.8E+02 3.3E−13 1.5E+04
50 4.2E−10 9.5E+06 2.5E−06 5.7E+10 7.8E−08 1.8E+09 8.5E−15 1.9E+02 4.3E−12 9.9E+04
55 4.2E−10 4.8E+06 1.6E−06 1.9E+10 2.3E−07 2.6E+09 1.3E−14 1.5E+02 3.1E−11 3.6E+05
60 4.2E−10 2.5E+06 1.3E−06 7.6E+09 4.9E−07 2.9E+09 1.7E−14 9.8E+01 1.6E−10 9.7E+05
65 4.2E−10 1.3E+06 1.1E−06 3.5E+09 8.4E−07 2.6E+09 1.7E−14 5.2E+01 8.0E−10 2.5E+06
70 4.2E−10 6.8E+05 9.3E−07 1.5E+09 1.3E−06 2.1E+09 1.5E−14 2.4E+01 4.2E−09 6.8E+06
75 4.3E−10 3.5E+05 4.8E−07 4.0E+08 2.3E−06 1.9E+09 1.5E−14 1.3E+01 2.5E−08 2.1E+07
80 4.3E−10 1.8E+05 1.7E−07 6.9E+07 1.3E−05 5.2E+09 3.6E−14 1.5E+01 1.9E−07 7.9E+07
85 4.3E−10 8.6E+04 4.9E−07 9.9E+07 2.5E−04 5.0E+10 1.6E−13 3.3E+01 1.2E−06 2.4E+08
90 4.3E−10 4.3E+04 1.6E−06 1.6E+08 2.0E−03 2.0E+11 7.0E−13 7.0E+01 2.0E−06 2.0E+08
95 4.3E−10 2.2E+04 2.5E−06 1.3E+08 6.6E−03 3.4E+11 1.8E−12 8.9E+01 2.3E−06 1.2E+08
100 4.3E−10 1.1E+04 3.2E−06 8.3E+07 1.5E−02 4.0E+11 5.2E−12 1.4E+02 2.4E−06 6.3E+07
105 4.3E−10 5.5E+03 3.0E−06 3.8E+07 3.9E−02 4.9E+11 2.5E−11 3.2E+02 2.5E−06 3.2E+07
110 4.4E−10 2.5E+03 1.1E−06 6.5E+06 8.0E−02 4.6E+11 1.4E−10 7.8E+02 2.7E−06 1.6E+07
619
Table A.6.2.d 620
log-palt. OH HO2 H2 O2 Cl ClO HOCl
0 1.2E−13 3.0E+06 5.6E−12 1.4E+08 7.8E−11 2.0E+09 1.6E−17 4.0E+02 1.0E−14 2.6E+05 1.7E−13 4.2E+06
5 7.0E−14 1.0E+06 3.4E−12 4.9E+07 2.8E−11 4.0E+08 1.1E−17 1.6E+02 1.2E−14 1.7E+05 8.5E−14 1.2E+06
10 7.0E−14 5.7E+05 1.0E−12 8.3E+06 2.6E−12 2.1E+07 4.0E−17 3.2E+02 6.0E−14 4.9E+05 7.6E−14 6.2E+05
15 9.6E−14 4.2E+05 5.3E−13 2.3E+06 1.4E−12 6.2E+06 1.9E−16 8.3E+02 1.3E−12 5.7E+06 6.9E−13 3.0E+06
20 1.8E−13 3.7E+05 1.5E−12 2.9E+06 1.2E−11 2.4E+07 4.0E−15 8.2E+03 1.7E−11 3.5E+07 8.6E−12 1.7E+07
25 6.9E−13 6.6E+05 6.0E−12 5.7E+06 6.2E−11 5.9E+07 1.5E−14 1.4E+04 6.4E−11 6.2E+07 5.1E−11 4.9E+07
30 3.4E−12 1.6E+06 1.6E−11 7.5E+06 1.2E−10 5.7E+07 8.5E−14 3.9E+04 1.1E−10 5.0E+07 8.8E−11 4.0E+07
35 1.7E−11 3.7E+06 3.2E−11 6.9E+06 1.0E−10 2.2E+07 5.1E−13 1.1E+05 2.2E−10 4.8E+07 1.4E−10 3.0E+07
40 6.7E−11 6.8E+06 5.8E−11 5.9E+06 7.8E−11 7.9E+06 2.1E−12 2.2E+05 4.3E−10 4.4E+07 1.8E−10 1.8E+07
45 1.8E−10 8.5E+06 1.0E−10 4.9E+06 6.2E−11 2.9E+06 1.1E−11 5.1E+05 4.6E−10 2.2E+07 1.4E−10 6.7E+06
50 3.0E−10 6.8E+06 1.7E−10 4.0E+06 5.9E−11 1.4E+06 3.2E−11 7.2E+05 2.3E−10 5.2E+06 7.4E−11 1.7E+06
55 4.0E−10 4.6E+06 2.4E−10 2.7E+06 5.3E−11 6.2E+05 4.9E−11 5.6E+05 1.1E−10 1.2E+06 3.4E−11 3.9E+05
60 5.5E−10 3.2E+06 3.2E−10 1.9E+06 5.4E−11 3.2E+05 6.0E−11 3.5E+05 7.4E−11 4.4E+05 2.3E−11 1.4E+05
65 9.1E−10 2.8E+06 4.9E−10 1.5E+06 8.2E−11 2.5E+05 6.7E−11 2.1E+05 7.1E−11 2.2E+05 1.9E−11 5.7E+04
70 2.0E−09 3.2E+06 9.4E−10 1.5E+06 2.2E−10 3.6E+05 7.1E−11 1.2E+05 7.2E−11 1.2E+05 1.5E−11 2.4E+04
75 4.1E−09 3.4E+06 1.8E−09 1.5E+06 5.3E−10 4.4E+05 8.7E−11 7.2E+04 5.5E−11 4.6E+04 1.1E−11 9.3E+03
80 7.5E−09 3.1E+06 3.0E−09 1.2E+06 9.2E−10 3.8E+05 1.6E−10 6.5E+04 1.0E−11 4.3E+03 4.0E−12 1.6E+03
85 2.4E−09 4.9E+05 6.3E−10 1.3E+05 1.1E−10 2.1E+04 6.5E−10 1.3E+05 3.3E−13 6.7E+01 1.5E−14 3.0E+00
90 5.9E−10 5.9E+04 4.0E−11 4.0E+03 3.6E−12 3.6E+02 1.4E−09 1.4E+05 2.1E−13 2.1E+01 7.7E−17 7.7E−03
95 2.6E−10 1.3E+04 6.5E−12 3.3E+02 3.5E−13 1.8E+01 1.9E−09 9.4E+04 1.2E−13 6.2E+00 2.4E−18 1.2E−04
100 1.3E−10 3.4E+03 1.6E−12 4.0E+01 6.9E−14 1.8E+00 2.1E−09 5.6E+04 7.3E−14 1.9E+00
105 4.8E−11 6.0E+02 2.3E−13 2.9E+00 8.8E−15 1.1E−01 2.6E−09 3.3E+04 3.2E−14 4.0E−01
110 1.1E−11 6.5E+01 3.7E−14 2.1E−01 2.2E−16 1.3E−03 3.3E−09 1.9E+04 8.0E−15 4.6E−02
Table A.6.2.e
log-palt. ClONO2 N NO NO2 NO3 HNO3
0 1.3E−12 3.3E+07 1.2E−10 3.1E+09 3.5E−10 8.8E+09 1.3E−11 3.2E+08 1.5E−11 3.8E+08
5 4.0E−13 5.7E+06 1.5E−11 2.2E+08 4.8E−11 6.9E+08 4.0E−13 5.7E+06 1.0E−11 1.5E+08
APPENDIX 6
10 5.0E−13 4.1E+06 4.5E−11 3.7E+08 5.9E−11 4.8E+08 3.0E−14 2.5E+05 2.3E−10 1.9E+09
15 5.9E−12 2.6E+07 6.7E−11 2.9E+08 7.6E−11 3.3E+08 3.5E−14 1.5E+05 3.1E−10 1.3E+09
20 2.2E−10 4.4E+08 9.1E−11 1.8E+08 2.0E−10 4.0E+08 6.5E−13 1.3E+06 2.7E−09 5.5E+09
25 7.3E−10 7.0E+08 2.4E−10 2.3E+08 7.2E−10 6.9E+08 3.9E−12 3.7E+06 4.8E−09 4.6E+09
30 9.8E−10 4.5E+08 1.8E−17 8.3E+00 1.1E−09 5.1E+08 3.3E−09 1.5E+09 1.3E−11 6.0E+06 4.9E−09 2.3E+09
35 6.2E−10 1.4E+08 6.7E−16 1.5E+02 3.5E−09 7.6E+08 7.6E−09 1.7E+09 2.8E−11 6.1E+06 2.6E−09 5.8E+08
40 2.4E−10 2.4E+07 4.6E−15 4.7E+02 5.1E−09 5.1E+08 7.3E−09 7.4E+08 6.3E−11 6.4E+06 5.6E−10 5.7E+07
45 4.1E−11 2.0E+06 1.3E−14 6.1E+02 4.6E−09 2.1E+08 5.1E−09 2.4E+08 6.8E−11 3.2E+06 5.3E−11 2.5E+06
50 4.0E−12 9.3E+04 3.5E−14 8.1E+02 3.3E−09 7.6E+07 4.2E−09 9.6E+07 1.9E−11 4.4E+05 6.8E−12 1.6E+05
55 9.8E−13 1.1E+04 1.2E−13 1.3E+03 2.4E−09 2.8E+07 4.1E−09 4.7E+07 3.6E−12 4.1E+04 2.4E−12 2.8E+04
60 5.6E−13 3.3E+03 4.2E−13 2.5E+03 2.8E−09 1.6E+07 4.2E−09 2.5E+07 2.6E−12 1.6E+04 2.2E−12 1.3E+04
65 3.6E−13 1.1E+03 1.6E−12 4.9E+03 5.0E−09 1.5E+07 5.7E−09 1.8E+07 7.0E−13 2.2E+03 3.8E−12 1.2E+04
70 8.2E−14 1.3E+02 5.8E−12 9.4E+03 1.2E−08 1.9E+07 7.1E−09 1.1E+07 7.0E−14 1.1E+02 5.6E−12 9.1E+03
75 4.5E−15 3.8E+00 3.6E−11 3.0E+04 2.6E−08 2.2E+07 3.1E−09 2.6E+06 3.9E−14 3.2E+01 3.6E−12 3.0E+03
80 6.9E−17 2.8E−02 1.2E−10 5.1E+04 4.9E−08 2.0E+07 2.6E−10 1.1E+05 3.8E−14 1.6E+01 7.0E−13 2.9E+02
85 3.1E−10 6.1E+04 9.3E−08 1.9E+07 2.1E−11 4.3E+03 5.6E−15 1.1E+00 1.1E−13 2.1E+01
90 6.8E−10 6.8E+04 2.0E−07 2.0E+07 2.3E−11 2.3E+03 3.9E−15 3.9E−01 3.9E−14 3.9E+00
95 1.4E−09 7.0E+04 4.8E−07 2.4E+07 3.2E−11 1.6E+03 3.3E−15 1.7E−01 2.4E−14 1.2E+00
100 2.6E−09 6.9E+04 1.3E−06 3.3E+07 4.7E−11 1.2E+03 2.9E−15 7.5E−02 1.9E−14 5.0E−01
105 5.7E−09 7.1E+04 4.7E−06 5.9E+07 8.5E−11 1.1E+03 2.6E−15 3.3E−02 1.4E−14 1.8E−01
110 1.4E−08 7.7E+04 1.4E−05 8.2E+07 1.1E−10 6.3E+02 1.4E−15 7.9E−03 8.6E−15 4.9E−02
621
Table A.6.2.f 622
log-palt. N2 O5 HO2 NO2 Br BrO HOBr BrONO2
0 1.2E−11 3.1E+08 3.6E−13 9.0E+06 9.1E−16 2.3E+04 1.3E−14 3.4E+05 6.6E−14 1.7E+06 1.7E−13 4.3E+06
5 2.3E−11 3.4E+08 9.5E−12 1.4E+08 3.2E−15 4.6E+04 3.0E−14 4.3E+05 6.2E−14 8.8E+05 1.2E−13 1.7E+06
10 1.5E−11 1.2E+08 2.4E−11 2.0E+08 2.0E−14 1.6E+05 6.3E−14 5.1E+05 3.0E−14 2.5E+05 1.6E−13 1.3E+06
15 2.7E−12 1.2E+07 1.5E−11 6.7E+07 6.9E−14 3.0E+05 2.6E−13 1.1E+06 4.3E−14 1.9E+05 4.7E−13 2.0E+06
20 3.6E−11 7.3E+07 7.0E−11 1.4E+08 9.9E−14 2.0E+05 1.7E−12 3.4E+06 4.1E−13 8.3E+05 3.0E−12 6.1E+06
25 2.2E−10 2.1E+08 1.7E−10 1.6E+08 9.9E−14 9.4E+04 3.1E−12 2.9E+06 1.2E−12 1.1E+06 5.2E−12 5.0E+06
30 9.9E−10 4.5E+08 2.0E−10 9.3E+07 1.5E−13 7.0E+04 3.1E−12 1.4E+06 1.3E−12 6.1E+05 5.4E−12 2.5E+06
35 1.1E−09 2.5E+08 7.2E−11 1.6E+07 3.9E−13 8.4E+04 3.3E−12 7.3E+05 1.6E−12 3.4E+05 4.0E−12 8.6E+05
40 5.3E−10 5.4E+07 1.3E−11 1.4E+06 8.8E−13 8.9E+04 3.6E−12 3.6E+05 2.3E−12 2.3E+05 1.9E−12 1.9E+05
45 1.8E−10 8.5E+06 3.5E−12 1.6E+05 2.3E−12 1.1E+05 2.8E−12 1.3E+05 3.1E−12 1.5E+05 4.7E−13 2.2E+04
50 1.1E−10 2.6E+06 1.4E−12 3.2E+04 4.0E−12 9.2E+04 1.7E−12 3.8E+04 3.5E−12 8.0E+04 1.1E−13 2.5E+03
55 5.0E−11 5.7E+05 5.9E−13 6.8E+03 4.8E−12 5.5E+04 1.7E−12 1.9E+04 3.2E−12 3.7E+04 4.0E−14 4.7E+02
60 9.9E−12 5.9E+04 2.8E−13 1.7E+03 5.2E−12 3.1E+04 2.2E−12 1.3E+04 2.8E−12 1.7E+04 2.7E−14 1.6E+02
65 6.2E−13 1.9E+03 1.4E−13 4.4E+02 5.6E−12 1.7E+04 2.7E−12 8.4E+03 2.3E−12 7.0E+03 2.1E−14 6.4E+01
70 9.2E−15 1.5E+01 3.7E−14 6.0E+01 6.8E−12 1.1E+04 2.5E−12 4.1E+03 1.4E−12 2.3E+03 3.3E−15 5.4E+00
75 3.2E−16 2.6E−01 5.9E−15 4.9E+00 9.4E−12 7.8E+03 1.1E−12 8.8E+02 5.0E−13 4.1E+02 1.8E−16 1.5E−01
80 3.1E−17 1.3E−02 8.2E−16 3.4E−01 1.1E−11 4.5E+03 4.4E−14 1.8E+01 5.4E−14 2.2E+01 1.4E−17 5.8E−03
85 5.5E−18 1.1E−03 6.3E−17 1.3E−02 1.1E−11 2.3E+03 2.3E−16 4.6E−02 1.0E−15 2.0E−01 2.6E−18 5.3E−04
90 1.9E−18 1.9E−04 1.8E−17 1.8E−03 1.2E−11 1.2E+03 4.5E−17 4.4E−03 2.4E−16 2.4E−02
95 1.1E−18 5.3E−05 1.1E−17 5.4E−04 1.2E−11 5.9E+02 1.7E−17 8.7E−04 1.6E−16 7.9E−03
100 8.5E−18 2.2E−04 1.2E−11 3.1E+02 7.5E−18 1.9E−04 1.3E−16 3.5E−03
105 6.9E−18 8.8E−05 1.2E−11 1.5E+02 2.7E−18 3.4E−05 1.2E−16 1.5E−03
110 5.7E−18 3.2E−05 1.2E−11 6.8E+01 9.4E−17 5.4E−04
Table A.6.2.g
log-palt. BrCl CH
H3 O2 O2 ClOx Cly Bry
0 3.5E−18 8.9E+01 5.4E−12 1.4E+08 2.1E−01 5.3E+18 1.8E−13 4.6E+06 2.8E−12 7.0E+07 2.9E−13 7.2E+06
5 6.9E−18 9.9E+01 2.3E−12 3.3E+07 2.1E−01 3.0E+18 1.0E−13 1.5E+06 2.0E−12 2.8E+07 2.3E−13 3.3E+06
APPENDIX 6
10 2.4E−16 2.0E+03 2.4E−13 2.0E+06 2.1E−01 1.7E+18 3.0E−13 2.5E+06 4.1E−12 3.3E+07 2.9E−13 2.4E+06
15 2.9E−14 1.3E+05 1.7E−13 7.5E+05 2.1E−01 9.1E+17 4.8E−12 2.1E+07 7.0E−11 3.1E+08 9.7E−13 4.2E+06
20 1.7E−13 3.4E+05 4.4E−13 8.8E+05 2.1E−01 4.3E+17 4.7E−11 9.5E+07 1.2E−09 2.5E+09 5.7E−12 1.2E+07
25 3.3E−13 3.2E+05 6.8E−13 6.5E+05 2.1E−01 2.0E+17 1.5E−10 1.5E+08 2.2E−09 2.1E+09 1.0E−11 9.7E+06
30 1.8E−13 8.3E+04 7.7E−13 3.5E+05 2.1E−01 9.6E+16 2.3E−10 1.1E+08 3.0E−09 1.4E+09 1.0E−11 4.7E+06
35 3.2E−13 7.0E+04 1.0E−12 2.3E+05 2.1E−01 4.6E+16 3.7E−10 8.1E+07 3.4E−09 7.3E+08 9.8E−12 2.1E+06
40 8.2E−13 8.3E+04 2.3E−12 2.3E+05 2.1E−01 2.1E+16 6.2E−10 6.3E+07 3.5E−09 3.5E+08 9.7E−12 9.8E+05
45 7.1E−13 3.3E+04 5.2E−12 2.4E+05 2.1E−01 9.9E+15 6.2E−10 2.9E+07 3.6E−09 1.7E+08 9.7E−12 4.6E+05
50 2.6E−13 6.0E+03 9.0E−12 2.1E+05 2.1E−01 4.8E+15 3.4E−10 7.7E+06 3.6E−09 8.2E+07 9.9E−12 2.3E+05
55 1.1E−13 1.3E+03 1.5E−11 1.7E+05 2.1E−01 2.4E+15 1.9E−10 2.2E+06 3.6E−09 4.2E+07 1.0E−11 1.2E+05
60 7.4E−14 4.3E+02 2.2E−11 1.3E+05 2.1E−01 1.2E+15 1.6E−10 9.3E+05 3.6E−09 2.1E+07 1.1E−11 6.2E+04
65 5.1E−14 1.6E+02 2.5E−11 7.7E+04 2.1E−01 6.5E+14 1.6E−10 4.8E+05 3.6E−09 1.1E+07 1.1E−11 3.4E+04
70 2.3E−14 3.7E+01 1.1E−11 1.8E+04 2.1E−01 3.4E+14 1.6E−10 2.6E+05 3.7E−09 5.9E+06 1.1E−11 1.8E+04
75 3.6E−15 2.9E+00 2.4E−12 2.0E+03 2.1E−01 1.7E+14 1.5E−10 1.3E+05 3.7E−09 3.0E+06 1.1E−11 9.3E+03
80 8.6E−17 3.6E−02 7.1E−13 2.9E+02 2.1E−01 8.6E+13 1.7E−10 7.1E+04 3.7E−09 1.5E+06 1.1E−11 4.7E+03
85 9.6E−18 1.9E−03 4.5E−13 9.0E+01 2.1E−01 4.2E+13 6.5E−10 1.3E+05 3.7E−09 7.4E+05 1.1E−11 2.3E+03
90 3.7E−18 3.7E−04 2.6E−13 2.6E+01 2.1E−01 2.0E+13 1.4E−09 1.4E+05 3.7E−09 3.7E+05 1.2E−11 1.2E+03
95 2.5E−18 1.3E−04 9.6E−14 4.9E+00 2.0E−01 1.0E+13 1.9E−09 9.4E+04 3.7E−09 1.9E+05 1.2E−11 5.9E+02
100 2.3E−18 5.9E−05 4.7E−14 1.2E+00 1.9E−01 5.0E+12 2.1E−09 5.6E+04 3.7E−09 9.6E+04 1.2E−11 3.1E+02
105 2.1E−18 2.6E−05 1.9E−14 2.5E−01 1.7E−01 2.2E+12 2.6E−09 3.3E+04 3.7E−09 4.7E+04 1.2E−11 1.5E+02
110 1.6E−18 9.0E−06 3.0E−15 1.7E−02 1.4E−01 7.7E+11 3.3E−09 1.9E+04 3.7E−09 2.1E+04 1.2E−11 6.8E+01
623
Table A.6.2.h
log-palt. NOx NOy HOx
0 4.7E−10 1.2E+10 5.2E−10 1.3E+10 5.7E−12 1.5E+08

5 6.3E−11 9.1E+08 1.3E−10 1.9E+09 3.5E−12 5.0E+07
10 1.0E−10 8.5E+08 3.9E−10 3.2E+09 1.1E−12 8.9E+06
15 1.4E−10 6.2E+08 4.8E−10 2.1E+09 6.3E−13 2.7E+06
20 2.9E−10 5.9E+08 3.4E−09 6.8E+09 1.6E−12 3.3E+06
25 9.6E−10 9.1E+08 7.1E−09 6.8E+09 6.7E−12 6.4E+06
30 4.4E−09 2.0E+09 1.3E−08 5.7E+09 2.0E−11 9.0E+06
35 1.1E−08 2.4E+09 1.7E−08 3.6E+09 4.9E−11 1.1E+07
40 1.2E−08 1.3E+09 1.4E−08 1.5E+09 1.3E−10 1.3E+07
45 9.6E−09 4.5E+08 1.0E−08 4.8E+08 2.8E−10 1.3E+07
50 7.5E−09 1.7E+08 7.8E−09 1.8E+08 4.8E−10 1.1E+07
55 6.5E−09 7.4E+07 6.6E−09 7.6E+07 6.7E−10 7.7E+06
60 6.9E−09 4.1E+07 6.9E−09 4.1E+07 1.0E−09 6.1E+06
65 1.1E−08 3.3E+07 1.1E−08 3.3E+07 2.2E−09 6.8E+06
70 1.9E−08 3.0E+07 1.9E−08 3.0E+07 7.1E−09 1.2E+07
75 2.9E−08 2.4E+07 3.0E−08 2.4E+07 3.1E−08 2.6E+07
80 5.0E−08 2.0E+07 5.0E−08 2.0E+07 2.0E−07 8.3E+07
85 9.3E−08 1.9E+07 9.3E−08 1.9E+07 1.2E−06 2.4E+08
90 2.0E−07 2.0E+07 2.0E−07 2.0E+07 2.0E−06 2.0E+08
95 4.8E−07 2.4E+07 4.8E−07 2.4E+07 2.3E−06 1.2E+08
100 1.3E−06 3.3E+07 1.3E−06 3.3E+07 2.4E−06 6.3E+07
105 4.7E−06 5.9E+07 4.7E−06 5.9E+07 2.5E−06 3.2E+07
110 1.4E−05 8.2E+07 1.4E−05 8.2E+07 2.7E−06 1.6E+07
FIGURE ACKNOWLEDGEMENTS
Academic Press: Figures 4.12, 4.23, 5.24a, 7.1, 7.6
American Association for the Advancement of Science: Figures 5.22, 5.30
American Geophysical Union: Figures 2.10, 3.4, 3.6, 3.7, 3.11, 3.12,
3.15a,b,c, 3.18, 3.24, 3.25, 3.27, 3.29, 3.32, 3.35, 4.8b, 4.9, 4.18, 4.25, 4.26,
4.27, 4.28, 4.29, 4.42, 5.4, 5.11, 5.13, 5.16, 5.19, 5.20, 5.21, 5.29, 5.40, 5.44,
5.45, 5.47b,c, 5.56, 5.60, 5.65, 5.66, 5.67, 6.2, 6.3, 6.5, 6.9, 6.10, 6.11, 6.12,
6.13, 6.14, 6.17, 6.18, 6.21, 6.22, 6.23, 7.11, 7.13, 7.15, 7.18, 7.25, 7.31. Plate 3.
American Institute of Physics: Figures 4.32, 4.33, 4.36
American Meteorological Society: Figures 3.21, 3.23, 3.31, 3.38, 4.19a, 5.33a,
5.62, 5.63
Birkhäuser Verlag: Figure 3.22
Clarendon Press: Figures 5.5, 5.8
Committee for Space Research (COSPAR): Figure 7.28
Elsevier: Figures 4.15, 4.16, 4.17, 4.31, 4.41, 4.44, 5.34, 5.35, 7.3, 7.8, 7.9,
7.26, 7.32, 7.34, 7.36
European Geophysical Union: Figures 4.32, 7.19, 7.20. Plate 6.
European Space Agency (ESA): Figure 7.24
Federal Aviation Administration (FAA): Figures 3.3, 3.19, 5.7, 5.15, 5.25
Federal Institute of Technology, Zürich, Switzerland: Figures 3.7, 3.9
International Global Atmospheric Chemistry Project (IGAC) of the

International Geosphere Biosphere Programme (IGBP): Figure 5.70
National Center for Atmospheric Research (NCAR): Figure 4.22
625
National Center for Environmental Prediction (NCEP): Figure 6.19
National Research Council of Canada: Figure 5.41
National Space and Aeronautics Administration (NASA): Figures 2.5, 3.2, 3.5,
3.17, 4.20, 5.6, 5.47a, 7.7. Plates 2, 4, 5, 9, 12, 15.
Nature Publishing Group: Figure 3.28
Nonprofit Publisher of Annual Review off Series: Figures 2.6, 4.6, 4.10
Ohio Academy of Sciences: Figure 4.21
Plenum Press: Figures 7.21, 7.23, 7.27, 7.30
Princeton University Press: Figure 5.53
Rand Corporation: Figure 2.8
Springer Verlag including Kluwer and Reidel, Publishing Companies: Figures

3.20, 3.37, 4.2, 4.8a, 5.1, 5.2, 5.26a, 5.36, 5.49, 5.52, 5.68, 7.5, 7.10
University of Bern: Figures 7.4, 7.12, 7.17, 7.22, 7.28, 7.29, 7.35
University of Chicago Press: Figure 4.7
University of Colorado: Figure 7.33
University of Wyoming: Figure 5.64
Van Nostrand Publishing Company: Figure 2.1b
World Meteorological Organization (WMO): Figure 3.30
World Scientific: Figure 4.13a

COLOR SECTION
COLOR SECTION 629
Plate 1. Aurora photographed on 30

March 2001 at 51°N and 119°W
(British Columbia, Canada). Variations
in the color arise from the variation in
the spectra of the incoming particles.
These energetic particles originate from
the Sun and are accelerated in the
magnetosphere. Green and red light is
produced by radiative emissions asso-
ciated with electronic transitions of
atomic oxygen.
Photo by James W. Hannigan, National
Center for Atmospheric Research,
Boulder, Colorado.
Plate 2. Zonally averaged methane (CH H4) mixing ratio (ppmv) and the Brewer-Dobson circulation
(shown by the arrows) in the middle atmosphere (15-55 km altitude) during October. Methane is
released at the Earth surface and is destroyed by chemical reaction with the OH radical, while
transported in the troposphere and the stratosphere. The ascending branch of the Brewer-Dobson
circulation carries high-methane air from the tropical troposphere into the stratosphere, while the
descending branch carries air masses with low methane mixing ratios from the upper stratosphere
and mesosphere to the polar troposphere. From NASA/GSFC.
630 COLOR SECTION
Plate 3. Distribution of the nitrous oxide (N

N2O) mixing ratio (ppbv) on the 10 hPa pressure level or
approximately 30 km altitude as measured on November 6, 1994 by the CRISTA instrument on board
the Space Shuttle. The data were interpolated by a Kalman filter. N2O can be regarded as a tracer of
dynamical motions: for example a tongue of Arctic air extrudes from the North polar vortex across
the United States towards the Pacific Ocean. At the same time a "streamer" of N2O rich air extends
from the tropics across eastern Asia towards the Aleutian Islands and the west cost of Canada and
the United States. Another "streamer" propagates along the eastern cost of the North America
towards Europe. From Offermann et al., J. Geophys. Res., 104, 16311, 1999.
Plate 4. Ozone volume mixing ratio on the

417 K isentropic surface (approximately 16 km
altitude) on 31 December 1991. Filaments of
ozone-rich air pulled from the polar region are
irreversibly mixed into the mid-latitude surf
zone. The net effect is to weaken the latitudinal
ozone gradient created by the Brewer-Dobon
circulation. From NASA.
COLOR SECTION 631
Plate 5. Ozone anomalies (ppmv) versus altitude (km) and time (years) in the equatorial region
(4ºN-4ºS) derived from observations made by the HALOE instrument on board the Upper Altitude
Research Satellite (UARS). Superimposed on ozone values are the zonal winds measured by the
HRDI instrument on the same satellite. Full lines are eastward winds and dashed lines westward
winds with intervals corresponding to 10 m/s. In the 20-30 km altitude range, the positive ozone
anomalies are associated with the westerly shear in the quasi-biennal oscillation (QBO), while the
negative anomalies are indicative of easterly shears. Above 30 km, ozone variability is associated
with temperature variability, which affects the photochemical source terms. Above 35 km, the ob-
served variations are due to temperature effects associated with the QBO and to the semi-annual
oscillation. Courtesy of Paul Newman, NASA/GSFC.
Plate 6. Atmospheric transmittance in the infrared (600-2500 cm-1 or 4-17 µm) observed on 4 April,
1997 over the South Pacific by the Interferometric Monitor for Greenhouse Gases (IMG) on board
the Advanced Earth Observing Satellite (ADEOS) (upper panel). The three lower panels show the
contributions of water, other strong absorbers (CO2, O3, N2O, CO, CH H4), and weak absorbers (NO,
NO2, OCS, HNO3, CFC-11 and HCFC-22). Adapted from C. Clerbaux, J. Hadji-Lazaro, S. Turquety,
G. Megie, and P.-F. Coheur, Atmos. Chem. Phys., 3, 1495, 2003.
632 COLOR SECTION
Plate 7. Thermal infrared cooling rates [10-3K(day)-1(cm-1)-1] by H2O, CO2 (355 ppm), and O3
represented as a function of wavenumber (cm-1) and atmospheric pressure (hPa) for mid-latitude
summer conditions. The band centered at 667 cm-1 (15 µm) is dominated by CO2, and exhibits
strong stratospheric cooling and a local warming at the tropopause. The band around 1043 cm-1
(9.6 µm) is dominated by O3 and exhibits cooling in the upper stratosphere and warming throughout
the lower stratosphere. Water vapor makes a strong contribution to upper tropospheric cooling near
300 cm-1, although its effects are moderated by CO2. Adopted from Clough and Iacono (1995) with
courtesy of WCRP/SPARC.
Plate 8. Global view of the ozone column abundance (Dobson) on 30 November, 1999 as measured
by the GOME instrument - Values have been assimilated by the Royal Meteorological Institute of the
Netherlands. Note the very low ozone column abundance over Antarctica. Courtesy of Hennie
Kelder, KNMI, Netherlands.
COLOR SECTION 633
Plate 9. Interannual variability in the Antarctic ozone hole. The figures compare the ozone column
abundance observed on September 24 in 2001 and in 2002. In 2001 as in most previous years of the
last 2 decades, the hole was centered approximately over the South Pole with low ozone values over
the Antarctic continent. In 2002, a strong wave-2 planetary wave disturbed the polar vortex and
produced a very peculiar distribution of the ozone column. The warming associated with this event
led to limited ozone depletion and a disappearance of the ozone hole at the end of September. From
NASA.
Plate 10. Vertical column density (molecules cm-2) of nitrogen dioxide (NO2) measured over
Antarctica by the spaceborne GOME instrument between early and mid-November, 1997. The low
values (blue) are associated with the polar vortex, which is displaced towards South America as a
large planetary wave develops in the region (final warming). Strong latitudinal gradients can be
observed in the abundance of NO2 (called “Noxon Cliff” after J. F. Noxon who discovered them).
Courtesy of A. Richter and J. Burrows, Institute of Environmental Physics, University of Bremen,
Germany.
634 COLOR SECTION
Plate 11. Nitric Oxide density variation in the lower thermosphere at 107 and 97 km as measured by
the Student Nitric Oxide Explorer satellite using the γγ-band fluorescence technique. Percent
differences from the latitudinal average in the intertropical region at ~11:00 local time are shown.
At 107 km, solar control dominates with the peak density shifting to the summer hemisphere. At 97
km, this effect can still be discerned, but the dominant perturbation is now dynamical, as vertical
winds driven by middle atmosphere tides control the distribution. The diurnal tide drives changes in
the nitric oxide density at the equator, particularly at the equinoxes when tidal forcing tends to be
high. Courtesy of Stan Solomon, National Center for Atmospheric Research, Boulder, CO.
Plate 12. Evolution between 1979 and 2003 of the zonally averaged aerosol optical depth at 1000
nm represented as a function of latitude. The timing and location of several large volcanic eruptions,
which have affected the sulfate aerosol load of the atmosphere, are also shown. The value of the
optical depth (typically 10-3 during quite periods) increased dramatically after the major eruptions
of El Chichon in 1982 and Mt Pinatubo in 1991. The surface area available for heterogeneous
processes was also dramatically enhanced. The recovery time following stratospheric perturbations
produced by major volcanic eruptions is of the order of 2-3 years. Courtesy of Larry W. Thomason,
NASA/Langley.
COLOR SECTION 635
Plate 13. Noctilucent clouds observed over Kühlungsborn, Germany (54°N) on 10 July 1997 at
01:00 GMT. These polar mesospheric clouds are found at approximately 90 km altitude and are
illuminated by the Sun, which at that time is located approximately 11 degrees below the horizon.
Courtesy of F.-J. Lübken, Leibniz Institute for Atmospheric Physics, Kühlungsborn, Germany.
Plate 14. Mother-of-Pearl clouds (also called Polar stratospheric Clouds or PSCs) observed in
McMurdo Antarctica. Photo by David J Hofmann.
636 COLOR SECTION
Plate 15. Satellite observations of chlorine monoxide (ClO) and ozone (O3) within the lower
stratosphere (465 K potential temperature level) of the Southern Hemisphere on 30 August, 1996.
The white line delineates the edge of the vortex, while the black line on the chlorine monoxide panel
indicates the edge of the dark region over the pole. Chlorine monoxide levels are greatly enhanced
in the region between the white and the black lines – that is, in sunlit cold air within the Antarctic
polar vortex. The bottom panel shows the corresponding ozone distribution. Ozone abundances have
begun to drop inside the vortex where chlorine monoxide is elevated. The region displaying reduced
ozone (purple) corresponds very closely to the region of enhanced chlorine monoxide. From NASA.
INDEX
Absorption cross-sections, 43, 169 Bimolecular reactions, 26, 28-30

Accomodation coefficient, 34 BrCl (bromine chloride)
Actinic flux, 159 chemistry of, 370-372, 375-377
Activity complex, 29, 31 spectrum and photodissociation,
Adiabatic processes, 59-60 246
Advection (definition), 51 Brewer-Dobson circulation, 94-98
Aerosol particles BrO (bromine monoxide)
climatology, 394 chemistry of, 374-376
formation of, 393-394 observations, 386
linkages to polar stratospheric spectrum and photodissociation,
clouds, 396-401 243
microphysics, 393 BrONO2 (bromine nitrate)
and multiple scattering, 177-180 chemistry of, 376-377
observations of, 394-396 spectrum and photodissociation of,
solubility in, 23 244
sulfate layer, 388-396 Brunt-Väisälä frequency, 66
and volcanic eruptions, 395 Brry (odd bromine)
Age of air, 88-90 chemistry of, 374-377
Age spectrum, 89 interaction with Clx and O3, 404-
Agricultural emissions, 4-5 406
Airglow, 44 partitioning, 380-383
Antarctic ozone depletion
chemical processes, 407-408,
469-474 Catalytic cycles, 28
depth and area, 478-481 CFCs (chlorofluorocarbons)
observations, 462-465 emissions, 361-363
role of dynamics, 462-464 spectrum and photodissociation of,
theories, 465-470 241
Arctic oscillation (AO), 123-125 trends, 503-504
Arctic ozone depletion CH2O (formaldehyde)
chemical processes, 484-486 chemistry of, 298-302
quantification, 486-488 spectrum and photodissociation of,
role of dynamics, 482-483 233
temperature variability, 488-491 CHBr3 (bromoform)
Atmospheric sphericity, 174-176 spectrum and photodissociation of,
241-242
CH3Br (methyl bromide)
Band models, 200-203 budget, 360-361
Beer-Lambert Law, 169 spectrum and photodissociation of,
638 INDEX
241-242 chemistry of, 364-372

CH3Cl (methyl chloride) spectrum and photodissociation,
budget, 360-361 243
spectrum and photodissociation of, ClONO2 (chlorine nitrate)
241 chemistry of, 374-372
CH3CCl3 (methylchloroform) in Antarctic stratosphere, 473
emissions, 361-363 lifetime of, 381-382
spectrum and photodissociation of, observations of, 384-385
241 ClONO2 (continued) d
trends, 503-504 spectrum and photodissociation,
CH3CN (methylcyanide), 335-336 244
CH3I (methyliodide) Clx (reactive chlorine)
budget, 360-361, 379 chemistry of, 364-372
CH4 (methane) observations of, 384-387
budget of, 296-298 partitioning of, 380-384
chemistry of, 298-302 Cly (odd chlorine)
lifetime of, 294-295 chemistry of, 364-372
observations of, 294-295 cycle, 6-7
oxidation of, 299-302 destruction of ozone by, 405-408,
spectrum and photodissociation of, 455-459
231-232 observations of, 384-387
transport of, 294-296 partitioning of, 380-384
Chapman function, 174-175 trends, 503-505
Chemical lifetime, 26-27 CO (carbon monoxide)
Chemical-transport models, 131-136 budget of, 303-304
Chemical algorithms, 269-271 chemistry of, 298-303
Chemical heating, 210-211 lifetime of, 304-305
Chemiluminescence, 44 observations of, 305-307
Chromosphere, 162 transport of, 305
Circulation CO2 (carbon dioxide)
diabatic, 99 absorption and emission of IR
Eulerian mean, 92-93 radiation by, 192
Lagrangian mean, 99 absorption of solar radiation by,
transformed Eulerian mean or 169-177
residual, 99 abundance of, in present
Cl (atomic chlorine) atmosphere, 305
chemistry of, 364-372 chemistry of, 302
diurnal variation, 372-373 effect of industrial activities, 6
lifetime of, 381-382 non-LTE conditions, 203-206
ClO (chlorine monoxide) on geological timescales, 3
chemistry of, 364-372 spectrum and photodissociation of,
diurnal variation, 372-373 232-233
in Antarctic stratosphere, 472 Column abundance, 267
lifetime of, 381-382 Continuity equation, 60, 84-85, 266-
observations of, 384-385 268
spectrum and photodissociation of, Coriolis force, 59, 67
242-243 COS (carbonyl sulfide), 388-390
Cl2O2 (dichlorine peroxide or ClO Cosmic rays
dimer) ionization by, 546-548
INDEX 639
production of NOx by, 334 F-region of the ionosphere, 534-535

Ferrell cell, 95
Fluorescence, 44
D-region of the ionosphere Free radicals, 11
definition of, 534-535 Fy (odd fluorine)
ionization in, 542-544 chemistry of, 372-374
negative ion chemistry in, 573- partitioning, 380-383
576
positive ion chemistry in, 562-564
Diabatic General circulation models, 129-131
circulation, 99 Geopotential, 67
definition, 60 Geostrophic approximation, 69-70
heating rates, 158 Gibbs free energy of reactions, 22
Diffusion Gravity waves 75-77
eddy, 58, 86, 128 Grid-point models, 131
molecular, 127-129
Doppler lineshape, 197
Downward control principle, 113 H (atomic hydrogen)
Dynamical barriers, 106-112 chemistry of, 317-325
lifetime of, 321-322
H2 (molecular hydrogen)
E-region of the ionosphere budget of, 313-316
definition of, 534-535 chemistry of, 313
ionization in, 539-542 lifetime of, 316
positive ion chemistry in, 554-555 observations of, 315-316
Eddies H2O (water vapor)
definition, 92 absorption and emission of IR
diffusion, 58, 86, 128 radiation by, 192
chemical eddy transport, 102 chemistry of, 310-311
Electromagnetic spectrum, 151-152 lifetime of, 312
Electronic transitions long-term trends in, 313
term symbols, 18-20 observations of, 311-313
selection rules, 20 spectrum and photodissociation of,
Eliassen-Palm flux, 100-101 227-231
Endothermicity, 21 and transport processes, 313
Energy budget, 216-218 H2O2 (hydrogen peroxide)
Energy equation, 59 chemistry of, 317-325
Energy levels of molecules, 12-20 lifetime of, 321-322
Energy of activation, 29 observations of, 325-327
Enthalpy of formation, 21 spectrum and photodissociation of,
Entropy, 22 231
Equation of state, 60 H2SO4 (sulfuric acid)
Eulerian mean, 92-93 and aerosols, 393-398
Evolution of the atmosphere, 2-4 chemistry of, 390-391
Excited species, 44-48 observations of, 393-396
Exothermicity, 21 Halons
Extratropical pump, 104-105 spectrum and photodissociation of,
241-243
640 INDEX
use of, 361-363 HO2 (hydroperoxy radical)

HBr (hydrogen bromide) chemistry of, 317-325
chemistry of, 376-377 lifetime of, 321-322
lifetime of, 321-322 observations of, 325-327
spectrum and photodissociation of, HOBr (hypobromous acid)
244-245 chemistry of, 376-377
HCFCs (hydrochlorofluorocarbons), spectrum and photodissociation,
361-363 245
HCl (hydrogen chloride) HOCl (hypochlorous acid)
chemistry of, 364-372 chemistry of, 364-372
in Antarctic stratosphere, 473 observations of, 384
lifetime of, 321-322 spectrum and photodissociation of,
observations of, 384-387 245
spectrum and photodissociation of, Homosphere, 53
244-245 HOx (odd hydrogen)
HCN (hydrogen cyanide), 335 chemistry of, 317-325
Heat conduction, 127 destruction of ozone by, 401-403
Heterogeneous processes, 338, 368, lifetime of, 321-322
377, 469 observations, 325-327
Heterogeneous reactions, 34-40 partitioning of, 321-325
Heterosphere, 53 production of, by ions, 585-586
HF (hydrogen fluoride) Hydrostatic approximation, 63
chemistry of, 372-374 Infrared (IR) radiation
observations of, 386 absorption and emission of, 191-
spectrum and photodissociation of, 206
244-245 source of, 191
HFCs (hydrofluorocarbons), 361-363
H+.(H2O)n (proton hydrate ions)
formation of, 562-563 Ion drag, 114-115
observations of, 559-562 Ionization
production of HOx bym 585-586 by solar radiation, 539-544
HNO2 (nitrous acid) by energetic particles, 544-552
spectrum and photodissociation of, Ionosphere, 533-539
239 Irradiance, 158
HNO3 (nitric acid) Isentropic coordinates, 68
chemistry of, 336-341 Iy (odd iodine)
lifetime of, 342-343 chemistry of, 379-380
observations of, 350 partitioning, 383
spectrum and photodissociation of,
239
HNO4 (peroxynitric acid) Kelvin waves, 80
chemistry of, 336-341
lifetime of, 342-343
spectrum and photodissociation of, Lapse rate, 65
240 Lifetime, 267
INDEX 641
Lorentz line shape, 197 observations of, 349
spectrum and photodissociation of,
239
Mass density (definition), 267 Negative ions
Mass fraction (definition), 267 in the D-region, 571-577
Mean thermal molecular velocity, 29, observations, 571-572, 579-581
34 ratio to electrons, 537-539
Mechanistic models, 130 in the stratosphere, 577-581
Meridional wind (definition), 57 Newtonian cooling, 215
Mesopause (definition), 53, 116- 117 Nitrogen cycle, 5
Mesosphere (definition), 53 NO (nitric oxide)
Metal ions, 555-559 chemistry of, 336-341, 353
Mid-latitude ozone depletion diurnal variation of, 346
gas-phase chemistry, 401-407 and the ionospheric D-region,
heterogeneous chemistry, 492-499 542-543
dynamical processes, 499-501 lifetime of, 342-343
Mie scattering, 178 observations of, 346-347, 355,
Mixed Rossby-gravity waves, 80 358
Mixing ratio (definition), 84, 267 spectrum and photodissociation of,
Molecular diffusion, 127-128 233-236
Mole fraction (definition), 267 in the thermosphere, 351-358
Momentum equation, 59 NO2 (nitrogen dioxide)
Montgomery stream function 68 absorption of solar radiation and
Montreal protocol, 7 heating by, 211
chemistry of, 336-341
diurnal variation of, 346
N(2D) (excited nitrogen atom) lifetime of, 342-343
emissions from, 47 observations of, 347-348
and thermospheric NO, 351-355 spectrum and photodissociation of,
N2 (molecular nitrogen) 237-238
abundance of, 266 NO3 (nitrogen trioxide)
spectrum and photodissociation of, chemistry of, 336-341
219-220 diurnal variation of, 346
N2O (nitrous oxide) lifetime of, 342-343
budget of, 329-330 observations of, 350
chemistry of, 328-329 spectrum and photodissociation of,
lifetime of, 330-331 238-239
observations of, 330-332 Noble gases (Ar, Ne, He, Kr, Xe)
production of stratospheric NO atmospheric mixing ratio, 266
from, 329 depletion on Earth relative to
spectrum and photodissociation of, cosmic abundance, 2
236-237 Non-acceleration theorem, 96
N2O5 (dinitrogen pentoxide) NOx (reactive nitrogen)
chemistry of, 336-341 chemistry of, 336-341
diurnal variation of, 346 destruction of ozone by, 403-404
lifetime of, 342-343 lifetime of, 342-343
642 INDEX
production of, by ions, 582-585 absorption of solar radiation and

stratospheric sources of, 328-335 heating by, 173-208
NOy (odd nitrogen) absorption and emission of IR
definition, 341-342 radiation by, 192
observations of, 351 and CH4, 413
partitioning of, 341-346 chemistry of, 273-281
stratospheric sources of, 328-335 depletion of, 505-506
Number density (definition), 267 general equation, 413-416
historical perspective, 272
hole, 462-464
OClO (chlorine dioxide) lifetime of, 276-277
in Antarctic stratosphere, 472-473 observations of, 281-292
chemistry of, 364-372 and particle precipitation, 448-
spectrum and photodissociation, 451
243 protection of living things by, 1
O(1D) (electronically excited oxygen and the QBO, 291
atom) and solar variability, 292, 451-
chemistry of, 273-281 455
lifetime, 276 spectrum and photodissociation of,
production of, 44 224-227
production of HOx by, 44, 311 and transport processes, 286-287,
red line emission, 45 290
O(1S) (electronically excited oxygen in the troposphere, 409-413
atom) and volcanic influences, 494-498
green line 45 Ox (odd oxygen)
O2 in the primitive atmosphere, 3 chemistry of, 274-281
O(3P) (atomic oxygen) lifetime of, 277
chemistry of, 273-281 partitioning of, 279
lifetime of, 276 Optical thickness, 170
OH (hydroxyl radical) Optical transmission, 170
chemistry of, 317-325 Order of reactions, 25
lifetime of, 321-322 Oxidizing power of the atmosphere,
Meinel bands of, 46 410-412
observations of, 325-327
O2 (molecular oxygen)
absorption of solar radiation and Particle precipitation
heating by, 173-208 effect on neutral species, 581- 586
abundance of, 266 ionization by, 551-552
chemistry of, 273-281 penetration depth of, 545
emission of O2(1ǻg), 46 Photodissociation frequency, 43, 218-
potential energy curves, 42 219
spectrum and photodissociation of, Photolysis, 40-44, 246-249
220-224 Photosphere, 160
O3 (ozone) Planetary waves, 78-79
INDEX 643
Polar mesospheric clouds (PMCs), Remote sensing of the atmosphere,
396 188-190, 195-196
Polar stratospheric clouds (PSCs) Rossby number, 70
composition and chemistry, 397- Rossby waves, 78-79
399, 474-478 Rotational energy, 16
effects on chlorine activation,
469-470 Scale height, 63
observations, 397 Semi-annual oscillation (SAO), 123
thermodynamics, 399-401 SF6 (sulfur hexafluoride), 364
Polar vortex, 107, 462-464 Single scattering albedo, 177
Potential vorticity, 71-74 Solar atmosphere, 159-162
Positive ions Solar constant, 162
in the D-region, 559-564 Solar proton events (SPE)
in the E-region, 554-559 ionization by, 549-550
in the stratosphere, 564-570 production of HOx by, 448-451
observations, 555-556, 559-562, production of NOx by, 334-335,
568-570 448-451
Potential energy curves, 42-43 Solar radiation
Potential temperature, 61 absorption of, 169-177
Predissociation, 43, 222, 234 heating by, 207-212
Primitive atmosphere, 2-4 scattering of, 177-180
Primitive equations, 59-69 spectral distribution of, 164-168
Quasi-biennal oscillation variation of, 167-169
description, 121-123 Solar variability
effect on ozone, 291 amplitude, 167-169
effects on middle atmosphere,
451-455
Radiance, 156 Solar zenith angle, 170
Radiative-convective models, 206 Solubility of gases, 23-24
Radiative equilibrium, 112 Sources of ionization, 539-552
Radiative lifetime, 215-216 Spectral model, 131
Radiative transfer Steady (photostationary state), 27
equation 180-181 Sticking coefficient, 34
multiple scattering 182-187 Stratopause (definition), 53
infrared absorption and emission Stratosphere (definition), 53
191-206 Structure of the atmosphere, 53-58
Radio waves, 586-589 Sudden stratospheric warming, 120-
Rate of chemical reaction, 25-26 121
Rate-limiting step, 28 Sulfur compounds
Rayleigh friction, 114 chemistry of, 389-391
Rayleigh scattering, 178 sources of, 388-389
Reaction probability, 34 Surface area density, 35
Recombination, 534, 576
Relativistic electron precipitation
(REP), 548-549 Tape-recorder effect, 106, 313
Temperature
644 INDEX
and definition of atmospheric Uptake coefficient, 34

layers, 53
and ozone coupling, 444-448
Termolecular reactions, 26, 32-33 Vibrational energy, 14-16
Terrestrial radiation Vibration-rotation transitions, 16
cooling by, 212-216 Voigt lineshape, 198
absorption and emission of, 191- Volcanoes
206 general, 389
Thermal wind, 70 and sulfate layer, 395-396
Thermosphere (definition), 53 and ozone depletion, 494-499
Tides 80-83 Vorticity, 71
Time constants
dynamical, 85-88
radiative, 215 Waves 74-83
photochemical, 26-27, 87-88 Wave drag
Tracer correlations, 90-92 by gravity waves, 117-119
Transport algorithms, 134-135 by Rossby waves 119-121
Tropopause Winter anomaly, 590
definition, 53
description, 111-112
Troposphere (definition), 53 Zonal wind (definition), 55
Unimolecular reactions, 25-26, 30-

32
1. F.T.M. Nieuwstadt and H. van Dop (eds.): Atmospheric Turbulence and Air
Pollution Modelling. 1982; rev. ed. 1984
ISBN 90-277-1365-6; Pb (1984) 90-277-1807-5
2. L.T. Matveev: Cloud Dynamics. Translated from Russian. 1984
ISBN 90-277-1737-0
3. H. Flohn and R. Fantechi (eds.): The Climate of Europe: Past, Present and Future.
Natural and Man-Induced Climate Changes: A European Perspective. 1984
ISBN 90-277-1745-1
4. V.E. Zuev, A.A. Zemlyanov, Yu.D. Kopytin, and A.V. Kuzikovskii: High-Power Laser
Radiation in Atmospheric Aerosols. Nonlinear Optics of Aerodispersed Media. Trans-
lated from Russian. 1985 ISBN 90-277-1736-2
5. G. Brasseur and S. Solomon: Aeronomy of the Middle Atmosphere. Chemistry and
Physics of the Stratosphere and Mesosphere. 1984; rev. ed. 1986
ISBN (1986) 90-277-2343-5; Pb 90-277-2344-3
6. E.M. Feigelson (ed.): Radiation in a Cloudy Atmosphere. Translated from
Russian. 1984 ISBN 90-277-1803-2
7. A.S. Monin: An Introduction to the Theory of Climate. Translated from Russian. 1986
ISBN 90-277-1935-7
8. S. Hastenrath: Climate Dynamics of the Tropics, Updated Edition from Climate and
Circulation of the Tropics. 1985; rev. ed. 1991
ISBN 0-7923-1213-9; Pb 0-7923-1346-1
9. M.I. Budyko: The Evolution of the Biosphere. Translated from Russian.
1986 ISBN 90-277-2140-8
10. R.S. Bortkovskii: Air-Sea Exchange of Heat and Moisture During Storms. Translated
from Russian, rev. ed. 1987 ISBN 90-277-2346-X
11. V.E. Zuev and V.S. Komarov: Statistical Models of the Temperature and Gaseous
Components of the Atmosphere. Translated from Russian. 1987
ISBN 90-277-2466-0
12. H. Volland: Atmospheric Tidal and Planetary Waves. 1988 ISBN 90-277-2630-2
13. R.B. Stull: An Introduction to Boundary Layer Meteorology. 1988
ISBN 90-277-2768-6; Pb 90-277-2769-4
14. M.E. Berlyand: Prediction and Regulation of Air Pollution. Translated from
Russian, rev. ed. 1991 ISBN 0-7923-1000-4
15. F. Baer, N.L. Canfield and J.M. Mitchell (eds.): Climate in Human Perspective. A tribute
to Helmut E. Landsberg (1906-1985). 1991 ISBN 0-7923-1072-1
16. Ding Yihui: Monsoons over China. 1994 ISBN 0-7923-1757-2
17. A. Henderson-Sellers and A.-M. Hansen: Climate Change Atlas. Greenhouse Simu-
lations from the Model Evaluation Consortium for Climate Assessment. 1995
ISBN 0-7923-3465-5
18. H.R. Pruppacher and J.D. Klett: Microphysics of Clouds and Precipitation, 2nd rev. ed.
1997 ISBN 0-7923-4211-9; Pb 0-7923-4409-X
19. R.L. Kagan: Averaging of Meteorological Fields. 1997 ISBN 0-7923-4801-X
20. G.L. Geernaert (ed.): Air-Sea Exchange: Physics, Chemistry and Dynamics. 1999
ISBN 0-7923-5937-2
21. G.L. Hammer, N. Nicholls and C. Mitchell (eds.): Applications of Seasonal Climate
Forecasting in Agricultural and Natural Ecosystems. 2000 ISBN 0-7923-6270-5
22. H.A. Dijkstra: Nonlinear Physical Oceanography y. A Dynamical Systems Approach to

the Large Scale Ocean Circulation and El Nino.˜ 2000 ISBN 0-7923-6522-4
23. Y. Shao: Physics and Modelling of Wind Erosion. 2000 ISBN 0-7923-6657-3
24. Yu.Z. Miropol’sky: Dynamics of Internal Gravity Waves in the Ocean. Edited by O.D.
Shishkina. 2001 ISBN 0-7923-6935-1
25. R. Przybylak: Variability of Air Temperature and Atmospheric Precipitation during a
Period of Instrumental Observations in the Arctic. 2002 ISBN 1-4020-0952-6
26. R. Przybylak: The Climate of the Arcticc. 2003 ISBN 1-4020-1134-2
27. S. Raghavan: Radar Meteorology y. 2003 ISBN 1-4020-1604-2
28. H.A. Dijkstra: Nonlinear Physical Oceanography y. A Dynamical Systems Approach to
the Large Scale Ocean Circulation and El Niño.˜ 2nd Revised and Enlarged Edition.
2005 ISBN 1-4020-2272-7 Pb; 1-4020-2262-X
29. X. Lee, W. Massman and B. Law (eds.): Handbook of Micrometeorology y. A Guide for
Surface Flux Measurement and Analysis. 2004 ISBN 1-4020-2264-6
30. A. Gelencsér: Carbonaceous Aerosol. 2005 ISBN 1-4020-2886-5
31. Soloviev: The Near-Surface Layer of the Ocean. Structure, Dynamics and Applica-
tions. 2005 ISBN 1-4020-xxxx-x
32. G.P. Brasseur and S. Solomon: Aeronomy of the Middle Atmosphere. Chemistry and
Physics of the Stratosphere and Mesosphere. 2005
ISBN 1-4020-3284-6; Pb 1-4020-3285-4
springeronline.com

Aeronomy of The Middle Atmosphere - 2005

Uploaded by

Copyright:

Available Formats

Aeronomy of The Middle Atmosphere - 2005

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Aeronomy of The Middle Atmosphere - 2005

Uploaded by

Copyright:

Available Formats

ATMOSPHERIC AND OCEANOGRAPHIC SCIENCES LIBRARY

Lawrence A. Mysak, Department of Atmospheric and Oceanographic Sciences,

Editorial Advisory Board

L. Bengtsson Max-Planck-Institut für Meteorologie, Hamburg, Germany

Library of Congress Cataloging-in-Publication Data

ISBN-10 1-4020-3284-6 (HB)

Printed on acid-free paper

All Rights Reserved

Printed in the Netherlands.

1. THE MIDDLE ATMOSPHERE AND ITS EVOLUTION 1

2. CHEMICAL CONCEPTS IN THE ATMOSPHERE 11

3. DYNAMICS AND TRANSPORT 51

5. COMPOSITION AND CHEMISTRY 265

6. OZONE PERTURBATIONS 443

7. THE IONS 533

Appendix 3: Rate Coeﬃcients for Second-Order Gas-Phase

Figure Acknowledgements 625

Color S ection 627

Since the industrial revolution, the chemical composition of the

of the atmosphere to destroy several chemical compounds, including

ozone hole is provided in this chapter. Finally, Chapter 7 provides an

at the National Center for Atmospheric Research (NCAR) for their

THE MIDDLE ATMOSPHERE

important role, particularly in the altitude region which we shall address

1.2 Evolution of the Earth’s Atmosphere

The terrestrial atmosphere has evolved considerably over the course

energy available, the planet’s average temperature was always higher

1.3 Anthropogenic Perturbations

The possible modiﬁcation of the chemical composition of the

ﬁgure is a signiﬁcant fraction of the 1 Tg N produced naturally from

lifetimes to represent a potential perturbation. This is the case, for

Numerical models have been used to predict the potential ozone

Almost all of the constituents that are present in the middle

2.2 Energy Levels of Molecules

The motions of molecules can be decomposed into translational,

Box 2.1 Some Useful Deﬁnitions

Box 2.1 (Continued)

Box 2.1 (Continued)

Figure 2.1.a presents a typical potential energy diagram for a

Figure 2.1.a. Typical potential energy diagram for a hypothetical diatomic

molecule, or as a result of a chemical reaction (see e.g., reaction 2.95).

Figure 2.1.b. Schematic diagram of the energy levels and vibration-rotation

Figure 2.2. Vibrational modes of diatomic and triatomic molecules.

2.2.1 Term Symbols

In general, term symbols indicate both spin angular momentum (S

(1s)2 (2s)2 (2p1 )2 (2p2 )1 (2p3 )1

The superscripts denote the number of electrons in each atomic

The two πg 2p electrons are unpaired, so that 2Σ + 1 = 3, and the

2.2.2 Selection Rules for Electronic Radiative Processes

molecules. This is generally the most important consideration in

if 0 → 0 (S → S). Changes of L or Λ of more than one (i.e., D → S)

but u → u or g → g transitions are diﬃcult.

Σ+ → Σ− transitions are unfavorable.

Examples of transitions which are of importance in the middle

2.3 Thermodynamic Considerations

2.3.1 Enthalpy, Entropy, and Free Energy

The basic principles governing the reactivity of chemical substances

where T is the temperature and Cp is the heat capacity at constant