3319256777

David S.
Stevenson
The Exo-
Weather
Report
Exploring Diverse
Atmospheric Phenomena
Around the Universe
Astronomers Universe
More information about this series at http://www.springer.com/series/6960

David S. Stevenson
The Exo-Weather Report

Exploring Diverse Atmospheric
Phenomena Around the Universe
David S. Stevenson
Carlton le Willows Academy
Nottingham, UK
ISSN 1614-659X ISSN 2197-6651 (electronic)

Astronomers Universe
ISBN 978-3-319-25677-1 ISBN 978-3-319-25679-5 (eBook)
DOI 10.1007/978-3-319-25679-5
Library of Congress Control Number: 2016943436
Springer International Publishing Switzerland 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole
or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical
way, and transmission or information storage and retrieval, electronic adaptation, computer
software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information
in this book are believed to be true and accurate at the date of publication. Neither the
publisher nor the authors or the editors give a warranty, express or implied, with respect to
the material contained herein or for any errors or omissions that may have been made.
Cover illustration: Cover photo by Flickr user Groman123 used under a CC BY-SA 2.0
license
Printed on acid-free paper
This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland
This book is dedicated to my sisters Avril
Stevenson-Davies, Karen Suzuki, and Mairi Allardice
and their families; and to my cousin Lesley Duncan.
Preface
My father fought in both World Wars. In 1915, aged 16 he joined

the British Navy, serving on HMS Orion. Training in communica-
tions, he ultimately took the surrender of the German Fleet, pass-
ing the terms of surrender to the German High Command (Fig. 1).
During the Second World War, my father served in the merchant
fleet and took part in the evacuation of Dunkirk. The science
of meteorology took off during the First World War, because it
became understood that the weather was a key variable in win-
ning battles. There were obvious patterns, such as the lowering
of the sky ahead of a rain, but the term front had yet to be coined,
except outside the unpleasant confines of the Trenches. Changes
in wind direction and cloud cover were noted as rain bands came
and went as well as the turning of the seasons (Fig. 1).
However, it wasnt until the late nineteenth century before
much of the underlying science was known and not until the
1920s that our understanding of frontal boundaries emerged. Later
still, in the 1930s and 1940s the jet streams were discovered and
the driving force behind much of the movement of our weather
became apparent. My father taught me the nature of frontal sys-
tems that he had learnt while at sea, along with the names of all
of the clouds. However, the idea that other planets might have
weather was still in its infancy.
Until the Soviet Venera probes landed on Venus, it was
thought rather likely Venus was a rather pleasant tropical world.
That Venus could experience a runaway greenhouse effect, driven
by carbon dioxide, was simply not understood. John Tyndall may
have done the first experiments with various gases as early as
1850, but the implications of such discoveries were not realized
until much later. Indeed, the idea that altering the concentration
of greenhouse gases can have an impact on terrestrial climate is
still rather contentious in some circles.
vii
viii Preface
FIG. 1 The original transcript of the terms of surrender of the German fleet
in 1918, taken by my father David Stevenson. My father rather mischie-
vously kept a copy of the transcript he recorded to be passed onto the fleet
admiral. In amongst the tragedy of the First World War, and in large part
because of it, the science of meteorology really took off. Respective Navies
and fledgling air forces needed to know how the weather would impact
their activities
We are at a point that we have gathered a lot of data on the

atmospheres of the planets in our solar system and are beginning
to gather information on planets hundreds or thousands of light
years away. Through a greater understanding of our planet, we can
see how the climate of other planets is affected by the composi-
tion of their atmosphere; the influence of their parent star and
the proximity of such worlds to their stars. Take Kepler 452b, for
example. This world is a fairly good match for the Earth a bil-
lion years into our future. As its Sun-like star slowly advances its
years, Kepler 452b is being stranded on the hot side of its stars
habitable zone. Our observations of this world will paint a more
accurate picture of how our world will eventually find its habit-
ability coming to an end.
Meteorology is a broad physical science, but this breadth is
its strength. It encompasses simple, generic observational skills,
through to complex mathematical modeling of fluids. Models are
best tested by comparison with observation and it often turns out
Preface ix
that simple observations can provide some very useful truths. In

Chap. 1 I describe a workable model for predicting winter weather
in the UK several months in advance. This is testable, which is the
hallmark of good science. However, it can be underpinned by solid
mathematics, something I could with brushing up on.
Ive arranged the chapters in increasingly hierarchical but
overlapping structures that provide clear links between phenom-
ena on all the worlds described. Lightning is introduced in Chap.
4, but links to all of the subsequent chapters (bar Mars) on the
worlds of our solar system. The greenhouse effect emerges in
Chap. 3 then links to the others, while terrestrial monsoons are
described in Chap. 2 before linking through to the climate of Mars
and extrasolar worlds. Hopefully, this approach provides clear
links and shows the interconnectedness of the underlying science.
Meteorology and climate science are truly fascinating aspects
of planetary science where we humble earthlings get to experience
firsthand how the physics of other worlds plays out.
Nottingham, UK David S. Stevenson

xi
Contents
1. What We Know About the Weather on Earth ................. 1

Why Do We Have Weather? ............................................. 1
The Highs and Lows of Meteorology ............................... 3
Wind Direction: Waterwheels and a Suspect Tale
from the Front ................................................................... 4
The Vertical Structure of the Earths Atmosphere .......... 10
The Language of War-Fronts ......................................... 15
The Jet Stream ............................................................... 19
Rotation, Rotation, Rotation............................................ 24
Rossby Waves .................................................................... 25
Other Jets That Drive Terrestrial Weather ...................... 34
How the Jet Stream Brought a Sting
to the Bergen Frontal Model ............................................. 40
Fastnet ............................................................................ 40
The New Zealand Bomb................................................ 42
The 1993 US Superstorm .............................................. 42
The Tropics: A Quick Guide ......................................... 44
Interconnections ............................................................... 47
Conclusions....................................................................... 53
References ......................................................................... 54
2. Climate Oscillations in Space and Time ......................... 55

Introduction ...................................................................... 55
The Present Monsoon ....................................................... 56
Shifting Continents, Shifting Climate............................. 59
Probing the Past Asian Monsoon ..................................... 60
xii Contents
Milankovi Cycles: Astronomical Influences

on Terrestrial Climate ...................................................... 64
The Death of the Tethys and the Birth of the Sahara ..... 66
Pangean Monsoons ........................................................... 68
Problematic Children: El Nio and La Nia ................... 70
The Future East Asian Monsoon Under the Cloud
of Global Dimming ........................................................... 79
How Global Dimming Caused Geopolitical Chaos
in the Dark Ages ............................................................... 85
Conclusions....................................................................... 90
References ......................................................................... 90
3. Tales of Mass Destruction ................................................ 91

Introduction ...................................................................... 91
Global Greenhouses: Eocene, Permian
and Anthropocene ............................................................. 91
The Eocene Climate Maximum.................................... 92
A Bad Day in the Permian............................................. 96
The Human Factor......................................................... 104
Anthropogenic Global Warming ................................... 105
Dissension ......................................................................... 105
Drifting Hurricanes Amid a Changing Climate........... 118
Endemic Misunderstandings ............................................ 119
The Medieval Tepid Period and the Luke-Warm
Little Ice Age .................................................................. 120
The Effect of Global Climate Catastrophes on Life..... 123
An Icy Future?................................................................ 129
Visions of Hell: Terrestrial Snowballs and Fireballs ....... 133
Snowballs and Slushballs .............................................. 134
Raindrops Keep Falling on My Head:
The Controversial Climate of the Early Earth ............. 139
Parallel Lives: The Formation of the Earth
and Venus ....................................................................... 142
The Last Icehouse ............................................................. 149
The Rise of Amasia ........................................................... 150
Geo-Engineering the Future ............................................. 154
Conclusions: The Big Picture ........................................... 156
References ......................................................................... 158
Contents xiii
4. Weird Weather ................................................................... 159

Introduction ...................................................................... 159
Strange Lights from Thunderstorms ................................ 159
Earthlights or Headlights? ................................................ 169
Fata Morgana or Something Else? .................................... 172
Earthquake Clouds and Lights ......................................... 174
Sprites, Jets and Other Luminous Atmospheric
Phenomena ........................................................................ 179
Gamma Ray Bursts on Earth ............................................ 182
Sky Quakes........................................................................ 185
The Red Rains of Kerala ................................................... 186
Flying Spiders .................................................................... 187
Conclusions....................................................................... 188
References ......................................................................... 188
5. Venus ................................................................................. 191

Introduction: The Twin That Isnt ................................... 191
The Venusian Day............................................................. 191
A Noxious Vision of Hell .............................................. 193
The Structure of the Venusian Atmosphere .................... 195
No Layer but Yes, Ozone Around Venus ...................... 201
The Edge of Space .......................................................... 201
Changes to Wind Speed at Venus.................................. 205
Snow on Venus? ................................................................ 206
Lightning on Venus ........................................................... 208
Why Does Venus Have So Much More
Carbon Dioxide than the Earth? ...................................... 212
Constructing a Dry Planet................................................ 212
Life in Hell? ....................................................................... 218
Future Venus: The Earth and Venus, Twins
Once More ......................................................................... 218
Conclusions....................................................................... 229
References ......................................................................... 230
6. The Wispy Weather of Mars ............................................. 231

Introduction ...................................................................... 231
The Structure and Formation of the Martian
Atmosphere ....................................................................... 231
xiv Contents
Dust Devils, Tornadoes and the Other Whirling

Winds of Earth and Mars .................................................. 240
Aurora ................................................................................ 248
Mystery Clouds Across the Southern, Morning
Terminator ..................................................................... 252
Martian Ozone ............................................................... 255
Martian Methane: Mars not Dead, Yet ......................... 256
The Lingering Death of Mars ........................................ 259
Future Mars ....................................................................... 268
Conclusions....................................................................... 270
References ......................................................................... 272
7. The Gas Giants ................................................................. 273

Introduction ...................................................................... 273
The Structure of the Giants ............................................. 273
The Color of Giant Planets in the Solar System ............. 277
Belts and Braces: The Bands and Storms of Jupiter ......... 282
Great Red Spot and Other Vortices .................................. 285
Derechos and the Storm That Ate Itself .......................... 290
The Polar Hexagons .......................................................... 302
Ice Fall from the Rings...................................................... 304
Ios Electrifying Connection to Jupiter ............................ 305
Is Mighty Jupiter Losing Any Atmosphere? .................... 309
Conclusions....................................................................... 310
References ......................................................................... 310
8. The Ice Giants................................................................... 311

Introduction ...................................................................... 311
Inside the Ice Giants ......................................................... 312
The Structure of Uranus and Neptunes
Atmospheres ..................................................................... 314
Seasons .............................................................................. 320
Dark Spots ......................................................................... 322
The Twisted Tale of Ammonium Metal .......................... 323
Conclusions....................................................................... 327
References ......................................................................... 328
Contents xv
9. Ice Dwarves: Titan, Triton and Pluto .............................. 329

Introduction ...................................................................... 329
Titan .................................................................................. 329
Introduction ................................................................... 329
General Structure of Titans Atmosphere .................... 330
The Weather Report for Titan ....................................... 332
Microbursts and Aeolian Features on Titan ................. 336
Titans Missing Vital Spark ........................................... 338
The Methanological Cycle ........................................ 340
Rat Poison ...................................................................... 340
The Loss of Titans Atmosphere ................................... 342
Triton ................................................................................. 347
Introduction ................................................................... 347
Tritons Atmosphere ...................................................... 348
Whats the Weather Like? ............................................. 350
Pluto .................................................................................. 352
Introduction ................................................................... 352
Pluto and Triton as Non-identical Twins ..................... 354
Conclusions....................................................................... 360
References ......................................................................... 361
10. Tales of Other Worlds ....................................................... 363

Introduction ...................................................................... 363
Exoplanets by the Bucket .............................................. 363
The Climate of Tidally-Locked Planets:
Assumptions and Expectations ..................................... 365
The Structure of the Atmosphere of Jupiter-Like
Worlds: Too Hot, Too Cold, or Just-Right .................... 366
The Mystery of Super-Rotation .................................... 372
Tidally-Locked Earths and Super-Earths:
Towards A Realistic Planetary Model .......................... 396
Earth-Like but not Tidally-Locked ............................... 406
Tipped Over Worlds ....................................................... 410
Atmospheric Gravity Waves ............................................ 415
Atmospheric gravity waves and Earthquakes .............. 420
Weather, Unbound: Brown Dwarfs, Stars,
Galaxies and Galaxy Clusters ....................................... 420
Solar Wind and Stellar Gales......................................... 422
xvi Contents
Galactic Winds............................................................... 424

Cold Fronts in Galaxy Clusters .................................... 425
Conclusions....................................................................... 427
References ......................................................................... 428
Glossary ..................................................................................... 431
Index .......................................................................................... 451

1. What We Know About the
Weather on Earth
Why Do We Have Weather?

Weather is such a ubiquitous part of our lives that most of us take
it completely for granted. Yet it controls nearly every aspect of
our life, from our route to and from work to the food we eat and
more. The weather is the talking point for the random meeting, or
the casual chat on the Metro. The weather is something that we
all have an opinion on; most notably the accuracy, or perceived
lack-thereof, of TV weather forecasts, or the dress sense of their
presenters. Yet, most have little understanding of why weather
happens at all.
Weather, for all its daily and geographical complexity, can be
broken down to one simple statement: it is the transfer of energy
from one area to another. Weather is a manifestation of our parent
stars inability to heat the Earth's surface evenly. A round, three
dimensional planet is notoriously difficult to deliver energy to in
any kind of even-handed manner. Areas under more direct sun-
light, such as at the equator, warm most strongly as heating is
more intense when the Sun is overhead. This is simply because
the energy is delivered to a smaller surface area than if it were
illuminated at an angle.
Worse still, the presence of an atmosphere ensures that energy
from our star is diluted or diverted on its multimillion kilometer
passage from star to planet. Much of the energy is reflected by
clouds or particles in the air, or absorbed by elements and com-
pounds that swirl within it (Fig. 1.1).
If the Earth were more like our Moon, with little or no atmo-
sphere, then the daytime temperatures would soar to nearly
175 C, while at night we would languish at 125 C: a range of
300 C. The atmosphere cushions the rise and fall of temperature,
Springer International Publishing Switzerland 2016 1

D.S. Stevenson, The Exo-Weather Report, Astronomers Universe,
DOI 10.1007/978-3-319-25679-5_1
2 The Exo-Weather Report
A: Light striking a planets

surface where the Sun is
overhead delivers its energy
over a small area.
Closer to the poles, the same

packet of energy is spread
out by the curvature of the
planet so that it covers a
much larger area than at the
equator.
B: The depth of air energy

passes through is proportional
to the latitude. The further
away from the equator (or
wherever the Sun is overhead)
the less energy reaches the
surface. Closer to a planets
poles much less energy reaches
the surface because of
scattering or absorption of
energy by dust, gases and
clouds.
FIG. 1.1 The effect of the curvature of a planet and any atmosphere on the
amount of heating different parts of its surface will experience. Were the
Earth a pancake facing directly into the sunlight all of it would be heated
to the same extent and weather, although it would exist, would be pro-
foundly dull. (a) Light striking a planets surface where the Sun is over-
head delivers its energy over a small area. Closer to the poles, the same
packet of energy is spread out by the curvature of the planet so that it
covers a much larger area than at the equator. (b) The depth of air energy
passes through is proportional to the latitude. The further away from
the equator (or wherever the Sun is overhead) the less energy reaches the
surface. Closer to a planets poles much less energy reaches the surface
because of scattering or absorption of energy by dust, gases and clouds
in part by absorbing and retaining some of the Suns energy, but

also by effectively transporting it from place to place. This transfer
of energy we call wind.
On any object with significant gravity energy can be trans-
ported vertically from the bottom to the top of the atmosphere
by convection. Advection is the transport of energy from place
to place in a (roughly) horizontal direction. This can happen in
response to convection pulling and pushing air around or, more
directly, in response to temperature differences and the spin of
What We Know About the Weather on Earth 3
our planet. On the Earth, convection dominates the transport of

energy at the hottest and (in our case) most equatorial regions,
but is prevalent elsewhere where conditions are suitable, partic-
ularly in the summer months when heating is strongest. By con-
trast, advection dominates energy transport between the tropics
and the poles.
The Highs and Lows of Meteorology

Where air rises, it exerts less pressure on the underlying layers
or the surface: this creates an area of lower pressure that draws
in air from the surroundings. Where the air is moist, clouds can
form and precipitation falls. Essentially: what goes up must
come down. Otherwise, when air is sufficiently cold and dense,
it descends under its own weight. As it falls it compresses the
air underneath, generating an area of higher pressurean anti-
cycloneand warms somewhat as it compresses under its own
weight. In addition, pressure will rise and fall if it is forced to do
so. Pressure falls where air is forced to rise over hills and moun-
tains, and conversely in the extreme environs of tornadoes and
hurricanes, as the air circulates rapidly around the central low
pressure core, inflowing air cannot keep pace with the air that
is being sucked out of the storms top. Some air then is forced to
descend inside the center of the storm: the eye of the hurricane
in particular is a region of descending, warming air that breathes
a brief window of calm in the otherwise violent storm.
In addition to this simple relationship overall, warm air exerts
a higher pressure than colder air. Thus, if you move vertically in
a column of warm air pressure falls more slowly with height than
if you are moving upwards through a column of cold air. This has
some interesting effects. Over Siberia (and North America) dur-
ing the winter, the air becomes extremely cold. This generates an
intense area of high pressure over Siberia. Yet, if you rise above
3000 m the high pressure area has been replaced by a finger-shaped
area of low pressurea trough. Conversely, Scandinavia is com-
monly the location of what is known as blocking anticyclones.
These are areas of high pressure, with a warm core, which extend
through the full height of the troposphere. Although the air may
be very cold at the surface, you dont have to travel far upwards
to encounter air that is much warmer than you would expect for
that latitude and altitude in the winter. Blocking anticyclones are
also common over the North Pacific in the winter, particularly in
La Nia years, and over the same region in the summer of El Nio
years (Chap. 2). It is the presence of the core of warm and relatively
high pressure air that makes these atmospheric features so stub-
bornly difficult to move. Indeed, the current climatic tribulations
taking Alaska and California by force, are largely down to the per-
sistence of a blocking anticyclone over the northern Pacific; this
is one block that has lasted, with only minor interruptions, for
several years.
Likewise, low pressure areas tend to become colder with
height. The major low pressure areas of the mid and high latitudes
develop cold cores as they deepen. However, in the summer, and
over the Tropics, many low pressure areas, including hurricanes,
are warm-cored throughout. Within these thermal lows pressure is
low at the surface but soon morphs into an area of high pressure at
greater heights, due to warmer air exerting a higher pressure than
cold air.
Wind Direction: Waterwheels and a Suspect

Tale from the Front
Air flow on any rotating planet is driven by two opposing forces:
pressure and spin. Pressure is the simplest to understand. Air
of lower pressure effectively draws air into it from regions with
higher pressure. Strictly speaking a force is exerted by the high
pressure air, which directs gases towards the regions with lower
pressure. This is the pressure force. Intimately linked to this idea
is temperature: air of a higher temperature has a higher pressure.
There are three fundamental gas laws describing this relationship
of pressure, temperature, volume and amount of gas. Boyles Law
dictates that the volume of gas increases as the pressure decreases:
gas under higher pressure has a smaller volume, if you keep the
temperature and mass of gas the same. Charles Law requires that
the volume of gas increases as the temperature increases. And

Avogadros Law tells us that equal volumes of all gases, at the same
temperature and pressure, have the same number of molecules: for
a given mass of an ideal gas, the volume and amount (moles) of the
gas are directly proportional if the temperature and pressure are
constant. Together, these laws guide much of modern science and
are integral to our understanding of how atmospheres work.
Charless law, incidentally, is often misattributed as Gay-
Lussacs Law, as it was published in 1802 by the French chemist,
Joseph Louis Gay-Lussac. However, the publication followed a
century after from the work of Guillaume Amontons and Jacque
Charles, whose work Gay-Lussac extended. Amontons work lan-
guished for a century before Gay-Lussac revamped it, along with
many unpublished results from Charles. Gay-Lussacs triumph lay
in the use of individual gases, such as oxygen and nitrogen, unavail-
able in Amontonss day and his ability to interpret, unite and extend
the work of Charles and Amontons, producing two of the three core
gas laws biologists, chemists and physicists use today.
If we look simply at pressure and temperature, particles of the
various gases that make up the air will exert a higher pressure at
a higher temperature because they move faster (they have greater
kinetic energy). Collisions between the particles and any surface
will, therefore, carry more energy and thus exert a greater force
than those in a colder and denser gas. Imagine a baseball hitting
a wall at 10 km per hour versus 80 km per hour. This idea has
important ramifications for the structure of the atmosphere as we
shall see later in this chapter.
Spin is a more complex concept. The Earth rotates on its axis
once per day. That means that a piece of soil at the equator has to
move at nearly 1700 km per hour to make a full rotation in this
interval. Conversely, standing at the Earths rotation poles, you
don't rotate at all (except slowly around your middle, of course).
That means that any stray packet of air trundling north or south
towards the poles will find itself, increasingly, moving faster than
the ground underneath; and this means that, relative to the sur-
face, it will steadily bend towards the east as it travels away from
the equator. Similarly, but in the opposite direction, air moving
from either pole towards the equator will curve towards the west
km/h Eastwards
km/h Eastwards
,km/h Eastwards
km/h Eastwards
km/h Eastwards
FIG. 1.2 The effect of a spinning Earth on the movement of air to and
from the equator. Air moving away pole-wards is deflected eastwards
while air moving towards the equator moves to the west. Numbers indi-
cate the rate of rotation of the surface around the polar axis
because the ground underneath it is moving more swiftly towards

the east than it is (Fig. 1.2). A similar thing happens when a person
steps off a moving bus or traintheir speed, doesnt match that of
the ground underneath.
Known as the Coriolis Effect, this made its presence known
during the latter part of World War I. In 1918, German troops
positioned a large gun 120 km to the northeast of Paris. The gun
was used to shell Paris from behind the German lines with the
aim of terrorizing the population rather than inflicting significant
damage. Because the distance was so far, the Earths rotation had
an impact on the shells trajectory; they reached higher than any
other projectiles up to that time, and their path was therefore sub-
ject to forces that had theretofore unaccounted for. Initially most
of the shells fell to the west of the French capital, because when
they began their journey they were moving eastwards at a slower
speed than Paris, which lay to the southwest and thus had further
to travel in its rotation around the Earths axis (Fig. 1.3).
FIG. 1.3 The 1918 Paris gun. Shells were fired at Paris from 120 km
to the north-east of the capital. Shells curved to the west of the city as
a result of the Coriolis Effect (red arrow). The gunners soon corrected
for this and repeatedly shelled Paris. French map courtesy of Wikipedia
Commons
Unfortunately, after some tweaking of the trajectory, Paris

was positioned firmly in the guns sights and was shelled up to
20 times a day, killing 250 people. The Paris Gun was notable
for one other meteorological first. The gun fired the first man-
made object 42.3 km into the stratosphere; a feat not beaten
until Germany launched the infamous V2 rockets in 1944. The
sheer height of the trajectory combined with the distance brought
Gustav Corioliss effect into play.
Otherwise, the Coriolis Effect is mostly experienced by air or
liquids that are moving northwards or southwards on any rotating
body. Its discoverer, GaspardGustave de Coriolis, first described
the Coriolis Effect while thinking about machinery. Coriolis
was interested in work, or more fully work done, a definition
of energy: the ability of a force to move a mass over a distance.

He investigated work in relation to objects that were rotating,
in particular water wheels, which were very much in demand
to drive the hardware of the early industrial revolution. Coriolis
had no interest in the dynamics of the Earths atmosphere, or
indeed any planetary atmosphere, although the title of his 1832
paper was inadvertently suggestive of this: On the equations
of relative motion of a system of bodies. In this work Coriolis
detailed various forces which afflict rotating bodies, including
planets, but it would take nearly 70 years before Coriolis name
became clearly associated with the meteorological effect he inad-
vertently described.
The Coriolis Effect is proportional to the speed that the body
spins and its overall diameter. Thus a small planet has a weak
effect, as does a planet that rotates slowly. Therefore, both Mars
(small) and Venus (slow) have weak Coriolis Effects, while mighty
Jupiter, with its 17 h rotation has an effect much stronger (about
three times) than that found on the Earth.
This effect works at the surface and throughout the bulk of
the atmosphere. On the Earth hot air rises above the Equator (or
more precisely where the Sun is overhead) and sinks at the poles.
You might then expect a simple pattern of air flow from pole to
equator. Easterly winds would dominate the planet1 because these
would be returning cold air from the poles towards the equator;
indeed, this pattern is evident on Venus (Chap. 5). However, this
is not seen in the Earths atmosphere, because the Coriolis Effect
prevents air from flowing in such a simple pattern. The issue is
the air rising above the equator. This air rises over 10 km through
the lowest layers of the atmosphere, until it hits a wall, called
the tropopause, the nature of which we shall return to later. At
the tropopause the air is diverted to the north and south and the
Coriolis Effect begins to work its magic.
1
An interesting consequence of this would be that the Earths spin would decelerate. Winds would
be blowing in the opposite direction to its spin and thus exert a frictional force on its surface.
High above the band of rising air, air flows north and south.
The Coriolis force diverts this towards the east, in a westerly
flow. Initially, the winds move northeastwards or southeastwards
towards the pole, but by the time the air has reached roughly 30
N or 30 S the air has been turned right angles to its original direc-
tion and flows strongly to the east. There is no further north or
south movement. Were this all that happened then the tropics
would be completely isolated from the poles and the only means
by which heat could be transported further north or south would
be via ocean currents. However, the planet has a trick to play
with this air. The air steadily cools as it initially rises near to the
equator. By the time it has traveled 30 north or south it is dense
enough to sink towards the ground once more. So down it falls
towards the ground in a steady stream, pouring over the Tropics of
Cancer and Capricorn. The planets steadiest belt of weather, the
dry Horse Latitudes, are created from this pattern, called Hadley
cells. Past the 30th parallel most of the air returns to the equa-
tor, while the rest streams northwards or southwards towards the
poles, entraining air from outside the Hadley cells.
Once more the Coriolis Effect comes into play as air moving
towards the North Pole is again bent towards the east, producing
the mid-latitude westerly winds. The southern hemisphere mirrors
this with the winds arcing round to form the roaring forties and
even more descriptively screaming fifties. The Coriolis Effect keeps
the air turning towards the east so once again it fails to reach either
pole and the atmosphere comes to the rescue. Cold air streaming
away from either pole, towards the west, collides with the westerly
belt and directs much of it upwards towards the tropopause.
Finally, after a fairly exhausting trip a portion streams
northwards (in the Northern Hemisphere) and southwards (in
the Southern Hemisphere) to reach the Polar Regions. The cir-
cuit completes when this now profoundly chilly air sinks under
its own weight towards the surface. From here the only way is
away from the pole to complete the loop, reuniting the air of our
planet in three great circulations between pole and equator. This
is illustrated in Fig. 1.4. It must be emphasized that this pattern is
broadly symmetrical around the equator, meaning that the same
overall pattern of air flow is seen in the Southern Hemisphere.
High
Low
High
FIG. 1.4 The general pattern of wind flow and pressure on the Earth in the
Northern Hemisphere. Geographical features such as oceans and moun-
tains strongly affect this generic pattern; as does the tilt of the Earth
throughout the seasons
The Vertical Structure of the Earths

Atmosphere
As well as horizontal transitions across the globe at different lat-
itudes, the atmosphere can be broken into different layers as it
increases in altitude. In each layer temperature and pressure vary
in predictable ways.
Extending through the first few kilometers of the atmosphere
is the region of densest and most turbulent air, known as the tro-
posphere. This is thickest (17 km) over the Tropics where the air is
warmest and exerts the greatest pressure. The depth of the tropo-
sphere decreases in a series of steps as you go further towards each
pole. Each step is marked by a current of fast moving air known
as a jet stream (returned to later in this chapter). Thus, where as

above the equator the top of the troposphere is 17 km high, it is
only 10 km or so above the ground over Spain, the bulk of the
US and southern Australia. Even closer to the poles and it drops
even closer towards the surface, ultimately lying 78 km above
the ground at the Poles.
Throughout the troposphere temperatures fall as altitude
risesindeed this is a defining feature of this layer. This can make
the air unstable and prone to convect. However, there is such vari-
ability in this region that simply describing it as a simple layer
would do it a great disservice. The troposphere is broken hori-
zontally into different regions that move under the influence of
pressure and the Coriolis Effect, leading to a very dynamic and
complex zone. Fortunately, as you go higher the complexities sub-
side and things become somewhat simpler to understand.
Above the troposphere lies the stratosphere. As the name
implies, air in this layer is stratified, displaying relatively little
vertical movement. The dividing line between the troposphere
and the stratosphere is a boundary called the tropopause. Here
temperatures stop falling with increasing altitude and begin to
rise. This change causes the air to stop convecting as from this
point upwards the air is warmer the higher you go. The reason the
temperatures rise is down to one gas: ozone. Ultraviolet light is
absorbed by oxygen, which is found in the form of a molecule (O2).
In this diatomic (two atom) form, the bonds can be broken and
then reassembled into ozone (O3). Ozone is superbly efficient at
absorbing ultraviolet light, particularly that in the range of longer
wavelengths closest to the visible part of the spectrum. It is these
wavelengths that are most harmful to life, thus ozone helps ensure
complex life can populate the solid surface of the Earth. Without
it, life would be confined to the ocean depths where UV is unable
to penetrate deeply (Fig. 1.5).
As you progress upwards through the atmosphere, the density
of gases steadily diminishes. The stratosphere begins where the
atmospheric pressure is approximately one fifth that found at the
surface (200300 mb) and ends with a pressure less than one tenth
this value. Although very dense compared to the vacuum of space,
the air is so rarified that any living organism (for example a stray
microbe) would be rapidly freeze-dried.
Height
(km)
Exosphere
500
Thermosphere
Aurora
100
90
Elves 80
Sprites Noctilucent Clouds 70
Meteors
Mesosphere 60
burn up
50
40
Stratosphere
30
Blue Jets
20
Upward Super-bolt
10
Troposphere Thundercloud
-100 -50 0 +50 +100 +200 +500/1,500
Temperature (oC)
FIG. 1.5 The overall structure of the Earths Atmosphere. The different
layers are separated by white lines. The thickness of the troposphere var-
ies with latitude and is thickest nearest the equator. Jet streams (small,
blue circles) mark the position of the jumps. The red line indicates how
temperature changes with height. The temperature of the thermosphere
varies from 500 C to over 1000 C when the Sun is most active. Various
phenomena are indicated. Sprites, Elves and Blue Jets that are associated
with thunderstorms are described in Chap. 4
Above the ozone layer, the mesosphere begins and the tem-
perature falls once more from around zero Celsius to around
100 C. The air is very dry but does hold a small amount of
water vaporjust enough to form rare noctilucent clouds out of
water ice crystal. These form at an altitude of around 7080 km,
predominantly in the early spring at the poles where the air is
coldest. Their frequency is increasing, which suggests that more
moisture is escaping through the stably stratified layer below
from the moist troposphere at the atmospheres base. This may be
a sign that rising global temperatures (Chap. 2) are driving more

water vapor into the higher atmosphere.
As you ascend through the mesosphere the abundance of mol-
ecules falls as ultraviolet light and increasing amounts of x-rays and
gamma rays break them apart. Atomic oxygen, nitrogen, and the
molecule hydroxyl, which is unstable at the Earths surface, domi-
nate the composition of the gases in the mesosphere. Hydroxyl is
produced when water vapor is split by UV releasing hydrogen to
space. The left-over oxygen and hydrogen remain bound together,
but are ultimately split up at higher altitudes. The atmospheric
pressure is less than 1000th that at the surface making the meso-
sphere a fairly good analog for the atmosphere that rests on the
surface of Mars.
The mesosphere is also home to some enigmatic electrical phe-
nomena known as sprites, elves and tendrils. These microsecond
long features appear above some, but not all thunderstorms and
appear to form electrical connections between the turbulent tro-
posphere and the thermosphere above (see Chap. 4).
Additionally, the mesosphere is also the region where most
meteorites are vaporized. This leads to a regional enhancement in
the abundance of silicates and metals, particularly iron and nickel.
All told, the mesosphere is one of the atmospheres most inscru-
table regions. It is too high for balloons and jet craft to probe,
yet too low for satellites to sample. Measurements are always
taken remotely from the surface or from space and are as yet rela-
tively piecemeal in nature. Some of these recent observations are
described in Chap. 10.
Above 80 km in height, the temperature begins to rise once
more. Gases are increasingly ionized forming bands of rarified
gases at different heights. These ionized regions are important for
ground-based radio communication as they efficiently reflect elec-
tromagnetic waves to allow humans to communicate with one
another around the curved surface of the planet. These gases also
give this layer its name, the ionosphere. Technically, the iono-
sphere encompasses both the mesosphere and the ionized layer
above which we now call the thermosphere. Whereas the amount
of ionization is limited in the mesosphere, all of the gases are ion-
ized to varying extents in the thermosphere. Temperatures rise to
over 500 C in this region. Not that you would feel it hot were you
able to experience it: the concentration of gases is so low that you

would still freeze solid if out of the glare of the Sun. The maxi-
mum temperature can exceed 1000 C when the Sun is active and
the Earth is subjected to the highest intensity of ultraviolet and
x-radiation.
Towards the top of the thermosphere you might bump into
passing orbiting craft, including the International Space Station
which orbits at 400 km altitude. As you went higher you would
enter the magnetosphere. In its regions above the equator the
Earths magnetic field guides the vapors and the solar wind into
warped donuts of plasma. Here the density of gas is trillions upon
trillions of times lower than the density found at the Earths sur-
face. It is effectively an extension of the near vacuum of space,
where hydrogen and helium pause on their way out into the Solar
Wind. The hydrogen comes mostly from the break-up of water
vapor that results from ultraviolet light splitting the molecule
up, while the helium comes from the Earths crust and interior.
Helium is synthesized by the radioactive decay of heavy ele-
ments, such as uranium, or was originally trapped there when the
Earth formed. In addition, small amounts of oxygen and nitrogen
are drawn outwards into space. This is partly as a result of their
entrapment within the flow of escaping hydrogen and helium,
something called hydrodynamic escape. Alternatively, oxygen and
nitrogen escape directly from the effect of ultraviolet light bom-
barding these gases and energizing them sufficiently to escape in
their own right.
This region where gases escape from the thermosphere to the
magnetosphere or the vacuum of space is known as the exosphere.
The exosphere is less of a region of atmosphere as much as it is a
transition from atmosphere to space. The density of gases isnt so
much measurable in millibars rather in terms of individual parti-
cles per cubic centimeter. The only thing that would exert notice-
able pressure upon you would be the impact of a passing meteor or
a man-made satellite.
As far as weather and climate are concerned most of the
action occurs in the troposphere, with its upper surface, the tro-
popause, keeping a lid on most meteorological activity. However,
within the stratosphere and mesosphere, meteorological phenom-
ena do occur that are at least visible from, if not having a direct
effect upon, the surface below. With our sortie of the atmosphere
complete we return to the troposphere to engage in some atmo-
spheric hand-to-hand combat. However, in Chap. 4 we will look
again at some of the more exotic atmospheric phenomena that stir
the layers above.
The Language of War-Fronts
During the First World War, and in it immediate aftermath, meteo-

rologists were trying to understand patterns within the flow of air
and the link to the weather in general. A group of scientists working
within the Bergen School of Meteorology harnessed the language
of the Great War to describe the observed patterns of temperature
and wind flow. The term front became established as a generic
term that described the boundary between two competing air
masses. In the Bergen scenario, known as the Norwegian Cyclone
Model, a low pressure first develops as a disturbance along the
so-called polar-front: this is the boundary between tropical air that
is moving pole-wards from the equator and polar air that is mov-
ing towards the equator. A wave develops along the polar front as
warm air nudges into it. Along this boundary, the overriding warm
air condenses as clouds and precipitation intensifies. As it does so
pressure falls along this portion of the boundary (Fig. 1.6).
To the rear of this surface feature, cold air begins to undercut
the warm air, again driving the warm air upwards as a second band
of precipitation develops. This region will become the cold front.
Between the two fronts is a region of warm air that has intruded
into the colder polar air: this is the warm sector. The idea of fronts
is most closely tied to Jacob Bjerknes at the Bergen School. Bjerknes
often referred to the warm front as the steering line, while the cold
front became known as the squall line (Fig. 1.7). Although these
terms are still often in use in the US they are generally not used
in Europe. Here, the terms cold and warm fronts have stuck and
most actively describe the changes in conditions associated with
the passage of each frontal boundary.
In the Bergen model, the low pressure moves along the frontal
zone as an intensifying wave. As it does so, the warm front advances
effectively at the same rate as the low pressure as a whole, while
the cold front progressively undercuts the warm air to the rear.
Ana Warm Front cirrus
cirrostratus
altostratus
nimbostratus
stratus
cirrus
Kata Warm Front cirrocumulus cirrostratus
altocumulus
stratocumulus
stratus
FIG. 1.6 Idealized warm fronts. In the top diagram, the warm air advances
quickly and is unstable, meaning that it is able to rise by convection,
as well as rise because it is being forced over the wedge of colder air.
This Ana warm front brings extensive precipitation, a strong rise in
temperature and a significant change in wind direction. On occasion
thunderstorm cells may be embedded (grey). Above the kata front air
is descending from higher up in the atmosphere. This may be associ-
ated with frontal boundaries in the upper air. Descending air causes it
to warm and evaporate much of the cloud layer. Clouds are typically
more broken and precipitation lighter (more dispersed blue dots). Red,
blue and graded arrows indicate overall direction of air flow. The vertical
scale is exaggerated relative to the horizontal scale. The frontal surfaces
are typically only a few degrees to the horizontal
After anything from one to several days, the cold front advances
so far that it undercuts the warm front, forming an occluded front
(Fig. 1.8). The low pressure center, which is often referred to as an
extra-tropical cyclone, then typically becomes isolated within the
cold air to the rear of the cold front. Denied access to the energy
contained in the warm air, the low pressure progressively fills in
and decays.
The Bergen Model was based exclusively on surface observa-
tions of clouds, pressure, temperature and wind direction. There
was little access to data from greater elevations; therefore, the
link to processes happening further up in the atmosphere was not
understood. Nonetheless, the model has worked extremely well
and accurately describes the processes occurring in frontal low
pressure areas. For most scenarios involving the formation and
development of low pressure areas, the Bergen model works more
than adequately.
cirrostratus
altostratus Ana Cold Front

cumulonimbus
nimbostratus
Kata Cold Front
cirrostratus cirrocumulus
altocumulus
stratocumulus stratus
FIG. 1.7 Cross sections through idealized cold fronts. The upper fig-
ure shows the stereotypical Ana cold front with strongly converging
air and strong uplift along the front. Rain typically falls immediately
after the cold front has passed from thick nimbostratus and occasion-
ally cumulonimbus. In the lower half of the figure the kata cold front
has air subsiding and warming aloft. This evaporates clouds, leading to
lighter precipitation. Kata fronts often form when they are overrun aloft
by upper level cold fronts (Fig. 1.13). However, they can form more gener-
ally when front meander into areas of subsiding air. The vertical scale is
exaggerated in each diagram with the slope of each front being at most
510 and usually much less
The frequently strong differences in temperature, humidity

and pressure associated with the air in front of and behind fron-
tal boundaries led to the concept of air masses. This idea is still
taught today and is, generally, a worthwhile method of defining
both different regional air types, as well as providing a predictive
measure of the weather that will follow when one of these air
masses advances overhead. Figures 1.9 and 1.10 show two other
types of atmospheric front. Figure 1.9 shows an upper level front
where changes in temperatures and pressure are only found above
510,000 feet. Many of these are formed when stratospheric air is
sucked downwards into the troposphere. Figure 1.10 shows the
dry line: a boundary between dry continental air and moist mar-
itime air. Such fronts often encourage the formation of vigorous
thunderstorms during the spring in the US that are often associ-
ated with tornadoes.
Advancing cold front occlusion
Less cold air

Coldest air
Advancing warm front occlusion
Less cold air Coldest air
FIG. 1.8 The occluded front. The cold front almost always moves faster
than the warm front, which it follows. This means that sooner or later
the cold front will catch the warm front and lift the entire intervening
warm sector aloftand with it the rain bearing cloud. When the advanc-
ing cold air behind the cold front is the coldest air in the storm system
the warm front and the cold air ahead of it are lifted en mass. The coldest
air follows the front so this is called a cold front occlusion. Precipitation
tends to be more intense and shorter in duration than in the opposite
case when the coldest air lies ahead of the warm front. In the winter in
the western States and the UK, air from the Pacific or Atlantic, respec-
tively, tends to be warmer than the air ahead of it so occlusions tend to
be of the warmer variety. Nimbostratus and altostratus dominate the
rain-bearing clouds
The air mass concept has its limitations, however. In particu-

lar, the system of fixed air masses needs tweaking to account for
the effects of direction of passage of each air mass and whether
or not the air mass has passed over water with temperatures that
differ from the bulk air mass. For example, dry continental polar
air can only reach the UK via the North Sea. In winter the sea is
considerably warmer than the air mass, so much like the passage
of continental polar air over the Great Lakes, this dry cold air mass
is heated from below and picks up considerable moisture. By the
time it reaches the UKs eastern shores, it frequently drops show-
ers of rain, sleet, hail or snow, depending on how much it has been
modified by its passage over water. Similarly, maritime polar air,
which starts out near Iceland, may take a fairly circuitous route
to the west of the UK (or the west coast of North America from a
starting point south of Alaska). If it travels far to the south before
Stratosphere
Jet
Upper Level
Troposphere Front
FIG. 1.9 Upper level fronts. Warm air flows over the mountains and with
sufficient momentum it continues onwards faster at height than at the
surface. This upper level front can continue forward while the remainder
of the air mass is modified as it pours through valleys and over mountain
tops. Air cools at a slower rate when moist than when it is dry, ensur-
ing that it is far warmer in the lee of mountains than on the windward
side. Such warm air is known as Fhn in Europe and Chinook in Canada
and the far north west US states. Upper level fronts can also form when
stratospheric air enters the troposphere at a jet stream. The synoptic
charts show an upper level cold front (arrowed) crossing the UK bringing
thunderstorms as it swept over the warm sector
arcing back northwards, the air is generally warmer than the

waters over which it travels. Subsequently, the air mass is cooled
at its base and becomes fairly stable. Rather than producing copi-
ous showers, such drab air brings banks of broken low cloud and
only slight precipitation.
The Jet Stream
Between 7 and 16 km above the surface of the planet lie narrow

bands of fast moving air. These air currents had been suspected
as early as 1883 when Krakatau erupted, throwing several cubic
kilometers of ash and sulfur dioxide into the air. It was apparent
to observers that the ash was initially carried quickly from east
to west along a narrow, equatorial band, referred to this as the
Dr air behind
the dry line
cools at 3o C
every 1,000 feet
(300m) so
although hotter
at the surface in Humid air in
the daytime , front of the dry
Dry cT cools more line cools at
air rapidly above. 1.5o every 1,000
The dry air is feet (300m) so
Dry Humid mT air also denser although cooler
Line than moist air at the surface in
at any given the daytime, is
temperature so warmer aloft
undercuts the
moist air
FIG. 1.10 A different kind of front. The dry line separates air with differ-
ing humidity. In the US, Argentina and India the dry line is responsible
for the formation of severe thunderstorms in the Spring and Summer
months. Dry continental (cT) air to the west undercuts and overrides
warm moist air from the south. The undercutting action lifts and heats
the air at the surface and generates instability. Because temperatures fall
at different rates in dry and humid air any storm that is developing near
the boundary may penetrate the overlying dry layer. As temperatures fall
more rapidly with height (in the daytime) convection can rapidly inten-
sify once the boundary is breached and warm moist air invades the cooler
air above. North American map courtesy of http://www.vectors4all.net/
vectors/north-america-map-vector
equatorial smoke band. The significance of this band of rapidly

moving air was not understood until the 1920s when Japanese
meteorologist Wasaburo Oishi detected another strong band of air
which swept westwards high above Mount Fuji during the winter
months. Oishi released a flotilla of balloons, known colloquially
as Pibals. These were allowed to drift upwards into the prevailing
winds. During the winter months some were carried towards the
United States.
Oishis work fell victim to his desire to promote Esperanto and
unwisely Oishi published his many studies in this artificial lan-
guage. Unfortunately, few shared his passion for Esperanto and his
work languished, with few readers outside of Japan. Consequently,
it is the American pilot Wiley Post who is normally credited with
the discovery of the Pacific jet stream. Wiley had invented a pres-
surized suit that allowed him to fly above 6200 m. During many
such flights Wiley observed that he was zipping merrily across the
surface of the planet at speeds far in excess of his air speed. Clearly,
something was giving his aircraft a boost. The transient nature of
the effect implied that these high winds were confined to narrow
bands only a kilometer or two thick and less than a couple of hun-
dred wide. In 1939 the term jet stream came into play, owing its
origin to Heinrich Seilkopf who coined the term Strahlstrmung;
the literal translation of which means jet streaming.
Air Masses
Air masses are defined as any large body of air with a spe-
cific temperature, pressure and humidity. In the UK we are
afflicted by five main types; in the US six or more. These can
be described as follows:
Maritime Tropical (mT) air: warm and humid air from the
Azores (UK), Caribbean (US) or the Indian and Pacific Ocean
(Australia and New Zealand). In the UK this air mass brings
relatively warm summer weather, but with moderate to high
humidity. In the winter, it usually brings cloud, damp condi-
tions with above normal temperatures.
Continental Tropical (cT) air is dry and warm in the winter or
hot in the summer. This is a rare but often welcome visitor
to the UK in the winter when it brings clear, mild conditions.
In the summer it is somewhat more prevalent and brings our
hottest weather from Spain and North Africa. In the US and
Australia this air mass originates over the continental inte-
rior, bringing the hottest and driest conditions.
Maritime Polar (mP) air: generally humid air that is cool in
the summer and relatively cold in the winter. Usually asso-
ciated with variable cloud and showery weather, this often
follows the passage of the cold front. This air mass is most
common in northwest North America, and Western Europe,
as well as southern Australia and New Zealand. The Atlantic,
Pacific and circum-Antarctic region sources this air mass.
Maritime Arctic (mA): a rare visitor to the UK in the sum-
mer, this brings cold, clear but showery conditions, along
with some of the coldest, most snowy weather in the winter

and spring. In the US this originates over the Arctic Ocean or
the northern north Pacific and Atlantic oceans in the winter,
along with air blowing across the sea from Greenland. In the
southern hemisphere, maritime (ant)arctic air rarely afflicts
Australia and New Zealand, being largely confined to the
regions closest to Antarctica.
Continental Polar (cP) air: this is a dry air mass that originates in
the continental interiors or Europe and North America. In the
summer it brings warm, dry weather, but in the winter it can
bring a deep dry cold. On the windward side of lakes (or in the
UK along the east coast) cP air can bring extensive snow show-
ers. cP air is responsible for extensive snows along the eastern
shores of the Great Lakes, particularly in the early winter.
Continental Arctic (cA) air. Largely unknown in the UK, but a
common part of the winter scene in Canada and the northern
States, this largely seasonal air mass brings a deep and largely
dry cold. Like cP it can bring snow along the windward sides
of lakes, if there is sufficient warmth in the water to generate
showers. Similarly, in continental Europe cA air brings deep,
dry cold from Russia during the winter.
Initially, Oishi and Posts work implied one jet stream,

located over the mid-latitudes, with the Krakatau eruption cloud
suggesting another easterly jet. However, during the Second World
War it soon became clear that there were numerous, distinct
jet streams lying at different latitudes. Fighter pilots, undertak-
ing long distance flightsparticularly between the US and the
UKencountered various bands of high velocity winds which var-
ied in position but were encountered most frequently at particular
latitudes (Fig. 1.11).
These newly discovered belts of fast winds were employed
to various effects during the Second World War. Aside from
pilots trimming their fuel consumption, the most notorious use
of the jet stream was Japans aerial bombing of the Pacific north-
west coast of the US between November 1944 and April 1945.
Using Wasaburos studies, at least 9000 incendiary bombs were

Polar Night Jet

(Winter only)
Height (kilometers)
Stratosphere Sub- Tropical

Tropical Easterly Jet
Jet (summer)
Polar
Front Jet Tropopause
Arctic Jet (Seasonal)
Hadley Cell
Ferrel Cell Trade Winds
Horse Latitudes
oN oN oN oN
FIG. 1.11 General circulation in the atmosphere of the Earth below

50 km. The tropopause marks the boundary between the increasingly
warm stratosphere and the unstable troposphere, which chills with
height. Black arrows indicate the general direction of winds in the tro-
posphere, while blue arrows indicate air flow in the stratosphere. At two
locations along the tropopause the level drops sharply. Here, the greatest
temperature differences are found and it is here that the two strongest
jet streamsthe polar front and sub-tropical jetsare located. Above
the rising band of air nearest the equator a third seasonal easterly jet
is found, while a fourth seasonal jet separates Arctic (or Antarctic) air
from warmer polar air. Higher up in the stratosphere a fifth, seasonal,
jet is located. Dry air can enter the troposphere from the stratosphere
near the sub-tropical and polar front jets, giving rise to upper level fronts
(Fig. 1.12). Not shown is air that leaks directly northwards, or south-
wards from the Tropical Hadley Cells to the Polar Cells
dispatched on stratospheric balloons. The westerly jet stream was

used to guide these onto the forests of California and Oregon.
Despite the prescient work of Wasaburo, his measurements of the
westerly jet stream were too imprecise to allow an intensive cam-
paign of bombardment. Rather than incinerate the western States,
most of the devices fell short of their targets having taken longer
than expected to reach the US coastline. Nonetheless, some of the
devices reached the US mainland. This was to be the last interna-
tional terror and it was to be the last international terror attack on
the US mainland until September 11th 2001.
Over time the true extent and nature of the jet streams has
come into focus, and with this improved resolution, a clearer
understanding of the role these ribbons of air play in determining
the weather (and climate) at the surface of the Earth. However, jet
streams are ubiquitous in the atmospheres of all rotating planets
and they even appear to have analogs inside many stars, including
our Sun. Its worth asking why planets generate these peculiar high
velocity streams of air, since they run counter to most of the other
observed air flows. However, a closer look at the movement of air
reveals that jet streams are an inevitable consequence on living on
a rotating body.
Rotation, Rotation, Rotation

Jet streams are powered by two competing forces: pressure and
temperature combined with the Coriolis Effect. The first of
these returns us to the work of the French physicists Guillaume
Amontons, Jacques Charles and Joseph Louis Gay-Lussac. Together
they showed how air of different temperatures occupies different
volumes and has different pressures. For any fixed mass of gas the
pressure and volume will be greater at greater temperatures. So,
as we observe, warm air expands because it particles are moving
more quickly and tend to spread out. This gives the warm air a
greater pressure than the same mass colder air. Therefore, pressure
exists between the warmer tropics on the Earth and the cold Polar
Regions. This force drives air from warm to cold areas.
Much as the air wants to move in this direction, the Coriolis
Effect tends to block it, causing instead the air to move with the
Earths rotation. The Coriolis Effect overpowers the movement
of warm air away from the equator, directing it into two broad
streams at about 30 N and 30 S of the equator: these are the
sub-tropical jet streams that overly the belt of high pressure areas
known as the Horse Latitudes that were mentioned above. The
Horse latitudes contain typically calm, hot and dry weather. It is
said that the weather was so calm that the Spanish galleons on
route to the New World were left with dead or dying horses that
the crew could not keep on board. In the Horse Latitudes, these
sick or dead animals would be tossed overboard. Vivid though this
imagery is, a more likely origin of the term relates to the finan-
cial state of the ship crews early in their voyages. Sailors were not
adept at managing finances and they would soon end up indebted
to their ships paymasters. The period over which this debt was
repaid became known as the dead horse time. This period, last-
ing one to two months, tended to coincide with the passage of the
ship over the 30th parallel, hence the association. Regardless of

the origin of the name, the weather in the Horse Latitudes has an
unflinching banality that is only punctuated in some locations, at
certain times, by the occasional tropical storm
North of the Horse Latitudes, the winds blow northwards
at the surface and are directed by the Coriolis Effect to the east.
Where these winds encounter the cold polar air streaming away
from the Arctic and Antarctic, there is a sharp difference in air
pressure because of the sharp difference in temperature. It is this
sharp transition in air temperature and pressure that generates the
planets strongest jets: the polar-front Arctic or Antarctic jets. The
rapid change in pressure causes air to move rapidly and the Coriolis
Effect directs this rapidly moving air into bands of high velocity
westerly winds. It is these jets that Oishi and Post encountered.
In general the Polar Front jet stream lies at a height of about
712 km above the Earths surface, while the sub-tropical jet lies
somewhat higher at 1016 km. The seasonal equatorial easterly
jet is the highest jet stream of the atmosphere's lowest layer the
troposphere, although other jets are recognizable at greater heights
within the stratosphere (Fig. 1.11).
Rossby Waves
This is where things get a little bit more complicated, but also cru-
cially important when we start to look at the atmospheres of other
planets. Rather than thinking about altitude in terms of kilometers
and miles, meteorologists use pressure as a proxy for it. For con-
venience, meteorologists pick a pressure value and then plot the
height above the surface at which this pressure is found (Fig. 1.12).
Looking at the atmosphere this way provides a kind of mir-
ror image of the differences in pressure that you would find if you
were looking at a set altitude. For example the 500 mb level rep-
resents the surface below which half the mass of the atmosphere
is found, while at 300 mb roughly two thirds of the atmospheres
mass is below you. In general, meteorologists prefer looking at
three levels in the atmosphere at differing distances above the
Earths surface. At the surface the average atmospheric pressure
is 1000 millibars (1000 mb or 1 bar). This is also the nominal
FIG. 1.12 Rossby Waves, jets and surface pressure. This view of the part
of the Northern Hemisphere shows the weather at different heights in
the atmosphere. Left, is a 300 mb view showing the jet streamor more
precisely jet streamsat the tropopause. The light blue line shows the
Arctic jet, the purple shows a very convoluted Polar-Front jet and the
red shows the sub-tropical jet just fringing the lower right of the image.
Between the Arctic and Polar-front jets, over eastern Canada, is a fourth
jet stream lying between both. The Arctic and Polar-front jets steer storm
systems from west to east, while the sub-tropical jet lies above the belt of
high pressure systems over the tropic of Cancer. Arrowhead indicate the
direction of air flow along each jet. At right shows the 500 mb level and
surface level pressure. Surface pressure features are shown in white, such
as a large high pressure area, blocking flow across the Atlantic, while the
500 mb surface is shown in various colors. Warm (and raised high pres-
sure) is shown in orange and red, while cold, and depressed surfaces are
shown in blue, purple and pink. 300 mb chart courtesy of Californian
Regional Weather Service; 500 mb chart courtesy of www.netweather.tv
surface used in extraterrestrial worlds that is used to describe

their sea level, particularly where there is no solid or liquid
surface, such as in a gas giant (Chap. 7). Above the surface, mete-
orologists pick a second level at 850 millibars (850 mb). This
pressure is found near 1500 m, or 5000 feet. At this level most of
the effects of friction, such as buildings, hills and forests are lost
and air is moving freely across most of the planets surface. Next
up is the 500 mb level, which corresponds to a height of roughly
5000 m, or 18000 feet. The 500 mb level represents the height
at which air broadly flows along pressure lines, driven by differ-
ences in pressure and the Coriolis Effect: it is neither diverg-
ing or converging across these pressure lines. Air flow at this
height also tends to be the driving force for most of the important
surface features, such as storm systems at mid-latitudes such as
Europe, the bulk of Asia and North America. Similarly, in the

Southern Hemisphere storm tracks across southern Australia,
New Zealand, southernmost Africa and South America are driven
by air movements at this altitude.
The final altitude that meteorologists concentrate on is found
at 200 or 300 mb. This is the level of the polar front jet stream that
organizes the flows beneath it and often drives the development
and decay of storm systems. The jet lies nearer 200 mb in summer,
when the air is warmer, and slightly lower at 300 mb in winter,
when air is colder, denser and hugs more closely to the ground. This
level is found at approximately 9300 m, or 30,000 feet (Fig. 1.13).
FIG. 1.13 A 500 mb chart for early February 2015. Rossby Waves are
broadly defined by the edge of the bluepurplepink regions centered on
the Arctic Ocean (outlined, approximately in red, with broader outline
in black). In this particular chart, three areas are identified that have
been cut-off from their sources. In the Pacific there are two blue circula-
tions (low pressure zones, sometimes called cold-pools) and one, warm
yelloworange regiona blocking anticyclone. Elsewhere, note how the
Rocky Mountains divert the flow of air from Seattle, south-westwards
towards Florida before it heads back up the east coast of the US and out
into the Atlantic. This is a characteristic feature of mid-latitude Rossby
Waves on the Earth, where mountains divert and anchor the overall flow
of air. This should also be true of other planets with varied surface terrain
(Chap. 10). Modified chart courtesy of www.netweather.tv
The temperature of the air is critical in determining the pres-

sure layer. Cold air tends to fall towards the surface because it is
denser than warm air. Therefore, the 500 mb level is lower in cold
air than it is in warm air. Warm air forms domes, while cold air
forms troughs and depressions in the 500 mb surface. The air at
500 mb and above is clearly organized into a series of ridges and
troughs which extend from the equator towards the Polar Regions
and vice versa (Fig. 1.13). The troughs are colder at any given lati-
tude than the ridges.
These troughs and ridges form features called Rossby Waves
(Fig. 1.14). These waves can move westward or eastward; and the
FIG. 1.14 Rossby Waves are a fairly complicated idea but one that we are
all familiar with. They consist of waves of different wavelengths that
can move in different directions. Imagine each blue wave is linked to its
neighbor in a group (red dashed outline) then this group of waves will
move together as a packet. This packet (or group) of waves can move
towards the right (east) or towards the left (west). Larger groups of waves
(those with the longest wavelengths) tend to move westward, while
shorter wave groups move eastward. A low pressure area (indicated by a
purple dot) moves with one wave inside the group. Rossby waves appear
in the jet streams, as well as the Tropics and between the Tropics and the
Polar Regions. Rossby waves are critical in organizing the atmospheric
flow on all planets and will be encountered in chapters 510. In Chap.
10 we will encounter how Rossby waves drive a phenomenon known
as super-rotation. Underlying wave image courtesy of Wikipedia Com-
mons: Oleg Alexandrov and modified by book author
way they do this is determined by the spin of the planet, its geo-
graphical features and the temperature of the air below. Within
the Earths troposphere three types of Rossby waves are discern-
ible. The most obvious waves lie along the Polar Front jet stream
and organize surface low and high pressure areas into alternating
east-west patterns. Another, very important set lie along either
side of the equator and are visible as alternating bands of higher
and lower pressure within the easterly trade winds. These extend
in a north-south pattern that is paired across the equator. A final
set of Rossby waves extends northwards across the pole. Both
of these north-south trending waves are most obvious during
El Nio weather phases (Chap. 2). A low pressure area over the
central Pacific has a paired high pressure area to its northeast over
the northeast Pacific (and also to the southeast of the equator).
Continuing further northwards, over the North Polar Region, sur-
face pressure tends to be low once more. This pattern was particu-
larly evident during the 2015 El Nio.
Now, lets look just at the eastwest trending waves. It has
to be said that the idea that these waves can be moving east and
west at the same time is probably confusing. However, things are
not quite what they appear. To get a better idea of whats going
on imagine throwing a rock into a pond. Instead of a single wave
moving outward away from the impact point, you get a package of
waves of different heights (amplitudes) moving outwards. Because
these move together, they are called a wave group. At the front of
the wave, smaller wavelets are moving forwards more slowly than
the whole group of advancing waves. This means that they appear
to move backwards into the body of the group, growing larger as
they do so, while new small waves appear in their place at the
front. Eventually the original forward-most wave passes to the rear
of the group and grows smaller once more. Atmospheric Rossby
waves are similar with one set of waves (those with the longest
wavelengthwell over 1000 km long) moving towards the west.
Within these large westward moving waves are smaller waves
(still a mighty 7001200 km long wavelength) that move towards
the east. Low pressure areas develop along their poleward face and
move from west to east (purple dots in Fig. 1.14).
Although planetary Rossby Waves may be more complex in
their structure and movement, the underlying rules apply, with
smaller waves moving through the body of the group, growing

first larger, then smaller as they traverse the group as a whole. In
reality the longest wavelength Rossby waves tend to be locked by
these geographical features and even without such barriers might
not move westward or eastward. Meanwhile, the shortest wave-
length features, associated with surface low pressure and high
pressure regions, tend to propagate from west to east along the
crests and troughs of the broader Rossby Waves. This is in the gen-
eral direction of the air flow within the jet streams and is driven by
them. However, on many planets the westerly-moving waves are
free to propagate. On Earth the westerly movement is often hard
to discern, except when the overall movement of surface lows is
blocked by warm areas of high pressure.
If Rossby waves are aligned along the direction of wind flow
then there isn't much of a problem as the mountains do not
obstruct the path of the air. However, mountains, such as the
Rockies, Andes and the Himalayas very much get in the path of the
waves. As a result waves will become locked around these features.
Perhaps most significant of these mountains are the Rockies. Air
flowing upwards, across the mountains, bends towards the pole on
the windward side, then bends towards the tropics, on the leeward
side. This pattern is fairly weakly defined for much of the sum-
mer, when the continental interior is hot, although the jet still dips
towards the south of the Great Lakes. However, during the winter
the jet stream and the air that immediately underlies it clearly bows
upwards towards the north of the Rockies, before bending south
across the Great Plains and out towards the eastern seaboard.
A Method of Predicting UK Winters at the End of August
Every Octoberusually in the second week of the month

there is a Coldest Winter Ever! headline in a well-known
British Tabloid. This is something of a British tradition, much
like egg custard or coronary heart disease. Despite a dismal
success rate the pattern persists year on year, much like the
Asian Monsoon.
Most of the UK population have learnt to scoff at this,
in part because it is invariably wrong; in part because it's an
amusing tradition. However, I am fairly certain I know the ori-

gin of the Tabloids erroneous system.
The system I use is well-troddenor at least it should
bebut there is little evidence it is put into practice
(Fig. 1.15, below). I look at the pattern of Rossby Waves at
the mid-tropospheric level of 500 mb in the summer or early
autumn. By late August the circulation of air over the Arctic
has formed patterns that is an excellent predictor of the pat-
tern in the North Atlantic in the following winter. Where
there is a well-established cold circulation over northern
Canada and Greenland, the following winter is invariably
dominated by westerly winds over Western Europe. The
best recent example was the winter of 20132014, which
brought a relentless tide of miserable wet weather to the
UK, France and the Benelux countries. The UK suffered its
wettest winter on record with severe flooding and extensive
storm damage. The average pattern month on month dur-
ing the autumn was of an increasingly well organized polar
vortex. Conversely, in the preceding year in the summer
of 2012 the polar vortex was poorly organized with warm
blocking high pressure areas punctuating its flow. The win-
ter of 20122013 was one of the UKs coldest in recent years
with extensive snow through to late March. Other recent
cold winters in the UK also follow the same pattern of air
flow above the Arctic during the late summer and early
autumn. Indeed, the more you look at the pattern the more
straightforward the correlation appears. That is not to say
that it has always held - only that it has over the last seven
years - pretty much without fail (six years out of seven pre-
dicted correctly using this method)2.
Does it apply to other regions? An obvious complication,
for North America is the continental landmass which cools
down dramatically in the autumn. While the UK, in partic-
ular England and Wales, drowned under the wettest winter
2
A short paper entitled An underlying predictability in the winter weather patterns in the
North Atlantic Basin will be published in the peer-review Journal Hypothesis by this
author in the summer of 2016.
FIG. 1.15 A surprising link between the northern summer weather pattern
in the Arctic and the ensuing winters in western Europe. See textbox for
details
on record in 20132014, North America was periodically hit

by a drifting Polar Vortexthe large pool of cold air normally
centered over the Arctic. This brought temporary spells of
severe cold and also helped pin the jet stream south of its
normal position, which in turn fired storm systems directly
at the UK, rather than allowing them to drive further north
towards Iceland and north Scotland. Similarly, while the win-
ter of 20142015 was downright average in the UK (a mix-
ture of mild and cold), the eastern States were pummeled by
a succession of severe noreasters. The Arctic patterns should
influence North America as it does here, but not necessarily
in the same way. Undoubtedly, some sense can be made of
the US patterns and link them to the summer pattern in the
same manner, but not being as familiar with the minutiae of
US meteorology I wont attempt to do that here!
So what of the Tabloid system, which appears to come
from Exacta Weather? Is that just a fanciful guess that routinely
is wrong or is there something to it? Well, it turns out there

is probably some method there. If you look at the same sys-
tem of Arctic air in October you get a picture which is similar
to the one I use earlier in the season. It appeared that every
October the Arctic goes a bit haywire for a while and appeared
to show a jumbled circulation that earlier on is indicative of
cold weather. Perhaps the Tabloid forecaster is looking at the
pattern here and judging it the same way each year. Instead he
or she should check the pattern in September and re-check in
November when the October glitch clears
Other websites make very strong claims about
their success. One site claim to beat the UK Met Office.
However, one of their June forecasts for July 2014 (http://
www.weatheraction.com/resource/data/wact1/docs/
BI1407JUL30dSlat9cWowContrastsProd27JunTypCor.pdf)
(for the UK) painted a picture of deluges, tornadoes and
torrential hail. July 2014, in the UK, was a rather warm, dry
monthsomething I got right with the method described
above. Similarly, claims of an astounding 52 % success rate
for gales (in 1 year) dont impress when by chance alone 50 %
might be expected. Be very careful of cherry-picked data. Ill
be clear I may get winters largely correct (6 out of 7) but my
success rate for predicting UK summers is hardly spectacular
(currently about 2 out of 3), so take my forecast for July 2014
with a pinch of salt; ditto that of others
In the winter, when the continental interior is cold, the jet

bends sharply to the Gulf of Mexico. It may break into one stream
tracking back along the Appalachians, while another branch
develops over the Gulf and tracks north eastward along the east-
ern seaboard. Thus the topography and temperature of the North
American continent strongly anchor the upper level Rossby wave,
producing either a mild trough in the summer or a much stronger
trough during the winter.
Similarly, the Himalayas divert a branch of the westerly polar
front jet stream to their south during the winter months. This helps
direct cool, dry air over India and maintains the dry winter mon-
soon. When the spring comes and India heats up, this southerly jet
stream is trapped and is eventually broken up. All the while a second
branch tracks east, along the northern edge of Tibet, towards Japan,
It was this jet stream that Oishi identified and, later, the Japanese
army used to deliver bombs to California and Oregon. Free from
a mountainous anchor, this northern branch is also free to move.
Thus when Asia moves into its summer, the jet drifts northwards,
eventually leaving Japan stranded in warm, tropical air.
In the Southern Hemisphere only the Andes really get in the
way of the polar front jet. There is a greater effect from ocean cur-
rents, in particular the cold Humboldt Current that flows up the
western side of South America. The polar front jet is diverted north-
wards towards the equator in the eastern Pacific before turning back
southeastwards over the southern Andes and out into the southern
Atlantic. In general it only clips the southernmost shores of Australia
during the coldest months of the year, but it frequently swings north
eastwards over the Tasman Sea to New Zealand, before heading out
over the Pacific. This configuration favors the formation of storm
systems to the west of New Zealand, during the winter.
Other Jets That Drive Terrestrial Weather

In addition to the broad, upper westerly jets and the Rossby waves
that they contain, are a myriad of lower level and seasonal jet
streams. These bands of concentrated airflow are defined as exist-
ing below 700 mb or about 12,000 feet (3500 m). The definition is
a little wobbly as jets at 600 mb are well below the height of the
main jets that blow at the top of the atmospheres lowest layer, the
tropopause. Perhaps these are mid-level jets?
Low level jets are found across the planet, particularly, but
not exclusively in the tropics, where they strongly influence
the seasonal patterns of rainfall. Normally, across the equato-
rial regions winds are deflected minimally by the Coriolis Effect,
which means that there are only weak (horizontal) forces direct-
ing rainfall. However, because the Earth is tilted on its axis the
area that is most strongly illuminated by the Sun moves north and
south between the Tropics of Cancer and Capricorn. On June 22nd
the Sun shines directly down on the Tropic of Cancer, leading to
the strongest heating, here. Conversely, in the southern summer,

centered on or around December 22nd, the Tropic of Capricorn is
most thoroughly heated. Wherever the Sun shines most intensely,
the belt of tropical thunderstorms follows: this is the inter-tropical
convergence zone, or ITCZ. However, particularly in the Northern
hemisphere, where there is a lot of land, there is a lag between the
Sun being overhead and the highest temperatures on the ground.
Therefore, instead of June being hottest in the Northern Summer
(or December in the south), July (or January in the southern hemi-
sphere) is the hottest.
Concentrating on the northern hemisphere, July sees the
peak of the summer (wet) monsoon over southern Asia and over
West Africa. It is also the season where most rain falls across
southwestern North America. On each continent the arrival of a
low level jet stream marks the onset of the peak rainy season.
As summer advances, the belt of equatorial easterlies drifts
northwards, strengthens and begins to be influenced by the Coriolis
Effect. At the surface the easterly winds begin to bend towards the
west, arcing around the southern tip of India and northeastwards
towards the southern Himalayas. However, that isnt all. The
strength of the Asian monsoon is linked to the strength of another
low level jet stream: the Somali jet. The jet is a consequence of the
topography of eastern Africa. Alongside the Great Rift Valley lie
the Virunga Mountains and the Ethiopian Highlands. As the sun
moves north and winds cross the equator winds from the south-
ern hemisphere initially move to the west south of the equator,
before, as we have seen then moving north eastwards towards
India. The mountains that flank the Rift Valley (in particular the
Ethiopian Highlands) intercept the low level winds and channel
them into a narrow (low level) jet stream. This, too, is diverted by
the Coriolis Effect to the east as it continues to travel northwards
towards India from the eastern coast of Somalia (Fig. 1.16).
High above all of these low level features a strong easterly jet
forms over northern India. This jet acts like an exhaust system,
removing the air that has been dragged across the surface towards
the continent. The combination of the Somali low level jet and
the upper easterly jet serve to guarantee the strength of the vital
summer monsoon. Without the integrity of these jets, India would
dry out and the rice paddies that fuel the country wither and die.
Polar Front Jet
Sub-Tropical Jet
Upper Summer
Easterly Jet
Mid-level West African

Easterly Jet
Somali Low-Level
Summer Jet
FIG. 1.16 The location of seasonal, low-level jets and upper level jets over
Africa in the summer. These jets, perhaps more than any others affect the
well-being of the greatest number of people on the planeteither directly
over Africa, or across Asia where they influence the summer monsoon
rainfall or in the Americas and Pacific where they influence the forma-
tion of hurricanes and typhoons. The underlying African map is courtesy
of Bruce Jones Design Incorporated
Meanwhile, back over Africa to the north of the seasonal east-

erly jet, over the Sahara, and somewhat further north, is a seasonal
westerly jet. Like the easterly jet over northern India this easterly
jet appears once the African interior has heated upparticularly
the southern Sahara. Air above this area begins to rise strongly,
helping draw moisture northwards from the African coastline and
into the Sahel. Here, it helps deliver welcome (if erratic) summer
rains. Like the Indian easterly jet the seasonal African mid-level
easterly jet and the more northerly westerly jet act like an exhaust
system, carting away the warm air that has been dragged in across
the surface. This westerly African jet ultimately directs moisture
eastwards into the interior of Asia where it influences the summer
weather of Russia.
Much like the Somali jet, another low level jet develops over
southwestern North America during the summer months. Again,
this jetdiverted this time by the Rockiesbrings moisture to

the south western States. In turn this helps drive the formation of
thunderstorms and larger clusters of storms that go by the grand
term, mesoscale convective complexes. Mesoscale systems simply
refers to size: these complexes of storms vary in size from 100
500 km across and are variously found in the tropicsforming the
cores of easterly wave ridgesor storm systems that spawn torna-
does in the Mid-Western States in the US.
To the east of India much of the summer rain is concentrated
in a narrow band called the Mei-yu (China) or Baiu (Japan) front
(Fig. 1.18). Here, another south westerly jet over southern China
meets the westerly winds blowing out of central Asiawinds that
may have started out with the monsoon rains over West Africa. As
the two air flows meet, storms are generated which move towards
the north east. Most of the heavy rain and thunderstorms lie to
the south and east of the front where very humid tropical air is
moving northwards. North of the front lie belts of low stratus
cloud and fog, with only light precipitation. Although the Baiu
front owes its origin to the tropics, it is structured very like a mid-
latitude cold front, with colder air to the north of the front under-
cutting the warmer air trying to advance from the south. Just like
a cold front, waves develop along the front at regular intervals,
probably initiated by waves running out of the tropical easterly
flow. These are foci of more intense precipitation bringing regular
pulses of heavy rain to northern China, southern Korea and Japan.
The Mei-Yu front persists throughout the summer months before
retreating back south towards Indonesia in the autumn, where it
decays away.
Meanwhile, over western Africa an easterly jet develops that
has a more insidious effect. As well as usefully directing mois-
ture and storms from the Cameroon coast, northwestwards, this
mid-level jet also directs the movement of squall lines called east-
erly waves. These begin as clusters of storms over West Africa,
which are embedded in the broad easterly flow. As these continue
to grow they are driven off the coast and out into the Tropical
Atlantic. From here, some become sufficiently well organized that
they develop into circular storm systems. It is these, from July to
September (or occasionally later) that may develop into Atlantic
Hurricanes. Thus there is a clear link between the success of the
rainy season in West Africa and the strength of the Atlantic hur-
ricane season.
Aside from bringing hurricanes, easterly waves bring the
Amazon, the Caribbean and Florida another benefit. Easterly
storms often kick up dust storms over the Sahel and southern
Sahara. Until the advent of comprehensive satellite monitoring, it
wasnt appreciated that these storms frequently carry Saharan sand
all the way across the Atlantic before dumping it on the oceans
western shores. Without these easterly propagating storms, the
beaches of Florida and the Caribbeanor, indeed, the lushness
of the Amazonian rainforest, for which they fertilizewould be
severely depleted (Fig. 1.17). One reason that these waves can reach
FIG. 1.17 NOAAs GOES-13 satellite took this image of 4 tropical systems
in the Atlantic on September 8, 2011. Hurricane Katia in the western
Atlantic between Bermuda and the U.S. East coast (A); Tropical Storm
Lee's remnants affect the northeastern U.S. (B); Tropical Storm Maria in
the central Atlantic (C); and newborn Tropical Storm Nate in the Bay of
Campeche, Gulf of Mexico (D). Credit: NASA/NOAA GOES Project. The
location of the polar front jet is also shown in purple and easterly waves
in turquoise. You can see where tropical air escapes northward into the
westerly flow near Nate in the Gulf and where Lees remnants impact
the polar front jet
across the full width of the Atlantic is the presence of another low
level, seasonal jet. This one is very low level, indeed. Extending
from the surface to around 1500 m this Caribbean Low Level jet
directs easterly waves into the Caribbean where they bring sum-
mer rainsand less fortunately hurricanes. The jet also sup-
plies extensive moisture towards the Caribbean coast of Central
America, where the rains sustain the rain forests of Belize, Costa
Rica and the Caribbean coast of Venezuela.
The easterly waves behave in much the same way as the
Rossby waves further north. The moist air pouring in across
Western, equatorial Africa during the summer monsoon season
is much cooler than the hot Saharan air to the north. The differ-
ence in air temperatures generates a strong pressure difference
and leads to the development of the easterly jet stream described
above. This is because the air is moving towards the west as it
crosses the equator and heads towards the center of the Sahara. A
major difference between the westerly jets and this easterly jet is
the pressure surface at which it occurs. The westerly jets occur at
the 200300 mb levelwhich lies between 10,000 and 13,000 m
or the tropopause. The equatorial easterly jet occurs at a lower
elevationapproximately 45000 m above sea level. At this level
the density of air is greater and there is a more direct drive on the
storm systems that develop over Africa and have cumulonimbus
towers that extend upwards of 8000 m.
Meanwhile in the Southern Hemisphere, during the summer
months, a low level jet transfers moist, warm air from the Corral
Sea across Australia towards low pressure areas over the conti-
nent's southwest flank. Again, this can drive the development
of mesoscale storm systems, just as the analogous jet does in the
southwest US.
Both the Somali jet and the US jet owe their existence to the
presence of mountains. In each case the northsouth aligned ranges
intercept the eastwest flow of tropical air and divert it northwards.
Without the activity of these jets normally dry areas would become
deserts and are otherwise sustained by the life-giving rains these
low level jets bring. That said, the activity of the US jet has been
marked by its absence (or at least its ineffectiveness) and a pro-
longed drought has afflicted the western States. This highly unfor-
tunate state of affairs has been linked to a persistent wobble in
the Polar Front jet to the north. Instead of migrating the overlying
Rossby wave has led to the persistence of a ridge of high pressure
over these States, which has served to maintain the dearth of rain.
How the Jet Stream Brought a Sting

to the Bergen Frontal Model
At the time of its conception the Bergen frontal model was based
exclusively on surface pressure, wind and temperature measure-
ments. Although it was obvious such surface features were linked
to processes occurring higher in the atmosphere, there was no
means to determine what was going on at these upper levels. The
jet stream, or rather the jet streams were unknown except in out-
line. Therefore, accurate though the Bergen model is, it does not
fully describe meteorological phenomena, much like Newtons
Law of Gravitation does not describe gravity in extreme environ-
ments.
Over the decades since the Bergen model was synthesized a
number of storm systems have required meteorologists to re-think
what is going on. In particular a class of mid-latitude storms is
known that appear to explosively develop, with a rapid decrease
in pressure and narrow walls of extreme winds that defy a straight-
forward explanation. Three of these are noteworthy, but by no
means unusual: the 1979 Fastnet Storm; the 2011 New Zealand
Bomb; and the 1993 US. superstorm. Although quite distinct
storm systems in their own rights, each system shared a com-
mon feature: a rapid drop in their central pressure alongside the
development of hurricane-strength winds around the core of the
low pressure core.
Fastnet
In August 1979, the international yachting race, the Admiral

Cup, was underway in the English Channel. The 605-mile course
ran from Cowes, direct to the Fastnet Rock and then west to
Plymouth, via the Scilly Isles. The Fastnet stretch was the climax
of the race and was, at the time, one of the most prestigious races
of its kind.
On August 11th a small low pressure area developed to the

south west of Ireland and began intensifying rapidly as it went
north eastwards towards the Irish Sea. Winds of 5060 km per hour
had been forecast, significant in their own right, particularly for the
month of August. However, by August 13th winds continued to
strengthen to near hurricane forcestorm force 1011 (80100 km
per hour). Caught in this cabal were 306 yachts. Five yachts were
sunk outright, with a further 100 suffering knock-downs (when the
mast and sail are tipped over at 90 or more and the sail fills with
sea water) and 75 were flipped over. Over the ensuing 24 h 15 sail-
ors lost their lives and 125 required rescuing. Of those 306 yachts
that began the raceonly 78 were to complete it. The racing disas-
ter forced a re-think of the industrys safety procedures.
Although the storm was fairly intense (979 mb) it was not
extraordinary. However, the Fastnet storm, as it is now known
had, quite literally a sting in its tail. Lurking a hundred or so kilo-
meters behind the cold front was a comma-shaped cloud swirling
around a largely cloud-free zone. Analysis soon showed that this
cloud-free feature was created by a rapid downburst of air from
the middle and upper levels of the troposphere. This downward-
directed jet is known as a sting-jet and is a frequent cause of extreme
winds in the mid-latitudes. These sting jets are not explained by
the conventional Bergen modeland this is hardly surprising as
the meteorologists of their day had no access to information from
higher altitudes.
The effect of the sting jet is twofold. As we have seen it
rapidly strengthens windsin that case of the Fastnet storm,
throughout a region around 150 km across. This is caused by the
rapid descent of cold, dry stratospheric air. However, the jet also
punches through the cold front and dislocates its southern portion
from the rest of the storm. Rapidly descending cold air essentially
drives the cold front so quickly that it moves independently of the
rest of the storm. The warm front then wraps around the center of
the low pressure forming a second front with the characteristics
of an occlusion to the storms rear. Over the space of a few hours,
the cold front and the associated low plow through the storm's
warm sector while the trailing occlusion wraps around the core
of the low, its heart driven by the descending flanking column of
cold air. It is along this curling frontal boundary that the strongest
winds are found. The standard Bergen model fails to predict this
kind of explosive behavior. However, a second model, developed

some 70 years later does fit the bill. The Shapiro-Keyser model, as
it is known, was developed in 1990, and adequately explains these
explosively developing storms. Warm air is trapped in a pocket
at the lows core, which is aggressively undercut by cold air that
is surging in on its southern flank. Much like a tropical hurri-
cane, within the warm core, air can ascend rapidly. This causes a
very rapid deepening of the low's central pressure, which, in turn,
intensifies the winds that wrap around it.
The New Zealand Bomb
In the Southern Hemisphere, a very similar set-up drove the explo-

sive development of a storm over New Zealand in March 2012.
This storm brought fairly widespread damage to the southern end
of New Zealands North Island. As the low, knick-named the
bomb, approached from the south west during the early morn-
ing of March 3rd, winds picked up to over 80 km per hour, with
one gust measured at over 120 km per hour at around 5 am. After
6 am all wind speed readings ceased at Hawera station when the
site suffered a power cut as a result of the storm. Meanwhile to the
north of the storm center, warm air dragged over the mountains
underwent the Foehn Effect. Warm moist air cools more slowly
and warms more slowly than dry air (49 C per kilometer ver-
sus 9.8 C per kilometer, respectively). As the air moves over any
mountains, the moisture is wrung out of it, leaving it bone dry.
Condensation also releases latent heat further heating the air mass
on its ascent. If the air cools by 10 C when it ascends the windward
side, it may then warm by 1321 C on the leeward side, depend-
ing on the amount moisture and its original temperature. Thus to
the west (leeward side) of the central chain of mountains, stations
recorded a sharp rise in temperature as the storm approached and
tracked away to the north east.
The 1993 US Superstorm
Although the 2011 New Zealand storm was fairly severe, the 1993
super-storm that afflicted the eastern seaboard of the US was
altogether more infamous and for good reason. This storm became
a classic Noreaster: a storm that while positioned off the east-

ern seaboard, drags very cold but moisture laden air from the Sea of
Labrador. Therefore, in addition to severe coastal gales, noreasters
also bring a dreaded harvest of snow during the winter and early
spring months.
The 1993 superstorm began its life rather innocently over
Cuba as a squall line on March 12th. Moving north eastwards from
the Gulf of Mexico, and driven along by a strong jet stream, the
storm rapidly intensified and brought devastation to many central
and eastern States of the US. in variety of distinct ways.
The key to understanding this storm's explosive devel-
opment is the geography of the eastern US. During the winter
months the continental landmass cools rapidly and the bulk of
it has a mean temperature well below zero. Meanwhile, drifting
north along the eastern seaboard are the warm tropical waters of
the Gulf Stream. When this particular storm struck, in March,
the continental interior was still very cold and the contrast in
temperatures between the land and sea, severe. As the storm
developed and moved north east with the overlying jet stream, it
dragged very warm, humid air with it. During the 13th of March
this mass of warm, humid and unstable air was progressively
undercut on its western flank by the cold continental air. Such a
configuration allowed the development of a very active cold front
and a rear-side sting jet that drove its northern portion rapidly to
the east. Like the Fastnet and New Zealand storms, this set-up
drove explosive development of the storm. By the time the low
center passed Washington DC it was packing storm force winds,
with gusts up to 180 km per hour.
At its peak the storm stretched from the Florida Everglades
northwards to southern Canada. Entrained to its rear was such cold
air that snow fell as far south as the Florida Panhandle: 10 cm were
recorded in places. Further north, in Alabama, snow fall exceeded
40 cm. Indeed, Floridas snow was the most severe since 1899.
Further north, the Superstorm dumped over a foot of snow over
much of the north east of the continent. Snowfall totals exceeded
1 m over the Appalachians, with blizzards whipping depths to
over a whopping 11 m in places. Indeed, calculations suggested
that up to 27 billion tones of snow fell over the day it took the
storm to pass.
As well as widespread hurricane-force winds, the active cold-

front spawned several lethal tornadoes. The combination of tor-
nadoes and storm surges across north-western Florida, dozens
of people were killed in that region alone. From Texas north to
Pennsylvania blizzard conditions were often mixed with thun-
der and lightning giving rise to the fairly exotic phenomenon of
thundersnow. Here, vigorous convection along the advancing
cold front produced thundercloudscumulonimbuswhile the
undercutting cold air ensured that the precipitation was in the
form of snow rather than the usual rain or hail.
In the storms immediate aftermath ten million people went
without power as storm force winds and heavy snow brought down
power lines. In all, 318 people were killed with over 40 % of the
countrys population feeling the effects of the storm. Many thou-
sands of homes remained without power for over 3 weeks. The
diversity and spread of damage, the number of people affected and
the severity of snowfallin places exceeding an entire winters
averageensured that the 1993 super-storm will be remembered
as the storm of the century, for many.
Like the Fastnet and New Zealand storms, the systems power
was enhanced by an upper level jet directing cold air rapidly down-
wards towards the ground to the rear of the cold front. As the storm
tracked northeast the cold front accelerated eastwards. With rapid
uplift of the warm air to its east, there was ample opportunity
for the initially shallow low to explosively deepen. The track of
the storm ensured that warm air was always available along the
systems eastern flank as the upper level jet drove it north east.
In the end this favorable track, coupled to the very constructive
topography of the eastern United States ensured that the system
had ample power to devastate the eastern States.
The Tropics: A Quick Guide
Lets now leave the chill of the mid-latitudes behind and exam-
ine the weather over the Earths tropics. Arguably, weather in the
Tropics is a simpler affair than more southerly and northerly lati-
tudes. Weather is driven predominantly by convectionthe rise
and fall of air that is dependent on its temperature and relative
density. In the tropics, air flows from the flanking high pressure
areas, centered around 30 north or south of the equator. The core
of the region is a belt of calm conditions and severe thunderstorms,
famed as the Doldrums. However, this apparent simplicity is decep-
tive. Aside from the seasonal northward and southward migration
of the storm belt that forms the systems core, the storms within
the belt are not continuous. Rather they form distinct pockets that
migrate from east to west, within the general airflow: some of these
are the easterly waves.
However, thats not all. Lurking within the general easterly flow
of air lies a pattern of variability akin to the Rossby waves embed-
ded within the mid-latitude jet streams. Named after its discover-
ers, Roland Madden and Paul Julian, the Madden-Julian oscillation
(MJO) is a slowly propagating wave that moves eastward against the
prevailing wind. These waves begin their life in the Indian Ocean,
before moving slowly towards the Pacific at 1429 km per hour. As
these waves move, they take with them a region of enhanced con-
vection on their leading edge, much like easterly waves. Linked to
them, but with a shorter wavelength and a faster easterly velocity
are so-called Kelvin waves. These regions have enhanced convec-
tion associated with their leading edge as well as strong westerly
winds lurking in their rear. These features often appear to kick-start
MJO events, but with faster movementup to 70 km per hour
pull away in front of MJO waves over the Pacific.
MJO waves manifest themselves first with enhanced rain-
fall, initially over the western Indian Ocean. This is followed
by a period of increased drought. In most cases they die out over
the cooler waters of the Eastern Pacific but on occasion they
travel right around the globe, weakened in the Atlantic, before
re-strengthening once again in the western Indian Ocean. Waves
emerge and propagate roughly every 12 months, predominantly
during the northern summer months. There is some hint that
these waves (unlike their speedier Kelvin Wave cousins) may help
initiate El Nio events (Chap. 2), but why some do and some dont
is as yet unclear. Certainly, the west to east movement of MJOs
would help encourage a reversal of the normal east to west pattern
of wind and rainfall, generating the reversed pattern that charac-
terizes the El Nio.
On extraterrestrial planets similar tropical waves are thought

to play a crucial role in the development of a phenomenon known
as super-rotation, where the equatorial atmosphere moves more
rapidly from west to east than the ground (or deeper atmosphere)
underneath (Chap. 10). Moreover, there are also some suggestions
that pairings of equatorial Kelvin Waves and westward moving
equatorial Rossby Waves3 not only kick-start MJO events but help
initiate tropical cyclones. They may also initiate trains of storms
that pummel the northern, mid-latitude, Pacific coastline during
winter months.
In the Pacific warm waters are driven towards the west by
the prevailing winds. If the Earth lacked landmasses, this water
would simply spread out evenly around the tropics near where
heating was strongest. However, Australia, Indonesia and Asia get
in the flow of this warm surface water. Therefore, it backs up in
the Western Pacific, with a similar, though less extreme parallel
development in the western Atlantic and Indian Oceans. Where
the warmest waters are you find the greatest extent of convection.
Warm, humid air rises most strongly here leading to the highest
rainfall totalsin most years. Over the western Pacific you also
find the greatest preponderance of tropical storms. These roughly
circular areas of low pressure may start out as easterly waves
over the Atlantic, before pushing westwards across the Pacific. In
the summer months, when these easterly waves are located fur-
thest from the equator they are influenced by the Coriolis Effect
and can become organized into circular storm systems. Over the
western Pacific where the summer monsoon rain belt intercepts
these waves the stage can become set for the rapid development of
storm systems. When these develop winds in excess of storm force
12 (110 km per hour) they take on the name Typhoon; in the
Atlantic Hurricane, Cyclone in the Indian Ocean and Willy-
Willy to the north of Australia.
Such tropical storms afflict the western edges of the Pacific
and Atlantic in particular, but as northern storms continue to
move away from the equator they become entrained in the overall
3
Equatorial Rossby Waves are indicated on surface pressure charts by pairs of high pressure and
low pressure that alternate along the equator and move from east to west like their larger mid-lati-
tude cousins. Each low or high to the south of the equator has a matching high or low (respectively)
on the northern side.
northerly flow of air and enter the mid-latitudes. Here they deliver
tropical warmth and moisture to countries as far north as Canada
and the UK. A similar, but less extensive process happens with
Australasian cyclones that enter the southern mid-latitudes north
of New Zealand.
The tropical storms are an obvious deviation from the regular
pattern of calm simplicity. However, they are not alone. Regional
topography plays a crucial role in the development of other tropi-
cal storms. Over Sumatra, during the summer, easterly winds
deliver moisture and vigorous storm systems develop over the
highlands that run along its length. These storms become orga-
nized into waves as they move out over the warm waters of the
Straits of Malacca, towards Singapore and south western Malaysia.
These Sumatras typically develop during the evening before slam-
ming into the coastline, bringing overnight storms. Sumatras bear
a striking resemblance to a number of other terrestrial and other-
worldly storms and we will hear more about these later (Chap. 7).
Finally, within the tropics and between the tropics and the
poles are Rossby waves. These are evident as paired areas of low or
high pressure that flank the equator. These waves not only orga-
nize convection along the equator but help transport momentum
to and from the equator. They will come into their own in Chap.
10, when we look at the circulation of the atmospheres of hot
Jupiters.
Interconnections
The tropical weather systems and those of the mid-latitudes and
Poles talk to one another in a number of ways. High above the sur-
face, at the level of the tropopause, air escapes the tropical Hadley
cells and leaks northwards into the mid-latitude Ferrel Cells and
the Polar Cells. At the surface, particularly during the summer and
early autumn, winds blowing around the limbs of the Horse Latitude
high pressure areas bring tropical warmth into the continental
interiors of China and the US in the northern hemisphere, and
across Australia and New Zealand during the southern summer.
These airflows carry tropical disturbances, including hurricanes,
across more polar latitudes. In the summer of 2014, the remains of
Hurricane Bertha swept across the Atlantic towards the UK. Along
the way the decaying storm became embroiled with a mid-latitude
frontal system and morphed into a rather aggressive frontal storm.
Although it was greatly reduced in its severity, it still packed a lot
of tropical energy. Fronts crossed the UK, from Southern Scotland
to northern France. Having driven south in it, I cant emphasize
enough the effect of all that tropical heat. Although winds were not
severe, the rainfall was impressive. Leaving Glasgow and heading
south over the Southern Uplands, the sky was nearly black, with
waves of torrential rain that made driving a rather interesting
experience. By the time we reached the Midlands, in the heart of
the UK, the rain had cleared east and a very blustery, autumnal
airflow sent clouds briskly across the sky.
Although Bertha did take a fairly southerly track, the passage of
hurricane remnants to the north of the UK is a more common expe-
rience in September and October. Such storms are largely welcome
across much of the UK as they usually redirect the prevailing winds,
bringing considerable warmth from the south. As autumn begins to
tighten its grip, these northerly-tracking extra-tropical storms bring
a return to summer temperatures across Western Europe.
Other connections include the link between Africa and cen-
tral Asia. As was mentioned earlier, air that began life over West
Africa may also be driven north eastwards towards southern
Russia during the summer months. This can bring excessive heat
and in some instances violent summer storms.
During the autumn and winter months cold fronts often pen-
etrate far to the south over Africa. Although these were previously
mentioned, their role in transporting sand and dust all the way to
the Americas cannot be understated. This delivery system affec-
tively cycles air from the mid-latitudes to the tropics and back
again over North America.
A more limited but sinister link connects South America
and North America. During the spring, when parts of the Amazon
Basin are set ablaze by farmers clearing forest for farming, smoke
fills much of the basin then becomes entrained in air moving
north across the Gulf of Mexico. Much of this activity is illegal
but goes on nonetheless, because the world demand for cheap beef
continues to rise with its growing population. Other forest areas
are cleared for palm oil or other cash crops.
Now, if you are sitting back watching the television (or read-
ing this book) and thinking that none of that matters to you,
then think about this piece of research. Published recently in
Geophysical Research Letters, Pablo Saide, an atmospheric scien-
tist at the University of Iowa, and his colleagues, used NASAs
Aqua satellite to monitor land clearance in the Amazon. Analysis
showed that when fires were lit across the Amazon Basin, the
smoke became entrained in air moving northwards across the
Gulf of Mexico and into the heartland of North America. Rather
surprisingly, the amount of smoke in the air traveling north over
the mid-west clearly enhanced the severity of the April 2011 tor-
nado outbreak: the deadliest on record. Modeling revealed that
black smoke particles from the burning Amazon enhanced low-
level cloud development over the Mid-West; and enhanced low
level wind-shear both of which are known factors in stimulating
the formation of tornadoes. During the spring months this warm,
humid air was given an extra kick from the aerosols hidden within
it. That cheap side of beef might just cost you more than a few
bucks
However, it must be stressed that aerosols have a variety of
effects. While they appear to ramp up supercells (and thunder-
storms in general) conversely, sand storms, over the Sahara appear
to depress the severity of Atlantic hurricanes. Over the Indian
Ocean aerosols decrease the intensity of the south west summer
monsoon and the easterly jet that overlies it. Ironically, this allows
for a greater intensity of tropical storms to form over the Arabian
Sea before and after the summer monsoon. This is apparently the
result of a weakening of the upper level easterly winds and the
southwesterly monsoon winds as a result of the effects of smoke
and other aerosols. The role of these aerosols is discussed more
fully in Chap. 2.
From September onwards the Asian monsoon swings into
reverse gear. As the vast bulk of Asia cools down during the
autumn, the westerly winds re-establish far above it. Over Siberia
a vast pool of frigid air develops that pours outwards, south east-
wards across China, India and Indonesia. By December this air has
flooded as far south as north Western Australia and the bulk of the
Indian Ocean. Over Indonesia and the Philippines the winds bend
from a northerly to a more easterly direction. As this air passes over
the warm waters of the western Pacific the air becomes unstable
and generates a lot of rainfall particularly across the Philippines,
Borneo and eastern Indochina. Indeed, much more rain falls over
this segment of easternmost Asia during these months than it does
throughout the summer monsoon. This humid cool air mass then
delivers rainfall along the northern territories of Australia. Some
of the air is sufficiently unstable that it generates tropical storms,
but the majority falls in smaller, so-called mesoscale systems,
similar in size and severity as the easterly waves that drive west
from Africa and into the Pacific.
Although the frigid air mass is considerably tempered by
the passage across the warm Indian Ocean, it serves to bring air
from the Arctic all the way south towards the Antarctic Ocean.
Although less extensive, the same process happens over North
America. Cold, Arctic air floods southwards behind storms, which
are moving up the continents eastern seaboard. This Arctic air,
again modified by its journey, crosses the Caribbean towards the
northern edge of South America.
In the Southern Hemisphere the lack of significant land
masses spanning several degrees of latitude limits the movement
of air in this manner. However, air still enters the tropical circula-
tion from the polar westerlies to the east of New Zealand and the
east of South America. In particular during the summer months a
strong area of converging air is found stretching south east from
Papua New Guinea to about 30 S, 120 W. This region, known
grandly as the South Pacific Convergence Zone is distinctly tropi-
cal in nature at its western end.4 However, as you track to the
south east, westerly winds increasingly mix into the tropical cir-
culation and instead of spawning tropical disturbances generate
broad mesoscale storms in a manner analogous to the Mei-yu/Baiu
front over Eastern Asia (Fig. 1.18).
There is one more connection worth mentioning from its
influence on the US and the UK. The Madden-Julian Oscillation
(described above) has a rather sinister connection to extreme win-
ter rainfall patterns in these parts of the world. In non-El Nio
years the MJO can generate large clusters of tropical thunderstorms
4
There is an equivalent zone of convergence over South America during the southern hemispheres
summer. Converging air in this zone brought severe ooding during the 20152016 El Nino event.
Mei-Yu Front
HIGH Low
Monsoon
Trough
HIGH
HIGH
Low Low
Low
FIG. 1.18 The major boundaries between air movement in the Pacific. To
the north west, over China, Korea and Japan lies the Mei-Yu front. South
of this lies the monsoon trough where tropical cyclones regularly develop
in the late northern summer and early autumn. South of the equator the
monsoon trough extends towards South America as the south Pacific
Convergence Zone (S.P.C.Z.). Like the Mei-Yu front this feature connects
the tropical circulation with the westerlies that lie further towards the
Poles. The major equatorial band of convergence is the Intertropical Con-
vergence Zone (I.T.C.Z.). Arrows indicate averaged direction of air flow
(mesoscale disturbances) in the western Pacific as the oscillations

propagate towards the east. Indeed, during these cooler years, MJO
activity is enhanced. Where the La Nia is established (Chap. 2),
or when otherwise a large blocking anticyclone develops over the
northern Pacific, warm moist air can become entrained within
the winter flow of air that runs eastward towards the Oregon and
Washington State coastlines. The jet stream, flowing just to the
north of the Hawaiian Islands picks up the moisture from the tropi-
cal MJO thunderstorm complex, then whisks it eastwards towards
the Oregon coastline. Within the westerly flow, the Rossby wave
that bears the blocking anticyclone drifts westwards towards east-
ern Russia, while the strong flow of air along its northern flank
weakens.
Blocked to the north, the jet stream then breaks loose and
drives along the blocks southern edge towards the Oregon
coastline. This super-charged air encourages the development of
deep low pressure areas off the northwestern coast of the lower
continental States. From here a vigorous train of low pressure
areas can develop which then batter the western coastline. In
particular, the winter of 19961997 brought $23 billion dollars
worth of damage to the Pacific Northwest, in the form of exten-
sive flood damage. These patterns, although not common, are
far from rare and have earned the title the Pineapple Express
after their origin near the Hawaiian Islands. The Express
brings a succession of rain-filled storms that have earned the
pattern its other namean Atmospheric Rivera stream of
densely packed moisture that can bring catastrophic floods as
far south as Central California.
The Madden-Julian Oscillation also influences the intensity
of summer monsoon rains across southernmost North America.
Eastward propagating waves intensify then weaken rainfall associ-
ated with the monsoon. Consequently, the MJO affects the mois-
ture content of the air across the entire continent and the intensity
of rainfall along the Polar Front jet stream further to the north.
Despite their origin in the equatorial Pacific, the flow of air
from these mesoscale storms can influence the weather much fur-
ther afield. A pattern, similar to the Pineapple Express developed
during the winter of 20132014, with a persistent blocking anti-
cyclone located over the northern Pacific. Once again clusters of
tropical thunderstorms developed in the western Pacific associ-
ated with the MJO and injected a large amount of moisture into
the atmosphere north of the Tropics. On this occasion, the block
remained anchored in the northeast Pacific and the injection of
energy and moisture drove the Polar Front jet stream into a persis-
tent pattern, arcing over Alaska then down across the Mid-West
then out into the North Atlantic. With the block locked in place
the jet to follow this path for months on end. Thus the Pacific
blocked pattern conspired with the intense tropical thunderstorms
to sweep storm after storm across the heart of the UK through-
out the three winter months. Flooding and storm damage were
unprecedented, even coming as they were on the back of record
breaking summer floods in 2007 and 2012. Much of south west
England lay under water for 3 months, while 50 m of the Victorian

train line that ran along the Devonshire coast was left hanging
mid-air, when storms washed away the embankment that it was
cited on. Thus distant effects in the Pacific can play havoc with
the weather half a world away.
Such connections are an essential means by which gases are
mixed throughout the atmosphere as well as ensuring the efficient
transport of heat from warm to cold regions. Much of the surface
and lower atmospheric connections on the Earth are driven by
topography that neatly diverts airflowin particular diverts air
between different latitudes. Without this the Earths atmosphere
would obviously still function; after all, you dont see Jupiters
tropics overheating while the poles become so cold that its
gases collapse out. Instead of these overt northsouth transfers of
heat that defy Gustav Corioliss effect, the atmosphere would dis-
play enhanced east-west transport of energy, while waves trans-
ferred large quantities of this energy to the north and south. The
remainder of the heat would be dissipated by convection so that
the atmosphere would retain its stability.
Conclusions
The Earths weather and climate is a very complex beast, indeed.
Only in recent years have the connections between different lati-
tudes and altitudes become apparent. The exchange of energy
and momentum between different climatic regions is perhaps the
most important process that happens on the Earth and probably
any planet. Such exchanges moderate the extremes of temperature
and on our planet at least ensure a relatively steady delivery of
moisture between regions.
Yet this pattern is prone to instability. The jet stream can
become locked into waves and troughs that deliver an often unwel-
come barrage of meteorological phenomena such as droughts or
floods. The underlying reasons for this change in behavior are as
yet unclear. However, the effects can be devastating.
On other planets, the same underlying principles will apply,
therefore, the more we understand about conditions within the
earths turbulent atmosphere, the more we will understand about
the myriad of new worlds that are being discovered, as well as the
other planets of the Solar System. Yet, this is a two-way street:
from our understanding of the greenhouse effect on Venus weve
learnt about how we can modify our own world through our activ-
ities. Once again, precisely how our behaviors lead to the changes
we observe requires further elucidation. It is to the extremes of
terrestrial climate that we turn to next as we begin our journey
outwards from the Earth to the rest of the universe.
References
1. Saide, P. E., Spak, S. N., Pierce, R. B., Otkin, J. A., Schaack, T. K., Heidinger, A. K., da Silva,
A. M., Kacenelenbogen, M., Redemann, J. & Carmichael, G. R. (2015). Central American bio-
mass burning smoke can increase tornado severity in the US. Geophysical Research Letters,
Retrieved https://www.researchgate.net/publication/271384910_Central_American_biomass_
burning_smoke_can_increase_tornado_severity_in_the_US. Available free on ResearchGate.
2. Evan, A. T., Ramanathan, V., Kossin, J. P., & Chung, C. E. (2011). Arabian Sea tropical cyclones
intensied by emissions of black carbon and other aerosols. Nature, 479, 9497.
3. Madden Julian oscillation. Retrieved from http://www.cpc.ncep.noaa.gov/products/precip/
CWlink/MJO/MJO_1page_factsheet.pdf.
4. Madden Julian oscillation impacts on the US. Retrieved from https://www.climate.gov/news-
features/blogs/enso/what-mjo-and-why-do-we-care.
5. Kelvin waves. http://www.ess.uci.edu/~yu/class/ess228/lecture.6.adjustment.all.pdf.
6. Stevenson, D.S. (2016) An underlying predictability in winter weather patterns in the North
Atlantic Basin. Hypothesis, 14(1): e3, doi:10.5779/hypothesis.v14i1.483.
2. Climate Oscillations in Space
and Time
Introduction
The East Asian Monsoon is perhaps the most important climatic
feature of the Earth. Across 60 of latitude and around 80 of lon-
gitude, winds reverse direction from summer to winter and back
again, year after year. The cycle brings rains to two continents:
Asia in the northern summer and northern Australia in the south-
ern summer. The present monsoon has been active for tens of mil-
lions of years and undoubtedly, systems like it have been active on
Earth since the first continents emerged from the dark blue waters
of the early Earth billions of years ago. This chapter examines the
broad nature of the monsoon and the broader context of global cli-
mate in which it exists. In Chap. 6, we will see how the Martian
landscape also experiences its own ghost-like version of the ter-
restrial East Asian Monsoon. Finally, in Chap. 10 we will see how
some planets may take a monsoon pattern of winds to extremes
with circulations that run from pole to pole, year after year.
The East Asian Monsoon resides in the broader canvas of
terrestrial climate and as such is subject to the influences these
other cyclical patterns of climate exert upon it. Of these the El
Nio Southern Oscillation, or ENSO for short, is perhaps the best
known, but others exist in the Atlantic and Indian Ocean basins.
The ENSO, and one like it in the Indian Ocean, has a global reach,
but unlike the East Asian Monsoon, does not seem to follow a
predictable pattern. While the Asian Monsoon is clearly driven
by the Sun and the tilt of our planet, the ENSO and Indian Ocean
Oscillations are more like the backwards and forwards sloshing
of water in a disturbed bath. The interaction of the two oscilla-
tions and the monsoon influences the well-being of billions of our
planets citizens and similarly, one must expect that such patterns

DOI 10.1007/978-3-319-25679-5_2
will exist on other Earth-like planets that we will be discovering

in years to come. As such, although this chapter may, on a univer-
sal scale seem rather parochial, it will undoubtedly relate to other,
foreign worlds in ways we can only imagine at present.
The Present Monsoon

Over nearly a quarter of our planets surface the pattern of wind
reverses direction every 6 months. The driving force for this
massive redistribution of air is the asymmetric distribution of
land on the Earth's surface. For the last 200 million years or so,
there has been a concentration of land to the north of the equa-
tor, and more importantly for much of this time the land has been
mountainous, to varying degrees. It is the combination of land and
mountains that really spurs on the monsoon pattern of airflow
that typifies many of the continents today.
The underlying physics of the monsoon is simple: land heats
up much faster than water. because land has a lower specific heat
capacity than water, it takes less energy per unit of mass to warm
land than it does water. Moreover, soil and rock are poor conduc-
tors of heat, so that when the Sun heats the land, most of that
energy is initially trapped in the topmost layers. This excess heat
is then available to warm the air above. Oceans, by contrast, not
only are made of water on our planet- which has a rather high heat
capacity (several times that of rock)they also have this annoying
habit of moving around. Such movement distributes what heat
energy they absorb from the Sun both vertically to some extent,
and horizontally to a much greater extent. Thus, although water
can store an awful lot of heat energy, it tends not to warm up very
much compared with rock exposed to the same solar battering.
This means that during the summer the land is much
warmer than the water around it; and, in the winter, land cools
down much faster than any neighboring water. As pressure var-
ies with temperature, the higher the temperature the higher the
pressure. This causes the air to expand, lower in density and then
rise under gravity. Such warm, low density air, rising over the hot
land lowers the pressure at the surface, but keeps it higher aloft.
Climate Oscillations in Space and Time 57
With low surface pressure air is drawn in from the surroundings,

in this case the neighboring oceans.
Over India, the air begins to warm in March and, by May the
atmosphere at the surface regularly cooks to over 40 C. Adding to
this mle is the presence of a jet stream and a zone of convergence
called the Inter-tropical Convergence Zone (or ITCZ, for short).
High above the developing surface low pressure area an easterly
jet stream develops and begins vacuuming away the rising air. This
jet continues westwards, ultimately crossing West Africa, contrib-
uting to the same effect here (Chap. 1). Although a surface low
pressure area develops as early as April, it doesn't begin drawing in
much moisture until the easterly jet has developed over southern
India in June. The ITCZ is the zone of maximum convergence, or
bringing together of hot moist air on the planet. To re-cap, this
band, which is experienced at the surface as the Doldrums, con-
tains the zone of maximum precipitation in the tropics and moves
north and south with the Sun. When the Indian vacuum cleaner
is working at its strongest in the summer, this band is pulled north-
wards across the sub-continent and brings the most intense rains
with it. Its arrival effectively marks the beginning of the summer
monsoon in June across the south; and its end in September when
it pulls back towards the equator, propelled in part by increasingly
strong north easterly winds to its north.
This combination of hot surface and upper level vacuum
cleaner ensures that air rises very efficiently: however, it is not
quite enough in itself to drive the full summer monsoon that we
see today. Instead it turns out that India's rather well-known range
of mountains to its north does the final trick: the Himalayas.
North of this lies a large expanse of land over 5000 m above sea
level: Tibet. Both of these immense structures serve to funnel air
upwards in a focused belt, intensifying further the surface low
pressure area. Having a plateau 5000 m up, means that during the
summer there is intense heating of the middle troposphere and
it is this that finally drags the summer monsoon winds upwards
across all of the Asian sub-continent. Together, all of this suction
(the heating of India and Tibet; the Himalayan mountain front
and the easterly jet above) that finally drags in cooler and much
moister air from the Indian Ocean.
You can see the effect of the Himalayas and Tibet by comparing
the northernmost extent of the summer monsoon rains. Although
Iran and the Gulf States are heated as strongly as India, it is only
India that experiences the full, wet monsoon. Heated by the land
and driven upwards by various highlands and the Himalayas to the
north, moisture-laden winds condense their cargo of moisture into
a raging belt of thunderstorms. This delivers a couple of meters of
rainfall over the summer months. This is the summer monsoon.
Weaker monsoons also occur in other parts of the world.
Over the south western states of North America, strong heat-
ing in the summer also generates a low pressure area that sucks
in air from the Gulf of California and the Gulf of Mexico. A low
level southerly jet stream develops along the eastern flank of the
Rockies which helps drag moisture northwards and develop spo-
radic but crucial rainfall in the hot summer months. Over Africa,
the summer monsoon is also marked by the development of low
pressure over the Sahara; an easterly jet (or rather two easterly jets)
over the Sahel and the arrival of the ITCZ over western Africa.
Ironically, as this moves north, a change in the circulation of the
South Atlantic, caused by the onshore monsoon winds, brings
cooler waters towards the equator and an end to rainfall across
the West African coast. This effect isn't as apparent over southern
India because the waters to its south are always warm.
As the position of the overhead Sun moves back south again
in September, the zone of maximum heating moves south with
it. Convection weakens over northern India and the easterly jet
stream weakens, falters and ultimately dies away. The moist south-
westerly airflow is replaced by a cool northeasterly flow and the
rains retreat back south towards the Indian Ocean. Over northern
India pressure rises and the polar front jet stream becomes estab-
lished south of the Tibetan plateau: the winter season has begun.
This pattern is now fairly well understood. However, embed-
ded within this greater understand there remain a number of
issues, such as patterns of strengthening and weakening that vary
on decadal timescales, along with a more generalized weakening
of the summer monsoon over recent decades. Some of these vari-
ations are caused by changes to the pattern of ocean and atmo-
spheric circulation over the Indian and Pacific Oceans, while
others appear to relate to the effects of soot and other pollutants.
Clearly, if we also seek to understand the past Monsoon we

can get a better handle on the factors that will affect the present
one as climate change takes hold. This chapter looks squarely at
the monsoon and how climatologists have begun to probe how
this great weather system operates. Through an improved under-
standing of these changes scientists, economists and politicians
can begin to plan more carefully for the future. For within the
hands of the Asian monsoon, lies the fate of more than two billion
people.
Shifting Continents, Shifting Climate

On the Earth, and presumably many other rocky planets, the sur-
face of the planet is in a state of constant flux. Heat is generated
within the mantle by a combination of radioactive decay and heat
delivered by a cooling core. On top of this is a cold, dense, upper,
rocky layer called the lithosphere, which includes the crust on
which we live. Comfortable though we are this is an unstable situ-
ation, where the dense, cold lithosphere wants to sink under grav-
ity towards the hot core.
While the oceanic lithosphere is easily dense enough to sink
into the mantle, the continental crust is far too light and fluffy
to descend far. Thus, as the lithosphere fractures and shuffles in
response to these forces, continents drift around like blocks of poly-
styrene on a pan of boiling water. Drifting continents mean that
occasionally they collide and when they do they generate chains
of mountains, such as the Himalayas or the Alps. Elsewhere, a
combination of strong pulling forces acting on the their edges and
hot currents rising from below can split continents asunder, form-
ing new ocean basins, much as you see in the Gulf of California or
the Red Sea.
Shifting continents bring an inevitability of concurrently
shifting climates. For the continents themselves, the process of
plate tectonics, as this is known, can carry them north or south
from one climatic belt to another. This is a fairly passive and rel-
atively sluggish process that can take tens of millions of years.
Alternatively, where a continental nugget moves across the path
of the prevailing winds it might direct warm ocean currents and
the winds themselves further towards the polesor it might end

up over one of the polar regions. This kind of movement can have
more sudden and profound effects on the climate of the globe as a
whole.
Alongside the slow dance of the continents are periodic larger
and more rapid redistributions of land called True Polar Wander
(TPW). Here, the Earth shuffles the land on its surface to bet-
ter balance its spin and momentum. Where the continents have
moved closer to the North or South Pole the spin of the planet
(and the pull of the Moon and Sun) tend to want to pull this extra
mass into alignment with the Earths spin. This involves a whole-
sale shifting of the continents on mass until they are more suit-
ably redistributed around the equator. Such movements probably
take around 1020 million years and will clearly have dramatic
consequences for the Earths climate and the life that populates
its surface. Although not too much will be said about these TPW
episodes, hold the thought in the back of your mind when we
examine the positions of the continents in each of the ensuing
geological periods in Chap. 3.
It is to these seismic shifts and to the impacts these have on
planetary climate that this chapter turns its attention. The focus
remains on the Monsoon but painted within a broader canvas of
continental drift.
Probing the Past Asian Monsoon

How does science begin to establish the pattern and driving force
of a climate system 50 million years ago? After all one can hardly
get a snapshot of cloud patterns in the fossil record Or can you?
As it turns out the geological record is incredibly helpful, if you
know where to look.
In general there are two very useful methods available to cli-
mate scientists (and geologists in general). The first is the most
obvious: fossil sea shells. A cursory examination of fossils tells
you what sorts of organisms were present and where they were.
This can tell you about the general conditions, and on closer exam-
ination sea shells can tell you a lot more. Sea shells are mostly
calcium carbonate (chalk/limestone). This compound contains
three oxygen atoms and is a product of the reaction between car-

bon dioxide, water and dissolved calcium. The water contains one
oxygen atom for every two hydrogen atoms, and it is this oxygen
atom which is of critical importance to climatologists. Oxygen
comes in three nuclear flavors: oxygen-16 (16O); oxygen-17 (17O)
and oxygen-18 (18O). Of these oxygen-16 is by far the most com-
mon and is present at roughly 495 times the abundance of oxy-
gen-18. Oxygen-17 is even rarer (only 0.037 % of the total oxygen
atoms on the Earth) and not generally used in any analysis.
Oxygen-16 has two fewer neutrons in its nucleus than oxygen-18,
which makes it lighter and easier to lift against the pull of gravity.
Therefore, water that contains oxygen-16 is lighter and easier to
evaporate than water which contains oxygen-18. This means that
on average water that contains oxygen-16 will evaporate under
cooler conditions than water containing oxygen-18.
When the weather is warmer more water containing oxygen-
18 evaporates and when it is cooler less oxygen-18 evaporates.
Moreover, when there is heavy rain the amount of light oxygen
(oxygen-16) is greater in surface water, including seawater, simply
because it evaporates easier and thus more of it ends up in clouds
which then produce rain.
At other times, when evaporation is high, but precipitation
low, the amount of heavier oxygen-18 rises in surface water. The
abundance of the different oxygen isotopes that remains in the
seawater will then alter depending on the temperature and on pre-
cipitation; and because the isotopes vary in the sea water, they will
also vary in the shells of organisms that are living in it and busy
assembling their shells. By analyzing the concentration of these
different oxygen isotopes in sea shells climatologists can probe the
temperature of the water in which the organisms grew and the
amount of rainfall that was occurring. So, where do you look for
these shells? The answer lies in layers of hardened mud that are
laid down in the oceans every year. In the right, and fortunate,
locations there are bands of sediment in the rock that clearly chart
the summer monsoon, either as bands of sea shells with alternat-
ing heavy (18O-rich) or light (16O-rich) rock.
By observing cyclical patterns in the deposition of soil and
isotopes of oxygen, evidence has accumulated that the modern
monsoon was established over 50 million years ago. At this time,
the Tethys Ocean, although shriveled markedly in stature, still

extended along to where the Tarim Mountains now lie to the north
of modern Afghanistan. The Himalayas were rising, but undoubt-
edly had lower elevation than they do now. Alexis Licht and
colleagues examined the ratio of different isotopes of oxygen in
fossilized invertebrate shells (gastropods) that lived in the Tethys
Ocean. Through the examination of oxygen-isotopes preserved
in their fossilized teeth and shells climatologists revealed a clear
monsoon-like pattern of drought and deluge that is characteristic
of modern Asia. Paralleling this were changes to the pattern of
dust movement across China.
Over China strong, cold and dry winter winds currently
deliver dust towards the south. Again, these winds only blow in
the winter, thus a seasonal pattern will be evident in sedimentary
rocks laid down at the time. Matching the pattern of isotopes seen
in the gastropods, the Chinese records clearly show a monsoon
pattern, with the annual pattern of deposition running in cycles
driven by the onset and termination of each winter monsoon.
Thus, Litch and co-workers showed that during much of the
Eocene, from 50 to 34 million years ago, a monsoon pattern of rain-
fall was apparent, despite a lower Himalaya and a limited Tibet.
Although the Tarim basin was flooded by the Tethys, the relatively
strong Eocene monsoon appears to have been driven by the higher
carbon dioxide levels of the day, which trapped more of the Sun's
radiation and propelled a stronger circulation. The lower eleva-
tion of the land, and the presence of the Tethys to the northwest,
would naturally decrease the strength of the summer monsoon
as these weaken the contrast in temperature and pressure needed
to drive the summer monsoon. However, with a higher concen-
tration of carbon dioxide in the atmosphereroughly twice what
is it nowthere was sufficient heating of the rising land to the
north of India to drive a monsoon circulation during the summer
months. The stronger greenhouse effect more than compensated
for the deficiency in the underlying geography.
However, things were about to change. Although Tibet was on
the rise, empowering the summer monsoon with extra lift, the rise
of Tibet had a less positive effect. As Tibet and the Himalaya rose
ever higher, they naturally intercepted more wind and, with the
incumbent rainfall, began to erode ever more fiercely. Enhanced
precipitation and erosion drags carbon dioxide out of the air. With
Antarctica also sliding into its modern position over the pole and
declining levels of carbon dioxide, the planetary greenhouse began
to fail. As the Eocene transitioned into the Oligocene, 34 mil-
lion years ago, the Asian Monsoon began to lose its way. Work
by Guillaume Dupont-Nivet and colleagues (who also worked
with Litch on the work described above) charted the decline in the
Monsoon. Their research indicates that the drying out, or aridifi-
cation, of Tibet happened as Antarctica assumed its modern posi-
tion and began to freeze over.
This was a bad time for the monsoon. For to the north of
Tibet, the Tethys had shriveled away (Fig. 2.1) and the supply of
moisture to central Asia declined. Drought intensified over Tibet
and central China, while declining carbon dioxide, coupled to
FIG. 2.1 An Oligocene (3024 Mya) Paleomap indicating the position of

the Inter-Tropical Convergence Zone (orange) and the likely direction
of the prevailing winds during the northern summer. Moist winds blow
directly across most of North Africa from the tropical Atlantic, keep-
ing the current Sahara a verdant shade of green. Over Asia, moist winds
blow from the Indian Ocean and from the Tarim Sea to the northwest of
India. By the mid-Oligocene the Tarim basin has closed and Tibet is ris-
ing, while during the Miocene 20 million years later, the closure of the
remnant Tethys ends North Africas summer monsoon. Map by author
Antarctica's freeze-up ensured that the rise of air over Tibet and
the Himalayas was weakened (Fig. 2.1), making the period between
34 and 20 million years ago India and Indochina's driest of the last
50 million years. In the end the ever-more lofty Tibet came to
the monsoon's rescue. Up until this point the Himalayan range
was providing much of the lift to the summer monsoon airflow.
However, as Tibet assumed its modern height and dimensions, it
boosted the overall lifting effect to the warm tropical air. Thus
from 20 million years onwards, despite significant changes to the
moisture supplyand declining global temperaturesthe mon-
soon was reinvigorated and began its current pattern.
Milankovi Cycles: Astronomical Influences

on Terrestrial Climate
In 1930 Milutin Milankovi published work that suggested that
the ice ages were caused by cyclical changes in a number of astro-
nomical factors: these were the shape of the Earths orbit; its wob-
bling, or precession, on its axis; and the degree of overall tilt of
the axis. The degree to which these different cycles coincided,
summated or detracted from one another determined whether
the world would warm or cool. Milankovi had been working
on these for the best part of a decade, motivated in part by close
partnerships with climatologist Wladimir Kppen and geophysi-
cist Alfred Wegener. Milankovis work built upon much earlier
observations and theoretical work from Johannes Kepler andand
from even earlierGreek philosopher Hipparchus. Kepler had
derived three successful descriptions and mathematical formula-
tions of the orbits of the planets around the Sun. These indicated
that the elliptical orbit of the Earth altered over long periods of
time (100,000 years as part of a 400,000 year-long super-cycle).
In 127 B.C. Hipparchus had noted, from even earlier work, that
the Earths axis underwent precession over a period of around
23,000 years. This was evident by a gradual change in the position
of the stars Regulus and Spica relative to the autumn equinox.
Milankovi triumph was to take these disparate observations and
synthesize an effective, holistic model of the astronomical factors
that govern the long-term evolution of the Earths climate.
To get more of a flavor for these cycles you need to think

of the Earth as a spinning top. Despite the Earths massive size,
it can still wobble about. In part this is caused by the pull of the
Moon, Sun and Jupiter in particular, but it is also a consequence of
having a surface that moves around through plate tectonics. The
Earths orbit around the Sun is an ellipse so that in January the
Earth is nearly 4 million kilometers closer to the Sun than it is in
July. The shape of this ellipse varies over a period of 100,000 years,
primarily under the influence of Jupiter. At present the Earth
receives around 6 % more energy from the Sun in January than it
does in July. This 6 % inflation ensures that the northern winters
are somewhat warmer than they would otherwise be. However, as
this long cycle progresses, the northern winter will coincide with
an even greater eccentricity that bequeaths the northern winter
with up to 20 % more sunlight than it would get if the orbit was
circular. Therefore, at the moment the northern winter is set up
for a cooling trend, not the warming one we currently observe.
Superimposed on this cycle are two other ones involving
the Earths tilt. The overall tilt of the axis varies, from 21.5 to
24.5, over the course of 41,000 years. When the Earths tilt is less
extreme, the planet experiences cooler summers but warmer win-
ters than when the tilt is greater. This set up favors more snow-
fall in the winteras the warmer air holds more moisturewhich
then thaws less in the cooler summer. Imagine a long autumn,
rather than a warm summer and cold winter as we have now.
Finally, the axis precesses, or wobbles, so that currently when the
Earth is furthest from the Sun the northern polar region is point-
ing most towards the Sun. Meanwhile in our northern summer
the North Pole is pointing at the Sun when it is furthest from it.
That process will reverse in 10,500 years so that during the north-
ern winter the Earth will be furthest from the Sun and thus be
considerably colder for longer.
These cycles only matter to the Earth now, because most of
the landmasses lie in the northern hemisphere. As land cools and
warms much faster than water, the appropriate combination of
trends will trigger cooling and glaciation across the northern land-
masses: a slight axial tilt; a less extreme eccentricity in the Earths
orbit; and precession that leads to the North Pole pointing away
from the Sun when the Earth is furthest from the Sun. While today
we think of Milankovi cycles in the context of cooling and ice

ages, during earlier epochs there was a variety of impacts ranging
from warming to cooling. In Chap. 3 we will see how Milankovi
cycles might explain three periods of extreme warmth of the
Eocene, but here we will look to cooling and the formation of the
Sahara desert.
The Death of the Tethys and the Birth

of the Sahara
Between 20 and 7 million years ago the world assumed its present
configuration of continents. Asia was already intact, with India
welded firmly to its southern flank. However, to the west, the
remnants of the Tethys and the modern Mediterranean Sea were
still a battle ground of micro-continents. Africa was drifting lazily
north as fragments of it shuffled around along the southern edge
of what would become Europe. The Black and Caspian Seas were
still part of a larger body of water called the Tethyan scraps while
the Arabian Peninsula still had some water to close to its north.
Although narrow, the Tethys was still able to supply a reasonable
amount of moisture to the north of Africa.
Around 11 million years ago a pivotal event happened and
the Sahara began to take shape. Zhongshi Zhang and colleagues
showed that shortly after 11 million years ago, the Tethys had
shrunk to such a point that the flow of warm moist air associ-
ated with the African summer monsoon effectively failed across
most of the northern half of the continent. The process took a few
million years to unfold but by 7 million years ago, most of the
present Sahara had dried out and was filling up with wind-blown
sand and other deposits.
As the Tethys shrank, the location of the summer doldrums
the Inter-tropical Convergence Zone (the I.T.C.Z.) began to retreat
back towards the West African coast, to where it now resides dur-
ing the summer months (Fig. 2.1). The ITCZ marks the zone of
greatest convection on the planet and consequently the zone of
heaviest tropical rainfall (Chap. 1). Before this time, the North
African region was periodically afflicted by drought, but this pro-
cess was cyclical, driven by the wobbling of the axis and changes
to the orbit of the Earth around the Sun (Milankovi Cycles).

However, over most years the summer monsoon brought rainfall
as far north as the southern Mediterranean coast. From 7 million
years ago, this pattern changed so that drought was the norm and
a verdant green the exception.
Prior to 7 million years ago there was no permanent drought.
However, once the Tethys had narrowed to dimensions similar to
those today, the flow of winds from the tropical Atlantic weak-
ened and eventually assumed their present northern limit in what
is now Chad: before this time the line separating the two Hadley
cells extended into what is now northern Libya. What changed?
The key moment appears to be the closure of the eastern end of
the remnant Tethysthe Arabian Sea. When this shut, the drier
air to the north of the modern Sahara was no longer able to retain
enough heat and the African monsoon began to waver. This was
caused by a reduction in the amount of moisture delivered from
the shrinking Tethys. With its critical support gone, Milankovi
cycles began to dominate the climate and deserts began to wax and
wane across the region. These began nearest to the Mediterranean
coastline, but steadily advanced towards the south over the ensu-
ing 5 million years. By 3 million years ago the modern Sahara was
largely in place.
Within the last few thousand years the Sahara has periodi-
cally turned green when Milankovi Cycles maximized Saharan
warmth. However, this is only when the Earth has its greatest
tilt (24.5) and its northern hemisphere is tilted towards the Sun
when it is closest to it. During this period of maximum summer
heatinglast experienced 60008000 years agothe normally dry
Sahara has blossomed, albeit briefly.
An intriguing additional factor in the Saharas development
must have been the closure of the Mediterranean Sea. As Africa
has continued to move northwards, the seaway at Gibraltar has
periodically been slammed shut, only to be breached by the waters
of the Atlantic. When the Mediterranean-Atlantic seaway first
closed around 6 million years ago it meant that the Mediterranean
soon dried out. Although the Mediterranean contributes a lot less
moisture than the ancestral Tethys did, its temporary loss (which
may have happened several times around 57 million years ago),
would further have limited the supply of both heat and moisture
to the north of Africa. Sooner or later, the progressive northward

march of Africa will close the gateway once more and lead to a
renewed loss of the Mediterranean Sea.
This evaporation of the sea will be a prelude to the final
act. As Africa continues its slow but steady march north, the
Mediterranean will finally close, along with the final remnants
of the Tethys: the Black and Caspian Seas. Their waters will drain
as the land rises and replaces their pristine blue with snow and
ice. Although these seas are small fry when it comes to supplying
moisture to Africa and Asia, their loss will exacerbate drought in
both continents over the next several tens of millions of years.
The Mediterranean mountains will block the flow of winter west-
erly winds from the Atlantic into the Middle East, further reduc-
ing the regions rainfall.
Pangean Monsoons
The slow dance of the continents has ensured that some parts
of the globe have had rain in abundance and drought in others.
During the late Permian and continuing into the Triassic Pangaea
undoubtedly experienced some form monsoon climate. The north-
ern flank of the Tethys was always in turmoil. Every few tens
of millions of years micro-continents were breaking away from
what is now northern India and Saudi Arabia and colliding with
what would eventually become southern Asia. Further west, the
suture between Africa and North America was also mountainous
(Fig. 2.2). This sort of set up would favor seasonal (summertime)
ascent of air and the inflow of warm, moist air from the Tethys.
Such monsoons were likely weak by todays standards, but they
would still be a factor.
Indeed, the situation could have been rather complex. For to
the south of the Tethys Ocean was Gondwanaland which undoubt-
edly experienced monsoon-like inflow of its own as it heated up
and cooled down with the progressing seasons. Like the Sahara
of today, it would seem likely that the ever-present underlying
Milankovi cycles would have driven periods of relative drier and
wetter climatic conditions.
HIGH LOW HIGH

LOW HIGH
HIGH HIGH
Low
HIGH
FIG. 2.2 Possible Late Permian northern summer monsoon pattern over
Pangea. The central Pangean mountains (the ancestors of the modern
Appalachians) act like the Himalayas and Tibet of today and focus warm
prevailing winds blowing in from the warm Tethys ocean. During the
summer, a strong area of low pressure develops and focuses rainfall on
the windward, eastern slopes. To the west, a combination of higher pres-
sure and prevailing winds blowing over the dry interior ensure that the
future North American region is dry. Winds are shown by thin arrows of
different colors, while a likely cold ocean current flowing up the west
side of what is now South America maintains dry conditions to the south
west of the Pangean Mountain range. Paleomap courtesy of Christopher
Scotese
The location of the Pangean Mountains, only marginally north

of the Equator would have meant that any Monsoon regime would
have been rather limited in extent. By focusing the formation of
seasonal low pressure areas, mountains are extremely effective at
organizing the flow of air across continents. On a different scale the
Rockies organize both a summer Monsoon across the south west-
ern United States but also organize the westerlies into a series of
waves that ensure that the central continental U.S. is very cold in
the winter (Chap. 1). On the Earth of today, Tibet and the Himalayas
direct a strong north-south flow of air, but one that spans more
than 50 of latitude (30N to around 20S over the southern Indian
Ocean). The Pangean Mountains, by virtue of their location, would

have directed an air flow over, perhaps 1020 of latitude, that
extended from around 510N of the equator to the northern half
of Gondwanaland in the south.
The arrangement of the continents is critical in directing
the movement of the air, but also the underlying ocean currents.
The natural inclination for the air and water is to move broadly
east-west or west-east under the direction of the Coriolis Effects
and this is true of the oceans as well. Continents get in the way
directing movement in more meridional (north-south) directions
(as opposed to zonal or east-west directions), which in turn directs
warmer or colder air in those directions. In general such masses of
land encourage the warming of the polar regions by giving them
access to warmer air from the tropics. Indeed, in todays atmo-
sphere the warming planet displays more rigid and extreme buck-
les in the jet stream that are directing greater warming towards
the Poles.
Superimposed upon this general trend are bumps and troughs
in the pattern of temperature change. Most significant are the
El Nio and La Nia patterns, which bring year-long alterations
in the flow of air across the Pacific and beyond (Chap. 3). In any
paleoclimate such patterns would also have been likely where the
arrangement of the continents permitted it. We look, now, at these
global phenomena and put them in the context of the broader pat-
tern of terrestrial climate.
Problematic Children: El Nio and La Nia

Were the Earth a bland ball of water things would be so much
simpler. Winds would blow according to the competing forces
of pressure and the Coriolis Effect. On regional scales air would
blow in a largely easterly direction towards the equator or in a
westerly direction towards the poles (Chap. 1). Differences in
pressure would ensure some north south movement, but primar-
ily winds would be orientated 90 to this direction.
Of course, the Earth is unfortunately dotted with continents.
Not only do these contain mountains that block and divert the
movement of air, they also block the corresponding movement of

water. This is no more apparent than in the Pacific, where the pre-
vailing easterly winds (the north east and south east trade winds)
that blow towards the equator drive a large mass of warm water
towards Indonesia and the Philippines. Unable to advance easily
through the myriad of islands that separate the Pacific and Indian
Oceans, the water backs up into an enormous pool of particularly
warm water. As long as the prevailing winds are easterly, the water
remains pooled in the Western Pacific (Fig. 2.3). To the east South
America intercepts the westerly winds and flow of cold water
around Antarctica. Much of this water manages to flow around the
southern tip of South America, but a large portion is forced to flow
northwards up the South American coast towards the equator.
Near the Galapagos the prevailing easterly winds divert this flow
of cold water west, towards the Central Pacific, where it eventu-
ally sinks below the surface. In a normal year the Western Pacific
is therefore warm, and the Eastern Pacific cold.
This prevailing pattern, with cold to the west of the basin
and warm to the east is part of a much larger pattern known as
the Walker Circulation. Thus, embedded within the prevailing
easterly flow towards the doldrums, is a west to east flow in the
Pacific Ocean. In physical terms the Walker Circulation is a bit
like a very large standing wave: a fixed wave-like pattern that per-
meates the entire flow of air across the tropical regions.
In the Atlantic, the same pattern exists, but on a smaller
scale, with cold currents flowing up the west coast of south-
ern Africa towards the equator and a cold current moving south
westwards past the Canary Islands to the north west of Africa.
These cold currents descend under the warm equatorial waters off
the west coast of central Africa leaving a warm pool of water west
towards the Caribbean. In the Atlantic, the draw of cold water is
enhanced in the northern summer when the African monsoon winds
(Chap. 1) move inland towards the Sahara from the southern hemi-
sphere. The Indian Ocean is forced into the opposite pattern with
a warm east and a cold west. The forcing factor is the warm pool
of warm water over Indonesia and the strong low pressure area
that is associated with it. The Walker circulation in the Indian
Ocean is strongly affected by the Indian monsoon which radically
B
Deep cold water
C Deep cold water
Deep cold water
FIG. 2.3 ENSO. (a) Normal (non-ENSO) circulation. Winds blow along
the equator towards the westthis is the Walker Circulation. Over the
Pacific, winds cause warm water to back up against Indonesia giving rise
to the greatest convection and heaviest rainfall. To the west of South
America cool, dry air descends and blows west. Winds cause cold water
to upwell which further enhances the sinking of air at the surface. (b) In
an El Nio year the Walker Circulation weakens while the strength of
the Hadley Cells supplying it strengthens. Warm water sloshes back east-
wards towards the west coast of South America, bringing the maximum
convection and rainfall with it. The cold upwelling along the coast stops.
On occasion the sloshing stops mid-Pacific leaving both Indonesia and
western South America relatively dry. (c) In a La Nia year the Walker
Circulation strengthens and there is greater than normal upwelling of
cold water along the west coast of South America. More warm water
than normal backs up north of Australia enhancing rainfall along the
east coast of Australia and over Indonesia. The pattern of circulation over
the equatorial Pacific, in turn, alters the tropical circulation across the
Atlantic and Indian Oceans with knock-on effects to the climate there
reverses wind flow across the entire ocean basin between winter
and spring. This makes the overall pattern of circulation very dis-
jointed with strong seasonal variation across the Indian Ocean
that is not strongly reflected in the neighboring ocean basins.
Every now and then, the prevailing easterly winds weaken

across the Pacific and the large pool of warm water near Indonesia
sloshes back east, over the top of the cold waters moving north
from Antarctica. Now, the waters lying in the Eastern Pacific
become very warm, while the water in the normally warm west
becomes cooler (though hardly cold). When the waters in the
Central Pacific become warm a weak El Nio is declared (Fig. 2.4).
The more warm water that sloshes to the east, the stronger the
El Nio is. NOAA will declare an El Nio only when sea surface
temperatures are at least 0.5 C above normal for 3 or more months
to the east of the 120W longitude. Strong El Nio conditions are
coined for El Nios that last 79 months. Anything longer and
the term El Nio episode is used. These reversals in the tempera-
ture of the ocean surface are reflected in changes in the air above.
On occasion the ocean sloshes warm water to the east but, as
a Normal Global Walker Pattern (Northern Winter)
LOW
HIGH LOW HIGH HIGH LOW
PACIFIC SAM ATLANTIC AFR INDIAN IND
b El Nio Global Walker Pattern (Northern Winter
HIGH
LOW HIGH LOW LOW HIGH
PACIFIC SAM ATLANTIC AFR INDIAN IND
FIG. 2.4 The effect of an El Nio on the pattern of airflow in the, tropics.
A comparison of normal and El Nio weather patterns. The Walker cir-
culation is the patterns of airflow along the equator and tends to produce
high pressure over the eastern pacific and low pressure over Indonesia (a).
This pattern broadly reverses during an El Nio event (b). Although the
most severe impacts are in the Pacific basin, the El Nio causes effects
further afield. The Amazon basin dries out with pressure generally rising.
This is also true of Indonesia, the Sahel and equatorial Africa. The Indian
Ocean has its own equivalent pattern to the ENSO (El Nio Southern
Oscillation) and may or may not run concurrently with the Pacific pat-
tern. When it does, the Indian sub-continent tends to get weaker rains in
the summer
Warmer
Wetter
Warmer water
Equator
than normal
Warmer: High
less sea ice
FIG. 2.5 Changes to the jet streams during El Nio events during the
northern winter. The largest change, outside the tropics, affects the
northern hemispheres polar front jet stream (maroon). This strengthens
along its southerly branch while the northern branch weakens. Frequent
low pressure areas move along this into Western United States bringing
large amounts of rainfall, while the south eastern states dry out. Across
the eastern seaboard milder than normal conditions prevail. Meanwhile,
the sub-tropical jet (orange) strengthens over Australia bringing drier
conditions. Further south, near Antarctica, high pressure areas become
more frequent west of the peninsula bringing warmer conditions. East of
the Peninsula, cooler conditions prevail
in 2014, the atmosphere above does not fully respond. In these

instances some of the weather phenomena associated with El
Nio occurs but not all.
In a full blown El Nio the easterly trade winds weaken and
die away and storm systems migrate towards South America.
The Peruvian coast gets a soaking, while northern Australia and
Indonesia suffer drought. The El Nio usually strikes in the late
autumn through to the early spring, causing a significant realign-
ment of the jet streams over the Pacific; however, in some years such
as 1997 or 2015 it begins in the early summer. During the winter
Colder
High Wetter
Drier
Colder than
normal water
Low
Colder: more
ice
FIG. 2.6 Changes to the jet streams during La Nia events during the
northern winter. Over the Northern Pacific the polar front jet stream
(maroon) breaks into two branches with a prominent blocking anticyclone
present south of the Aleutians. During La Nia episodes the northern
branch has more energy than the southern branch. The southern branch
wobbles more and brings heavier rainfall to the Pacific Northwest, while
California and Mexico experiences drought. This set-up can become pro-
longed, establishing mega droughts in the South Western States. Over
the southern hemisphere, the Polar Front jet stream bulges northwards
affecting more of South America while a cooling trend affects Antarctica
west of the peninsula. East of the Peninsula the situation is reversed with
warmer conditions
months the northern sub-tropical jetwhich normally lies across

southern California is diverted southwards, while to the north, the
polar front jet, with its attendant winter storms, also moves south
and pummels the western states (Figs. 2.5 and 2.6). California
often experiences flooding as a succession of low pressure storms
swing south eastwards from the central north Pacific. The Indian
monsoon is usually weaker during El Nio years and much of the
Amazon experiences lower than normal rainfall. These changes are
associated with changes to the Walker Circulation (Fig. 2.4).
Further afield, these oscillations have a more subtle effect

on climate, with milder winters in Western Europe more typ-
ically associated with La Nias. Figure 2.7 shows how the El
Nio can establish broader changes to the circulation of the
atmosphere over the northern hemisphere through the forma-
tion of far reaching Rossby wave. Even within that pattern is an
interesting sub-pattern. La Nias often bring blocking high pres-
sure areas over Western Europe from late November through
December. In many instances these bring relatively cold but dry
conditions to the U.K.
FIG. 2.7 Rossby Waves associated with the El Nio. Low pressure (red)
over the central Pacific (paired on either side of the equator) is linked
through a series of south-west to north-east trending high and low pres-
sure areas that extend across the North Pole. These trans-latitudinal
waves interact with and organize the longitudinal Rossby Waves that
are moving around the North Pole within the polar front jet stream (Fig.
1.14). Similar patterns extend to the south east of the equator over the
southern hemisphere. Map courtesy of Paul Anderson (http://www.csiss.
org/map-projections/index.html)
When the Arctic Oscillationan episodic pattern afflicting

the circulation of air around the Arctic basinis in its negative,
high pressure dominated phase, a La Nia can work with it to
produce extreme cold. The La Nia of 2010, for example, brought
the coldest December on record to the UK. Blocking developed to
the north of the UK in late November which directed a succession
of winter storms across the UK with abundant snowfall. At the
close of December, normality resumed and the La Nia helped
maintain a very mild pattern, with February ultimately becoming
one of the mildest on record. What appears to matter is how many
waves separate the equatorial Pacific from the Atlantic. This can
vary from three to seven and ultimately it may well come down to
which wave number is present when the El Nio or La Nia kicks
in. It would seem the latter tend to lock the system down in what-
ever mode it is in. So for example, during the 2015 El Nio there
was already a 3 year long pattern of blocking over the northern
Pacific and Alaskasomething more normally associated with La
Nia conditions (Fig. 2.6).
Although December 2010 was extreme, blocking patterns
are common in December in La Nia years. The only difference
is the position of the block, which more often than not lies over
the UK or just to its east. December 2000 also saw a very cold
pattern in the UK following a weak La Nia, and again this was
associated with a negative Arctic Oscillation. Conversely, the
winter of 2007 had a La Nia blocking pattern over the UK but
this time the Arctic was positive and December, although fairly
cold, was far from remarkable, with a very mild winter ensuing.
Thus, in Western Europe La Nias have effects, but they are
embedded in much broader patterns, dictated by what is going
on over the Arcticand that is only partly linked to patterns
further south.
Most significantly, El Nios stir up the Pacific waters, releas-
ing heat from deeper layers as well as redistributing heat that are
otherwise locked to the west. For this reason El Nio years tend
to be the warmest ones in any one decade. The grand-daddy of the
lot was the 19971998 El Nio which set records that were only
matched if marginally broken in 2014, but shattered in 2015.
What is clear is that the ENSO seems to run to its own beat
and has (thus far) shown little if any influence from the rising
temperatures caused by anthropogenic global warming. However,
within this end of the story is a growing uncertainty regarding
the influence of rising CO2 and (over decadal timescales) rising
temperatures on the strength of the Hadley circulation. Models
predict and observations suggest that as temperatures rise the
Hadley cells expand towards the poles. Although the movement is
relatively slow on human timescales it will influence the patterns
of rainfall across the globe. Expectations would include enhanced
rainfall across the southern Sahara and northern India where
abundance wouldnt cause concern. However, an expansion of
the Hadley Cells would reduce rainfall to the north of the Sahara,
in the Mediterranean; reduced rainfall across the mid-West and
also throughout parts of southern Australia where conditions are
already marginal. Superimpose the effects of El Nio and La Nia
and some parts of the globe could really suffer in terms of food
production or the availability of fresh water.
Whether these expectations are met, is something we will
have to wait and see, as we continue our experiment with the
Earths climate.
The triggers for El Nio events are far from understood. There
are clear connections between Madden-Julian Oscillation (Chap.
1) and the El Nio pattern. Both are westward propagating sys-
tems, with the MJO driving westerly winds in its wake, which in
turn drive a pool of warm water eastwards to a depth of 100 m.
The intensity of MJOs is enhanced 612 months ahead of El Nio
events, but once the El Nio is established MJO events tend to
melt away, becoming almost non-existent. The 1982 El Nio, one
of the strongest in decades, was driven by a strong MJO event that
began in May 1982. By July of that year, it had initiated the El
Nio which persisted into the following spring.
While MJO activity is lower during El Nio events, it is height-
ened during La Nia events and observations suggest that prior
to El Nios the westerly winds running in the wake of MJOs are
stronger, while the easterly winds running in their path are stron-
ger ahead of La Nias. Together, this would tend to suggest MJO
activity plays a significant role in determining the broader pattern
of weather in the tropics by influencing the movement of surface

and upper level winds around the storm complexes. However, the
precise nature of the connection remains to be determined.
The Future East Asian Monsoon

Under the Cloud of Global Dimming
The far-reaching effect of seemingly innocuous presence of air-
plane contrails was revealed when terrorists attacked the United
States on September 11th 2001. Then, in the attacks aftermath,
planes were grounded and temperatures across the entire 48 States
changed by more than a full degree Celsius. This is an astonish-
ing observation. If you were to add together the entire area corre-
sponding to a days worth of flights, it would be a tiny fraction of
the total area of the United States. Yet these 50 m long aluminum
tubes generate trails of water vapor that can be kilometers long.
As there are several thousand flights a day the high icy clouds that
the combined flights generate is substantial. The authors, David J
Travis (University of Wisconsin-Whitewater) and colleagues con-
cluded that the 1.8 C change in the regions diurnal temperature
profile was the result of less cooling at night and more warming
in the daytime.
Therefore, the effect of contrails depends on what time of
day they appear. During the daytime, contrails primarily reflect
incoming visible radiation and have a cooling effect. At night
time, the reverse is true with the contrails serving to trap out-
going infrared radiation giving rise to a warming effect. The net
effect (the balance between the day and nighttime influences) is to
cause warming as the trapping of infrared radiation is greater than
the cooling, reflective effect. This explains the observations made
after the 9/11 attacks. There was a greater diurnal range in tem-
peratures when there were no contrails: less cooling in the day and
less warming at night giving rise to greater temperature maxima
and lower temperature minima. This observation has a bearing on
what will happen to the Earth in a billion or so years. If warming
produces more cirrus clouds then the Earth might overheat faster
than if it produces more reflective cumulus clouds. This is looked
at again in Chap. 3; and again in Chap. 10, when we look at extra-

solar planets.
Beyond, the confirmed influence of airplane contrails, there
are other chemical pollutants that have had a varying influence on
global climate over the last hundred years. During the first half of
the Twentieth Century industries in Europe and the US belched
copious amounts of black carbon soot into the air. We loved our
fossil fuels, but not enough to burn them completely. Currently,
Russia, China and the progressive incineration of tropical rainfor-
ests are the biggest sources of such particulate pollutants (Fig. 2.8).
Aside from the negative effects of soot on our health, such par-
ticulates absorb incoming radiation and cause very diverse effects
on temperature that depend critically on what sort of material is
present and at what height in the atmosphere it is found. Even one
type of pollutant, black carbon which is a mixture of unpleasant
carbon-rich compounds, has diverse effects. Black carbon absorbs
radiation and causes warming within the layers in which it is most
concentrated. This is simply because it absorbs infrared radiation
and visible radiation efficientlyafter all, thats why its black.
Beneath that layer, radiation that would have been received on the
ground is blocked, resulting in it being cooler than if the soot was
absent.
Particulates, as these particles are known, may also stimulate
cloud development, which can further reduce temperatures on the
ground. Black soot resulted in extensive smog across many indus-
trial areas in the West in the 1950s and now in the East. The same
chemicals we in the West pumped into the air six decades ago can
be seen poisoning cities in India and China today. Sadly, this is
a lesson in public health that has not been learnt. The drive for
wealth apparently supersedes the necessity of preserving human
health. In China, alone, there are 1.3 million deaths per year asso-
ciated with pollution from burning fossil fuels.
Black soot has considerable effects on rainfall. Through their
warming effect on the middle and upper atmosphere particulates
will raise the air pressure in this layer relative to less polluted
surrounding air. By raising the temperature in the middle lay-
ers of the atmosphere, the air column in which the soot parti-
cles are suspended in becomes also more stable and less prone
to convection. Areas of warm air at middle and high levels in
FIG. 2.8 Computer models showing the heating and cooling effects of
black carbon soot. The upper panel shows the heating effects of these
compounds in the middle and lower troposphere, caused by the absorp-
tion of energy from the Sun. The lower panel shows the cooling effect at
the surface caused by the same materials. These effects alter wind flow
across many regions of the globe as well as directly impacting human
health. Not shown is the effect of soot on snow cover, which has an addi-
tional heating effect. Figures adapted from: http://earthobservatory.nasa.
gov/Features/Aerosols/page3.php
the atmosphere can also have a blocking effect on the jet stream
which can pin it into particular patterns which, in turn, can bring
an over-abundance of precipitation in some areas but a deficit
in others.
The clearest impacts of black soot pollution are seen in

areas of snow and ice, and in Asia, where complex changes to the
regional weather and climate have unfolded in recent years. While
in Asia the greatest impacts appear to be on the Asian Monsoon,
further north and south black carbon soot has very negative effects
on the temperatures of the Polar Regions. Measurements show
that by darkening the surfaces, black carbon soot has increased the
amount of snow melt in the Arctic (and presumably the Antarctic),
as well as on mountain glaciers. Increases in melt have reduced
the albedo of the Earth, which has led to a net warming of the
planet as a whole.
Significant changes have been observed the strength of the
East Asian Monsoon which parallels the growth of cities and indus-
try. Modeling work around the middle of the first decade of the
millennium, in principle, suggested that the East Asian Monsoon
should strengthen because soot blowing onto the snow cover on
Tibet causes it to melt faster. This in turn should make Tibet
warm up faster in the spring and summer. In turn, this should lead
to an earlier and stronger monsoon over India and southern China.
However, observations actually show a decrease in the strength of
the Asian Monsoon since the 1970s. Unlike in previous geological
periods, such as the Eocene warmer conditions are not strengthen-
ing the monsoon.
The weakening monsoon is particularly evident over north-
ern China where rains often fail altogether or arrive much later
than normal. This failure, or delay, is a significant contributor to
the decrease in air quality over Chinas northern and eastern cities,
primarily because the monsoon airflow pattern does not become
established until later in the summer. This leaves a stagnant pond
of heavily polluted air over the north of the country. When the
summer pattern does emerge, the south westerly winds blow the
pollutants out into the Pacific.
Interestingly, where modeling did show a failure of the mon-
soon, it was over eastern China. These models indicated once again
that a warmer Tibet should bias any trends and lead to a stronger
monsoon. In this instance the increasing warming power of Tibet
sucks air more evenly in from the south and south east so that less
moisture blew northwards over mainland China. This left parts of
the Yangtze basin dry. The researchers concluded that more soot
should give a warmer plateau and stronger summer monsoon, yet,
that this was likely offset by the effect of a Pacific ocean warmed
by the enhanced concentration of greenhouse gases. Thus even
though Tibet was warming, the Eastern Pacific was warming
faster, which reduced the temperature gradient between land and
ocean and thus weakened the Asian monsoon.
The waters were muddied further by research carried out by
Veerabhadran Ramanathan (Scripps Institution of Oceanography,
University of California) and colleagues and published in
Proceedings of the National Academy of Sciences. Their research
indicated that even though soot should intensify temperature dif-
ferences between land and sea, the effect was drowned out by the
effect of particles on the region as a whole. These particles caused
more overall cooling at the surface that drowned out the effect of
the soot lying on Tibet. This effect dominated the impact of soot
and thus, overall, also weakened the monsoon.
Thus, the initially simple picture of more soot means more
warming and a stronger Monsoon is clearly insufficient to explain
the changes to the climate of southern and eastern Asia. This is no
trivial observation. The climate of the region critically determines
the supply of food for more than two billion people. Moreover, the
climate also can enhance or mitigate the health effects of the soot
and other noxious gases liberated by the burning of fossil fuels.
Therefore, it is vitally important that we understand the discrep-
ancy between observation and theory. Not simply for the prosaic
reason that we the need to understand our climate better, but also
because the well-being of our species depends upon it. Moreover,
the impact of improved modeling will clearly extend our under-
standing of the climate of planets elsewhere.
Finally, another critical issue is the effect of dimming on the
transpiration of plants. As well as the obvious negative conse-
quences of destroying forests directly, reducing the intensity of
sunlight reduces the rate of transpiration. The fewer trees that
there are transpiring over Indonesia the less moisture is returned
to the atmosphere upwind of India. Indeed, experiments show that
fully 8090 % of the water vapor entering the atmosphere over
land comes from plantsnot from simple evaporation from the
soil. Although the plants give us oxygen story may not quite
hold up to scrutiny (because most oxygen comes from bacteria and
algae in the oceans), the contribution forests make to the plane-
tary hydrological cycle is undeniable. Fewer trees will mean lower
levels of atmospheric moisture and hence less cloud and precipita-

tion downwind. Burn them and the risk rises for extra tornadoes
in vulnerable regions (Chap. 1). Humans chop the forests down at
the peril of climatic balance.
Thus while extra carbon dioxide should cause a stronger
airflow across Asia, the effect of carbon dioxide is swamped out
by the negative effects of soot and other particulates, especially
sulfatesand there may be other effects caused by changes in
the regional hydrological cycle. Sulfates tend to reside in the
atmosphere for longer periods particularly, when they are high in
the atmosphere (stratosphere and upper troposphere) and above
the level at which rain can wash it out. Sulfates cause cooling by
reflecting incoming radiation from the Sun. In the past, sulfates
in the high troposphere and stratosphere, primarily originated
from volcanic activity and from the oceans. However, since the
industrial revolution sulfates also come from combustion of fos-
sil fuels, in particular coal. Around 150 million metric tons of sul-
fur, mostly in the form of sulfate, is mobilized by human activity
each year; and around two thirds of the sulfate in the air is from
man-made activity. Although most of this contamination is low
leveland thus is washed out fairly readily as acid rainsome
does work its way higher and contributes to global dimming. On
the positive front, overall in the West, most of this sulfate is now
removed from the flue gases so that it has less impact now than
it did 30 years ago. Despite this the problem is not solved and
sulfates continue to plague many developing nations where less
investment has been put in place to remove them from the efflu-
ent from coal-fired and oil-fired power stations.
The West cannot claim to be out of danger either.
Notwithstanding nearly three or four decades of cleaner air, the
lower levels of soot and other particulates produced by cars and
other vehicles is still an issue. One only has to look across the
skyline of Los Angeles, New York or cities in the UK to see a hazy
layer of soot and dust which we all breathe in. This material ulti-
mately blows around the globe affecting the formation of clouds
and consequently precipitation. Combustion may be a profitable
business but it is also a very messy and damaging one. The best
way to deal with it is to replace it as the longer-term impacts of
humanitys love of combustion becomes clear.
To put things in a broader historical perspective, an unusual

and damaging event around 1500 years ago can help illustrate cur-
rent risks.
How Global Dimming Caused Geopolitical

Chaos in the Dark Ages
Can global dimming bring down an empire? That was the ques-
tion raised by David Keys at the close of the last millennium. Keys
presented a very interesting and rather persuasive thesis that sug-
gested something rather catastrophic had set about the end of the
Roman worldand the birth of our modern one. Although much
of Keys 1999 book Catastrophe is clearly speculative, it makes for
an utterly fascinating ride. Moreover, Keys is very careful through-
out to underpin his hypothesis with some very real and ultimately
quite disturbing science.
To summarize, in 535 AD there were reports of loud detona-
tions to the south east of China. In 536 AD, thick yellow dust
began to fall from the skies. Although such precipitations are not
uncommon in the winter months (wind-blown loess from west-
ern China) these precipitations were followed by snow in July and
frost in August: hardly something normal in sub-tropical south-
ern China. Meanwhile, in Europe the Sun was reported to be dim,
often taking on a bluish tint. This effect is something typical of
airborne dust, which scatters the blue light. The winters in the
following years were extremely harsh, while the summers were
uncommonly cool.
Ice cores in Greenland and Antarctica reveal the presence of a
spike in the levels of sulfuric acid (hydrogen sulfate) at this time. As
we have seen, this key chemical is produced by the burning of fos-
sil fuels and by volcanic activity. Given that Middle Age humans
were not accustomed to burning fossil fuels, its presence in both
polar locations in such abundance indicates that a large volcanic
eruption occurred near to the equator around 535 or 536 AD. The
location of the eruption was important, for only an equatorial erup-
tion could scatter large quantities of sulfate into the atmosphere of
both hemispheres, and as such only an equatorial eruption could
affect the climate and geopolitics of both hemispheres.
An equatorial location narrowed down the likely culprits.

The eruption had to be large and explosive, which further nar-
rowed the search. Keys suggests that Krakataufamous for its
1883 eruptionwas the likeliest candidate. Working with Keys,
Ken Wohletz (Los Alamos National Laboratory) produced some
likely scenarios for an eruption that would produce the desired
effects. Geographically, what is interesting about Krakatau is that
the island that was obliterated in the massive 1883 eruption itself
lies within a much larger calderaa volcanic structure produced
by the collapse of a much greater volcanic edifice. This arrange-
ment, flanked by enormous eruption deposits suggests that prior
to the 1883 eruption Krakatau had at least one other, even more
powerful eruption that had obliterated the earlier, and much larger,
volcano. The famous 1883 eruption was a mere blip that occurred
after this. Volcanologist, Harald Sigurdsson, took part in the pro-
duction of the TV series1 that accompanied the book and although
he was unable to definitively date the deposits, he did narrow the
eruption to a (broad) window that encompassed 536 AD using
radiocarbon dating.
Bathymetry reveals that the original volcano that gave rise
to this massive outer caldera was at least 50 km across and per-
haps, given the scale of the later volcano, 1000 m or more high.
To produce the observed caldera, the eruption would have had to
release over 200 cubic kilometers of rockor 200 times that of Mt
St Helens in 1980. Although dwarfed by much larger eruptions,
such as the Siberian Traps (Chap. 3), this is still a substantial erup-
tion that would have devastatingly global impacts today.
Wohletzs models show an eruption column that is perhaps
50 km in height, discharging rock and gases at a rate of over 1 billion
kilograms (1 million metric tons) per second. Given the amount of
rock, Wohletz suggested that the eruption was sporadic and took
place over several weeks or monthssimilar to the 1883 eruption.
After an initial violent phase, the volcano would have begun to
collapse, generating a lower, broader eruption column and more
violent explosions. Perhaps it was these that were recorded by the
1
Secrets of the Dead; Channel 4 UK, 1999.
Southern Han Dynasty. Up to 150 cubic kilometers of seawater

would have been vaporized either through contact with hot flows
of volcanic debris or through direct contact with the underlying
magma when the volcano began to collapse. This would drive a col-
umn of very hot steam high into the stratosphere, where it would
combine with sulfurous gases to produce even more sulfuric acid.
An eruption of this magnitude could produce a cloud layer
over 20 m thick in the troposphere, with a finer but thicker spray
of sulfates and ash higher up in the stratosphere. This would have
lasted for months, dimming the sunlight reaching the ground
underneath. Therefore, beneath this translucent layer, global tem-
peratures would have plummeted for several years leading to the
cooling evident in tree ring data.
A more recent and precise analysis of ice core data paints and
even grimmer picture to the one initially envisaged by Keys and
Wohletz. Work by Michael Sigl (Desert Research Institute, Nevada)
as part of a larger collaboration, indicates that there was more than
one large eruption. Around 536 AD, the ice cores show a signature
of possibly three large eruptions that probably went off in North
Americamost likely Alaska or British Columbia. These left a
signature in Greenlands ice cores, but not Antarcticas and low-
ered temperatures by 1.62.5 C. Devastating though these erup-
tions would have been to life in the Northern Hemisphere, they
barely touched the Southern Hemisphere because the Hadley cells
of the north and south barely communicate. However, these erup-
tions were followed 4 years later by an eruption in the tropics,
possibly in Indonesia at the site of Krakatau. All four of these erup-
tions were very largedwarfing Tambora and adding a significant
baggage of dust and sulfate to the atmosphere. This final assault
on the climate lowered temperatures by a further 1.42.7 C. This
cooling then lasted until 550 AD and making the newly arrived
Dark Ages truly dark.
Quite aside from the obvious effects of prolonged cooling
and general drying out of the atmosphere caused by more lim-
ited evaporation, cooling could have a more insidious effect. Keys
pointed the finger of cooling as a causative agent of an outbreak
of the Black Death which decimated the Justinian Roman Empire.
He identified an effect of prolonged drought followed by abundant
rain as a mechanism that would have allowed plague bacteria to
jump from their resistant endemically infected rodent hosts to

other rodents, including Black Rats. A drought would have led
to famine and a drop in the resident population of rodents. The
prolonged period of cooling would have caused drought because
less water was evaporating from the oceans. Several years after
the global catastrophe, when the planet warmed up once more,
the rains would have returned and allowed the rodent population
to boom. This larger population of rodents would then have come
into contact with black plague and then humans. The Justinian
trade routes then became the unwilling vectors that carried their
lethal bacterial passengers around the Roman Empire.
There is another, more recently discovered means by which
climate change could facilitate the spread of disease. Many patho-
genic bacteria, such as plague (Yersinia pestis) or cholera (Vibrio
cholerae) derive their pathogenicity from resident viruses. These
viruses bequeath their bacterial hosts with additional traits that
allow the bacteria to cause untold misery. For example the cholera
bacterium contains circular pieces of DNA called plasmids which
carry the gene encoding the poison that causes the potentially fatal
diarrhea. This plasmid is clearly derived from a type of virus called
a bacteriophage that infects these cells. Strains of the cholera bac-
terium that lack this plasmid are effectively harmless to humans.
Similarly, plague bacteria harbor a set of related plasmids that
encode features allowing the bacteria to invade tissues and survive
in the bloodstream. The original Yersinia bacterium was more like
the largely harmless E. coli bacterium that is a resident of the guts
of every human on the planet. However, with a handful of addi-
tional genes, the Yersinia pestis bacterium became an aggressive
invader of rodent then human tissues, which ultimately leads to
the characteristic infections of Bubonic, Pneumonic or Septicemic
Plague. The interesting thing about these bacteria is that tempera-
ture appears to play a big part in which of these disease-causing
genes function. Some gene appear to work best in the flea with
a body temperature of around 25 C, while others work best at
mammalian blood temperatures of 37 C. Now, although the key
driver of these genes will be the temperature of the host organism,
playing with environmental temperatures will undoubtedly also
impact on the working of the bacteria, as well as the behavior of
the host insect and mammals.
Of greater concern is related research by an international

team led by Martha Clokie (University of Leicester). Their research
focused on the behavior of a bacteriophage that infects patho-
genic bacteria in the tropics. Both the bacterium and accompany-
ing viruses are present in soil throughout the region. When these
viruses infect cells that can either go through what is known as a
lytic cycle and bursts open (lyse) the cells; or they can go through
what is referred to as a lysogenic cycle, where the virus remains
dormant and only replicates with the cell. At 25 C the viruses
tend to replicate with the cell in a lysogenic cycle, while they pref-
erentially use the lytic cycle at 37 C.
You might think that having a virus kill the cells at higher
temperature would be a good thing for us and that global warming
would tend to favor the virus lifecycle where the pathogenic bacte-
ria is killed. However, this is too narrow a view of evolutionand
of climate. For one, global warming might not make the area in
which the virus and bacteria live warmer. Indeed, if the area was
tropical, at altitude, it might become wetter. This would make it
cooler favoring the persistence of the virus within the bacterium.
If that virus made the bacterium more pathogenic then humans
could be in trouble.
Alternatively, at higher temperatures there is a clear evolu-
tionary benefit to the infected bacterial pathogen if it can avoid
death by lysis. This can happen in a number of ways that lead to the
persistence of the virus and work against the virus killing its host
cell. This change can happen through mutations to the virus or
changes to the inner workings of the bacterial cell. If that virus hap-
pens to produce any toxins, or can otherwise affect the immunity of
the human or animal the bacterium infects, that infected pathogen
can become a lot more deadly. Indeed, it is clear that many human
pathogens, cholera included (and perhaps others like the plague)
have become modified by the addition of viruses that then became
adapted to live harmlessly in the bacterium at higher temperatures.
The discovery that certain pathogenic bacteria might play
host to their own pathogens certainly complicates out view of
the natural world and should make us wary of experimenting
with its climate. This is a new and active area of research and one
worth keeping an eye on. At present the precise outcome of each
of the complex evolutionary interplays needs further scrutiny.
No organism exists in isolation and the laboratory investigations

will not reveal the full repertoire of behaviors these organisms
show in the wild. These investigations do reveal a new possible
impact of changing global temperatures. Bear this in mind in
Chap. 3s exploration of the planetary turmoil we and other spe-
cies have coped with over past millennia.
Conclusions
This chapter and the next set weather and climate within a
broader terrestrial framework to examine the broader context of
climate before moving on to discuss the climate of specific plan-
ets. Climate is the greatest natural phenomenon on the Earth
and its impact on us, via natural catastrophes and more routine
patterns, is as large as humanitys own influence on climate has
become today. By considering these issues together a broader can-
vas of planetary climate is created, one that we can apply to the
many new worlds we are exploring and discovering.
References
1. Modelling of the possible 536 AD Krakatau eruption. Retrieved from http://www.ees.lanl.gov/
geodynamics/Wohletz/Krakatau_6th_Century.pdf.
2. Knirel, Y. A., Lindner, B., Vinogradov, E. V., Kocharova, N. A., Senchenkova, S. N.,
Shaikhutdinova, R. Z. et al. (2005). Temperature-dependent variations and intraspecies diversity
of the structure of the lipopolysaccharide of Yersinia pestis. Biochemistry, 44 (5), 17311743.
3. Ramanathan, V., Ramanathan, V., Chung, C., Kim, D., Bettge, T., Buja, L. et al. (2005).
Atmospheric brown clouds: Impacts on South Asian climate and hydrological cycle. PNAS,102,
53265333.
4. McKitrick, R. R., & Michaels, P. J (2007). Quantifying the inuence of anthropogenic surface
processes and in homogeneities on gridded global climate data. Journal of Geophysical
Research, 112, D24S09.
5. Zhang, Z., Ramstein, G., Schuster, M., Li, C., Contoux, C., & Yan, Q. (2014). Aridication of the
Sahara desert caused by Tethys Sea shrinkage during the Late Miocene. Nature 513, 401404.
6. Licht, A., van Cappelle, M., Abels, H. A., Ladant, J. -B., Trabucho-Alexandre, J., France-Lanord,
C., et al. (2014).Asian monsoons in a late Eocene greenhouse world. Nature, 513, 501506.
7. Dupont-Nivet, G., Krijgsman, W., Langereis, C. G., Abels, H. A., Dai, S., & Fang, X. (2007).
Tibetan plateau aridication linked to global cooling at the EoceneOligocene transition.
Nature 445.
8. Shan, J., Korbsrisate, S., Withatanung, P., Adler, N. L., Clokie, M. R. J., & Galyov, E. E. (2014).
Temperature dependent bacteriophages of a tropical bacterial pathogen. Frontiers in
Microbiology, 5, Article 599, 17.
9. Sigl, M., Winstrup, M., McConnell, J. R., Welten, K. C., Plunkett, G., Ludlow, F. et al. (2015).
Timing and climate forcing of volcanic eruptions for the past 2,500 years. Nature 523, 543549.
3. Tales of Mass Destruction
Introduction
This chapter illustrates the fluid nature of planetary weather
and climate. Climate is a dynamic creature that changes with
the activity of the planets central star, the Sun, and with other
dynamic forces that operate on and within planets. The climate of
the Earth has repeatedly changed dramatically over the eons since
it has formed.
Moreover, humans have become particularly adept at alter-
ing their environment. These activities have taken the planet
down a new and at present unpredictable path that will impact
all other life forms. Yet, dramatic though these changes are,
they will be transient in nature with Earth driving the final out-
comes. At the end of course the planet will meet an inevitable
death inside the dying Sunbut there will be lots of excitement
on the way.
Global Greenhouses: Eocene, Permian

and Anthropocene
Utter the words greenhouse effect and suddenly attitudes polar-
ize in the United States and United Kingdom, though not as much
elsewhere. Conversations change from idle chatter to politically
divisive personal opinion. In order to get to the bottom of all this
social and political angst we will look at global greenhouses in
three different contexts: the Permian Mass Extinction; the Eocene
Hyperthermals and most recently anthropogenic warming. As the
underlying mechanism is the same in each case, we should and
can expect similar outcomes, at least in terms of process if not
(hopefully) the degree of change.

DOI 10.1007/978-3-319-25679-5_3
The Eocene Climate Maximum
Fifty five million years ago the worlds climate experienced a rather
puzzling incident. Starting rather abruptly, a period of around
86,000 years ensued when global temperatures rose between 6 and
8 C above their current value. The warming, known as a hyper-
thermal, was truly global in its complexion. Occurring at the end
of the Paleocene and start of the Eocene, the Paleocene-Eocene
Temperature Maximum or PETM was followed two million years
later by a second somewhat smaller temperature surge called
the Eocene Thermal Maximum 2 (ETM2). Finally, 1.6 million
years later at 52.6 million years ago a third hyperthermal (ETM3)
occurred. The origin of these warming periods initially appeared
to be utterly mysterious. Each period is associated with an appar-
ent surge in the levels of carbon dioxide, without a clear reason in
the biological record. So what happened?
Two largely competing, but successful, models have been pro-
posed. Initial hypothesis suggested a geological signature in the
hyperthermal, while the most recent proposal looks further afield
to the cosmos. In 2004 Henrik Svensen and colleagues (University
of Oslo) suggested that the event, which is superimposed on a lon-
ger period of global warming, was linked to the opening of the
North Atlantic. Around 55 million years ago, large continental
rifts were extending northwards, severing Europe from North
America and Greenland. The climax of this process was the erup-
tion of large volumes of basaltic magma from volcanoes in the
area. The remnants of this outburst stretch from Northern Ireland,
through western Scotland and northwards to Greenland and west-
ern Norway. Here, a magnificent series of basaltic plateau and
extinct volcanoes play testament to the arrival of a particularly
large batch of magma that would ultimately give rise to Iceland.
As this magma torched its way through the crust, Svensen
and colleagues proposed that it interacted with large amounts
of buried organic material in the form of sedimentary rocks. As
these were progressively cooked by the intruding magma, around
3003,000 billion metric tons of methane was released into the
atmosphere. Methane is a potent greenhouse gas in its own right
and this surge of methane would readily explain the sudden rise in
global temperatures. The methane would then oxidize to carbon
Tales of Mass Destruction 93
dioxide, which although a weaker greenhouse agent than methane,

would still perpetrate a period of more prolonged warming, lasting
tens of thousands of years. Together with volcanic carbon dioxide,
the total input of CO2 was probably double this (Table 3.1).
The alternative model is less dramatic but perhaps ultimately
more compelling. In 2012 Robert M DeConto (University of
Massachusetts) and coauthors published an alternative model that
owed its origin to the work of the Serbian astronomer, engineer,
mathematician and geophysicist Milutin Milankovi. In Chap. 2
how the Milankovi cycles influence the climate of North Africa
in the last few million years was discussed, but their effects can
be traced even further afield. In the Eocene their cyclical effect on
warming in the Arctic is apparent in reconstructed temperature
records.
In DeConto and his colleagues work, it is suggested that
Milankovi cycles triggered the hyperthermals by themselves.
Rather than an extreme bake-off involving volcanism, DeConto
suggested that the alignment of tilt, precession and orbit tipped
the Earth across a climatic divide where the northern permafrost
catastrophically melted. They quote the apparently rapid rise in
temperature over a short 10,000 year window, as well as the cycli-
cal pattern of the events that appear to repeat with a two million
year period. This period, corresponds to a longer cycle over which
there are changes to the shape of the Earths orbit and its axial tilt.
The Milankovi cycles, mentioned above, which are relevant to
todays climate, are embedded within these (Fig. 3.1).
In the world of 55 million years ago, carbon dioxide concen-
trations were higher than at present to begin with. With 700900
parts per million (ppm), the level then is more than twice the con-
temporary value of 400 ppm. More importantly, the arrangement
of the continents allowed warm waters to circulate much more
efficiently from the tropics to the poles. The combination of higher
greenhouse gas concentrations and an underlying climate that was
biased towards warmth ensured that the planet was ice free at the
poles. Temperatures in the Arctic may have been as high as 24 C
during the summer months. Antarctica, although far from frigid,
was at least cold enough to have permafrost underlying its forest
and tundra. This was also true of northernmost North America
and Eurasia, which were in the process of fissuring.
FIG 3.1 A plausible, but speculative representation of wind flow during

the northern summer in the early Eocene. The continents were assuming
their modern positions and the rising Himalaya were beginning to gener-
ate something akin to the modern Asian Monsoon. Most of the moisture
for this came from the south and from the western Tarim Sea. Tropi-
cal conditions extended 50 north and south of the equator. Presumably,
westerly (zonal) winds were weaker than present to allow much stronger
transport of heat to the Poles from the tropics. Paleomap courtesy of
Christopher Scotese
As the Eocene began, the planet crossed a threshold where

the right combination of Milankovi cycles and underlying
atmospheric carbon dioxide would thaw the permafrost at both
Poles. When this first happened, 55.3 million years ago, around
3000 trillion grams (or 3000 billion metric tons or gigatonnes) of
methane would have been liberated over a period of ten millen-
nia. This was enough to raise global temperatures for a few hun-
dred thousand years, until the permafrost was exhausted and the
methane converted first to carbon dioxide then inorganic carbon-
ate rocks on the ocean floor.
With the ebb and flow of the Milankovi cycles, the perma-
frost would partly reform before thawing out once more. Yet each
refreeze was less than the time before because the underlying cli-
mate remained warmer than it had been previously, thanks to the
carbon dioxide that remained from past thaws. Eventually, the
permafrost was largely eliminated and the process ended at around

52.6 million years ago. For the remainder of the Eocene, tempera-
tures remained several degrees warmer than today, but gradually
lowering levels of carbon dioxide and continued reshuffling of the
continents would eventually bring the hothouse Earth to an end.
By 34 million years ago, India was fully engaged with Eurasia
and the Tibetan plateau began to rise in earnest (as seen in Chap. 2).
Around the same time Australia uncoupled from Antarctica, open-
ing up the southern ocean and allowing the circum-Antarctic cur-
rents to become established. This largely cut Antarctica off from
the rest of the planet and allowed the air over it to chill even more
starkly. Antarctica then swung round so that it sat directly over the
South Pole. As Tibet rose, carbon dioxide was flushed out of the
atmosphere as Tibets rocks were eroded. A chilling Polar Antarctica
reflected more of the Earths radiation in the winter and released heat
more efficiently than a Polar Ocean ever could. Over the next 20 mil-
lion years the continent would slide into a near eternal deep freeze.
As it disappeared under highly reflective ice, and carbon dioxide con-
tinued to fall, the stage was set for the next wave of ice agesthe
underlying level of carbon dioxide is the driver of the change. The
Milancovi cycles are superimposed on top of these and provide the
icing on the meteorological cake.
Were something akin to the Eocene Hyperthermals to occur
today, the outcome would be disastrous. Such a rapid rise in global
temperatures would cause significant melting of the ice caps
and the flooding of much of the planets most habitable regions
for humans. Sea levels were a good 200 m or more higher in the
Eocene than at present. Although part of that rise is to do with
changes to the depth of the ocean floor, much of it is clearly linked
to the melting of the polar ice caps. In this regard, it is important
to remember that we are increasing carbon dioxide at an unprec-
edented rate. From 1800 until present we have added nearly a third
extra mass of this greenhouse gas to the atmosphere. At 400 ppm
(and with a rise of nearly 15 ppm per decade) we are well on our
way to recreating the pre-ice age hothouses of the last 60 million
years. While it seems highly unlikely that we will see a tempera-
ture rise of over 5 C in the next 85 years, a still sizeable rise of at
least 2 C is likely, given our predilection for generating this green-
house gas. Since 1900 sea levels have risen by 20 cm and this rate
is accelerating. The most recent measurements from NASA sug-

gest that sea levels are rising faster than predicted and will rise a
further minimum of 30 cm by the end of the Twenty first Century.
A rise of a meter is now increasingly likely.
Critical though such an event would be for modern man
today, 55 million years ago, it was also crucial in an altogether
more profound way. For around this time, the Eocene warmth con-
verted much of continental Eurasia into a lush tropical swamp,
despite it lying at a latitude similar to that of today. Somewhere
within this swamp the first primate evolved. Thus, while much
of the planet drowned under warm tropical waters, our immedi-
ate ancestor was taking its first steps along the branches of some
ancient, greenhouse gas-driven tropical paradise.
If we inadvertently create a similar climate in todays world,
then would it be so bad? Well, the rise time for carbon dioxide
was around 10,000 years in the Eocene. Its happening in around
one quarter to one eighth that time-span now. Evolution through
natural selection is a remarkable thing, but its capacity for deliver-
ing rapid change is not boundless, and while the resulting ecology
could be a fruitful one for new species, it might prove ruinous for
humans and many other existing life forms.
A Bad Day in the Permian
Although the Eocene saw a rise of around 7 C over the space of

around ten millenniavery much a blink in the cosmic eyethe
rise appears to have been embedded in a much more gradual rise
in temperature that dates to the end of the dinosaurs, ten mil-
lion years earlier. Indeed, during most of the time the dinosaurs
were around, the same combination of favorable plate tectonics,
high sea levels and higher than present carbon dioxide ensured the
planet was a few degrees warmer than it is now. Thus the Eocene
warmth was not catastrophic in its outcome. However, 200 mil-
lion years earlier things were a little different.
251 million years ago the planets continents were arranged in
one large supercontinent, encircled by a vast ocean, Panthalassa. A
wide tongue of water extended inwards broadly along the equator,
known as the Paleo-Tethys Ocean. This ocean was the immediate
precursor of the Tethys Ocean but had broadly the same area and
shape. The Paleo-Tethys was partly separated from Panthalassa

along its eastern flank by a string of micro-continents and island
arcs that would ultimately become China, Indochina and Indonesia.
Over much of what is now Asia, warm waters connected the Arctic
basin with the Tethys.
At the heart of the continent lay a string of recently formed
mountains that separated the future Africa from North America.
These would have intercepted the prevailing tropical easterly trade
winds and delivered abundant rainfall to an area roughly the size
of the modern Amazon basin. North of the mountains, was a large
rain-shadow, covering most of what is now Western Europe and
continental U.S.A. Further south, under the belt of Horse Latitude
high pressures would lie another arid region, which would grade
through semi-arid steppe to taiga and tundra far to the south. Over
the southernmost portion an ice cap remainedthe remnant of a
much larger continental ice sheet that had extended across much
of Gondwanaland 50 million years earlier.
The region that is now Siberia was partially covered in shal-
low seas, with much of the remainder forested, much as it is now.
Figure 3.2 shows a speculative, but plausible arrangement of high
and low pressure areas associated with the climate of the day.
The micro-continents that slice across the eastern end of the
Paleo-Tethys are important as they will isolate this basin from the
wider deep circulation that filled Panthalassa. Despite a slightly
dimmer Sun (96 % its current luminosity), higher levels of carbon
dioxide and a favorable arrangement of continents, which directed
warm currents north and south around their edges, ensured that
the Permian was a largely warm epoch.
Four million years after the planetary snapshot shown above,
something dramatic was to afflict the planet. For reasons, largely
unknown, that may link to the formation of the supercontinent
50 million years earlier, a hot plume of rock rose from the base of
the Earths mantle and began to whittle away at the crust under
Siberia. Perhaps after doming the crust somewhat, Siberia split
open and, as the late, great volcanologist Maurice Kraft would have
said, the blood of the Earth poured out. An area roughly bounded
2000 km from north to south and 4000 km from south west to
north east was covered in hundreds of meters of molten rock and
ash. The eruptions began around 251 million years ago and lasted
FIG 3.2 A re-iteration of the late Permian paleomap to illustrate the

diversity of Pangean landscapes. Rumors of a vast uninterrupted con-
tinent dominated by deserts are vastly exaggerated. Equatorial regions
were dominated by the Appalachians, formed when Africa and North
America (Laurasia) collided. On the western flank of these mountains
would be warm tropical rainforest, grading into savannah then desert to
the south. A monsoon climate similar to India, would be likely. To the
northwest of these mountains, central North America and Europe would
have been hot desert, grading into steppe and tundra to the north. The
Eurasian portion was largely submerged under fairly warm seas, giving
it a climate similar to western Europe. The large southern ice cap was
largely gone with steppe, taiga and tundra surrounding the pole. Underly-
ing paleomap courtesy of Christopher Scotese
for a few hundred thousand years. Amongst the litany of disasters

that ensued, lavas poured into the Arctic Ocean vaporizing meth-
ane clathrates on the ocean floor; they disintegrated and inciner-
ated developing coal beds and also likely explosively decimated
the permafrost along Siberias northern edge.
Vast in their own right, such eruptions could not directly do
much damage outside Siberia. However, at the end of the Permian
almost all major land animals and much of those in the oceans were
extinct. The planet was cooking in its own juices and lifeaside
from microbial lifewas nearly obliterated in its entirety. How then
do we go from the eruption of lavas in one corner of the globe to near
obliteration of the Earth as a habitable planet?
The key to understanding the near annihilation of life was the

temperature of the globe and the arrangement of the continents:
both of these factors appear to have conspired with the greenhouse
conditions established by the eruptions to change the composition
of the oceans and the atmosphere.
Lets picture the scene. At the end of the Permian, dinosaur-like
animals dominated the land along with many reptiles and the hum-
blest beginnings of the mammals. The newly assembled Pangaea has
a diversity of landscapes from a sizable ice cap in the south, through
deserts, tropical swamps and open grassland. The levels of carbon
dioxide are roughly three times that at present, giving rise to a very
temperate world. Into this world 251 million years ago, poured over
1,700,000 cubic kilometers of basalt and pyroclastic material1in
places over 2 km thick. Such a volume of basalt would cover all of
China to a depth of around 300 m, or the punier UK to a depth of
more than 12 km. The eruption would have been accompanied by
the steady release of various gases, including gigatonnes of sulfur
dioxide, hydrochloric acid, hydrofluoric acid and much more sub-
stantially 12001800 gigatonnes of carbon dioxidealthough the
final amount of this may have been far greater (Table 3.1).
Table 3.1 Comparative changes in carbon dioxide levels, temperatures and

the time over which these happened at two natural and one man-made green-
house
Averaged
Period for Total mass rate for
tempera- of CO2 CO2
Temperature ture change (Gigatonnes/ addition
Period change (C) (years) Gt) (Gt/year)
Permian (PT) 810 over 60,000 5400 0.09
boundary 60,000 years
Ecocene 58 over 20,000 27000 0.10.35
10,000 years
Anthropocene 0.9 over 200 200 ~340 since 3.4
years 1901
Changes in the Permian were gradual, occurring over many millennia. Those in the
Eocene, although less extreme were faster (perhaps 100 centuries). Currently, humans
are setting the pace with the fastest rate of carbon dioxide rise and the fastest rate of
temperature change (approximately 1 per century versus 0.050.08 per century for
the Eocene and 0.016 C per century for the close of the Permian). The figure for the
anthropocene is an averaged one for the last 116 years. It hides the fact that most of
the input of carbon dioxide has occurred since 1940
1
This is an absolute minimum with the likely volume exceeding three million cubic kilometers, or
more than twice the Deccan Traps in India.
Seth Burgess (M.I.T.) and colleagues narrowed down the

Permian extinction to a window only 60,000 years wide sand-
wiched between 251 million 942 thousand years ago and 251 mil-
lion 880 thousand years in the past. Through the examination of
carbon isotopes, the authors identify a major change in the way car-
bon was being cycled through the ecosystems of the planet imme-
diately before the extinction event. It takes another 500,000 years
before the system appears to return to normal. In that 60,000 year
window, the biosphere contracted rapidly as species after species
became extinct. The 500,000 year period that followed marked the
recovery of the planet, where the surviving species begin to radi-
ate and fill the vast bank of free habitats that were now available.
The method of annihilation appears to have been anoxia:
a lack of oxygen. Almost certainly this was accompanied by an
altogether smellier killer: euxinia, a lethal combination of lack of
oxygen and a wealth of hydrogen sulfide. The arrangement of the
continents ensured that there was very limited cold, deep water
circulating from the poles. In the modern Earth, this method drives
cold, oxygen-rich waters from the Polar Regions towards the equa-
tor through the ocean depths. This guarantees a supply of oxygen
to the planets deepest waters maintaining an oxygen-rich ocean.
Whenever the waters get too warm, ice will obviously be limited
and this deep, cold flow of oxygen-rich water will cease. The Paleo-
Tethys, in particular, was vulnerable to becoming anoxic as it was
partly cut-off from the deep circulation along its eastern flank.
Instead of temperature differences driving circulation, in the
relatively isolated Paleo-Tethys circulation would have been driven
by evaporation from the surface. In this system a dense, warm,
brine-rich current develops that takes warm, oxygen-poor water
down into the ocean depths. Once the deep ocean has been filled
with this warm, oxygen poor water, anaerobic (oxygen-hating) bac-
teria thrive. These generate noxious hydrogen sulfide gas as well
as methane and carbon dioxide. Hydrogen sulfide is directly toxic
to most oxygen-breathing organisms and it reacts with dissolved
oxygen forming sulfates that rain out onto the ocean floor. Thus
the oxygen levels are kept low by those organisms that hate it.
Investigations by Paul Wignell (University of Leeds) and
Richard Twitchett (University of Plymouth) revealed that as the
extinction got underway, oxygen levels, even at depths as shallow as
1 m, fell below those necessary for the survival of marine animals

that lived there. Except at the very shallowest levels, the entire vol-
ume of ocean water on the planet appears to have become anoxic
and there are many indications that it was probably flooded with
hydrogen sulfide. The nail in the coffin of the Permian was the
release of thousands of gigatonnes of carbon dioxide. Although car-
bon dioxide dissolves in rain water and ultimately precipitates out
into the oceans, it has a fairly long residency time in the atmo-
sphere. This ensures that with a sufficiently large and sustained
supply, the amount in the atmosphere will build up. The release
of carbon dioxide appears to have compounded pre-existing levels
which were already high: perhaps three or four times the current
level. Adding even more carbon dioxide was a critical event. Such
a large amount of greenhouse gas ensured the ocean circulation
would shut down and allow bacteria to ferment the gases that
would eliminate life.
Although sulfate aerosols from the eruptions, as well as
soot from incinerated forests, would have provided a temporary
cooling blanket, soon thereafter the excess carbon dioxide would
have taken charge of global affairs and driven the temperatures
sky-high. Fossil evidence indicates that global temperatures rose
sharply at the end of the Permian and this almost certainly relates
to the massive release of carbon dioxide by the Siberian eruptions.
With temperatures likely 10 C higher than they are at pres-
ent, the oceans became so hot at the surface that they ceased to
circulate in all but the feeblest of manners. Deep waters became
effectively oxygen free and the marine extinctions that ensued
were as inevitable as they were catastrophic. These conditions
would favor the generation of more hydrogen sulfide, which is
produced under anaerobic conditions. Warm water also holds less
dissolved gas, so warmer waters would favor the release of the gas
to the atmosphere. Further compounding this effect would be the
addition of large quantities of sulfate from the volcanic eruptions,
which can be chemically reduced to hydrogen sulfide by bacteria
living in the oceans. Therefore, soon after the cooling blanket of
sulfate was washed out, even more hydrogen sulfide would have
been available from the oceans to further poison the atmosphere.
Meanwhile, on land animals were cooked, while what
remained of the polar caps melted and flooded many other
habitatsnot with clear ocean waters, but with a noxious brew

containing sulfurous gases. That complex life survived at all is a
testament to the strength and ingenuity of evolution. Were such
a catastrophe to unfold today it would be far from clear if humans
would survive. Adding to the potential woes, halides such as
hydrochloric and hydrofluoric acid, which were released by the
extensive volcanic eruptions, would deplete atmospheric ozone.
Potentially, this would have exposed surviving species on land to
lethal amounts of ultraviolet radiation.
At the time, mammals were far from the top species. Most large
animals were reptilian in nature. The problem these animals have is
their metabolic activities, the level at which their cells tick over is
dependent on the external temperature. Such cold-blooded, or poiki-
lothermic, animals cannot maintain a constant internal temperature
and merely respond to that of their environment. This spells dou-
ble-trouble for the reptile. If the environmental temperature exceeds
its safe level and it cannot do enough to cool down, the chemical
machinery in its cells will cease to function. The high temperatures
of the Permian would have been bad enough in this regard, but the
high temperatures would also have sped up the metabolism of the
animal. Metabolism in all complex animals depends on respiration
and in such organisms this requires oxygen. While oxygen would
likely have been available in broadly the same quantities as it is now,
the gas hydrogen sulfide competes with it. High levels of hydrogen
sulfide would have prevented the cells from generating the energy
they needed to keep the organism alive in such high temperatures:
many organisms would simply have asphyxiated under these com-
peting demands.
Yet, much like the fallout from the milder Eocene greenhouse,
as far as evolution was concerned, catastrophe was the mother of
invention. While many of the dinosaur-like large animals went
extinct, the gaps that opened up were soon filled by the ances-
tors of the Dinosauria and most importantly for humans, by the
modern mammals. Were it not for this short but violent period of
annihilation, we would not be here today.
What is most interesting is the suggested rate at which tem-
peratures rose. You might imagine that such a catastrophe was
associated with that 10 C rise over a couple of hundred years,
but apparently not. The vast release of carbon dioxide happened
over a fairly protracted period of time. Instead of rapid change, the
increase was 1 C every 6000 years, on average. Hold that figure in

your head as we move on to look at the current rise in global tem-
peratures. If you have a head for another figure it is worth compar-
ing the outpouring of carbon dioxide in todays world with that
in the Permian mass extinction: burning fossil fuels is currently
releasing around 7 gigatonnes (Gt) per annum. Compare this with
0.018 Gt released on average by the Siberian Trapp eruptions
every year of the 300,000 years or so that the event took place.
The numbers for our contemporary period are rather frightening
in comparison.
Perhaps, as a group of M.I.T. and Chinese researchers have
suggested, alongside the global spectacle of extreme volcanism,
there was an altogether more pervasive yet unseen killer. The
researchers suggested that a lone group of bacteria may have
helped annihilate much of the contemporaneous life on Earth.
Although this idea may seem out there in terms of its impact,
it is far from ridiculous. These Methanosarcina species have the
capacity to consume abundant organic chemicals called acetates
that are found in decomposing organic matter and convert it into
methaneour very potent greenhouse gas. Although widespread,
the activities of this bacterial species are largely held in check
by a lack of a crucial metal cofactor: nickel. Without this ele-
ment the enzyme that converts acetate to methane is inactive.
Unfortunately, so the story goes, the basaltic lavas prevalent in the
Siberian Traps are loaded with this element. The researchers pro-
pose that as the eruptions unfolded, nickel that was abundantly
present in the ash fertilized the ocean depths. As this element
rained down, the Methanosarcina set to work, growing rapidly in
number and releasing vast amounts of methane. In concert with
the carbon dioxide from the eruptions, the methane added greatly
to the greenhouse effectturning the planet into the hellish oven
that led to the wipe-out of plant and animal species that demar-
cated the end of the Permian.
In the end the Permian mass extinction probably had many
immediate causes, but with very little doubt the final squeeze of
the trigger was the release of the Siberian Trap lavas. This set about
initiating fatal anoxia in the oceans; the emission of gigatonnes of
carbon dioxide; the release of copious hydrogen sulfide; and pos-
sibly runaway growth of a particular strain of oceanic bacteria that
vented methane. A planet that was already close to its temperature
maximum was shoved over the edge and 95 % of ocean species

and greater than 70 % of those on land were wiped off the planet.
That the Permian mass extinction is the only one known to have
drastically affected insects is also likely to be significant and a
testament to the events severity. Insects are very sensitive to oxy-
gen concentrations in the atmosphere (and while larvae, in water).
This could be another hint that oxygen levels fell, while hydrogen
sulfide rose. Could such an event happen again? Perhaps, given the
manner we are fiddling with our climate and taking into account
the vagaries and unpredictable nature of our planets geology. It
must also be a consideration when we look further afield at life
elsewhere in the universe.
Finally, we turn our attention next to the anthropocene global
greenhouse: the era where humans are affecting the planet at a rate
far in excess of any natural cause. How does our global greenhouse
measure up to the two that we have already discussed?
The Human Factor
Its an interesting and perhaps unavoidable fact that most peoples

belief systems are based on what those around them subscribe to,
rather than being derived from their own personal experiences
and learning. After all, why bother going to the trouble of deriv-
ing something from first principles when you can draw on the
accumulated knowledge of your peers. Such a system of deriva-
tion works fine for pretty much everything we do in everyday life,
whether it is choosing the best brand of pizza or finding the cheap-
est deals online. However, when it comes to systems that require
more than daily experience, or ones which require derivation over
time intervals longer than a human life-span, the collective body
of current peer knowledge is often a poor guide.
Instead of consulting peers for opinion, when it comes to most
of the subjects that actually impact on our longer term survival we
substitute our opinion with those of experts in whatever field is
relevant. Therefore, if we have chronic pain we consult a physi-
cian. Although we might make some initial diagnosis ourselves,
ultimately if the pain persists we out-source our opinion for that
of an expert who has been formally trained to diagnose and treat
our malady. Similarly, if the lights dont work in our home, we
hire an electrician. In each case the person we consult is formally

trained in the relevant field.
Now, aside from matters of life and death, most of us like to
think we are experts in popular science and dont look beyond the
information that often validates our personal view point. Most peo-
ple simply dont have access to the original data or peer-reviewed
work that is published. Instead we access the canned views that are
available in popular media outlets or the Internet. This is hardly
surprising when the language of primary research is fairly impen-
etrable to the armchair scientist, and more importantly may cost
an arm-and-a-leg to access in the first place. This limited access is a
serious problem, for it is the peer-reviewed work that is most likely
to be valid and reliabletwo keywords in science. After all anyone
can say what they like on the Internet: there is absolutely nothing
to guarantee that it is even vaguely true. Thus many inaccurate,
untruthful and often downright bizarre interpretations of climate
science appear, which will be summarized in the next section.
Anthropogenic Global Warming
There are four different viewpoints that people of different per-

suasions adopt when looking at the climate data. By far the most
commonly adopted in the scientific establishment is that the
Earths atmosphere and oceans are warming mostly as a conse-
quence of rising levels of carbon dioxide and, to a lesser extent,
methane and other greenhouse gases produced by human activi-
ties. The Sun and other factors, such as the tilt of the Earth (dis-
cussed in Chap. 2), contribute to, but are no longer driving the
observed changes.
The second viewpoint is that the Earth is warming but that
its part of a natural cycle, most probably driven by the Sun.
The third viewpoint, which partly overlaps the second, is that
the Earth is no longer warming and that this climate pause dem-
onstrates that rising carbon dioxide plays no part in global warm-
ing (or in extreme views) the global greenhouse at all. Given that
temperatures have now risen above the 1998 peak for 2 years in a
row, this view has perforce somewhat fallen by the wayside.
The fourth point of view is probably the most radical, and
posits that the Earth is in fact cooling, not warming, and that we
16
15.5
15
14.5 Global Number of

Pirates (approximate in
14 thousands)
Mean Global
13.5 Temperature (oC)
13
12.5
12
1800 1850 1900 1950 2000 2050
FIG. 3.3 Myth 1Global surface temperatures are driven by sunspot

number/changes in the output of the Sun. (Solar output is linked to the
number of sunspots through complex changes in surface and atmosphere
of the Sun driven by the suns magnetic field.) The data clearly show that
there is no correlation between the twohowever, from the matching
pattern of waves in both curves, the Sun appears to add or detract from
are heading for a mini ice-age. The cooling Earth, in this view,
is driven by lower solar output. Although this point of view is
rarely heard now with the current El Nio in full swing, it was a
dominant skeptic viewpoint throughout the last 15 years. We can
expect this view to become more pronounced from 2017 onwards
if the current El Nio is followed by a generally cooler La Nia, as
is often the case.
Dissension
Lets tackle the second viewpoint: that there is warming but that
it is part of a natural cycle and linked (presumably) to the Sun
and also to a lack of volcanic eruptions that would otherwise
cause cooling. The problem with this thesis is that the solar out-
put declined over the last four decades in terms of sunspot num-
bers, yet temperatures continued to climb with each decade being
warmer than the one preceding it (Fig. 3.3). The warming trend
goes back at least until the 1960s. Before this point, the trend
continues but it is more complex (Fig. 3.4). When one compares
the solar output and the temperature record there is an apparent
match between the two when one looks at low amplitude changes
in temperature. The Sun may well be modifying the temperature
variation by a small fraction of the total (roughly 0.010.02 C) but
it is clearly not a driving influence in dictating temperature.
What about volcanic eruptions? We looked at the potentially
serious effects of these in Chap. 2, since the Dark Ages were truly
dark as a result of up to four catastrophic eruptions around the
530s540s AD. However, to claim that a lack of eruptions would
produce a continuous upward trend is clearly untrue. Volcanic
eruptions produce dips that are superimposed on the background
trend. They do not drive long term trends. Even massive historical
FIG. 3.3 (continued) the overall pattern in temperatures by a fraction of a

degree (the three arrows). Below lefta well known negative correlation
between the number of pirates (only an approximate value) and global
temperatures. Maybe pirates hold back global warming? For more on this
idea visit: http://sparrowism.soc.srcf.net/home/pirates.html Many more
spurious correlations can be found at: http://www.tylervigen.com/
1
!
Temperatuere Difference from mean ( C) 0.8
o
0.6
0.4
0.2
0
1860 1880 1900 1920 1940 1960 1980 2000 2020 2040
Year
-0.2
-0.4
-0.6
Period of declining Period of static

temperature temperature
FIG. 3.4 Myth 2There is (or was) a standstill in global temperatures.

Human perception is a fickle thing. While most of the media prevaricated
over the so-called climate slowdown of 19982013 and its alleged sig-
nificance, a brief sortie into the past shows that the alleged flat trend,
shown by a green bar and exclamation mark is nothing of the sort. The
slow down, even when restricted to 19982013, is virtually unrecogniz-
able in the trend. From 1885 to 1915 temperatures fell relative to the
mean for the period shown (18802014). Temperatures were static (with
variation) from the early 1940s until the mid-1970stwice the length of
the recent slow-down. However, there is absolutely no doubting that
over the period, where most records are available (over one billion read-
ings), temperatures have increased significantly by 0.96 C, relative to
the mean for the period
eruptions such as those of 535550 AD only lowered temperatures

for a couple of decades. They did not suppress it for over a century.
Therefore, one really cannot claim the current upward trend in
temperature is down to a lack of eruptions.
The last climate skeptic positionand it is perhaps the most
widespread and popularis that there is a cooling trend. A very pop-
ular British Tabloid newspaper and one of its columnists strongly
advocate this position, particularly when there is a cold winter.
After all, if you start your record at 1998, between then and 2013
there does indeed appear to be a slight but statistically significant
downward trend in the data. There are many websites dedicated
to this pattern and each has its own slant. The problem with this
interpretation is twofold. Firstly, 1998 was an anomalously warm

year, powered by the strongest El Nio on record at that time. The
19971998 El Nio brought temperatures as high as 29 C across
the central and eastern Pacific, with a rise of 4 C above normal
along the South American coast. Most of North America enjoyed its
warmest winter on record in that yearparticularly the Northeast
of the country: this is a common El Nio effect (see earlier in this
chapter). Temperatures were on average 5 C above normal for the
December to February period. Similarly, the UK basked in a very
mild and wet winter, which was followed by (at the time) record
breaking wet March. The El Nio lasted into the following sum-
mer and ensured that 1998 had record warmth, globally.
Clearly, if you start from an anomalously high positionand
there can be no contesting that 1998 was anomalousthen sub-
sequent years may fail to peak above this. Indeed, with a lack of
significant El Nios since then (2006 and to a lesser extent 2014,
excluded) it is hardly surprising that global temperatures struggled
to exceed this large rise. However, in 2014, without an El Nio,
global temperatures surpassed those of the extreme 19971998 El
Nio. That should be an eye-opener. Many skeptics disputed the
numbers for 2014 when they were first published as there was a
small, but significant possibility they were simply on a par with
1998, but later analysis confirmed the numbers. 2015 went on to
be broadly equivalent to the temperatures of 19971998, with a
moderate to strong El Nio. Before the year was even out, and the
El Nio at full strength, the 1998 record had been smashed, along
with that set in 2014.
As Fig. 3.4 shows, there is clearly no slow-down: the trend
continues upwards as predicted by Anthropogenic Greenhouse
Warming (AGW) models. If you want to argue that inclusion of
2015 isnt fair, as its an anomalous El Nio year, then you must
also remove 1998. Do this and the trend clearly persists: El Nio
or no El Nio there is a consistent upward trend in global tem-
peratures in defiance of any natural background factor. Reflecting
this, in 2015 the NOAA released comprehensive analysis which
thoroughly debunked the cooling hypothesis.
If still unconvinced by this argument, house prices are a use-
ful analogy. From 2008 until 2013 or 2014 house prices declined
in most parts of the western world. If you take the climate skeptic
approach and apply it to house prices then the trend in house

prices is downwards. While that might be true over a period of
5 or so years, no one would argue the trend in house prices has
been down for the last 130 years. Nor would they argue that the
trend in house prices will continue downwards for the foreseeable
future. The recession of the last few years, now largely over, was
simply the latest blip in a series of blips, which have punctuated
the upward trend in house prices over the last few hundred years.
The cooling trend interpretation is perhaps personified by US
meteorologist Joe Bastardi who has argued for astronomically
driven cooling since at least 2011. Bastardi also claims that the
southern ice cap has increased to record levels, which in his and
many other climate skeptics minds indicates that there is a nat-
ural cycle in the amount of ice between the north and south polar
regions.2 However, the data (Fig. 3.5) clearly tells a different tale.
This brings us to another argument used in favor of no
global warming, that the amount of sea ice around Antarctica
has increased. Therefore, the argument goes there can be no global
warming as this would surely decrease the amount of sea ice. Yet
a casual examination of the full data on ice caps reveals some-
thing rather different (Fig. 3.5). There is no doubting that the trend
in Arctic sea ice cover is down and has been for at least the last
four decades. Antarctica paints a similar, if slightly more complex
picture. The volume and hence mass of ice on West Antarctica
is in serious declinebroadly matching the Arctic sea ice. The
mass and volume of ice in East Antarctica did show a smaller
increase, but this too may now be in reverse. Overall, the change
in Antarctic ice is downward and has been for at least 10 years
(Fig. 3.5). Therefore, far from increasing the amount of ice on
Antarctica is decreasing and this is clearly in line with Arctic sea
ice. Thus the skeptics claims that there is a seesaw cycle of ice
between North and South is clearly at odds with the observations
of declining ice at both Poles.
If Antarctic ice is in decline, then why is the amount of sea
ice increasing? Take a lot of ice perhaps 1 km thick, melt it and
dump the slush and fresh water in the ocean and there will be two
2
http://www.forbes.com/sites/larrybell/2013/05/26/meteorologist-joe-bastardi-blaming-turbulent-
weather-on-global-warming-is-extreme-nonsense/2/
a Greenland Ice Mass Change b Antarctic Ice Mass Change

1000 1000
500 500
Ice Mass (Gigaton)
0 0
500 500
Monthly Data Monthly Data
5year Doubling Time 5year Doubling Time
1000 10year Doubling Time 1000 10year Doubling Time
1500 1500
2002 2003 2004 2005 2006 2007 2008 2009 2010 2002 2003 2004 2005 2006 2007 2008 2009 2010
c Northern Hemisphere Extent Anomalies Sep 2014

20
20
%
40 1981-2010 mean = 6.5 million sq km
1970 1980 1990 2000 2010 2020

slope = 13.3(+/2.8) % per decade
FIG. 3.5 Myth 3Ice cover is decreasing in the Arctic but increasing
in the Antarctic. These three graphs show the change in ice since 2000
recorded by satellite. Graph (a) shows the change in mass for Greenland;
(b) the change in mass for the Antarctic as a whole, and (c) the extent of
Arctic sea ice cover. In no instance is the mass (or extent, in the case of
the Arctic) increasing. East Antarctica does show an increase in mass
(not shown) but the Antarctic, as a whole, shows a decrease in mass.
West Antarctica is losing mass at a rate roughly four times the gain in the
east. Skeptics pointed to the increase in Arctic cover from 2012 to 2013
and suggested the ice cover was recovering. Do you agree? Arctic data is
from the National Snow and Ice Data Center website (http://www.arc-
tic.noaa.gov/detect/ice-seaice.shtml) and Antarctic and Greenland data
from GRACE
effects. Firstly, the slush can cover a much larger area than the
intact ice on land and secondly, the influx of fresh water, which
naturally floats on top of the denser salt water, raises the freezing
point of the water. This allows for more extensive freezing come
the winter. Thus sea ice can increase in total area even as the total
mass of ice declines. Moreover, the spread of fresh, icy water over
the surface of the ocean alters precipitation and evaporation lead-
ing to greater snowfall, which in turn increases sea ice. However,
the gain in thin sea ice in no way compensates for the extensive
loss of the thicker ice sheets on and around the continent.
Another very common misconception is that carbon diox-
ide and methane are unimportant greenhouse gases because the
wavelengths over which they absorb infrared radiation overlaps
water, which is far more abundant. Therefore, the argument goes,
even increasing carbon dioxide and methane levels has no effect
because there is no more energy left to absorb because water vapor
has already absorbed it. But if the current concentration of water
vapor was all that was needed to set the global temperature, then
changing its concentrations would have no effect on temperature.
This would be regardless of the presence or absence of any other
greenhouse gas. This is blatantly untrue and it can be shown in
the laboratory that changing the concentration of water vapor will
alter the absorption of energy. This is through line broadening and
other effects (Fig. 3.7). If this argument were true nor would we
have to worry about the Earth overheating in a billion years or so
(Chap. 5): increasing the concentration of water vapor would have
no effect. Furthermore, that water vapor has already absorbed all
of the energy makes a very testable prediction. If this is true then
the amount of energy escaping the Earth will be the same now as
it was in 1970 or 1990. We can test this as we have satellite data
of our planets outgoing radiation. Unfortunately for the skeptics
the amount of outgoing long wave radiation has decreased at pre-
cisely at those wavelengths over which both carbon dioxide and
methane absorb. This is in agreement with the idea of greenhouse
gases contributing to warming but contradicts the nonsense that
water vapor has absorbed all the radiation that there is to absorb
(Fig. 3.6). If you are unhappy with that argument then the First
Law of Thermodynamics is not in your favor. Energy is conserved:
less energy escaping the planet means that it must be retaining
that difference in energy and thus be warming up. Stratospheric
temperatures are also on a decline. Again, this is perfectly reason-
able if more energy is being trapped in the troposphere and is con-
sistent with the First Law of Thermodynamics.
The problem with water vapor is that under terrestrial condi-
tions it is very close to its triple point where it will either freeze
or condense. As such water vapor responds to temperature rather
than drives it. Lower the temperature and water condenses out.
+
Change in Outgoing Radiation (DTBK)

CFCs
O
HO
-
CO
-
-
CH
-
, , , , ,

Wavenumber (cm-) (/wavelength)
FIG. 3.6 Testing predictions. Most skeptic sites say that although meth-
ane and carbon dioxide are greenhouse gases, their effect is insignificant
compared to water; or that all of the available radiation has already been
absorbed by water so these gases can have no effect. This is testable.
If this is true then increasing the amount of these gases will not affect
how much radiation escapes the planet. How could it, if it had already
been absorbed by water vapor? However, by comparing data from satel-
lites over the last few decades it is abundantly clear greenhouse gases
are absorbing more long wave (infrared) radiation (making the line dip
into the negative portion of the graph). Methane has a notable effect in
the portion of the spectrum shown here, despite absorbing radiation at
wavelengths that completely overlap with water vapor
Raise it and water evaporates. Consequently, water vapor has a

short residency time in our atmospherearound 9 days on average.
This is how long a water molecule has in flight before it condenses
with others and precipitates out. As such water vapor responds to
temperature rather than drives it. Indeed, water vapor only really
becomes a driving greenhouse gas at high temperaturesabove
70 Cwhen conditions become difficult for liquid water to con-
dense at terrestrial atmospheric pressures.
On planets with temperatures above 100 C, liquid water is
impossible unless the atmospheric pressure exceeds ours. Here,
the vapor pressure will depend solely on the temperature and
water will become a strong greenhouse agent, just as carbon diox-
ide does at terrestrial temperatures. It is likely that such a moist
greenhouse was critically important just after the Earth formed;

and, unfortunately, will come into play once more on the Earth
in the distant future. The grisly fate of the Earth is examined in
more detail later in this chapter. For now, it is carbon dioxide that
is well above the temperature at which it can condense (or rather
re-sublimate). Therefore, it is this gas, with its long residency time
in our atmosphere, which has the greatest role to play in modulat-
ing climate.
If unconvinced by this argument then we could apply the
same logic to Venus. If the argument for water vapor having
absorbed all the available radiation was applied to Venus, then if
we were to suddenly remove 99 % of this planets atmospheres
carbon dioxide the temperature wouldnt go down. This would be
because the carbon dioxide left over will be absorbing all of the
available energy. Indeed, if the skeptic argument is correct then
the trivial amount of water vapor in the Venusian atmosphere
(Chap. 5) should override the carbon dioxide and poor old carbon
dioxide wont do anything to Venuss heat. Or imagine youve got
10,000 dollars in earnings and you earn another $500. Do you tell
the IRS that the $500 is irrelevant because the $10,000 is much
bigger? No, you add the figures together and reluctantly fill in your
tax return as appropriate. Similarly, the effect of carbon dioxide and
methane is additive to that of water. Its not that water overrides
absorption, for as was said before if the current water had absorbed
all that was available then the planets temperature would not
respond to the overall concentration of greenhouse gases: it would
simply reflect a baseline concentration of water, which it does not.
Carbon dioxide therefore is a greenhouse gas, which has not
been in question since John Tyndall experimented with various
pure gases in 1859. Tyndall observed that carbon dioxide, ozone
and water vapor all strongly absorbed radiant heat, yet oxygen,
nitrogen and hydrogen did not. Various measurements since then
have refined the contribution carbon dioxide makes to absorption
of heat (Fig. 3.7 below).
Figure 3.7 also illustrates the effect of changing the concen-
tration of a greenhouse gas. Even if we could guarantee that every
available photon of light at the appropriate wavelength that could
be absorbed is being absorbed, increasing the concentration of
greenhouse gases will still elevate temperature. At higher pres-
Comparison of Transmittance upto Tropopause for 280ppm vs 560ppm CO2

1
280ppm
560ppm
0.9
0.8
0.7
0.6
Transmittance
0.5
0.4
0.3
0.2
0.1
0
500 550 600 650 700 750 800 850
Wavenumber, cm1
FIG. 3.7 Collisional broadeningthe absorption of infrared radiation

at two different concentrations of carbon dioxide: 280 and 560 ppm
(0.028 % and 0.056 %, respectively). At the higher concentration car-
bon dioxide absorbs energy across a broader range of wavelengths,
increasing its potential as a greenhouse gas (green curve). Graph
sources: http://scienceofdoom.com/2011/04/30/understanding-atmo-
spheric-radiation-and-the-%E2%80%9Cgreenhouse%E2%80%9D-
effect-%E2%80 %93-part-twelve-curve-of-growth/and https://www.cfa.
harvard.edu/hitran/ for the source data
sures there are more collisions between greenhouse gas molecules

(of any type). This has the effect of altering the energy of the elec-
trons that link the atoms on these molecules together. The net
effect is that this causes the molecules to absorb energy across a
broader range of wavelengths. Effectively, at higher concentrations
the gases behave more efficiently as greenhouse agentsand this
effect on greenhouse gases can result from an increase in the con-
centration of any gas that these greenhouse gases are mixed with.
So, if we were to crank up air pressure on the Earth, but not change
the concentration of greenhouse gases, the strength of the global
greenhouse would also increase. This process is called collisional
broadening and will be returned to in Chap. 5 when we look at

how Venus and the Earth will twin-up once more, in the future.
Finally, many skeptics take to doubting the temperature
record effectively in its entirety. The reason for this endemic view,
cited since the 1960s, is the urban heat island effect. Essentially,
the argument boils down to the location of many thermometers or
other devices used to record temperature. The argument is that if
you site these devices in built-up areas, the extra heat generated
within the city by the combination of solar heating, respiration,
air con units, vehicles and planes will naturally raise the tempera-
ture. Moreover, as cities have grown in size the effect as naturally
increased over the last century or so. This is a good argument, but
with one very obvious flaw (quite aside from all those rural and
latterly satellite measurements): global temperatures did not rise
significantly from 1998 to 2013and weve seen many skeptics
argue jhmtf that they declined slightly. Hopefully, you can see the
problem: the human population, and the size of cities, has clearly
not declined since 1998 (there are around 1.3 billion more people
since 1999). Therefore, wouldnt one expect a rapid increase in the
heat-island effect, rather than a leveling off? Whats worse for pro-
ponents of the heat island effect is that when measurements are
made of cities 42 % of them show a lowering of averaged tem-
peratures, not a rise, relative to their surroundings. This is because
temperature measuring readings are increasingly taken from park
locations rather than actual concreted areas.
There are over one billion measurements in the global tem-
perature record, and they cant all be wrong. Neither can you pick
and choose data points in order to make a case for or against man-
made global warming. If you accept the validity of a climate hiatus
in the last decade, despite the many studies disputing it requires
accepting a generic rise in temperature since 1885. The record can-
not be right only in the bit you choose to validate your argument.
More importantly, in most analyses that are done, readings from
those stations that are most prone to the heat-island effect are dis-
counted in favor of any nearby rural measurements, thus the effect
would only ever have limited impact, if any at all.
There is one peculiar historical perspective underlying these
arguments: the birth of the industrial era, which most historians
would put at around 1800. However, many skeptic websites erro-
neously suggest that the industrial era began as late as the 1940s.
If you subscribe to this view, then there can be no link between
historical rises in carbon dioxide and global temperatures since
carbon dioxide levels began rising around 1800.
Another dubious argument by climate skeptics is a rather odd
mathematical one. Some skeptics will say that natural variability
is around 0.1 C per decade, which is reasonable. They will also say
that temperatures have only risen by 0.10.2 C or so over the last
two decades. Therefore, the current rise is supposedly within the
bounds of natural variability. But the temperature record supports a
rise of roughly 1.0 C (0.96 C if you want to be precise) over the last
130 years or so: this is roughly ten times faster than the rate of rise
during the Eocene Hyperthermals, the nearest comparison to the
period of current warming that we know of in the geological record.
The rise is erratic but is utterly inconsistent with any natural cycle,
solar or otherwise. The only factor that varies in line with this rise is
the level of greenhouse gasesnotably, but not exclusively, carbon
dioxide. Table 3.1 puts things into a somewhat grim perspective.
Given all of this, despite likely upward and downward bumps
along the road, the general path of the global temperature will be
an inevitable rise unless carbon dioxide levels begin to fall sharply
soon. Assuming a worst case scenario society will continue to burn
all fossil fuels until these are exhausted, or at least become uneco-
nomical to harvest. In this case, assuming we dont turn our atten-
tion to methane hydrates, or experience some form of runaway
where there is a catastrophic release of methane, the temperature
peak will happen in around 300 years. Carbon dioxide levels will
be back at where they were in the Precambrian, but with a more
luminous Sun, the temperatures will be rather more torrid. Bjorn
Lomborg has even argued in The Skeptical Environmentalist that
it may well be more cost productive to ride out the effects of global
warming rather than stop it happening in the first place. Although
Lomborgs opinions are controversial, riding out the storm seems
a more likely political and social outcome than cutting production
or investing in new means of generating energy. So, if we decide
not to tackle warming by changing energy production what will
the likely costs be? The principle cost to the West will be increased
insurance costs tackling the output of climate change: increased
rainfall in some places causing flooding and increased drought in
others, which limits food production or the availability of water.

There is obviously cost associated with loss of coastlines through
sea levels that are rising faster than predicted. Rising seas also
bring enhanced coastal flooding, even when the land itself isnt
permanently flooded. To illustrate this trend, there is now a data
set for coastal flooding in the UK which covers the last 100 years.
This was produced by David Haig and colleagues and is referenced
at the end of this chapter. There is an obvious upward trend in
the numbers of floods as well as their severity during this period.
The problem with these costs is that you can make guesstimates
upfront, but the precise regional impacts on damage to infrastruc-
ture, loss of food production, etc. are hard to determine and clearly
prone to a lot of debate. This doubt affords people the flexibility to
do as little as possible. However, just to underscore the point that
climate is changing and we are the likeliest culprits, well look at
a couple more pieces of evidence.
Drifting Hurricanes Amid a Changing Climate
The Earths climate is certainly changing, but it is true that the

specifics of this change are still being established. Within the
mle of data a few trends are becoming clearer. Take tempera-
ture extremes, for example. Analysis of the temperature records
for Europe reveal that on any given day temperatures are far more
likely to break record highs than they are record lows. Martin
Beniston (University of Geneva) has analyzed the temperature
records of the last 60 years from 30 different climate stations
across the continent. The ratio of days per year with record high
temperatures to record low temperatures increased from 1 to 1 in
the 1950s to 1 to 5.1 in 2013. Indeed, despite the alleged climate
slowdown over the last 15 years, the rise in the ratio of record
highs to record lows has accelerated. It appears a slight rise in aver-
age temperatures has been masking a much more marked rise in
(high) temperature extremes.
Thats not all. Of rather more concern is a slow but steady
shift in the distribution of tropical storms. Although much was
erroneously made about changes in the frequency of such storm
systems, something which did not hold up to scrutiny of longer
term records, there has been a shift in the location of tropical storms
over the last 31 years and in particular the latitudes at which they
reach their maximum intensity. James Kossin (NOAA) and co-
workers analyzed the hurricane track data for the northern and
southern hemispheres.
Analysis reveals that there has been a gradual migration north
and south of the equator amounting to approximately 1 of lati-
tude (53 km north and 62 km south) per decade. This movement
is consistent with the gradual expansion of the Hadley cells that
deliver hot, humid air to the Doldrums and drier, cooler air back to
the Horse Latitudes (described in Chap. 1). The expansion of these
large cells, that dominate the central 60 flanking the equator, is an
observation that matches predictions for global warming caused by
changes in the levels of carbon dioxide and other greenhouse gases.
As these cells expand the amount of wind shear reduces further
north and south, which in turn enhances the conditions favoring
the formation of tropical storms. Most of the migration is appar-
ent within the Atlantic and Pacific basins, which host the greatest
number of tropical storms. The northern Indian Ocean shows the
least migration of storm intensityand interestingly this follows
the opposite pattern, with storms becoming more intense towards
the equator. However, given the relatively small number of storms
the Indian Ocean spawns, the effect of this pattern is drowned out
by that in the Atlantic and Pacific basins. Storms, meanwhile to
the south of the equator in the Indian Ocean, follow the same pat-
tern as the Atlantic and Pacific Oceans. Quite why the northern
Indian Ocean bucks the trend for the rest of the planet isnt clear,
although changes to the strength of the Asian summer monsoon,
which has generally weakened, may be a factor (Chap. 2).
Endemic Misunderstandings
Within the skepticism fraternityand much more widely within
the scientific communitylie a number of underlying miscon-
ceptions that unfortunately help to crystallize the ideas of natu-
ral variability as a cause of the current temperature trends. Most
notably within the literature are many comments relating to the
historically recent climate swings that are perceived erroneously
as being global events. In particular the Medieval Warm Period
and the subsequent Little Ice Age bring a lot of attention to

apparent global trends. If you subscribe to these ideas then the
current upward trend in temperature is clearly not going to be any-
thing new.
However, when these climate swings are examined in more
detail, rather large holes appear in the natural variation picture
and the local and regional reality of these swings emerges. The
Medieval Warm Period is endemically quoted in most climate
skeptic literature, but perhaps more significant is the so-called
Little Ice Age. For it is from this point onwards that anthropogenic
global warming ensues. Lets examine each of these very interest-
ing climatic periods in more detail and unpick the myth from the
climatic narrative.
The Medieval Tepid Period and the Luke-Warm

Little Ice Age
Around 1000 years ago the Viking Empire was extending from
northern Europe across to Iceland, Greenland and eventually
North America. As the true European discoverers of the western
continent, they took advantage of relative warmth that brought
a largely ice-free North Atlantic and a climate suitable for the
growth of some crops and livestock, such as cows and pigs, in
southern Greenland. The climate over the southern and eastern
side of Greenland was more typical of western Scotland with tem-
peratures up to 2 C higher than they are at present in that region.
Although by no means warm, conditions were benign enough
to allow the formation of permanent Viking settlements for more
than 100 years.
In all the Medieval Warm Period, or MWP for short, as this
strange climatic blip is known, lasted approximately 300 years
from 950 to 1250 A.D. its presence leading many to conclude that
the current warm period was simply another upward blip in tem-
perature equivalent to the MWP. However, like so many things,
appearances are deceptive.
Initially, the climate used to determine the extent of the period
was obtained from Europe and the periphery of the North Atlantic
basin. This data revealed that the warmest region, which rivaled or
even exceeded the temperature maxima of the 19602000 period.
However, with further, and more extensive, analysis of global tem-

perature data it soon became clear that the MWP was restricted to
Southern Greenland and the neighboring Atlantic basin. Elsewhere,
globally there was clear evidence of a significant cooling event.
For example, the so-called Medieval Warm Period did not
affect the Anasazi tribes of south western North Americaor
the Maya in Central America. Indeed, the same pattern of North
Atlantic warmth coincides with an extreme mega-drought in the
region encompassing southern United States and Mexico. Prior to
the MWP, the Anasazi had developed a wonderful cultural system,
with complex cities, many carved into and constructed within cliff
faces, such as the famous Mesa Verde Complex in New Mexico. As
the Vikings made their way across the Atlantic, the Anasazi civili-
zation collapsed: the underlying reason was persistent drought asso-
ciated with regional cooling. Temperature records from tree rings
reveal that while the Vikings enjoyed warmth, the Anasazi endured
temperatures that were consistently lower than they are now.
Similarly, but somewhat earlier, the Mayan civilization col-
lapsed in the period 850950 AD and there is at least circumstan-
tial evidence that they too were reduced in power by increasing
drought in Central America. This begs the question, what meteo-
rological phenomenon united the demise of the Anasazi and the
success of the Vikings? The answer was a prolonged La Nia
(Chap. 2).
For much of the period from 950 to 1250 AD the central and
eastern Pacific basin was afflicted by La Nia conditions. This
brought cooler, drier air across southern North America and much
of Central America. La Nias bring very predictable changes in
global temperatures, with the coolest conditions around the
shores of the eastern Pacific (Fig. 2.5). However, these conditions
also redirect the polar front jet stream so that after dipping south-
wards across the Western States region, it arcs it back northwards
towards Greenland. This brings considerable warmth to the north
western Atlantic basin and Western Europe. Thus, while the
Anasazi lands cooled down and dried out, the Europeans enjoyed
warmth (and Viking raids).
Across the rest of the globe it is apparent that much of cen-
tral Eurasia was colder than average, while the La Nia brought
warmer and much wetter conditions to the western Pacific and
New Zealand, with temperatures in the latter region up to 0.75 C

higher than the average for 19601990. There is circumstantial evi-
dence that Australia was also affected by prolonged La Nia con-
ditions, but climate records for this region during this period are
scarce. However, it is clear that what was meat for some was mur-
der for others. The Medieval Warm Period was simply a regional
event, not a global one: the current warm period is global and has
no natural counterpart.
Similarly, a lot is made in Europe and North America of the
Little Ice Age (LIA, for short)the period lasting roughly 2300
years between 1600 and 1900 where severe winters and cool
wet summers afflicted Europe and the Atlantic seaboard of the
U.S. However, once again, more extensive analysis of global tem-
perature records (and we are primarily looking at tree-ring evi-
dence and in some instances isotopic evidence) the timing and
extent of the cooling clearly varies dramatically from region to
region. China had its coldest period of the last 1000 years around
1200 A.D., while the mainland US was coldest in the nineteenth
century. 1200 A.D. is, as you may notice, when Western Europe had
the Medieval Warm Period Scandinavias lowest temperatures of
the last millennium were around 1600 A.D. while the UK suffered
the most extreme cold of the period in the 1700s. Indeed, climate
data indicates that while central England shivered, parts of North
America and Labrador were as warm as the average for the period
19601990. The Little Ice Age, it seems, was more a scattering of
ice cubes, rather than a global freeze.
The sporadic nature of this period unfortunately disposes of
the supposed link between global cold and the lack of sunspots,
known and the Maunder Minimum that took place during the
1600s. Much was and still is made of the lack of sunspots and
the cold period. Since there was no global chill, but instead a
period of enhanced climatic instability that spanned well beyond
the sunspot minimum, there can be no causative link. The cur-
rent warming trend is then even less likely to be associated with
a change in sunspot numbers as it clearly wasnt in the past, as
the sunspot numbers would not affect Scandinavian temperatures
alone. Currently, there are a few skeptic websites that suggest we
are (or will be) heading into a mini ice age (one even claims that we
are already there). The driver, it is suggested, is a weak solar cycle.
Yet, as historical records show there is no link between the two.
There is something of a link, albeit a weak one between warm sum-

mers in the northern hemisphere and solar maxima, but there are
plenty of summers that really dont fit the bill, at least in Europe.
The moral of these two tales is to beware of cherry-picked data.
Closer analysis of the temperature record also reveals some
interesting regional biases that seem to cause similar mispercep-
tions. The temperatures across continental US (and this was a sur-
prise to me) fell so far in the 1930s that it took nearly 60 years
for them to reclaim the ground. This was not a pattern that was
repeated elsewhere on the planet. Quite why the continental US
dipped more strongly, then warmed more slowly than the rest of
the planet is unclear but it is likely to relate to the overall pat-
tern of regional climate and airflow, or perhaps to the release of
additional pollutants that dimmed the surface. For example weve
already seen in Chap. 2 how the tragedy of the terrorist attacks
on New York and the Pentagon in 2001 affected regional tempera-
tures. To reiterate, immediately after the events of 9/11 planes
were grounded across the entire continent. Over the following 3
days, trans-continental temperatures showed an increase in the
diurnal (daynight) range of 1.8 C compared to the days running
up to the attacks. Clearly, something as seemingly insignificant as
airplane contrails could drastically affect temperatures.
All of this evidence shows that the idea that temperatures
will not be affected by increasing amounts of methane is clearly
flawed. By making a straightforward comparison of the amounts of
carbon dioxide and the rate of temperature change with the Eocene
and Permian Hyperthermals, hopefully the current changes can
be put into perspective. With the three tales of climate warm-
ing behind us, its time to look at how drifting continents led to
changes in the strength of Monsoons across Asia and North Africa.
These illustrate the volatile nature of the global climate and the
underlying role for carbon dioxide and plate tectonics in regulating
the terrestrial climate system.
The Effect of Global Climate Catastrophes on Life
Each of the climatic catastrophes had radically different effects

on life on Earth. The snowball, or slush-ball, episode coincides
with the emergence of complex life. Although this may be just
that, coincidental, a global glaciation event would eliminate many
forms of life, clear niches of competition and create new environ-

ments for lifes re-emergence and diversification. Although one
would certainly regard the connection as circumstantial at pres-
ent, there is a clear case for the glaciations having led to the devel-
opment of complex, multi-cellular life. As such, we may owe our
existence to this period.
Similarly, the Permian-Triassic boundary marks a near
wipe-out of complex life on Earth and is clearly contemporane-
ous with the eruption of the Siberian Traps. Although it appears
clear, the eruption of vast amounts of carbon dioxide appears to
have conspired with an increasingly anoxic ocean to drive the
global mass extinction. However, despite the severity of The
Great Dying once again, although somewhat protracted, within
ten million years of the event life had regenerated, firmly estab-
lishing both the mammals and the dinosaur lineages and hence-
forth the birds.
The first Eocene event is perhaps the closest analogy with
the current phase of warming. While the end-Permian event may
have taken 60,000 years, much of the PETM warming was concen-
trated within 10,000 years (although 20,000 years probably elapsed
for the full warming to unfold). The Eocene saw the extinction of
some marine species, particularly those cold-loving species found
at greatest depth. Like the PT extinction, this was probably as a
result of the deeper oceans becoming anoxic as a result of poor
circulation. However, the Eocene also saw the evolution of many
large species of mammal and the first primate. Undoubtedly, at
least from our perspective, the Eocene Hyperthermals were criti-
cal for our evolution. Without the Eocene warmth, there would be
no modern humans or the large mammalian fauna we preyed upon
at the close of the last ice age.
Strange Tales from the Frontiers of Political Climate Skepticism

The UK Independence Party Climate Policy 2012Keeping
the lights on had some wonderfully silly material in it. The
document contained a graph that purported to show the tem-
perature records for the last ten thousand years. This indi-
cated that temperatures now are no higher than they were
in the Medieval Warm Period around 1000 years ago. Sadly,

the data was shown to be flawed 3 years before UKIP used
it. The Medieval Warm Period has been shown to only have
been warm in the north west Atlantic. It was in fact cooler
elsewhere. The 2014 document states that the current 0.8 C
rise is part of a natural cycleand they favor the Suneven
though the pattern of solar radiation and temperatures do not
match.
UKIP also describe carbon dioxide as a trace gas essential for
plant growth (a common theme). Yes, but they dont seem to
understand simply whacking up the concentration will not
increase crop yield. Plants do not respond in this manner to
changes in carbon dioxide: other factors are at play, which
limit the beneficial effects of extra carbon dioxide. More
importantly, cereal crops, upon which the world depends for
the bulk of its calorie intake respond poorly to increasing car-
bon dioxide, whereas competing broad-leaved weeds do rather
better. Worse still, geological evidence (PETM and ETM2)
suggests that increasing the level of atmospheric carbon diox-
ide results in a stunting in the size of mammalswhich now,
of course, includes us.
As the science above reliably shows, increasing carbon diox-
ide increases its potential to absorb infrared radiation. UKIP,
a potential party of government, do not seem to understand
this, grossly underrating the effects of carbon dioxide and
other greenhouse gases and openly seek to expand the con-
sumption of coal.
Worse still was the bizarre claim by Maurice Newman, who
chairs the Australian prime ministers Business Advisory
Council. Newman had adopted the world is heading for
a mini-ice age scenario, also touted by Piers Corbyn, of
Weather Action in the UK1 and several others on the
Internet. Newman claimed that there was evidence that the
world is set for a period of cooling, rather than warming. This
would leave the nation ill-prepared for the coming climate
catastrophe. There are several pundits in the cooling camp
that even suggest that we should burn more coal to avert
this coming disaster. That begs the question (particularly

in a country blessed with a (relatively) benign climate like
Australia), are you burning the coal to make carbon dioxide to
avert cooling because you realize CO2 is a greenhouse gas (and
causes warming), or are you expecting such a downturn in
temperatures that the average Australian will have to invest
in a coal fire to keep warm? Strange indeed.
1
Weather Action pdf: WANews13No5.pdf ((2013) On some
publications, Piers Corbyn has claimed that the mini-ice age
is upon us.
Yet, this apparently rosy tale for mammals comes with an

interesting caveat; one which impacts on the most currently fash-
ionable skeptic position that extra carbon dioxide is a boon for
plants. While the Paleocene (the immediately preceding geological
phase) saw a spurt in the size of mammals, as did the late Eocene
and Oligocene, the warmest period of the Eocene saw mamma-
lian sizes decrease. Indeed, this was not an isolated event limited
to a few species: this shrinkage in mass was across the mamma-
lian board and occurred not once but twice, first 55 then 53.5 mil-
lion years ago. Philip Gingerich of the University of Michigan has
attributed this temporary shrinkage to a loss in the nutritional
value of the abundant vegetation that the hyperthermals stimu-
lated. While many originally attributed the observed reduction in
the mass of mammals to the increase in heat, clearly mammals
evolved during the Mesozoic when the planet was warm and sur-
vive today in places such as Africa. Heat, per se, is not an issue for
mammals. However, when one looks at the effect of increasing the
amount of carbon dioxide on plant growth, yes plants increase in
mass, but their nutritional value does not increase in turn, mean-
ing any animal feeding on plants will need to eat more in order
to prosper. Thus a herbivore will succumb to the contrary need
to eat more to get nutrients, but be unable to do so because the
sheer mass needed for survival will exceed what it can usefully
gather and process. Thus, mammals shrank in size even though
there was more food available. Bear this in mind when you see the
latest skeptic argument: more carbon dioxide means bigger plants.

Yet beyond the immediate concerns that global warming will
undoubtedly have for us humans, more carbon dioxide wont do the
planet (as a whole) any harm. If we look beyond the next 50 years
and assume that we will do at best too little to avert a temperature
rise of a few degrees for the planet, such a rise in temperatures
might be beneficial over time to the future evolutionary history of
life on the planet as a whole, though not likely to humans.
One must bear in mind that a projected three to four degree
rise (if nothing is done to limit carbon dioxide emission) will be suf-
ficient to melt the West Antarctic ice sheet. We know this because
during the Pliocene, when carbon dioxide concentrations were
comparable to those of today (400 parts per million, or 0.04 %),
temperatures were 3 C above those of 1800. Two papers published
by Tim Naish (Antarctic Research Centre, Victoria University of
Wellington), David Pollard (Pennsylvania State University) and
Robert DeConto (University of Massachusetts) and colleagues
demonstrate that during the Pliocene, between three and five
million years ago, the West Antarctic ice sheet periodically col-
lapsed and reformed over intervals around 41,000 years long. Each
collapse took, at most, a few thousand years to complete. The cycli-
cal nature was driven by, you guessed it, Milankovi Cycles, that
continued to drive the longer term cyclical glaciations of the last
Ice Age a couple of a million years later. The 41,000 year cycle is
linked to the angle of tilt of the Earths axis and is embedded within
the longer 100,000 year-long cycles that were evident during the
Eocene that were described earlier in this chapter. Interestingly,
after 800,000 years ago the dominant 41,000 year cycle gave way
to the current 100,000 year-dominated cycle. The origin of this
changethe Mid-Pleistocene Transitionis currently unknown
but may well relate to (what was overall) declining carbon dioxide
levels in the atmosphere.
Remember that in the Pliocene changes to the amount of
heating in the Arctic underpinned the warming: there was not a
global increase in the amount of radiation the planet received, so
the 23 C rise that caused the polar melt was not a global phe-
nomenon in the sense that it is today. Thus we should be wary
of driving a global increasing in insolation and making a direct
comparison with something that primarily impacted the North
and South Poles. The effects of our activities are likely to be much
broader in their impact.
The cyclical rise and fall in sea levels during the Pliocene
amounted to roughly 10 m. Seven meters of that was from the
collapsed West Antarctic Sheet, with a further three meters from
the melting of ice on East Antarctica. Ten meters is rather a lot.
Imagine all of our coastal cities progressively drowninga process
that is already underway as sea levels rise at their accelerating rate.
Already somewhat alarming are measurements made by Fernando
Paolo (Scripps Institution of Oceanography) of the floating por-
tions of Antarcticas ice sheets. Overall, the rate of loss increased
from 25 cubic kilometers per year from 19942003 to 300 cubic
kilometers per year in the period since then. Meanwhile in East
Antarctica the process of increasing ice thickness ceased. Now, for
the first time since the Pliocene, Antarcticas ice is undergoing a
profound net melt. For some floating shelves, in West Antarctica
the rate of melting is now so high that it means they will have
melted in their entirety by the year 2100. Now, remember, fortu-
nately, these floating shelves do not contribute much to sea level
rise as they are already part of the ocean. However, their loss will
facilitate enhanced movement of the ice shelves that lie behind
them, increasing their rate of flow into the ocean and thus their
subsequent melting. This is because these ice shelves are partly
grounded on the ocean floor, below and buttressed against islands
and mountains in the neighboring ocean.
So while the entire West Antarctic Ice sheet wont have melted
in the next 50 years, the effects of melting Antarctic ice will be
progressive, with positive feedbacks accelerating the melting pro-
cess. Even London is vulnerable, not only because of accelerating
sea level rises, but because ever since the northern ice cap began
melting 15,000 years ago, the whole of the UK has been tilting,
upwards in the north and downwards in the south. Undoubtedly, a
capital city such as London will be defended for some considerable
time, but smaller towns may be less fortunate.
If we limit temperature rises to 2 Ca worthy but a likely
unachievable targetthen the West Antarctic sheet will probably
fray around the edges but otherwise persist. The world will be
warmer, the climate, perhaps, but not definitely more fractious, but
nothing dramatic will have happened. Several billion dollars extra
on coastal and river defenses might well do the trick. For exam-
ple, in the UK we currently spend 1.1 billion (around $1.7 billion
US) per annum on flood damage costs. This amounts to roughly
1 penny per pound on income tax in the UK. The Parliamentary
committee overseeing the spending projects an extra 1 billion
cost per annum over the next 20 years.3 This is regarded as finan-
cially acceptable by The Market and thus will continue. You can
easily see that, globally, the cost of flood defenses will exceed $500
billion by 2030: this figure is from the World Bank. Thats a lot of
extra taxationas well as increased insurance costsmerely to
defend coastline many of us dont live on.
The long-term cost of defending the coastlines and river mar-
gins is likely to be prohibitive in the longer-term and when we
decide that we are unwilling to keep subsidizing those directly
affected by global warming we will surrender those lands. Areas
such as peninsular Florida are already vulnerable to sea level rise.
In the case of this low-lying state its not so much flooding that is
the problem (although it should be of concern); its more the effect
on the supply of fresh water. Here, the States supply of drink-
ing water is effectively confined to a narrow band sitting atop the
saltier waters that permeate the deeper ground. As sea levels rise
at a few millimeters per decade this layer is being progressively
displaced by the rising tide of salty water. Residents of the State
are already aware of the growing problem caused by sink-holes. As
water levels riseand this is increasingly acidic sea waterthe
limestone foundations of the State are dissolving away, opening
up sink-holes that threaten property or cost increasing amounts of
money to shore-up.
An Icy Future?
Heres where the current debate on global warming gets interest-

ing. Imagine we burn most or all of the current available supply
of carbon-rich fossil fuels. What will be the effects on the planet
in the immediate and longer term? We can look at two scenarios:
one where climate skepticism is overwhelmed and we restrict
3
Source: http://www.parliament.uk/business/publications/research/brieng-papers/SN05755/ood-
defence-spending-in-england
further excessive input of carbon dioxide to the atmosphere. In

this scenario, energy production switches from predominantly
fossil fuel incineration to greener technologies, so that we only
release 1000 Gigatons (see Table 3.1 for a comparison with the
Eocene and Permian-Triassic events). Carbon dioxide levels are
restricted to 500600 ppm for which the peak level occurs around
the year 21002300. There is a delay in reaching the peak even if
we turn off the supply around the year 2050 because of feedbacks
within the Earths climate system that will cause the release of
more carbon dioxide from buried stores, including the oceans and
thawing permafrost. Meanwhile, global temperatures continue to
climb for another hundred or so years after carbon dioxide peaks,
due to feedbacks involving the oceans.
With a cut-off of the supply by 2100 temperatures max
out around 34 C above those of present in around the year
2300 AD. Since carbon dioxide has a residency time running into
thousands of years, the subsequent decline in global temperatures
lasts for tens of thousands of years. Indeed, it could take over
50,000 years for the amount of carbon dioxide and global tempera-
tures to reach pre-industrial (pre-1800) levels. During this time,
humans successfully hold off the next glaciation of the northern
hemisphere, which is due in 50,000 years (Fig. 3.8), until at least
130,000 years from now.
In the more extreme scenario we profligately consume all the
fossil fuels we can and make no effort to switch to cleaner tech-
nologies until we absolutely have to. In this scenario we obliterate
enough fossil fuels (around 1300 Gt) to release 5000 Gt of carbon
dioxide. Consequently, carbon dioxide levels keep rising until at
least the year 2150 to 17001900 ppm or higher. The final level
depends sensitively on the nature and extent of the inevitable
impacts that such profligacy will have on other carbon stores: these
include clathrates on the ocean floor; permafrost carbon stores; and
the Amazonian forest (whatever remains of this in 200 years). This
scenario is altogether more unpleasant. Carbon dioxide levels max
out 5001500 years in the future, then take more than 300,000
400,000 years to fall to levels comparable with those of today.
Indeed, one million years into the future they will still be above pre-
industrial levels. The very slow decline is simply a consequence of
Irradiance (mulples of 10W/m2)
51
48
45
-200,000 -100,000 Today +100,000 +200,000 +300,000

Time (Years)
FIG. 3.8 Changes in the amount of radiation received at the Earths sur-
face at (the insolation) a latitude of 65 N, 1 month after the summer
solstice. The value varies in a complex, yet utterly predictable way deter-
mined by the overlapping effects of the angle of precession; the axial tilt
and the amount of eccentricity in the orbit of the Earth around the Sun.
These Milankovi cycles accurately describe the patterns of cold and
warmth back at least as far as the Eocene and can be used to predict
when the Earth should return to glaciation in the next few million years
at least. Cold periods are indicated by blue, translucent bars overlapping
the curve. Data used to construct this figure came from NOAA paleocli-
mate databases
the inability of the natural sponges for carbon dioxidethe oceans

in the short term and weathering of continental rocks in the longer
termto cope with such a massive rise in the levels of this gas.
Consequently, the planet would be looking at a rise in carbon diox-
ide not seen since the end of the Permianand over a much shorter
interval of time than the Siberian Trap event. Warming would be
so severe that the planet would lose almost all of its ice. Even the
mighty sheets of East Antarctica could not hope to cope with such
a prolonged thaw.
What then would be the immediate impacts for humans?
Well, at the moment a sea level rise of 710 m over the next few
hundred years is probable with a modest 23 C rise in tempera-
ture, if West Antarctica does what it has done under similar condi-
tions in the past few million years. The precise timing, as was said
earlier, is unclear but has a very definite geological precedent. Sea

levels rise because of three factors: extra melt-water; extra expan-
sion from higher temperatures; and finally a gravitational boost
from the redistribution of mass. The Antarctic ice sheets are so
massive that their combined gravity holds water pinned up against
the continent. Melt this and the gravitational pull of Antarctica
declines and the waters slosh outwards to cover more of the globe.
As was said previously, such a melt will have significant financial
implications for all of the nations of the planetat least those
with coastline.
Take a more severe scenario with a 2000 ppm carbon dioxide
level and you look at a 59 C rise in global temperatures lasting
centuries, which will be followed at 30005,000 AD by a peak rise
in sea level of around 70 m (approximately 230 ft). Such a rise
would inundate much of the low-lying areas of the continents and
necessitate a massive reorganization of human infrastructures,
including roads, cities and the food supply. Now, thats a long way
off, but once again, the geological record is clear: increase carbon
dioxide and you raise temperatures. Raise temperatures and you
melt ice and raise sea levels. The amount of carbon dioxide that is
released would vastly exceed the capacity of the natural systems
to absorb it, therefore, temperatures would take nigh on five mil-
lennia to fall by a single degree Celsius.
At least once we consume whatever all available fossil fuels,
the inevitable (eventual) trend in carbon dioxide levels is down-
wards. Carbon dioxide is vulnerable on any planet with water, as
weve seen. If it is not continually supplied it will be taken up by
the planets rocks and oceans and slip silently out of the atmo-
sphere. Thus, regardless of whatever point we choose to stop burn-
ing fossil fuels the supply of the gas in the atmosphere will decline.
The driver of this change is the Sun. Since the planet and Sun
formed, 4.56 billion years ago, the Sun has gradually become more
luminous. As this occurs, the amount of radiation received by our
planet has naturally increased. As it has done so the planets meteo-
rological systems have slowly pepped up. As theyve done this, the
rate of weathering and erosion has increased, in turn. More heat
means more evaporation and more evaporation means a stormier
and wetter planet. This in turn accelerates erosion, which removes
more carbon dioxide from the atmosphere. The two forces, declin-
ing carbon dioxide and increasing solar luminosity, have worked
in unison to keep the planets temperature roughly constantand
well within the range of habitability. In summary: more Sun; less
CO2; stable conditions. Even in the extreme scenario where we
incinerate all of our coal reserves and cause a massive release of
methane from clathrates on the ocean floor; from the permafrost
and from the tropics, we will still face declining carbon dioxide
levels from 400,000500,000 years in the future.
Regardless of whether we decide to maintain a global green-
house or not, our ancestors will face the challenge of dwindling
carbon dioxide and rising temperatures. The long-standing cool-
ingwarming, Milankovi cycles will be overwhelmed by the
effects of the brightening Sun. A steadily brightening Sun will
raise temperatures faster than a declining level of carbon dioxide
can balance it: the planet will then steadily warm. This process
will begin somewhere in the next couple of hundred million years.
For other examples of how the planet can really (and ulti-
mately) irrevocably change in response to changes in the concen-
tration of greenhouse gases and the intensity of radiation the Earth
receives from the Sun, lets look further forwards and backwards
in time.
Visions of Hell: Terrestrial Snowballs

and Fireballs
As continental drift continues for the next several hundred mil-
lion years at least, the current pattern of drought and deluge
will continue until the oceans finally boil away. But before we
contemplate the rather torrid death of planet Earth, we must turn
the clock back 750 million years to when planet Earth was dis-
appearing under a sea of ice. This intriguing phase of terrestrial
climate saw the planet utterly transformed in its appearance, but
also saw the greatest wave of evolutionary innovation that would
ultimately lead to us. The snowball Earth marked both the planets
greatest (identified) freeze and a critical role for a carbon dioxide as
the driver of its eventual release.
Snowballs and Slushballs
During the period from 800700 million years ago the planet
appears to have done something unpleasant to its resident life
forms. There were significant changes to the abundance of the
two stable isotopes of carbon: carbon-12 and 13 during this period
which imply photosynthesis took something of a knock. Plants
prefer carbon-12-carbon dioxide over the rarer but still relatively
common carbon-13 (about 1 % of the abundance of carbon-12).
During the periods of apparent freeze, the level of carbon-13 fell by
14 parts per thousandnot a large figure in itself, but one which
is extremely significant for life on Earth. Such a change implies
one of two things: either the planet suddenly released a very large
amount of carbon-12 from hidden stores, as in the Eocene hyper-
thermal (earlier in this chapter) or photosynthesis largely ground
to a halt and little carbon-12 was being taken out of the atmo-
sphere. These excursions in the abundance of carbon-13 were the
largest in the preceding 1.2 billion years: this event, whatever it
was, had to be severe.
Moreover, during this interval, 750 million years ago, the
planet began to form deposits called banded-iron formations. These
iron-rich rocks form when oxygen concentrations in the oceans
are low. Iron sulfide becomes more common and the iron remains
in a less oxidized state. When oxygen periodically rises, these iron
compounds effectively rust, becoming more oxidized. Such rust
is less soluble in water than the less oxidized forms of the metal
and they rain out on the ocean floor, forming alternating bands of
rust. On the Earth, most banded iron formations were produced
around 2.45 billion years ago when the planets oceans and atmo-
sphere began to fill with oxygen. In the intervening years, oxygen
concentrations were too high to allow much iron to build up in the
oceans. Yet, 750 million years ago, something happened that made
the oceans more anoxic, or oxygen-poor. Life, it seemed, was in
trouble and the evidence was in the rocks. The only way by which
photosynthesis could so extensively fail would be if the oceans
froze, temperatures became too low for the process to happen, and
the availability of light to the surface ocean was restricted. An
extensive glaciation, in which much of the Earths oceans were
frozen tight, would do the job.
The idea that the entire planet could freeze over might seem
rather fanciful, however a substantial, if still controversial body of
evidence suggests that this happened several times during the late
Precambrian, and possibly much earlier, as well. The idea that the
planet could freeze over developed over the last half century from
a mixture of diverse observations that had implied glacial condi-
tions existed near to the equator.
The original proposal goes back to the work of Sir Douglas
Mawson in the first half of the twentieth century. Mawson had
examined glacial deposits in southern Australia and assumed,
without knowledge of plate tectonics, that Australia had been gla-
ciated at its current latitude. It is now understood that the geo-
logical formations Mawson described were formed when Australia
lay further to the south than it does at present. Later, in 1964, the
idea of a global glaciation was re-examined by W. Brian Harland.
Harland identified Proterozoic deposits in Svalbard and Greenland
that also implied that glaciation had occurred while these regions
lay within the equatorial region. Harland identified interwoven
glacial and carbonate rock formations that implied alternating
periods of deep freeze and deep heat. However, Harland lacked a
model to explain how these alternating periods of climate could
have occurred. It then fell to the Russian climatologist Mikhail
Budyko to develop a climate model whereby ice sheets would
grow until the planet was largely covered. His positive feedback
loop sent ice over enough of the planet that the reflectivity, or
albedo, of the planet rose to the point at which it reflected enough
energy to cause further cooling.
Budyko concluded that although the planet could theoreti-
cally cool by such a mechanism, it couldnt possibly have hap-
pened. This was because Budyko had no means of envisioning a
reversal: given his premise, the planet would cool, freeze over and
remain so indefinitely.
In the 1990s Joseph Kirschvink resurrected and extended
the idea further, coining the term Snowball Earth. But how
did planet earth freeze over if it happened? Kirschvink proposed
that around 750 million years ago the supercontinent Rodina lay
close to the equator, but extend Pole-ward along its southern flank
(Fig. 3.9). Such a large area of land would cool effectively during
the southern winter and serve as a repository for snow and ice.
FIG. 3.9 Possible global wind patterns for the latest Precambrian period
around the supercontinent Rodina. Like today westerlies dominate the
mid to high latitudes with easterly trade winds dominating the tropics.
The presence of a mountain chain down the middle of this belt would
encourage the formation of a warm tropical low pressure over the middle
of this supercontinent. With a faster spin (21.5 h day) the largely unim-
peded westerlies flowing around the northern Polar Region would be
stronger than those seen around Antarctica today (the Roaring Forties).
An ice cap dominates the southernmost portion of the continent much
as it does today. Underlying maps courtesy of: C. Scotese
Around 750 million years ago, Rodina began to fragment as pieces

that would ultimately form Asia, Europe and North America broke
away on their journey towards the north. As the break up ensued,
these continental fragments began subducting increasing amounts
of ocean crust. As mountains rose around their flanks, enhanced
weathering set about lowering the concentration of carbon diox-
ide. Meanwhile, the relatively weak Sun (around 94 % its current
luminosity) conspired with the lowering atmospheric concentra-
tions of carbon dioxide to set the global temperatures tumbling.
Temperatures were on the fall and increasing snowfall meant
that the land became even more reflective. With less heat absorbed
by the Earth, the planet cooled further and the process accelerated
in a spiraling period of positive feedback: enhanced snowfall meant
more ice; more ice meant lowering temperatures and lowering
temperatures meant more snowfall at increasingly low latitudes.
This was the planetary freeze of Mikhail Budyko. Estimates sug-

gest that these periods of glaciation lasted between 4 and 30 mil-
lion years and were repeated.
So if the planet froze, why did the cooler climate not persist?
The key to solving this puzzle is our favorite greenhouse gas, car-
bon dioxide. When the ocean was sealed off and the land frozen,
carbon dioxide was not taken up by weatheringor by photosyn-
thesis. Moreover, carbon dioxide was being released by volcanic
activity. Consequently, the level of the gas built up in the atmo-
sphere. When it reached 10100 times its current value, the warm-
ing effect of carbon dioxide overwhelmed the cooling caused by
the presence of the global ice sheet. Within a few hundred years
(at most), the ice melted. With melting, the planetary albedo
decreased, which further accelerated warming. This time positive
feedback drove melting and warming. Estimates suggest that in
the immediate aftermath of the glaciations, carbon dioxide might
have raised the temperatures to an average of 32 Csomething
that will not be seen again for another billion years.
Evidence for the surge in carbon dioxide levels is found in
the form of the carbonate rocks (cap carbonates) that were found
in Svalbard. These carbonate deposits are formed in contempo-
rary tropical conditions, thus their presence immediately above
the glacial deposits implies that temperatures rapidly soared from
frigid to torrid over a very short interval of time. In particular, the
deposits in the Neoproterozoic are thick. This implies that a par-
ticularly large supply of carbon dioxide gas was present.
Although the evidence for a global freeze is strong it is not
infallible. For one the presence of carbonates implies that the
oceans were fairly alkaline. For if the concentration of carbon
dioxide was high, the increased acidity it would cause would
prevent carbonate rocks from forming. One way around this is
if during periods of deep freeze, the oceans filled with calcium
and magnesium ionspossibly from underwater volcanic activ-
ity. These would then react with the atmospheric carbon diox-
ide once the ice melted and the environments came back into
contact. However, this remains a problem, particularly as a
rather different mechanismthe oxidation of methane that was
released abundantlywould not only fit the observed changes in
the ratio of carbon-isotopes but could also produce the observed
carbonate rocks.
Moreover, analysis of the carbon record implies a cyclical

eventmuch like the modern day (Pleistocene) ice ages where peri-
ods of glaciation alternated with periods of comparable warmth.
Indeed, perhaps four or five periods are identifiable where the ratio
of carbon isotopes varies in a pattern suggestive of repeated clima-
tological catastrophe and recovery. This pattern implies that the
entire snowball period was driven and controlled by changes to the
planets tilt and orbital shape (a la Milankovi Cycles). Although
there is nothing wrong with this idea in principle, the problem is
whether the planet can drive a massive spike in carbon dioxide
levels only to rapidly removing the excess over a period of a couple
of hundred thousand years. As carbon dioxide has a fairly long resi-
dency time in the atmosphere, this process seems less tenable.
Finally, the evidence for a global glaciation is piecemeal.
Glacial deposits are found in some places but not others. The pres-
ence of a global freeze depends on where those pieces of conti-
nent lay at the time of the glaciation. Resolving this depends on a
good geological record and unfortunately, nature is not kind: plate
tectonics has scrambled and destroyed much of the rock record.
Moreover, much of the geological record can be interpreted in
different ways. For example, rocks called dropstones are usually
formed when glaciers carry rocks from mountain to sea (or lake)
and when the ice melts, the rocks fall to the bottom of the water.
However, these can also form in so-called debris flows: massive
currents of rock and mud that flow down-slope, underwater. Other
deposits, which can be interpreted as glacial, can also be formed
in this environment. This introduces some degree of ambiguity in
the rock record.
All of these caveats accepted, the prevailing view driven
by the greatest body of evidence still favors some form of global
freeze, but that the Earth was completely frozen is certainly in
doubt. Instead of a snowball, evidence favors more of a slush-ball.
For one, life was not completely erased: photosynthesis was main-
tained, albeit at a lower level than was seen before the glaciations
occurred. Computer models certainly struggle to produce a com-
pletely frigid Earth. Instead they cover roughly two-thirds to three
quarters of the surface in a solid sheet of ice, with the remain-
ing land more like Alaska or the Antarctic peninsula: chilly but
not rigidly frozen. There is evidence of material being rafted over
open water, probably on icebergs. This further suggests a planet

that is largely frozen, but one retaining substantial areas of open-
waterthat is open-water dotted with icebergs. In these oases life
persisted, somewhat chilled but still robustly intact.
Raindrops Keep Falling on My Head:

The Controversial Climate of the Early Earth
Lets keep journeying back in time. If we go back to the end of the
Archean era we can begin to answer some deeper questions about
the nature of our planets atmosphere. How can we find out what
the Earths climate was like 3 or ever 4 billion years ago? What we
do know is that by 3.5 billion years ago photosynthesizing bac-
teria known as Cyanobacteria were busy in the planets shallow
seas and lagoons constructing structures called stromatolites.
These living buildings of colonies of bacteria are still found today
in places like northern coastal Australia. Their structures are so
unique that they are easily recognizable when preserved in rocks.
The presence of these macroscopic fossil structures tells you that
there was abundant liquid water and that, globally, temperatures
must have been moderate enough for liquid water to persist. So
far, all seems well: except that the Sun was only around 75 % its
current luminosity. This means that the Earth should have been
rather chillyso chilly in fact that the later Proterozoic snowball
would seem positively balmy by comparison: picture present day
Mars to get some idea.
Faced with this conundrum scientists will first turn to our
favorite greenhouse gas, carbon dioxide. This isnt unreasonable
as it should have been abundantindeed it could have been a
hundred times as abundant as it is now, given its source in what
should have been a sea of volcanic activity. The problem with
this scenario is that such a high level of carbon dioxide will make
its presence known through the formation of the cap carbonates
described above. Huge amounts of carbon dioxide will dissolve in
the oceans and form carbonate rocks. Such abundance is not seen
in the geological record. Next scientists will turn to line broaden-
ing. This was mentioned previously in the context of the effect
of raising carbon dioxide concentrations and its capacity to act as
a greenhouse gas. The higher the concentration the broader the
range of wavelengths carbon dioxide can absorb infrared, because

the molecules bash into one another and alter the properties of the
molecule (specifically, the energy of its electrons).
Now, you can broaden the absorption of carbon dioxide with-
out increasing its concentration if you have a higher density of
atmospheric gas, overall. The additional, non-greenhouse gas will
boost the absorption power of whatever greenhouse gas is pres-
ent. Nitrogen can serve this purpose and it is clearly very com-
mon within the Earths atmosphere. Therefore, if nitrogen or other
gases were more common and the overall pressure of the Earths
atmosphere higher, carbon dioxide could act more effectively as a
greenhouse gas than it does now, because it will be able to absorb
heat energy over a wider range of wavelengths. This would be a
simple solution to the conundrum of the missing CO2, except that
it isnt true.
Sanjoy Som (University of Washington) and co-workers used
an ingenious method of determining ancient atmospheric density:
fossilized raindrop impact craters. Raindrops falling on fresh volca-
nic ashin this case laid down by very hot flows of materialturn
into concrete-like material which can be preserved. The research-
ers examined a group of South African volcanic rocks called the
Ventersdorp Supergroup. Within these rocks the impact craters
produced by raindrops were preserved and indicated how quickly
the drops fell to the ground. The faster the drop falls, the larger the
crater, for any given size of raindrop. The speed of descent is pro-
portional to the density of the air. The denser the air is, the lower
the raindrops terminal velocity will be. Analysis soon showed
that the size of the craters was the essentially the same as modern-
day onesunless of course Archean raindrops were much larger,
and hence more massive, than modern day ones. This meant that
the density of the Archean atmosphere was effectively the same as
todays: bang goes that idea. The result also confirmed that carbon
dioxide levels could not have been much higher than present as
this, too, would clearly have raised the density of the atmosphere.4
So, what are we left with is a warmer planet and, apparently,
no way to make it any warmer. Methane produced by anaerobic
bacteria could be the answer, but ultimately this methane would
4
While going to press, Som has published another research article in the journal Science that sug-
gests that the density of the atmosphere was actually a lot lower than it is now.
have to be at a fairly high concentration and there is, at least, lit-

tle direct evidence to support its presence. Indeed, the presence of
some minerals such as magnetite and banded iron indicates that if
there was methane present it couldnt have been very abundant. If
it were more common in the atmosphere it would have prevented
these materials from forming.
Enter the most prosaic and likely correct hypothesis that the
ancient Earth was less reflective than it is at present. If it was
darker in color than it is now then, much like black card heat-
ing faster in the Sun than silver card, then the early Earth would
absorb more energy and warm more strongly than it does at pres-
ent. To understand this idea further we need to look a little more
closely at the history of the planet.
At some undetermined time, several tens of millions of years
after the planet formed the temperature of the atmosphere fell
below the condensation point of water. A deluge would then have
begun that soon, thereafter, covered the bulk of the Earth in liquid
water. At this point the only visible land, if there was any, would
have been a scattering of volcanic islands. Modern plate tecton-
ics was likely non-existent as the super-hot mantle would need
to cool further before an organized system of convection currents
could begin. The Earths crust and topmost solid mantle, the litho-
sphere, was likely too hot and buoyant to subduct as it does now.
And, although a different system of plate tectonics wasnt far off
in the earliest and hottest Earth this wasnt quite able to operate.
Run the clock forward a couple of hundred million years and
things in the cooling mantle begin to work more like they do now.
Hot currents rising from the core generate massive waves of (likely)
underwater volcanism. These eruptions produce large undersea pla-
teau that serve as the starting points for continents. Even so, at 4.3
4.2 billion years ago most of the land is still under the dark oceans.
Now, take a step back and look at the modern Earth. The lightest
areas, those that reflect the most light, are clearly the continents
particularly those areas free from vegetation. The deserts and ice
caps are very reflective in comparison to the dark, blue oceans. 4.3
billion years ago there were no land plants to the land was barren
and thus reflective. There was so little of it that the planet as a
whole was dark in color in comparison to that of today. A darker
surface absorbs more light and hence can warm more thoroughly. A
dark early Earth, as Minik Rosing from the Nordic Center for Earth
Evolution proposed, would neatly circumvent the problem of not

having sufficient greenhouse gases to warm the planet. The lack
of living organisms is also of benefit early on. Life, in particular
plants release chemicals that form a photochemical smogclearly
visible as a haze over forests during the day. In todays atmosphere
this haze serves as a source of particles that can condense water
vapor to form clouds. Without life such a haze would not have been
present and the planet likely less cloudy. Fewer clouds would also
mean that the planet would appear darker than it does now and
would be more able to absorb energy from the Sun. A difficult prob-
lem appears to have found a simple solution. Though one should
always bear in mind H.R. Menckens adage, A complex problem
has a simple solution. And its wrong, it does appear as though the
issue of the faint early Sun paradox is probably solved in a very
prosaic manner.
Parallel Lives: The Formation of the Earth and Venus
To conclude this chapter we bring the Earths climate into close

parallel with that of our sister world and continue backwards in
time to the era in which they formed.
The Earth and Venus began their lives in very similar ways:
neither was at all pleasant. Both planets were likely bequeathed
with approximately the same abundance of light elements and con-
sequently would have had approximately the same amount of green-
house gases at birth. What of the earliest atmosphere? That would
have been interesting. Within the first one to two thousand years
(and before the Moon formed) the atmosphere of both the Earth and
Venus would have been supercharged by frequent large impacts.
At such high energies transient atmospheres of vaporized rock and
metal would have cloaked both planets. However, by 2000 years
or so of their formation the most frequent bombardment stopped
and the atmosphere would have cooled. From this point the hellish
brew of materials that formed the early atmosphere would have
rained out into the magma ocean underneath. Over the next two
million years the surface would have solidified. At this point the
surface temperature would still have been around 500 K (227 C)
about half that seen on modern day Venus. An increasingly dense
atmosphere of water vapor, carbon dioxide and nitrogen would have
maintained hellishly high temperatures while the surface cooled

slowly underneath. The atmospheric pressure would also have been
similar to that on present day Venus: around 100 times the pressure
of the Earths current atmosphere at the Earths surface.
At some point, perhaps 60 million years after the Earth took
shape a Mars-Sized object, side-swiped the Earth. Likely eject-
ing the earliest atmosphere, this giant impact blasted most of the
Earths surface off into space along with a sizable amount of man-
tle. In the collision the Earths greater gravity pulled the denser
portions of the impactorits metallic core in particularback to
the Earths tormented surface. A thick soup of rocky debris formed
a cloud then ring around the Earth. Within 1000 years of the catas-
trophe the Moon began to take shape. During the time the Earth
would have re-gained its atmosphere of gassified rock, at least
transiently. The temperature of this atmosphere would have once
again hovered somewhere above 1800 K (1523 C). The Earth also
received a kick to its spin and a new tilt on the universe. The stage
was set for the formation of todays climate systems. Hovering
only 25,000 km over the Earth, the rapidly orbiting Moon gener-
ated tides up to 30 m high.5
Modeling by Kevin Zahnle (NASA Ames Research Center)
indicates that once againwithin a few hundred yearsthis rocky
atmosphere would have cooled and rained out to the surface. Left
behind was a very dense, hot atmosphere dominated by water and
carbon dioxide. The surface pressure would have dwarfed the con-
temporary Venus and returned to the pre-impact value of more
than 100 bars of pressure from carbon dioxide and several hundred
bars of pressure from steam. This Venusian atmosphere was, once
again, kept hot by heat radiated upwards from the magma ocean
underneath. Additional heat was also being pumped into it (and the
rocky sea beneath) by the infant Moons immense tides. Such tidal
heating still pumps a whopping 10 trillion, trillion Watts of power
into the Earths oceans today and drives the tides. Tidal heating
would have been orders of magnitude greater when the Moon first
formed and hovered over the Earths atmosphere like some vast
5
Finding any denitive value for the height of tides seems to be more trouble than its worth! Values
range from 1 to several tens of meters with few references giving anything conclusive. Take 30 m
as speculative
bloody orb. It is interesting to note, that at this time Venus would

have likely been the more habitable of the two worlds, with a cool-
ing atmosphere and a solid surface.
This primordial vision of hell was unstable. The Sun couldnt
supply sufficient energy to keep the atmosphere this warm, and
by 20 million years into the Earths history, the surface cooled and
solidified. This cut off the supply of heat energy that was required
to sustain the global greenhouse. With the energy input per second
reduced from over 140 W per square meter to a value more similar
to that of today (0.5 W per square meter) the water vapor cooled
to the point at which it could condense as rain. Now, dont go
getting all dreamy-eyed about the Earth suddenly becoming habit-
able. The loss of the water from the air still left over 100 bars of
carbon dioxide which kept the temperature high. Thus, despite
an increasingly deep ocean, the atmospheric temperature was still
around 500 K (227 C). This is far above that at which water boils
at the Earths surface today. This was possible because the atmo-
spheric pressure was that much higher, and with it the boiling
point of water.
This final Hadean greenhouse was also unstable. Without
a constant input of energy from the surface (or the Sun) heat
was gradually lost to outer space. All the while, carbon dioxide
was dissolving in the oceans and precipitating out as carbonate-
rich rocks on the ocean floor. This appears to have been true
for Venus, although as Chap. 5 describes, this is still controver-
sial. If there was no subduction only around 10 bars worth of
carbon dioxide could be removed from the Earths atmosphere
and dumped into the interior at any one time. However, once
the interior was cool enough to allow some form of subduction,
greater masses of carbon dioxide could be dumped into the man-
tle. The main issue is how long this took. Current estimates put
the loss of the carbon dioxide-rich atmosphere at some point
between 10 and 100 million years after the Moon formed. The
Earth would still have had an atmosphere far richer in carbon
dioxide than today (a few percent of its total mass rather than
the current fraction of one). Yet, even this relatively low level of
carbon dioxide couldnt be maintained and as the atmospheric
temperature continued to fall carbon dioxide levels fell to levels
comparable with that of today. With this final decline in carbon
dioxide levels, down went the temperature from more than

200 C to around 1015 C. Thus, by 4.4 billion years ago the
atmosphere was rather similar to that found around usexcept
that it was effectively oxygen-free. The dark ocean compensated
for low levels of greenhouse gases and ensured that the planet
remained largely ice-free.
Venus being closer to the Sun would have been warmer but
there is a good chance that it, too, had extensive oceans and rela-
tively benign temperatures at 4.4 billion years ago. Current views
of this world suggest that it lay somewhere between having a hot
(50100 C) ocean or one that was cooler and more similar to the
Earths. Were the ocean hot, it probably didnt last more than a
few tens of millions of years before evaporating. However, a cool
ocean could have been stable for hundreds of millions or perhaps
even a couple of billion years. The precise figure depends on how
much cloud cover Venus had and what kind of clouds were most
prevalent. A lot of cumuliform clouds would have kept it cooler
for longer by reflecting solar radiation back to space.
Returning to the Earth, things werent quite as pleasant as
the vision of a cool, planet wide ocean first suggests. Punctuating
this broad sweep of time, through to the beginning of the Archean,
600 million years after the planet formed, were a succession of
massive bombardments. Although there appears to have been a
spike in the rate of such collisions around 394.2 billion years
ago (the Late Heavy Bombardment), the entire period is likely to
have been rather violent. Impactorsmostly asteroids initially,
but later massive comets, would have brought objects up to a few
hundred kilometers across to the Earth at tens of thousands of
kilometers per hour. Although such monsters were likely rare,
compared to hundreds of objects tens of kilometers in size, these
would still have packed enough energy to boil substantial portions
of the oceanand penetrate the crust through to the mantle. A
single 300 km-wide projectile would vaporize a significant depth
of ocean directly and generate a thick water-filled greenhouse
above what remained. Anything above 500 km wide would have
boiled the ocean away in its entirety. Smaller objects obviously
generate less steam on impact, but would still generate spikes in
the Earths atmospheric temperature as the strength of the moist
greenhouse temporarily stepped up.
Frightening though such impacts would have been, their

effects would have been transient. Even a brutal 300 km wide
impactor (something that would leave a scar up to 3000 km wide
on the ocean floor) would only par-boil the surface for 10,000 years
or so: certainly long by our standards, but nothing compared to the
length of the geological period. Water would rain out of the ini-
tially hot (80 C) atmosphere at a terrific rateperhaps equivalent
to a global depth of one meter per year.
However brutal all this sounds, there is sturdy geological evi-
dence that the Earth was relatively balmy by as little as 100 million
years after the Moon formed. Tiny crystals of zirconium silicate
(zircons) were produced in sufficient abundance that they eventu-
ally became preserved in 3.3 billion year old rocks in Australia. The
Jack Hills zircons tell of a time when the Earth was able to form
granite. Granite only forms when there is sufficient liquid water in
the mantle to allow the separation of the rocks minerals (includ-
ing the zircons) from the hot brew. As the mantle is relatively dry,
the major source of water comes from the subduction of oceanic
crust. Moreover, the trace elements in these zircons indicate that
the granite formed in part from melted clay-like minerals and these
only form in the presence of liquid water. Although the manner in
which this crust was delivered to the mantle is a matter for consid-
erable and quite contentious debate, it must have happened in some
form so that enough water and other minerals were present to broil
the granite out and form early continental crust. The implication is,
that even as little as 100 million years after the planet formed, the
Earth hosted at least some aerial land masses and experienced ero-
sion, which formed sediments like those of today.
By contrast the crust of Venus is rather telling. Most of the
crust is basaltic and we know this from two sources: the mul-
tiple Soviet Venera landers of the 1970s and early 1980s; and more
recently from Venus Express. Radar and other measurements of
the surface show a primarily dark, flat rocky bowl and the likeli-
est volcanic rock that would produce this is basalt. Basalt is abun-
dant on all of the rocky planets, the asteroids, and probably Io, too.
Indeed, basalt may be the most common extrusive volcanic rock
in the universe. Granite needs water to form and the only likely
areas for granite on Venus are two prominent highlands, Ishtar
Terra and Aphrodite Terra. If we assume that these are continents
then we can make some useful inferences about the climate under
which they formed and how long it lasted. On Venus although
most of the surface lies within 1 km of the lowest points around
14 % can be classified as highland. Again, if we assume that these
regions are all equivalent to terrestrial continents (or the inert
remains of associated regions called island arcs) then we can infer
how long this took to form. If plate tectonics was about as active
on Venus as it was on the Earthand their similar mass, grav-
ity and composition would make this a reasonable assumption
then it is possible to infer that plate tectonics likely persisted for
around 5001000 million years. On Earth the preserved Archean
continental crust is limited in extent and of comparable global
coverage to the highland regions of Venus. There is certainly a
lot of hand waving going on here, but its not unreasonable. If, as
on Earth oceans supplied the water needed to make the granite
we can assume that there were oceans for at least this time. On
Earth, within ten million years of subduction ceasing volcanism
and the production of granite also stops: by this point the mantle
above the subducting plate has dried out and no more granite can
be produced. The highland crust of Venus, therefore, forms a kind
of clock that can be used to infer the presence and persistence of
its oceans.
When the late heavy bombardment ceased 3.9 billion years
ago, the Earth was able to settle down into a more sedate period.
During the next 300 million years it was able to convert organic
slime into recognizable bacterial fossils and the story of life took
off. Yet, this transition from hell to heaven was not quite that
straightforward. For one the proliferating life conspired with the
steady growth of continents to remove carbon dioxide from the
atmosphere. With the Sun still faint, planet Earth was vulnerable
to global freezes. The earliest detectable freezethe Huronian Ice
Ageoccurred just over 2.2 billion years ago. Details are scant as
little continental rock remains to produce a legible tale.
Shortly before the Huronian freeze, the Earths atmosphere
began its dramatic final transition to its modern composition.
Sometime around 3.2 billion years ago cyanobacteria learnt to
process water into free oxygen gas and usable hydrogen. The latter
was already used to convert carbon dioxide into glucose, but until
this time the bacteria were unable to harvest sufficient energy
to split water. Instead they used the abundant hydrogen sulfide.

When water became usable, suddenly one of natures most reac-
tive gases, oxygen, was free to trample all over the planets geol-
ogy and pre-existing life. For around 800 million years oxygen was
soaked up by hydrogen sulfide and metals in the oceans and the
atmosphere saw little of this noxious element. There came a time
when those bacteria that were able to produce oxygen broke free of
whatever geological chains were present and took over the oceans.
Iron sulfide was converted to iron oxide rust and rained out onto
the ocean floors. Billions of years later we would dig this up to fer-
ment our industrial revolution. Most bacterial life in the oceans
would have been quickly exterminated, with what remained,
forced to adapt to this new poison or retreat to the muddy depths
of the ocean floors. Oxygens rise undoubtedly eliminated huge
swathes of life on Earth but in doing so, cleared the way for the rise
of more complex multicellular life.
At some point during the terrestrial Archean or early
Proterozoic, Venus and the Earth diverged. While the Earth was
vulnerable to extreme cooling, Venus was moving in the opposite
direction. The ever-brightening Sun would have taken Venus past
a tipping point where its warming oceans would have first given
up much of their store of carbon dioxide then their bulk water
through evaporation. In concert, these greenhouse gases would
have begun to rapidly warm the planet. Somewhere between 3.5
and 1 billion years ago the last drop of Venuss oceans drifted
into the atmosphere. Most likely this would have been sooner
rather than later. With its moderating oceans gone and an atmo-
sphere stuffed to the hilt with greenhouse gas, Venus began its
thermic ascent to its current position as the hottest body in the
Solar System, other than the Sun. The once twins were no more
identical.
The history of the Earths atmosphere has shaped the evolu-
tion of life as well as the deeper machinations of the planet. Our
Goldilocks world owes its benign quality to the interaction of the
deep interior with the wet and now oxidized surface. Liquid water
remains because the pace of geological change broadly matches the
opposing change in the luminosity of the Sun. Moreover, changes
to the surface, brought about by plate tectonics, have facilitated the
preservation of liquid water. An early Earth was deficient in terra
firma was darker and absorbed more sunlight keeping it warm.

As the land rose and carbon dioxide was drained from the air, the
Sun grew brighter at just the right pace. Plate tectonics probably
also maintains the planets geodynamo by helping stir the pot of
molten iron in its core. This is a two-way street as plate tecton-
icsat least in its present guiseis likely only possible because
we have a wet, cold surface. Remove the water and not only would
the plates gum up, because the mantle would be more viscous, but
the steady addition of continental crust would end as well. The
planetary interior and exterior are fundamentally entwined and it
is this that maintains the habitability of the Earth.
This state of affairs will not last forever. Although the Sun
has been a benign accomplice in the evolution of the Earth and its
atmosphere, this relationship will break down in the not too distant
future. It is to this eventual divorce, and its lethal consequences,
that this chapter now fatefully turns; but not before we encounter
one lastand possibly rather longfreeze.
The Last Icehouse

The current planetary climate is controlled in large part by the
greenhouse gas carbon dioxide. Although exact measurements are
uncertain, the amount of this gas may have been 100 times greater
when the Earth formed than it is now. Over geological time,
despite blips like the Permian and Eocene, the overall concentra-
tion of this gas in our atmosphere has been in decline. Although
photosynthesis has played a significant role in enhancing the rate
of decline, the primary driver has been a series of reactions called
the carbonate-silicate cycle.
In essence the carbonatesilicate cycle is a process driven by
two competing forces: erosion and burial versus volcanism and
uplift. When silicate rich rocks are exposed to carbon dioxide, par-
ticularly in the presence of water, chemical reactions convert the
silicate rocks to carbonates and free silicon dioxide. This lowers
the amount of carbon dioxide in the air. The greater the rate ero-
sion occurs, the greater the rate of decline in the amount of car-
bon dioxide in the atmosphere. This, as you may recall, was the
critical link between the rise in the Himalayas and the decline in
atmospheric carbon dioxide levels that would ultimately take the

planet into the most recent Ice Age (Chap. 2). However, the Earth
is a dynamic planet, where the crust is buckled, twisted and in
places drawn down into the mantle.
Where the temperatures exceed several hundred degrees, the
carbonate-rich rocks decompose: the process is grandly known as
thermal decomposition. Primarily, on Earth, the process happens
where oceanic crust is subducted. But wherever temperatures are
suitably high, carbonate rocks will cook, decompose and release
their store of carbon dioxide.
Over time, geological activity has declined on Earth, as the
planets interior has cooled down. Moreover, more and more of the
planet has become covered in continental rocks, which are too light
to subduct. Consequently, when carbonates become washed onto
their margins, it is more difficult for these rocks to fall into the ocean
depths where they can then be subducted. As a result more and more
carbon dioxide is being lost from the system, forever trapped as lime-
stone and other carbonates on the continental margins.
Thus, human activities aside, carbon dioxide is becoming (or
would have become) a very insignificant player in the global green-
house over the next several tens of millions of years.
The Rise of Amasia

In 250 million years time, the Earths continents will have reassem-
bled into a twisted mirror of the Permian Pangaea. Although this
may not be the planets final resting place for the continents vast
bulk, their assemblage will coincide with a turning point in car-
bon dioxide levels. The Sun will be around 23 % more luminous
than it is nownoticeable to our eyesand, consequently, global
temperatures will have risen from an average of around 15.5 C
now to around 20 C. Carbon dioxide levels will have dropped
from 400 ppm now to around 50 ppm. At this critical point, broad-
leaved plants will no longer be able to collect sufficient carbon
dioxide to successfully photosynthesize. Those that cant adapt
will become extinct. The grasses, on the other hand will continue
unaffected as they can collect sufficient carbon dioxide down to
levels of approximately 10 ppm. Perhaps the giant forests of today
will be replaced by thick stands of bamboo. Regardless, existing

floras will begin a final and irreversible decline as carbon dioxide
levels become critical.
What will conditions be like in this future Pangaea, or
Amasia as some have christened it? In general, the popular lit-
erature paints a grim picture resembling the end of the Permian
with hot deserts dominating the interior and a much more stormy
sea battering its shores. In part this may be true but it is, like
it was for the Permian Pangaea, unlikely to be the whole story.
For one, with an extra 5 C or thereabouts in global temperatures,
much of the planets ice is likely to have melted, but probably not
all. Remember an extra 5 C is something we might experience in
100200 years with continued consumption of fossil fuels.
Amasia is most likely to reside over the planets North Pole
(Fig. 3.10). The most recent models show the future Pangaea will
form through the closure of the Arctic Ocean and part of the
North Atlantic basins. This will create a large landmasspossi-
bly a rather mountainous one over the North Polar Region. The
North Pole, itself, is likely to lie roughly along the eastern edge
of Greenland. Land, being more reflective than water and with a
much lower heat capacity (Chap. 2) will chill rather nicely. Thus,
even with globally higher temperatures, there is a good prospect
that some form of polar cap will be present in the north, particu-
larly as much of this region will be fairly mountainous following
the continental collision that closed the basin.
Likely extending towards the south will be the remnant
Atlantic Ocean. The most recent models have this looking very
much like the Tethys, but turned 90 on its side (Fig. 3.10). This
pattern, if it is produced should facilitate the movement of cold
waters from the North Polar Region southwards. Thus, unlike the
Permian, the oceans could still be thoroughly mixed, if sufficient
ice is retained in the north. Indeed, a Tethyan-like set-up would
have subduction zones extending along one or both of the Atlantic
margins, generating yet more mountain building. All of these high
terrains would readily allow glaciation along the edges of part of
the supercontinent. Certainly, this is speculative but definitely not
unreasonable. We can only really know by surviving and observ-
ing this future world directly, which is rather out-with our current
technological level
Australia
Eurasia
North
America North
Pole
Africa
South
America
45oN
FIG. 3.10 A plausible view of the continents looking down on the North
Pole, 250 million years, hence. Darker browns indicate mountains pro-
duced by the continental collisions that formed Amasia; lighter brown,
the somewhat older belt produced by the collision of Australia with Indo-
china. A plausible ice cap is shown in translucent lighter blue overlying
the northern mountains and extending somewhat over Greenland to the
North Atlantic. Further mountains along the western Atlantic seaboard
represent new belts formed by likely subduction of the Atlantic ocean
crust under the Americas
Meanwhile, the situation of the South Polar Region is a bit more

complicated. The most recent models do not clearly reveal where
Antarctica will lie. The island continent could equally move north
and close the Indian Ocean, leaving a body of open water over the
South Pole. Alternatively, other models also suggest that Antarctica
stays roughly where it is now, stranded. However, given that by 250
million years from now, Antarctica would be surrounded by a vast
expanse of old, cold ocean crust, one would have thought subduc-
tion would become extensive around at least some of its shores,
causing it to move northwards to warmer climes. If it does, and
the Indian Ocean becomes its target, then the Indian Ocean may be
reduced to a relatively small inland ocean, while Antarctica finally
thaws out after 280 million years in the freezer.
Given its isolation from the rest of the planets oceans the
interior ocean, if it forms, could well become anoxic because its
warm waters will be unable to mix with the greater body of cir-
culating water on the planet. Regardless, the Indian Ocean or its
descendent will be a source of moisture to the supercontinents
interior, along with the remnant Atlantic Ocean. Moreover, in a
warmer world, with less water likely stored as ice, sea levels will
likely be higher than present, which will leave open the prospect of
much of the low-lying continental interior being flooded. Exactly
where this will be depends on many factors which are, once again,
hard to predict.
Given this set-up, what will the prospects be for future life?
On the face of it probably rather good. Species will have to adapt
to a radical change in plant life, but the supercontinent is unlikely
to be a bland desert, given the predicted arrangement of the conti-
nental masses. Thus there should be plenty of different ecological
niches for life to occupy.
One of the more dramatic assumptions about the climate of
Amasia is its propensity for ravaging by tropical storms. Many
published climate models for Amasia assume the greater heat
from the Sun will drive the formation of proportionately more
brutal tropical storms than are found on Earth today. In real-
ity this depends on a lot of unknowns and assumptions. Firstly,
although the planet will be warmer overall, it is the distribution of
heatand in particular the warmth in the oceansthat is critical.
Future Earth might well generate more tropical storms rather than
more brutal ones. Or the layout of the continents or of interven-
ing cold ocean currents might mean that few access the shores of
the future landmass. We can still get a few clues about the planets
likely tropical targets. Looking at the likely geography of Amasia
one can speculate as to which areas would be prone to the effects
of tropical storms. The least-favored (most heavily afflicted by
tropical storms) would be the western shores of Amasiaareas
that are now Eastern Africa, Eastern South America and, further
east, southwestern Australia: these will all likely face the future
Panthalassa. A good stretch of warn water is just what a tropical
storm needs to develop and the set up of the continents will facili-
tate a thorough battering of these parts of the globe. The Northern
oceans are largely subdued in extent, although the Atlantic basin
might spawn a smaller share of tropical storms, with these affect-
ing what is now Northern Canada and northern-most South
America (the area that is currently Venezuela).
Will Amasia be the last supercontinent? Again, thats hard to
predict but given the secular cooling of our planet, the continents
should eventually cease wandering sometime between 250 million
and 800 million years into the future. Once the mantle becomes
too cold and viscous to allow subduction the internal works of the
planet will begin to gum up. The timing of this phase is important
as when subduction ceasesor becomes intermittent, the return of
carbon dioxide to the atmosphere will drop further. Moreover, less
water will cycle into and out of the mantle and that may have pro-
found consequences for the way the atmosphere is maintained and
the supply of nutrients to organisms. Again, this is speculation.
Looking beyond the bounds of Amasia and into the more
distant future we will begin to see the futures of the Earth and
Venus converge once more. Evan as Venus and the Earth began
life in a very similar state, but then diverged, the ever brighten-
ing Sun will bring the climate of these two worlds back together
once more. Before we examine the fate of the Earth and Venus
in Chap. 5, we will branch off into more speculative territories.
Chapter 4 will look at the weird and wonderful atmospheric phe-
nomena on Earth, but we close this chapter with a view to the
distant future of the Earth. In particular we ask: can humanity or
its descendents hold back the tide and keep the Earth habitable
until the Sun consumes us?
Geo-Engineering the Future

Life on Earth, in terms of its diversity, has likely peakedindeed,
it probably peaked during the Carboniferous, around 300 million
years ago. Although life will continue to evolve to fit its new sur-
roundings, over the ensuing billion years, or so, the trend for life
on Earth is irrevocably downwards. That is if we let nature take its
course and let the planet heat up.
Is there another way for the Earthbeyond the next few million
years to at least until the Sun becomes a red giant? Could this idea
be extended to look for intelligent life elsewhere in the universe?
Think about it: conditions for life on Earth may become intolerable
once the Sun is a mere 10 % brighter than it is at present. If you
could block out that extra 10 % and keep the shield up for as long
as you required it, then the planet could remain positively balmy,
even as the other planets began to overheat. This could be main-
tained right through the entire main sequence lifetime of the Sun.
There are two obvious possibilities for shielding, neither cheap and
neither cost-free in terms of the environment, but compared with
the alternative, a costly move elsewhere in the solar system, then
a spot of geo-engineering might just come in handy.
The first option is to put sulfates into the atmosphere (or
something similarly reflective) to block out the radiation from
reaching the planets surface. That might work, but you would
need to constantly replenish the sulfate or other atmospheric
shield as they would tend to wash out or precipitate out of the
atmosphere. That might get tricky. Sulfates also alter the distribu-
tion of rainfall, so you would need to plan ahead where your crops
grow and people live.
Likely more expensive, but a better long-term option would
be to physically build a reflective mesh around the planet.
Undoubtedly something of an engineering nightmareand hardly
cheapsuch a feat would be possible with a bit of asteroid min-
ing and space-based construction and it might be cheaper than
building a fleet of exodus craft to take you to the nearest habitable
world. Over time, such a shield would need repair and enhance-
ment, but it could provide a long-term strategy to keep the planet
habitable. With less sunlight now reaching the surface, solar fur-
naces could be used to cook carbon dioxide out of rocks to replen-
ish the atmosphere and maintain plant life.
If we wanted to look for a truly advanced civilization else-
where in the cosmos we might want to look for the signature of
such a planetary shield. Depending on the planets composition we
could search for enhanced reflectivity (an unusually high albedo)
for a planet that lay just inside the orbit at which one might expect
a runaway greenhouse to commence. Now, Venus already has this
and no one would suggest Venus harbors highly intelligent life.
However, something a bit more subtleand with a chemistry not

dependent on sulfuric acidmight just imply, if not definitively
prove, intelligence lay behind the high albedo. Markers would be
an oxygen-rich atmosphere implying photosynthetic life and a
spectrum compatible with metallic or other reflective shielding.
Free metals are not commonly found high in planetary atmo-
spheres.
Conclusions: The Big Picture

In this chapter we have built up a broader picture of the Earths
climate system, importantly illustrating how the planets climate
has changed over evolutionary time. Although there have been
some monumental swings the planets in-built regulatory systems
have brought the climate as a whole back into line with habit-
ability repeatedly. The underlying players are the Sun and carbon
dioxide.
While carbon dioxide is by no means intelligent, its ability to
respond to changes in the amount of radiation the planet receives
from the Sun is crucial to the well-being of the planet. For most
of the planets history carbon dioxide has been the regulator of
planetary temperature. It was only in the last 34 million years that
the level of carbon dioxide fell so low that the underlying astro-
nomical cycles began to show their effect. Previously, these cycles
gave small boosts to the global temperature but never took it low
enough for global glaciations to be an issue. From here until around
1800 A.D. the level of carbon dioxide was sufficiently low that it
then responded to the change in temperature. This meant that as
temperatures fell, the hydrological cycle slowed and with it the
rate of erosion. This in turn allowed more carbon dioxide to build
up and persist in the atmosphere, which then warmed the planet
more. When the Sun was brighter, as a result of both changes to the
Earths orbit and angle of tilt (the Milankovi Cycles) or to its secu-
lar brightening with age, the rate of erosion stepped up and more
carbon dioxide was lost from the atmosphere. This in-built ther-
mostat has operated successfully for 4.5 billion years until human
activity uncoupled the gas pedal from the clutch and the brake.
Indeed, the accelerating rate of melting of the northern and

southern caps is unprecedented in geological history. Even in
the closest match, the Eocene, the rate of temperature increase
was roughly ten times slower than the currently observed rate.
No natural cycle in recorded geological history (barring aster-
oid impacts and a Yellowstone-style super-volcano eruption)
has changed the temperature and the mass of carbon dioxide so
quickly. Even the relatively speedy Milankovi cycles operate
on longer timescales measured in tens of thousands of years: not
decades.
Putting this in context, if we assume that we are not alone
in the Universe, it is likely that other intelligent species face
the same issues elsewhere in the cosmos, assuming that all
life is governed by the same underlying laws. Species have to
fit within the broader canvas created by the geochemistry and
geophysics of their home world. There are going to be many
themes that transfer from the Earth to other celestial bodies as
clearly the underlying astrophysics is universal. Thus, if we find
a planet with roughly the same mass and composition as our
world orbiting a G-class main sequence star at a distance of 150
million kilometers, expect it to look rather like the Earth, if not
as it is now than as it once was or will be. However, dont for-
get that this exoplanet might not look like the Earth does now,
simply because it shares our physical location. Our planet has
changed significantly and with it the advent of life has radically
altered the planetary atmosphere. The temporal dimension is as
important as the spatial dimensions when we think about habit-
ability or simply the gross features of the planets atmosphere.
Were the Earth lifeless, there would be little free oxygen and
the atmosphere likely more noxious. Without human activity,
carbon dioxide levels would be on a downward trend. It is from
the perspectives laid out in chapters one and two that we can
now move on to examine the workings of the atmospheres of the
other planets that we know about. This is a growing family and
thus far few are clones of the Earth. With improving technol-
ogy and our ability to resolve greater and greater details, expect
a clone or three of the Earth to pop up soon, providing another
planet on which to observe climate.
References
1. Zahnle, K., Arndt, N., Cockell, C., Halliday, A., Nisbet, E., Selsis, F., et al. (2007). Emergence
of a habitable planet. Space Science Reviews, 129, 3578.
2. Fernando S. Paolo., Helen A. Fricker., Laurie Padman (2015). Volume loss from Antarctic ice
shelves is accelerating. http://www.sciencemag.org/content/early/2015/03/31/science.aaa0940.
3. Svensen, H., Planke, S., Malthe-Srenssen, A., Jamtveit, B., Myklebust, R., Eidem, T. R., et al.
(2004). Release of methane from a volcanic basin as a mechanism for initial Eocene global
warming. Nature, 429, 542.
4. Leconte, J., Forget, F., Charnay, B., Wordsworth, R., & Pottier, A. (2013). Increased insolation
threshold for runaway greenhouse processes on Earth-like planets. Nature, 504, 268271,
doi:10.1038/nature12827.
5. Som, S. M., Catling, D. C., Harnmeijer, J. P., Polivka, P. M., & Buick, R. (2012). Air density
2.7 billion years ago limited to less than twice modern levels by fossil raindrop imprints.
Nature, 484, 359362.
6. Rosing, M. T., Bird, D. K., Sleep, N. H., & Bjerrum, C. J. (2010). No climate paradox under the
faint early Sun. Nature, 464, 744749.
7. Mitchell, R. N., Kilian, T. M., & Evans, D. A. D. (2012). Supercontinent cycles and the calcula-
tion of absolute palaeolongitude in deep time. Nature, 482, 208211.
8. Hansen, J., Ruedy, R., Sato, M., Imhoff, M., Lawrence, W., Easterling, D., et al. (2001). A
closer look at United States and global surface temperature change. Journal of Geophysical
Research, 106, (D20) 2394723963.
9. Travis, D. J., Carleton, A. M., & Lauritsen, R. G. (2002). Climatology: Contrails reduce daily
temperature range. Nature, 418, 601.
10. Pollard, D., & DeConto, R. M. (2009). Modelling West Antarctic ice sheet growth and collapse
through the past ve million years. Nature, 458, 329333.
11. Naish, T., Powell, R., Levy, R., Wilson, G., Scherer, R., Talarico, F., et al. (2009). Obliquity-
paced Pliocene West Antarctic ice sheet oscillations. Nature, 458, 322328.
12. Kossin, J. P., Emanuel, K. A., & Vecchi, G. A. (2014). The poleward migration of the location
of tropical cyclone maximum intensity. Nature, 509, 349.
13. Michael E Mann. (2002). The Earth system: Physical and chemical dimensions of global envi-
ronmental change Volume 1, (pp. 504509). In Michael C MacCracken & Dr John S Perry
(Eds.), Encyclopedia of global environmental change. New York: John Wiley & Sons. ISBN:
0-471-97796-9 Editor-in-Chief Ted Munn.
14. Hoffman, P. F., Kaufman, A. J., Halverson, G. P., Schrag, D. P., Hoffman, P. F., Halverson,
G. P., et al. (1998). A Neoproterozoic snowball Earth. Science, 281, 13421344.
15. Peterson, T. C., Gallo, K. P., Lawrimore, J., Owen, T. W., Huang, A., & McKittrick, D. A.
(1999). Global rural temperature trends. Geophysical Research Letters, 29(3), 329332.
16. DeConto, R. M., Galeotti, S., Pagani, M., Tracy, D., Schaefer, K., Zhang, T., et al. (2012). Past
extreme warming events linked to massive carbon release from thawing permafrost. Nature,
484, 8790.
17. Wignall, P. B., & Twitchett, R. J. (1996). Ocean anoxia and the end Permian Mass extinction.
Science, 272, 11551158.
18. Daniel H. Rothmana., Gregory P. Fournierc., Katherine L. Frenchb., Eric J. Almc., Edward
A. Boyleb., Changqun Caod., & Roger E. Summons. (2014). Methanogenic burst in the end-
Permian carbon cycle. Proceedings of the National Academy of Sciences, 111 no. 15 > Daniel
H. Rothman, 54625467, doi: 0.1073/pnas.1318106111.
4. Weird Weather
Introduction
In principle this chapter concerns luminous phenomena, some of
which are natural, while others might well be tricks of the light.
Around 2000 thunderstorms are active within the Earths atmo-
sphere at any one time. Most of these occur in the afternoon when
the land or sea is at its hottest. Although almost anyone reading
this book will be familiar with thunder and lightning, thunder-
storms generate a diverse family of other phenomena, which will
be explored in the following pages.
Note that whereas the other chapters in this book are built
on reliable and verifiable evidence, this one is altogether different.
While most of what is said is verifiable, there will be some stories
which are, perhaps, more anecdote than hard science. Yet embedded
within this raft of diverse and peculiar tales are some scientific gems
that lie at the frontiers of our understanding. A few will be misiden-
tifications and will be debunked, but others will develop into truly
odd manifestations of our planets atmospheric and internal engine.
There is a splattering of new science and a lot of wonder in these
phenomena, some of which might just have counterparts on other
worlds that we are discovering today. Read on and imagine.
Strange Lights from Thunderstorms

Ball lightning is a good, as in verifiable, starting point. Ball light-
ning was known and recorded either in drawings, painting or in
stories for hundreds of years. Figure 4.1 reportedly shows a paint-
ing made by an unknown artist around 1888 of three luminous
balls traveling along a small ravine during a thunderstorm. In
those pre-UFO ages we can be pretty sure the Russia artist was
documenting a natural phenomenon and was less likely to be let-
ting his imagination wander.

DOI 10.1007/978-3-319-25679-5_4
FIG. 4.1 The first known image of ball lightning. This painting was
published in La Physique Populaire in July 1888. The artist remains
unknown but this work is believed to have been painted it following a
thunder-storm near St Petersburg, Russia. As with any artistic rendering
how much artistic license was employed is unclear, but differences in
the color of different balls are obvious
Descriptions of ball lightning vary considerably. Some

describe it as moving upwards, others downwards, sideways or
unpredictably in different directions. These can even be against
the direction of the prevailing wind. Some ball lightning appears
to move towards metal objects, or along the surfaces of walls,
while others indicate free movement in the air. Some ball light-
ning appears capable of penetrating solid objects, including glass,
Weird Weather 161
wood or masonry and in some instances appears to pursue peo-

ple. In general it is luminous white to yellow or even red in color
and adopts round or oval shapes, with some more cigar-shaped
or flanked by sparks. In most cases the balls are opaque but
a few accounts exist that suggest balls are more translucent in
nature. Some balls are associated with the smell of sulfur, ozone
or nitrogen dioxide, but this is uncommon. As you can see there
are almost as many descriptions of their appearance as there are
accounts of their presence. This diversity of these accounts has
led to a lot of confusion and suspicion that ball lightning only
existed in the imaginations of the observers. Indeed, one working
hypothesis for the phenomenon was that electromagnetic fields
in thunderstorms made people hallucinate. Although this sounds
outlandish, the effects of electromagnetic fields on our percep-
tions are well known and are suspected to lie behind many alien
abduction stories. Therefore, maybe, just maybe, thunderstorms
do cause hallucinations.
Despite the many differences in the reported incidents, there
are a few clues that might help scientists uncover the secrets of
ball lightning. Aside from the clear association with lightningin
particular cloud-to-ground lightningand thunderstorms in gen-
eral, there are a few other reports of ball lightning or something
like it being reported in quite different contexts. In World War
II submariners often reported consistent accounts of small, lumi-
nous balls of light emerging when battery banks were switched
on or off. This suggests one of two things: either the balls were
produced directly by electrical discharges that ionized the air, or
the electrical discharge caused an electromagnetic pulsea wave
of microwaves that energizes the air indirectly. Both are plausible
in thunderstorms.
Another interesting observation was made by pilot Lieutenant
Don Smith during a routine flight to Hawaii in the 1960s. His
cargo plane was flown with its radar turned up to its maximum
intensity while the plane penetrated dense fog. Two horns of
St Elmos fire appeared on the random (the radar cover). St Elmos fire
is a luminous electrical discharge seen around pointed surfaces
in the presence of strong electrical fields. St Elmos fire is now
very well understood, however in this instance, what was more
+ + + + + + + + + + + + + + + +
+ + + + + + + +
+ + + + + +
a b
- -
- -- - - - - - - - - -- - - - --
-- - - - -
-- -
--- --
- -
--
+++
+ + ++ + ++ + + + ++
FIG. 4.2 A simplified anatomy of lightning. Convection within a cumu-

lonimbus cloud allows charge to separate. Although not clear, it appears
formation of ice is key. During the formation of ice, small particles break
off and carry positive charges upwards, while negative charges remain
with the larger ice particles and water droplets lower down. During a
cloud-to-ground strike (shown) the clouds strongly negatively charged
bases induces a positive charge in the ground underneath by repelling
negatively-charged electrons. These charges are most strongly concen-
trated near the tips of objects (a). A faint, low current stepped leader
stroke advances relatively slowly towards the ground while correspond-
ing positively charged feelers (red) move upwards from any suitable
point. When one makes contact with the stepped leader the circuit is
completed and the cloud discharges a powerful electrical current to the
ground (b). Where there is a large spread of negative charge in the base,
further areas of the cloud discharge in succession, leading to a prolonged
spider-like display. Intra-cloud discharges drain positive charge from the
cloud top, along with rain or hail
interesting was the subsequent appearance of ball lightning in the

cockpit between the two horns of St Elmos fire. Clearly there was
some form of communication between the two phenomena sug-
gesting the strong electrical field that had generated the St Elmos
fire also initiated the ball lightning event.
Slightly creepier was the appearance of ball lightning in
another plane flying over Alaska near a thunderstorm in 1962. The
crew of a P3 Orion were joined, mid-flight, by a red translucent
ball of light. The crewmen reported that the large ball appeared
near the planes rear bulkhead before slowly passed down the gal-
ley of the plane towards its front. After a few moments the bas-
ket-ball sized orb reached the front section of the plane only to
Weird Weather 163
- - - + - - + - -+ - - - + -+ - - - -
+ - + + -+
+ + + + + + + + + + + + + + + ++ + + ++ + +
+ + + +
- -- - - -
-
- --- -- - - - - - -
FIG. 4.3 Positive lightning. In some instances, during particularly pow-

erful thunderstorms, cloud-to-ground lightning can originate directly
from the positively charged anvil of the thundercloud. This requires a
much higher voltage than conventional lightning as the distance the
bolt needs to travel is greater. The very strongly positively charged anvil
again induces the separation of charge on objects on the ground. How-
ever, in this case the ground becomes negatively charged. The same pro-
cess happens as before with a positive leader moving downwards then
making contact with a negative feeler. Electrons then flow up the path,
rather than down it, to cancel out the charge in the cloud. Interestingly,
in these storms the connection with the ground appears to result in fur-
ther electrical discharge between the cloud and the upper atmosphere
(light blue squiggle). This process appears to require the presence of a
negatively-charged cap above the top surface of the positively charged
anvilalthough the details are far from clear
disappear through the fuselage near the planes electrical distribu-

tion center.
Both of these observations suggest a link between ball light-
ning and strong electromagnetic fields generated naturally or
through inadvertent man-made activities. But there are other
observations that suggest a connection between this enigmatic
phenomenon and the direct contact of lightning with surfaces.
Two dramatic accounts both involving churches in renais-
sance England appear to be the first documented instances of ball
lightning in the world. The first occurred in Wells, Somerset,
England, in 1596. Here, during a violent thunderstorm, a lumi-
nous ball of light entered the church through a window during a
sermon. The ballabout the size of a soccer ballpassed along
the wall of the church near the Pulpit, before violently exploding
and scattering parishioners across the ground. The explosion was
hot enough to melt the metal parts of a nearby clock and leave
odd star-shaped burns on some parishioners clothing. Aside from
that, and structural damage to the interior of the church, no one
was harmed.
The second event, which is now the more famousand is
consequently often erroneously referred to as the first recorded
ball lightning observationwas made on October 21st 1638. This
vivid account, similar to that in Wells, tells of the Church of St
Pancras in Cornwall being struck by an almighty bolt of lightning
during a powerful thunderstorm. Incandescence then entered the
church, through the window, where 300 parishioners were attend-
ing a service. The ball of light was hot enough to burn several of
those attending the service, in some cases underneath apparently
un-marked clothing. Soon thereafter the ball was said to have vio-
lently exploded throwing many of the churchgoers to the ground,
one so violently as to smash his skull apart on impact with the
ground. In all four were killed and 60 injured; with severe damage
caused to the structure of the building. Accompanying the storm,
a small tornado appears to have subsequently done away with a
dog that ran from the explosion in the church. The evidence of this
unusual event is still visible on boards within the church.
Another tale comes from 1944 Uppsala, Sweden, where
a thunderstorm is reported to have created a ball of light that
penetrated a glass window, leaving a 5 cm-wide hole. But it was
reports from the 1960s that ultimately led to ball lightning being
taken seriously by the scientific community.
Perhaps most peculiar in ball lightning folklore is the man-
ner by which the phenomenon went from fantasy and delusion to
accepted scientific factor at least onto the record of Natures let-
ters page in a 1969 issue (Nature 224, 895). Professor R. C. Jenninson
of the Electronics Laboratories at the University of Kent was flying
in an Eastern Airlines Jet from New York to Washington when ball
lightning apparently emerged in the gangway of the plane, lazily
bouncing down the aisle past the scientists and other passengers,
before winking out. The main problem with ball lightning despite
these first-person accounts, as it is with many of the other lumi-
nous phenomena described here, is that they have proved to be
Weird Weather 165
impossible to recreate convincingly in the laboratory. This tends

to inflame skepticism in the scientific community. Various expla-
nations that are touted tend to err on the side of science fiction
and fantasy, rather than reality. Explanations of ball lightning vary
from the unlikely (antimatter meteors, mini-black holes or spin-
ning electric field vectors) to the surreal (aerogels, heavy neutrinos
or self-confined light)it seems that there are more explanations
than there are validated observations of the phenomenon. That
said a few models do appear to hold water. Were going to look at
a couple of these, one of which has some observational evidence
to support it.
The two front-runners are the silicon vapor hypothesis and
the microwave cavity hypothesis, with a third known as the light-
ning ball hypothesis a close third.
The silicon vapor model posits that when lightning hits the
ground in some instances the composition of the soil allows the
formation of a ball of combusting gases. The soil has to have the
following constituents: silica glass (sand) and a source of organic
carbon, such as dead leaves, animals, bacteria and fungi. When
the lightning strikes all of this mulch of material is vaporized and
the organic material reacts with the silicon dioxide to form car-
bon dioxide and silicon vapor. All of this happens in an instant
and the hot silicon vapor that is produced then begins to burn in
the air, reforming the silicon dioxide. In this hypothesis, it is this
chemical reaction that makes the ball lightning glow. Though this
couldnt readily explain ball lightning seen onboard aircraft, the
theory did gain some backing from observations made of the phe-
nomenon during a thunderstorm in China.
In 2012 a group of Chinese scientists from Northwest Normal
University in Lanzhou, China were observing a thunderstorm on
the Tibetan Plateau. During the storm a cloud to ground strike gen-
erated a short-lived luminous ball. Using a spectroscope attached
to their camera they managed to obtain the chemical composition
of the lightning and the accompanying ball. As expected the light-
ning showed the presence of oxygen and nitrogen from the air,
while the spectrum of the ball showed silicon, calcium and iron,
as well as carbon, oxygen and nitrogen (Fig. 4.4). The 5 m wide
ball itself had a lifetime of less than 2 s but it did manage to move
around 15 m before winking out. Interestingly, the balls behavior
FIG. 4.4 A representation of the first spectrum taken of ball lightning.

The spectrum was captured during a Chinese thunderstorm in 2014 using
a modified camera. The spectrum is of the emission typeindicating
that it is a low density, ionized gas. Chemically, the gas broadly matches
the silicon-vapor model, which suggests ball lightning is produced when
lightning hits and vaporizes silica-rich soil. Elements: Si silicon, Fe iron,
Ca calcium, N nitrogen, O oxygen; with wavelengths in nanometers
indicated. Interestingly, carbon, which is a prediction of the vaporized
silica model, is missing. Image courtesy of Wikipedia commons universal
license author: Olli Niemitalo
showed that it was influenced by nearby power lines. There was a

clear 50 Hz flicker that was associated with the oscillating electric
field in the cables. This suggests that the ball contained ionized
gas that the electric field from the power cables could influence.
Thus clearly, the ionized silicon vapor theory does work, at least
for some ball lightning. Both in the laboratory and in this instance,
in the field, these balls are very short lived. So, are there any other
means by which ball lightning could come about?
The next most favored theory involves radio or microwaves.
In the 1920s Nikola Tesla is said to have produced ball lightning
routinely while investigating sending electrical energy through the
air using radio waves. Similarly, scientists have created ball light-
ninglike phenomena in the lab using beams of microwaves. Y. H.
Ohtsuki and H. Ofuruton succeeded in making ball lightning by
interfering opposing beams of microwaves. It is not unreasonable
to think that lightning could generate pulses of microwaves that
might do the trick to create ball lightning. A derivative of this is
Weird Weather 167
the Handel Maser-Soliton theory in which the microwave source

is a maser with a diameter up to 5 km. Masers are microwave-
wavelength equivalents of optical lasers where molecules in a gas,
typically water or carbon dioxide can amplify a microwave signal
producing a coherent beam of radiation. If that sounds outlandish
bear in mind astronomers have been aware of masers hundreds,
thousands or even millions of kilometers across that lurk in places
like the atmospheres of aging red giant stars, collapsing nebulae or
the dusty torus of gas that surrounds many super massive black
holes. Thus, unlikely as it sounds, there are clear precedents for
such phenomena in nature: our atmosphere might just behave like
an enormous microwave laser on occasion, and those 5 km wide
sources might just be thunderstorms.
One of the best observational sources comes from a 1954 ball
lightning sighting, or as Domokos Tar (Swiss Federal Institute of
Technology) likes to call his observation, a lightning ball. During
a thunderstorm Tar was stuck in Budapest. Lightning struck the
ground close to where Tar was standing. Domokos Tar later wrote,
Immediately a very strong wind began to blow. I saw the bushes
bend to the right under the wind. After the lightning, it got rela-
tively dark, because of the dark clouds. In the background there
was a dark building. The bushes were still moving in the wind and
on the left wet leaves and grasses were whirling through the air.
After about 2 sec of darkness, suddenly a very beautiful sphere
with a diameter of about 30-40-cm appeared about 1.2 m above
the ground, at the same distance (about 2.5 m) from the lightning
impact point as the bush, but in the opposite direction, to the left.
The Lightning Ball (LB) was very brilliant, like a little sun. It was
spinning counter clockwise. The axis of the rotation was horizontal
to the ground and perpendicular to a straight line drawn between
the bush (trees), impact point and the ball. The LB had one or two
plumes or tails. The plume was not as brilliant as the sphere, and
reddish. Very strange was that the plume was not on a perpendicular
plane to the ground, but to the north of this plane. In other words,
the plume had a component in the YZ-plane. After a very short time
about 0.3 s the plume merged in the ball. The LB was moving at
slow, constant speed to the left on straight line mentioned. It had
absolutely sharp contours. Its brightness was constant across the
entire surface. At that moment, I no longer saw any rotation.
Tars observations were recounted at an annual symposium

on ball lightning in 2006 that provided a fascinating overview of
the phenomenon. Tars lightning ball model only holds about a
hundred or so Joules of energy so approximates a domestic light
bulb. This is shown in Fig. 4.5. The model itself seems perfectly
reasonable and Tar even suggests that such a collapsing system
might be applicable to the generation of energy by nuclear fusion.
There are a few variants of this model around involving rings of
plasma generated through lightning strikes, but Tar seems keen to
emphasize that his lightning ball model is distinct from ball light-
ning because it operates at lower energies. Whether this distinction
is real or not is as open to question as the model or models for the
observed balls that are identified. It seems obvious that the vapor-
ized silicon model could work in concert with the ring models,
or each could operate separately. Likewise the maser hypotheses
could work with either, as well. Either way ball lightning, no mat-
ter how peculiar it seems, is an established occurrence.
FIG. 4.5 Dominic Tars Lightning Ball Model. A lightning strike (a) gener-
ates a rotating charged ring of air. This ring expands outwards until on one
side it encounters an object (b). The ring breaks at the contact point with
the object then collapses along its length (c). Eventually, the ring collapses
to form a rotating ball (d). This glowing mass of air is the lightning ball.
A hybrid ball involving vaporized soil particles could also give rise to the
silicon vapor identified in the Chinese observations in Tibet (see text)
Weird Weather 169
Why do some thunderstorms appear to make ball lightning

and not others? This might come down to the nature of the storm.
Most thunderstorms produce negative lightning bolts. A precious
few, five in every ten million, lightning bolts are much grander and
likely bear a positive, rather than a negative charge (Figs. 4.2 and
4.3). Such super-bolts last around ten times longer than conven-
tional negative lightning and carry proportionately greater charge.
Its certainly possible that ball lightning only forms in those rare
situations where positive cloud-to-ground strokes occur.
In support of this is the curious tale of the 1978 Bell Island
Boom. Bearing some connection with the Seneca Guns described
later in this chapter, this loud explosion was heard over 100 km
from Bell Island, a small island near Nova Scotia. Residents of the
island reported a loud explosion (or possibly three closely spaced
explosions) and one resident reported seeing a bright flash followed
by the appearance of a luminous ball. This faded after a few sec-
onds. There was considerable damage to nearby electrical equip-
ment, suggesting a large electrical discharge had occurred. Many
interesting conspiracy stories were kicked up by this, including
Soviet woodpecker radio signals or a Tesla device that beamed
radio waves into objects to induce massive electrical discharges.
Using data from the Vela gamma ray satellites, which were used
to monitor nuclear tests, John Warren and Robert Freyman from
the Los Alamos National Laboratory (then called the Los Alamos
Scientific Laboratory) detected a superbolt over the island at about
the time the reports of the explosion or explosions came in.
Earthlights or Headlights?
Across the globe there is a bewildering array of other types of
lights besides ball lightning that appear to come and go at will in
particular locations. These other Earthlights can be nagar lights,
igniting marsh gas or earthquake lights, described later, but there
are also less well explained ones that appear when the weather
is largely fair and in non-seismically active areas. Burning marsh
gas, a mixture of phosphine and methane, rises upwards as one
would expect since it is hot and moving by convection. Swampy
areas with related decomposition of plant material give rise to
these flammable gases. The Naga fireballs on Thailands Mekong

river are light, small rapidly rising bursts which appear at the end
of the annual Buddhist lent, khansa, in October. Their recorded
history is spotty but they are observed by many each year at the
festival season, whether originating as tracers set by humans, as
phosphine gas, or as plasma orbs.
The Paulding Lights are another intriguing case. Believed to be
supernatural in origin by some in the region where they occurred,
a group of students at the Michigan Technological University
Society of Photo Optical Instrumentation Engineers (SPIE) used a
spectroscope to find the scientific explanation. How these lights
came to be understood says rather a lot about our perceptions and
what we are prepared to believe.
In order to understand the Paulding Lights you need to pic-
ture the scene. They appear in the distance along the line of an old
dirt track flanked on its right by power cable-bearing pylons. The
lights are mostly white or off white; but red lights are common,
as well. To the unaided eye the lights magically appear flicker
then disappear, only to reappear once more. There are several very
good videos on YouTube replete with people whooping and crying
at the seemingly magical appearance of the lights. With a small
pair of binoculars an observer at one end of this track will see
lights appearing before moving up or down, in part hidden by the
trees. These observations are always indistinct and taken from a
bridge overlooking the path or on that section of the path under
the bridge. On a good day many observers will gather, sometimes
even film crews, all trying to catch a glimpse. So far so good: the
Pauling Lights are very well documented and are clearly a genuine
phenomenon. The problem is what is their cause?
UFOs and psychic energy are front runners in some quar-
ters, as well as lights associated with tectonic strain (earthquake
lights) and ball lightning. However, clearly none of the latter natu-
ral phenomena fit the bill. Enter the SPIE team from Michigan
Tech University armed with a powerful spectroscopic telescope.
Training the scope towards the direction of the Paulding Lights
they could clearly see the junction of the main interstate (Highway
45) and a side-road that runs continuously with the end of the dirt
track that the lights are seen on.
Weird Weather 171
As dusk descends the team continues to spy on the junc-

tion until the famous lights appear. Again, by naked eye the same
configuration of white and occasionally red lights are seen. Some
flicker; some appear to descend a short distance then wink out.
However, under the gaze of the telescope all becomes clear: the
lights are definitely head and tail lights of cars. The single, flicker-
ing lights are resolved into pairs of headlights or less frequently
red tail lights on cars moving onto or off the interstate in an exact
spectroscopic match.
In 2005 students from the University of Texas carried out a
similar investigation with the even more famous Marfa Lights.
Once again, under the gaze of a spectroscopic telescope the Marfa
Lights are resolved into car headlights or camp fires, distributed by
refraction through the atmosphere. In both cases the lights are only
clearly visible to distant observers under specialized conditions. In
these instances a layer of warm air sits above the ground with a
cooler layer below. Such a layer readily forms in warm weather
when the sun goes down and the ground cools quickly through
radiation. Light from distant cars (up to 8 km away) is bent by the
change in the density of the air as it moves towards the observer.
In some instances this causes it to bend enough so that it is vis-
ible over longer than normal distances. What is obvious about
many of these light phenomena is that they appear to exist only
in the distance, near the horizon. They never seem to appear near
the observer. In a few documented instances observers report that
the lights appear to play tag with them, keeping pace with their
movements. This can be explained if cars are driven by law-abiding
citizens that are keeping to the speed limit and hence remain equi-
distant. The same phenomena has been observed, and explained, at
the so-called Spooklight near Hornet Missouri; again, spectroscopic
investigation showed that these were refracted car headlights.
Refraction can occur in two ways. Where the air is warmest
above the ground, but colder upon it, light rays from an object
bend downwards towards the observer. This is known as a superior
image. The converse, with warmer air near the surface is called an
inferior image and is responsible for conventional mirages that are
seen on hot days. In the case of the Paulding and Marfa lights, the
colder air at the surface has this effect leaving light rays apparently
coming from higher up in the air than they truly are. At sunset
the Suns rays are refracted through the air in a similar way, with
colder air at the surface. This allows the Sun to appear as though
it is still above the horizon when in fact it has set.
In the opposite situation, the warmest air lies at the surface.
This most commonly happens on hot summer days. Light rays are
then bent upwards through the hot overlying air. This most com-
monly forms mirages, when there appears to be water lying on a
dry surface. This water is simply a reflection (or rather a refrac-
tion) of light from the sky above.
More complex superior mirages, called Fata Morgana, can also
occur when the air has a more complex structure, but is still pre-
dominantly coldest near to the surface. Named after the Arthurian
sorceress Morgan le Fay, these were first noted in the Straights of
Messina. According to legend this wicked sorceress hung out in
the straits, luring sailors to their deaths by leading them to believe
land was near. In a typical Fata Morgana image, the original object,
which has its image refracted, is rather distorted and unrecogniz-
able.
In Australia, Professor Jack Pettigrew examined the famous
Min Min lights of Queensland. Pettigrew confirmed that the lights
there, similar in appearance and behavior to those in Marfa and
Paulding, had an origin in refraction, which in turn was caused
by changes in the temperature of the air with height. Pettigrew
observed that a great number of natural and man-made lights can
contribute to the observed Fata Morgana and that these are not
always evident in the light of day. However, these odd light phe-
nomenon are not confined to distant lights.
Fata Morgana or Something Else?

The Hessdalen lights are perhaps the most famous Earthlights on
the planet. This is certainly down to their frequency, at one time
in the early 1980s being visible 1020 times per week. They have
been the subject of prolonged scientific investigation involving an
international array of investigators.
Hessdalen is located in a broad valley around 120 km south
of Trondheim, Norway. For decades unusual lights have been
reported regularly at various locations within the valley, some-
times near to or apparently on the ground, with others higher up
Weird Weather 173
in the air. Most are white or off white in color and may appear
randomly or appear to move at high speed.
The lights associated with this small, isolated village have
become something of a national phenomenon for Norway and
are certainly the most researched Earthlights on the planet. Their
nature as of yet defies simple explanation; many appear in mid-
air and can be tracked on radar before becoming visible. Most
impressively many of these lights appear to be moving at very high
speeds, typically 10,00020,000 m per second, or over 25 million
miles per hour. Whatever the radar is detecting must be very low
in mass otherwise it would suffer spectacular drag and frictional
heating in the Earths atmosphere. A simple bundle of electric or
magnetic field lines could do this and reflection of radio waves by
aurora has been known for some time. In the late 1960s George
Millman (General Electric Corporation) wrote about the reflec-
tion of radar by magnetic fields at low latitudes in The Journal
of Geophysical Research, Space Physics.1 While these are largely
visible phenomena, radar reflections from some form of magnetic
field is certainly possible as an explanation for those phenomena
that appear to have unreasonably high velocities and are at least
initially invisible.
Other suggestions for the phenomena involve a diverse array
of phenomena. These include: natural batteries in the valley that
involve iron-rich rocks on one valley side and copper-rich rocks
on the other; piezoelectric discharges from strained rocks; even
odd combustion reactions involving the element scandium that
is somehow kicked up in clouds of dust from the valley floor.
Piezoelectricity could explain the phenomenon, but the rocks of
Hessdalen are not under geological strain, so how that process
would work in this location is unclear.
The problem with the Hessdalen Lights is the likely diverse
nature of their origin. Although many Earthlights will turn out
to be either man-made lights or perhaps mirages of distant lights,
such as those at Paulding or Marfa, some, such as those in Norway,
are likely to be genuinely novel phenomena and worthy of thor-
ough investigation. Often from such interesting science comes
something unexpected and useful.
1
Journal of Geophysical Research, Space Physics, Volume 74, Number 3 (1969) Wiley Online
Earthquake Clouds and Lights

Moving on from the genuinely odd, or somewhat suspect, phe-
nomena of Earth Lights we enter still mysterious but extremely
well documented earthquake lights (EQLs). Like Earthlights these
lay on the fringes of mainstream science for decades but their
existence is now becoming more widely accepted. There now
exists a very well populated library of accounts of atmospheric
lights accompanying, preceding or following a minority (perhaps
0.5 %) of major earthquakes. Indeed, some very thorough scientific
reports accompany their sightings before, during and after earth-
quakes such as the 2009 LAquila earthquake in Italy or the 2008
Sichuan quake in China; however, many more are known.
Moreover, many of these odd and rather beautiful lights are
accompanied by the formation of clouds which may, on some
instances, be visible from space (Fig. 4.7). Just like the Hessdalen
Lights, the manner in which these lights and clouds are generated
appears to be utterly mysterious. Figure 4.8 illustrates some earth-
quake lights that have been recorded on film. From the ground
some of these clouds appear to be luminous and somewhat similar
to aurora (upper right Fig. 4.6). While others appear to be just light
rather than illuminated clouds (lower left Fig. 4.6).
Although it is certainly believable that earthquakes generate
luminous phenomena, you have to be rather careful about which
are genuinely connected with quakes and which are simply coinci-
dental observations. The main problem with the phenomena is the
broad dispersion in their timings and their relative rarity. There is
no doubting the photographs and occasional video evidence are
real. The question is are they associated with the earthquake or
are they misidentifications of man-made or natural phenomena,
or just a very interesting and unusual coincidences? Given the
broad nature of the observations, from balls, to lightning-like dis-
charges, to diffuse glows, and that some of these lights appear well
in advance of the earthquake or even several weeks afterwards,
proving any connection is fraught. Some luminous phenomena
are clearly iridescent clouds, a well-known and very beautiful
rainbow-like pattern caused by sunlight refracting and scattering
within ice clouds. These iridescent clouds could have connections
Weird Weather 175
FIG. 4.6 Earthquake lights and clouds from different perspectives. Upper
left (a) shows a photograph taken of a luminous ball near LAquila on June
20, 200810 months before the quake. Photograph by Bruno Chiarelli.
Photograph (b) is one of many ground photographs taken by an observer
around 30 min before the deadly 2008 earthquake in Sishuan Provence in
China. Photograph (c), lower left was taken near the Sakurajima volcano
in Japan. As well as the prominent light near the foot of the volcano there
are numerous glowing balls (one is arrowed). When the brighter light
turns off in the YouTube video this is taken from, the balls disappear.
The wide spread in the timing of lights leaves open the door that they are
unconnected to earthquakes
with the quakes or they could just be coincidental in their appear-

ance. Iridescent clouds are certainly not that rare. Other flashes
during earthquakes could be downed and blowing transformers.
Observations of earthquake lights made around the time of
the 2009 LAquila earthquake are certainly unusual and suggest
some interesting phenomena is going on. Whether these are truly
linked to the earthquake is the point in question. At LAquila,
before the quake, countless observers reported flame-like erup-
tions from the ground and flashes as well as increasing reports
of ball-lightning-like phenomena (Fig. 4.6). These tended to give
way to broader atmospheric glows both during and after the main
shocks had passed. However, some observers reported lightning-
like discharges coming from the ground after the after the main
shocks had passed. Other observers noted odd clouds; many like
those seen in the moments before the Sichuan quake, looking like
aurora, while afterwards odd, stratified clouds with a violet com-
plexion covered the ground near the regions mountains.
How might we explain these phenomena, first of all assum-
ing that all have connections with earthquakes? As with earth-
lights we enter territory with fairly wild speculation. Explanations
included the piezoelectric effect, suggested also for Hessdalen,
which involves shocked or strained quartz. Alternatively, some
suggest earthquakes cause disruption of the Earths magnetic field
leading to flashes or other luminous phenomena.
Robert Thriault and colleagues have suggested a new phe-
nomenon to explain EQLs (Fig. 4.7). This is similar in some regards
to the piezoelectric effect, but somewhat more radical. In their
FIG. 4.7 Cross-cutting basaltic dyke at Innellan, western Scotland, slic-

ing through older, metamorphosed limestone. The left edge of which is
indicated in blue. Is this 60 million year old feature the sort of rock that
generates earthquake lights? If Robert Thriault and colleagues are cor-
rect, the inhabitants of this British village could be in store for some
UFO activity should the earth violently shudder
Weird Weather 177
new model it isnt quart-filled rocks like granites that are responsi-
ble but instead their more iron-rich cousins, basalts, dolerites and
gabbros. In particular, where these rocks are arranged in vertical
or near vertical structures called dykes, the build up or release of
strain associated with earthquakes generates ionized gases within
the rock. These are then captured by the Earths magnetic field
and funneled along the length of the dyke at high speed, eventu-
ally erupting out of the rocks and into the atmosphere above. The
source of all this activity is defects in the formation of iron-rich
crystals within the rocks. Positive charges uncouple from their
surroundings when the crystals are stained then escape along the
axis of the channel. This eventually brings them to the surface.
Although it seems quite convincing that some EQL phenomena,
such a erupting blue jets or lightning-like discharges, would seem
to be explicable in the context of this model, others, such as irides-
cent clouds, diffuse glows or ball lightning-like phenomena would
be harder to ascribe. That some luminous phenomena are associ-
ated with violent underground events seems hard to refute. And
the figure of 0.5 %, mentioned earlier? This low figure, you may
recall, is the proportion of earthquakes to which earthquake lights
are associated through observation. The low number, Freund and
colleagues suggest, is down to the geology of the local rocks. Only
in those places where there are steeply-dipping layers of basaltic
rocks will earthquake lights be found. This should be a testable
prediction as most of the planet is sufficiently well-mapped to
identify those rocks the authors suggest are associated with earth-
quake lights (EQLs).
What of earthquake clouds? Figure 4.8 shows a train of clouds
originating above the epicenter of the 2003 Bam earthquake that
trail in the prevailing winds east southeastward towards western
India. Other such clouds have been documented for nearly 400
years. Like the EQLs definitively linking clouds to earthquakes
will be fraught with risk. In particular the definition of what con-
stitutes an earthquake cloud seems to vary and the mechanism of
their formation is speculative, to say the least. It is suggested that
such clouds may form when ground water percolates into rocks
that are cracking ahead of the main earthquake. As the rocks crush
and grind together the water is heated and escapes as invisible
clouds of steam. These then condense downwind when they have
FIG. 4.8 An earthquake cloud? This satellite photograph taken shortly

before the 2003 Iranian earthquake centered on Bam shows a large cloud
originating above the earthquake epicenter and drifting downwind across
southern Iran and Pakistan. The cloud was visible for the best part of a
day. Photo taken by the IndoEx satellite/University of Dundee
risen to suitable height. In some instances steam is seen to escape

from some areas, however, this is not common. Alternatively,
clouds could form in more complex ways when crushed rocks
release clouds of dilute ions into the air. These ions then attract
water molecules which condense to form clouds. None of this is
scientifically unreasonable: the question is, is any of it true?
What this area of research needs is a mechanism that can be
directly tested on site. Given the sporadic nature of earthquakes
this is a bit like looking for a needle in a haystack. That said, 20
years ago supernova hunting was in a similar state. The intro-
duction of automated searches soon revolutionized the science;
opening the door to much more systematic, observational astron-
omy while providing a far clearer understanding of the diverse
phenomena called supernovae.
For now, it is fairly positive some of the EQL phenomena are
real, like the lightning-like discharges and blue jets erupting from
the ground as well as some of the broader luminous displays. Are
the others really associated with the earthquake or are they just
happening coincidentally? For those which we have a reasonable
Weird Weather 179
degree of certainty are we seeing a new connection between the

planets interior and atmosphere, one far more subtle but no less
impressive than volcanic eruptions or other geothermal activity?
If these observations are confirmed, we shouldnt be surprised
to find similar lights in the atmospheres of Mars or extra-solar
worlds. Will these worlds be oddly illuminated by fleeting phos-
phorescences?
Sprites, Jets and Other Luminous

Atmospheric Phenomena
Another observation dismissed for decades was the appearance of
brief blue or red discharges above thunderstorms. A lot of pilots
sightings of these enigmatic lights were grouped with UFOs and
consigned to the dustbin. As there are some glorious sightings,
photographs and videos of these illuminations, the combined body
of evidence became overwhelming in the early 1990s when they
were finally added to the list of phenomena associated with thun-
derstorms. Figure 4.9 shows a particularly wonderful photograph
taken of sprites above a tropical thunderstorm taken from the
International Space Station.
The origin of sprites is partly understood and appears to be
linked to particularly large thunderstorm cells with an extensive
anvil structure. Sprites appear simultaneously with cloud-to-
ground strikes and in particular the discharge of the positively-
charged core of the anvil or upper cumulonimbus cloud. The upper
portion of most thunderstorms carries a strong positive charge,
while the corresponding negative charge is found lower down,
where the air is warmer.
Around the outer surface of the positively charged anvil
is a halo of negative charge that is produced by induction. The
strong positively charged ice particles cannot conduct to the air
as it is a strong insulator so instead they attract negative charges
towards them in the surrounding air. Thus the entire upper por-
tion of the cloud is shielded from the stratosphere by a shell of
negative charge. What happens next is unclear, but the trigger for
these novel luminous phenomena appears to be the discharge of
the positively charged portion of the cloud to the ground below.
b c
Elve (E.M.P.)
Mesosphere
Sprite
Stratosphere
Blue jet of ionized particles into stratosphere

Strat
Troposphere
FIG. 4.9 A red sprite (central image, above) fluorescing 60 km above a

thunderstorm over Myanmar near the Indian Ocean (a). The sprite appears
over the center of the thunderclouds anvil just as lightning discharges
below. Image courtesy of NASA/ISS. Lower image is a sprite caught by
the Southern Observatory in Chile (b). The sprites are located above very
distant thunderstorms over Brazil. Below (c) sprites (red tendrils) tend to
form above the positively charged anvil when it discharges directly to the
ground (yellow forked arrow). Elves (the blue ring) are rapidly expanding
blue rings of light caused by electromagnetic pulses associated with the
sprites reaching the base of the thermosphere (Chap. 1)
This, in turn, triggers two other processes. In the first, a breakdown

occurs between the upper, positively-charged region of the storm
and the screening negative charge that surrounds it. This process
appears to trigger the formation of smaller blue jets that are seen
above the anvils of some storms and extend into the stratosphere.
In some storms, there is instead an upward directed collapse of the
negatively-charged region found in the middle of the storm. In these
events, the bubble of negative charge explodes upwards through the
top of the cloud and into the stratosphere, eventually reaching the
base of the mesosphere, a good 3080 km higher up. This generates
much larger blue jets, unimaginatively called gigantic jets.
Weird Weather 181
Sprites are better understood. These are directly associated

with the discharge of the positively charged anvil to the ground
and are presumably a response of the upper atmosphere to the rapid
change in the voltage (the potential difference) between the top of
the anvil and the ground underneath; or to a change in voltage
between the top of the anvil and the screening charges that sur-
round it. They are exclusively mesospheric phenomena (4090 km
up) while the jets are confined to the stratosphere and appear to
behave in a similar way to cloud-to-ground lightning with leader
and return stroke components.
On closer analysis the gigantic jets appear to be a hybrid
between the smaller blue jets and the red sprites. Smaller blue
jets give rise on occasion to gigantic jets when the smaller blue
jets have initiated the breakdown of the electrical resistance
through the full depth of the stratosphere. Where they carry suf-
ficient oomph, the discharge of the blue jet triggers a sprite, which
expands upwards towards the mesosphere. This, in turn, creates
an ionized pathway for further electrical discharge, which extends
from the cumulonimbus cloud, directly to the base of the iono-
sphere. With the insulating layer disrupted, a negatively charged
bubble of gas erupts out of the cloud and punches through the
stratosphere. A gigantic jet is born.
Finally, there is one further mesospheric phenomenon known
as elves that is fairly well understood. These are rare, faint concen-
tric shells of expanding luminosity that can be seen at the base of
the thermosphere (Chap. 1). Elves form when an electromagnetic
pulse, associated with the lightning discharge, reaches the base of
the thermosphere and expands outwards through the highly ion-
ized gas. The electromagnetic pulses associated with lightning
also heat the upper layers of the mesosphere and layers of ion-
ized gas in the ionosphere. These pulses can be up to 20 billion
Wattsa rather formidable output. As well as generating gamma
ray flashes (below) there is the suggestion that such electromag-
netic pulses have sufficient energy to accelerate electrons from
the atmosphere out into the surrounding space. Here, they are
captured by the Earths magnetic field and travel through the Van
Allen radiation belts. Electrons from here can then be scattered
back into the upper atmosphere, completing a vast circuit that
links our atmosphere and interplanetary space.
Gamma Ray Bursts on Earth

The repertoire of phenomena associated with terrestrial thunder-
storms took a major leap up when, in 1991, the Compton Gamma
Ray telescope detected millisecond-long pulses of gamma rays
originating within the Earths atmosphere. Although gamma ray
bursts were well known from astronomical sources (and from ter-
restrial nuclear tests) it came as something of a surprise when the
orbiting telescope detected these highly energetic (20 MeV) bursts.
The bursts, known as Terrestrial Gamma Ray Flashes (of TGFs),
were as brief as they were violent with each lasting between 0.2
and 3.5 milliseconds. The origin was clearly connected to thun-
derstorms with the few detected bursts at the time coming shortly
before lightning was detected from the ground.
Although the precise mechanism remains elusive, the likely
scenario involves the involvement of cosmic rays. These highly
energetic particles bombard the Earths atmosphere minute by
minute and will generate sprays of secondary particles in the
stratosphere. Some of these liberate electrons and it is these elec-
trons that are the eventual trigger for the gamma rays. What hap-
pens next is a powerful lightning strike close to where the spray of
electrons has formed. Some researchers have the electrons sourced
within the top of the thundercloud while others require only that
it appears close to a storm. When lightning discharges it generates
a powerful electrical field that accelerates the cloud of electrons,
which then whiz upwards towards space. As these interact with
the nuclei of neighboring atoms, the electrons rapidly decelerate
and release what is known as Bremsstrahlung, or braking radia-
tion. This can include gamma rays if the electrons have enough
oomph to begin with.
More amazing still was the discovery in 2011 that thun-
derstorms can also liberate jets of antimatter. The scenario is
rather well understood, but their detection relied on a bit of luck
and help from the Earths magnetic field. The antimatter elec-
trons, known as positrons begin their life as a gamma ray. These
gamma rays originate in thunderstorms through the mechanism
that was discussed above. Once generated, these gamma rays
Weird Weather 183
can interact with further atomic nuclei in the atmosphere. In

the right circumstances these produce electronpositron pairs.
Ordinarily these pairs of particles might wander around before
the positron interacted with, and then annihilated, a nearby elec-
tron. However, in the rarefied confines of our upper atmosphere,
both the electron and positron can escape. This involves these
charged particles being rescued by our planets magnetic field.
Born within our planets magnetized umbrella each particle is
whisked away from its partner and shot out into space. Normally
at that point they would either escape the Earth in its entirety or
become trapped in our magnetic field, but thanks to our technol-
ogy some have a grizzlier fate. High above the Earth, some of the
positrons happen to whack into the orbiting Fermi Telescope.
When they do so, they immediately annihilate. Their demise
results in the formation of a characteristic pair of gamma rays,
each with the energy of 511,000 eV. The plucky antimatter par-
ticle evades many hazards throughout its flight and escape is just
on the horizon when they are annihilated in an instant. This con-
voluted journey involved a thunderstorm, a cosmic ray, and the
Earths magnetic field. As with the discovery of gamma rays by
Compton Observatory, the detection of antimatter by the Fermi
Gamma Ray Observatory was entirely fortuitous. Given the con-
voluted path to make antimatter, does ball lightning seem so odd
by comparison?
Terrestrial gamma ray bursts and the formation of antimat-
ter are also interesting because only a minority of thunderstorms
seem able to produce these. Over 1000 storms may rage across
the planet mostly in the tropics each day, but less than 100 TGFs
have been detected over this period. This might be because only
a minority of thunderstorms are energetic enough or perhaps tall
enough to launch gamma rays through the overlying atmosphere
into space, or it might be because the gamma rays are focused into
beams rather than broad sprays of radiation. Most of these beams
simply evade our detection. More work is clearly required.
Aside from the dramatic displays of lightning associated with
thunderclouds that can deliver 10,000 Amps of current in every
strike, the atmosphere also delivers a steady stream of charge,
amounting to 1000 Amps across the entire surface of the planet.
This is delivered through a voltage (potential difference) of 120 V

per meter. One thousand Amps would be rather unpleasant if
focused on one spot, but across the entire face of the planet this
1000 Amp flow amounts to a drizzle of 10 trillionths of an amp per
meter square. Clearly, this is not enough to light even the dimmest
of bulbs. Measly though this is, this arm of the circuit is merely
the closest part of an even larger circuit that extends through the
ionosphere to the magnetosphere and ultimately outwards to the
stars through the solar wind. Within this vast electrical track thun-
derstorms play their part by using the Suns energy to charge up
the ionosphere from below. When lightning discharges a torrent
of electrons to the ground, the cloud can then discharge positive
charges upwards towards the base of the ionosphere. This process
helps maintain the voltage between the Earths surface and the
ionosphere above.
While thunderstorms are a dramatic player in the global elec-
trical circuit they are also a profuse source of radio waves. Each
lightning bolt generates an electromagnetic pulse, as we have
seen, and this is detectable as annoying interference on our tele-
visions and radios when storms are nearby. Storms also generate
much lower frequency radiation as part of this repertoire. The
electromagnetic pulses that race away from lightning discharges
have a range of wavelengths. Those in the 50 Hz range are able
to propagate right around the Earth numerous times before dying
away. As they do these waves can generate so-called Schumann
Cavity Resonance. This phenomenon is caused by the electrically
insulating properties of the lower atmosphere, which separates
the charged ionosphere from the chargeable surface of the planet.
Schumann Cavity Resonances are effectively standing waves
that oscillate, or vibrate, through the depth of the stratosphere
and troposphere. These vibrations release radio waves at various
low frequencies that can be detected by radio antennae. Because
lightning can excite the lower atmosphere in the right range of
wavelengths, it can generate these Schumann Cavity Resonances.
This allows this low energy radiation to be used to monitor the
frequency of lightning strikes across the globe as a hum of radio
waves. Observations suggest that the overall frequency of light-
ning is increasing and at least in theory this could be direct conse-
quence of rising global temperatures.
Weird Weather 185
Sky Quakes
One further phenomenon is truly global in nature and just as
mysterious. So-called sky quakes have been identified in various
locations, dotted around the globe, each with its own regional
name. Along the banks of the Ganges in Bangladesh the booms are
called Barisal Guns; along the eastern shores of the North Sea in
Belgium and the Netherlands they are called mistpoeffers; while
in Seneca and other parts of the south eastern US they are referred
to as Seneca Guns. In Connecticut Valley they are called Moodus
Noises, while in the Philippines the name retumbos is used. In each
location these wonderfully evocative names describe the rumbling
booms that are sometimes seconds long that occur without warn-
ing and at any time in the day. They are said to resemble distant
thunder or artillery fire and may be accompanied by low rumbling
which shakes buildings. Although this is clearly reminiscent of
earthquakes scientists trying to assign the reported noises to seis-
mic activity often (but not always) come up short.
Although some of these booms have been definitely linked to
the sonic booms of military aircraft, others are clearly not man-
made with accounts going back over 100 years in locations as
far apart as Victoria in Australia and Seneca in the US. So, what
causes the Seneca Gun to fire?
Meteors entering the atmosphere might work as a cause, but
the frequency and their seeming preference for some geographical
locations would make that hypothesis unlikely. Gas erupting from
underwater is possible and would tie with the frequency of these
lights near the coast or lakes. Similarly, collapsing underground
limestone caverns might work as well. In some instances, in much
like the manner Fata Morgana is a mirage of light caused by atmo-
spheric refraction, the Seneca Guns and related sky quakes might
be distant thunder refracted over the observers horizon. In at least
some instances this idea has been investigated using regional
meteorological reports and there is no evidence that any thunder-
storm lies close enough to the purported sounds to have created it.
That said thunder refracted from further afield could explain some
of these sky quakes.
In a few instances where small earthquakes appear to be asso-
ciated with similar noises the arrival of the earthquake P-wave
may, on occasion be registered as a rumble by observers. However,

none of the current explanations quite fits the bill. The Seneca
Gun or sky quake might be a combination of different noises,
including gas burps, collapsing caverns and super-sonic military
jets. At Seneca some booms are definitely accompanied by visible
rumbling which would indicate either atmospheric sonic booms
or earthquakes. Hopefully, time will tell what causes these kinds
of phenomena and whether any are related to geological turmoil or
atmospheric disturbances.
The Red Rains of Kerala

Finally, turning from the surreal to the sublime, we look at the red
rains of Kerala. In the southern Indian province of Kerala blood-
red rain fell sporadically across much of the region during the late
summer of 2001. Some reports tied the appearance of this lurid
precipitation to the sound of thunder, which was taken by some
to imply an atmospheric boom, perhaps caused by an incoming
meteor. Analysis of the rains revealed they were colored by cellu-
lar structures. The assertion that these had an association with the
boom and supposed meteor led to the idea that this was some sort
of panspermian event with alien cells having arrived from outer
space. Early analysis suggested that the spores had an unusual
composition. However, later investigation confirmed that the red
cellular material was some form of algal spore.
In the same region black, green and yellow rain has also been
reported. The same phenomenon was recorded in Kerala in 1896
and again in 2012, suggesting an interesting local source of these
spores. Similar rains have been reported to the south in Sri Lanka,
implying a localized but generally widespread geographical origin.
At present the source is fairly mysterious, but there is no doubt
that the coloration is down to terrestrial rather than extraterres-
trial sources.
As areas only meters apart can have red rain or clear and quite
normal rainfall, the source of the spores must be very localized
and somehow remain isolated within the atmosphere so that it
doesnt all mix into a bland baby-pink. Quite how this is achieved
could have implications for issues as diverse as the circulation
Weird Weather 187
within the atmosphere and, potentially, the spread of pathogens

by air. The red algal spores could thus be used by scientists as a
natural tracer for air movement presenting a novel and quite color-
ful means to probe the dynamics of the Earths atmosphere.
Flying Spiders
Even animals can take advantage of atmospheric irregularities.
Charles Darwin observed the phenomenon of flying spiders
onboard HMS Beagle in the early 1830s, and found hard to explain
how spiders could launch themselves from objects to fly horizon-
tally when the air was apparently still. Convection would tend to
lift the spiders vertically, not sideways. Confounded by his obser-
vations Darwin went as far as to suggest that the observed fanning
out of fibers from the spinnerets might best be explained by elec-
trostatic repulsion. Indeed, it is now thought that the pervasive
120 V per meter field, permeating the Earths atmosphere might
be the force at work.
On a calm, clear day, as weve seen there is a persistent field
generated by various phenomena. This field and the accompany-
ing charge separation forms the basis for the formation of thunder-
storms under differing circumstances. Spiders might well utilize
this field to launch themselves, for only 30 nanoCoulombs (30 bil-
lionths of a Coulomb of charge2) is sufficient to provide enough
repulsion to drive uplift of the spiders silk fibers for a spider that
is 12 mm across.
As Darwin noted it is certainly suggestive that when spiders
produce the threads for ballooning, as this process is known, the
fibers immediately fan out and rise upwards, which would imply
they are repelled by themselves and the underlying surface. This
implies that they have acquired the same electrical charge as the
underlying surface. Thus, it seems reasonable that spiders are ini-
tially launched from the surface by electrostatic repulsion, before
any light air currents can catch and transport them further.
2
1 Coulomb is one unit of electrical charge that is equivalent to one amp of current owing in one
second: Q = It, for the math fan; where Q is the charge in Coulombs; I is current in Amps and
t is time in seconds. It is often more customary to dene the Coulomb in terms of electrons, where
1 C is equivalent to the charge on approximately 6.241 1018 electrons.
Conclusions
While the general outlay of meteorological and related phenom-
ena is clearly fairly well understood, there still exists a battery of
observations on Earth that retain a considerable air of mystery.
Some, like the lights at Marfa and elsewhere, clearly fall into the
realm of misidentified human activity, but there are clearly oth-
ers that have truly natural origins that we still do not understand.
Similarly, we are bound to encounter phenomena on other planets
that simply dont fit current theories. Getting a better grip on the
weird end of terrestrial weather will certainly aid our understand-
ing of phenomena on truly alien worlds. Look at Chap. 6 (Mars)
or Chap. 7 (Jupiter and Saturn) for a clear indication that this sup-
position is true. The universe throws all sorts of apparently inex-
plicable phenomena at us. The beauty of these observations does
not lie in presupposing that these are mysterious and inexplicable,
but instead reveling in the joy of unpicking them and finding out
how they work. Meteorology and climate science are universal
phenomena and we can only understand the climate of our world
and of alien worlds if we focus on both. It will be interesting to
complete the full inventory of weird and wonderful atmospheric
phenomena: undoubtedly it will be far greater than we imagine.
References
1. Dimitar Ouzounov, Sergey Pulinets, Alexey Romanov, Alexander Romanov, Konstantin
Tsybulya, Dimitri Davidenko, Menas Kafatos, Patrick Taylor. (2011). Atmosphere-Ionosphere
response to the M9 Tohoku Earthquake revealed by joined satellite and ground observations.
Preliminary results. Retrieved from http://arXiv.org/pdf/1105.2841.pdf.
2. Dijkhuis, G. C., Callebaut, D. K., Lu, M. (Eds.), (2006). Observation of lightning ball (ball
lightning): A new phenomenological description of the phenomenon. In Domokos Tar.
Proceedings of the 9th International Symposium on Ball lightning (ISBL-06), 1619 August
2006, Eindhoven, The Netherlands, pp. 222232.
3. The Society of Physics Students at the University of Texas at Dallas. (2005). An experimental
analysis of the marfa lights progress report (Submitted). Available as a PDF online.
4. Pettigrew, J. D. (2003). The Min Min light and the Fata Morgana. An optical account of a mys-
terious Australian phenomenon. Clinical Experimental Optometry, 86(2), 109120.
5. Castelvecchi, D. (2015). Rogue antimatter found in thunderclouds. Nature, 521, 135.
6. Fidani, C. (2010). The earthquake lights (EQL) of the 6 April 2009 Aquila earthquake, in
Central Italy. Natural Hazards and Earth System Sciences, 10, 967978.
7. Park, S. K., Johnston, M. J. S., Madden, T. R., Dale Morgan, F., & Frank Morrison, H. (1993).
ELF precursors to earthquakes in the ULF band: A review of observations and mechanisms.
Reviews of Geophysics, 31, 117132.
Weird Weather 189
8. Zhonghao Shou, Darrell, Harrington. (2000). Bam earthquake prediction & space technology.
www.earthquakesignals.com/zhonghao296/copies/harrington.pdf.
9. Stolyarov, A., Klenzing, J., Roddy, P., Heelis, R. A. (2005). An experimental analysis of the
Marfa Lights. Progress Report submitted by: The Society of Physics Students at the University
of Texas at Dallas. Retrieved from http://www.spsnational.org/wormhole/utd_sps_report.pdf.
10. Domokos Tar. (2008). New revelation of lightning ball observation and proposal for a nuclear
fusion reactor experiment. In: Vladimir L. Bychkov & Anatoly I. Nikitin (Eds.), Proceedings
of the 10th International Symposium on Ball Lightning (ISBL-8), Kaliningrad, Russia,
pp. 135141.
11. Hayakawa, M., Nakamura, T., Iudin, D., et al. (2005). On the ne structure of thunderstorms
leading to the generation of sprites and elves: Fractal analysis. Journal of Geophysical Research,
110, 2735. D06104.
12. Siingh, D., Singh, A. K., Patel, R. P., Singh, R., Singh, R. P., Veenadhari, B., Mukherjee, M (
2009). Thunderstorms, lightning, sprites and magnetospheric whistler-mode radio waves.
Retrieved from http://arxiv.org/abs/0906.0429.
13. Thriaulta, R., StLaurentb, F., Freundc, F. T., & Derr, J. S. (2014). Prevalence of earthquake
lights associated with rift environments. Seismological Research Letters, 85, 159178.
14. Tavani, M., et al. (2011). Terrestrial gamma-ray ashes as powerful particle accelerators.
Physical Review Letters, 106, 17.
15. Gorham, P. W. (2013). Ballooning spiders: The case for electrostatic flight. http://arxiv.org/
pdf/1309.4731v2.pdf.
16. Schneider, J. M., Roos, J., Lubin, Y., & Henschel, J. R. (2001). Dispersal of Stegodyphus dumi-
cola (araneae, eresidae): they do balloon after all! The Journal of Arachnology, 29, 114116.
5. Venus
Introduction: The Twin That Isnt

Take two planets of roughly the same mass and composition
and place them at similar distances from their parent star. Youd
clearly expect them to end up as near twins. However, what the
Solar System shows is that it doesnt take much of a butterfly to
drastically alter the fate of worlds.
Venus undoubtedly was bequeathed with the same amount
of volatile materials as the Earth and it is supposed that early on
the two planets might have looked rather similar: immense water
worlds under a thick, hot atmosphere of water vapor and carbon
dioxide (Chap. 3). However, Venus received just that little bit
more radiation from the Sun than the Earth did and never quite
cooled off the way the Earth did. What happened next would set
the fate of Venus and turn it into the noxious hell it is today. This
chapter describes how a world much like our own can end up so
very different to ours despite a similar start; and how the fate of its
atmosphere became tied to the fate of the planet as a whole.
The Venusian Day

Venus is a thoroughly odd planet. A Venusian sidereal daythe time
it takes a fixed, distant object like a star to culminate each rotation
period is 243 Earth days: moreover, Venus does this backwards (its
rotation is retrograde). Consequently, the Sun rises in the west on
Venus and sets in the east. The Earth, by contrast, has a sidereal
day of 23 h 56 min and rotates on its axis in a prograde direction,
matching the direction of its orbit around the Sun. Moreover, while
the Earth rotates on its axis at 1670 km per hour at the equator,
Venus crawls around at 6.5 km per houra modest walking speed.

DOI 10.1007/978-3-319-25679-5_5
Meanwhile, the Venusian year is 224.7 Earth-days long. Therefore,

the retrograde rotation on its axis means that the Sun rises and sets
on a period considerably less than the sidereal day. Instead of 243
days, the Venusian solar day is only 116.75 Earth-days long, as on
every rotation on its axis, the retrograde spin brings the Sun to the
same point of culmination far earlier than it takes the planet to spin
on its axis. These odd characteristics are as stable as they are nor-
mal. In the 16 years separating the arrival of the Magellan and Venus
Express Probes, the Venusian sidereal day lengthened by 6.5 min for
every sidereal day that had elapsed. This means that the length of a
Venusian sidereal day was 156 min or so longer in 2006 than it was
in 1990. How did this strange state of affairs arise and why does it
continue to change? Just imagine what it would be like if we gained
the best part of 3 h a day every decade.
The likely principle factor directing this odd state of affairs is
likely the shear bulk of Venuss atmosphere and its relative prox-
imity to the Sun40 million kilometers closer than the Earth.
The Sun exerts strong tidal forces on Venuss thick carapace of gas
and on its (probably) plastic mantle. Many calculations show that
this should have been sufficient to tidally-lock Venus to the Sun
within a few hundred million years of its formation. If this had
happened, Venus would now present one face permanently to the
Sun, while the other resided in darkness: this is something that
we shall look at in Chap. 10 when we examine the planets that
orbit red dwarfs. Venus, however, is not (quite) tidally locked to
our star. So, how do we explain the odd rotation pattern of Venus?
Venuss massive atmosphere has clearly had an enormous
influence on the evolution of the planet as a whole. When one
looks at both worlds you can see that the atmosphere on Venus
and the Earth carry vastly different amounts of momentum and
this, in turn, grossly affects the interplay of energy between the
spinning planet and its atmosphere. Although the solid part of
each planet is around one million times more massive than the
Earths atmosphere, this figure is around 100 times smaller for
Venus. Although the masses of atmosphere and underlying planet
are still very different, the bulk of Venuss atmosphere is still suf-
ficient to have significant effects on the rotation of the planet.
Clearly, over time this massive atmosphere must drag on the
Venus 193
planet, both through the bulk gravitational pull and through the
effect of friction. Moreover, with the atmosphere acting as a giant
ocean of liquid-like material, the solar tides can also exert a sig-
nificant pull, which in turn influences the planets spin.
For many decades it was thought that the retrograde spin
of Venuss surface, relative to its orbit around the Sun, implied
that Venus was likely hit by a giant impactor, much like the
one thought to have formed the Earth and Plutos large moons.
However, although such an impact is possible, and perhaps even
probable, the chance that it would have hit Venus in just the
right way to flip it over is unlikely. Instead more recent calcu-
lations place the origin of Venuss retrograde spin squarely with
its massive atmosphere. In 1978 Andrew Ingersoll and Anthony
Dobrovolskis (Caltech) showed that solar tides acting within the
massive atmosphere were more than sufficient to slow the planets
spin to the point that the planet now completes an orbit of the Sun
faster than it rotates on its axis. This process was dependent on a
number of other factors, including the pull of the Earth. However,
in principle tidal forces could slow the rotation of the planet to its
current, decelerating, level. Venus may be in a brief window where
it the interplay of forces have led the planet to abandon tidal lock-
ing. Thus, if you were to look at Venus one billion years or so
ago it might have been locked, facing the same way to the Sun
for hundreds of millions of years. However, repeated gravitational
nudges from the Earth, or other planets, accelerated Venus out of
its locked position. Now, it may be that Venus is slowly returning
to a life presenting only one face to the Sun. This, too may last for
another few hundred million or few billion years before repeated
nudges from the Earth allow Venus to slip free once more.
A Noxious Vision of Hell
When Mikhail Lomonosov first spotted the planets atmosphere

during a solar transit in 1761 it might have been hoped that Venus,
with its permanently obscured face was some sort of tropical para-
diseif a rather cloudy one. Indeed, this idea persisted until rela-
tively recently when the atmosphere began to be probed by Earthly
machinery.
However, set against the idea of a tropical paradise were cal-

culations made by Rupert Wildt in 1940. Wildt showed that the
amount of carbon dioxide in Venuss atmosphere would lead to
temperatures far in excess of the boiling point of water. Recall
the climate skeptic argument (Chap. 3) that there is some sort of
upper limit to which carbon dioxide could heat the Earth when its
absorption bands become saturated. However, as the concentration
of greenhouse gas rises, the range of wavelengths, over which the
carbon dioxide (or any greenhouse gas) will absorb radiation, will
increase in kind. Such collisional broadening (Fig. 3.7) ensures that
Venus, and presumably the early Earth (Chap. 3) are (or were), very
hot indeed. All of this thanks to the absorption of infrared radia-
tion by carbon dioxide gas. Wildts calculations were confirmed
by the fly-by Mariner probe in 1962. When the Soviet Venera 4
arrived in 1967 it confirmed that carbon dioxide and its associated
massive greenhouse effect dominated the climate of Venus. The
idea that Venus might be some form of lush tropical jungle was
thoroughly incinerated.
The vision of hell was confirmed when the Soviet Union
landed probes in the late 1960s and early 1970s. During their short,
unpleasant lives the Venera landers indicated a barren volcanic
hell, where a dim light from the Sun refracted into an odd appari-
tion of Hades. Venus is a planet boiled and crushed under an atmo-
sphere 92 times more massive than that surrounding our world.
So dense was the atmosphere that the pressure at the surface was
roughly equivalent to being under a depth of 1 km of water. Were
you able to survive the temperature (and suit yourself up appropri-
ately) you would be able to swim in the dense carbon dioxide-rich
atmosphere, which behave much more like a dense liquid than a
gas under Venusian conditions. Even the modest 10 m per second
breeze would be sufficient to knock you off your feet were you
attempting to walk.
More interesting are the effects of refraction. Refraction is
the bending of rays of light (and all electromagnetic waves) as
the radiation travels from one material to another. In Chap. 4
we saw how refraction can generate some odd and quite confus-
ing effects, such as the Marfa Lights. If Venuss atmosphere was
transparent, from your position on its surface, the horizon would
appear to bend upwards around you. Moreover, were you at a
Venus 195
height of 35 km, light would be totally internally reflected around

the planet, so that Venus would appear to extend infinitely in all
directions. However, sadly in practice atmospheric haze prevents
such an optical wonder. So, although the bowl-like effect would
be apparent, you wouldnt see as far before the light was scattered
and absorbed.
Disappointing though this is you might still see an interesting
effect after Sun set. On Earth the atmosphere still remains light for
perhaps half an hour after the Sun had gone down. This is because
the light from the Sun reflects off particles in the atmosphere and
because light refracts around the edge of the Earth. On Venus the
latter effect is far stronger as a result of the thick sea of carbon
dioxide on the surface. As a resulthaze dependentthe Sun
might appear as a thin sliver of light all through the night, until it
rose again the following morning. In the case of Venus, remember
that this night is more than 121 Earth days long. Finally, this thick
layer of carbon dioxide is thought to behave as a super-critical
fluid, meaning that although strictly speaking a gas, it has many
of the properties of a denser liquid. Of these possible properties
the key one is its ability of the dense carbon dioxide atmosphere
to conduct heat very rapidly. This property ensures that there is
little difference in temperature between the day and night sides of
the planet at the surface, even at the surface where winds are light.
The temperatures at the surface of Venus are sufficient to
melt lead, tin and zinc; and probably also induce fairly vigorous
chemical reactions between the rocks. Clearly, the extreme heat
will also drive out any moisture that was present in these rocks
when they reached the surface. After all, on Venus it is only when
you reach around 5060 km up that temperatures fall consistently
below the boiling point of water and whatever moisture is present
would be able to condense.
The Structure of the Venusian Atmosphere

As on the Earth the atmosphere can be divided into distinct lay-
ers. The lowest, extending up to around 60 km, is equivalent to
the terrestrial troposphere. This has a lid at 60 km defined by a
small temperature inversion (a region where temperatures rise
Table 5.1 Temperature and pressure within Venuss atmosphere at various

heights above its surface
Height above Pressure (as a fraction of
surface (km) Temperature (C) that at the Earths Surface)
0 462 92.1
10 385 47.4
20 306 22.5
25 264 14.9
50 75 1.0
55 27 0.60
70 43 0.037
100 112 0.000027
Of note at about 55 km the pressure and temperature are suitable for human survival
with height) that separates the troposphere from the mesosphere

above. As with the Earth air circulates from the equator polewards
in two large Hadley Cells (Fig. 5.1). The slow rotation of the planet
ensures that these reach far closer to each pole than they do on
the Earth. Fully 120 of latitude is covered by these vast, slowly
over-turning flows of air. This is double the latitudinal range that
is seen on the Earth, leaving only a relatively thin sliver of surface
covered by a different mode of air flow (Table 5.1).
The fast and counter-clockwise flow also means that most of
the atmosphere moves in a retrograde direction, like the surface
underneath, but that it moves faster than the planet rotates. Such
motion is known as super-rotation. This phenomenon appears
to have slightly different origins, depending on the planet that is
examined. In the case of Venus the atmosphere completes a circuit
of the planet in far less time than the planet takes to rotate on its
axis: 4 Earth days compared with 243 Earth days, respectively, or
one sixtieth the length of the sidereal day. Compare this to the
mean wind speed on Earth which is roughly one tenth the rotation
speed of our planet.
While super-rotation in itself is remarkable, most of this
super-rotation is accomplished above 50 km where winds blow up
to 400 km per hour in a band south of the equator. How is this feat
accomplished on a planet that rotates so slowly? Led by Hctor Javier
Durand-Manterola, scientists from the Universidad Autonoma
de Mexico have analyzed winds flowing across the terminator of
Venus at altitudes of 150800 km. These trans-terminator flows
Venus 197
FIG. 5.1 Atmospheric circulation on Venus. Two broad patterns are dis-
cernible. Nearer the surface the atmosphere is broken into a number of
large cells which shrink in size towards each pole. The largest of these is
equivalent to our Hadley Cells but extends across twice the latitude the
equivalent cells do on Earth (060 on Venus). A smaller polar collar
may be thought equivalent to our Ferrell Cells (Chap. 1). Meanwhile,
nearest to the poles are two anticyclonic vortices shown as blue donuts.
Overlapping these in the mesosphere and lower thermosphere (90
150 km up) is a simple cell that transfers air at height from the sub-solar
point to the anti-solar point on the night side. This thermal tide is driven
directly by solar heating and ultimately leads to oxygen loss to space on
the night side (brown arrows). Some of this ends up captured by the Earth
contain very little mass but move at several kilometers per sec-
ondfar higher than the air movement below. Thanks to differ-
ences in the speed of these winds in the dawn and dusk sides of
Venus, complex pressure waves are formed as the faster moving
dusk terminator flow pushes across the night-side of the planet and
into the outgoing dawn terminator flow.1 As a result of the col-
lision atmospheric waves are generated that transfer energy and
momentum into the lower atmosphere. Similar waves are exam-
1
Measurements made by Venus Express conrm that atmospheric gravity waves (chapter 10) trans-
fer momentum from the ionosphere to the cloud decks and drive the super-rotating ow.
ined in Chap. 10. However, in this case these atmospheric gravity

waves transfer momentum from the Earths surface to its upper
atmosphere.
The scientists speculate that it is this transport of energy and
momentum has accelerated the atmosphere of Venus to dizzying
speeds. Compare this mechanism to that proposed for the genera-
tion of super-rotation within the atmospheres of hot Jupiters in
Chap. 10. Although both planets might appear superficially simi-
lar, with atmospheres dominated by super-hot gases, the origin
of this heat is quite distinct. In the case of Venus heating of the
upper atmosphere is effectively delivered from below, as the thick
carapace of greenhouse gases captures the Suns radiation. In the
atmosphere of a hot Jupiter, the heating is strongest at the top of
the atmosphere, where direct heating from the parent star is most
intense. Thus, it shouldnt be surprising that both planets have
super-rotation but that it is driven by such different mechanisms.
Returning to Venus, ultimately the origin of the pressure
waves that drive super-rotation can be traced back to the solar
wind. Thus, if correct, it is the Suns hot breath that forces the
atmosphere of Venus to move so swiftly. A similar mechanism
may operate out at Titan (Chap. 9) causing its atmosphere to
super-rotate as well.
Air speeds are low at the surface but increase to around 100 m
per second (360 km per hour) at the tropopause. Thick clouds of
sulfuric acid, with a smattering of sulfur, sulfur dioxide and water
vapor fill this layer and obscure the surface almost in its entirety.
Above this region the cloud layer rapidly clears, leaving a haze of
sulfuric acid in a layer perhaps 20 km thick that can be thought of
as the mesosphere. Unlike the Earth there is no intervening strato-
sphere where temperatures briefly rise again with height. This is
because Venus lacks a distinct ozone layer that would absorb solar
energy and cause warming.
In the mesosphere temperatures fall much more slowly with
height and there are slight blips in the overall trend before we
enter the thermosphere above. Indeed the lowest part of the meso-
sphere, which extends from 65120 km above the ground. Here,
the temperature remains roughly constant in the lowest 8 km of
this layer, at around 43 C. However, as we continue upwards
and clear the upper cloud deck we enter the upper mesosphere
where temperatures continue downwards to around 110 C. On
Venus 199
the dayside of Venus, this is the coldest region in the atmosphere.

Above 95 km we cross the mesopause: the broad transition region
between the mesosphere and thermosphere. Temperatures rise
slowly in this region to temperatures between 27 C and 120 C at
120 km up. However, the Sun plays a very big part in controlling
the temperature of the thermosphere. In the shade of the planet,
the thermosphere in the night-side of Venus is even colder than
the mesosphere with temperatures lowering towards 120 C.
The layers above the tropopause also circulate from the sun-
lit side to the dark side of the planet. There is very little north
south movement of air at all and the entire process is driven by
solar heating which drives gases away from the area that is most
strongly heated. In this layer carbon dioxide is broken down to car-
bon monoxide and oxygen, some of which later escapes to space at
the anti-solar point on the planets night side.
In the troposphere the circulation immediately above the sur-
face, poleward of 60, the Hadley cells give way to two smaller
cells. The first cell, known as the polar collar is broadly equivalent
to terrestrial Ferrel Cells and is a zone of upwelling gas on Venus.
Astronomers can track the air flow here by the distribution of all
that noxious carbon monoxide which is concentrated within these
polar bands. Clouds generally rise highest in these zones, driven
by convection from below. Presumably the cooler air sinking at
either end of the Hadley cells drives the upward movement of air
in the polar collar a bit like a cold front kicking up warm air to its
front. Near the northern and southern flanks of the Hadley cells
the Coriolis Effect is sufficient to drive airflow into two circum-
polar jets that circulate at roughly 140 km per hour at 5060 north
and south of the equator.
In the late 1970s the Pioneer probe detected an anticyclonic
storm located over the North Pole. Consequently, it was not
entirely surprising when Venus Express detected the correspond-
ing circulation over the South Pole in 2006. However, while the
North Pole revealed a single storm that over the South was alto-
gether more striking and unusual: an s-shaped pairing of two vast
anticyclones, swirling air back downwards to the planets surface.
Each swirling eye counter-rotates around the other, forming a pair
of hurricane eyes. However, like Jupiters red spot, each Venusian
anticyclone is an area of relative calm in an otherwise turbulent
sea of poisonous gas. Each vortex rotates around their common
center over a period of about 3 days in the same general direction

as the rest of the atmosphere: i.e., from west to east at this lati-
tude. Within and around the storm edges winds drop from around
50 m per second around the eye to zero within the eye, where
winds are descending. In many regards these storms are analogous
to the Polar anticyclones that form over Antarctica (and to a lesser
extent) over the North Pole on Earth during the winter in each
respective hemisphere.
Within the polar collar clouds can form at the top of the
troposphere or within the mesosphere on a very rapid and short
term basis. Venus Express saw this kind of event in 2007 and
was thought to have happened when a blast of sulfur dioxide was
injected into the mesosphere by a storm lower down. Such nacre-
ous clouds, composed principally of sulfur dioxide and sulfates,
are fairly rare phenomenon on Earth but are probably relatively
common in the middle atmosphere of Venus, where such gases
are abundant.
As you ascend higher through the atmosphere you reach the
ionosphere. Here, as on the Earth and other planets ultraviolet,
x-ray and gamma radiation from the Sun, together with a barrage
of cosmic ray particles, blast electrons from the nuclei of the
abundant oxygen and carbon atoms. This breaks carbon dioxide
down to carbon monoxide and free oxygen. Gases continue to
flow away from the sun-lit side towards the opposing darkness.
Meanwhile at the point where the solar wind reaches the iono-
sphere a turbulent boundary forms called the ionopause. Here
the solar wind is largely deflected by a combination of increas-
ing gas pressure and the effect of magnetic fields created as the
wind meets the ionized top of the atmosphere. This, and a bow-
shock formed somewhat further upwind, towards the Sun, gives
Venus some limited protection from the erosive power of the
solar wind. However, as the European Space Administrations
Venus Express showed, a considerable mass of atmosphere is still
removed on a continuous basis by the constant stream of gases
from the Sun. In all Venus loses roughly 1000 kg (1 metric ton)
of oxygen per hour: however, the surprising thing is that this is
roughly the same as the amount of mass lost by the Earthand
by Mars. An absence of a magnetic field is apparently no great
problem for Venus.
Venus 201
No Layer but Yes, Ozone Around Venus
Soon after its arrival in 20006, Venus Express began sampling the
atmosphere of Venus at different altitudes. It used measurements
in the infrared and microwave portions of the electromagnetic
spectrum. These were complemented by an array of measurements
at radio wavelengths, which among other things monitored the
planets magnetic field and lightning.
Above the Earth ozone is formed in significant quantities
from molecular oxygen. Ultraviolet light interacts with the nor-
mal two atom diatomic form of the gas that is produced primarily
from bacterial photosynthesis in the oceans. The diatomic form of
the gas is converted into the unstable three atom combination we
call ozone. Ozone is proficient at absorbing ultraviolet light in the
window at which this would otherwise penetrate the air and reach
our planets surface.
On Mars and Venus there is no photosynthesis (as far as we
are aware) so the oxygen needed for the formation of ozone comes
from the photochemical break down (photolysis, or light-splitting)
of carbon dioxide gas, which both planets have in abundance. On
Venus carbon dioxide is split by light to form carbon monoxide
and a free oxygen atom. These free atoms are then swept around
to the night side of the planet by Venuss thermal tide (Fig. 5.2),
where they can then combine with one another to form first
diatomic oxygen (the stuff we breathe) or triatomic ozone. High
above the obscuring haze and clouds, in the mesosphere of Venus
carbon dioxide gives birth to ozone with some help from the Sun.
At no place are concentrations sufficient to shield the atmo-
sphere below, but if there were life to be had on Venus, its thick
sulfuric acid clouds and haze would do the trick and prevent the
ultraviolet from harming it. Ozone or not, the place to look for
any kind of life will therefore be within the clouds, noxious to our
senses, though they are.
The Edge of Space
Continuing 100 km above the layer containing the maximum

concentration of ozone we reach the edge of the atmosphere.
Immediately under the Suns glare at the subsolar point the solar
Height
(km)
Exosphere
500
Thermosphere
130
120
110
Mesosphere 100
Nacreous Clouds
90
80
70
Sulphuric Acid Haze
60
Sulphuric Acid Clouds
50
40
Troposphere 30
20
Sulphuric Acid Haze
10
-150 -100 -50 0 +50 +100 +200 +300 +400 +500
Temperature (oC)
FIG. 5.2 The structure of Venuss atmosphere. Overall it has a much sim-
pler structure than the Earths with a steady drop in temperature and
pressure to a height of over 120 km. There is no comparable stratosphere
with a rise in temperature as is found above the Earth. Above 120 km
temperatures rise again as we leave the thermosphere and enter interplan-
etary space. The height of the boundary between the thermosphere and
exosphere depends on which side of Venus faces the Sun, being lowest on
the Sun-lit side at around 220 km above the surface but rising to 350 km
on the night side. Moreover, while on Earth most of the thermosphere is
ionized this is only true of the sunlit side on Venus
wind blows the bow shock inwards to 1900 km above the sur-
face of Venus, which is roughly one third the radius of the planet
(Fig. 5.3). However, when Venus Express made this measurement
it was at the solar minimum in 2007 when the solar wind was at
its weakest. Between the bow shock and the ionosphere lie two
further boundaries: the ionopause, which marks the top of the
ionosphere; and the magnetopause, which is the upper boundary
of the induced Venusian magnetic field. The former lies at roughly
Venus 203
M
a
g
n
e
t
i
c
F
i
e
l
d
L
i
n
e
s
FIG. 5.3 The gross magnetic field and ionopause around Venus. Interac-
tions of the solar wind (red arrow) with the planets ionosphere result in
the formation of a bow shock (black) and a magnetic barrier between
the ionosphere yellow-orange. When the solar wind is stronger the bow
shock approaches within 1900 km of the surface. However, when the
wind is weak the ionosphere balloons outwards and forms a long comet-
like tail in the wake of the planet. Oxygen atoms, with others, escape in
the tail
250 km but is absent on the night side of the planet. The magneto-
pause lies immediately above this at 300 km. Again, like the bow
shock, these distances vary with solar activity, somewhat counter
intuitively being highest when the Sun is most active. Between
these two boundaries there lies a magnetic barrier which prevents
any of the solar wind penetrating more deeply into the ionosphere.
Beyond the planetary terminator, on the night side of Venus,
the magnetic field of Venus is drawn outwards downwind to radii
exceeding ten times that of the planet (Fig. 5.3). It is here that ener-
gized oxygen ions, neutral oxygen gas and hydrogen ions, as well
as a dose of helium venting from the planets interior, ultimately
escapes into interplanetary space. As Fig. 5.3 shows gases from
the ionosphere can leave Venus on the night side of the planet.
When the Sun has very low activity and the speed of the solar
wind is low, gases within the ionosphere balloon outwards and
more escape into interplanetary space. At higher solar wind speeds
the ionopause wraps more fully around the planet and less gas is
removed. This is mostly oxygen that has been released from the
break-down of carbon dioxide, but hydrogen (from water and sul-
furic acid) and helium also escape here.
One of the surprises from the Venus Express mission was the
discovery of reconnection events. Here, within the magneto-
tailthe region downwind of Venus where the field is dragged
outthe magnetic field lines can flip backwards and launch par-
ticles back towards the night-side of Venus. Here, they can induce
air glow or even aurorae as happens at the Earth.
One problem with these observations is the obvious differ-
ence in the amount of water on each planet. While the Earth loses
approximately the same amount of oxygen as Venus per second,
this rate may not have been the same in the past. On Earth most
of the oxygen comes from water vapor that is wafted up into the
stratosphere and mesosphere. Here ultraviolet light splits it and
releases first hydrogen, then by a complex route, oxygen. On Venus
the present loss of oxygen comes from the splitting of carbon diox-
ide. Almost all of the water is gone. Carbon dioxide is much denser
than water vapor and clings more closely to the planets surface.
This property makes it harder to shift than lighter water vapor. At
best this implies that although oxygen losses are comparable, in
the past the lighter water would have been much easier to dispose
of than the carbon dioxide is today.
Overall the dominant process by which Venus loses atmo-
sphere is through a mechanism with the grand title electric force
field acceleration. High in the atmosphere, electrons are energized
by ultraviolet and x-ray photons from the Sun. As electrons are less
massive than other particles, they are less tightly held by Venuss
gravity and thus more likely to escape from the top of Venuss ion-
osphere. As these move outwards and captured by the Suns solar
wind and magnetic field, they leave behind a net positive charge.
This, in turn causes the remaining ions to repel one another, caus-
ing the upper atmosphere to balloon outwards and ultimately driv-
ing the ions outwards, where they too can be captured by the Suns
Venus 205
rasping wind. This, plus the effect of photochemical reactions

releasing oxygen and the drag of the solar wind, led to the steady
loss of Venusian air. However, as we shall see later, the rate of loss
is not as dramatic as it first seems.
Changes to Wind Speed at Venus
One of the surprising discoveries made by Venus Express where

long and short-term changes in wind speed. These indicate that for
utterly mysterious reasons the average wind speed at 50 km has
been increasing over the 7 years Venus Express was observing the
planet. When the probe arrived in 2006 the average wind speed was
300 km per hour. By the demise of the craft in 2014 wind speeds
had increased to nearly 400 km per hour. Igor Khatuntsev (Space
Research Institute in Moscow) and lead author of the Russian-led
paper to be published in the journal Icarus carried out one study
using the data from Venus Express, while an independent Japanese
group carried out a separate set of observations made from the
Earth. Khatuntsevs team calculated the changes in wind speed by
recording how cloud features in images moved between frames.
In all over 45,000 features were methodically tracked by hand and
more than 350,000 further features automatically compared using
a computer program. The Japanese team used a similar process,
using an automated process to monitor the position of different
cloud features over successive image frames and using some sim-
ple math to derive the velocities.
Over shorter periods of time both teams also recorded rapid
variation in cloud velocities with some features circumnavigat-
ing the globe in under 4 days, while a few orbits later the speed
had decreases so that over 5 days were needed for clouds to cover
the same distance. Such short term changes must reflect altera-
tions in the amount of momentum shared by Venuss core, man-
tle, crust and atmosphere. Changes in Venusian wind speed might
be caused by an exchange of momentum between the accelerat-
ing atmosphere and the decelerating planetary spin. Perhaps more
surprising is the possibility that the atmosphere may be becom-
ing spun up by an exchange of momentum between the Earth and
Venus as they pass one another in their orbits around the Sun. As
such, the Earth also exchanges momentum with its atmosphere.
Therefore, changes to the wind speed on Venus might even be

caused by changes in the circulation of our atmosphere. What a
weird thought.
Finally, as Venus was being observed the Sun was moving
from solar minimum to maximum: perhaps the increase in the
strength of the solar wind was pumping more energy into the
upper atmosphere. From here, the mechanism that is believed
to pump up super-rotation (and is described above) might simply
have been transporting the greater momentum that was present in
the solar wind to the lower atmosphere. If that is true, expect the
speed of the atmospheric winds to decrease with the decline in the
Suns output. However, if the pattern is linked to an exchange of
momentum with the planets surface, expect a much more grad-
ual change in the spin of the planet that runs independently of
the solar cycle. However, maybe we should think again: could the
spin of Venus (as a whole) change with the solar cycle? Does the
Suns magnetic cycle drive changes to the movement of its atmo-
sphere, which ultimately exchanges momentum with its surface?
Wouldnt that be an idea: Venusa planet driven by its weather.
Snow on Venus?
In 2012 Venus Express reported that along the day-night terminator
of Venus temperatures could fall as low as 175 C at an altitude
of 125 km (Fig. 5.4). Scientists, led by Arnaud Mahieux (Belgian
Institute of Space Aeronomy) determined the atmospheric pres-
sure and carbon dioxide concentration before calculating the tem-
perature. This very cold layer is sandwiched between two warmer
layers which should lead to the trapping of any carbon dioxide ice
particles and perhaps lead to the formation of clouds analogous
to noctilucent clouds, here on Earth. On occasion Venus Express
noticed bright patches of cloud along the limb of Venus and its pos-
sible that these are the proposed carbon dioxide ice clouds. Future
explorers might just see snow on Venusor more accurately high
above Venus. Clearly such a frosty phenomenon wouldnt stand a
snowballs chance in hell anywhere else on this planet (Fig. 5.4).
Yet, despite Venuss renowned extreme heat, there is another
possible location for snow on Venus, and this is on Venus. When
Venus 207
Height
(km)
Exosphere
500
Thermosphere
130
Carbon 120
Dioxide Ice 110
Clouds Mesosphere 100
90
80
70
Sulphuric Acid Haze
60
Sulphuric Acid Clouds
50
40
Troposphere 30
20
Sulphuric Acid Haze
10
-150 -100 -50 0 +50 +100 +200 +300 +400 +500
Temperature (oC)
FIG. 5.4 Unexpected phenomena in the Venusian atmosphere. Near the

daynight terminator (photo inset) temperatures fall to low enough tem-
peratures for carbon dioxide to freeze out at around 125 km up. Here,
carbon dioxide ice clouds might be dense enough to allow the formation
of dry ice snowfall. Purple straight lines indicate cosmic rays streaking
through the atmosphere. Although not yet clear these may create ioniza-
tion in the sulfuric acid clouds that ultimately charges them enough to
create lightning. Venus image from Venus Express (ESA)
Magellan mapped the surface during the early 1990s it detected

some highly reflective material on Venuss highest peaks. Although
no one would suggest conventional snow and iceeven at the rela-
tively cold surface temperatures of 380 C found at the summits
of these mountainsa peculiarly Venusian snow might just be
possible. Although youd be hard-pressed to go skiing on this pres-
sure cooked (45 bar) toasty material, what might be either elemen-
tal Tellurium or perhaps lead sulfide, would make the mountains
seem rather terrestrial in appearance. Of course, all of this has to
be envisaged through the peculiar magnifying lens of the lower

Venusian atmosphere, but nonetheless if you could see Maxwell
Montes rising upwards in the distance, its top gleaming in the faint,
reddened sunlight, you might at least be temporarily reminded of
the Earth (before Venus crushed, fried and dissolved you).
Lightning on Venus
On the Earth there are over 40 flashes of lightning per second,
amounting to over 1.4 billion flashes per year. On a particularly
stormy, early autumn day across Europe, in 2015, there were over
230,000 strikes. This rather large figure puts the Earth in one
of the most electrically active planets known. However, within
the Solar System at least three other planets are known to host
thunderstorms. Jupiter has been known to host thunderstorms
since Pioneer 1 soared past in the 1970s. Saturn also hosts a large
population of thunderstorms, primarily associated with specific
large storm systems that periodically dot the planets atmosphere
(Chap. 7). It is also likely that Uranus and Neptune will join the
fraternity but Venus remained a curious outsider. Well, it turns
out that it isnt (Fig. 5.5).
FIG. 5.5 Left above, a typical thundercloud (cumulonimbus) moving

briskly eastwards over Crieff in Scotland. The fibrous top part of the
cloud consists of ice particles, which develop a positive charge (Chap. 4).
The base of the cloud, which is largely hidden by the trees, retains a
negative charge, although this is grossly simplified. Above snow is fall-
ing from the anvil of another thundercloud, on a chilly late May day in
Scotland. Photographs by author
Venus 209
How do we know that Venus has lighting? Well, aside from the
possibility that it might be visible through a telescope or through
the eye of an orbiting satellite, most lighting would be invisible
as it occurs within Venuss dense, acidic and highly reflective
clouds. However, what clouds hide in visible light they cannot
hide at radio frequencies. Lightning produces very characteristic
short bursts of radio waves, lasting roughly 1 s. The effect is quite
unmistakable (Chap. 4). Using such radio frequencies, the Soviet
Venera probes obtained data that suggested Venus does indeed
have lighting. This was subsequently confirmed by observations
made at similar frequencies by Magellan and later Venus Express.
Venus Express also detected lightning through its effects on plan-
etary magnetism. Each discharge generated bursts of magnetism
associated with the movement of electrons along the lightning
bolt, and through the effects of magnetic fields associated with
the electromagnetic pulses (EMPs) lightning generated (Chap. 4).
Interestingly, around the Earth such EMPs also affect the magne-
tosphere surrounding our planet (Chap. 4). Lightning discharges
shove the terrestrial Van Allen radiation belts upwards creating
safe zones at altitudes of a several hundred kilometers. Although
Venus does not have the equivalent radiation belts, as it lacks
the equivalent magnetic field, it does still generate its own micro
magnetic shield as we have already seen. To recap this is formed
by the interaction of the solar wind with the thermosphere, which
generates a bow shock around the planet. This region is not only
influenced by the strength of the solar wind, but thus will also be
affected by EMPs generated deeper down within the clouds of the
planet. As Chap. 4 suggested, terrestrial and other lightning links
the atmospheres of planets to the broad galactic and intergalactic
magnetic fields. The effect may be slight but it is still a fundamen-
tal link between the cosmologically tiny (a lightning bolt) and the
vastness of the cosmos as a whole. Just imagine the beating of the
wings of a butterfly on Earth affecting the formation of a distant
star, by the subtle but measurable effect it has on terrestrial light-
ning and henceforth the planetary magnetic field.
That Venus had lightning was something, but the question
remained how a planet with an apparently dry atmosphere pro-
duces the separation of charge needed to synthesize lightning? On
the Earth, Jupiter and Saturn (Chap. 7) lightning forms in a similar
way with charges separating within large cumulonimbus clouds.

However, this mechanism cannot happen in quite the same way
on Venus as dry Venus does not host an icy cloud structure in
its main cloud deck: temperatures are just too high. For exam-
ple at the cloud tops in Venus temperatures are just about freez-
ing and the clouds are almost exclusively made of sulfuric acid.
However, the underlying force that produce lightning is still con-
vection as it is on the Earth, but with a generous hand from cos-
mic rays (Fig. 5.4). Christopher Russell (University of California,
Los Angeles) and colleagues have analyzed Venuss electrical dis-
plays from data acquired by Venus Express. This reveals that the
Earth and Venus are rather similar in terms of the strength of the
lightning. As on Earth the majority of lightning on Venus occurs
on the day lit side of the planet and convection within the reflec-
tive cloud layer is probably required to drive charge separation.
However, although cosmic rays certainly help generate electrical
charges within terrestrial clouds, they probably play a subordinate
role to charge separation on ice particles and water droplets. On
Venus, without ice, cosmic rays probably play the dominant role
in generating charge differences in Venuss cloudsperhaps with
an additional input from ultraviolet light (Fig. 5.4). Convection
may simply drive clouds to different altitudes, with those reaching
the greatest heights becoming the most strongly ionized. Recall
that the energy of cosmic rays attenuates (weakens) with decreas-
ing altitude. Therefore, cosmic rays have the greatest energies at
the greatest altitudes in the atmosphere.
Most lightning on Venus also occurs in its tropics (within the
120-wide strip flanking the equator). The principle difference is
that convection probably does not originate at the ground, as it
does on Earth, which is almost uniformly hot and poorly illumi-
nated. Instead the relevant convection is likely confined within
the cloud layer itself taking energy conducted and convected from
the dense, fluid-like layer of carbon dioxide below. The sulfuric
acid that forms the clouds is of low density in the deepest layers
of the atmosphere as here temperatures are so high that the com-
pound evaporates before it reaches the ground (Fig. 5.6).
On the Earth around 90 % of the lightning within the tropics is
between or within clouds, but this value drops as you approach the
poles. Over Norway roughly half the detected lightning is of this
variety, with the other 50 % striking the ground. The proportion is
Venus 211
FIG. 5.6 Fall-streak virgaa.k.a. jellyfish clouds. On a warm summers

day ice falls from mid-level altocumulus castellanus and trails behind
the clouds, which are moving on a modest south-east wind. The ice sub-
limates long before it hits the ground. Such clouds might be common on
Venus where sulphuric acid rain fall is both less dense than the carbon
dioxide sea into which it is falling and boils at 337 C, a temperature
reached above the summits of Venuss highest mountains. Virga photo-
graphed by author
set mostly by the level at which water can freeze. This is clearly
going to be higher over the Tropics than it is nearer the Poles.
On Venus we have no real idea where the lightning is, but the
majority is likely to be of the intra-cloud variety, in part because of
the effect described above but also because the cloud bases are so
much higher than they are on Earth. To strike the ground a much
larger voltage would be needed to overcome the insulating effects
of the greater depth of air. We know that on Earth, the much more
powerful super-bolts, which strike from higher up, are also much,
much rarer (Chap. 4).
Moreover carbon dioxide is a very poor conductor of electric-
ity, meaning that to strike the ground a bolt of electricity that was
generated within the clouds 40 km up would need to be extraor-

dinarily long and thus of an extraordinary voltage. This assumes
that Venusian air does not carry some form of highly conducting
gas or other particle and this is not easy to determine. However,
from terrestrial observations of our storms the effect of tempera-
ture should make most (if not all) Venusian lightning of the intra-
cloud variety.
Why Does Venus Have So Much More Car-

bon Dioxide than the Earth?
Appearances can be deceptive. When you look at Venus you see an
atmosphere dominated by carbon dioxide with very little nitrogen,
compared to the Earth. However, this is by comparison of percent-
ages alone. Remember that Venus has over 90 times the mass of
carbon dioxide than the Earth has, so that value of 3.5 % nitrogen
actually corresponds to nearly four times the mass of this gas in
Venuss atmosphere compared to that found on Earth. As for the
carbon dioxide, recall that on Earth most of this is taken up as
carbonate rocks (Chap. 3) leaving only 1/250,000 the total mass of
carbon dioxide as free gas. Solid carbonate rocks form when carbon
dioxide and water interact with dissolved ionsmostly calcium
and magnesium. Thus, on the Earth this process, known as seques-
tration, has accounted for the loss of most of our atmosphere to
the surface. Of course, sequestration has also removed water vapor
and carbon dioxide. On the Earth if you could resurrect this carbon
dioxide as gas, the Earth would have roughly three times the mass
of carbon dioxide in its atmosphere as is found on Venus: the Earth
would have a thicker atmosphere than Venus has.
Constructing a Dry Planet

When we look at water and the Noble Gases something else
becomes obvious. Clearly Venus is very dry indeed and the differ-
ent noble gases reveal that much of Venuss original quotient of
some of the noble gases such as Argon. Quite apart from assuming
that the Earth and Venus might have started out roughly the same,
Venus 213
with the same budget of water as our world, there is clear evidence
that Venus has lost almost all of its water and would therefore
have originally been a rather wet planet just like the Earth. The
clue to this planetary dehydration is the ratio of two isotopes of
hydrogen. If you recall from Chap. 2 chemical elements come in
a variety of flavors called isotopes. Some are heavier than others
and thus easier or harder to remove by evaporation or by other
processes. For example as Chap. 2 showed, oxygen comes in three
flavors of different masses: oxygen-16, oxygen-17 and oxygen-18.
Oxygen-18 has two more neutrons than oxygen-16 making it 2 g
heavier for every 600,000 million trillion atoms that are present.
Likewise hydrogen comes in three flavors: hydrogen-1 (boring old
hydrogen); hydrogen-2 (or deuterium); and hydrogen-3 (tritium).
Tritium is unstable and radioactively decays in a short time, so
we shall ignore it; but deuterium is completely stable and roughly
twice the mass of bog-standard hydrogen-1. Because it is heavier it
is harder to shift from the atmosphere than hydrogen-1.
Now, deuterium and hydrogen are normally found in a
very similar ratio throughout the universeor at least in water.
Therefore, if that ratio is very skewed in favor of deuterium it
tells you straight away that there was a lot more of the original
hydrogen-1 that has now been lost. Quick calculations indicate
that Venus has lost more than 99.9 % of its original store of hydro-
genand most of this would have been found as water. The con-
clusion is clear: Venus was just about as wet as the Earth but then
lost almost all of its water. At present, if we assumed that the
amount of oxygen we observe leaving Venus is from water, as it is
on Earth, and to a lesser extent on Mars, then Venus would only
have lost about 8 cm depth of water if we assume an ocean had
completely covered its surface. For comparison the Earth would
have lost about 9 cm in depth, with Mars losing around 30 cm: the
difference in depth of water lost is a reflection in the difference in
surface area of each planet. The problem is this isnt enough. Venus
has lost far more water than the Earth and clearly something much
more dramatic than the current rate of loss must have occurred.
In Chap. 3 we compared the early histories of Venus and the
Earth. Both planets likely began life in the same way and both
likely had identical (or near identical) conditions for the first hun-
dred million years: a hot pressurized greenhouse. From here on
two models are possible. In the hot early scenario Venus stays
hot and any early ocean is unstable and soon succumbs to evapo-
ration. In the cool early scenario Venus, like the Earth becomes
cool enough, for long enough to form stable oceans and run plate
tectonics.
In the hot early model water vapor is held in the atmosphere
as a hot gas for tens of millions of years, with much of it rising
upwards from a high tropopause to the stratosphere. Here the
more intense ultraviolet light of the Sun (both because Venus is
closer to the Sun and because the early Sun released more ener-
getic, ultraviolet-rich flares and a stronger stellar wind) steadily
broke the water vapor down to liberate its hydrogen. This light
gas steadily boiled off into space leaving a more oxygen-rich
atmosphere. After maybe, 100500 million years all of the water
was lostthe hydrogen to space and the heavier oxygen largely
retained. What happened to that gas remains a mystery, but pre-
sumably it reacted with materials on Venuss surface or sulfur-rich
gases in its atmosphere. What was left was an increasingly dense
and hot carbon dioxide atmosphere. The final straw for Venus
might have ironically come from the failure of plate tectonics.
Measurements taken by the Visible and Infrared Thermal
Imaging Spectrometer (VIRTIS) on Venus Express made in the
infrared portion of the spectrum shows that the rocks on the
Phoebe and Alpha Regio plateaus are lighter in color and look old
compared to the majority of Venuss surface. On Earth, such light-
colored rocks are usually granite and form continents; and because
granite is a lot less dense than basalt it floats higher in the man-
tle than basalt. The presence of granite would naturally explain
why these regions are the only noticeable highlands on Venuss
surface. As was suggested in Chap. 3 that parts of Venuss crust
appear to be granite implies that Venus held oceans for at least
a few hundred million years. That such continental crust cov-
ers only a trivial proportion of Venuss surface also implies what
oceans Venus once held must have been transient and probably
lost within one billion years. This hotter version of Venus might
well have hosted early life, giving rise to the possibility that early
in the Solar Systems history there were three Earth-like worlds: a
steamier Venus, a Goldilocks porridge-like Earth and a more frigid
and possibly effervescent Mars.
Venus 215
However, by one billion years, Venus was likely overheating;

its oceans on the wane and the atmosphere steadily filling with
water vapor and carbon dioxide gas. Whether this led to a thermal
runawaya thoroughly unpleasant phase where water vapor drive
the temperature to over 8001000 Cis unclear. As temperatures
rose to over 70 C then 100 C the remnant oceans would have
boiled away while the air filled with a torrid sea of broiling thun-
derclouds. Extending up to 100 km in height, these monstrous
storms pumped water vapor high into the stratosphere where
ultraviolet light broke it down. While the Earth maintained a cold
and stable lower stratosphere, Venuss was overwhelmed by vigor-
ous convection extending from the hot surface below. Although
ozone would have been forming from the oxygen that was released
from the split water, this would not have been sufficient to shield
the water below. Once so much water is split the hydrogen that is
released limits the stability of ozone and the stage was set to dry
out the planet. The escaping hydrogen can also drag the oxygen
(and ozone) away with it in a process called hydrodynamic drag.
Thus by 3.54 billion years ago Venus was drying out and
perhaps initially cooling off a little. As water vapor split the
greenhouse effect may have waned slightly and the clouds cleared
revealing a hot desert-like planet. However, such relief would
have been temporary. While much of the planets carbon dioxide
was likely locked up in carbonate rocks on the planets surface, the
steady output of volcanic activity would have once more caused
the levels to increase in the air. With every extra gigaton the tem-
perature would have increased, even as the abundance of water
vapor declined.
The question mark lies with the state of Venuss volcanic
activity. With a similar mass and energy budget to the Earth it
should have retained active volcanism. However, as the oceans
vanished and the crust and mantle steadily dried out the process of
tectonic activity would have soon ground to a halt. What happens
next is speculation. With so much internal heat, volcanism cant
simply stop as not enough heat would have been lost through the
crust by conduction. While hot spot volcanism would have con-
tinued, an unstable situation was developing that would finally
condemn Venus to the state of a torrid hell. Without plate tec-
tonics to cool the mantle it would soon have overheated. As well
as causing extensive melting of the uppermost layers, the crust

would have begun to soften and buckle. At least once in the inter-
vening three billion years the thick Venusian crust appears to have
foundered. On Earth plate tectonics depends on the formation of
a rock called eclogite. This metamorphic rock forms from basalt
that is heated and compressed around 100 km under the surface.
This transformation only happens nowadays when the crust is
cooled, thickened and buckles downwards. On Venus this process
is unlikely as the crust is too hot and dry. However, if the crust
thickens up the base of it might become transformed into eclogite
at its base and the resulting denser slab of rock would then sink
directly into the mantle. Such a process might have happened on
the early Earth when the mantle was much hotter and volcanic
activity led to the formation of a thicker crust.
Why is this important? Well, if Venus developed a thick crust
that then foundered into the mantle the process would have rapidly
consumed the entire basaltic crust of the planetprobably leaving
the island continents that formed early on. Just imagine the scene
unfolding. Initially volcanic activity steps up as the mantle begins
to overheat. Copious volcanic activity begins to resurface the
planet, adding a lurid red glow to the already superheated world.
After a few million years the crust has thickened up and begins to
founder under its own weight, plunging catastrophically into the
mantle. Piece by piece the crust would crack and whole segments
would begin a rapid subduction under a fresh, thinner, molten sea
of lava. With the atmosphere already hot, the extra input of heat
would raise temperatures further. Worse still, as volcanic activity
increased the amount of water vapor and carbon dioxide, tempera-
tures would have begun to sky-rocket. However, that wouldnt be
the end.
On the surface of the foundering basaltic crust would have
laid thick deposits of carbonate rocks, created while Venus still
had oceans. As this was dragged down into the mantle with the
foundering basaltic crust, it would have broken down releasing its
store of carbon dioxide, as well. At this point the oceans worth
of carbon dioxide would have been vented to the atmosphere.
While initially, the wet greenhouse might have exerted roughly
100 times the surface pressure, this would have decreased to much
more manageable levels as sunlight broke the water vapor down to
Venus 217
oxygen and hydrogen and the oxygen became chemically bound to

rocks through chemical reactions. However, as volcanicity stepped
up, and in particular the massive amount associated with resurfac-
ing got under way, the atmospheric pressure would have increased
once more as massive amounts of carbon dioxide accumulated. In
turn this would have enhanced the greenhouse effect until tem-
peratures may have once again reached 1000 C. If they did get
this high, any remaining carbonate rocks would have decomposed
and released their store of carbon dioxide, adding a further wave of
greenhouse gas to the atmosphere.
The only saving grace for Venus would have been the increas-
ingly dense burden of clouds. If Venus ever had a largely trans-
parent atmosphere (to visible light) it would have rapidly clouded
over once all of this volcanic activity kicked in. Clouds would
have reduced the amount of solar energy that made it through
and allowed temperatures to fall to their present levels. At some
point around 500 million years ago, most of the volcanic activ-
ity declined to its current, apparently, low levels. The atmosphere
then stabilized at around 95 bars of pressure with a temperature
around 450500 C, remaining there until the present day.
Today, the largely dry planet is still venting hydrogen.
Although much of this may come from the photolysis of sulfu-
ric acid some of it is coming directly from the break-up of the
small amount of free water found in the atmosphere. The Venus
Express magnetometer instrument (MAG) detected hydrogen gas
being stripped from the day-side. Until Venus Express arrived at
the planet, this process had been suspected but there was no real
evidence that it occurred, with most hydrogen thought to exit the
planet through its magnetotail. It is clear that Venus is still drying
out todaymost likely this is residual water from the mantle that
was dumped there during the planets formation or subsequently
during the brief era of plate tectonics.
So, how much mass is Venus losing from its atmosphere now?
And perhaps more pertinently how long will it take the solar wind
to strip poor old Venus bare? You might imagine that the relatively
exposed Venus might not be able to hold onto its atmosphere very
well. However, as weve seen the combination of high atmospheric
mass with a dominance of the relatively weighty carbon dioxide,
and the protective effect of the bow shock and ionopause means
that Venus currently loses only one metric ton per hour (roughly
8,700,000 kg per year). That might sound a lot but with an atmo-
spheric mass of roughly 500,000 million trillion kilograms it means
that Venus can retain a reasonable atmosphere for the best part
of 55 trillion years. The Earth has a similar mass loss rate and if
nothing else were to change the Earth would be stripped bare in a
comparatively trivial 591 billion years. Of course, long before that
happens both planets will be vaporized by the expanding Sun
Life in Hell?
Heres an interesting idea: Venus is still habitablealbeit not on
its hellish surface. Recall, that at about 55 km up the temperature
is around 2530 C with a pressure comparable to sea level on
Earth. Moreover, photolysis of carbon dioxide gas (the chemical
break up of carbon dioxide caused by ultraviolet light from the
Sun) also releases fairly reasonable amounts of oxygen. In principle
life could survive in these conditions as long as they didnt mind
the low pH and the relative dryness. On Earth bacteria are known
to survive and reproduce within our cloudsand many species of
bacteria survive hot, acidic conditions. Thus, maybe, just maybe,
there is life in the middle atmosphere of Venus.
Future Venus: The Earth and Venus,

Twins Once More
Although Venus and the Earth are very different now, they both
began their lives with hot, carbon dioxide and water vapor-rich
greenhouse atmospheres. Perhaps for the first 50100 million years
these encased a molten ball of rock and metal. However, soon after
100 million years had passed the surface would have solidified and
lowering temperatures would have allowed water to begin falling
from the skies (Chap. 3). Under a dense carbon dioxide-rich atmo-
sphere, water began to accumulate in oceans that would then have
begun to pull down the carbon dioxide.
As we have already suggested, perhaps for a few hundred
million years, both the Earth and Venus presented a deep blue,
Venus 219
clouded face to the universe. However, as soon as Venus began to

lose its oceans the planet would once again have clouded over and
the long, dark separation of these two worlds would have begun.
Yet despite this schism, the parting is only going to be temporary.
At some point in the next two to three billion years Earth will
pursue its twin along a path to hell. Before the Earth succumbs
to this fate it will pass through a process of slow decline where
falling carbon dioxide, but rising temperatures force the demise of
first land plants then land animals.
800 million to 1 billion years into the future carbon dioxide
levels in the Earths atmosphere will have fallen because the gas
will have been absorbed into the oceans and into the planets crust.
Less and less carbon dioxide will be returned by volcanism, as this
too declines with the secular cooling of the planet.
While photosynthesis in higher plants will cease when the
concentrations of carbon dioxide fall lower than 10 parts per mil-
lion (0.01 %), bacteria in the oceans will continue to operate pho-
tosynthesis and generate oxygen as a byproduct for a long time
thereafter. As long as life can persist, feeding on these bacteria
life on Earth will survive. I speculated in Under a Crimson Sun
that plants could survive if they mined carbonate rocks for car-
bon dioxide. At present most plants (and their symbiotic fungi)
actively secrete acids and other chemicals into their environment.
In principle, any plant growing on carbonate-rich rocks, such as
limestone, or lime-rich soils could capture carbon dioxide from the
rock using acids secreted around its root system. There certainly
would be evolutionary pressure for such a move, but whether it
would be sufficient to maintain a terrestrial biosphere is anyones
guess. A more grisly alternative is that plants take up carnivo-
rous lifestyles, obtaining their carbon from animals. In terms of
the biosphere, they couldnt all adopt this change as this would
be akin to a living perpetual motion machine with animals sur-
viving by eating plants and plants surviving by eating animals.
Thermodynamics will step in and prevent such practice becoming
widespread amongst plants.
More importantly for our planet will be the ever-brightening
Sun. At one billion years from now the Sun will be 10 % brighter
than it is now. This will be sufficient to raise the temperature of
the planet to around 30 C, or a good 7 C higher than its been

in any previous greenhouse phase, such as the Permian-Triassic
boundary. The planet will be then be ice free for the first time in
around 5.4 billion years and much of the continental surface is
likely to be under water. This state of affairs will be exacerbated
if plate tectonics has ceased and mountains have largely eroded
down. Venus, meanwhile will hardly notice the difference with its
thick, highly reflective atmosphere keeping the increasing glare of
the Sun at bay.
While Venus will remain largely unscathed for the ensuing
few billion years, the Earths biosphere and atmosphere will be
forced to change. Its at this one billion year mark that two differ-
ent fates emerge. It was thought that the most likely fate would be
a rapid rise in temperature over a few tens to hundreds of millions
of years. This could result in temperatures escalating to the point
at which the surface of the Earth was completely melted. This
was because if water vapor was allowed to trap sufficient energy
the surface temperature would need to rise to the point at which
it was hot enough to release energy at visible wavelengths. Then,
and only then, would the atmosphere be able to release energy
trapped by water vapor. However, in new work, described recently
by Jrmy Leconte and co-workers (Laboratoire de Meteorologie
Dynamique,) is that the Sun will begin a protracted but accelerat-
ing boil-dry of the oceans. This will likely take over a billion years
to complete. With the planet at an average temperature of 35-45 C
water vapor will act much more effectively as a greenhouse agent.
Remember that 45 C is an average for the planet, which means
that some areas will see temperatures well above 70 C. As more
water evaporates the planetary greenhouse will strengthen and so
temperatures will rise, further accelerating the effect. In this sce-
nario the oceans, and with them life, will be eliminated by 2-3
billion years into the future.
The fate of the water that boils away will mirror events on
Venus perhaps five or six billion years before. At present the Earths
water is largely contained within the troposphere by the temper-
ature inversion at the tropopause: although the number seems
large, only a small amounts of water escapes into the stratosphere.
This is then split by ultraviolet light to hydrogen and oxygen.
Venus 221
From here, somewhere in the region of 95,000 metric tons of hydro-

gen is then lost to space, per annum. Most of the accompanying
oxygen retained because it is heavier. A comparison of terrestrial
isotopes of hydrogen in some archean rocks with those in pres-
ent day equivalents implies that the Earth has lost nearly a quar-
ter of its water since it formed. Fortunately, above the Earth the
stratosphere is cold and hence dry. Moreover, as we saw in Chap.
1, above the tropopause air begins to warm with height because
of the absorption of ultraviolet light by ozone. This gas is con-
centrated within the stratosphere. Such a warming trend blocks
convection from carrying moisture any higher and thus preserves
most of our planets riches in the troposphere, where it is shielded
from ultraviolet light by the abundant ozone above.
However, as happened on Venus when the terrestrial tropo-
sphere warms the lid of the tropopause is forced upwards until
the moist air below becomes accessible to much more abundant
and energetic ultraviolet light. As increasingly energetic and
intense ultraviolet radiation splits the Earths water, hydrogen
will drift off into space and be blown away in a majestic comet-
like apparitionthough one visible only if you could see in the
ultraviolet. As with Venus, the Earth could, in principle develop
an oxygen-rich atmosphere. However, whether this happens
will depend very much on how much material is available on
the planets surface to oxidize. This will include dead and dying
organisms, as well as any sulfur-rich compounds thrown out by
our planets declining volcanic activity. As with Venus a consid-
erable amount of oxygen will also be lost from the upper atmo-
sphere, either dragged out by escaping hydrogen, or split into
single oxygen atoms and ions that can be grabbed by the solar
wind and escorted into the deep, near vacuum of space. Recent
work by James Kastings (Pennsylvania State University) and col-
leagues suggests that the critical surface temperature is 350 K
(77 C). At this point the troposphere leaks so much water vapor
that the stratosphere effectively becomes saturated and warms
up, thus allowing all of the planets water to escape.
By the time the planet has aged a further two billion years, in
this slow-cook scenario, the Earth will have boiled dry. Without
water to sequester carbon dioxide gas, this greenhouse gas will
begin to accumulate once more within the atmosphere. Although

our water will be split by ultraviolet light at a pace sufficient to pre-
vent the complete meltdown of the planets surface, temperatures
will begin to rise inexorably to levels approaching the contem-
porary Venus. Although this will avert a catastrophic greenhouse
effect, any life that remains in the crust will still be cooked out of
existence.
One interesting consequence of these newer models is that
Venus was more likely to have held onto oceans for a significant
part of its early life. Instead of an aggressive, early thermal run-
away (Fig. 5.7), Venus would have had a more benign climate and
possibly life for up to a billion years or so from its formation.
Early hot, 1A 1B 1C 1D
wet
atmosphere
overlies
Venus
pressurised
ocean 2A 2B 2C 2D
-4.5 -4.0 -3.0 -2.0 -1.0 Today 1.0 2.0 3.0 4.0 5.0
Billions of Years
Early hot, 3A 3B 3C
wet
atmosphere Earth
overlies
pressurised
ocean 4A 4B 4C
FIG. 5.7 Key events in the histories of Earth and sister Venus. 1ATran-
sient deep oceans evaporate; 1BRunaway wet greenhouse; 1CWater
lost to space and dry greenhouse begins; 1Dglobal resurfacing event(s)
fill atmosphere with carbon dioxide and water establishing modern
Venus. In the alternative model early on Venus has cool oceans (2A).
These survive for at least a few hundred million years before they begin
to evaporate (2B). During this time plate tectonics makes Venusian high-
lands. In 2C a hot, wet greenhouse is established before the water is lost
to space; 2D global resurfacing occurs as in 1D. By contrast the cooler
Earth maintains oceans after an early, hot, wet greenhouse (3A/4A).
These persist until 3B when a thermal runaway occurs or much later in
4B when they have evaporated. In 3C and 4C a final hot, dry greenhouse
is established, finally pairing the Earth and Venus once more
Venus 223
The Earth can maintain habitability for longer if the Earths

magnetic field falters in the next billion yearssomething that could
happen when plate tectonics ends in the intervening time2the
Earths atmosphere could be exposed to the erosive power of the solar
wind. Now, we already know that this stellar breeze has minimal
impact on the planets atmosphere at present. If the rate of removal of
gas (and the water vapor will do) can keep pace with the brightening
Sun, the global greenhouse effect might be kept largely in check for a
little longer, ensuring that the period from 2 to 2.8 billion years from
now is more habitable than in the simplest scenario. With significant
reserves of gases and liquid water, a steady loss of these volatile mate-
rials readily lowers the atmospheric pressure, which in turn weakens
the greenhouse effect. You may recall from Chap. 3 that increasing
the concentration of carbon dioxide enhances its greenhouse effect
because at higher concentrations particles of gas collide and cause a
broadening of the range of wavelengths over which the gas can absorb
infrared (heat) radiation. Therefore, if we can remove gas from the
atmosphere (mostly nitrogen) then the overall pressure will decrease
and with it the strength of the planetary greenhouse.
During this extended phase the Earth would steadily dry out,
but sufficient water could be maintained on the planets surface to
keep it habitable up to three billion years into the future. Beyond
this point the atmospheric pressure would fall to less than a third
its current level until liquid water became unstable at what would
still be higher temperatures than are found on average today. What
remained of the oceans would evaporate and the planet would
finally run dry. Venus would have been unable to pull off this trick
for two reasons. Obviously, being closer to the Sun it grew hot-
ter faster than the Earth until its greenhouse became unstable.
However, the geology of Venus also played its part. If Venus had
been geologically inert with little volcanic activity, the brighten-
ing Sun might have split and driven off sufficient water to keep
the planet cool and thus habitable for longer. However, while
the young Earth stayed cool, with its greater distance to the Sun,
Venuss atmosphere was soon overwhelmed by a combination of
2
Plate tectonics may be essential for the maintenance of a global magnetic eld. Dense cold sub-
ducting rock might stimulate convection in the planets core, which in turn might drive the forma-
tion of the magnetic eld.
evaporating water from its oceans and the constant addition of

more carbon dioxide and water vapor from its abundant volcanoes.
Venus never stood a chance.
Yet, if the Earths volcanism is sufficiently subdued a billion
or so years hence and the loss of water can keep pace with the
brightening Sun, the Earth might cling onto a faltering habitability
until 23 billion more years have elapsed. Certainly the Earth will
be hotter and drier than we are used to, but lifequite abundant
and complex lifecould hang on.
For the sake of argument lets stick with the more benign fate
for our planet as that gives us three billion years to play with. How
will the Earths climate systems alteraside from getting warmer
and drier; and at what point will the Earth and Venus come to resem-
ble one another once more? Aside from the ever-growing influence
of the Sun the next most important factor will be the spin of the
Earth. This is coupled to the orbit of the Moon, with the exchange
of momentum between the two driving the Moon outwards at
roughly 4 cm per year. This process will continue until the length
of a terrestrial (solar) day equals 1 month. At this point the moon
will orbit the Earth around 540,000 km away. Meanwhile this
gain of momentum by the Moon means that the Earth must be
losing the equivalent amount, as momentum is always conserved.
For our planet, the effect is causing the Earth to rotate more slowly:
roughly 1.7 milliseconds are added to the length of the day each
century. Were the Earth fully solid this value would be greater, but
after the last Ice Age the mantle rebounded pulling mass inwards
from the equator towards the poles, and this has caused a slight
acceleration, partly off-setting the effect of the retreating Moon.
Thus, the Earth will gain around 4 h to its day over the next bil-
lion years. Deceleration will continue after this point, until, as was
said above, the length of the terrestrial day matches its month.
However, the Earth will be truly dead by that point
A decelerating Earth means a weaker Coriolis Effect and more
sluggish westeast movement of the air in response to north
south differences in pressure (Chap. 1). When the Earth formed
the planet likely rotated with a period of around 56 h. During the
Proterozoic snowball (Chap. 3), a terrestrial day was approximately
21 h long. That is known with reasonable precision because tidal
deposits are preserved in some sedimentary rocks that were laid
Venus 225
down at the time. The to-ing-and-fro-ing of the water left very

characteristic patterns in the deposits which allow the length of
day to be established. Although the rate of slow-down decreases as
the Moon pulls away from the Earth, the length of the terrestrial
day will increase and with it the Coriolis Effect will decrease.
As the Coriolis Effect weakens the strength of terrestrial
winds will decrease over time and will blow more northsouth (on
average) than they do now. This will facilitate a greater exchange
of heat between tropics and poles, ensuring that the Polar Regions
are more thoroughly heated than they are at present. This, coupled
to a brighter Sun will make the planet more thoroughly cooked
throughout. Embedded within the weaker westeast flow of air
will be a grossly expanded pair of Hadley cells. A hotter planet
with less spin will allow these to reach ever closer to the plan-
ets poles, compressing the more westerly mid-latitude flow of air
towards the poles. You can see this on Venus where the equivalent
Hadley cells extend to 60 north and south of the equatortwice
the width of the domain on the Earth. On Venus the belt of west-
erly winds is confined to a narrow collar around the Poles and
this is the way the Earths winds will come to resemble when the
planet has grown hotter and spins on its axis more lethargically.
Inside the domain of the Hadley cells tropical storms will still
generate assuming there is a sufficient width of ocean to supply
them with moisture. Although tropical storms rely on the Coriolis
Effect to gain spin, with more territory over which to form and
more heat energy available than it is today, it might mean that
much more of the planet will be afflicted by such tropical monstros-
ities. In Chap. 3 we saw how human activities already appear to be
having this effect on the distribution of such storms, with a north
and south migration of maximum storm intensity of 5060 km per
decadethis matching the steady expansion of our planets Hadley
Cells. Whether the tropical storms of tomorrows Amasia will be
the near-legendary mega-storms of the popular scientific media,
remains to be seen, but the storm intensity will be a product of the
temperature, distribution of land and the overall spin of the planet.
More heat means more energy for the storm; while a slower spin
means a weaker Coriolis Effect and less spin. Before the Earth boils
dry tropical storms might just rule the meteorological roost over
much of the planet.
From two to three billion years any residual volcanic activity

would slowly refill the atmosphere with carbon dioxide and the
greenhouse would accelerate. Although unlikely to get as hot as
contemporary Venus, given that the Earth is further from the Sun,
it would still be hotter than the boiling point of water. In this des-
iccated state the planet would remain until the Sun left the main
sequence, expanded into a red giant then blew the atmosphere
away. This process would take around 700 million years, until the
swollen Suns surface came so close to the Earth that tidal forces
pulled the Earth in to be vaporized 200 years later. En route, the
Earth may get a rather interesting atmosphere. When the Earth
enters the last few million years of its existence, the solar surface
will be less than one million kilometers from the Earths surface.
At this point, the Moons orbit will take it half way to the surface
of the expanding Sun. Perhaps it will come to resemble to a flam-
ing molten orb, with a cometary tail of vaporized rock periodically
showering the Earth as it swings to and fro around the molten orb
of the Earth.
Although the dying Earth will rotate slowly it will still spin
on its axis with a day length perhaps 100-200 h long. Finding a
definitive figure for the length of the Earths day is rather tricky
but were the Earth to survive the Suns red giant phase the day
length would eventually max out at 1000 h long in 50 billion years.
Immediately prior to the red giant phase the Sun will be
around three to four times its current luminosity, depending on
where you draw the line for the Suns main sequence. This will be
more than enough to make the Earth a close twin of the present
Venus. Surface temperatures will easily exceed 500 C. However,
unlike the Venus of today the Earths atmosphere could be rel-
atively cloud freeif volcanism is very subdued on our rather
decrepit planet and little in the way of aerosols are available. The
slowly rotating Earth may have a dense but clear carbon dioxide-
rich atmosphere that stirs slowly. This atmosphere, free of reflec-
tive clouds will absorb and retain much more heat than Venus
does today and thus be proportionately hotter. Indeed, both Venus
and the Earth should have broadly parallel geologies at this point
meaning that Venus and the Earth could appear nigh on identical
from afar. One significant difference is that the Earth actually has
more carbon dioxide gas in the form of carbonate rocks than pres-
Venus 227
ent day Venus. Most of these rocks are held at relatively shallow
levels within the crust. As such, once global temperatures exceed
850900 C these carbonate rocks will break down (through ther-
mal decomposition) releasing their store of greenhouse gas. As
such there may come a time 34 billion years from now when the
Earth is as hot, or hotter, than Venus at a comparable time. Rather
than vying for habitability, Venus and Earth will vie for inhospita-
bility. Aside from a few hardy fossils there will be no evidence that
the Earth ever hosted life, never mind intelligent life.
However, if the Solar wind has had its way with our planet,
the terrestrial atmosphere will remain relatively thin and the
greenhouse effect proportionately weaker. Although still hot, it
wont be hot enough to melt the surface. However, that (relatively)
benign phase wont last much longer; for six billion years hence,
the Sun will begin its final expansion and begin to melt the Earth
and Venus.
Some seven billion years from now, the Earth and Venus will
face-off with a Sun that is nearly 260 times its present diameter
and around 2000 times as luminous. For the final five million
years of its first red giant phase the Sun will expand most rapidly,
vaporizing first Mercury then Venus. Both of these inner worlds
will be directly swallowed, however, the Earth has a different fate.
Although the Earth will likely be spared direct annihilation, the
word spared will be a relative term. Basking, or perhaps bast-
ing, under a glare that raises the surface temperature to around
2000 C might just permit the kind of hellish weather seen on The
Chronicles of Riddicks planet Crematoria. If you havent seen
the film, Riddick has to escape his underground dungeon and run
the surface to the nearest spacecraft, before sunrise. The emerg-
ing Sun not only blasts the surface with the hellish light, but also
drives a superheated wind of vaporized rock and debris across the
surface. Riddick, obviously escapes, but not before two of his col-
leagues have been turned into toast. Imagine sunrise: a vast red orb
rises to fill nearly 70 % of the sky. The ground is a molten soup
of rock, cast against a stark, black void above. As the Sun rises
slowly to its zenith the molten stew at your feet starts to boil;
not the whole surface but rather the lighter components. A thick
dense wind of vaporized rock lifts upwards and flows in a massive
thermal tide across the daynight terminator and into the dark
side of the planet. Here, under a chill as deep as 180 C the rock
condenses and falls as a molten rain or a powdery, ashen snow.
Although unlikely, if the Moon has been ripped away at this
time, then the Earth will tidally lock to the Sun. Then the constant
erosion of the surface will strip away the crust on the exposed
sunlit side. Progressively deeper layers of material will boil away,
only to rain out on the dark, cold side of the battered Earth leaving
the Earth resembling a lurid version of the contemporary Moon,
with a thin molten crust on its Sunlit side and a thick ashen crust
hidden in the perpetual night of its dark side. Unlike the present
Moon, there really will be a dark side of the Earth. The Earth will
be a bit of a mess, really.
Interesting though this fate might be much more likely is
that the Earth will retain the Moon in a high, but, now, increas-
ingly unstable orbit. As the Earth and Moon boil, the pair will
have a brief (and fatal) final encounter. As the Moon orbits the
Earth the denser solar wind blowing outwards will act like molas-
ses on the orbital motion of our Moon. This will slow the Moon
down, which, in turn, will cause it to spiral inwards. Interestingly
as the two move ever closer, and the Moon speeds up, tidal forces
between the molten Earth and Moon will spin the Earth ever faster.
In the process, increasingly vast tides of molten rock will slosh
across the Earths surface. When the Moon reaches 11,400 km or
so above the Earths surface, the Earths tides acting on it will tear
it apart. For a brief moment a shower of molten rock will splash
across the heavens. Under the glare of the incandescent Sun, this
molten ring of material will gradually boil it awaya fate that
mirrors that of the Earth, soon, thereafter. Once again, as with
nearly every section of this chapter, there is an alternative. Should
the Suns bloated surface drag on the Moon more strongly than the
Earth, the Moon might get pulled off its orbit around our world
and into the depths of the Sun, before its orbit shrinks causing it
to collide with the Earth. In which case, the fate of the Moon will
simply presage the fate of the Earth a few million years later.
Earths last weather a blisteringly hot rain of molten rock
and ash will succumb to an increasingly dense, though no less
torrid, fog. Timing is everything: Venus will last until around five
million years before the peak of the first red giant phase; while the
Earth will probably last for another 4.5 million years after Venus
Venus 229
is vaporized. However, at some point, around 500,000 years before

the end of the first red giant phase, the Earth will begin a 200 year-
long death spiral into the Suns outer layers. Temperatures will
steadily rise until all of the rock begins to boil. Streaking, invis-
ibly downwards towards the Suns core, the cometary remains of
the Earth will follow Venus and dissolve into nothingness, adding
roughly 0.01 % extra heavy elements to its bulk. Only the sullen,
molten face of Mars and the hot gaseous orbs of the outer giants
will be left to witness the fate of the Suns twins.
Conclusions
Venus and the Earth: the twins that arent, but once were and will be
once more. This terrestrial pair are so similar, in terms of mass and
overall composition. However, thanks to a subtle difference in the
location of their births, both worlds rapidly diverged early in their
histories. Whether the Solar System hosted two Earth-like worlds
(and possibly a third in Mars) remains unclear. There are tentative
signs that Venus was cool and wet enough for at least a few hun-
dred million years of its history to form small continental regions.
However, it might be that the earliest ocean or oceans of Venus
were simply pressure-cooked pans of boiling water that resided tran-
siently before the greenhouse effect lifted them into outer space.
What is clear is that by 23 billion years ago Venus was boiled dry
and the modern greenhouse atmosphere began to take shape.
Contemporary Venus paints a portrait of the future Earth. At
some point between one and three billion years into the future
the ever brightening Sun will catch up with the Earth and drive
its oceans into out space. Steadily rising solar luminosity and
increasing carbon dioxide levels from residual volcanic activity
will raise the global temperatures until they approach the melting
point of the crust. Although this could be a fairly protracted affair
for our world, if the Earth can hang onto enough water vapor its
atmosphere will superheat much earlier (11.5 billion years from
now). Either way, a global greenhouse effect will bring the atmo-
spheres and histories of the Earth and Venus back into line with
one another. For the remaining few billion years Venus and Earth
have left, they will present a hot, but increasingly cloud-free face
to the universe. These twins will continue to overheat until they

both melt under an increasingly severe glare from our aging star.
Both worlds will then end their lives, with the Earth follow-
ing Venus into the Suns fiery maw roughly seven billion years
from now. The twins will be no more.
References
1. Mller, N., Helbert, J., Hashimoto, G. L., Tsang, C. C. C., Erard, S., Piccioni, G., Drossart, P.
(2008). Venus surface thermal emission at 1 micron in VIRTIS imaging observations: Evidence
for variation of crust and mantle differentiation conditions. The Journal of Geophysical
Research, 113, Issue E5, doi:10.1029/2008JE003118.
2. Ingersoll, A. P., & Dobrovolskis, A. R. (1978). Venus rotation and atmospheric tides. Nature,
275, 3738.
threshold for runaway greenhouse processes on Earth-like planets. Nature, 504, 268271.
4. Kasting, J. F., Chen, H., & Kopparapu, R. K. (2015). Stratospheric temperatures and water loss
from moist greenhouse atmospheres of earth-like planets. http://arxiv.org/pdf/1510.03527v1.pdf.
6. The Wispy Weather of Mars
Introduction
Mars was born unlucky. Thanks to an unfortunate proximity to the
proto-Jupiter, the debris from which Mars coalesced was severely
depleted in mass. This left Mars something of an undersized runt,
setting the tone for its future evolution and ultimate atmospheric
demise. Unlike the Earth and Venus, Mars has a relatively slight
gravitational pull, which is somewhat less than half that of its
larger brethren.
The Mars of today is a cold, dry, barren world, but it wasnt
always this desiccated. How it became so is an interesting tale
that ties the deep interior to events happening millions of kilome-
ters away. Much remains to be understood about the weather and
climate on Mars, but there is already a wealth of interesting and
mysterious phenomena that parallel many seen on Earth. Like the
Earth, Mars has a diverse climate with an atmosphere structured
much like ours. The large difference in the mass of our atmosphere
and that of Venus and Mars means that the Martian climate is as
different from Earths as Earths is from that of Venus.
The Structure and Formation

of the Martian Atmosphere
The contemporary Martian atmosphere is a thin, dry affair with
less than 1/100th the mass of Earths. The atmospheric pressure
averages around 6 millibar, although this figure varies consider-
ably across the planets surface. After all, there is a 22 km dif-
ference in height between the bottom of the Hellas impact basin
and the summit of Olympus Mons. Pressure thus varies by more
than one order of magnitude, being densest at lower elevations,
where it is around 12 millibar. Meanwhile at the lofty summit

DOI 10.1007/978-3-319-25679-5_6
of Olympus Mons atmospheric pressure is only 0.02 millibars, or

150,000th of the pressure at sea level on Earth. As it turns out this
is rather similar to the atmospheric pressure on Pluto (5 microbars
or 100,000th of the pressure at the sea level on Earth).
Mars has such a thin atmosphere that vertical and horizontal
heat transport by convection and advection is much more limited
than it is in our atmosphere: on Mars direct transport by radia-
tion dominates. That is not to say that convection and advection
are not important, and convection is really rather rapid on Mars.
With less mass to transport, and the atmosphere more transpar-
ent, overall, radiation is a far more effective means of getting
heat from the surface to the top of the atmosphere than it is on
Earth or Venus. This difference leads to some odd meteorological
effects that we shall look at later. Most importantly for the idea
of habitation, while the atmosphere is not as noxious as that of
Venus, the pressure is so low that were a human to be exposed to
it, the water in their blood would boil, leading to a rather unpleas-
ant death. Occasional UFO conspiracy stories paint the surface as
habitable if only we could stick a few plants there to make oxy-
gen. Unfortunately the altitude would dessicate them unless they
were in a pressure-controlled environment (as seen in the recent
hit movie The Martian).
Temperature-wise, Mars shows a lot of variability. The aver-
age temperature is around 20 C, with surface temperatures that
can be as high as 25 C or more on the surface nearest to the equa-
tor. Meanwhile, in the long polar winters, temperatures fall well
short of 100 C, low enough for carbon dioxide to precipitate
out as frost.
Figure 6.1 illustrates the general layout of the thin Martian
atmosphere. Aside from the low overall pressure, perhaps the most
striking thing is how dispersed it is. While on Earth the bulk of the
terrestrial atmospheres mass is found within 7 km of the surface
that on Mars is far more diffuse. The troposphere extends up to
70 km, with a narrower stratosphere and mesosphere forming the
upper layers of the Martian pancake. The troposphere is broader
than that on the Earth because, on this occasion there is such a
big temperature difference between the Martian surface and the
tropopause. There is also no significant ozone layer. These factors
allow the atmosphere to convect over a broader range of altitudes.
The Wispy Weather of Mars 233
500
Thermosphere
175
150
Aurora Mesosphere Mystery Dust Clouds 100
Height (km)
Stratosphere
Diffuse airglow 75
Troposphere
50
Carbon dioxide Ice Clouds
Dust
25
Water Ice Clouds
-200 -150 -100 -50 0 +50 +100 +150 +200
Temperature (oC)
FIG. 6.1 The overall structure of the Martian atmosphere. Aside from
being much colder than that of the Earth, the rarefied atmosphere is far
more dispersed than above the Earth, with the troposphere, for exam-
ple, spanning more than three times the height of the same layer above
our planet. The newest feature is sporadic, seasonal clouds seen at 100
120 km above the surface. The nature of these are unclear. Diffuse air-
glow (low level aurora) occurs at 6080 km
Overall, the temperature profile varies rather like that of

Venus, but it is obviously a lot colder in the Martian troposphere.
Above the troposphere, temperatures are rather similar to that
above Venus (Fig. 5.1). This is largely set by two properties: the
intensity of solar radiation and the composition of the air which
is predominantly carbon dioxide. Carbon dioxide maintains the
surface temperature on Mars as it does on Venusbut the far less
concentrated greenhouse gas is less able to retain much heat. The
uppermost layer, the thermosphere, has a temperature controlled
by the intensity of ionizing radiation from the sun, but also by
the still present, but even more dilute, carbon dioxide and carbon
monoxide gas that still functions as a greenhouse gas.
Despite a Venusian temperature profile, surface temperatures

excepted, and composition, Mars shares more climatic and mete-
orological phenomena with the Earth than it does with Venus.
The more rapid Martian rotation and low temperature favors a
terrestrial-like circulation driven by differential heating across the
surface and by the Coriolis Effect (Chap. 1). But there are impor-
tant differences. For one, the smaller radius of Mars (roughly half
that of the Earth) means that the Coriolis Effect is weaker. Why?
Well, imagine you are moving at 100 km per hour, northwards
from the Martian equator. Mars rotates at about 300 km per hour
around its polar axis. So, for every 100 km the air moves north,
it also moves 300 km to the east. On Earth, the rotation speed is
1700 km per hour at the equator. Here, for every 100 km the air
moves north, it moves 1700 km to the east. Thus air is going to
be deflected by the rotation of our planet far more than it will be
on Mars. Thus, although both the Earth and Mars rotate on their
axis every 24 h or so, the Coriolis Effect is far stronger on the Earth
than it is on Mars. Yes, air is still deflected in the same way as it
is on Earth, but the effect is far weaker, so air moving northwards
or southwards on Mars can travel far further (about five times fur-
ther) towards the pole before being swung eastwards. Likewise, air
moving towards the equator is deflected less strongly towards the
west on Mars than it is on the Earth.
The low density and mass of the Martian atmosphere sets the
overall theme for its weather and climate. Like a worn out sponge
with many large holes, the Martian atmosphere allows most of the
energy it receives from the Sun to pour back out into space. It has
what atmospheric scientists call a low thermal inertia; and this
creates some odd effects.
For one, although the Martian atmosphere can get rather
warm at the surface these temperatures are restricted to the air
closest to the surface. Ascend a few hundred meters and the tem-
peratures will have fallen close to or below the freezing point of
water. This strong gradient in temperature means that Martian air
is very unstable and convects air rapidly. Thus strong heating near
the surface can lead to some dramatic winds.
Moreover, when the Sun goes down Martian air loses heat
very rapidly at night. This, in turn means that there are dramatic
differences in temperature between the day and night hemispheres.

With temperatures low enough for gases to condense out of the
air, there can be rapid changes in mass of gas that is in the atmo-
sphere. Rapid changes in temperature and in the mass of gas in the
air result in large pressure differences between the day and night
hemispheresand between the summer and winter hemispheres
(Figs. 6.2 and 6.3). These phenomena are largely unknown on the
Earth: any effect thermal tides have is overwhelmed by the effect
of our planets rotation.
Mars thus, displays a very complex pattern of circulation
driven by competing effects. The Coriolis Effect tries to organize
the winds in a manner like the Earth (Figs. 6.4, 6.5 and 6.6), while
the low density of gas results in flows from warm to cold (Figs. 6.2
and 6.3). The resulting pattern shows overlapping effects which
work together in different ways in different seasons.
Sun
FIG. 6.2 Thermal tide flow in the Martian atmosphere. Thermal tides are
airflows driven by strong differences in heating and cooling of air in the
atmospheres of planets. Although Mars rotates at a similar rate to the
Earth and has a modest Coriolis Effect, the atmosphere is so thin that
radiation can easily escape. This leads to strong heating at the surface of
the planet in the day and very strong cooling at night. This leads to a gen-
eral flow of air from the day side towards the night-side of the planet: this
is a thermal tide. The same process happens in the atmosphere of Venus
but this is driven more by much stronger heating and a much slower rota-
tion. As Mars rotates the area afflicted by cooling and heating migrates
across the surface, interacting with surface features such as the Tharsis
bulge, as well as with the Coriolis, and other, effects
Summer
Winter
FIG 6.3 Bulk seasonal flow on Mars. Despite a modest Coriolis Effect and
thermal tides driving air flow on Mars, there is an underlying seasonal
flow of gas within the Martian atmosphere. As both carbon dioxide and
water are at or below their condensation and re-sublimation points. This
means that in the summer hemisphere both of these gases migrate from
frost to the atmosphere, raising the overall atmospheric pressure. How-
ever, this freezes out once more in the winter hemisphere. This gives
Mars a bulk, cyclical motion of gas from one hemisphere to another over
the course of the Martian year. This is exacerbated by the Martian orbit
which takes Mars more than 42 million kilometers further from the Sun
during the Martian southern summer than it does during the Northern
summer. Thus southern summers are warmer and northern winters
colder than the corresponding season in the opposite hemisphere
For example, during the southern summerwhich is also,

currently when Mars is closest to the Sunstrong heating along
the flanks of the southern highlands and Tharsis Bulge lead to
strong updrafts. These updrafts draw in air from the northern
hemisphere, which pours southwards, much like the strong Asian
monsoon flow on Earth (Chap. 2). Bereft of oceans to add mois-
ture these winds dont bring rain to Tharsis, but they can draw
sand southwestwards towards the equator, then south eastwards
towards the southern highlands. As Mars is a good 42 million kilo-
meters from the Sun during the northern summer than it is in
Acidalia
Planitia
Amazonis Olympus Arabia
Planitia Mons Terra
Equator
Argyre Planitia
FIG. 6.4 Topography-driven Rossby waves within the northern winter

hemisphere of Mars. Air is forced around and over the Tharsis Bulge and
associated mountains. Just like the Rockies on Earth, winds dip south-
wards over the Tharsis region and then northwards again to the east of
the rise. Such topography driven waves are limited to 13 over the entire
planet, while on Earth there may be up to six (Chap. 1). Such waves help
to organize seasonal storms, including planet-wide dust storms that
afflict northern Mars in this hemispheres spring. Ground topography
image: NASA
the southern summer, the northern Martian summer is therefore

cooler and longer than that in the south. Consequently, the north-
ern summer monsoon is much weaker than the southern summer
one. This discrepancy leads to Martian sand dunes aligning in the
direction of the prevailing winds of the southern summer.
The Martian atmosphere mimics that of the Earth in other
respects. As well as organizing monsoon flows like those on Earth,
the Martian atmosphere hosts large circulations like the Earth. Air
rises strongly over the regions under the strongest illumination in
the summer, moves Pole-wards at height, before descending over
the Martian sub-tropics. These Hadley cells are directly analogous
to those on Earth. North or south of these, at least in the winter
and spring hemisphere, a second cell analogous to the terrestrial
Spring Northern Hemisphere
Fe Fe
Ha Ha
Ha
Ha
Winter Northern Hemisphere
Fe Fe
Ha
Ha
FIG. 6.5 The general, seasonal circulation on Mars. In the Spring, with
the Sun over the Equator there are two Hadley Cells over the Martian
Tropics. The Northern cell is weakest. Topography helps enhance the
strength of the southern cell. Interactions between the westerly winds to
the north; thermal tides (Fig. 6.2); topography (Fig. 6.6) and the reversing
tropical Hadley cells can lead to the formation of planet wide dust storms
that begin life as frontal-like disturbances within the Ferrel cells. In this
season an easterly jet overlies the division between the Hadley cells as it
does on Earth. The Martian atmosphere has a radically different structure
during the northern and southern winter seasons. During the northern
winter a single, large Hadley Cell (Ha) carries air from the warm southern
hemisphere to the north while the northern Ferrel Cell (Fe) strengthens.
Carbon dioxide is carried by this flow and precipitates out as frost on
the northern ice cap along with much of the atmospheric water vapor. A
westerly jet stream overlies the Ferrel Cell-Hadley Cell boundary much
like the sub-tropical jet does on the Earth. This pattern reverses in the
northern summer, with the Hadley cell carrying air southwards at height.
As on Earth the Coriolis Effect bends the winds eastwards away from the
equator and westwards towards the equator (Fig. 6.6)
Ferrel Cell organizes a belt of westerly winds. Above the axis of

the twin Hadley cells lies an easterly jet stream; while above the
junction of the Hadley and Ferrel cell lies a westerly jet stream,
analogous to the sub-tropical jet stream on Earth (Chap. 1).
Although this is superficially like the terrestrial situation,
Mars is constrained by its weaker Coriolis Effect and thin air,
Equator
Polar Westerlies
FIG. 6.6 Monsoon winds on Mars. As with the Earth, Mars has a seasonal
monsoon (Chap. 2). However, unlike the Earth there are obviously no
oceans, so the temperature gradient is entirely due to the position of the
Sun. Like the Earth, Mars is closest to the Sun when the planet has its
southern summer (shown above) and furthest from the Sun, currently
during the northern summer. However, on Mars the effect is far greater
as the Martian orbit is far more elliptical than ours. Consequently, the
strongest monsoon winds blow to the south, an effect exacerbated by the
large Tharsis bulge, which act much like the Himalayas and Tibet (Chap. 2).
On Mars the Monsoon leaves its mark on the Martian surface in the form
of sand dunes that align with the southern summer monsoon wind
which cant retain much heat. Thus, on Mars, air flows symmetri-
cally around the equator in the spring when the Sun lies above the
Martian equator. During the summer in the northern or southern
hemisphere air flows through one cell to the winter hemisphere,
leaving a single giant Hadley cell. The summer hemisphere also
lacks a Ferrel cell, leaving Mars with a stripped down version of
the terrestrial circulation system.
The transition from winter to spring has important conse-
quences for the Martian circulation. Not only does a rising Sun
begin to vaporize the ice cap and raise atmospheric pressure in
the spring hemisphere, but it rapidly warms the ground increasing
the instability of the air. The axis of the Hadley cell moves with
the area of strongest heating and the Ferrel cell is overwhelmed.
In Chap. 10 we will see how this pattern of airflow might happen

in an exaggerated form on an Earth-like planet that was tipped
over on its side. However, this never really happens on the Earth,
primarily because the Coriolis Effect causes a far stronger devia-
tion in air flow. However, we do see seasonal differences that are
somewhat reminiscent of the situation on Mars. In the northern
summer, in particular, as the land warms, the belt of westerlies
becomes unstable and partly breaks up. The Asian summer mon-
soon only really kicks in when the westerlies running along the
front of the Himalayas break down (Chap. 2). Over the northern
oceans the jet stream buckles in response to the difference in tem-
perature over the land and sea. During the northern summer the
normally strong and broadly circular circulation around the Arctic
also tends to break up into smaller cells. Although there is still a
Ferrel cell per se, it is grossly weakened and is very erratic in its
circulation pattern.
Dust Devils, Tornadoes and the Other

Whirling Winds of Earth and Mars
On Mars the large differences in surface temperature from day to
night and season to season generate large differences in atmospheric
pressure. This leads to strong movement of air across the surface
of Mars leading to some interesting weather.
At a local level, strong heating and limited retention of heat
by the thin air leads to strong convection. On the Earth, such
strong, localized convection combines with the effects of friction
and airflow around obstacles to produce small, whirling columns
of air. Although the Coriolis Effect has a role in these small dis-
turbances, they are so limited in horizontal extent that this effect
can be neglected. The terrestrial myth that the Coriolis Effect
determines whether an emptying bathtubs swirl is clockwise or
counter-clockwise springs to mind. Rather than the hemisphere
being the determining factor in flow direction, minor differences
in friction, water flow or the shape of the bathtub cause the water
to preferentially spiral in one direction or another. With the right
shaped bath and the plug withdrawn upwards, smoothly, the water
will spin in either direction: friction does the work on its own.
The Coriolis Effect has no measurable effect unless you have a

large tub and are very, very careful how you set up the experiment.
On Mars localized convection leads to spinning columns of
air called dust devils, frequently seen on the deserts of the Earth.
These spinning columns form from the ground up and grow as
entrained debris are carried upwards by the spinning, ascending
column of air. Rather more insidious and unpleasant versions of
these dust devils form in terrestrial fires, where combustion rap-
idly heats the ground at the surface. Violent heating then leads to
the formation of narrow columns of fire which twist in response
to the movement of incoming air. Again, friction rather than the
Coriolis Effect is at work. Such mini-twisters frequently assist
in the spread of the fire across non-combustible surfaces, such as
roads. Figure 6.7 shows such a mini-twister during an Australian
bushfire.
Rather more violent than these miniature fire twisters are vio-
lent tornadoes that are spawned from thunderclouds, themselves
generated by large bush fires with flame fronts several kilometers
long. With a very large fuel supply, extensive fires can release so
much heat and moisture that they form fully-sized cumulonim-
bus clouds. Such pyro-cumulonimbus clouds are fortunately
rare, but these can generate hellish weather phenomena of their
own. As well as lightningwhich may start further fires, pyro-
cumulonimbus have been known to produce black hailsoot
encrusted and encased grape sized lumps of iceand, as was men-
tioned, genuine tornadoes.
On the Earth, tornadoes and water spouts at sea are primar-
ily associated with spinning cumulonimbus thunderclouds called
supercells. Unlike dust devils, which plague Mars and the des-
erts of the Earth, convection plays an indirect role. Rather than
growing tornadoes from the ground upwards, convection generates
large cumulonimbus cells, and from these the funnels grow down-
wards. Growth of the funnel is a consequence of lowering atmo-
spheric pressure that is associated with the development of the
storm. As the condensation point of water decreases with pressure
the decreasing pressure will cause moisture in the air to condense
into a funnel-shaped cloud, the cloud densest where the pressure
is lowest. As the storm intensifies and pressure continues to fall,
the condensation level descends towards the ground.
FIG. 6.7 Meteorological effects caused by fires. In (a) a miniature fire

devil joins the litany of catastrophe in a large bush fire in Australia. (b)
shows a man-made fire devil at the Magna Centre in Sheffield. (c) (above)
shows a pyro-cumulonimbus: a thundercloud formed above a large bush
fire from the excessive heat and moisture released by combustion. In
January 2003 this particular cloud hit Canberra, Australia. A short while
after this photograph was taken the incendiary thundercloud generated
a genuine tornado. As with conventional tornadoes this spinning vortex
formed from the cloud down. Journalists, accompanying the Fire Service,
filmed it flattening trees and buildings within the heart of the fire
Supercell storms dont spin because of the Coriolis Effectand

indeed, some storms have two cores spinning in the opposite
directions. Spin is obtained when a horizontal swirl of air becomes
entrained into the developing storm. In these systems winds mov-
ing at different speeds and at different altitudes, generates a spin-
ning vortex of air. If this becomes drawn into the rising stream
of air, within the storm, the spin becomes aligned vertically and
the storm as a whole begins to spin. Where the spin becomes con-
centrated into a narrower spiral a tornado can result. Here, the
collapsing column of spinning air conserves its momentum by
spinning faster, just like an ice-skater. Supercells are common
in many parts of the world, including Australia, central Europe,

Russia, Argentina and southern Brazil. However, they occur most
abundantly in the US, in the region known as tornado alley. In
these regions there is the appropriate combination of meteorologi-
cal conditions: a supply of warm moist air to fuel storm formation;
a nearby source of warm dry air that caps convection for prolonged
periods; and most importantly active cold fronts that can penetrate
both air masses and displace their bulk rapidly upwards generating
powerful storms. By virtue of these requirements, these phenom-
ena only come together at certain times in certain locations.
Supercell thunderstorms have relatively long lifetimes as
they can maintain a strong central updraft. Smaller thunderstorms
tend to choke off their supply of warm moist air as soon as the rain
begins to fall and cold air flushes downwards. This kills the storm
within an hour of its formation. By contrast supercells can last for
hours. In the case of the 2003 Canberra storm, the large bush fire
generated the cumulonimbus cloud that morphed into a supercell
storm. In turn, this supercell generated the tornado that repeat-
edly touched down in the suburbs of Canberra and causing further
devastation. Just like a regular tornado, but unlike a dust devil
or smaller fire devil, this vortex, colloquially called a firenado,
was formed from the cloud down. This firenado included fire in its
arsenal of destruction, demolishing houses and incinerating much
of their remains.
Returning to Mars, dust devils are most common in the
Martian spring and summer when heating is strongest on the
Martian surface. As with tornadoes on Earth, dust devils on Mars
help redistribute the heat within the thin Martian atmosphere.
They can combine in waves and deliver sufficient dust to the
atmosphere that it blocks solar radiation from reaching the sur-
face. This cools the surface and warms the atmosphere and, over-
all, makes the atmosphere more stable. In the process gangs of
dust devils can generate larger Martian dust storms, although in
general the largest of theseplanet-wide affairs form when frontal
disturbances form nearer the Martian poles in the spring.
Why is the spring the key season for dust storms? This is an
interesting question that has only recently been solved. During
the spring certain key atmospheric phenomena combine to gen-
erate ideal conditions for dust storm formation. Storms become
global phenomena when frontal storms combine with the effects

of Martian thermal tides to deliver extensive dust into the tropical
Hadley Cells. Huiqun Wang (Harvard University) and colleagues
combined observations of the 2001 planet-wide storm and com-
puter simulations. Wang identified how different scale atmo-
spheric processes collaborated in generating a storm that was to
eventually blanket the entire planet (Fig. 6.8). Topography was key
to their development. Dust storms develop and progress initially
across the low elevation portions of the planet: Acidalia-Chryse,
and some within Arcadia-Amazonis and Utopia. The arrangement
of wide basins, flanked by regions of (broadly) northsouth aligned
topography helps generate Rossby waves (Chap. 1) within the
northern atmosphere within the Ferrel cell. This is analogous to
the effect the Rockies and Andes have on the westerly circulation
on Earth. These topographic rises are critical to delivering energy
and dust southwards.
In the southern spring and summer a number of different fac-
tors are coming together. The northern cap is cooling down and
increasing in mass. Strong winds are moving mass (mostly dense
carbon dioxide gas) into the cap and the dominant monsoon flow is
flowing to the south of the equator. During the northern autumn,
the Hadley cell is shrinking north of the equator and is eventu-
ally sub-served to a single cell extending from the south (Fig. 6.5).
Thermal winds are still aligned southwards in the direction of the
southern hemisphere since as this hemisphere warms, winds blow
from the cold hemisphere towards the warm hemisphere at the
surface, but away from it at height (Fig. 6.5).
FIG. 6.8 (continued) (b) On Earth most dust storms in the temperate
regions and the sub-tropics begin their life like those on Mars. Cold
fronts penetrate southwards during the autumn, winter and early spring.
Here, over Africa and the Middle East (and similarly over the US in the
spring or Australia during the winter) they encounter the dry continental
interior. Abrupt changes in wind speed and direction can kick up enor-
mous clouds of dust (brown arrows) ahead of the advancing cold front.
Unlike the desiccated Martian storms, these cold fronts may have frontal
thunderstorms associated with them and, in many cases, it is downdrafts
from these thunderstorms that initiate the dust storms
L
Acidalia
Planitia
Syrtis
Major
Low
PFJ
Low
FIG. 6.8 (a) Developing dust storms on Mars in the Martian spring of 2002.
A frontal storm (blue arrow) developing and moving south-eastwards
along a dip in the Rossby wave (purple) towards the Martian equator.
Although broadly analogous to a terrestrial cold front, the cloud along
this boundary is dust, not water vapor. To the fronts rear a second frontal
dust storm is developing (dark blue arrows). Where these frontal storms
work together with the daily thermal tide and reach the tropical Had-
ley cell dust can be carried around the entire globe forming planet-wide
storms. Underlying Mars photograph NASA/JPL (Mars Global Surveyor).
Periodically strengthening this bond are the Rossby waves that

develop along the westerly jet stream that forms the divide between
the Hadley cells and the westerly winds further north. As the north
cools down, these waves become stronger as the temperature con-
trast grows between the northern polar cap and the rest of the planet.
Like the Rocky Mountains on Earth, the Tharsis bulge promotes
the development of waves that dive south over Acidalia Planitia
(Fig. 6.4). Here, they drive the formation of low pressure areas and
cold fronts, which migrate southeastwards across the flat plain
towards Syrtis Major (Fig. 6.8). These fronts move rapidly: around
70100 km per hour and are accompanied by smaller, regional
storms. Such dust storms are frequent during the Martian spring
and autumn (around 12 per season) but on the whole are limited in
extent. Generation of a global storm requires these wave-generated
dust storms to align with the daily tidal flow as the front is cross-
ing one of the low elevation plains. When it does, dust is advected
into the equatorial Hadley cell. Successful stormsthose that
generated planet-wide dust stormswere those that crossed 0 E
between 9 am and 7 pm. This was when tidal winds across Acidalia-
Chryse were also blowing to the south. At other times, the tidal
flow was reversed and storm propagation was blocked.
In the northern winter, atmospheric waves are stronger and
this, too, facilitated the progression of dust-bearing fronts to the
equatorial regions. Interestingly, although the distance the cold
fronts penetrate is a factor in getting dust near to the equator,
this appears to have no direct link to the formation of global dust
storms. Wangs group found the only factor that mattered was the
speed of the front, which was solely linked to the tidal flow and the
season. A deeply penetrating front that couldnt add its flow to the
tidal flow wouldnt successfully deliver its load of dust to the tropi-
cal cells. Instead, most of the dust associated with these storms
stays put, north of the equator. In the Mars Global Surveyor data
the largest, global, storms were associated with a fortuitous move-
ment of multiple frontal storms across Acidalia-Chryse. Such mul-
tiple introgressions give the atmosphere more chance align with
the tidal flow and to deliver dust to the tropical circulation.
What happens next was demonstrated in research carried out
in 1982 by Robert Haberle (currently at NASA Ames) and a decade
later R. John Wilson (then at Princeton). Research indicated that
there would be a dramatic change in the structure of the atmo-

sphere. Dust absorbs heat energy far more efficiently than the thin
carbon dioxide gas on its own. Able to hold and transport more heat
the Hadley cell suddenly has a massive injection of energy. Just as
extra carbon dioxide is slowly enhancing the strength and reach of
the tropical Hadley cells on Earth, dust has the same but far stron-
ger effect within the wispy air of Mars. Instead of extending from
the southern mid-latitudes to just north of the equator, the south-
ern Hadley cell will now extend all the way to the North Martian
Pole, bulldozing the Ferrel cell out of the way. As soon as this hap-
pens dust fills most of the atmosphere. Surface temperatures drop,
while the dust layer warms by 510. Indeed, during the 2001 global
storm the atmosphere warmed by a rather stunning 45 C, causing
it to puff upward, enough to threaten orbiting spacecraft.
Remember that on Mars, the atmosphere is so thin that it
has what you might call a low buffering capacity: it cannot adjust
smoothly to changes that occur within it. Were the Earths atmo-
sphere to be warmed regionally by dust, the circulation within its
denser bulk would rapidly even out the differences in temperature
and pressure. Were this the even denser, soup-like atmosphere of
Venus, the atmosphere would hardly notice the difference in
conditions.
There is an interesting link with terrestrial climate-
skepticism, here. As Mars warned in the 2000s a small but signifi-
cant number of skeptics suggested that this was evidence for the
Sun driving our warming. After all, how could terrestrial carbon
dioxide warm Mars simultaneously? Obviously, it couldnt: there-
fore, man-made global warming was a myth and the Sun must
be warming both planets more. The problem, or rather problems,
with that tale was that the Sun wasnt getting more luminous.
Moreover, Venus, Jupiter and the other planets werent getting
warmer, either. So much for the Sun Instead, Mars was simply
getting warmer because of dust and changes to its albedo (reflec-
tivity) at the same time as the Earth was getting warmer as a con-
sequence of a greater burden of greenhouse gas. Unlike the Earth,
Mars eventually cooled back down once the dust settled. Indeed,
changes in Martian temperature, aside from seasonal differences,
are largely caused by changes in surface albedo. The brighter Mars
is, the colder it is and vice versa (Fig. 6.9).
FIG. 6.9 Sunset on Mars, taken by the Late Spirit Rover at Gusev crater
in 2005. It shows a couple of unique phenomena caused by atmospheric
dust. The Sun is pink while the region immediately above the setting
Sun glows blue. Both phenomena are caused by the scattering of blue
light. Ubiquitous Martian dust scatters the blue light forward, with the
greatest scattering at lower altitudes where the concentration of dust is
highest. This creates a focus of blue light above the setting Sun. Scatter-
ing leaves red light projected directly from the Sun towards the Spirit
rovers cameras. This combined with the color of dust in the air leaves a
pinkish glow
Indeed, Lori Fenton (Carl Sagan Center), Paul Geissler (USGS)

and Robert Haberle (NASA Ames) have examined more cyclical
changes in Martian surface temperatures. Like the more drastic
changes induced by global dust storms, seasonal changes in wind
direction can enhance surface winds. Where the ground is bright-
est (has the highest albedo) the winds are lightest and where the
ground is dark the winds strengthen. Wind scouring can remove
lighter upper dust, exposing deeper, darker soils, which, in turn
causes the winds to blow harder. These changes are also associated
with temperature rises of up to 0.65 C over the darker areas. Rises
of this scale increase the formation of dust devils simply because
the amount of convection increases. More dust devils mean that
there is more transport of dust and heat into the atmosphere,
which, in turn, increases the temperature of the lower atmosphere.
Aurora
Auroras are an inevitable consequence of having an atmosphere on
a planet around a star. Therefore, all of the planets, bar Mercury
and probably Pluto (if we can include this) host auroras. These
luminous displays occur when charged particles from the Sun
crash into the atmospheres of the planets. In the case of the Earth,
Jupiter, Saturn, Uranus and Neptune, such particles are funneled

onto the planets Polar Regions by the action of strong, internal
magnetic fields. This usually isnt a direct dive from solar wind to
atmosphere. Rather, magnetic fields first escort the particles into
a large magnetized domain that surrounds the planet, called the
magnetosphere. From here changes in the direction and strength
of these field lines cause particles to be dumped periodically into
the atmosphere of the planet, nearest the magnetic poles. Such
reconnection events are fast and follow magnetic field lines
from the Sun linking and re-working the planets field.
Consequently, with the exception of Venus and Mars auroras
are confined to the polar regions and it is only when the Sun ejects
a particularly energetic burst that aurora can spread further afield
cross the disc of the planet (Fig. 6.10).
Ultraviolet 500
Aurorae
Thermosphere
175
A Ultraviolet
A Aurorae 150
Mesosphere B
100
Stratosphere Height (km)

75
Troposphere
Belts of 50
localised
magnetic
field 25
S N
FIG. 6.10 Auroras on Mars. In 2004 Mars Express found ultraviolet auro-
ras associated with localized regions of more intense magnetic fields in
the planets southern hemisphere (A). Such auroras are unique in the solar
system. In December 2014 MAVEN found more widespread Christmas
Light auroras over much of the northern hemisphere (B). Although spot-
ted in the UV part of the spectrum these auroras may be visible as a dif-
fuse, green glow in the night sky, as well. Moreover, Mars hosts a diffuse
airglow, much like Venus, which is also found in the ultraviolet part of
the spectrum on the night-time hemisphere of the planet
Venus and Mars are differentlacking an appreciable mag-

netic field of their own. The average present magnetic field at
the surface of Mars is less than a few tens of nanoteslas, or about
1/1000th that of the Earth (25,00065,000 nanoteslas or 0.25
0.65 Gauss). For comparison, a decent fridge magnet has a field
that is about 10 million nanoteslas. Consequently, the solar wind
is able to interact with the full sunlit disc of the planetwell
almost. Above Venus the solar wind energizes the carbon dioxide-
rich atmosphere on the sunlit hemisphere and breaks apart some
of the atmospheres constituents. Atomic oxygen and ionized oxy-
gen are directed to the night-side of the planet by thermal winds
flowing above the cloud tops. The movement of these ions is anal-
ogous to an electrical current and results in the formation of an
induced magnetic field (Chap. 5). Once in darkness, the oxygen
can recombine with electrons or with carbon monoxide and emit
light. Although much of the energy released is in the ultraviolet
part of the spectrum, a faint greenish glow might also be apparent
when oxygen ions recombine with electrons.
A similar phenomenon occurs on Mars, but at much lower
altitudes simply because the wishy-washy Martian atmosphere
allows far deeper penetration of the solar wind than happens on
Venus. This is due to a more limited braking of incoming particles
by the lower density of the gas, but also because the thinner gas is
less able to develop its own internal and repulsive magnetic field
in response to the solar wind slamming into it.
In some ancient regions of the southern Martian crust there
exist broad, linear strips of magnetized crust with a field strength
up to 1500 nanotesla (Fig. 6.11). Their origin predates the large
Hellas impact basin and likely formed within the first 400 million
years of the planets formation. Only, here, where magnetic fields
generated within the crust is sufficiently strong, is the Martian
atmosphere afforded a limited degree of protection. However, it is
here that magnetic fields funnel charged particles into particular
regions of the atmosphere. Thus, the protection that is provided is
certainly limited and the net result of these regional fields is that
more of the atmosphere is energized than might otherwise be the
case were Mars surrounded by a smooth bow-shock like Venus.
In 2004 the ultraviolet spectrometer, onboard Mars Express,
discovered ultraviolet emission from localized regions above
FIG. 6.11 The effect of localized magnetic fields around Mars. Unlike the
Earth, Mars has localized fields which partly encase regions of the plan-
ets surface (purple ovals). Extending out into space, these fields distort
the bubble of space around the planet located within the bow shock (blue
line). This not only protects parts of the Martian surface from the impact
of solar wind particles (yellow arrows) and extra-solar cosmic rays, but
also directs the solar wind into other area causing aurora and influencing
the stripping of gases from the Martian atmosphere. The pattern of mag-
netic stripes is shown in the graphic from Mars Global Surveyor, lower
(NASA)
Mars. This was clearly distinct from the diffuse emission as it was
localized to one geographical region. Jean-Loup Bertaux (Service
dAeronomie du CNRS/IPSL) and colleagues had identified the
first true aurora on Mars. Through the channeling of charged par-
ticles from the Sun, the localized magnetic fields on Mars were
producing aurora over the southern hemisphere of the planet. In
essence this had been expected, yet their discovery was important,
for it still remains the only example of regional aurora above any
planet, where the crust is generating the magnetic field that guides
the particle flow.
Eleven years later there was a slightly more surprising dis-
covery by MAVEN. Again, observing in the ultraviolet portion
of the spectrum, MAVEN discovered very widespread aurora
over the northern hemisphere of Mars. Like the southern aurora
these occur at significant depth within the Martian atmosphere.
Whereas, terrestrial auroras occur exclusively above 100 km, the
Martian auroras manifest more deeply; down to 6080 km from
the Martian surface. Again, this is a consequence of the dilute
Martian air, which affords limited protection from energetic solar
and cosmic particles.
Mystery Clouds Across the Southern, Morning Terminator
From 1995 through to 2014 there were a number of amateur astron-

omer (archived), Hubble observations and later MAVEN observa-
tions of odd clouds seen near the Martian terminator. All of these
were seen above the same geographical region, 195 West and 45
South, near Terra Cimmeria. The problem with these clouds is
that they appeared to lie at great height within the Martian atmo-
sphere: up to 250 km above the surface. This is almost twice as
high as many auroras seen across the portion of the Martian globe
(above) and clearly places them in the Martian ionosphere and
exosphereeffectively out into interplanetary space. Each cloud
that was seenand the best observed occurred in March and April
2012 was several hundred kilometers across and lasted for about
10 days. More perplexing the clouds were only seen at the morning
terminator but had vanished by the time Mars had rotated bring-
ing the elusive clouds around to the evening side. Inspection of
images (Fig. 6.12) of Mars does appear to imply that these clouds
FIG. 6.12 Mystery clouds on Mars. Appearing at only one location, these
clouds of what appear to be dust extend across 5001000 km of Mar-
tian terrain and only appear in the morning. They occur at extraordinary
heights, greater than aurora. Possible explanations include some sort of
unknown and very elevated aurora and dust clouds that have somehow
reached the edge of space. Images: Donald Parker (Association of Lunar
and Planetary Observers, Pennsylvania)
do extend some way across the day side. This is evident from the
parity in the colors of features seen at that latitude on the Martian
surface and in the clouds. However, they appear to dissipate dur-
ing the Martian day. So, what are these mysterious, seasonal and
periodic clouds and how do they get so high in the atmosphere?
There are a few possibilities based on their location above
the Martian surface. Firstly, they occur where magnetic fields are
strongest, so they might be auroras. However, the color is wrong
(they look dusty rather than bluegreen or red) and they only
appear in the early morning sky. Moreover, they would have to
over 1000 times brighter than terrestrial aurora to be visible from
Earth. Alternatively, they might be dustthe color is rightbut
how did dust get so high up: wind strength is going to be negligible
as the air density is miniscule.
Martian dust is likely ruled out, but another possibility is dust

from the small Martian moon, Phobos. The authors of the 2015
Nature paper, Agustin Snchez-Lavega and colleagues suggest that
the clouds, if thats what they are, would be best modeled as car-
bon dioxide or water ice. Getting the ice to this height100 km
and above the surfaceis problematic, but not impossible; but
why is it only visible across the morning terminator?
The suggestion is that these odd clouds were created in the
Martian mesosphere where temperatures can be lower tan 150 C
at night. Here trace amounts of water vapor and/or carbon dioxide
might freeze out forming tiny particles (around 100 nanometers
(100 billionths of a meter) across that eventually assemble into
clouds. The location, nearest where Mars hosts its strongest mag-
netic fields might be an important factor. The particles will need
to be levitated to higher altitudes (and warmer altitudes) where
they can be seen above the terminator. As is likely, air density is
too low for wind to be a factor. However, particles at the heights
the clouds are seen will become charged through the action of
charged particles from the Sun and through the action of cosmic
rays. Once charged, magnetic fields in the southern hemisphere of
Mars could transport them to great heights. Enhanced and mag-
nified images imply some form of filamentary structure to the
clouds. This would favor the action of magnetic fields in structur-
ing the gas. However, at present there is no real evidence to sup-
port any of the possible mechanisms that levitate clouds to such
great altitudes.
If the clouds are ice we can begin to understand why the
clouds vanish during the Martian day. As the planet rotates and
the clouds are swept into the full glare of the Sun, they will soon
sublimate once more and disappear from view. Presumably, much
of the material that makes up these clouds will become lost to
space as at 250 km they are well within the grasp of the solar wind.1
MAVEN has also detected dust at high altitudes. There
appears to be rather a lot about the depleted atmosphere of Mars
that we do not yet understand.
1
David Andrews of the Swedish Institute of Space Physics suggests that these clouds are levitated
by coronal mass ejections. These impact the ionosphere and generate enough ionization to lift dust
or ice grains from further down.
Martian Ozone
Like Venus, Mars has ozone within its atmosphere; and like Mars
this is produced not by the reaction of abundant oxygen with
ultraviolet light, but indirectly from oxygen released when carbon
dioxide is split by ultraviolet radiation. On Venus, ozone is largely
confined to the night side of the planet at an altitude of about
100 km (Chap. 5), but on Mars a much more complex pattern of
ozone production and destruction is found. This difference, once
again, relates to the tenuous nature of the Martian atmosphere.
With a low atmospheric density, ultraviolet radiation pene-
trates deeply and ozone can be produced low down. It is circulation
within the atmosphere ultimately determines where the greatest
concentrations of this will be found. Nowhere does the concentra-
tion of ozone come anywhere close to that in our atmosphere. At
its densest it is 300 times less dense than found in our ozone layer
and thus similar to that above Venus. This is enough to somewhat
limit the ultraviolet radiation that reaches the Martian surface.
The greatest concentration of ozone occurs near the South Pole in
the winter. Mars has a complex distribution of this gas with height.
Not far above the Martian surface is a persistent layer that covers
most of the Martian surface. Above this, in the northern spring
and summer, lies a second ozone layer in 3060 km up. Over the
southern pole lies the third and most plentiful layer at an altitude
of 4060 km. This third layer never forms near the northern Pole
during its winter and more unusually, decreases in altitude with
decreasing latitude, being lowest nearest the Martian South Pole.
This pattern of ozone creation and destruction relates to
changes in the circulation of the Martian lower atmosphere.
Remember, that during the southern summer a large circulatory
cell expands until it delivers air from the Martian equator almost
the entire way across the southern hemisphere (Fig. 6.5). This
sweeps up any ozone and not only delivers it more generally to
the rest of the planet, but brings it into contact with hydrogen gas
that has been released from the breakdown of water vapor by ultra-
violet light. Ozone is then destroyed. During the Martian winter,
the developing Ferrel Cell isolates the polar atmosphere from the
rest of the planet. It also cools so much that any moisture that is
present freezes out as frost. This leaves any free oxygen the chance
to reform more ozone. This brings us a gentle reminder about

developing an extensive hydrogen economy on Earth. Hydrogen is
notoriously difficult to store safely in large quantities. It is highly
likely that in our attempts to use hydrogen we will inadvertently
release yet another agent that will deplete our ozone.
Martian Methane: Mars not Dead, Yet
Methane is a contentious gas. On Earth, as a greenhouse gas it

is primarily produced by anaerobic bacteria in paddy fields or
the large colons of mammals, and as such is quite literally the
butt of jokes. Methane is also stored in large quantities in three
terrestrial locations: prehistoric gas fields dating back several tens
or hundreds of millions of years; historic to prehistoric meth-
ane in permafrost; and finally as frozen, water-caged complexes
known as clathrates on the deep ocean floor. In every one of these
cases methane has a biological origin, having been produced by
anaerobic fermentation. Methane may also be produced through
chemical reactions between the mineral olivine, water and carbon
dioxide. During the geological process known as serpentinization,
olivine takes up water and in the process hydrogen is released that
can chemically reduced carbon dioxide to leave methane.
In the coldest reaches of the universe, hydrogen ions that are
produced by ultraviolet irradiation of deep, dark nebulae can also
initiate a bewildering array of chemical reactions. Amongst the
myriad of products is methane in abundance. Comets and meteor-
ites can deliver this to the terrestrial planets, but it is thought that
they were probably not born with much of this volatile gas. On a
planet like Mars conditions on its surface are strongly oxidizing:
that is methane effectively wouldnt stand a chance of surviving
longperhaps a few years at most. When scientists at the Goddard
Flight Center discovered methane in the Martian atmosphere in
2003 it caused something of a stir. What was all the more inter-
esting was that the methane seemed to come and go in a seasonal
pattern. Enter the Curiosity rover.
During the first year of the rovers operations it found little
evidence of methane in the Martian atmosphere. Stringent upper
limits were placed on the gas of less than 1 part in 5 billion. For
comparison oxygen comprises 0.1 % of the Martian air which is
1 part in a thousand, or a million times denser. Thats where the

situation remained for the best part of a year, thus flying in the
face of the terrestrial measurements. In the winter of 20132014
NASA reported that Curiosity had discovered a tenfold increase in
atmospheric methane. While not quite a surge, the level had gone
up so much that it implied some sort of source on Mars. Moreover,
the rapid changes in the concentration clearly implied some
sort of local source for the gas, which is then rapidly destroyed.
Immediately ruled out were photochemical reactions between
water and carbon dioxide in the Martian air.
Meteorite impacts or atmospheric detonations might do the
trick but researchers at Imperial College in London ruled out this
as their calculations suggested such an effect could not explain
the observed variability. At present, researchers are left with two
possibilities: Martian microbes (presumably underground) or some
form of subterranean volcanic or hydrothermal activity that is pro-
ducing methane. Sadly, in each case there is no evidence to sup-
port either. Subterranean microbes are not impossibleand there
are clearly many examples on the Earth, as deep as 3 km under-
ground: we simply cant tell if this is true. The chemical reactions
they employ to produce methane are very similar to that involving
olivine. Given that serpentinization would be limited to deeper,
warmer terrains, discriminating between the two processes will
be challenging.
Worse still is our inability to identify the process or processes
that is rapidly removing the vented methane. Methane only peaks
for a few weeks on Mars implying a very rapid process of elimina-
tion. What this is also remains in the realm of speculation and
could involve chemical reactions or sequestration, where clath-
rates or other compounds absorb but do not necessarily destroy
the methane.
Experiments done by Viking in 1976 offer a clueas do mea-
surements made by the Phoenix Lander in 2008. Viking went
looking for life, essentially by offering up lunch to any waiting
microbe. Glucose was mixed with Martian soil and a rapid chemi-
cal reaction ensued. Carbon dioxide and water vapor were released
as though the glucose had been respired. However, on repeated
exposure nothing else happened. It appeared that the observed
reaction was a one of chemical reaction involving materials in
the Martian soil, meaning it looks like highly oxidized iron com-
pounds or hydrogen peroxide may be the source of the chemical
reactions that ate Vikings meal. The discovery of perchlorate
compounds by the Phoenix Lander in 2008 might offer an alter-
native explanation for the observations. These compounds make
very handy components of capssmall explosive devices used
by children throughout the western world to annoy their parents
and bystanders. These can explode on impact with solid surfaces,
particularly when mixed with sugar. In the chemical reaction, the
perchlorate oxidizes the sugar, surrendering its oxygen atoms to
power the reaction. On Mars similar reactions would readily do
away with methane and any organism unwitting enough to be
near them when it was venting the gas.
Perchlorates are something of a double-edged sword on Mars.
Aside from aggressively oxidizing organic compounds, perchlo-
ratesand this is most likely calcium perchlorate on Marswill
help stabilize liquid water. Under standard Martian conditions
water is either solid ice (as frost or buried ice) or a gas. Exposure
of solid ice results in sublimationthe direct conversion of solid
to gaseous water. However, add some perchlorateand the obser-
vations suggest 0.6 % by massand water is stable in its liquid
form. Now, although such water would be a tad aggressive to
terrestrial microbes, it doesnt necessarily mean it would be to
Martian microbes.2 With some irony, perchlorate might make
lifes existence rather unsavory but might also allow the presence
of liquid water that allows it to develop in the first place. If an
organism could eke out a living, perchlorate would make a very
handy alternative to oxygen gas. Still not convinced, look at life
on Earth now. Life on our little blue worlds surface is bathed in
one of the universes most noxious and reactive gases. Oxygen is
a brutal gas to organic compounds; just ask any anaerobic bacte-
rium. Yet life on Earth thrives. A dash of perchlorate and a whiff of
peroxide might just be the tonic to perk up a long-frozen Martian
microbe. Never doubt the resilience of life to withstand and adapt
to harsh conditions.
2
In September 2015 NASA conrmed that liquid water does periodically ow on the surface of
Mars and that this is likely stabilized against evaporation by dissolved perchlorates.
Regardless of the source the presence of methane in the

Martian atmosphere is almost certainly the most interesting dis-
covery of the last 15 yearseven dwarfing the observation of mys-
tery dust clouds. If methane has a biological source or a geological
one, it implies Mars is not quite dead yet. The biological signifi-
cance is obvious, but the geological significance is certainly on
a par: the Martian giants, the volcanoes of Tharsis and beyond,
might still hold a lit candle to the steadily ticking clock of Martian
senescence.
The Lingering Death of Mars
The Martian atmosphere of today is a pale reflection of the one it

once had. Mars once held a deeper, denser atmosphere, the means
by loss of which remains controversial. Early on Mars likely had
an atmosphere similar to the early Earth and Venus. After all, each
of these worlds was born from roughly the same materials. Thus
we would expect Mars to have had a fairly hot, wet atmosphere
for at least the first 50100 million years, or so. This might well
have been similar enough to the Earths for it to have hosted hot,
pressure-cooked oceans much like the Earth and probably Venus.
As Mars is roughly 50 % further from the Sun than the Earth, this
early wet super-greenhouse would have rapidly broken down with
any moisture and much of the carbon dioxide raining out to be
sequestered in Mars early oceans. All of that is fairly uncontrover-
sial. Its what happens next that a mystery.
Measurements by Curiosity reveal that most of the Martian
atmosphere has been lost to space: what is left on Mars is a pale
and tattered imitation of the planets former blanket. At some
point between 4.4 and 1 billion years ago Mars dried out. The gen-
erally accepted but poorly constrained thesis is that the solar wind
stripped the Martian atmosphere away leaving the currently desic-
cated world. The problem is how to do this.
There are several different ways in which gases can escape
the atmosphere of a planet. Some are straightforward and relate to
the temperature of the gases: such mechanisms are called thermal
mechanisms and they are driven by the temperature of the gas
and the gravitational pull of the planet as a whole. Other mecha-
nisms involve more complex reactions or physical extractions
by the solar wind. These rather more robust pathways are called
non-thermal mechanisms and include: stripping by the solar wind;
chemical reactions within the gases in the atmosphere; extraction
by impacting comets or asteroids; or electrical levitation.
The manner in which mechanisms have the greatest effects
depends strongly on the mass of the planet; the mass and tempera-
ture of the gases in the atmosphere and the strength of the solar
and more broadly stellar winds that impact on it, as well as the
chemical composition of the gases in the atmosphere. For Venus
the relatively large gravitational pull of the planet, combined
with the relatively massive principle gas, carbon dioxide, means
that relatively little gas can escape through simple Jeans escape.
However, Mars is a different matter altogether. Two papers in the
journal Science summarized Curiositys findings. In the first Paul
Mahaffy (NASA Goddard) and colleagues described the skewed
ration of carbon-12 to carbon-13.
As with oxygen in the water in the Earths atmosphere, there
is a fixed abundance of carbon isotopes in the universe that is set
by nuclear reactions in stars and the processes that give rise to
the formation and isolation of compounds. Carbon-12 is far more
abundant than carbon-13, but because it is lighter, compounds,
such as carbon dioxide, which are made from it are easier to lift into
space than those containing the heavier isotope. On Mars there
is an enrichment of about 5 % carbon-13 relative to the amount
on Earth. This suggests that Mars has lost significant amounts of
carbon (most likely as carbon monoxide) through a process called
sputtering. Here, cosmic rays impact the top of the atmosphere
and energize carbon dioxide. This splits the carbon dioxide and in
most cases the oxygen atom (or ion) escapes because it is lighter
than carbon monoxide. However, in the right circumstances car-
bon will escape as carbon monoxide ions leaving an uncharged
oxygen atom behind.
In the second paper by Mahaffys colleague, Chris Webster
(Jet Propulsion Laboratory) looked at oxygen isotopes. Like carbon,
oxygen was strongly enriched in the heavier isotope, again suggest-
ing that Mars had lost a significant amount of oxygen. Comparison
of Curiositys results with those obtained from examining Martian
meteorites suggested a rather violent scenario. Pieces of Mars
often end up on Earth as meteorites. These will be delivered when
large impacts blast rocks into space. The most famous Martian
visitor is ALH84001, the meteorite reputed to possibly show signs
of Martian life. This rock dates to four billion years ago. Other
iron-rich basaltic rocks are called Shergottites and are often
younger, dating to a few hundred million years old. Comparisons
of ALH84001 and these younger Shergottites implied that Mars
had effectively attained its present atmosphere as early as four bil-
lion years ago. Curiositys measurements broadly agree with this
but allow for an additional slower loss in the intervening years.
It looks like Mars took on its present aspect a long time ago.
How might this have happened? Consider a few factors: firstly
Mars has a low mass so its relatively easy for Mars to shed gas.
Secondly, a low mass means that Mars would have cooled down
faster and subsequently lost its early volcanism that would oth-
erwise replenish gas. That is not to say the Tharsis volcanoes are
utterly dead; but they certainly cant and arent supplying the atmo-
sphere with gas at a detectable level now. Sulfur dioxide gasa
key gas from volcanoesis effectively undetectable at Mars, now.
Thirdly, Mars receives less radiation from the Sun because it is
further away from it than the Earth or Venus.
Picture the scene. Volcanic activity on Mars is waning and
becoming episodic. Although there are oceans, Mars shows no
evidence for plate tectonics. Those planet-wide stripes are cer-
tainly reminiscent of the magnetic striping pattern on the ocean
floor but they are not the same and occur in crust far too thick
to subduct. As far as we can tell, without subduction there are
no plate tectonics. Without plate tectonics carbon dioxide drains
out through sequestration into the oceans and ultimately into the
Martian crust as carbonate rocks. Simultaneously, without exten-
sive volcanic activity there is insufficient gas being returned to the
atmosphere. Like Titan (Chap. 9), with a low gravity the Martian
atmosphere is relatively puffy and accessible to the solar wind.
Once the geodynamo on Mars failsapparently within the first
few hundred million years, water vapor is attacked by ultraviolet
light and cosmic rays and is lost to space.
Once sufficient water vapor is split, and once most of the
atmospheric carbon dioxide has either reacted with rocks to form
carbonates, or been split high in the atmosphere to liberate oxygen
and ultimately carbon monoxide, the temperature of Mars falls.
There is little or no cycling of gases from the crust into the mantle
and back again so once these have gone into the crust there is
no coming back. As temperatures continue to decline what water
vapor that remains first rains then snows out on the Martian sur-
face. By four billion years ago, despite ongoing volcanic activity,
the Martian atmosphere has largely been lost. There will be waves
where the density of gases periodically increases. These will coin-
cide with larger bursts of volcanism. On Mars once the crust stabi-
lized this was confined to the Tharsis bulge, where an underlying
hotspot appears to have fueled volcanic activity in bursts separated
by calm intervals lasting 200 million years. These volcanic out-
bursts could have raised the density of gas high enough to allow
liquid water to flow. Although there is good (but still controver-
sial) evidence that much of the Martian northern hemisphere was
initially flooded by an ocean, most of the Martian record implies
only periodic inundations. These could have been caused by vol-
canic activity melting glacial deposits, or indirectly thorough
volcanic out-gassing raising the pressure and temperature of the
Martian atmosphere to the point where surface ice melted. This
still remains unclear.
With all of the evidence in hand, what is (or are) the likeliest
explanations for the loss of the Martian atmosphere?
Jeans Escape
The simplest to understand, but the least important mechanism

for most planets is Jeans Escape. Particles pick up energy from
solar and other radiation sources. This is manifest as an increase
in the kinetic energy of the particle. As kinetic energy scales with
mass, it is the lightest gases, hydrogen and helium that are easi-
est to accelerate through this mechanism. Heavier gases, such as
oxygen (32 g for every 6 hundred billion trillion atoms) or carbon
dioxide (44 g per 6 hundred billion trillion molecules) are usually
too massive to be moved efficiently by this route. If the gravita-
tional pull is low, because the mass of the planet is low, then the
velocity that the particles need to achieve in order to escape is
also correspondingly low. Within the Earths atmosphere oxygen
atoms have a mean velocity less than 1.1 km per secondor less
than one tenth the escape velocity. Therefore Jeans escape has no
measurable impact on the loss of gases similar to or more massive
than oxygen on Earth. On Mars the escape velocity is only about
4 km per second and although this is still higher than the mean
velocity of oxygen atoms, it is closer. The key term is mean veloc-
ity: many oxygen atoms have the mean velocity, but a significant
proportion will have velocities that are higher than this and can,
therefore, escape easily to space.
Both Venus and the Earth are sufficiently massive that very
little of the particles making up their atmospheres are energetic
(or hot) enough to escape in their own right. In Jeans Escape (ther-
mal) mechanism gas particles gain sufficient kinetic energy from
the Sun that they can escape the planets gravitational pull. With
the exception of hydrogen and helium, most of the particles in our
atmosphere and that of Venus are simply too massive and hence
too slow moving for this mechanism to really work. For this mech-
anism to be efficient the gravitational pull of the planet has to be
low or the temperature of the gases higher. For Mars the gravita-
tional pull is sufficiently low enough that the Jeans mechanism
can work, however, it isnt quite enough to strip Mars down on its
own.
Charge Exchange
If an energetic ion trapped within the planets magnetic field col-

lides with a slow moving neutral atom, the atom can pick up the
charge of the ion. This leaves the fast moving ion free of charge
but still with enough kinetic energy to escape. This is not an issue
for Mars or Venus, which lacks an appreciable field, but it does
contribute to the loss of gases above the Earth.
Hydrodynamic Escape
In this situation the atmosphere is heated so strongly by its star

(or by tidal forces) that it simply flows off its parent world, much
like a wind into space. The high temperatures generate high pres-
sures (Chap. 1) that are sufficient to lift the gas free of its gravita-
tional shackles. Given enough kinetic energy and this gas escapes
beyond the gravitational limit of the planet and is captured by

the solar wind. This process is very evident in a few, unfortunate
extrasolar worlds, which lie perilously close to their parent stars.
HD 209458b, or GJ 436b, the latter discovered by the now defunct
French craft CoRoT, show strong ultraviolet absorption in a large
comet-like cloud that extends away from each world. In both cases
Hubbles ultraviolet camera was used to examine the cloud of gas.
Although GJ 326b is likely to survive its ordeal largely intact and
is already six billion years old, the younger HD 209458b is not
going to be so fortunate. Losing at least 10,000 metric tons per
second (or 10,000 times the rate of loss of gas on Earth, Mars and
Venus), HD 209458b is likely to be stripped down to its core before
its parent star becomes a red giant (Chap. 10).
Solar Wind Stripping
Like Venus, Mars lacks any appreciable magnetic field, and what
field there is localized primarily to regions in the southern hemi-
sphere. Consequently, like Venus the solar wind can directly inter-
act with the top of the Martian atmosphere. Like Venus, its not
entirely a one-way street. Sunlight and energetic particles from
the Sun ionize gases at the top of the atmosphere forming an iono-
sphere. In turn these ionized gases move around under the influ-
ence of the solar magnetic field and under the influence of light.
As a result a magnetic field is induced within these flowing, ion-
ized gases, which then shield the bulk of the atmosphere from
the scouring action of the solar wind. Around this region a bow
shock forms that also deflects the bulk motion of the solar wind,
before it can whittle away at Marss tenuous gases. However, as
the Martian atmosphere is less dense than that of Venus, the effect
is weaker. Moreover, with a weaker gravitational pull, many of the
gases that Venus could hold onto, Mars cant and it loses a more
substantial portion of its atmosphere per second than either the
Earth or Venus.
Like Venus, Mars does suffer from wind stripping. With a
weaker gravitational pull and a less dense atmosphere, the Martian
air is more prone to stripping than that of either Venus or the Earth.
This is particularly true when the solar wind is running at its
strongest and the Martian bow shock is shoved closer towards the
surface of the planet. At these times, the denser Venusian atmo-
sphere develops a stronger induced field and hence is more thor-
oughly shielded. However, Mars is unable to pull off this trick as
its thin atmosphere cannot respond to the change in the strength
of the wind and induce a stronger magnetic shield in the way that
the Venusian atmosphere can.
Mars has another chink in its armor. As weve already seen,
during the largest Martian dust storms the middle and upper por-
tions of the troposphere warm so much that the entire atmosphere
puffs up. This, in turn, exposes more of upper atmosphere to the
erosive effects of the solar wind. Mars, it seems, has a rather self-
destructive side, for the more atmosphere that is lost, the greater
the difference in surface temperature and this leads to greater
storms at particular times of the year.
If we look at the Martian atmosphere today, in all solar wind
stripping accounts for less than a third the rate of gas loss from
Mars through these non-thermal processes. However, MAVEN
data suggests that it was this process, and the additional erosive
effects of solar storms, that ripped the Martian atmosphere out of
Mars puny gravitational clutches. Today, other non-thermal pro-
cesses, including stripping by asteroid and comet impacts; and the
effects of charging of the atmosphere by solar and cosmic radia-
tion. None of these are relatively important. Mars, thus, lost most
of its atmosphere early on to the young Suns more powerful solar
wind. MAVEN data suggests that this happened between 4.2 and
4.0 billion years ago. Meanwhile, today it loses most of its gas
through thermal effects: the Jeans Mechanism (described above),
with non-thermal mechanisms accounting for the remaining loss.
Remember, that in the Jeans mechanism gas particles have to
attain sufficient energy to overcome the gravitational pull of the
planet. Very little gas in the atmosphere of Venus comes anywhere
close, despite lying being closer to the Sun and its gases being
heated more strongly. Mars loses out simply because it is a small
planet with a low gravity. The lack of an appreciable magnetic
field is simply the icing on the cake. Could you combine Mars
with Mercury and youd get a world with enough gravity (and with
a magnetic field) that would be far more habitable in the longer
term? I am not aware anyone is planning to maneuver these two
worlds together. However, there is a small, but reasonable chance

that repeated gravitational pulls from Jupiter might lead Mercury
to come off its orbit and collide with Mars in the next few billion
years. Earth: watch out!
While Venus loses gas through the charging and repulsion,
Mars is actively stripped. Venusian gases also undergo more chem-
ical reactions within their atmosphere than do those of the Earth
or Mars. This is because the particles within the Venusian atmo-
sphere are closer to the Sun and, therefore, receive more energy,
which in turn encourages chemical reactions amongst them.
These reactions can release further energy that helps whisk them
out of the atmosphere.
The atmospheres of Mars, Earth and Venus lose mass at a
similar rate. Unfortunately, Mars has less to lose (2.5 1016 kg
for Mars versus 5.15 1018 kg for the Earth) and, was its current
rate of loss to be maintained, would bleed dry in less than 2.9 bil-
lion years. Compare that to the Earth, where we wouldnt lose
our atmosphere for another 591 billion years or so (Chap. 5)if
the Sun didnt get us first. Mars wont be a place to go when the
Sun begins to parboil the Earth. However, this figure assumes that
the atmosphere isnt replenished from surface or interior stores
which, of course, it is. As the atmosphere bleeds away to space,
iceboth water and carbon dioxide, sublimates on the Martian
surface and partly compensates for the loss. However, Mars will
never approach the conditions it might have had when the solar
system was young. The planet has lost too much gas and has dwin-
dling stores from which to replenish this.
There is one final possible mechanism through which Mars
lost most of its atmosphereand might explain why it hap-
pened four billion years ago. Enter the joker: the Late Heavy
Bombardment. From 4.2 to 3.9 billion years ago, the Earth and
Moon appear to have suffered catastrophic impacts, with objects
up to 100 km across. These massive objects appear to have begun
their life far out near the current orbit of Neptune. During this
interval, Jupiter and Saturn appear to have booted Uranus and
Neptune out of their earlier and warmer orbits to the outer fringes
of the solar system. During these maneuvers, vast quantities of icy
debris were scattered in all directions, with a significant amount
falling in towards the Sun. Mars would not have escaped this bom-
bardment. In the case of the Earth, our substantial gravitational
pull would have held onto most of the impact materialand more
importantly much of the gas energized and blasted upwards by the
impact.
In such an event, the blast wave expands radially in all direc-
tions from the impact point and most of the material that escapes
(at least temporarily) into space vents upwards like an enormous
rocket exhaust through the hot channel the incoming asteroid or
comet has carved through the atmosphere. This forms an enor-
mous mushroom cloud that is centered on the impact point. With
a low gravity, not only can more of the impacting object escape
into space, but also a larger volume of hot gas can be vented, as
well. Unable to hold onto this, such material is captured by the
solar wind and vented into interplanetary space. Such impact ero-
sion was likely more significant to early Mars than any gains it
would have made from the impactors themselves. However, the
precise contribution of atmospheric erosion by impacts is still
unclear. All we can currently say with confidence is that Mars
was bled dry early on. From four billion years ago, Mars underwent
a slower process of attrition that was only delayed by the slow
sublimation of its remaining icy inventory and a declining input
from volcanic activity. This process appears to have been primar-
ily caused by the Jeans Mechanism and Solar Wind stripping of
what little atmosphere, Mars retained.
How then would we sum up the manner in which Venus,
Earth and Mars lose atmosphere? On Venus, gas loss is primar-
ily through two non-thermal processes: electric field acceleration
and photochemical reactions (Chap. 5). In the former case, elec-
trons are driven off and this creates a strong enough electric field
to accelerate ions. In the latter sunlight splits carbon dioxide into
carbon monoxide and free oxygen. When these flow to the night
side and recombine they release sufficient energy to propel many
of the oxygen molecules into space. Mars, by contrast loses its
atmosphere to the solar wind, when particles achieve sufficient
energy to escape. Thus, Mars is stripped by a combination of ther-
mal and non-thermal processes that lift gases gradually out into
space, with a likely additional contribution to atmospheric loss
from impacts during the Late Heavy Bombardment.
Despite our perception of a deep, impenetrable atmosphere,

the Earth has already lost most of its original bulk. Much hydro-
gen has escaped through the Jeans mechanism, but most of our
atmosphere has been lost to our planets surface. Vast quantities
of carbon dioxide, along with most of the atmospheres water have
condensed onto its surface, or have chemically reacted with it.
Such processes do happen on Mars and a substantial body of water
remains frozen. Carbon dioxide also freezes out in a seasonal pat-
tern. Venus will have also lost some of its atmosphere to its sur-
face through chemical reactions with erupted rocks. How much
this is remains in the realm of mysteryand will stay that way
until someone comes up with a credible means of sampling the
Venusian surface.
All in all, Mars suffered solely because it has a relatively low
mass. For not only is Mars unable to maintain a magnetic field,
but it incapable of generating a sufficiently strong gravitational
pull to hold its atmosphere in place. With no means of replenish-
ing its mass, Mars bled first to space and then later to its surface.
That Mars can maintain any real atmosphere today is a testament
to the slow sublimation of what remains of its former riches from
its planets icy surface.
Future Mars
Mars is a desiccated world, but it is not one without potential.
Large reserves of water ice and carbon dioxide ice carpet the planet.
One of the more interesting discoveries of the last few years has
been the discovery of glaciers and glacial deposits buried in dust
near the Martian equatorial highlands. Quite how much water
remains is unknown but it may be enough to refill most of the
Martian northern basin were it to be thawed out.
Many people have hypothesized about the potential to
reclaim Mars as a habitable world. In order to realize this dream
the pressure within the atmosphere must rise to at least one tenth
that at the surface of the Earth. Any less than this and water will
not be stable in its liquid form. To put this in perspective, at the
tropopause the average pressure is around one fifth that at the
Earths surface. This is still 20 times the pressure currently found
at the surface of Mars. So how will we raise the pressure to the

point at which water could be stable at reasonable (020 C) tem-
peratures?
In a paper published more than 20 years ago in Nature,
Christopher McKay and James Kastings outlined what would be
the simplestand fastestscenario. This was revisited by McKay
and Margarita Marinova a decade later, with a consideration of
the ethical issue associated with terraforming. In essence terra-
forming Mars will all be about thawing it out. In order to do this
cheaply and sensibly future terraformers would likely flood the
lower atmosphere with fluorinated carbon compounds. The sim-
plest and possibly the best would be tetra fluorocarbon (CF4). Not
only is this easy to makeand could be manufactured on Mars
but it is sufficiently stable that the fluorine wont escape its chem-
ical bonds and erode any future, protective ozone layer.
After a few million tons of tetrafluorocarbon are released
the Martian greenhouse will strengthen and the temperature will
begin to rise. Get it right and in about 100 years the rising tem-
perature will sublimate carbon dioxide and water from the frozen
Martian crust. Moreover, the water will react with the copious
amount of perchlorates and decompose them. This will release
oxygen gas and begin the process of making the planet suitable for
microbial and more complex lifenotably the construction of a
thicker ozone layer.
The process should become self-sustaining with the increas-
ing concentration of greenhouse gases driving higher tempera-
tures, more thawing and thus the release of more greenhouse gas.
Although the next steps would take millennia, eventually the bal-
ance of energy trapped by carbon dioxide, carbon tetrafluoride and
water vapor will match that escaping and hold the temperature
at 1020 C. Atmospheric pressure will never equal that on the
Earth, but it could reach 300400 millibar, or roughly the pressure
found just below the summit of Mount Everest. Humans would
need to wear oxygen masks and any large animals would need to
be kept under pressurized domes, but otherwise Mars would be
somewhat akin to life on the Tibetan plateau. Complex plants
could take over the job of creating oxygen and maintain habitabil-
ity, as well as producing food for the future Martians although the
low atmospheric pressure would likely slow their growth.
If we decide to let nature do the work we will have some time

to go. For it wont be until the Sun is around 15 % brighter than
it is at present that Mars will receive enough radiation to begin
to thaw it out. For us, this will be around 1.41.6 billion years in
the future. If we are lucky, just as the habitable phase of the Earth
wanes and finally expires, Mars will blossom. Imagine watching
the Earth turn from blue to brown from the greening surface of
Mars? However, the word blossom should not be taken too liter-
ally. At its temperate peak Mars will be about as warm as ice age
Earth If one wished to wait until Mars was as warm as the current
Earth, we would need to wait another 4.5 billion years. For only at
this point would the Sun be as bright to make Mars feel truly like
a second home.
How long could Mars remain habitable? With no interven-
tion and no magnetic field to protect its low gravity and, conse-
quently, puffy atmosphere, the same damage will begin to destroy
what humans or nature have created within a few hundred million
years (at most). Stripping of gases by the solar wind; sputtering by
cosmic rays and Jeans Escape will strip the water down to hydro-
gen and drag it away into space. Carbon dioxide will follow; and
there will then begin a protracted phase where Mars returns to its
current dry, desertified state. Unless we chose to intervene and
continually replenish the planet with water, perhaps from comets,
there will be no going back. Mars will be left bone dry and as bereft
of air as Mercury by the time the Sun leaves the main sequence.
Long before Mars could hope to attain the terrestrial temperatures
of today it will be utterly airless and lifeless.
Conclusions
Mars is a complex world. Its atmosphere has undergone a trau-
matic series of changes since the planet formed 4.5 billion years
ago. Soon after this time an extraterrestrial visitor to our Solar
System might have seen three, pristine habitable worlds: although
free from oxygen, Venus, Earth and Mars would have hosted dense,
cool, moist atmospheres with extensive oceans carpeting their sur-
faces. Although Venus is somewhat problematic in this scenario
(Chap. 5), that the solar system could have been habitable in this
regard is truly fascinating. The loss of two of these worlds was due
to the success of greenhouse gases in one and the failure of them
in the other. Venus spawned a moist, then dry super greenhouse
and its atmosphere boiled dry, while the atmosphere of Mars was
whittled down and eventually froze dry. Only the Earth had a com-
bination of sufficient mass and greenhouse gases to maintain its
habitability.
While Venus paints a picture of how the Earth will be, Mars
reminds us the planetary ecosystem is fragile and malleable in
equal measure. Add extra greenhouse gases to Mars and it will
warm and thaw. By the same token adding additional greenhouse
gases to the Earths atmosphere is currently causing it to warm.
The fatal fallacy that adding extra greenhouse gases will have no
effect is a virus which permeates the internet and leads many to
believe carbon dioxide and methane are irrelevant (Chap. 3). Close
examination of our terrestrial cousins, Mars and Venus exposes
the flaw within this argument. For if it were true all those ideas
of terraforming Mars through the addition of more greenhouse gas
would clearly come to nothing. Moreover, there would be no link
between high density of greenhouse gases and the high tempera-
tures found at Venus.
That the Earth has an average temperature of +15 C instead
of 18 C, while Venus has a mean temperature of 470 C instead
of 43 C is a testament to the additive power of greenhouse warm-
ing. By trapping much, but not all of the outgoing radiation these
gases maintain planetary temperatures. Should we wish to terra-
form Mars (or even Venus) it will only be through the manipula-
tion of the concentration of these gases that these dreams become
possible.
A well-known terrestrial meteorologist is keen to equate the
ability to dream with burning fossil fuels. I might agree but not in
the way he envisages the power of dreams. Surely, the only way
we can ensure our future prosperity is by imagining about how
we make it sustainable; not by keeping our fingers crossed and
daydreaming our way into an irretrievable situation. Dreaming
is all well and good only for as long as you can separate what is
achievable and what is merely a wasteful desire. There is a vast
world of available energy out there: why restrict ourselves to those
forms which are most harmful for our future?
References
1. Fenton, L. K., Geissler, P. E., & Haberle, R. M. (2007). Global warming and climate forcing by
recent albedo changes on Mars. Nature. http://humbabe.arc.nasa.gov/~fenton/pdf/fenton/
nature05718.pdf.
2. Wang, H., Richardson, M. I., John Wilson, R., Ingersoll, A. P., Toigo, A. D., & Zurek, R. W.
(2003). Cyclones, tides, and the origin of a cross-equatorial dust storm on Mars. Geophysical
Research Letters, 30(9), 1488. http://www.gfdl.noaa.gov/bibliography/related_les/hqw0301.pdf.
3. Haberle, R. M., Leovy, C. B., & Pollack, J. B. (1982). Some effects of global dust storms on the
atmospheric circulation of Mars. Icarus, 50(23), 322367.
4. John Wilson, R. (1997). A general circulation model simulation of the Martian polar warming.
Geophysical Research Letters, 24(2), 123126.
5. Bertaux, J.-L., Leblanc, F., Witasse, O., Quemerais, E., Jean, L., Stern, S. A., et al. (2005).
Discovery of an aurora on Mars. Nature, 435, 790794.
6. Leshin, L. A., Mahaffy, P. R., Webster, C. R., Cabane, M., Coll, P., Conrad, P. G., et al. (2013).
MSL science team. Science, 341(6153), 1238937. doi:10.1126/science.1238937.
7. Pamela G. Conrad., Dan Harpold., John J. Jones., Laurie A. Leshin., Heidi Manning., Tobias
Owen., Robert O. Pepin., Steven Squyres., Melissa Trainer., MSL Science Team. Science
341(6143), 263266.
8. Webster, C. R., Mahaffy, P. R., Flesch, G. J., Niles, P. B., Jones, J. H., Leshin, L. A., et al.
(2013). Isotope ratios of H, C, and O in CO2 and H2O of the Martian atmosphere. Science,
341(6143), 260263.
9. McKay, C. P., & Marinova, M. M. (2001). The physics, biology, and environmental ethics of mak-
ing Mars habitable. http://www.ifa.hawaii.edu/~meech/a281/handouts/McKay_astrobio01.pdf.
7. The Gas Giants
Introduction
Saturn and Jupiter have more in common than they have in iso-
lation, thus Chap. 7 compares and contrasts the two giant mas-
ters of the Suns planetary realm. Despite some issues of scale,
the broad brush strokes of each planet are rather similar. Within
this broad canvas there are specific differences in the appearance
of each world and the behavior of its weather. The underlying
reasons such as differences in the amount of sunlight each world
receives are understood, but how this manifests as differences in
the appearance of each world remains largely in the realm of spec-
ulation. As more probes venture into their territory there remains
much to be learned and more than enough to discuss, pontificate
and speculate over. Welcome to our giants.
The Structure of the Giants

While the interiors of the terrestrial planets can largely be ignored
when it comes to weather, primarily locked in a solid crust as they
are, the atmospheres of all four of the giants are so tightly and inti-
mately intertwined with their cores that they must be addressed.
In particular, the gas giants, Jupiter and Saturn, have atmospheres
that emerge from the sea of liquid hydrogen and helium that forms
the bulk of their interiors. Additionally, there can be a strong
transfer of energy by convection from their liquid centers to the
atmosphere above (Fig. 7.1).
Jupiter has an atmosphere that is roughly 5000 km thick,
and Saturns is similar. Of this depth, only around 100 km lies
below the cloud tops on Jupiter (Figs. 7.2 and 7.3). At Saturn this
figure is somewhat higher at around 320 km. On each world the
clouds lie where the pressure is conducive to their formation at

DOI 10.1007/978-3-319-25679-5_7
Atmosphere
Super Critical Hydrogen
Core Core
Metallic Hydrogen
FIG. 7.1 The internal structure of the gas giants. The bulk of their vol-
ume (and mass) consists of liquid hydrogen and helium surrounding a
possible rocky-metallic core. Both planets have diameters in excess
of 100,000 km. However, Jupiter as a little over three times the mass of
Saturn the degree of internal compression is far higher. Jupiter and Sat-
urns atmospheres are roughly 1000 km thick, with most of this above
the cloud tops. Roughly 100 km beneath the ammonia cloud layer on
Jupiter the pressure reaches 10 bars and the atmosphere gives way to an
increasingly thick soup of hydrogen. This transition is somewhat deeper
on Saturn at around 320 km, due to Saturns lower gravity. Beneath this
hydrogen slowly morphs into a thick fluid. 20,000 km down on Jupi-
ter and 30,000 km down on Saturn the liquid hydrogen becomes metal-
lic and its motion generates each planets magnetic field. Temperatures
reach 35,00040,000 K at the core of Jupiter and around 11,700 K inside
Saturn, where helium rain may explain some of Saturns extra heat
around 0.11 bar for ammonia and up to 10 bars for water vapor
(Fig. 7.2). These conditions are reached at slightly deeper levels on
Saturn with water vapor condensing below 10 bar and the domi-
nant ammonium hydrogen sulfide layer at around 7 bars of pres-
sure. The differences are a consequence of the lower temperature
of the Saturnian atmosphere and the difference in gravity: mighty
Jupiter squeezes its layers more tightly than fluffier Saturn due to
its larger mass.
At about 10 bars of pressure (equivalent to around 100 m
depth in our oceans, or one ninth the pressure at the surface of
Venus) molecular hydrogen gradually takes on the appearance and
properties of a liquid. This is not an abrupt change but a grad-
ual one. On Jupiter this layer is 20,000 km deep and for the first
The Gas Giants 275
1/billion 1,000
Altitude relative to terrestrial sea level

1/10,000 1/hundred million
750
Thermosphere
Pressure (bar)
500
Stratosphere-Thermosphere Boundary (320 km)
(km)
250
Stratosphere
Tropopause (50 km)
A 0
1
B
C
Troposphere
25
-250
-150 -100 -50 0 50 100 150 200 1,000
Temperature (oC)
FIG. 7.2 The structure of Jupiters atmosphere. Clouds occur in three

decks in the troposphere at or below 1 bar in pressure. 1 bar corresponds
to terrestrial atmospheric pressure at sea level. Cloud deck A consists
of ammonia ice crystals at a temperature of 150 C and an altitude of
23 km above terrestrial sea level. Below this (B) are clouds of ammo-
nium hydrogen sulfide at around 1 bar and water (C) at 50 km or so.
Above the tropopause there exists a stratosphere with broadly constant
temperature (red line), and above 320 km we reach the thermosphere
where charged particles from the Sun cause temperatures to rise rapidly
(Chaps. 1, 5 and 6)
10,000 km or so the mix of hydrogen and helium is more gas than

liquid. Beneath this youd be hard-pressed to think of this sub-
stance as a gas. On Saturn, the liquid hydrogen layer extends from
320 km down to a depth of 30,000 km, or half the way to Saturns
core. Hydrogen, in this state, is known as a super-critical fluid
something we encountered in Chap. 5. To recap, in this odd state,
hydrogen adopts the ability of a liquid to dissolve substances, but
retains an ability to effuse through solids like a gas. It is also an
excellent conductor of heat.
Within this layer there are some important processes taking
place. On Jupiter and Saturn, when temperatures exceed 1000 K
(727 C) and around 40,000 bars of pressure methane begins to
transform into other compounds. In the absence of other materials
this tends to be longer chain of hydrocarbons and eventually graph-
ite, but in the dense interior of the gas giants the outcome is less
certain. The presence of ammonia and hydrogen sulfide as well
200
Stratosphere Stratosphere
100
Altitude relative to terrestrial sea level

Haze Tropopause
Tropopause 0
Haze
(km)
-100
-200
-300
Troposphere Troposphere
-150 0 150 -150 -50 0 50 100
Temperature (oC) Temperature (oC)
FIG. 7.3 A comparison of the lower atmosphere of Jupiter (left) and Sat-
urn (right). The cloud layers on Saturn have the same composition as
those above Jupiter (Fig. 7.2), but lie at lower altitudes. This is simply
because Saturn has a colder atmosphere so the condensation levels are
lower down. The lower atmospheric pressure and lower temperature gra-
dient means that the cloud decks are more spread out The upper atmo-
sphere of Saturn is also filled with a haze of carbon-based compounds
which is largely absent above Jupiter. This mutes the appearance of the
banded cloud structure on Saturn. The tropopause marks the lid of the
troposphere as on Earth, while 0 km marks 1 bar of atmospheric pres-
sure, or sea level on Earth
as water favor the production of more complex compounds, until

at even greater temperatures and pressures these break down into
others. This means that while the bulk of both Jupiter and Saturn
is hydrogen and helium, there is a substantial amount of heavier
elements that cant be ignored.
In some speculative models these heavier compounds form a
kind of hot icy mantle around the core of these planets. Indeed,
the presence of a core at all is unclear on Jupiter. Even if it began
life with one several tens of times the mass of the Earth, it could
well have dissolved by now into the metallic hydrogen ocean that
lies beneath 20,000 km inside Jupiter. The presence of a core in
Saturn is more likely and measurements by Cassini suggest some-
thing with a mass between 9 and 22 times the mass of the Earth
may well lie at Saturns heart.
The Gas Giants 277
While Jupiters metallic hydrogen layer is too hot to allow

stratification, or the settling of materials, the layer on Saturn is
not. It is suspected that the helium which should compose around
24 % of the material has condensed or is condensing. This gener-
ates a rain of helium falling through to the metallic layer below.
As this material descends it releases gravitational potential energy.
This is realized as heat which warms Saturns interior and allows
this smaller world to mimic Jupiter, releasing more than twice the
heat energy the planet receives from the Sun. Jupiter, meanwhile,
has remained hot from its formation due partly to the steady
release of gravitational potential energy as the planet contracts.
Between the core and the liquid hydrogen ocean lies a thick
layer of metallic hydrogen. Under a pressure three millions of times
that found at terrestrial sea level (3 million bars) the super-critical
fluid hydrogen undergoes a more profound rearrangement, forming
an ordered structure more akin to a metal. In this state the elec-
trons that are normally bound one per nucleus can break free and
become delocalized, making the hydrogen a super-efficient con-
ductor of electricity. Stirred by both convection and the rapid spin
of both planets, this metallic stew can generate magnetic fields
which enshroud each world. Jupiters is particularly showy, extend-
ing outwards over 700 million kilometers to the orbit of Saturn.
The Color of Giant Planets in the Solar

System
On Earth the color of a cloud contains information about its nature.
Aside from the ubiquitous whites and greys there are a host of
other colors that can give a clue to impending weather. For exam-
ple Fig. 7.4 shows a small sample of the possible variety of shades
caused by the effect of water and dust on the light that ultimately
enters our pupils. In the dawn or evening sky a yellow hue usually
signifies the presence of abundant water vapor and the likelihood
of coming rain. A very pale, turquoise blue between clouds has a
similar warning during the middle part of the day. Green is rare,
but if a green cumulonimbus thundercloud is approaching you can
expect severe hail. A copper tint around the edges of clouds or along
part of their base (Fig. 7.4a) tends to indicate electrical charging
and hence the threat of lightning. In each case it is the scattering
of light around dust or water droplets and ice crystals; or the refrac-
tion of light through droplets of water that produces these different
colorations. In the atmospheres of the giant planets there exists a
bewildering array of additional chemicals that can add their own
absorptions to expand this array of colors even further.
The chemical composition of Jupiter and Saturns atmo-
sphere is very similar. The cloud layers are made of water vapor,
overlain by ammonium hydrogen sulfide with wispier ammonia
clouds on top (Figs. 7.1, 7.2 and 7.3). On Jupiter the ammonia
clouds sit in a hazy layer of hydrocarbon compounds and com-
pounds called tholins that are produced through the reaction
of ammonia and the abundant methane. Methane itself does not
form clouds simply because in the atmospheres of both Jupiter
and Saturn it is too warm.
How do scientists know what the clouds are made of, particu-
larly since the water clouds are effectively hidden from view? To
some extent the presence is inferred because oxygen is one of the
universes most abundant chemical elements and most of it will be
found as water vapor or ice at Jupiters orbit and beyond. Moreover,
the presence of lightning within the atmospheres of both planets
atmospheres suggests that water clouds are present. This assump-
tion is because on Earth water vapor is needed to form the ice
that ultimately helps separate charge in thunderclouds, however,
ammonia has a chemical structure not too dissimilar from water
vapor, with molecules able to separate charges just as water does,
so this is a little dubious as evidence.
Fortunately, the Galileo Orbiter probe was able to detect water
vapor from measurements made in orbit. Additionally, the violent
collision of Comet Shoemaker-Levy 9 with Jupiter in 1994 blasted
deeper material upwards, beyond the cloud tops. Spectra taken from
Earth and from Galileo indicated that there was water present but
that it was at much lower level than was expected. Quite why this
is so remains unclear. In the case of the Galileo Orbiter, it appears
that there was a bit of bad luck involved in the trajectory as the
probe descended through a relatively dry belt with limited cloud
cover. In the case of the comet, its plausible that the comet may
have vaporized above the 10 bar level where water clouds would
be formed. In this case water clouds were present, but the comet
was unable to dredge them up to the surface where terrestrial
The Gas Giants 279
spectroscopes, and those on Galileo, could observe them.

Alternatively, our assumptions that Jupiter has a thick layer of
water clouds is wrong and this layer is so thin that it rapidly gives
way to the sea of hydrogen and helium immediately underneath.
However deficient Jupiter appeared to be in terms of water, the
cometary impacts did reveal a host of other interesting chemicals in
Jupiters atmosphere. Amongst these were hydrogen sulfide, ammo-
nia, sulfur, carbon disulphide and it was thought that it was some of
these sulfur-containing compounds that gave Jupiters deeper layers
their color. Yet despite many attempts the reddish color of these
bands could not be reproduced with sulfur chemistry. Instead these
deeper belts and the eye of Jupiters Great Red Spot are red because
of some complex carbon-chemistry involving ultraviolet light. The
Great Red Spot is red (or rather a varying shade of pink) because the
storm carries ammonia high into the atmosphere (Fig. 7.4).
High at the tropopause, ultraviolet light can react with the ubiq-
uitous methane haze to produce acetylene (ethyne)the gas used in
welding. This is carried downwards into the belts or the eye of the
great red spot where it encounters ammonia that has been driven
upwards. With a bit of a spark from ultraviolet radiation the two gases
combine to produce a pinkish compound. Without the influence of
solar ultraviolet light the clouds would mostly be white as they are
on Earth. Experiments suggest that were ammonium hydrogen sul-
fide irradiated, the clouds would appear slightly green. One can think
of the pink coloration as a kind of Jovian sunburn. Wherever the
atmospheric circulation brings ammonia and acetylene together and
are exposed to sufficient ultraviolet radiation they take on the charac-
teristic pink coloration. One can see this process in action with some
of Jupiters new storms that are joining the Great Red Spot (GRS) to
pockmark Jupiters otherwise stripy complexion.
Out by Saturn the colder and lower pressure atmosphere
leads to condensation of ammonia and ammonium hydrogen
sulfide at deeper levels (12 bars instead of 0.5 bars at Jupiter).
Consequently, there is a greater burden of haze above the clouds
which obscures them, much like smog above a terrestrial city.
Saturn, thus, appears rather muted in color with less pronounced
belts and zones (Fig. 7.5).
Even further out in the Solar System, by Uranus and beyond,
methane clouds contribute white coloration, but it is the deep
FIG. 7.4 The color of clouds. Jupiters Great Red Spot (center) is a distinct
pinky-orange color, with distinct differences depending on the depth to
which we see the cloud. Is this a true color caused by a chemical sub-
stance, or is it caused by something more prosaic, such as scattering of
sunlight? An approaching thunder-storm delivers the usual dark grey but
also a distinctive copper color indicative of charge separation (a, above).
(b, below) shows a partial circumzenith arc produced by refraction of sun-
light through high, thin ice clouds. (c, above) shows cirrocumulus clouds
demarcating the Polar Front Jet Stream. Scattering of light from the set-
ting Sun gives an orangepink glow on the horizon. (d, below) shows a
pervasive yellow glow caused by scattering of light on water vapor. Image
credits: Great Red Spot (NASA/Voyager 2); Remaining photographs by
author ((a)Hunstanton, UK; (b)Glasgow, UK; (c)Nottingham, UK
and (d)Kyle of Lochalsh, UK)
expanse of methane in general that colors these planets blue or

greenish-blue. The red and yellow bands that are visible in the atmo-
spheres of Jupiter and Saturn are simply buried far below the deep
ocean of blue methane. Saturn forms a yellow transitionary mem-
ber, Uranus appears somewhat green, while the transformation
is complete at Neptune with only the deep, and oddly terrestrial,
blue visible. Why blue? This comes down to Rayleigh scattering.
The Gas Giants 281
This process contributes to our reds and yellows of sunset and the
blue of our daytime sky, but out at Uranus and Neptune, methane
also preferentially scatters blue light. The longer wavelengths are
absorbed, meaning Uranus and Neptune are blue for the same rea-
son that Earth has a blue sky during the daytime.
At Jupiter and Saturn the yellowish haze also shares its origin
with ultraviolet light. Carl Sagan and Bishun Khare christened the
chemicals that were thought to color Saturn and Jupiters upper
atmosphere, tholins. Originally, the term was restricted in use to
describe those chemicals thought to provide the haze in the atmo-
sphere of Titan, but the term can be more generally applied to the
atmospheres of all cold worlds, as well as the reddened surface of
Pluto and other Kuiper Belt objects. The precise composition of
these chemicals is less well defined, though they are more than
likely produced in a series of complex photochemical reactions
between ammonia, methane and ethane. With the aid of cosmic
rays, ultraviolet restructures the bonds between these molecules
and produces a complex range of much larger nitrogen and carbon-
based molecules. These are able to absorb a considerable range of
wavelengths, but their low density means that they merely mute
the colors of Jupiters clouds rather than obscure them. At Saturn,
with deeper cloud layers and consequently, the greater depth
of tholin haze, means that the banded pattern of clouds is blander
in its complexion. Above Venus sulfates, principally sulfuric
acid (Chap. 5), are the driver of Rayleigh scattering. Within the
atmospheres of the Earth and Mars Rayleigh Scattering is predom-
inantly caused by silicate dust (Chap. 6). Scattering is bolstered
within the Earths atmosphere by organic molecules produced by
living organisms.
Belts and Braces: The Bands and Storms

of Jupiter
In the atmospheres of both gas giants, clouds are organized into
colorful bands called belts and zones. The belts are thought to
represent areas of down-welling, with the zones as regions of up-
welling. Where air rises ammonia is taken up several kilometers
above the level of the belts. Here it cools and condenses forming
North Temperate Belt (low, warm)

North Temperate Zone (High, Cold)
North Tropical Belt (low, warm)
North Tropical Zone (High, Cold)
North Equatorial Belt (low, warm)
Equatorial Zone (High, Cold)
South Equatorial Belt (low, warm)
South Tropical Zone (High, Cold)
South Tropical Belt (low, warm) GRS
South Tropical Zone (High, Cold)
South Temperate Belt (low, warm)
South, South Temperate Belt (low, South Temperate Zone (High, Cold)
warm)
Zone Belt Zone Belt Zone Belt Zone Belt Zone
FIG. 7.5 The organization of Jupiters cloud belts and zones. Belts rep-
resent areas where cooler air is descending and warming and any clouds
tend to evaporate except at deeper levels. In zones upwelling air cools
and ammonia clouds condense that overly the deeper ammonium hydro-
gen sulfide layer. You can imagine each zone represents the top of a Had-
ley cell, while the belts represent terrestrial Horse Latitudes (Chap. 1).
The rapid rotation of Jupiter generates the much stronger Coriolis effect
that drives the formation of this highly banded atmosphere. Colors in the
lower rendering are not representative. Infrared image of Jupiter: NASA/
JPL/CALTECH
white stripes across Jupiters face (Fig. 7.5). In the deeper zones,
air over-spilling from the neighboring zones descends; warms and
the clouds within it evaporate, exposing the ammonium hydrogen
sulfide layer. A contributing factor to the freshness of the zones
is their altitude. At Jupiter these zones are more clearly visible
because there is less obscuring haze high up. The belts and zones
also counter-rotate around Jupiter and Saturn at speed (Fig. 7.6), so
as air rises and falls it also moves to the east or west with substan-
tial speed giving an overall corkscrew motion.
What is the underlying mechanism that generates these belts
and zones within the giant planet atmospheres, and why do Jupiter
and Saturn differ in their organization of these cloud patterns? On
the Earth, Venus and Mars the driving cells are the Hadley cells
that overlie the equator. These pump energy into the atmosphere
The Gas Giants 283
100
-10
100
-10 150
140 -20 480
-20 450
-20
140 140
-60 -10
120
-20
40
-15
35
FIG. 7.6 The direction and speed (velocity) of the winds on Jupiter (left)
and Saturn (right). On both planets winds are predominantly east-
wards. Air rises in zones and sinks in belts, giving the overall flow a
corkscrew-like pattern as the air barrels across the face of each world.
Airflow around the Great Red Spot (GRS) is westerly along its northern
flank, at around 180 km per hour, and easterly at 140 km per hour on
its southern flank. Airflow around the GRS peaks at 120 m per second
(432 km per hour), while it is roughly stagnant, with modest upwelling
in the heart of the GRS. At Saturn, one of the most interesting atmo-
spheric features is the north polar hexagon. It turns out the 100 m per
second westerly jet at the top of the image is key to the formation of the
north polar hexagon. Meanwhile, further south, the 20 m per second
retrograde (20) jet drove the great storm of 2011. Images: NASA/HST
from the sun-warmed surface. As air flows north or south, away

from the equator, the Coriolis Effect directs the air towards the east
(westerly winds). Air moving in the opposite direction towards the
equator has less eastward velocity than the ground underneath so
appears to curve to the west (as seen in Chap. 1).
North of the Hadley cell in the Martian and terrestrial atmo-
sphere lie the Ferrell cells. Venus has a more subdued equivalent
polar collar, which is broadly equivalent to these. Air flow in
the Ferrel cell is overspill from the Hadley cells and its motion
to the east is driven by air flowing down and around the Hadley
cells. If you need an analogy, run a bath and watch the water cir-
culating. A fast cell spins where the water moves outwards away
from the tap. Further away, a second cell spins in the opposite
direction carrying momentum towards the tail of your bath. On
Saturn and Jupiter this effect alone cant explain the observed
airflow. For one, equatorial winds on Jupiter and Saturn are
westerlyand very strong. Indeed, on both Jupiter and Saturn

the predominant wind direction is westerly at the level of the
cloud tops. This super-rotationwhere air moves more rapidly
in the direction of the planets spin than the underlying planet
is unexpected and was looked at in Chap. 5. It will be looked at
again in Chap. 9 (Titan) and 10 (exoplanets). Models by Tapio
Schneider and Junjun Liu (Caltech) indicate that superrotation is
driven by convection at the equator. Convection from the deep
interior generates eddies and Rossby Waves (Chap. 1), which
allow momentum to flow into the equatorial regions, accelerat-
ing winds into two prograde jets.
There are some differences in direction with latitude and con-
siderable variation in speed (Fig. 7.6). On Jupiter an easterly jet
stream is found roughly 15 north and south of the equator, with
the same feature found at 40 north and south in the atmosphere of
Saturn. Polewards of these latitudes the atmosphere is organized
into alternating easterly and westerly bands with the frequency
increasing with increasing latitude. As the equatorial air on both
worlds is prograde (rotating in the direction of the planets rota-
tion) unlike the Earth, movement must be driven not by conver-
gence at the equator but divergence. Immediately above the axis
separating each Hadley cell there is an easterly jet stream (east
to west) on Earth, while both Jupiter and Saturn have a westerly
(west to east) jet stream.
There are more bands of circulation within each planet com-
pared to the Earth simply because both Jupiter and Saturn rotate
faster. The Coriolis Effect is more than ten times as powerful so
winds deflect to the east or west over proportionately less of the
planets face. Remember when looking at Jupiter that each belt or
zone is roughly 13 times the width of our planet. Thus, the air is
still moving a considerable distance to the north or south before
it is deflected eastwards or westwards. If Jupiters rotation slowed
to that of Earth (1700 km per hour at the equator) it would take
more than 10 Earth days to rotate on its axis whereas now Jupiter
accomplishes this feat in less than 10 h (9.6). Saturn is only mar-
ginally slower.
One must also remember that the cloud decks form very
inconsequential layers within the atmosphere of Jupiter and
The Gas Giants 285
Saturn. Jupiters atmosphere is more than ten times deeper than

ours. The cloud decks form a visible shell only one hundredth that
depth. Most of Jupiters atmosphere is the upper stratospheric and
thermospheric layers (Figs. 7.1 and 7.2). At present very little is
known about these other than their bulk composition (mostly
hydrogen) and their temperature.
One thing that sets all the giants apart from their solid,
diminutive cousins is the intimate connection between the inte-
rior and the atmosphere. Jupiter releases a little more than twice
the energy it receives from the Sun and this energy is the prin-
ciple driver of atmospheric movement. Heat can convect directly
from the interior to the top of the cloud deck. Diagrams in most
books imply a series of distinct layers, but in actuality the liquid
interior gradually morphs from one state to another. The atmo-
sphere is merely the outer veneer where chemicals from within
the interior broth can condense against the frigid chill of outer
space. Chemicals, such as water vapor will break apart and reform
within this stew, depending on the temperature and pressure, but
the chemicals are still present, simply mixed within the giant
planets depths. The super-rotating atmosphere can be considered
the surface manifestation of the circulation within the planet as a
whole, even if they are not moving in exactly the same way. The
weather on Jupiter and Saturn is more a product of heat coming
from within than heat coming from without. The Sun still plays a
role in their weather, but it is likely subordinate to the release of
internal heat from the planetary interior.
Great Red Spot and Other Vortices

Perhaps the greatest meteorological icon in the Solar system is
Jupiters Great Red Spot (GRS). This anticyclonic storm measures
12,000 by 14,000 km across, first spied by Galileo, has apparently
been raging for over four centuries. Indeed, despite the rather
fractious environment in which the storm rotates, mathematical
models suggest that this storm should be stableparticularly if it
is sustained by devouring smaller storms that wander too close.
Perhaps because the GRS inhabits the highly turbulent environment

of Jupiters stormy equatorial belt, the storm now appears be wan-
ing. Just as with the common misunderstanding of the storms
likely nature are misunderstood, the apparent rumors of the spots
inevitable demise may be greatly exaggerated.
Before we consider the likely fate of the spot, lets look at how
the spot is structured and how it obtains its red color. In 2010 the
GRS was imaged in the far infrared portion of the spectrum. This
more than any other set of observations gave a clue to the storms
structure and perhaps its origin.
Observations of the GRS and flanking white ovals suggest
that these are always cold cored and high in the atmosphere: up
to 8 km higher than the surrounding cloud decks. The GRS and
white ovals are a few degrees colder than their surroundings sug-
gesting that they are produced by uplift of air. Within this column,
like the lower zones, air cools until ammonia condenses and it is
this that produces the white coloration. However, the GRS and
later the GRS Junior (below), all develop pink hearts. Given what
we now know about the chemistry of this pigment, these can only
form where air is subsiding and bringing ammonia into contact
with acetylene from higher in the atmosphere. Around the flanks
of the GRS air is warmer than in the core, again implying subsid-
ence. As air descends it compresses and warms and in both terres-
trial and Jovian systems this causes the evaporation of clouds. Far
below the level of the visible red eye, lies an area of warmth, 34
hotter than the eye wall (Fig. 7.7). Whether this is warmed by con-
vergence or is an underlying plume of warmer material is unclear.
Measurements of air flow imply that air is not spinning within
the eye of the GRS. Instead it is roughly stagnant, much like the
air in the Venusian polar anticyclones in Chap. 5. Vertical motion
could be present but has been hard to determine directly. This led
researcher Leigh Fletcher to a rather complex model (Fig. 7.7). In
2000 the merger of three small ovals produced a larger oval which
gradually reddened and warmed. These eventually formed The Red
Spot Junior (or more prosaically Oval AB). These observations
imply that the white ovals are high in the atmosphere near the
tropopause then, when they have generated sufficient vorticity, or
spin, they extend downwards into the deeper cloud decks and tap
warmer layers. The introduction of acetylene to the ammonia-rich
The Gas Giants 287
c Warm Warm
a b Annulus Annulus
Tropopause
HIGH LOW
Cold Core
HIGH HIGH
HIGH
d GRS
A
Hot Spot
Hot
Core
FIG. 7.7 Not all anticyclones are alike. Terrestrial anticyclones are
regions of calm, descending and warming air. Anticyclone means a
region where the air rotates around the core in the opposite (clockwise)
direction to that seen in cyclones. Terrestrial anticyclones can have a
warm core (a) with air circulating in a clockwise direction, throughout,
or have a cold core with high pressure at the surface and low pressure
aloft (b). The Great Red Spot (GRS) appears to be even more complex
(c). In Leigh Fletchers model (right), pressure is highest at the level of
the red cloud deck where air converges from all sides. This air is then
sucked slowly upwards and outwards by the storms rotation. Here, it
cools and ammonia clouds condense (see A in photo, d). Thermal imag-
ing by ground-based infrared telescopes identified a warm ring (annulus)
of descending air where clouds evaporate and a much deeper hot spot
below the visible red spot. GRS Image Voyager 2/NASA
clouds then sets in motion the reddening of their color. However,

this may be misleading.
You can envisage a number of scenarios that fit all of the data
together. Storms begin as regions of upwelling from deep down
like terrestrial, mid-latitude cyclones (Chap. 1). Warm to begin
with, as they deepen they develop a cold core. Ammonia clouds
form above and within these domes due to the drop in pressure
and temperature. If the storms can draw additional energy from
flanking jet streams they continue to intensify.
Consider terrestrial hurricanes, although the analogy is not
precise. When the velocity exceeds a certain limit it induces down-
welling in its center. In a terrestrial hurricane this is in the center
of the eye, which soon transforms from a cold to a warm core

through the transfer of heat from the surrounding cumulonimbus
towers. In the case of the GRS and Oval AB, the area of maximum
down-welling is within the eye wall, while the center of the eye
has broadly stagnant air. This may mean that the GRS and hurri-
canes are fundamentally different or it may simply be a reflection
of scale. Remember, each storm is roughly the size of the Earth. If
eye-wall winds exceed a critical threshold the storm will open out-
wards through the effects of inertia (or what we mistakenly call
centrifugal force). Winds of several tens to hundreds of meters per
second will exert considerable force on their surroundings. The
critical event in the lives of these storms appears to be the opening
of the eye. When the threshold is reached air begins to descend
from the upper troposphere to the top of the underlying cloud
deck. Air is descending along the eye-wall and passively filling
the space vacated by the parting cloud wall. Within the eye of the
GRS a counter flow becomes established that brings air upwards
in the center of the eye, maintaining the high dome under the
tropopause that was observed by Voyager 2, Leigh Fletcher and
others (Fig. 7.7).
Further down, beneath the cloud deck, the presence of the
warm core implies either a region of hot upwelling or possibly,
as Leigh Fletcher, conjectures, another counter flow bringing air
downwards where it warms through convergence, much like a ter-
restrial anticyclone.
The GRS is maintained through its ability to drain energy
from surrounding jets, but also by absorbing smaller storms that
add spin to its bulk. Red Spot Junior has grown through the same
process: merging of smaller, cold-core white ovals increases the
vorticity, or spin, to the point that they can generate their own
internal circulation. Once it is established the resulting storm is
a rather passive feature within the atmosphere. Air barely moves
within them and then only in response to the circulation of air
around them.
The only question that then remains is what kick starts the
white ovals in the first place. At present the only clue is the warm,
lower core. There are two possible sources of this warmth. In the
first the closest terrestrial analogy is a warm, blocking anticyclone
(Fig. 7.7a). In these terrestrial systems, a pool of warm air meanders
The Gas Giants 289
with a kink in the polar front jet stream (Chap. 1). This has the
effect of enhancing the kink further until the Rossby Wave (Chap. 1)
expands to breaking point leaving a pool of warmer air cut-off on
the cold, Polar side of the jet. This cut-off feature then acts to fur-
ther block the passage of the jet stream, which then has to flow
around it. Such blocks are the most persistent features of terres-
trial mid-latitudes. Depending on their initial warmth, they can
last for weeks or even months in what is otherwise a very chaotic
flow. Such terrestrial blocks can be maintained by the continued
reintroduction of more warm air and only fade when their supply
wanes causing their internal temperature to fall.
Alternatively, the warmth is associated with a warm-cored
thunderstorm complex bringing heat from deeper down. This
forms a large storm with a cold cap. This is the white oval stage.
These release latent heat as various gases like ammonium hydro-
gen sulfide, water and ammonia condense, remaining warm-cored
deep down, but cold on top. This warm core then blocks or re-
directs the flanking jets. When the storm is sufficiently deep the
core opens up revealing its pink heart. Unlike terrestrial low
pressure areas, however, these tap their energy from the planets
interior, rather than the Sun. With a warm lower core they can
continue as stable, features spinning between the planets jets.
The upper, visible and very cold feature is a passive veneer over
an otherwise warm core. This would explain the updraft of cold,
dense air within the storm, in apparent defiance of basic phys-
ics that would demand cold, dense air sinks. Deciding which, if
any, of these hypotheses are correct will await more data that will
hopefully come with the arrival of Juno in 2016.
Remarkable though the GRS is, it is certainly no longer as
stable a feature as it was once thought. The GRS appears to have
been as much as 40,800 km across in 1800, if contemporary paint-
ings are accurate. A grainy black and white photograph taken in
1879 appears to vindicate this, but since at least this time the spot
has been on the decline. In 1979 when Voyager 1 passed the spot
was 23,200 km across, and today it is approximately 16,400 km in
diameter along its longest margin. It would seem that there is a
distinct downward trend that has accelerated to over 928 km per
year. The cause of the shrinkage is unknown but may relate to
the absorption of smaller storms into the GRS. Rather than adding
momentum they may be draining momentum from it, so the storm

shrinks. Is the storm then going to vanish for good? Before we write
the obituary for the spot we should remember that the spot has
disappeared before. During the 1680s the GRS vanished but reap-
peared nearly 20 years later in 1708. The spot may be taking a well
deserved rest before emerging reinvigorated in a few decades time.
One final observation: while the Red Spot continues to shrink
and become more circular, its color is morphing from pink to
orange. Indeed, at the time of writing, it would be more accurate
to refer to the GRS as the GOS.
Derechos and the Storm That Ate Itself

Jupiter is pock-marked with multiple storms, but if we cast the
Great Red Spot aside, nothing much else springs to mind in terms
of intensity or shear grandeur. If instead we turn our attention to
Saturn we see storms on a vast scale, both temporary and long-
lasting, that put the Great Red Spot to shame. These storms bear
a striking resemblance to features we see on Earth only on a far
grander scale.
Two Saturnian storms in particular were impressive: the long-
lasting polar vortices, one shaped like a hexagon; and the northern
spring storm of 2011. The polar hexagon appears to be a long-lasting
event with its first detection in August 1981 by Voyager 2. There
are some lovely ideas regarding the origin of this storm, but before
we look at the storm in more detail its worth looking back again
at the structure of the Saturnian atmosphere (Fig. 7.8). After all, the
storm owes its development to the interaction of two layers and
the transfer of energy between them.
Like on Jupiter clouds of water lie deep down, just above the
liquid hydrogen ocean, at around the 10 bar level. Ammonium
hydrogen sulfide lies at 36 bar around 100 km higher, while
ammonia clouds lie at 0.51 bar, just below the tropopause. It is
thought that the large Saturnian storms originate in the water
layer and presumably tap the internal heat of the liquid interior,
just as they do at Jupiter. It is clear that these storms only appear in
response to the slow change in seasons at Saturn. Aside from some
deep oval storms Saturn was relatively storm free when Voyager
The Gas Giants 291
200
0.001
Stratosphere
100
Altude relave to terrestrial sea level

0.01
Tropopause
0
0.1 Haze
Pressure (bar)
(km)
Ammonia
-100
1
Ammonium
Hydrogen Sulphide -200
10 Water
-300
Troposphere
-150 -50 0 50 100

Temperature (oC)
FIG. 7.8 The structure of Saturns atmosphere. This is very similar to

that of Jupiter, except that it is colder at ay particular altitude relative to
Jupiters. The lower gravity also allows for the clouds to be more spread
out. The same bands of cloud are seen here as at Jupiter, although the
lower temperature and pressure means that they occupy a greater spread
of height than above Jupiters liquid hydrogen sea. 1 bar is sea level pres-
sure on Earth and occurs just under the Saturnian tropopause (pale blue
line). Storms likely originate in the deeper layers then punch through and
mix with the upper visible ammonia layer
passed. Since the arrival of Cassini, many more larger storms have
been spotted and a deeper connection between the atmosphere and
the interior has been uncovered.
The 20112012 storm was particularly illuminating. This
storm appears to have begun as a cluster of small dark storms as
early as late 2010, but these initially remained stuck within the
deeper water and ammonium hydrogen sulfide layers. In 2015
Cheng Li and Andrew Ingersoll (Caltech) published a model that
could reconcile the slow, early development of this storm, as well
as the apparent periodicity of storms on Saturn. For observations,
tend to suggest Saturn undergoes waves of storm development

with a period of 2030 years. Storms seem to occur during the
northern summer; and this was has been the case since 1876 with
storms seen in 1903, 1933, 1960 and 1990. The 20102011 storm
followed the same trajectory as one observed in 1903, but was sub-
stantially larger. This time the storm came nearly 9 years ahead
of schedule and appeared in the northern spring, rather than the
summer. Why this storm chose to arrive early is unclear.
Each of these Earth-sized white storms is situated between
the equator and the mid-latitudes in the northern hemisphere.
Smaller oval storms are found at all latitudes except in the imme-
diate neighborhood of the poles. Li and Ingersoll propose that
Saturn spawns many small storms every year. The problem for
these little storms is that they cannot penetrate the full depth of
the atmosphere. It is suggested that water vapor at the 10 bar level
is the problem. This gas bubbles, as heat is conducted or convected
from deeper layers. The high density of the gas then prevents these
plumes from penetrating the less dense upper ammonia cloud deck.
Li and Ingersoll suggest that without the effective transfer of heat
from the water layers far below, the upper ammonia deck becomes
so cold and dense that it effectively subducts into the deeper warm
layer. As it descends, it shoves the warm air up from below and
the formerly deep storm erupts through the ammonia cloud deck.
Li further suggests that one reason that such large storms arent
seen on Jupiter is because the planet is intrinsically dry. There is
insufficient water to form a layer of air that blocks convection. Air
simply rises through the entire depth of the atmosphere on Jupiter,
distributing energy and preventing it from building up to the point
which it generates large, periodic storms. This would jibe with the
observations made by Galileoand in the wake of the Shoemaker-
Levy 9 impactsboth of which implied Jupiter has a surprising
lack of water vapor. The arrival of Juno at the Jupiter system in
2016 should help settle that one.
One problem with this model is that it doesnt explain the
apparent link to the northern spring and summer when all these
storms erupt. I suggest a possible solution in Fig. 7.9. Here, heating
of an upper layer that separates the upper cold layer from the deeper
warm layer generates convection within it. This only occurs when
sunlight heats the layer appropriately in the northern spring and
The Gas Giants 293
A
Sunlight heats layer
Stable, Stratified Layer

Ammonia Cloud Layer
a c
b
Deep, hot, moist layer supplies heat
B C
FIG. 7.9 Initiation of the 2011 storm. This possible model has this sea-
sonal storm beginning with deep, moist convection that is trapped below
a temperature inversion (A, left). A combination of heating of the layer
from above (center) and intense convection below eventually breaks the
inversion and allows ammonia-rich convection to rise higher, generating
the visible storm. On Earth altocumulus castellanus clouds mark con-
vection within a high layer (B). When surface convection breaks through
the intervening (C) inversion rapid ascent and thunderstorms can ensue.
Photos by author
summer. At this stage convection in this layer can link up with

deeper convection or destabilize the upper cold layer. Disturbance
of the upper layer then initiates subduction of the upper cold air
into those below. On Earth analogous situations arise frequently
in the spring and summer in the mid-latitudes. Cold air often finds
itself trapped above an increasingly hot lower layer. If the middle-
layers of the troposphere are heated sufficiently, they spawn further
mid-level convection characterized by the appearance of altocu-
mulus castellanus clouds or so-called high base storms (Fig. 7.9).
When sufficiently hot and unstable, convection in the lower layer
can penetrate any temperature inversion that is present and fuse
with the middle layer. On its own this can cause thunderstorms,
but throw an upper layer of cold air into the mix and the situa-
tion can become explosive. This is a frequent occurrence over the
Mid-West of the US in the spring, where the introduction of an
advancing cold front, steered by an upper level jet, sets in motion
the formation of supercell storms that often spawn tornadoes.
The 55 day-long 2011 storm was unique in that it spread right
around the globe, eventually consuming its own tale. Winds on
Saturn are predominantly westerly, the 2011 storm erupted north
of the equator where there is an easterly jet (Fig. 7.6). Like the ter-
restrial supercell storms the jet helped steer the storm to the west
this jet directed the 2011 storm to the west. On its own this could
only have provided enough oomph if it connected more deeply
with the storms base. It is here that we can turn our attention
to long-ranging thunderstorms on Earth. These are the infamous
derechos. These unique storms provide a clue to the longevity and
underlying power of the Saturnian global storm.
Derechos are a world apart from humdrum thunderstorms.
These fascinating phenomena are effectively unknown in the
United Kingdom but are fairly common in the United States
and eastern Europe and Russia. Similar phenomena also occur in
Australia, Argentina and South Africa. Derecho roughly trans-
lates as horizontal wind in Spanish and describes a violent, gale
force wind generated from vicious and fast moving thunderstorms.
Derechos require a lot of energy in the atmosphere, both to form
and then to propagateadvance at speed away from the originat-
ing disturbance.
The basic set up for these monster storms is the formation
of a violent thunderstorm in very unstable, hot, energy-rich air.
The initial thunderstorm can be really rather small and isolated,
but given the right forward push, perhaps because the storm has
become entwined in an upper level jet-stream, and the storm will
become self-sustaining.
Most thunderstorms have a life measured in minutes, with
individual cells lasting less than 35 min on the whole. Storms die
when rain chills the air within the storm and generates strong
downdrafts. Normally, these fan outwards in all directions and
choke off the supply of warm air needed to keep the storm alive.
Derechos circumnavigate this problem by moving quickly forward,
The Gas Giants 295
through the warm air, so that the axis of the storm is tilted back-
wards. As their rain-cooled wall of air descends it drives upwards
an accompanying wall of warm, moist air at its leading edge and
since most of this downdraft is behind the leading edge of the
updraft the storm can continue to advance without cutting off its
supply of warm air. This pattern also allows the rapid formation
of further storm cells along its leading edge (Figs. 7.10 and 7.11).
In turn these replace the trailing, rain-cooled, storm cells at the
rear and a wave is launched through the warm, humid air mass.
In front of the wave, warm air is blasted upwards by the advanc-
ing wall of cold, rain cooled air at the rear. As long as there is a
100 km
Anvil Over-
Shoot Region Anvil Region
Storm
direction
rain-cooled and stabilised lower layer
FIG. 7.10 The planet-wide springtime storm on Saturn in 20112012. The

storm emerged as a cluster of small dark storms before morphing into
an enormous, white monstrosity which encircled the planet. The storm
bears a striking resemblance to advancing squall lines, such as Malaysia
and Indonesias Sumatras (radar image lower left of figure): These arise
as smaller clusters of storms over the north coast of Sumatra during the
summer monsoon, before fusing into advancing lines of storms. Images:
SaturnNASA/JPL/Cassini; Sumatra satellite images (left) http://www.
nea.gov.sg/training-knowledge-hub/weather-climate/sumatras
Direction of storm advance shown by pink arrow
Rain-
cooled
air
Rain-cooled
air in shadow
of storm
Cold, ammonia-rich air Direction of

Guiding subsides rapidly in the storms storms advance
upper level wake driving it forward
wind
Unstable air ahead
More stable, stratified air behind the wave of wave
FIG. 7.11 Derechos, Sumatras and the Saturn Springtime Storm are
all examples of advancing waves. These storm systems are initiated in
unstable air and if conditions are right will advance in the general direc-
tion of the air flow, swallowing unstable air along their leading edge,
uplifting, cooling and releasing its energy. The June 29th 2012 derecho,
in the US, began as a cluster of storms that rapidly organized into a wave.
This swept across nearly half of the northern US from the Great Lakes
to the east coast. It might have gone further if the cool Atlantic waters
hadnt been in the way (lower left radar image). Similarly, the Saturnian
storm began as a cluster of dark storms and could have gone on indefi-
nitely had it not run into its tail and run out of unstable air. Derecho
radar image courtesy NOAA
supply of warm unstable air in front of the waveand as long as

the storm cell can maintain its forward momentumthe Derecho
continues. In its wake the warm, unstable air has been replaced
by a deeper, colder and much more stable air mass, generated by
the advancing storm. The atmosphere then becomes more stable,
overall with warm air aloft, and cooler, drier air below.
A couple of Derechos are notable for their violence, lon-
gevity and for their degree of analysis and are worth describing,
here, for comparison. On July 5th, 2002, a derecho struck east-
ern Finland generating winds of hurricane strength, which in turn
produced over 400 storm-related wind damage events. The Finish
derecho was the most northerly event of its kind ever observed.
Like many derechos in the States, the storm originated when
a cold front stalled over the western portion of the country.
The Gas Giants 297
FIG. 7.12 The 2002 Finish derecho and 2012 US derecho. Both storm sys-
tems formed ahead of stalled cold-fronts: the Finish storm to the north
of the Bay of Finland and the US storm to the SE of Lake Michigan.
However, the Finish storm moved perpendicular to the prevailing wind,
driven by upper level flows of air. The US storm moved more conven-
tionally along the direction of the wind-field from NW to SE. The Fin-
ish storm damaged a track 450 km long, while the US system affected a
track over 1100 km in length. Both storms died out when the supply of
suitably unstable air was cut-off. The derecho tracks are shown as light
shaded areas. Underlying US and Finish maps courtesy of http://www.
proteckmachinery.com/stats.php?p=us-map-blank-vector and http://
www.vector-eps.com/finland-vector-map/ respectively
Here, things diverged from the usual scenario that afflicts the
States. On July 4th the unstable, warm and humid southerly flow
of air was moving out of Germany and Poland towards the Arctic
Circle. Over Lithuania a severe thunderstorm developed a classical
comma-shaped (bow-echo) appearance that is typical of derechos
(Fig. 7.12). This storm generated winds of over 100 miles per hour
before dying out over the Gulf of Finland.
The following day this feature had moved north before merg-
ing with the systems warm front over northern Lapland. In its
wake a narrow ridge of higher pressure poking into the warm
sector over southern Finland into which air was descending from
above. A belt of thunderstorms developed ahead of this region and
soon these became organized into a classical comma (or bow-echo)
shape and advanced quickly towards the northeast. The trajectory
followed that of the Lithuanian storm system taken the day before.
You can see, here, the link to the Saturnian storm. Descending air
kick starts uplift in the air along its forward edge. This generates
further storm cells along the forward edge of the downdraft.
Over the next several hours the Finnish system surged north-
wards, scooping up warm, unstable air in its path and replacing
it with drier air descending in its wake. The storm crossed the
length of Finland over the ensuing 12 h, leaving a trail of damage
450 km long.
The Finnish storm was severe in its own right, but the US
storm of 2012 was even more remarkable. Spawned along the
southern margin of another stalled cold front, this system devel-
oped during the tail end of the June 2012 heat-wave that brought
temperatures into the 40s Celsius across much of the mid-west.
Like the Finish system, this derecho began as a cluster of thunder-
storms. These formed to the south of the Great Lakes along a wave
in the otherwise stationary cold front. Given the vast amount of
heat energy that was available, the cluster of storms soon orga-
nized themselves into another bow-shaped structure that was
driven aloft by northwesterly winds.
What was extraordinary about this storm was its longevity.
Powered by the high temperatures, high humidity and the strong
upper level winds, the derecho crossed ten states from Indiana to
Washington D.C. and finally out into the North Atlantic, moving
at an average speed of 100 km per hour. Thirteen people lost their
livesprimarily killed by trees felled by winds up to 130 km per
hour (80 mph). One might imagine that such a violent storm would
have at least drained the heat from the baking North American
interior: there was no such relief, once the clouds cleared, the sun
continued to bake downthe causative frontal boundary waving
back northwards once more. The storm had crippled many power
lines, leaving four million people without power. This was a criti-
cal effect adding insult to injury: with no power, cooling systems
in buildings were rendered useless. Consequently, a further 34
people died as a result of heatstroke, or from its related effects.
Derechos are fairly well understood and relatively common in
the US, with 50 or more occurring every year. In most cases storms
begin on the northern side of an eastwest trending frontal bound-
aryalong its upper surface. As these develop they migrate south-
wards into the deeper, warm air and begin to draw energy from it.
In this location, the storms soon organize into the classical bow-
shaped structure that drives eastwards with the prevailing winds.
The Gas Giants 299
Although the mechanism underpinning their formation is

distinct, derechos share many common features with other line
squallsorganized systems of rapidly advancing thunderstorms.
Amongst these are the tropical Sumatras (Fig. 7.11). Like the
derechos of the mid-latitudes, these begin as a cluster of thunder-
storms over northern Sumatra (Indonesia). These storms form dur-
ing the northern summer monsoon as southwesterly winds draw
very humid warm air northeastwards towards the west coast of
Malaysia. The storms usually form in the late afternoon before
crossing over the Straits of Melaka during the first half of the night.
At night time the waters retain a lot of heat and readily give this up
to the overlying air, causing the storms to intensify. After a short
time the storm systems impact the west coast of Malaysia and
Singapore. Here, deprived of further moisture, the Sumatras decay.
Before they expire, the storms can generate winds up to 100 km
per hour, along with 12 h of torrential rain, thunder and lightning.
The link to the Saturnian springtime storm can only be
inferred (Figs. 7.9 and 7.10), but the appearance and propagation of
the storm to encompass the planet are very similar to the propaga-
tion of line squalls on Earth. Saturns springtime storm emerged
at 37 N, with Cassini detecting lightning ahead to the appearance
of the white cloud top. The storm tracked the trajectory of a sea-
sonal easterly jet, and propagated faster to the west than the jet
could have carried it. That these white storms are seasonal implies
the Sun helps in part determine their development. Measurements
show that these storms originate well below the level that the Suns
energy can reachat a depth where pressures are 1012 times that
on the surface of the Earth. Here water and ammonia clouds con-
dense and form a deep layer that rises and falls by convection.
Inevitably, one wonders how the Sun can control the forma-
tion of a storm that lies outside its reach. The answer might lie in
the way the Sun controls the top of Saturns atmosphere. As Saturn
moves round its orbit and enters springtime once per 29.5 years, the
top of the northern hemispheres atmosphere warms and the weak
westerly jet stream emerges. As on Earth this might mix colder
air down into Saturns deeper layers and break up any stable layer
lying on top of the deeper clouds. Once this process has begun,
storms that are normally trapped deeper down in the atmosphere
can begin punching upwards, much as thunderstorms do on the
Earth, once overlying stable layers have been breached. In the
US the development of supercell stormsand, in a slightly dif-

ferent manner, Derecho-bearing stormsaccompany the breach-
ing of stable layers above the embryonic storm cells. The storms
remain small and isolated until the upper stable layer is degraded.
At this point convection becomes explosive and a major storm cell
is initiated. On Earth the Sun provides all of the energy to carry
out this process, while on Saturn the Sun may merely breaks the
seal and allows Saturns hot interior to do the rest. Another link
between terrestrial line-storms, such as derechos and Sumatras,
is the alignment of the winds at different levels. On Earth where
the upper winds blow in the opposite direction to the movement
of the storm, any instability is sheared off and storms tend to die
away. Align the two movements and storms can propagate aggres-
sively. The upper wind helps get them going, much like giving a
car a nudge on a steep hill, when the break is released.
What of Saturns other storms? The white ovals are essen-
tially the same as those on Jupiter with cold cores that project
a few kilometers above the surrounding belts and zones. Unlike
Jupiter, but continuing a trend that extends out to Neptune, the
belt of equatorial westerlies is broader than it is in Jupiter. In gen-
eral, the wind flow pattern is much simpler with a broad westerly
belt overlying the equator and tropics. Only north of 30 N and
30 S do the westerlies give way to easterlies. Polewards of 50
or so from the equator narrower belts of easterlies and westerlies
are found (Fig. 7.6). Within the southern hemisphere, one rather
interesting and lightning-generating storm has been spotted. This
Dragon Storm seems to be a semi-permanent feature, much
like Jupiters Great Red Spot. Although the storm appeared in July
2004 and vanished a few weeks later, the location seems to match
a long-term region of storm generation and provides a further clue
to the working of Saturn and Jupiters atmospheres.
The Dragon Storm was illuminating for three reasons. Firstly,
the storm generated bursts of radio waves in precisely the same
fashion as terrestrial thunderstorms. Although lightning was not
directly seen, the storm is believed to have generated a consider-
able battery of electrical discharges. Secondly, its location where
multiple other convective storms have formed suggests that there
is an underlying plume that regularly generates visible atmo-
spheric storms. In this regard, it should originate with rising warm
The Gas Giants 301
gas from the deeper interior. Infrared measurements of the GRS

also indicate it has a deep warm core. Thus, both the GRS and the
Dragon appear to have a warm heart. Within the Earths atmo-
sphere the largest storms have warm coresas do the strongest
anticyclones (above). The key is the pressure warm air creates. By
pushing on its surroundings it creates pressure that can divert air-
flow around them. Both the GRS and perhaps the Dragon Storm
appear to have this pushiness.
Finally, the Dragon Storm presented atmospheric scientists
with a view of how such storms pump energy into the surrounding
atmosphere. Unlike the GRS of Jupiter, in March 2004 a precursor
to the Dragon Storm was seen shedding smaller, dark stormlets.
Some of these peeled away from the main storm area and migrated
with the surrounding jet streams until they were eventually con-
sumed. Others merged to produce larger storms that eventually
morphed into larger, white ovals. This suggests that at least at
Saturn, warm-cored deep storms eventually drive the formation of
much higher, cold features. Again, this implies that perhaps fea-
tures like the GRS are driven first by the generation of a warm core
deep down, which then diverts the jet stream.
Aside from illuminating how storms develop and contribute
to the overall running of Saturns atmosphere, the Dragon Storm
presented an odd puzzle. Radio signals, associated with lightning,
were only detected when the storm was below the nightday ter-
minator. As soon as the storm rose over the horizon Cassini saw
the radio signals dropped off. This was repeated over many rota-
tions of Saturn over several weeks. Some researchers concluded
that the source of the lightning was related to the visible storm
but that they were separated by such a great distance that the light-
ning source was in view while the cloud was not. Perhaps instead
the emissions are beamed, much like a laser, in the direction of
Cassini (this would be coincidental); or perhaps illumination
changes the structure of the cloud so that it no longer produces
sufficient charge separation.
One of the intriguing general features of the gas and ice giant
atmospheres is the steady increase in mean wind speed with
increasing distance from the Sun. This is also reflected in the
increasing simplicity of the airflow. One suggestion is that nearer
to the Sun solar radiation helps generate more turbulence and such
turbulent motion robs the atmosphere as a whole of its energy.

This leaves less to power the jet streams. Perhaps this is weakly
analogous to the situation on Earth. When the winds blow most
directly from west to east, they also blow strongest. When the
winds become, as meteorologists term, more meridional and less
zonal, with a stronger north-south component, the general speed
decreases. At Uranus and Saturn (Chap. 8) there are effectively five
belts over the entire globe. A large easterly belt overlies the equa-
tor, with two broad westerly belts extending to each pole and a
relatively quiet polar region with slower westerly winds. As more
energy is available to blow the zonal winds, Saturn has stronger
winds (around 500 m per second, or 1800 km per hour), and these
only accelerate as we move further out from the Sun.
The Polar Hexagons

Discovered by Voyager 1 in 1980 was a large hexagonal cloud fea-
ture that appeared to rotate with Saturn. Superficially, the struc-
ture appears utterly shocking: this is an apparently perfectly,
hexagonal feature, thousands of kilometers side-on-side that sits
above Saturns North Pole. Infrared observations show that the
hexagon is rather deep; extending to several bars of pressure
essentially to the base of the atmosphere. From the perspective
of Cassini or Hubble each edge seems perfectly straight with per-
fect 120 angles. However, scale and distance are important; lets
look at this storm and reconsider its geometry. A side length of
13,800 km is wider than the diameter of the Earth. Each corner is
roughly 500 km across, meaning what appears to be a sharp turn is
actually rather smooth.
Even without the sharp turns, how then do we recreate a hex-
agonal feature? Recall Chap. 1s atmospheric features called Rossby
Waves. Due to the Coriolis Effect winds rotate counterclockwise
around the poles (west to east). Because air is trying to move north
and south to even out pressure, the atmosphere organizes itself
into a series of waves. Now, on Earth surface features, such as the
Rocky Mountains, tend to cause buckles and kinks that alter the
flow, but on Saturn there is no surface to impede movement. There
is still a northsouth pressure difference because the atmosphere
The Gas Giants 303
is warmer across the more intensely illuminated latitudes, and

because the spinning motion of the planet influences convection
from the deep interior. As Fig. 7.6 indicates, at 78 N where the
hexagon is found there is a sharp change in wind velocity with
latitude from 100 m per second east to near stationary or slightly
to the west. This abrupt change is all that it takes to generate the
waves. Indeed, it is the absence of such an abrupt change in wind
velocity near the South Pole (and in the atmosphere of Jupiter) that
prevents a similar hexagon from forming here.
Put this together and the result is Rossby waves within
Saturns atmosphere. In the atmosphere of a planet with a sizable
spin, like Saturn, the atmosphere will naturally set up jet streams
that rotate around the poles in a series of these Rossby waves.
Within the core of this six-wave pattern is the polar vortex that is
largely cut-off from the rest of the atmosphere.
The Rossby Waves model has been successfully recreated
in the laboratory. Ana Aguiar and colleagues (Oxford University)
continued their research using a large tank of fluid, in which they
could alter the velocity at different distances from the tanks center.
The research clearly showed that where there was a suitably steep
gradient in the velocity of the winds, a polygonal pattern rapidly
becomes established. In experiments shapes with up to 8 sides were
appeared, with six sided polygons the most common. In experi-
ments where tanks of fluid were used, rotating the middle of the
tank faster than the fluid on the outside resulted in the formation of
turbulence in a narrow zone between the periphery and the center.
The polygonal shape arises when the turbulence begins to organize
into a fixed number of smaller vortices that form the motion of the
fluid on the slower outside of the tank. As the vortices mature they
naturally space themselves out and then organize the rest of the
flow around them. Ultimately, each vortex forms a corner of the
polygon. If this sounds abstract, remember that on the Earth the jet
stream will naturally generate identical waves with sides thousands
of kilometers long (Chap. 1). Abrupt changes in direction arent
visible merely because Earth is a smaller world where the bends in
the jet take up a proportionately larger length of the overall jet.
What of the southern vortex? Cassini images are telling. Here,
the circumpolar winds organize white oval storms into a broad
belt around what is clearly an eye-wall. Earlier we used this term
for the wall of clouds around Jupiters GRS, but strictly speaking
this is more of a gap in the main cloud decks. Saturns southern
vortex has, instead, a clearly defined eye wall: a region of clouds
that rises above the main deck (Fig. 7.10). Within this wall the
level of the cloud deck bottoms out. Inside this lower shield of
cloud rests a smaller polygonal feature analogous at least in terms
of shape to the northern hexagon. Quite why the two hemispheres
show distinct patterns is unknown but conceivably might be due
to the season. It could be that the southern storm and northern
storm alternate structures depending on the relative amount of
solar heating. That wont be clear until Saturn has completed one
of its 29.5 Earth-year long orbits.
Speaking of heating, oddly enough if one was planning a
balmy vacation in Saturns cloudy atmosphere, the southern polar
region would be the place to be as it is a good 63 C warmer than
anywhere else on the planet. While the cloud tops drop out at
185 C on average, the South Pole is almost as warm as a Martian
night at 122 C. Certainly not tropical by any standards, it is still
significantly warmer than one would expect for distant Saturn.
Quite why the South Pole is so much warmer than the rest of the
planet is mysterious. Solar heating is unlikely, as this temperature
is far above those reached at Jupiter, and right now, the south pole
remains warm, while the Sun is heating the northern hemisphere
most strongly. Instead this extra heat is almost certainly coming
from Saturns interior. If so the difference in each vortex might
reflect differences in the delivery of heat to the atmosphere. Saturn
might always have a warm south pole regardless of the season.
Ice Fall from the Rings

While the South Pole is the place to be for an artificial summer
on Saturn, the place to go for snow lies at the equator if you are
prepared to go high enough up. Saturns rings orbit the planet at a
distance of 7000 km from the cloud tops. Although they seem per-
manent, various processes, from collisions to the action of Saturns
magnetic field, are constantly whittling away at them. While shep-
herd moons do their best to keep the rings in shape, material will
naturally spread outwards and inwards from the main body of the
rings. The icy material that falls inwards is captured by Saturn.
The Gas Giants 305
These icy debris drizzle down into Saturns upper atmosphere.

Although this persistent snowy fall is miniscule and evaporates
long before it reaches Saturns cloud-tops, it does contribute water
vapor to the upper atmosphere of Saturn.
If you want proper snow on Saturn then look just beneath the
tropopause. Here, ammonia ice crystals barrel around the planet in
the most severe blizzard you can imagine. 100 km lower and this
blizzard morphs into an ammonia rain, as temperatures climb.
Jupiter has much the same, but here the snow has considerably
less far to fall before it encounters the ammonium hydrogen sul-
fide layer and chemically merges into it.
Ios Electrifying Connection to Jupiter

The giant planets all generate strong magnetic fields that actively
deflect the solar wind (Fig. 7.13). Jupiters field is roughly 14 times
stronger than that of the Earth, while Saturns is just approximately
half the strength of our planets. The Jovian fieldits magneto-
spherecompletely encases the Galilean satellites, with Io, Europa
and Ganymede buried deep within it. Callisto, lies nearly two mil-
lion kilometers from Jupiter and is effectively separate from the
remaining, inner three moons. As each lies within Jupiters mag-
netic field they are subject to a constant pummeling from charged
particles that are trapped within the field. In turn each moon sheds
copious charged particles into the magnetosphere. Most of these
remain trapped along the orbit of the satellite, but as the moons
sweep through the cloud of ions, some are driven out of the ring
and along the magnetic field line towards Jupiter (Fig. 7.14).
Io adds another dimension to this relationship. Io is also tor-
tured by strong gravitational forces that keep its interior molten.
Consequently, Io is racked by violent volcanic activity. Blasting
hundreds of kilometers above the satellites surface, these eruptions
contribute a wealth of sulfur, oxygen, sodium and chlorine ions at
a rate of approximately 1000 kg (1 metric ton) per second. This is
approximately the same rate of loss as the Earth, Venus and Mars
lose from their atmospheres. Like those oxygen and hydrogen ions
shed from the outer Galilean satellites, this broth of charged par-
ticles is swept in a pair of arcing columns of ions along Jupiters
magnetic field lines (Fig. 7.15).
FIG. 7.13 The Polar vortex storms of Saturn. The North Polar storm (left)
is clearly hexagonal with edges over 13,800 km long and rotates with the
same period as Saturn as a whole (10 h 39 min). Within the hexagon there
is a more circular elevated storm region, separated from the walls of the
hexagon by a lower moat. Over the south pole there is another storm
with broadly circular eye walls. However, superficial inspection show
that within this structure is an inner vortex which is roughly hexagonal
with roughly straight walls (right). Images by Cassini (ESA/JPL/NASA)
Planet Earth Jupiter Saturn Uranus Neptune

Tilt of Rotaon axis 23.5o 3o 27o 98o 30o
Tilt of Magnec 12o -10o 0o -59o -47o
Field
Oset of Field Axis 8% 10% 5% 31% 55%
Field at Equator 31,000nT 428,000nT 22,000nT 23,000nT 13,000nT
Magnetosphere 10 Earth Radii 65 Jupiter Radii 20 Saturn Radii 18 Uranus Radii 25 Neptune Radii
FIG. 7.14 A comparison of the magnetic fields of the giant outer plan-
ets. Aside from Jupiters, which dwarfs all the others, most have field
strengths comparable with the Earth. Jupiter and Saturn generate theirs
in their core, while Uranus and Neptune generate theirs further out.
Image credits: Jupiter (HST/NASA); Saturn Cassini/NASA; Uranus:
HST/NASA; Neptune Voyager 2/NASA
The Gas Giants 307
Dayside Main Oval Io Spot

Aurora
Flux
Tube
Plasma
Torus
Europa Spot Ganymede Spot
FIG. 7.15 The complex interaction of Jupiters large Galilean moons with
its atmosphere and magnetosphere. Jupiters powerful magnetic field
sweeps ions across the surface of its moons and also traps ions released
by these satellites. Such ions are directed into two flows, one following
the orbit of the satellite through the planets magnetosphere, and the
other flowing along the planets magnetic field lines to the cloud tops.
Within this second flow, known as a flux tube. A giant electrical circuit
links the moons to the planet. Io has a particularly aggressive interac-
tion, as its volcanoes release as much mass each second as the Earth loses
to the solar wind. Ios flux tube thus carries a whopping 5 million Amp
current. With 5 trillion Joules of energy transferred each second. As the
flux tubes from Io, Europa and Ganymede reach Jupiter they generate
intense foci of aurora (inset photo, NASA)
As these particles move along the lines they accelerate until

they smash into the thermosphere of Jupiter. Here, ionization
causes permanent aurora focused into spots and arcs (Fig. 7.15).
Within Ios flux tube runs a 1 trillion ampere current that flows
back and forth along the flux tube as though it was a giant copper
wire. More than 2 trillion joules of energy is transferred every sec-
ond. In many ways the flux tube forms a far more visible version
of the circuit that links the terrestrial ionosphere (and the thun-
derstorms below) to the magnetic field. Were you able to hover
in the vicinity of where the flux tubes connect with Jupiter, as
well as impressive aurora, beneath this shimmering blue canvas
would flicker and flare hundreds of lightning bolts as Ios flux tube
probed down towards the metallic hydrogen mantle. The presence
Jupiter
Rotation Axis
Magnetic Field
Orientation
Diameter of Sun
to scale with
magnetic field
Saturn
Uranus
Neptune
FIG. 7.16 The scale of the magnetic fields around the giant planets. Jupi-
ters dwarfs the others and stretches 700 million kilometers to Saturn.
Those of Uranus and Neptune are steeply inclined to the oncoming
solar wind and whip the field around in different directions as the planet
rotates. Planetary field lines are in yellow while the solar field lines are
in red. The spin axes and magnetic field axes are indicated above along
with a comparison to the diameter of the Sun
of Jupiters Galilean satellites adds far more conducting material

than is available from the exosphere and solar wind; while the
rapidly rotating powerful magnetic field acts as a super-charged
dynamo driving the flow of charge into and out of Jupiters upper
atmosphere. Out at Saturn, Titan also sheds particles into the mag-
netosphere of Saturn, though this is not enough to power similar,
punctuated auroral displays as are seen at Jupiter.
The flux tube connection points are embedded in more gen-
eral auroral emission (Fig. 7.14). The day lit side of the polar region
glows permanently with a constant barrage of charged particles
arriving from the Sun that are grabbed by Jupiters powerful mag-
netic field and like the Earth, some aurora are focused into a large
ring centered around the magnetic pole. Here, the interaction of
the magnetic field lines with the atmosphere generates another
circum-polar current in Jupiters ionosphere.
The Gas Giants 309
Filled with ions from its satellites and from the solar wind,
Jupiters magnetosphere is a powerful source of radio emission
in the 0.630 MHz range. However, Jupiter also generates maser
emission as so-called Alfvn waves carry pockets of ionized gas
into Jupiters Polar Regions. As this radio emission is beamed it
is not always detectable from the Earth. When it is, it emphasizes
Jupiters regal status by exceeding radio emission from the Sun.
One must remember that with the exception of sunspots, Jupiter
generates the strongest magnetic field of any object in the Solar
System. Jupiter really is the King.
Is Mighty Jupiter Losing Any Atmosphere?

Previous chapters showed that Venus, Earth and Mars are shed-
ding their atmospheres to space at a rate of about one metric ton
per second. This is mostly through mechanisms involving ion-
ization of their gases and the subsequent loss of them to their
magnetic fields or the solar wind. Mars also loses gases directly
as their kinetic energy can overcome the gravitational attraction
of the planet. By contrast, Jupiters strong gravitational field and
its relatively cold upper atmosphere guarantees that most of its
atmosphere will remain bound for hundreds of trillions of years.
Some gas is shed through its magnetic field. Like Venus, particles
become charged high in its upper atmosphere and here, where the
gravitational pull is weakest, these particles can become embroiled
in the strong magnetic field. From here, they can be accelerated
into the magnetosphere and hence slowly out into space. A small
amount might find its way to Saturn, where the immense field of
Jupiter occasionally sweeps past. Jupiter and Saturn thus form a
frigid pairing, much like Venus and the Earth (Chap. 5).
However, the rate of loss is trivial. Even compared to Venus,
where trillions of years would be needed to bleed the atmosphere
dry, Jupiter and Saturn will remain impressive, if frigid, gas giants
until the distant dark future of the universe. The only moment of
risk for these impressive worlds will come during the few hundred
million years that the Sun is a red giant, when cloud top tempera-
tures will rise above 100 C at Jupiter and well above 0 C at Saturn.
This will be enough to drive enhanced mass loss, but still nothing
near enough to drain these vast worlds of much of their substance.
Conclusions
Jupiter and Saturn are the true giants of the solar system. Each
manifests weather on a scale dwarfing the small terrestrial planets
in their entirety, but when one looks closely at their super-charged
phenomena, one returns to familiar territory. Saturns mysterious
hexagon is a grossly expanded manifestation of our polar front jet
stream; the Great Red Spot shares some aspects of its structure
with warm cored high or low pressure areas; Saturns northern
spring time storm of 2011 behaved like a derecho (or tropical
Sumatra). Weather, it seems is the same the universe over, with
each planets atmosphere flowing to the same underlying laws.
This is not surprising in itself: what is amazing is how these laws
play out and how scale can make such an impressive difference at
least to our eyes.
This is a theme that will continue as we head outwards to the
ice giants, Uranus and Neptune, and to their icy satellites. Beyond
our cosmic shore we will encounter Rossby waves and circular
storms once more, forming a unified picture of weather across
the universe.
References
1. Snchez-Lavega, I. A., del Ro-Gaztelurrutia, T., Hueso, R., Gmez-Forrellad, J. M., Sanz-
Requena, J. F., Legarreta, J., et al. (2011). Deep winds beneath Saturns upper clouds from a
seasonal long-lived planetary-scale storm. Nature, 475, 7174.
2. Li, C., & Ingersoll, A. P. (2015). Moist convection in hydrogen atmospheres and the frequency
of Saturns giant storms. Nature Geoscience, 8, 398403.
3. Transient internally driven aurora at Jupiter discovered by Hisaki and the Hubble Space
Telescope. (2015). Geophysical Research Letters. 42(628), 16621668.
4. Barbosa Aguiar, A. C., Read, P. L., Wordsworth, R. D., Salter, T., & Hiro Yamazaki, Y. (2010).
A laboratory model of Saturns North Polar Hexagon. Icarus, 206(2), 755763.
5. Emily Lakdawalla. (2010). Saturns hexagon recreated in the laboratory. Planetary.org.
6. Godfrey, D. A. (1988). A hexagonal feature around Saturns North Pole. Icarus, 76(2), 335.
7. Snchez-Lavega, A., Lecacheux, J., Colas, F., & Laques, P. (1993). Ground-based observations
of Saturns north polar SPOT and hexagon. Science, 260(5106), 32932.
8. Sayanagi, K. M., Dyudina, U. A., Ewald, S. P., Fischer, G., Ingersoll, A. P., Kurth, W. S., et al.
(2013). Dynamics of Saturns great storm of 20102011 from Cassini ISS and RPWS. Icarus,
223(1), 460478.
8. The Ice Giants
Introduction
Superficially the ice giants, Uranus and Neptune, are simply
smaller cousins of the gas giants, Jupiter and Saturn, but it doesnt
take much examination to realize that this analogy is only skin
deep. Aside from their smaller size and much lower tropospheric
temperature, the interiors of these worlds are composed chiefly
of the ices ammonia and water. Hydrogen makes up nearly 83 %
by mass of their atmosphere, with helium second at 15 %, but
this is a relatively thin veneer over a thick sea of liquid domi-
nated by water and ammonia. This structural difference has fairly
profound consequences for the internal workings of each world,
including how its atmosphere and magnetosphere behave. Their
relatively low mass compared to Jupiter and Saturn means that
their upper atmospheres extend for thousands of kilometers. Most
surprisingly, that of Uranus extends out past its rings and affects
their behavior as well.
The outer two giants also play host to the solar systems
strongest winds. Yet, despite many similarities, there are perplex-
ing differences between them. Neptune, like Jupiter and Saturn,
releases more energy than it receives from the Sun, while Uranus
appears dead, releasing only what it receives. Quite why these two
worlds have evolved differently remains unclear. Uranus had also
appeared utterly bland when Voyager 2 swept past it in 1986, until
in recent years Uranus has had something of a meteorological
renaissance on showing an atmosphere that appears as dynamic
and interesting as our solar systems other pale blue dot, Neptune.
Yet many gaps remain in our understanding and will likely do so
for decades to come unless more missions are launched their way.
This chapter is a story of what is known and what remains to be
discovered.

DOI 10.1007/978-3-319-25679-5_8
Inside the Ice Giants

Both ice giants have a very similar structure. The visible atmo-
sphere extends fully 1000 km below the cloud tops. Until Yohai
Kaspi (Weizmann Institute) and co-workers completed studies of
the gravity fields of both planets this layer was poorly understood.
Using various observations of the wind-driven cloud patterns and
shape of the planet it was possible to work out the distribution of
mass and density for both worlds. In each case Kaspi concluded
that the weather on both worlds was confined to a layer above
2000 bars pressure on Uranus and 4000 bars pressure in the more
massive Neptune. In terms of depth, this meant that each planet
had winds confined to approximately 1100 km from the visible
surface. Those mighty jets, which are seen pushing the weather on
Uranus and Neptune around, were fairly shallow in nature encom-
passing less than 1 % of the planets mass. Kaspis team obtained
the same result as that of earlier modeling of the jet stream, which
also implied that weather was limited to the outer skin of these
large planets. Given the apparent differences in terms of the heat
that these worlds emit and their appearance, such a strong physi-
cal connection might seem surprising. The bottom line is that they
each are very similar in mass and density, and probably formed at
similar positions in the solar system. This is most likely roughly
where Saturn is found today.
A different set of models were produced by Jonathan Aurnou
and Moritz Heimpel. In these the circulation was made to extend
deeply towards the core. When this manipulation was carried out
the number of jet streams that was seen in the lower atmosphere
decreased from 20 to around 35. This is clearly at odds with
observations of Jupiter and Saturn, which both have around 1020
jet streams. It also implies that they have shallow circulations.
Indeed, most models of giant planet atmospheres suggest that the
lower boundary for circulation occurs at the depth where the gases
become cool enough for silicates and metal oxides to condense.
This occurs at temperatures of around 2300 C, which, in Uranus
and Neptune is broadly at the depths Kaspi concludedaround
1100 km below the visible surface.
The Ice Giants 313
Beneath the meteorological domain at a little over 1100 km

down inside both Uranus and Neptune, the atmosphere gives way
to a thick layer dominated by hydrogen and helium gas. This is
broadly analogous to the liquid hydrogen and helium layer seen in
Jupiter and Saturn, except that it is at a lower, overall density. As we
continue deeper at a depth of 5000 km the hydrogenhelium soup
gives way to what might loosely be called a deep, hot ocean. The
interior of both worlds is never dense enough to make hydrogen
become metallic. Instead this hot icy ocean forms the lowermost
layer above the planetary core. Like the liquid hydrogen layer inside
Jupiter and Saturn, this material gradually morphs from gas to liq-
uid state and then, possibly to a plastic-like semi-solid layer nearest
to the core. Dominated by water and ammonia, this extends down
to a depth of roughly 20,000 km, where it should be compressed
further into a type of solid ice. Pressure in this layer rises steadily
from 300,000 bars to 6 million bars at the core-mantle boundary.
It is possible that somewhere within this deep partly liquid shell,
methane is compressed and transformed into diamondat least in
the more massive Neptune, giving Neptune a diamond rain onto
a jewel-encrusted core. At the heart of each world is thought to lie
a small hot, rocky-metallic core with a mass perhaps a half that of
the Earth. The core probably simmers at around 5000 C in Uranus
and somewhat hotter at 7000 C in Neptune.
Given their similar mass and history, why does Uranus appears
so much colder than Neptune from the vantage of its cloud tops
and why does Uranus radiate so much less energy than its near
identical twin? Two possible answers emerge: either Uranus expe-
rienced a catastrophic collision which stirred up its interior and
allowed heat to escape, or Uranuss interior is structured in such
a way that heat is prevented from escaping. The latter hypoth-
esis seems more credible as you would imagine that a catastrophic
impact would release more heat into the interior, causing warm-
ing rather than cooling. If this insulating blanket is a seasonal
feature associated with layering in the atmosphere, it could well
alter as Uranus moves from one season to the next.
What of the diamond layer? Aside from the jewel-encrusted
possibility of diamond rain, at the pressures and temperatures
concerned, the diamonds that fall will most likely melt into a
rather interesting diamond sea that overlies the small, rocky core.
The Structure of Uranus and Neptunes

Atmospheres
Uranus has an atmosphere structured much like that of Saturn
and Jupiter. However, the lower gravitational field means that this
is much more extended. It is also much colder, lying almost twice
as far from the Sun as Saturn. This means that the layers of cloud
we have encountered before (ammonia, ammonium hydrogen sul-
fide and water) lie at much greater depth and are all but invisible
beneath a thick layer of hydrocarbon haze at 130260 km above
the 1 bar level. Figure 8.1 shows this in more detail.
Despite the bland-seeming appearance Voyager 2 encountered,
which might suggest a permanent and pervasive fog, the atmosphere
of Uranus is otherwise really clear. Indeed, Uranus has one of the
clearest atmospheres in the solar system. Only Neptune exceeds
its clarity. Like Jupiter and Saturn before it, and Neptune subse-
quently, the atmosphere of Uranus can be divided into broad lay-
ers analogous to those above the Earth. Extending downwards from
56 km at the 1 bar level to 300 km below is the troposphere. The
atmosphere is at its most dynamic in the troposphere with exten-
sive cloud decks and very strong winds powering westward flow
around the tropical latitudes and eastward flow nearer the poles.
Winds exceed 240 m per second, or the best part of 870 km per
hour, near the mid-latitudes on Uranus. On Neptune the pattern
of wind strength with latitude is broadly reversed, but the general
pattern of direction is maintained. The same equatorial easterlies
and mid-latitude westerlies are found, but on Neptune the winds
are strongest in the easterly jet along the equator, with winds in
excess of 400 m per second or a nifty 2,100 km per hour. These are
the fastest winds in the solar systema good three times faster
than winds on Jupiter and more than four times faster than the
fastest tornadic winds on Earth (Fig. 8.2).
Uranus has the coldest tropopause of any planet in the solar
system, with temperatures barely above 220 C. Oddly, the cold-
est region of the atmosphere currently lies 25 south rather than
at the dark southern pole, but this regime is changing as Uranus
moves from northern summer to autumn. Uranus, like Jupiter and
Saturn has an atmosphere dominated by hydrogen and helium.
The Ice Giants 315
Thermosphere
4,000
3,500
Altude relave to terrestrial sea level
3,000
Ionosphere
2,500
(km)
2,000
1,500
1,000
Stratosphere
500
Troposphere
-500
-200 -150 -100 -50 0 50 100

oC)
Temperature (
FIG. 8.1 The overall layout of the atmosphere of Uranus. The atmosphere
is very extended compared to that of the Earth (and much colder overall).
Extending out to around 10,000 km is an ionosphere, which may, like the
Earth, have different layers embedded within it. This overlaps a thermo-
sphere and stratosphere with temperatures that steadily fall from around
523573 C at 4000 km up to 220 C at the tropopause which is situated
at 56 km above the 1 bar level (equivalent to sea level on Earth). It is cur-
rently unknown why the upper atmosphere is this hot. The stratosphere
has a hydrocarbon haze layer at an altitude of 130260 km above the
1 bar level. Within the troposphere lie four cloud decks. The coldest and
highest is composed of methane at 12 bar; ammonia or hydrogen sul-
fide clouds at 310 bar; ammonium hydrogen sulfide clouds at 2040 bar
and finally water clouds deep down at 50300 bar. Thus Uranus (and
Neptune) have cloud decks much like Jupiter and Saturn, but these are
found at greater depths and pressures where temperatures are sufficient
to allow their condensation
200 200
250
50
100 -200
-50 -300
100 -200
230 50
150 200
FIG. 8.2 The circulation at the cloud-tops of Uranus (left) and Neptune
(right). Both planets show three jet streams in their atmospheres, but
the overall strength is different. While Uranus has only a weak easterly
jet stream above its equator, with winds of 180 km per hour, Neptunes
winds scream around its equator at well over 1000 km per hour (left).
Both planets have strong westerly jet streams at mid-latitudes with Ura-
nus having its peak winds in its northerly (Sun-facing) jet. The southern
jet is around 20 m per second slower
Unlike the former two worlds, the amount of helium is broadly

the same as found in the Sun, indicating that little if any helium
has settled into the planets deep interior. The remainder of the
atmosphere (around 2 %) is methane. Rayleigh scattering by this
compound gives Uranus its deep aquamarine color and Neptune
its vivid blue. Above the methane clouds the atmosphere has far
less methane, most of it having condensed out at lower levels.
The upper troposphere and lower stratosphere of both Uranus
and Neptune hosts layers of hydrocarbon haze produced, as
with Jupiter and Saturn, when ultraviolet light and cosmic rays
impact and energize molecules of methane. The broadest of these
bands lies in the stratosphere between 160 and 320 km above the
nominal 1 bar level. Overall, it is the narrow distribution of these
haze layers that ensures that the remainder of the atmosphere is
clear (Fig. 8.3).
On Uranus, the methane cloud deck was found to lie at
1.21.3 bars with deeper ammonia, ammonium hydrogen sul-
fide and water clouds implied from spectral measurements, and
through implication from the general similar (overall) composi-
tion of Uranus and Neptune to Jupiter and Saturn. Water clouds
are thought to lie at 50300 bars, ammonium hydrogen sulfide at
610 bars, although this is uncertain (Fig. 8.4).
The Ice Giants 317
400 0.00002
300
Altitude relative to terrestrial sea level (km)
Stratosphere
Pressure (relave to terrestrial sea level)

0.0002
200 Hydrocarbon Haze
100 Hydrocarbon Haze 0.02
Tropopause (56km) 0.2

0
A 2
Hydrocarbon Haze
-100 B
Troposphere C
20
-200
D
-300 200
-150 -50 0 50 100

o
Temperature ( C)
FIG. 8.3 Zoom-in on the lowermost stratosphere and troposphere of Ura-

nus. Temperatures rise across the tropopause as we descend through the
troposphere, the base of which lies at around 320 km below the 1 bar level
(dashed blue line). Various layers of hydrocarbon haze fill the upper tropo-
sphere and lower stratosphere. These hazes consist of ethane and acetylene
mixed with more abundant methane. Methane clouds (A) occur within
50 km of the 1 bar level, with ammonia (or hydrogen sulfide) (B), ammo-
nium hydrogen sulfide (C) and water clouds (D) at greater depths (Fig. 8.1).
Temperatures rise from around 220 C at the tropopause to over 200 C at
the base of the troposphere. More limited heat flow from the Uranian inte-
rior means that convection, and hence clouds, are more limited in extent
in the atmosphere of Uranus than the other giant outer planets
As well as bright clouds, which appear to becoming more

common, Neptune and Uranus also share a propensity for dark
spots, the most famous was the anticyclonic Great Dark Spot
that was visible when Voyager 2 visited in 1989, but subsequently
shrank and faded.
400
300
Altude relave to terrestrial sea level (km)
Stratosphere
200 Hydrocarbon Haze
100
Tropopause (56km)
Hydrocarbon Haze
0
A
-100 B
C
-200
Troposphere D
-300
-150 -50 0 50 100

Temperature (oC)
FIG. 8.4 The structure of Neptunes lower atmosphere. Like Uranus the
temperatures at the tropopause are extremely low at around 220 C. Con-
vective clouds of methane (A) are visible in Voyager 2 and Hubble Images.
The deepest layers, visible in the larger storms are dark but the composi-
tion of this layer is unknown. It is likely that like Uranus, underneath
the visible methane clouds lie deeper layers of ammonia (B), ammonium
hydrogen sulfide (C) and water vapor (D). These are too deep to currently
observe directly. Like Uranus various layers of methane and other hydro-
carbons form hazes at higher altitudes
Lawrence Sromovsky and Patrick Fry used HST and Keck

images to analyze cloud movements in the atmosphere of Uranus.
There appeared to be no overall change in wind speeds from 1986 to
2004. However, they did find a wealth of new information. Storms
came in a wide variety of forms: some lasting hours; while another,
dubbed S34, appeared to be a long-term feature throughout the
The Ice Giants 319
period. S34 also appeared to oscillate in latitude from 32 to 36.5

S suggesting that it was entrained in a Rossby Wave (Chap. 1) as it
pummeled around Uranus.
Above the tropopause is a stratosphere with steadily rising
temperatures (Figs. 8.1 and 8.3). Heating occurs by conduction
from the layer above and through the action of cosmic rays and
ultraviolet radiation from the Sun. Although not detected thus
far it would seem likely that atmospheric gravity waves exchange
energy between the different atmospheric layers, as they do on
Venus (Chap. 5) and other worlds. Well say more on these phe-
nomena in Chap. 10.
Finally, above 4000 km is the thermosphere. At 850 K
(577 C) this is by far the warmest in the outer solar system, bar
Jupiter. At present the reason for this intense heat is unclear, but
it may relate to the shape and activity of Uranuss magnetic field.
Overlapping the stratosphere and thermosphere is an ionosphere
where energetic radiation has stripped atoms and molecules of
one or more of their electrons, leaving them positively charged
(Chaps. 5, 6 and 7). It is through such action that particles can
attain enough energy to escape both Uranus and Neptune.
Uranus has an odd thermosphere in other regards. For one,
it is vast: stretching 50,000 km, or several times Uranuss radius,
deep into space, it is more like the vast coma of a comet rather than
an atmospheric layer. The thermosphere encompasses the planets
dark rings and exerts drag on them. Ultimately this causes par-
ticles from the rings to rain into the upper atmosphere. The ther-
mosphere will also bombard the rings with energetic particles and
could well contribute to chemical reactions that drive their darken-
ing. Uranuss extensive thermosphere also necessitated NASA car-
rying out course corrections for Voyager 2, which plowed through
it en route to Neptune. The underlying cause of this puffed up ther-
mosphere is unclear but it may relate to heating from charged par-
ticles that are trapped within and energized by the planets peculiar
and ever-changing magnetic field (Figs. 8.6 and 8.7).
Neptune does not share such an extensive or warm thermo-
sphere, ensuring that it hugs more tightly to its cloud tops and
experiences a lower rate of loss to interplanetary space. Neptunes
somewhat higher gravitational pull also contributes to this effect.
However, where Neptune holds a cooler thermosphere, Voyager 2
detected something odd in its stratosphere. Here, temperatures rise

to over 480 C over the equator, while further north and south over
the mid latitudes they remain resiliently below zero. At either Pole
the air appears to warm again, however, not to the heights seen over
the equator. Why is the atmosphere so warm here? The suggestion
is that over the mid-latitudes, where winds blow from west to east,
air is rising and cooling. High in the stratosphere this reverses and
arcs back, mostly towards the equator but also, to a lesser extent,
towards the Poles. As this air descendsand one must remember
that it is moving at hundreds of meters per second to the east as
well as downwards under gravityit compresses and warms up.
This makes the equatorial regions of Neptune particularly warm,
high above the frigid, tropospheric cloud layer. At present, with
fairly limited data on Neptune available, this remains conjecture.
Finally, clouds on Neptune are more apparent in part because
it has more dispersed cloud decks. In many images methane cir-
rus casts particularly striking shadows on the lower, hazier lay-
ers. These filamentous clouds are situated at altitudes 50100 km
above the deeper tropospheric haze layer. Belts are tens of
kilometers to a couple of hundred kilometers wide and thousands
of kilometers long, blown across Neptunes face by the powerful
tropospheric winds.
Seasons
Despite these similarities there is one obvious difference between
Uranus and Neptune: the angle of tilt. Uranus lies tilted at an
angle of 97.7, while Neptune has a much more modest tilt of 29.
This has obvious implications for its climate, seasons and for their
wild magnetic fields.
When Voyager 2 arrived Uranus had its south pole tilted
towards the Sun, while Neptune was experiencing its southern
spring. Over the last 30 years Neptune has been growing steadily
brighter at near infrared wavelengths (Fig. 8.5). This appears to be
the result of an increase in the abundance of methane clouds high
in Neptunes atmosphere. Sunlight, striking the upper troposphere,
is warming it as spring advances. This allows methane, which
is normally frozen into ice, to sublimate (turn directly to gas).
The Ice Giants 321
FIG. 8.5 Neptune in the infrared. This doesnt show heat so much as
where the high, cold methane clouds are formed. Methane clouds appear
to be confined to the temperature latitudes where the westerly jet streams
are found. This forms two nearly continuous bands. Nearer the equator
methane clouds are much patchier. Since 1980 Neptune has been grow-
ing brighter at near infrared wavelengths because the cover of methane
clouds has been increasing in the southern hemisphere. This probably
relates to the change in season
Gaseous methane then leaks out into the stratosphere where it

influences the chemical make-up of Neptunes upper atmosphere.
This pattern should continue for another 30 years or so while
the southern hemisphere moves into summer and the upper atmo-
sphere is warmed more thoroughly by the distant Sun. Meanwhile,
the southern hemisphere of Uranus has now moved into its
autumn, while the northern hemisphere appears to be perking up.
Where would you find the warmest conditions along the
cloud tops in the troposphere of Uranus? You might expect that
after 20 years the southern hemisphere would be the warmest, but
on average the equator is the sultriest place to go. By contrast, the
warmest place on Neptunes cloud tops is found across the South

Pole where the Sun has warmed the region for 40 years. Quite
why each planet shows such contrasting responses to heating is,
as with so much of these worlds, unclear at present. Regardless,
it further adds to the idea that there is something odd about the
way Uranus releases heat from its interior and transfers this (and
energy from the Sun) through its atmosphere.
Dark Spots
Both Uranus and Neptune display deep, dark spots. Perhaps the
most iconic feature of Neptune was its rather gothic take on
Jupiters Great Red Spot (GRS). First observed by Voyager 2, this
dark storm gradually migrated towards Neptunes equator then
petered out in the early 1990s. Although this anticyclonic feature
didnt persist, it has been followed by many more similar features
on Neptune and on Uranus in the last few decades. The Great
Dark Spot appeared at the same latitude as the Great Red Spot on
Jupiter and the Storm Alley on Saturn. Perhaps, just perhaps,
there is a common structural feature of the atmospheres of these
planets that preferentially creates storms at this latitude. It could,
however, just be coincidence. Clearly, some atmospheric model-
ing is needed.
In 2006 the Hubble Space Telescope spotted the first Uranian
dark spot north of the equator. Overall, Uranus is now showing a
much more turbulent atmosphere, rich in methane clouds, as its
seasons move on with the unrelenting advance of Uranus along its
84 year-long orbit. It seems Voyager 2 arrived when Uranus was at
its least interesting.
Each storm reveals a window into the deeper interior of these
planets. However, while air may well be descending in the heart
of these anticyclonic storms, the presence of high, methane clouds
around their periphery suggests that, like the GRS, air may well
be rising at higher altitudes within the troposphere (Chap. 7).
Despite differences in longevity, these anticyclonic storms may
share a common structure and perhaps a common origin at deeper
levels within each planets lower atmosphere1.
1
The L-class red dwarf W1906+40 also hosts a long lived dark spot with a size similar to Jupiters
red one.
The Ice Giants 323
Such anticyclonic storms indicate volatility in the atmo-

spheres of these planets where otherwise east-west, westeast
winds dominate. This brings them somewhat closer to Jupiter and
Saturn, both of which have atmospheres studded with similar vorti-
ces (Chap. 7). Given what we know about the Earth we might expect
circular storms to develop at the interfaces of the westerly and east-
erly jet streams, where air is forced to change velocity abruptly.
Such dramatic differences in air flow would be expected to set up
waves in the atmosphere which would favor the development of
storm systems. Why these features appear to be more persistent on
Jupiter and not Uranus or Neptune may simply be a reflection of
scale. Winds on Uranus and Neptune are far faster than those on
Jupiter and Saturn and this might shear storms apart. It might be
that there is a deeper connection between the Great Red Spot and
Jupiters interior than the dark spots share with the icy mantles of
their worlds. It is remarkable that, despite a large difference in the
amount of heat escaping Uranus and Neptune, they look so very
similar otherwise. Indeed, their similarities seem to grow as both
worlds move coincidently towards their respective equinoxes.
The Twisted Tale of Ammonium Metal

Deep within the interior of Uranus and Neptune, water and ammo-
nia dominate. These two chemicals can react with one another to
form ammonium hydroxide. This chemical breaks into two ions,
ammonium and hydroxide in solution. These make the entire icy
layer strongly conducting, a process exacerbated by the effect pres-
sure on ammonium ions. While both Uranus and Neptune are
too insubstantial to form metallic hydrogen, they can compress
ammonium ions so that they behave in much the same way. In
this hot, watery soup electrons are free to wander around between
the ions that make up its chemical structure. At very high pres-
sure it is thought that water also behaves in an odd way, with the
oxygen atoms forming a solid crystalline lattice through which the
hydrogen ions and electrons wander. Both of these processes make
the entire interior layer of ice highly conducting. The observed
fields are off-set from the center of each world, implying that they
are produced in relatively thin shells within the ice layer. This, in
turn implies that either only a small fraction of the icy mantle
is in a liquid form, or that only a small fraction has ammonium
and water in the appropriate form to conduct electricity. Whatever
the mechanism, each ice giant is able to stir this region to produce
a large circulating magnetic field that wraps around each world.
The magnetosphere is generated by the spin of Uranus and
Neptune. As Uranus spins once every 16 h, its rapid rotation drives
the movement of ions through the planets interiors. In turn, this
creates a powerful electrical and magnetic field that envelopes
the planet. This off-center generator has some odd effects. As
each planet spins rather than the field moving uniformly with the
planet, the field whips around at uncomfortable angles as though
the planet was holding its field generator at arms length. The
effect is less pronounced at Uranus but has quite marked effects
out at Neptune. Uranus, with its strongly tilted axis, has its mag-
netic field broadly aligned with its spin but lies on its side relative
to the plane of the solar system. By contrast, Neptune sits more or
less upright, with its magnetic field highly inclined to its spin axis
(Figs. 8.6 and 8.7). In both instances this means that the magnetic
field points towards and away from the Sun.
Consequently, as both Uranus and Neptune rotate, their
magnetic fields are contorted into cork-screw shapes that whip
out behind each planet in the direction of the solar wind. What
is even weirder is the strong variation of the magnetic field with
latitude; something we do not experience much on Earth. Because
the field is not located in the geometric center of Uranus, the field
you would experience at Uranuss cloud tops would vary from less
than a quarter that of the Earth (0.1 Gauss) in the southern hemi-
sphere and up to twice as strong as ours in the northern hemi-
sphere (1.3 Gauss).
Neptunes field is very similar as a result of the greater off-set
of the magnetic engine from the center than is found in Uranus
(Fig. 8.6). Both Uranus and Neptune also display two different
kinds of magnetic field with one embedded within the other. The
dominant field is a strong dipole fieldan Earth-like field emerg-
ing from the north magnetic pole and flowing towards the south in
a donut-like configuration. Both worlds also show an underlying
quadrupole field with other blebs and folds punctuating the dipole
The Ice Giants 325
Planet Earth Jupiter 200 Saturn Uranus Neptune

Tilt of Rotaon axis 23.5o 3o 27o 98o 30o
Tilt of Magnec 12o -10o 0o -59o -47o
Field
Oset of Field Axis 8% 10% 5% 31% 55%
Field at Equator 31,000nT 428,000nT 22,000nT 23,000nT 13,000nT
Magnetosphere 10 Earth Radii 65 Jupiter Radii 20 Saturn Radii 18 Uranus Radii 25 Neptune Radii
FIG. 8.6 A comparison of the magnetic fields of the giant outer plan-
ets. Aside from Jupiters, which dwarfs all the others, most have field
strengths comparable with the Earth. Jupiter and Saturn generate theirs
in their core, while Uranus and Neptune generate theirs further out.
Image credits: Jupiter (HST/NASA); Saturn Cassini/NASA; Uranus:
HST/NASA; Neptune Voyager 2/NASA
field. This means that there is substantial variation in the strength

of the magnetic field an observer would experience at Neptune just
as there is at Uranus. Like Uranus the field strength varies, but in
this case this variation is a temporal one that is associated with
Neptunes 16.1 h rotation. Thus while there is a varying magnetic
field strength with latitude on Uranus, this would be fixed from
hour to hour. Meanwhile at Neptune the field changes rapidly
every hour. Imagine you traveled to Neptune only to find that your
compass steadily pointed to a new north over the course of a day.
At 23 Uranian radii beyond the center, the magnetic field
forms a bow shock that deflects the bulk of the solar wind. With a
weaker solar wind at Neptunes orbit Neptune is able to stem the
flow at 35 Neptunian radii. Within the magnetosphere, charged
particles fill much of the space, as they do around the other giants.
While Jupiter and Saturn gets most of their charged particles from
the solar wind or from the satellites embedded within it, Uranus
Jupiter
Rotation Axis
Magnetic Field
Orientation
Diameter of Sun
to scale with
magnetic field
Saturn
Uranus
Neptune
FIG. 8.7 The scale of the magnetic fields around the giant planets. Jupi-
ters dwarfs the others and stretches 700 million kilometers to Saturn.
Those of Uranus and Neptune are steeply inclined to the oncoming
solar wind and whip the field around in different directions as the planet
rotates. Planetary field lines are in yellow while the solar field lines are
in red. The spin axes and magnetic field axes are indicated above along
with a comparison to the diameter of the Sun
appears to populate its field with charged particles from its highly
extended ionosphere. This is supplemented by material blasted
from the surfaces of its icy satellites and its rings, as well as a little
material from the solar wind. Uranuss magnetic field has wide
swathes carved into it as its moons sweep through the magnetized
sea of particles. In turn, these particles appear to be responsible
for darkening (or space weathering) the surfaces of the icy moons.
Do Neptune and Uranus experience aurora? In 2011 the
Hubble Space Telescope detected aurora in spots over the Uranian
equator. Why here? Again, this is thanks to the oddly aligned mag-
netic field. At particular locations across the equator the mag-
netic field lines slice through the cloud tops, through the planet
and back out towards the Southern Pole. This forces particles to
The Ice Giants 327
slam into Uranuss atmosphere at distinctly odd locations. By

contrast, in an archived image of Uranus, taken in 1998, there is
also evidence of a ring of aurora near the planets Northern Pole.
A weaker region of emission also appears in the southern hemi-
sphere. Similarly, an observer on Neptune can experience aurora
at various latitudes and longitudes depending on the angle of the
fast rotating magnetic field compared with the observer, which, as
weve seen, alters throughout the Neptunian day.
Aurorae are far less intense at Neptune than they are on
Earth. While terrestrial aurorae generate over 100 billion Watts
of power, those at Neptune are wimpy 50 million Watts (50 MW)
affairs. Compare this to the output of the Three Gorges hydroelec-
tric plant in China, which produces over 22,000 MW. You could
power 450 Neptunes worth of aurora with that monster.
Conclusions
Uranus and Neptune form an interesting pairing of icy worlds
in the Solar Systems deep, dark outskirts. Receiving as little as
1/400th and 1/900th the solar energy we receive, respectively, each
world is a domain of ice. Most of this ice is hidden and is distinctly
warm. While Neptune releases over twice the energy it receives
from the Sun, Uranus remains frigid. Their cyan atmospheres are
oddly Earth-like in appearance yet mask a deep, frigid sea of gases.
At present relatively little is known about each world. This
wall of ignorance is perpetuated by the enormous orbital peri-
ods of each planet. Uranus takes the greatest part of a human
lifetime to complete one orbit, while Neptune takes twice this
long. Consequently, it will take several generations of humans
to observe the full seasonal range of each world. That said it is
already apparent that each planet displays marked contrasts from
summer to autumn or winter to spring. Uranus, with its wild tilt,
will undoubtedly experience the greatest variation. By the time
a Uranian year is complete, which of our models will be vindi-
cated and which will find themselves consigned to the trash can?
Already we are playing catch up to a world in transition. Hopefully,
NASA or another agency will have sent probes back to one or both
of these worlds long before their birthday comes round once more.
References
1. Kaspi, Y., Showman, A. P., Hubbard, W. B., Aharonson, O., & Helled, R. (2013). Atmospheric
connement of jet streams on Uranus and Neptune. Nature, 497, 344347.
2. Lawrence Sromovsky & Patrick Fry. (2015). Dynamics of cloud features on Uranus. http://
arxiv.org/pdf/1503.03714v1.pdf.
3. Hubbard, W. B., Nellis, W. J., Mitchell, A. C., Holmes, N. C., Limaye, S. S., & McCandles,
P. C. (2013). Interior structure of Neptune: Comparison with Uranus. Science, 253, 648651.
4. Lamy, L., Prang, R., Hansen, K. C., Clarke, J. T., Zarka, P., Cecconi, B., et al. (2012). Earth-
based detection of Uranus aurorae. Geophysical Research Letters 39(7). http://hubblesite.org/
pubinfo/pdf/2012/21/pdf.pdf.
5. Aurnou, J. M. & Heimpel, M. H. (2004). Zonal jets in rotating convection with mixed mechani-
cal boundary conditions. Icarus. 169, 492498.
9. Ice Dwarves: Titan, Triton
and Pluto
Introduction
Titan, Triton and Pluto form a triad of similar icy bodies in the
outer solar system. Each has a diameter around 25003500 km
across with an interior dominated by ice and rock. Each plays
host to an atmosphere dominated by nitrogen. Here, however, the
similarity ends. The atmospheres of Triton and Pluto are tenu-
ous affairs, a thin veil of gases that loosely cling to the surface
of each world, while Titan hosts an atmosphere thicker than the
Earths. Titans atmosphere has been rich enough to bequeath the
diminutive world with a rich hydrological cycle based on liquid
methane and ethane. Consequently, Titan displays many of the
complex eroded landforms found on Earth.
Triton and Pluto, while far colder than Titan, still host dynamic
though tenuous atmospheres that circulate in a manner that would
be familiar to us. Both worlds show some aeolian features that are
found on the Earth, Mars and Venus, though in smaller quantities.
Indeed, as data continues to pour in from the New Horizons probe,
the surface and atmosphere of Pluto appear to be far more dynamic
than was presupposed on the basis of what we knew of Triton.
This chapter, therefore, represents a snapshot of what is otherwise
a rapidly evolving picture of weather in our outer solar system.
Titan
Introduction
Titan stands alone in the solar system as the only natural satellite
that hosts a significant atmosphere. Triton shares an atmosphere
dominated by nitrogen, making Triton to Titan as Mars is to the

DOI 10.1007/978-3-319-25679-5_9
Earth: a pale reflection of this cloaked world with only the barest
wisp of gas. Titans atmosphere, predominantly nitrogen in con-
tent (98.5 %), also contains an abundance of different hydrocar-
bons with the majority of this being methane. These proportions
change as we descend from the stratosphere towards the surface
because methane condenses out as clouds at low altitudes. Were
one at the surface, nearly 5 % of the atmosphere by mass would
be methane. Around 0.10.2 % is free hydrogen, but this is mostly
found higher up. The hydrogen comes from the breakdown of
methane through the action of ultraviolet light. Near the surface
Titans atmosphere is denser than that at the surface of the Earth.
General Structure of Titans Atmosphere
Titan has an atmosphere layered much like that of Earth (Fig. 9.1).
The lowest level is the troposphere where most of the weather
action occurs, but unlike our atmosphere where the two retain
fairly distinct identities, Titans troposphere is coupled very
strongly to the stratosphere above. Recent evidence from Cassini
indicates that this coupling extends all the way through to the
thermosphere where a vast circulation extends from near the top
of the atmosphere to its base hundreds of kilometers below.
The atmosphere is cold throughout, with temperatures lowest
around 50 km above the surface where the troposphere gives way
to the stratosphere. The lowest 30 km contains most of the plan-
ets true clouds. These are mostly made of methane with an admix-
ture of more complex hydrocarbons. As in the Earths atmosphere
the stratosphere shows steadily rising temperatures caused by the
absorption of ultraviolet light by complex chemicals. While ozone
has this effect in our atmosphere, a complex brew of hydrocarbons
causes the effect above Titan. These compounds form hazes that
fill the space from 100 to 210 km above Titans frigid surface.
Titans mesosphere shows cooling but never gets as cold as
the troposphere thanks to heat is conducted downwards from
the upper thermosphere. Once again, hazes form distinct layers
around 300 and 450500 km up, while in the winter hemisphere
the polar hood begins its life this high up before extending down-
wards into the troposphere.
Ice Dwarves: Titan, Triton and Pluto 331
1,500
10-11
1,250
Pressure (bar)
Altude (km)
1,000 Thermosphere 10-9
750
10-7
Polar Hood Clouds

500
Hydrocarbon/Tholin Haze 10-5
Mesosphere
Hydrocarbon Tholin Haze
250 10-3
Hydrocarbon/Tholin Haze Stratosphere
50 0.1
Methane Clouds Troposphere
-200 -150 -100 -50 0
Temperature (oC)
FIG. 9.1 The general structure of Titans frigid atmosphere. Temperatures

fall from around 180 C at the surface to 200 C near the tropopause.
Like the Earth the atmosphere is divided into four principle layers, tro-
posphere, stratosphere, with rising temperatures, mesosphere, with low-
ering temperatures and a (relatively) warm thermosphere with rising
temperatures. Methane clouds are confined to the troposphere, while
layers of tholins, ethane, other hydro-carbons and cyanide hazes fill lay-
ers within the stratosphere and mesosphere. Methane clouds form two
distinct decks: one at 2030 km and a lower deck at 815 km. Polar Hood
clouds can form as high as 300600 km above the winter Pole
Titan possesses a very extended atmosphere, thanks to its low

gravity. The upper atmosphere consists of a thermosphere extend-
ing above 520 km, and this strongly overlaps with an ionosphere
the region where much of the gases are ionized. Ionized layers are
found as low as 63 km in the mesosphere, but the main region lies
above 1400 km, giving Titan an atmosphere more than quarter the
width of the moon.
The Weather Report for Titan
Whats the weather like on Titan? For starters it is extremely cold,

with average temperatures around 180 C. Its also fairly cloudy,
and though clouds as they would be recognizable on the surface of
the Earth are relatively rare, there is a persistent high level haze.
This is produced when a combination of ultraviolet light from the
Sun and cosmic rays energize methane and nitrogen gas. These
create a sea of complex hydrocarbons and nitrogen-containing
organic compounds that block out much of what is already a dis-
tant and dim Sun. That is not to say Titan doesnt have its fair
share of clouds, only that they are more diffuse in nature and often
localized to specific regions at certain times of the Titan year.
Titans surface temperature is far lower than one would natu-
rally expect for an object at Saturns distance. This is the result of
two opposing factors: the methane that is present acts as a green-
house gas, while the low temperature means that much of it is fro-
zen out as ice particles high in Titans atmosphere. This, coupled
to the pervasive haze, blocks much of the sunlight that strikes
the satellite. As a result Titan is a good 1020 or so colder than
it would otherwise be. Titan is thus a fairly good model of Earth,
where aerosols produced by the burning of fossil fuels impact on
our surface temperatures and climate.
What of the circulation? Titan turns out to be an interesting
hybrid of Mars, Earth and Venus. Above 20 km the winds are orga-
nized into broad westerlies that circulate around the entire globe,
much as happens in the stratosphere of the Earth. However, unlike
the Earths stratospheric winds, these move faster than the under-
lying surface: they super-rotate. Weve already encountered super-
rotation in Chaps. 5 (Venus) and 7 (Jupiter and Saturn). In each
case winds are moving considerably faster than the rotating globe,
beneath. The difference with Titan is that the winds on Venus
super-rotate from east to west, while Titans winds super-rotate
with the moons rotation from west to east.
Figure 9.2, shows some that we would expect to find in the
atmosphere of Titan. At altitudes of 2030 km the winds change
direction with the planetary Hadley cell (or cells) broken into two
layers. Again, this pattern is seen on Venus, but here an upper
cell carries heat from the day to the night side, while the lower
Methane cirrus Northern winter 2004

Polar
Hood 20
Altude (km)
10
-90 -60 -30 0 30 60 90

Northern Spring 2009
20
Altude (km)
10
-90 -60 -30 0 30 60 90

Northern Summer 2016
Polar
20
Altude (km)
Hood
10
-90 -60 -30 0 30 60 90
FIG. 9.2 General circulation in Titans troposphere. Titan has seasons

similar to the Earth as it orbits Saturns tilted equator. Seasons last sev-
eral years as Saturn completes its 29.5 year orbit of the Sun. The middle
and upper troposphere is dominated by a single Hadley Cell (Chap. 1)
much like Venus. This takes warmer air from the summer hemi-sphere
to winter hemisphere. In the spring the atmosphere splits into two Had-
ley Cells centered on the equator. The lower atmosphere contains one
or more additional cells where heating is strongest. Images: Cassini/
ESA/ NASA
cells carry heat from equator to pole. On Titan the two cells are
organized broadly from the moons warmest to coldest regions.
Interaction between these two layers of cells generates turbulence
and the temperature differences help form banks of cirrus and cir-
rostratus clouds. Unlike their terrestrial analogues, these are made
of frozen methane crystals. As the air is expected to be saturated
with methane these crystals can grow until they are heavy enough
to fall towards the surface.
Methane snow does not have the same crystal structure as

typical terrestrial snow. Water vapor tends to form hexagonal
structures that are dictated by the formation of weak chemi-
cal bonds between water molecules. Methane cannot form such
bonds (Fig. 9.4) so the crystals adopt a different shape. Experiments
suggest that methane snow would be more like the rare terres-
trial diamond dust. Diamond dust is occasionally seen on Earth,
mostly within the dry continental interiors, during the deepest,
coldest parts of winter, when the air becomes super-chilled and
ice forms directly in the cold, dry air. On Titan, the air in the
region methane snow forms is also likely to be far colder than
the temperatures normally required to form methane crystals.
This super-cooled methane probably produces eight-sided octahe-
dra or even fourteen-sided cuboctahedra diamond dust that drifts
lazily down under Titans low gravity to the cloud decks below.
In amongst these odd methane crystals will be ethane hexagonal
crystals, which although structured like water refract the light
much more strongly.
As these crystals descend into warmer air the crystals melt
forming a fine and very frigid drizzle. By the time these droplets
penetrate the lower cloud decks at 816 km up, they have absorbed
a considerable amount of nitrogen. Although much of this subse-
quently diffuses back out before the drops hit the surface, it does
alter the density of the drops and reduces the speed that they ulti-
mately hit the ground. What would Titan rain be like? In short:
rather odd to our eyes. Methane is less dense than water so this,
alone, will make it fall more slowly than terrestrial rain. Factor in
Titans lower gravity (roughly one third that on the surface of the
Earth) and denser atmosphere, which provides greater air resistance
and would marvel at the spectacle of large rain drops drifting lazily
downwards. These would have little capacity to erode surfaces
as they do on Earth, but they could collect on the already sodden
surface and pour downwards under gravity. Erosion could occur
here, though this mechanism does not appear sufficient to produce
Titans widespread surface features.
Not a fan of rain? What about snow? In 2006 Cassini discov-
ered a 150 km long, 1.5 km (4800 ft) high mountain range that
appeared to be capped with methane snow. The lower clouds are
likely most similar to terrestrial stratus or stratocumulus clouds:
shallow in extent with only limited vertical movement of mate-

rial upwards or downwards. Like a dull winters day on Earth, a
thick bank of stratus delivers a cold and wet countenance to Titan.
It is this widespread drizzle that maintains Titans boggy surface,
which Huygens detected when it landed in 2005.
Despite the likely frequent nature of this slight rain on Titan
there is clear evidence that Titans atmosphere must deliver a
stronger punch at least occasionally. When Cassini-Huygens
arrived at Saturn in 2004 Titan was still in the grip of its northern
winter (Fig. 9.2). There was little evidence of the large, convec-
tive clouds that would have appeared necessary to spawn more sig-
nificant precipitation. However, the surface of Titan shows clear
and abundant evidence for large flows of what is likely methane.
There exist countless dry river channels, much like a terrestrial
desert. These cannot have been eroded by fluids deposited as a
steady drizzle. Instead, these require a strong burst of heavy rain-
fall and this in turn requires strong convection and deep clouds.
Computer modeling had shown that such convection was possible
on Titan where topography or solar heating drove uplift. On Titan
the main driver will be solar heating. With a surface saturated in
liquid hydrocarbons the atmosphere is fully saturated with meth-
ane within a few kilometers of the surface. Under these conditions
only a slight increase in temperature (less than 1 C) is enough
to drive the formation of storm clouds over 20 km high. As with
water on Earth, when the gaseous methane cools and condenses,
it releases latent heat which encourages further uplift. Therefore,
given the right push at the surface, methane storm clouds can rise
through half the height of the troposphere and generate clouds
massive enough to produce substantial rain fall.
Where would we find these flash floods? Figure 9.2 provides
some answers. Topography can provide uplift and there are small
pockets of convection likely in the mid-latitudes, the most likely
locations for rain storms are where the air is rising most strongly.
These are the Polar Regions in the summer hemisphere or across
the equator during the spring. Indeed, the simplest analogy
between Titan and the Earth are the tropics. On Titan the area of
greatest uplift and precipitation is the Inter-tropical Convergence
Zone (or ITCZ for short: Chap. 1). On Earth this forms the heart
of two broadly symmetrical Hadley cells that span 30 north to
30 south. While this belt of convergence moves seasonally north

and south with the overhead Sun, it is confined to the mid-riff of
Earth. On Titan, with very slow rotation (15 Earth-days, 22 h), the
Hadley cells are global in extent and run from Pole to Pole. Thus
as the seasons progress, the region of strongest convection moves
from the South Pole to North Pole and back again, once every
29.5 Earth-year long Saturnian year. At present (late 2015), Titan
is moving into its northern summer and the band of strongest pre-
cipitation will move with it, reaching the northern pole in 2016.
These storms are thus expected to be the agents that create the
seas of Titan, which are most prevalent within the Polar Regions.
Microbursts and Aeolian Features on Titan
Figure 9.3 shows the mean wind directions on Titan from season
to season. Winds blow weakly from east towards the west across
the equator during the spring season with virtually no wind dur-
ing the northern winter or summer. When Huygens arrived and
descended immediately south of the equator, it encountered first
2004 2009 2016
FIG. 9.3 Average wind directions at the surface and at height in Titans
troposphere in different seasons. Large arrows show the Hadley cells that
carry warm air from the areas that are most strongly heated to the poles
or from the summer to the winter hemisphere predominantly at altitudes
less than 20 km. The more regional circulations indicated in Fig. 9.2 are
not shown. Of particular significance are the equatorial easterly winds
that kick in only during the equinoxes. Also not shown are the general
movements of air in the upper troposphere (above 20 km), which appear
to flow from west to east during the northern winter and summer. The
upper cloud deck at 2030 km appears to separate these two belts of
wind. 2004 was the northern winter; 2009, the northern spring; and 2016
will be the start of the northern summer
a strong (120 m per second or 432 km per hour) westerly storm.

These winds decreased with height in some surprising ways. Below
7 kmwell into the tropospherethe probe encountered two
changes in wind velocity. Immediately beneath seven kilometers,
the winds changed from westerly to easterly with the speed drop-
ping to around 20 km per hour. Just before touch-down, Huygens
detected a change in wind direction back to a westerly breeze, at
only 0.3 m per second or roughly 1 km per hour. Tetsuya Tokano
(University of Cologne) interpreted this change in winds as caused
by the probe descending through the Hadley Cell. Between 7 and
1 km up Huygens sliced through the upper part of the cell; while
in the lowest 700 m, Huygens encounters the cross-equatorial
flow from north to south that was westerly. It is above 7 km that
super-rotating winds are encountered.
Winds dont always follow this pattern on Titan. As with
Venus, there is a strong tidal influence on wind direction. Although
Titan experiences no significant tidal pull from the Sun, it does
get pulled to and fro by Saturn and this effect alters from day to
day and season to season as Titans orbit is elliptical. As Titan
orbits Saturn a tidal wave moves across the surface. Like ter-
restrial ocean tides, this atmospheric tide can periodically reverse
the average wind direction from west to east by a few kilometers
per hour. Compare this to the Earth where atmospheric tides are
primarily caused by heating by the Sun. Such tides migrate across
the sky in a variety of different waves as the sun heats the surface
and greenhouse gases within the atmosphere. The large sun-driven
bulge moves generally westward through the day, with smaller
bulges moving in different directions depending on what kind of
surface lies underneath. By contrastand this may be surprising
the Moon, which drives ocean tides, has very little effect on the
atmosphere. At Titan, the impact arises because Saturn is so much
larger than Earth and the Sun far more distant, reversing their con-
tributions to Titans tides.
Tides are not the only factor that appears to drastically alter
Titans winds either. Across the central 30 or so, north and
south from Titans equator are large dune fields that cover an area
roughly equivalent to the continental United States. These peter
out moving into the mid-latitudes but can be up to 300 m high.
The odd thing is, these all point from east to west and equatorial
winds on Titan do not generally persist in this direction, nor are

they apparently strong enough to drive the formation of dunes in
the first place.
Titans dunes are either made entirely from rather sticky but
frozen hydrocarbon globules, or water ice covered in a frozen, yet
sticky, hydrocarbon gloop. These are not expected to move at the
sluggish wind speeds experienced at Titans surface. How, then do
Titans winds move its sand into such large dunes? The sugges-
tion is that during the equinox, strong storm systems develop over
the equator. As these rise over 5 km in height they are directed
eastwards by the super-rotating upper winds. The super-rotating
winds appear to be the source of the dune orientation. Moving
against the prevailing surface winds, convective storms travel
from west to east. As they move these storms generate powerful
downbursts of wind along their leading edges. Much like thunder-
storms on Earth (or indeed the Derechos that were encountered in
Chap. 7), these eastward-moving downbursts can exceed 20 m per
second and, over desert regions, kick up enough sand to generate
a dust storm.
The key idea is that these storms have an initially high base
which allows them to move with the prevailing, super-rotating
westerly winds at height, rather than those weak easterly winds
at the surface. This allows the storm to transfer momentum
from the upper troposphere (and stratosphere) to the surface of
Titan and kick up otherwise static sand that lies across much of
Titans equator.
Titans Missing Vital Spark
You might imagine that an atmosphere that hosts very active

storm cells, more than 20 km high would also play host to some
significant lightning. However, on repeated passes Cassini failed
to find the characteristic whistler-mode radio noise produced by
electrical discharges. The Earth, Venus, Jupiter and Saturn all dis-
play these characteristic radio bursts (Chaps. 4 and 5) that signify
a discharge of electrons through the planetary atmosphere. Titan
does not. Cassini swung past Titan on seventy-two occasions and
detected copious radio emission from Saturn. However, when
Titan swung in front of Saturn and eclipsed it, the radio receivers
fell silent: there was not a single lightning detection from Titan.
The inescapable conclusion was that Titan does not host thun-
derstorms. Its methane cumulonimbus clouds produce precipita-
tion but nothing else. Why should this be? Figure 9.4 suggests one
answer to do with the chemical structure of methane.
Because methane can neither readily separate charges nor dis-
solve substances that are available, charged clouds made from it
will have a fairly uniform charge throughout. The only exception
would be if ultraviolet light or cosmic rays were able to split the
molecules of methane up into charged fragments. As this does not
happen in the atmosphere of Titan, its clouds cannot build up the
a
-
O O- H
H+ H+

- H+

+ +
O O
H+ H+ H+ H+

b -
N H
H+ H+ H+
H N +
N
+
H+ H+ H+ H+
H
H
+ +
c H
C
H H
H
FIG. 9.4 Methane versus the polar covalent molecules ammonia and
water. The atoms in water (a) and ammonia (b) bear small charges (+ or
). These help ammonia and water molecules cohere to one another and
adhere to surfaces. These molecules can also gain hydrogen ions from
other molecules or in the case of water directly through interactions with
other water molecules (a). Both these processes allow ammonia and water
to separate charges in clouds, most commonly when water or ammonia
freeze. Methane (c), on the other hand cannot indulge in these sorts of
reactions as its atoms have approximately equal charges: the molecules
are non-polar) and it cannot dissolve charged substances such as salts in
atmospheric dust, easily. Therefore methane clouds might look spectacu-
lar, but will never play host to lightning
kinds of charges needed to initiate lightning. Sadly, all of those

artistic renderings showing lightning flickering through a sullen
sky will need to go into the trash. Lightning remains the preserve
of our planets, not their moons.
The Methanological Cycle
Methane is a problem for Titan. At Titans distance from the Sun,

methane should be able to react with itself and other chemicals
under the influence of the Suns ultraviolet radiation, leading to its
breakdown. The rich stew of chemicals in Titans atmosphere pays
testament to this, but there is still an abundance of the chemical
in Titans air, leading to the belief is that methane can circulate
from Titans interior to its atmosphere.
Under Titans thick and very rigid water and ammonia-ice
crust, there is believed to lie a relatively balmy ocean. The word
relatively is important here. Hovering at somewhere between
50 and 90 C this ocean is kept fluid through a mixture of weak
tidal heating from Saturn and a small, residual amount of radioac-
tive heating, left over from the satellites formation. This heating
is not enough to thaw pure water, with a sizable dash of ammonia,
this ocean can remain liquid down to 97 C. Although the precise
route is unknown, it is thought some of the surface features on
Titan resemble cryovolcanoes. These plus cracks in the icy shell
are sufficient to allow methane and other gases to leave the ocean
and enter the atmosphere, where they can replace the methane
that is destroyed through the action of ultraviolet radiation. The
remaining methane circulates between the atmosphere and the
surface much as water does on Earth.
Rat Poison
What does Titan have in common with a ubiquitous poison found

on Earth, and what does the detection of this volatile and highly
noxious gas tell us about Titans atmospheric circulation?
In 2009 Titans southern hemisphere was heading into autumn.
By 2012 the upper atmosphere was cooling dramatically and a
cap of cloud began to form over the South Pole (Figs. 9.2 and 9.5).
FIG. 9.5 Hydrogen cyanide and the southern polar vortex of Titan. The
polar vortex stands out above the rim of the moon in the Cassini/JPL/
NASA image at center. A color zoomed image of the vortex is shown at
top right. Meanwhile, at left is an ALMA (Atacama Large Millimeter/
submillimeter Array) image taken in the microwave portion of the spec-
trum, which shows the distribution of hydrogen cyanide. This is centered
on each pole but skewed in an east-west direction by the circulation of
air hundreds of kilometers above the surface of the satellite.. However,
instead of winds smearing out the concentrations, there are still high
concentrations near each Pole. ALMA image credit: NRAO/AUI/NSF;
Martin Cordiner et al./NASA
This happened far earlier than predicted by climate models. Within

it various gases began to condense. Amongst these was the highly
toxic compound hydrogen cyanide (Fig. 9.5). Cyanide is perhaps
best known as the lethal agent of the gas chambers, or as rat poi-
son; but it also adds that acerbic bite to cherry stones, apple pips
and banana skins. Within Titans atmosphere, hydrogen cyanide
is produced when methane is split apart by ultraviolet light then
partly recombines with nitrogen. At high altitudes the air is very
rarefied and will only permit condensation when it becomes satu-
rated. This requires extreme cold. In 2009 the South Polar Region
was anomalously warm and there seemed no chance that clouds
could ever form here. Within 3 years, at an altitude of 300 kma
distance that would put terrestrial gases into outer spacea cap
cloud began to form over the top of the vortex, far above the tropo-
pause, in the heart of the mesosphere (Fig. 9.1).
The picture that emerges is that from 2009 onwards the upper
atmosphere began to cool very rapidly, in contradiction of the mod-
els that held this off until 20122013. Air in this region dropped
below 145 C and hydrogen cyanide began to freeze out forming
micrometer (millionth of a meter) sized ice grains. Measurements
by Cassini had shown warming at greater depths, around 100 km
up. This suggests that air cools rapidly once the Sun sets on the
polar region. This air begins to descend slowly from above under
its own weight, but warms through compression as it falls. In
turn, this caused brief warming in the stratosphere, before these
gases, too, cool in pitch blackness. The cyanide and other cyanide-
related gases such as acetonitrile, enhance the cooling of the air,
overall by releasing latent heat when they re-sublimate and form
ice grains. This, in turn, accelerates the downward motion of the
air above the Polar Region, helping drive the change in the circula-
tion of the air far below.
What is amazing is that these observations suggest that the
entire atmosphere to an altitude of 600 km above the ground, or
more than one tenth the diameter of the satellite, is involved in
the global changeover in circulation. This is well into the ther-
mosphere (Fig. 9.1). If we compare Titan with the Earth, although
there are certainly seasonal changes in the mesosphere and ther-
mosphere with season, these are not thought to directly change
circulation within the troposphere, where most of our weather
action occurs. Titans atmosphere, therefore, demonstrates a
degree of connectedness not apparent in the atmospheres of the
gas giants and the terrestrial planets, with the likely exception of
Venus (Chap. 5).
The Loss of Titans Atmosphere
Before we leave Titan for the solar systems outermost realm,

its worth looking at just how stable Titans atmosphere is. After
all, Titan is a fairly low gravity world, with a mass and diameter
roughly twice that of our Moon but less than one fifth the mass of
Mars. Its atmosphere extends over 1000 km above its surface. Just
compare its troposphere to the Earths: Titans is roughly 56 km
thick or eight times the average depth of ours. Although Titans
atmosphere squeezes in somewhat more mass (1.19 times the
mass) over a smaller surface area, the greater thickness of its atmo-
sphere is a consequence of its low gravity, which cannot compact
it down towards the surface as strongly as the Earth can to its own.
Titans low mass and its extended atmosphere should, therefore,
make it vulnerable to escape, if we apply the same rules as we did
to the other planets.
As with the other planets, there are two main routes for
escape of gases: thermal (or Jeans escape) and several non-thermal
mechanisms. Titans low mass should be sufficient for hydrogen
and methane to escape through the Jeans mechanism, but heavier
gases, such as nitrogen could be maintained. Despite this measure-
ments of the ratio of different isotopes of nitrogen (nitrogen-14 and
nitrogen-15) imply that rather a lot of atmospheric nitrogen has
been lost. Being lighter, nitrogen-14 should escape more readily
than nitrogen-15. As nitrogen is found as N2 (diatomic nitrogen),
molecules can be purely 14N or 15N or consist of one atom of each
form. Nitrogen-14 is by far the more abundant and this abundance
is set by nuclear reactions in stars. Nitrogen can escape as single
atoms (or ions) from the top of the atmosphere when molecules
are struck by ultraviolet radiation or cosmic rays. (See Chap. 5 for
a comparison with Venus.) Any charged nitrogen ion can then be
picked up by Saturns magnetic field and whisked off into space.
Exchanges of electrical charge between particles can also give
nitrogen atoms enough energy to escape Titans gravitational hold.
Examination of the ratio of these two isotopes of nitrogen reveals
that most of Titans original atmosphere must already have been
lostyet Titan retains a deep atmosphere.
It is possible that Titan lost a lot of its early atmosphere within
the first 50 million years of the satellites formation. Stronger heating
from a young, hot Saturn and more energetic radiation from a youth-
ful Sun, may have whittled its atmosphere down through the process
of hydrodynamic escape whereby heating of the atmosphere is strong
enough that it creates sufficient pressure to blow much of the atmo-
sphere off into space (described in Chaps. 5 and 6). Hydrodynamic
escape works best on light gases such as hydrogen, but as these blow
off into space they can carry heavier gases, such as nitrogen, with
them. This is known as hydrodynamic drag. Hydrodynamic escape
might have been accelerated by heat released from the nascent Saturn
or from tidal heating caused by changes to the orbit of Titan around
Saturn, or simply heat released from the formation of Titan, itself.

Hydrodynamic escape would have left a greater proportion of the
heavier isotope of nitrogen. A lack of significant argon points to
the present atmosphere having come from the photolysis (light-
splitting) of ammonia that was accreted as ammonia. This process
would make nitrogen, but also remove much of it. As the ammonia
was split the hydrogen that was also liberated would have escaped
Titans low gravity to space. This process has the potential to drag
nitrogen atoms with it and these are, preferentially, the lighter
nitrogen-14 atoms, which are easier to lift. The argon data also
points to the Saturnian system being rather warm during its for-
mation. This could be because of heat radiating from Saturn as it
coalesced or because, as some models of the developing solar sys-
tem suggest, young Saturn wandered around: at one point Saturn
came well within the region now occupied by the asteroid belt.
This would have greatly modified the materials available to form
its moons.
Current measurements suggest Titan is still losing rather a
lot of its atmosphere to space or to its surface. Using Cassini data,
Titan is shedding around seven metric tons of hydrocarbons per
day into interplanetary space through the process of electric force
field acceleration that we encountered in Chap. 5.1 While much
of the atmosphere appears to have been lost early on through
hydrodynamic escape (Chaps. 5 and 6), nowadays this process is
subordinate to the electric force field route. To recap, electrons
are energized by solar energy and are pulled away by Saturns
magnetic field. This leaves the upper atmosphere positively
charged, which then causes it to puff up to the point that it either
blows out into Saturns magnetosphere or is pulled into it by the
departing cloud of electrons. This process occurs most strongly
above Titans Polar Regions, earning the name a polar wind. Now,
this data only relates to hydrocarbons: hydrogen is also being lost
when it is split from methane by sunlight and then drifts off into
1
Getting a precise gure here has proved very frustrating with numbers varying from 0.4 kg per
second (34 metric tons per day), to as much as 300 metric tons per day. My calculations, based
on a different published unit (amu), were closer to, but a lot higher than, the UCL gure I went
with the UCL gure, for the described mechanism, in the end and remain vague on the total loss.
https://www.ucl.ac.uk/mathematical-physical-sciences/maps-news-publication/maps1535
space and much of this loss takes nitrogen with it. Methane reacts
to form other compounds under the influence of ultraviolet light.
These rain out onto Titans surface and are ultimately lost, as well.
The total loss is on the order of several tens of metric tons per
daybut again getting a consistent figure has proven challenging.
Using a ball-park figure of 3040 metric tons per day, thats prob-
ably reasonable and within an order of magnitude of the current
rate of loss of gases at Mars (approximately 100 metric tons). With
an atmospheric mass marginally higher than the Earth there is a
chance Titan can expect to hold onto its atmosphere until the Sun
becomes a red giant.
While the atmosphere as a whole seems secure for billions
of years to come, the same cannot be said to be true for one of its
components: methane. Methane is not stable even in the extreme
distant cold of Titan. Ultraviolet light is continuously depleting
it by converting it into more complex hydrocarbons, which then
rain out onto the moons surfaceor escape to space. Calculations
by Christopher Sotin (University of Nantes) and collaborators sug-
gest that the methane was probably released a few hundred mil-
lion years ago; possibly as a result of a giant impact or perhaps
cryovolcanism. Within a few tens of millions of years hence, this
will all have been destroyed through the action of ultraviolet light.
Titan will run dry and its current active geology might freeze out
with it, leaving a dry and desert-world: our ancient Earth analogy
will become a frigid mirror of Mars, instead. The current wealth
of atmospheric phenomena thus appears to be a temporary blip in
the evolution of Titans otherwise dry atmosphere.
It appears therefore that Titan shed much of its early atmo-
sphere in a dense wind. At some critical point the rate of mass loss
slowed to its present rate, probably as Saturn and Titan cooled
down and the orbit of Titan stabilized. Convenient though this
is, it still begs the question if Titan has already lost a substantial
amount of atmosphere, why then, is its atmosphere still dense,
while the more massive Mars struggles to retain a dribble of gases
despite a similar rate of loss? Moreover, why do Jupiters large sat-
ellites have no meaningful atmosphere at all? Jupiter plays host
to two satellites that are more massive than Titan: Callisto and
Ganymede, yet neither of these has an atmosphere other than a
sparse cloud of oxygen atoms and ions that are kicked off their icy
surfaces. There are a few possible reasons. For one, Ganymede and
Callisto were born in an intrinsically warmer part of the solar sys-
tem and, therefore, may not have had access to the same amount
of volatile gases as Titan. Secondly, heat from proto-Jupiter would
have been significantly greater than that released from Saturn.
Running from Io outwards, there is a clear trend in the abundance
of light, volatile gases: Io is essentially dry; Europa has a thin, icy
shell, Ganymede is a massive combination of metal, ice and rock;
while outermost Callisto is mostly ice. This implies Jupiter was
hot and radiated enough heat to drive volatile materials away.
This would have included those gases needed to form an atmo-
sphere, such as ammonia. Thirdly, Jupiters greater mass would
have meant that impacts from asteroids and comets would have
had more energy (and potentially a greater frequency) and thus
been able to blast any youthful atmosphere off into space. Mars
has lost its atmosphere through an unfortunate combination of
low mass and its proximity to the Sun (Chap. 6). Its low mass and
low atmospheric density means that it simply cannot effectively
hold onto what gases it has (Chap. 6). Venuss higher gravity is
more than sufficient to retain a thick veneer of gas (Chap. 5).
All of these effects would have meant Ganymede remained
largely airless while less massive Titan was able to maintain a
rich, hazy firmament. There is one more factor worth considering:
magnetic fields. Thinking back to Venus there is a perception that
a planet must have a magnetic field if it is to retain an atmosphere
against its stars stellar wind. This is a misnomer, as Venus read-
ily proves. In the case of Jupiter and Saturn magnetic fields may
have had contrary effects. Ganymede has its own, relatively strong
field but this is embedded within Jupiters enormous magnetized
blanket. Titan is also within the field of Saturn, but it lies further
out and Saturns field is around 2530 times weaker than Jupiters.
In Jupiters system the magnetic field energizes particles that have
been trapped from the solar wind. These are accelerated to enor-
mous energies before they slam into the surfaces of the Galilean
worlds. The effect of these is profound. Secondary particles are
then blasted off the surfaces of these icy worlds and driven into
Jupiters magnetic field, each forming a torus around the giant
planet. The distress of these satellites is manifest as bright auroral
spots in Jupiters atmosphere (Chap. 7). There is clearly sufficient
energy in Jupiters magnetic field to bulldoze any atmosphere that

these satellites may once have had, into deep space. Under the aus-
pices of Saturns field Titan, by contrast, enjoys some protection
from the solar wind. Yes, there is some erosion of its atmosphere
from charged particles, but this effect is far less than experienced
by Jupiters satellites. Thus, a magnetic field may not be the pro-
tective, nourishing blanket it is always assumed to be. Venus does
just fine without one, while Ganymede may have suffered because
of one. Mars didnt lose its atmosphere so much for the lack of one,
but for its proximity to the Sun and its low mass.
Does Titan retain an atmosphere simply because it is further
from the Sun than Ganymede? In part yes. But it seems it had a
higher starting point: it was simply born with more volatile mate-
rials. Its nitrogen almost certainly came from ammonia. This was
vented into the atmosphereas it may still be nowthrough cry-
ovolcanism. Even though much of this appears to have been lost
to space, it is replenished through the release of more ammonia
from an underlying ocean. Ammonia seeps out through cracks or
is vented by volcanic activity. In the atmosphere this is split to
hydrogen and nitrogen. Some of this is lost to space along with
most of the hydrogen, enough is retained to maintain a volumi-
nous atmosphere. Whether this can be maintained for the entire
main sequence lifespan of the Sun is unclear. Either way when the
Sun does expand to become a red giant, whatever Titan has left
will be rapidly removed and ultimately, see its rocky core exposed.
Triton
Introduction
Twenty AU further from the Sun than Titan lies Triton, an odd,
massive satellite with a tenuous atmosphere and an orbit that
runs retrograde around Neptune. Triton almost certainly began its
life further out in the EdgeworthKuiper Belt and was captured;
perhaps as a binary object, much like Pluto-Charon, before one of
the pair was ejected. During this time Triton was undoubtedly tid-
ally heated allowing it to manifest some of the activity we see on
its surface today.
Triton is 2,700 km acrossnearly identical to, but slightly

larger than Pluto. Like Titan and our Moon, Triton is tidally
locked to Neptune. However, because it orbits Neptune in a ret-
rograde orbit (one that has an opposite direction to Neptunes
spin), tidal bulges on Neptune are gradually pulling the satellite
down. In around 3.54 billion years Triton will cross Neptunes
gravitational Roche Lobe and be tidally shattered. Here, Neptunes
tidal pull will act more strongly on one side than the other until
Triton comes apart, much like desiccated spaghetti. Before it dies,
tidal heating will once again rejuvenate this icy world and perhaps
give it an atmosphere as dense as Titans is today. For now, Triton
resides in the deepest freeze the solar system can provide.
Tritons Atmosphere
Tritons atmosphere in some regards resembles a frigid version of

that of Mars. However, the surface pressure is 10,000 times lower
at one hundred thousandth that found at the surface of our world
(around 0.017 millibar). Unlike Mars, the atmosphere is dominated
by nitrogen, rather than carbon dioxide: for, at the temperatures
found at Tritons surface, carbon dioxide would be as hard as steel.
Indeed, fully 99.99 % of the atmosphere is molecular nitrogen (N2)
with carbon monoxide and methane making up a fraction of one
percent. Of these, measurements made from the Earth in 2010,
suggest that carbon monoxide is more common.
At 237.6 C, the surface of Triton is too cold for many gases
to remain in their free-wheeling forms; and the only reason we see
nitrogen, methane and carbon monoxide is that they sublimate
from their frozen form under the action of sunlight and cosmic
rays. Remember, that on Earth ice will sublimate at temperatures
lower than freezing: you dont have to be hot to boil. The con-
stant action of sunlight (and cosmic rays) provides just enough
energy to drive some of Tritons icy surface into the surrounding
space where it forms the barest of atmospheres. Remember the
term temperature is effectively an average measurement of parti-
cle energy. So even at 237 C some nitrogen molecules, for exam-
ple, will always have a temperature greater than that registered
by a thermometer.
Compared with Triton, Pluto is positively balmy with tem-

peratures soaring to 229 C: this is still below the freezing point
of nitrogen (210 C); carbon monoxide (205 C) and methane
(182 C), but high enough that we also expected to find a reason-
able atmosphere above Pluto.
Tritons atmosphere can be broken into three layers, based
on measurements from Voyager 2, terrestrial measurements and
from circulation models. The lowest 8 km is largely clear and still
air. Triton sports geysers on its surfacesomething we will return
to shortly. These rise vertically for 8 km and clearly indicate that
there is no meaningful wind on Triton. There is an expectation
of two general and very light flows of air: one is a tidal flow from
the far to near side of Triton, with respect to Neptune as Triton
completes its 141 h orbit, and the other is a sublimation and con-
densation flow that runs from the day side to the night side of the
satellite. These are clearly so slight that neither seems to deflect
the steady, upward flow of gases from Tritons surface. Above 8 km
the jets of dusty gas are strongly deflected to the east, indicating a
light westerly wind begins abruptly at this point.
More general measurements indicate that above the majority
of Tritons surface there is a general easterly flow (prograde with
respect to its orbit around Neptune). The term jet stream has been
loosely applied. For in Tritons equivalent atmospheric flows, air
moves at a very leisurely 34 m per second (or less than 20 km per
hour). The fastest flows are in the equatorial easterly jet (Figs. 9.5
and 9.6), with more sluggish eastward movements further north
and south. These jets mark the base of Tritons next atmospheric
layer: its thermosphere. Here temperatures rise, but only to around
178 C. Nowhere on Triton (or above it) is warm.
The lowest 8 km can be loosely thought of as a troposphere,
although there is no pronounced rise in temperature as we move
into the thermosphere, above. Unlike on every other world we
have encountered so far, the troposphere is not bounded by a tem-
perature inversion (the tropopause), simply an increase in wind
shear. Triton is, therefore, unique in this regard. The thermosphere
is lightly warmed by the steady impact of charged particles from
the Sun and from Neptunes magnetosphere, which whips across
it every 16.1 h (Chap. 8).
Altitude
Altitude
(km)
(km)
Exosphere
1,000 Exobase 1,000
800 Cosmic rays heat 800

-182oC thermosphere 40o higher
than surface
600 600
Thermosphere
400 400
(microbar)
Pressure
200
Thermosphere
E 200
E
0.1 W
W W Hazes up to 83 kilometers E
1
8
E 10
10
-237oC Haze Troposphere Troposphere
90 60 30 0 -30 -60 -90 90 60 30 0 -30 -60 -90
Latitude Latitude
FIG. 9.6 A comparison of the atmospheres of sister worlds Triton

(left) and Pluto (right). Although superficially similar, there are differ-
ences. Tritons troposphere is the lowest 8 km, above which wind sheer
increases because of weak easterly or westerly jet streams. Pluto, on the
other hand has a troposphere capped by a strong temperature inversion.
The troposphere on both worlds has virtually no wind due to low tem-
perature gradients and friction. Plutos exobase (where the thermosphere
ends) may be higher than 15,400 km from the surfacehigh enough that
Charon may orbit within it
High above Tritons surface the atmosphere is complete

when the thermosphere gives way to the exosphere at 950 km.
At this point molecules, atoms and ions from the tenuous atmo-
sphere below can escape into Neptunes magnetosphere. There is
no aurora marking the upper boundary, just the whisper of gases
making their way into the void.
Whats the Weather Like?
Obviously it is cold, with largely clear skies that gives a virtu-

ally unimpeded view of the cosmos beyond. There are no clouds
to speak of, only the barest of hazes that permeates most of the
troposphere (Figs. 9.5 and 9.6). These hazes are thought to be com-
posed of nitrilescompounds we encountered at Titan. These
are compounds of carbon and nitrogen that are produced when
Haze Layer
Wind streaks
FIG. 9.7 Global winds on Triton above 8 km. Thankfully, with such a
tenuous atmosphere winds are very light and follow a simple pattern.
There is virtually no wind at the surface because the air is so thin and
the temperature gradient across the surface, and vertically, is limited.
However, above 8 km winds increase, with speeds up to 15 km per hour
at an altitude of 10 km. These blow from an easterly direction across the
equatorial and mid-latitude regions. This is evidenced by plumes of dark
material that streak to the west at 53 and 57 south (see inset image).
Over the more polar regions there are gentle westerly winds blowing at
roughly 7 km per hour. Given that at this altitude Tritons atmospheric
pressure is roughly one millionth that at the Earths surface, you wouldnt
feel this gentle breeze blowing at all. Plutos atmosphere is warmer but
expected to be similar. Triton image: Voyager 2/NASA
ultraviolet light and cosmic rays causes the trace methane and
nitrogen to react. Elsewhere there are small, localized clouds of
nitrogen ice between 1 and 3 km above its surface (Fig. 9.7).
Aside from this the only real excitement comes from occa-
sional eruptions as nitrogen gas from under Tritons surface. The
origin of these is unclear, but it is believed to result from a surface
greenhouse effect. Here nitrogen is largely frozen into a broad series
of solidified pools or seas. Sunlight can penetrate these and warm
them from the base upwards. This, with limited geothermal heat,
causes the nitrogen to melt. Normally, the nitrogen would simply
sublimate, but if melting occurs at sufficient depth, pressure can
rise until the warming solid turns into a liquid. Continued warm-
ing raises the pressure until some of this evaporates and is explo-
sively released through the icy shell. Jetting through the lower
atmosphere, these nitrogen geysers will shower out a curtain of
nitrogen snow, downwind of the jet. Entrained within these jets
are any other solid materials and these produce the dark streaks
that have proved so useful in tracking Tritons winds.
At present this is all we can say about Triton. The only real
data was recorded during the Voyager 2 flyby or through stellar
occultation, where Triton passes in from of a distant star. Then,
Earth-bound scientists can grab a quick look at a portion of
Tritons atmosphere as it interferes with the light from the distant
orb. More complete data will require another fly by, or perhaps the
insertion of a Neptunian orbiter, forming a more frigid analogue of
Cassini. As there are no plans to do this at the time of writing, the
forecasters will have to content themselves with the wispy data
that exists, or resort to models.
Pluto
Introduction
New Horizons visit to Tritons near-twin, Pluto, has provided a lot

of clues about these distant, icy worlds. It is to Pluto and Charon
that we now turn to provide an unexpectedly lively end to this
chapter. Both worlds are small, icy orbs that pinwheel around their
common center of gravity. Both are likely to have formed in a man-
ner analogous to the Earth and Moon, with a giant impact gener-
ating a large debris cloud around the former Pluto. However, the
original Pluto was shattered in the impact and the debris scattered
with enough momentum to allow the formation of a binary dwarf

planet. Although Pluto has the lions share of the pairs mass,
Charon remains the most massive satellite, relative to its parent
world, of any partnership in the solar system.
Both worlds have a center of mass lying between them that
means that they are tidally locked to one another. Initially, this
arrangement would have ensured that a substantial amount of
heating would have occurred within each early on. However, 4.5
billion years later, most of this energy should have escaped such
small worlds. A small amount of radiogenic heat: energy released
by the slow decay of long-lived radioactive isotopes of elements
such as potassium and uranium. Rough calculations suggest that
around 2.4 milliwatts per meter squared (or around 1/40th of that
at the surface of the Earth) will currently be available and this
might be enough to maintain a partly liquid interior. Now, by
liquid we are not talking terrestrial peridotitethe rock domi-
nating the Earths upper mantle. Rather, we are talking nitrogen
or ammonia ices; materials that will melt under very low tem-
peratures. Even water ice is as hard as steel at the temperatures
of Pluto. However, because Pluto may retain a partly molten (or
at least soft ice) interior it will be able to convect and transport
gases into the atmosphere above, much as volcanoes do on Earth.
Indeed, the first images of Pluto appear to confirm this.
In terms of seasons, Pluto is an extreme version of Mars. The
dwarf planet and Charon are tilted over at 119.5 with the North
Pole pointing downwards (as defined by the direction of spin).
And while the Martian orbit varies from 1.4 to 1.7 AU, that of
Pluto changes radically from 30 to 50 AU. At its closest approach
to the Sun in 1990, the Sun shone down most strongly over the
Plutonian equator. With Pluto making a slow retreat to its most
distant position (aphelion) 99 years from now, the southern hemi-
sphere is pointing more strongly away from the Sun and cooling
down. Within the next 20 years the southern polar cap should
become so cold that most of the neighboring atmosphere will con-
dense out onto it. Meanwhile, albeit in a less extreme manner, the
northern cap will experience its summer and begin to thaw out.
As this will occur when Pluto is furthest from the Sun, the release
of gas from sublimation will still not compensate for the rate of
loss onto the dwarfs surface. It will only be when the southern
cap experiences spring once more (after the year 2114) that the
overall mass of the atmosphere will begin to rise again. Here, the
combination of greater insolation from the tilt and Plutos closer
position to the Sun ensure the planet can begin to warm up and,
at least partially, thaw out.
Pluto and Triton as Non-identical Twins
Before we delve into the small amount of data that currently exists
for Pluto (soon to increase further from New Horizons) lets look
again at Pluto and Triton together. Pluto has a diameter roughly
the same as that of Triton (2370 versus 2700 km, respectively) and
both were expected to host similar geological features when New
Horizons arrived at Pluto in the summer of 2015. Despite areas of
similarly young terrain, Pluto was still a considerable surprise to
observers. Not only did its surface play host to some utterly unique
features, such as the vast, flat Sputnik Planum (Figs. 9.8 and 9.10), it
also had a dynamic multi-layered atmosphere (Fig. 9.8) that defied
both expectation and observations made from the Earth. At the
time of writing much of this information remains preliminary but
FIG. 9.8 A view of Pluto from New Horizons, 11,000 km from its surface.
The edge of Sputnik Planum is visible as the flat area near the center of
the image. Norgay Montes is the highest peak to its west. The moun-
tains appear oddly flattened which might indicate erosion, subsidence
or perhaps cloud cover? Ice appears to be flowing onto Sputnik Planum
from the mountains. Extending over 100 km above the surface are layers
of hazes. These appear to reach down into the valleys nestling within the
mountains where they form fog-like banks lit by crepuscular rays. New
Horizons has confirmed that these layers are produced by atmospheric
gravity waves (Chap. 10) as air flows over the mountains
FIG. 9.9 The Solar Systems Final Pale Blue Dot. Hazes extend over
80 km from the surfacemuch higher than expected. Sunlight refracts
through the atmosphere, over the upper right edge, partly illuminating
the night side and giving rise to twilight. Haze layers are far more com-
plex than was suspected from Earth-bound observations. Right: Predicted
wind speeds 94 km above the surfaceand above the haze layers. Little
atmospheric motion is predicted in the lowest 50 km except the barest
of breezes caused by the slow sublimation and re-sublimation of surface
frosts as sun rises and sets. Pluto image: New Horizons/JPL/NASA
Smooth
terrain
Smooth
H
terrain
H H H
H H
H H
H H
H
H
Rough (older?) D
and cratered H
terrain Rough, highly cratered (and older?)
H
terrain
FIG. 9.10 The Surface of Pluto is dynamic. The large and relatively flat area
is called Sputnik Planum and appears to be a frozen sea, possibly of nitro-
gen ice. This abuts a region of more highly cratered, and presumably older
terrain along its southern edge (left). In a close-up (right) of part of this mar-
gin there is folded darker terrain with ripples (D) aligned broadly parallel
to the edge of hummocky terrain (outlined in red). This latter terrain lies
all along the margin of the two different regions. On Earth Hummocky ter-
rain is formed during large landslips, such as that at Mt St Helens during
its 1980 eruption. This suggests the margins of the older terrain are foun-
dering into the flooded basin (red arrows). Original, unmodified, images:
New Horizons, NASA. Image modifications by author
what is known and conjectured is discussed below. Like Triton,

Pluto shows considerable evidence of resurfacing and this may
relate in each case (at least in part, or initially) to energy released
from tidal interactions with its neighbor: Neptune in the case of
Triton; Charon in the case of Pluto. Both worlds also have suffi-
cient internal reserves of radioactive materials to keep them just
about ticking.
Looking first at the surface, superficially Pluto is rather like
Mars, with some important differences that are only apparent on
closer inspection. For one, the pink color of Pluto is reminiscent
of the rusty color of Mars. However, while Mars has a fairly old,
cratered terrain covered in iron oxides and iron-rich basalts, Pluto
has a predominantly water (and other)-ice covered surface, col-
ored pink by compounds suspected to be organic tholins. Spectra
appear to suggest that most of the methane2 ice is concentrated in
Sputnik planum and other low-lying regions, while highland areas
appear to be water ice, but largely covered in other material.
One of the first surprises was that the atmospheric pressure
was far lower than was measured from Earth. Whereas terres-
trial measurement, using radio waves and observations of stellar
occultation suggested a pressure of 22 microbars (22 millionths the
pressure at the Earths surface), measurements by New Horizons
were around 5 microbars at the surface. Why the difference? One
suggestion is that the terrestrial measurements are taken for
greater altitudes (around 5075 km) than those made by the pass-
ing probe. The latter were made directly by observing how radio
waves that were sent from Earth refracted as they passed through
the tenuous Plutonian atmosphere to reach the New Horizons
probe. These measurements were made at or just above the dwarf
planets surface. However, this discrepancy remains only partly
explained at the time of writing .
One of the most intriguing terrains is Sputnik Planum
(Fig. 9.10), a vast frozen and very flat and broadly circular basin
with an extension that makes the whole feature look rather like a
heart. Its broadly pink coloration only added to the analogy. The
south western edge of the basin is semi-circular, implying that it
is an impact feature that has been flooded by some lava. New
2
At the time of going to press, it is thought that this methane layer is only a thin covering over what
is mostly convecting nitrogen ice.
Horizons has determined that the ice in the basin is methaneat

least on its surface. Elsewhere the crust in the highland regions is
probably water ice but it is covered almost completely by another
material that is unidentified at the time of writing. However,
water ice is visible along some of the cracks in Viking Terra. The
basin edge also appears to have a broadly consistent geology along
its entire edge. This is where my interpretation comes in. The
elevated highland region (largely) uniformly gives way to a lighter
colored smoother area than then morphs into a hummocky region,
adjoining the edge of the flat basin. Some of the nearby, smaller
craters are also flooded, perhaps to the same depth (or altitude)
as the neighboring basin. This may imply a water-table exists
across the region with whatever frozen liquid that is present lying
at a consistent depth across and through the entire terrain.
So, what of the hummocky terrain: how might one interpret
that? One simple explanation that may or may not be true, but
that is consistent with terrestrial geology is that the basin mar-
gins have either been undermined by liquid and collapsed directly,
forming the hummocks, or the material is utterly saturated at a
particular depth and this has weakened it so that it has then col-
lapsed. Close examination of the region shows lakes embedded
between some blocks implying the region is thoroughly infiltrated
with liquids at least at depth.
On Earth you see the same hummocky terrain in places
such as land adjoining Mt Shasta in California or Mt St Helens in
Washington State, as well as numerous others. Formerly mysteri-
ous in origin, nowadays, hummocky terrain is thought to form
during massive collapses of mountainous structures. Nearest the
source the landscape may be relatively smooth, while at different
distances away the landscape is pockmarked by hillocks embed-
ded in a flatter surround. These hillocks are large blocks of the for-
mer mountain that have slid downwards and become jumbled. In
the case of Pluto there are two other suggestive features of regional
collapse: one, an area suggested to be dunes, while the other is sim-
ply the pervasive nature of the hummocky terrain that flanks the
entire high basin margin and is consistent with collapse (Fig. 9.10).
The supposed dune field may be a misnomer. The dunes
are aligned with the crater wall and the area broadens towards the
basin and more distant hummocky terrain. The pattern is again
reminiscent of collapse but in this case a more sluggish one where

the land has crept downwards and fanned out as it came to rest.
In this interpretation the dunes are simply folds in the material
caused by its movement over the underlying landscape. Such folds
form on terrestrial hillsides where the soil is moist and slips over
the underlying bedrock. This suggests collapse is ongoing but per-
haps now is more restrained in scope than the events that formed
the hummocky terrain. The hummocky terrain and "dune-field"
would happen if the subsurface was strongly undermined by flu-
ids. On Earth, this is almost always water; on Pluto it is likely to
be nitrogen. As these infiltrate the fractured bedrock lining the
impact basin, the walls become unstable and collapse catastrophi-
cally into the basin floor. Either simultaneous with this, or subse-
quently, the soft methane and nitrogen ice then flows around the
margins of the blocks, softening their edges.
Within the basin the flat icy terrain has an unusual blobby-
kind of appearance. Its possible that this pattern reflects regions
of upwelling and down-welling within the icy material (or in liq-
uid or slush) underlying the icy surface.3 Thus, Sputnik Planum
may be similar to the frozen surface of the Arctic Ocean. However,
unlike the Arctic Ocean, the depth is likely relatively shallow and
the contents are probably at least partly frozen to the base allow-
ing the pattern of upwelling or down-welling to become apparent
(Fig. 9.10).
Why is a frozen (or partly frozen) ocean under Sputnik Planum
important for Plutos weather? It suggests ubiquitous fluids, and at
the pressures within Plutos atmosphere, nitrogen, carbon monox-
ide and methane can vaporize forming the observed atmosphere
(Figs. 9.8 and 9.9). With an atmosphere, no matter how thin, we
have weather. What weather there is probably reminiscent to that
found in Antarcticas driest valleys. Here, katabatic winds and
sunlight cause ice (in this case water ice) to sublimate. Now free
within the atmosphere, the prevailing westerly winds carry the
moisture to neighboring areas, where it either condenses as frost
or re-joins the general mle as atmospheric moisture or ice grains.
On Pluto the ice is nitrogen, with some methane and carbon
3
At the time of going to press this has been conrmed and convection within a layer 3 km thick
generates the observed pattern in Sputnik Planum.
Hazes
B
C Fog
A D
E
Sputnik Planum
FIG. 9.11 The methane/nitrogen-cycle on Pluto. Not to be confused

with the biological nitrogen cycle on Earth, here nitrogen and methane
gas, liquid and solid cycle from atmosphere to surface and back again in
complex ways. At Sputnik Planum (A) nitrogen and methane ponds in a
large icy layer, in which heat from the interior or solar energy may liq-
uefy underneath. Nitrogen and methane sublimate from here and enters
the atmosphere, along with carbon monoxide. Light, prevailing westerly
winds at this latitude (Fig. 9.8) blow these gases towards the 3,000 m high
mountains (B). Here, some condense as freezing fog and frost, or perhaps
on occasion when Pluto is further from the Sun, form clouds, precipitat-
ing snow (C). Such snow either compacts or partly melts and flows back
to Sputnik Planum (D) in glacier-like flows. These may be liquid deeper
down where geothermal heat melts the ice without sublimation. At E
liquid nitrogen and methane underlying the western basin rim and desta-
bilizes it leading to collapses (Fig. 9.9)
monoxide. It too sublimates under the action of sunlight. Carried

by the prevailing, light westerly winds, this gas ultimately reaches
the highlands to the east of Sputnik Planum. Here, the gases re-
sublimate as frosts or perhaps (at least on occasion) precipitate as
nitrogen and/or methane snow. After building to sufficient depth,
these ices either flow directly back into the basin or partly melt
at their base, with a mixture of ice and bedrock liquid nitrogen
and methane (held as a liquid under pressure) flowing back into
Sputnik Planum (Fig. 9.11). Some of these flows strongly resemble
terrestrial water ice glaciers, suggesting a cycling of methane ice,
perhaps as snow, from basin to mountains and back again. In the
basin such liquids form the flat layer, which may undermine the
western wall of the basin leading to its periodic collapse.
Pluto, overall, should be rather similar in terms of weather to

Tritonat least for part of its 248 Earth-year-long year, since when
it is closest to the Sun, it receives roughly the same amount of
solar energy as Triton. However, Pluto has a more elongated orbit
around the Sun, which ultimately takes it far beyond the orbit of
Neptune. 120 or so years from now temperatures should be low
enough for the entire atmosphere to cool to the point at which
it collapses onto the dwarfs surface. Some 200 years hence, the
process will repeat, with the Sun vaporizing enough of the dwarfs
surface to regenerate an atmosphere.
Initially, this first atmosphere might be dotted with clouds
of nitrogen, methane or carbon monoxide ice. These might erupt
from Plutos frozen surface, much like the geysers seen on Triton.
As more ice sublimates and pressures rise, solar ultraviolet light
and cosmic rays will be able to act on these gases and produce
complex tholins. These materials and related nitriles will then
form hazes at different layers within the atmosphere. Some of
these precipitate out in a manner analogous to terrestrial frost,
coloring the landscape pink. As the density of methane and carbon
monoxide rises, their property as greenhouse gases will come into
play, trapping more outgoing radiation and raising the temperature
of Pluto even higher. This should reach equilibrium a few years
into the thaw with a balance between trapping of energy, warming
and further sublimation of ices.
Even then, the low gravity of Pluto is insufficient to retain
most of its newly constructed atmosphere. As pressure and tem-
perature rise, more and more of the atmosphere reaches the escape
velocity of the dwarf world. New Horizons detected a cometary
tail of material stretching down (solar) wind of the world, suggest-
ing that a combination of Jeans escape and solar wind stripping is
sufficient to remove much of the atmosphere that is vented from
Plutos warmed surface. Interestingly, some of this material appears
to be raining out on its sister-world, Charon. Charon wears a red
cap over its northern Pole. It is suggested that this was formed (and
continues to form) when gases from Pluto that are rich in tholins,
spew across the face of Charon. Only at the poles are conditions
cold enough for the tholins and other materials to condense out
forming a thin veneer, much like the icy cap of Earth or Mars.
Conclusions
Pluto, Triton and Titan form an intriguing triad of dwarf planets.
While we think of Titan (quite rightly) as Saturns greatest moon,
it has rather a lot in common with further out Triton and Pluto.
All three worlds have nitrogen-dominated atmospheres, the only
difference being the temperatures involved, which then causes
substantial differences in pressure. If one could shift Titan out to
where Triton and Pluto reside, in the twilight limb of our solar
system, most of its methane and nitrogen atmosphere would rain
then snow out onto its surface, leaving a wispy ghost of its former
self. Conversely, move the frigid pair inwards and much of their
bulk would vaporize forming (at least briefly) a thick atmosphere.
Briefly, because their low gravityless than half that of Titan
would be unable to retain such a thick firmament for long. Most
would escape through the Jeans mechanismand ultimately a
much denser hydrodynamic flow (Chaps. 5 and 6) over the course
of a few hundred million years. What would remain would be an
icy lump more akin to Europa or Callisto, than the effervescent,
but still alive worlds we see today.
Ultimately, none of these worlds will survive in their present
form. None retain gases strongly and all will be depleted in mass
before the Sun leaves the main sequence. Their end will come
when the Sun becomes a red giant. Then, temperatures will rise
far beyond the melting point of water at Saturn and easily above
temperatures sufficient to sublimate gases out by Neptune. By the
time the Sun ceases to be a red giant only their rocky cores will
remain. Triton has an even grizzlier fate in store. In a few billion
years, its orbit will decay and send it crashing towards Neptune.
Tritons loss will be Neptunes gain: a new set of icy rings that will
slowly disperse under the increasing glare of the dying Sun.
References
1. Tokano, T. (2009). The dynamics of Titans troposphere. Philosophical Transactions A, 367,
1889. Retrieved from http://rsta.royalsocietypublishing.org/content/roypta/367/1889/633.full.pdf.
2. Burr, D. M, Bridges, N. T., Marshall, J. R., Smith, J. K., White, B. R. & Emery, J. P. (2014).
Higher-than-predicted saltation threshold wind speeds on Titan. Nature, 517, 6063.
3. Teanby, N. A., Irwin, P. G. J., Nixon, C. A., de Kok, R., Vinatier, S., Coustenis, A., Sefton-
Nash, E., Calcutt, S. B.& Flasar, M. F. (2012). Active upper-atmosphere chemistry and dynam-
ics from polar circulation reversal on Titan. Nature, 491, 732735.
4. Fischer, G., & Gurnett, D. A. (2011). The search for Titan lightning radio emissions. Geophysical
Research Letters, 38, 8. Retrieved from http://onlinelibrary.wiley.com/
doi/10.1029/2011GL047316/pdf .
5. de Kok, R. J., Teanby, N. A., Maltagliati, L., Irwin, P. G. J. & Vinatier, S. (2014). HCN ice in
Titans high-altitude southern polar cloud. Nature, 514, 6567.
6. Tokano, T., McKay, C. P., Neubauer, F. M., Atreya, S. K Ferri, F., Fulchignoni, M & Niemann,
H. B. (2006). Methane drizzle on Titan. Nature, 442, 432435.
7. Hueso, T. R. & Snchez-Lavega, A. (2006). Methane storms on Saturns moon. Nature 442,
428431.
8. Lorenz, R. D., Wall, S., Radebaugh, J. Boubin, G., Reffet, E., Janssen, M. et al. (2006). The
sand seas of Titan: Cassini RADAR observations of longitudinal dunes. West Science, 312 no.
5774, 724727.
9. Charnay, B., Barth, E., Rafkin, S., Narteau, C., Lebonnois, S., Rodriguez, S. et al. (2015).
Methane storms as a driver of Titans dune orientation. Nature Geoscience 8, 362366.
10. Zalucha, A. M., Michaelsa, T. I. (2013). A 3D general circulation model for Pluto and Triton
with xed volatile abundance and simplied surface forcing. Retrieved from http://arxiv.org/
pdf/1211.0009v2.pdf.
11. Schneider, T., Graves, S. D. B., Schaller, E. l., & Brown, M. E. (2012). Polar methane accumu-
lation and rainstorms on Titan from simulations of the methane cycle Nature, 481, 5861.
12. Erwin, J. T., Tucker, O. J., & Johnson R. E. (2012). Hybrid uid/kinetic modeling of Plutos
escaping atmosphere. Retrieved from http://arxiv.org/pdf/1211.3994v2.pdf.
13. Gilliam, A. E., Lerman, A. & Wunsc, J. (2015). Evolution of titans atmosphere in relation to
its surface and interior. Retrieved from http://www.hou.usra.edu/meetings/abscicon2015/
pdf/7772.pdf.
14. Sotin, C., Lawrence, K. J., Reinhardt, B., Barnes, J. W., Brown, R. H., Hayes, A. G. et al.
(2012). Observations of Titans northern lakes at 5 microns: Implications for the organic cycle
and geology. Icarus 221, 768786. Retrieved from http://c3po.barnesos.net/publications/
papers/2012.11.Icarus.Sotin.Northern.Lakes.pdf.
15. Johnson, R.E., Tucker, O. J., Michael, M., Sittler, E. C., Smith, H. T., Young, D. T. et al. Mass
loss processes in Titans upper atmosphere. Retrieved from http://people.virginia.edu/~rej/
papers09/TitanChap15MassLossJohnson09.pdf.
10. Tales of Other Worlds
Introduction
Since 1992 astronomers and planetary scientists have been able
to feast upon a bewildering variety of planets that lie outside our
parochial system of eight, and a bit, worlds. We now have plan-
etary systems orbiting white dwarfs, red giants, neutron stars, as
well as a bewildering array of more mundane main sequence stars.
Kepler doubled the size of the planetary menagerie in the few
years it was running at top speed. Even now, with reduced stabil-
ity, astronomers have managed to keep it throwing out new plan-
etary landscapes, by using the solar wind to keep it pointing in a
fixed direction
In turn this array of planets has begun to allow planetary scien-
tists to begin testing ideas about how atmospheres workoften in
extreme environments. This chapter explores what is known and
what is yet only imagined. It will be apparent where the now rou-
tine topics such as Monsoons, Rossby Waves and the ubiquitous jet
stream guide our thoughts as they guide our weather. It will also
become apparent where there are some surprising contradictions
in models that affect how habitable a planet might be. In turn this
reflects on how we view the running and habitability of our world.
Exoplanets by the Bucket

Exoplanets fall into a number of different categories based upon
their size and location. In general we can recognize planets with
masses in the range of several Jupiters, through Neptune-like plan-
ets to Earth and Mars-sized orbs. Many of these worlds are locked
in star-grazing orbits that toast their surfaces and atmospheres to
hundreds or thousands of degrees Celsius.
In general giant planets are confined to more massive stars
with only one example known to orbit the universes most

DOI 10.1007/978-3-319-25679-5_10
abundant stars, the red dwarfs. Red dwarf stars seem to favor plan-
ets with masses a few times that of the Earth. Such super-Earths,
or super-terrans, may be largely rocky worlds or ones whose entire
surface is submerged in a global oceans or a deep atmosphere. At
present we cannot discern their true nature, but the presence of
such worlds within the habitable zones of their stars suggests that
the universe may favor life on such worlds.
One of the first discoveries, and the first planet known to
orbit a Sun-like star, was 51 Pegasus b. This by and large set the
tone for many of the subsequent finds: a Jupiter-like world in a
tight, hot orbit around its star. Such planets are clearly unknown
in our solar system and suggested that our orderly view of the uni-
verse needed some refining. Instead of planets forming in situ they
are prone to migration inwards or outwards. Indeed, this theme
was subsequently applied to our seemingly orderly solar system:
the so-called grand track hypothesis, where Jupiter and Saturn
have migrated great distances during their formative years, and
consequently grossly modified the layout and subsequent evolu-
tion of the solar system as a whole.
At the time of writing over 1890 worlds are known, with
over 4690 other worlds listed as candidates. Of these, over 470
fall into multi-planet systems: planetary systems with more than
one planet orbiting its star (or in a few occasions, stars).1 These
fall into some broad and interesting categories. Of those that have
a known mass (and these were primarily found by radial velocity
or transit-timing methods), the majority are Jupiter-class worlds.
However, when we look at planetary radiuswhere planets are
found by transits of their host stars, the vast majority are so-called
super-Earths (or super-terrans) with radii (and probably masses) a
few times that of our world. Next most common are Neptune-
class (or ice giant) worlds.
The difference in numbers of each type of planet that is found
is down to the manner of how the planets are detected. The radial
velocity method is best-suited to find massive planets in short
period orbitsso-called hot Jupiters and hot-Neptunes. While the
transit and transit-timing methods only require that the planet
crosses the face of its star. The latter method, therefore has less
1
Updated lists can be found at: http://exoplanetarchive.ipac.caltech.edu/docs/counts_detail.html.
Tales of Other Worlds 365
bias but is clearly limited to the few percent of star that have
orbits suitably aligned with ours so that we can observe a transit.
Amongst these largely biased searches there are still some
apparent trends. Red dwarfs rarely host Jupiter-class worlds, favor-
ing super-terrans or warm to hot-Neptune class worlds. This is
probably a function of the mass of the star and planet forming
cloud, which is typically around one tenth that thought to have
spawned our solar system. As the mass of the star increases so
does the likelihood of hosting planets: only 3.5 % of red dwarfs
host planets while 14 % of A-class, Sirius-like, stars do. Moreover,
planets forming around stars with low metal contents also tend to
have fewer massive Jupiter-class planets.
Of the planets we are interested in throughout this chapter
there are a few types that stand out. First are the ubiquitous hot
Jupiter and Neptune-like worlds that whip around their host stars
in a matter of hours or a few days. All of these are expected to be
tidally-locked to their host star, meaning that they always present
one face to their star. The next most interesting are similarly short-
period planets, but those which orbit less luminous red dwarf stars
and thus could be habitable. Finally, there will be planets orbiting
less common, but still ubiquitous orange K-class stars. These plan-
ets are likely to rotate relatively slowly as they experience strong
tidal forces, but not necessarily strong enough to cause them to lock.
It is these three classes of world that will be the focus of this chapter.
The Climate of Tidally-Locked Planets:

Assumptions and Expectations
Any planet that lies sufficiently close to its host star will lock
to it. A young planet will experience tides within its hot, plastic
mantle (and within any atmosphere and ocean it has). These tides
are communicated between the planet and its star at the speed
of light, which, being finite, takes time. Imagine then the planet
orbiting its star but rotating niftily on its axis. As the star pulls on
the planet it generates a tidal bulge. Given that light has a finite
speed, the bulge will always be pointing towards the star in slightly
the wrong direction because while the bulge is pulled one way, the
rotation of the planet will carry it forward: the bulge will lead the
pull of the star. With this happening, the star will pull on this in
a direction pointing slightly backwards. This pulls on the bulge,
slowing down its motion around the polar axis of the planet. After
anything between a few tens of millions of years to a few hundred
million years, the planet will have had its rotation slowed until
one face permanently faces its host star. Now, the planet is still
rotating on its axis, but now only so much as to rotate once per
orbit. Like our Moon, the planet may librate, or appear to wobble,
due to an eccentric orbit. It is thought that most tidally-locked
worlds will have roughly circular orbits thanks to tidal locking.
What does this mean for its climate? Without going into
detail just now, you would expect one side to be very warm,
being permanently illuminated, while the opposing hemisphere,
being permanently dark, was cold. The presence of even a Mars-
like atmosphere will modify temperatures substantially. In 1998
Martin Heath and colleagues made some initial models which
explored the effect of adding an atmosphere to the temperatures
of planets orbiting red dwarf stars; and the subsequent impact on
habitability. In the habitable zone, a planet without an atmosphere
has a day side with temperatures approaching the boiling point of
water. Meanwhile, temperatures on the dark hemisphere languish
at less than 120 C: pretty much what you find on the Moon. Add
a Martian quota of atmosphere and the temperatures on the day-
side fall to 50 C, while the night-side warms to 70 C. Add still
more gas and the dayside maximum falls to a tropical 35 C and
the night-side rises to 20 C. Finally, in more recent analysis by
Yongyun Hu and colleagues, once you start modifying the amount
of carbon dioxide and other greenhouse gases, the temperatures
even out event more (as well as rise, overall).
The Structure of the Atmosphere

of Jupiter-Like Worlds: Too Hot,
Too Cold, or Just-Right
Before we look at the climate of tidally-locked planets we need to
see how changing the temperature affects what gases are actually
present in their atmosphere. Figure 10.1 shows the likely composi-
tion of clouds in the atmospheres of a planet with the composition
TiO/VO
CH4 Stratosphere CH4 Stratosphere CO Stratosphere

-150oC 300oC 1,000oC
Haze Tropopause
Tropopause Tropopause Magnesium Silicate
NH3
Fe Liquid
NH4SH H2O
CH4 KCl
CsCl
LiF
RbCl
KCl Na2S
Troposphere LiF
Magnesium Silicate
Na2S Perovskite
CO Magnesium Silicate CO Fe Liquid CO
Fe Liquid Perovskite
Perovskite Troposphere Troposphere

Jupiter Warm Jupiter Hot Jupiter
FIG. 10.1 Clouds and layers in the atmospheres of Jupiter-like planets.

Jupiter, left, has visible clouds of ammonia (NH3), Ammonium hydrogen
sulfide (NH4SH) and likely water (H2O). Beneath this at higher tempera-
ture and pressure these substances are gaseous but more refractory com-
pounds (those with higher boiling points) can condense to form clouds.
These include Lithium fluoride (LiF); cesium chloride (CsCl); potassium
chloride (KCl); sodium sulfide (Na2S). At still deeper levels magnesium
silicates and iron metal become liquids and gases. At the deepest levels,
with the highest pressure a compound called perovskite is found as a gas.
On Earth perovskite is only found beneath 1000 km of rock. As atmo-
spheric temperatures rise (center and right) it becomes too hot at all lev-
els for ammonia and water, and these are replaced by clouds from deeper
down. Methane (CH4) gives way to carbon monoxide (CO). This process
continues as temperatures rise until it is hot enough for silicate clouds at
the highest levels. Titanium and vanadium oxides may form clouds high
in the atmospheres of the hottest Jupiter-like planets
of Jupiter, but heated to different extents by its parent star. Our

cold Jupiter has icy clouds of ammonia and ammonium hydrogen
sulfide, with water expected deeper down below the 5 bar level.
Convention puts the upper lid on the troposphere (the tropopause)
at 0.1 bars of pressure (Chap. 7). In the case of Jupiter this has a tem-
perature of around 150 C and this is the condensation point for
ammonia at the pressures found there. If you lower the temperature
to 220 C, methane will form ice crystal clouds at this pressure, as
is seen in the atmospheres of Uranus and Neptune (Chap. 8).
Conversely, raise the temperature to around 300 C and now

it is too hot for ammonia, water or ammonium hydrogen sulfide.
Instead dusty clouds of halide compounds such as potassium or
cesium chloride are formed. Hotter still and clouds of silicate
dust and molten droplets of iron are found. Its not that water isnt
found at these temperatures, rather that it is just too hot for it
to condense to form clouds. So, in the first instance the types of
clouds are determined by the available temperatures. Indeed, if
you wanted to look for silicate vapor on Jupiter (but probably not
clouds, per se) youd have to drop down well below the 10 bar level
until temperatures were at the appropriate temperature for these
compounds to condense.
Thats not quite all. In the cooler Jupiters the background
gas (other than hydrogen and helium) is methane. This is because
methane is only stable, chemically, at temperatures less than
1000 C. As temperatures rise, methane will react with a variety
of other oxygen containing compounds to form carbon monoxide.
Chemical equilibria: the balance between different chemical com-
pounds in a mixture of gases changes with pressure and tempera-
ture. In practice this means that methane picks up oxygen from
silicates, water or other oxygen-containing compounds. This oxi-
dizes the methane to carbon monoxide, while releasing the hydro-
gen. One of the benefactors of these reactions is oxides of iron.
These chemically react with the hydrogen are then chemically
reduced to iron metal. Therefore, in very hot Jupiters, methane
reacts with available iron oxides to form carbon monoxide and
liquid iron, which then rains out into the planetary interior. Hot
Jupiters also show sodium, potassium and other metal halides
such as sodium chloride in their upper atmosphere. The firs detec-
tion of this was made by Sarah Seagers group in the atmosphere of
HD 209458b (Fig. 10.4) confirming the basic model.
One notable difference to this pattern of change was the
detection of titanium and vanadium oxide clouds in the upper
atmosphere of HD 209458b. This compound, common in the
lower atmosphere of red dwarf stars, only appears at temperatures
between 20002500 C implying that the upper atmosphere is
not only hot enough to host this chemical, but dense enough so
that these compounds can condense in the first place. This clearly
implies a strong temperature inversion in the middle atmosphere of
this giant planet. These surprises aside, the general picture, which
is emerging, matches expectations. Clouds of different composi-
tions will form depending on the overall temperature and pressure.
If we increase the mass of these objects eventually we meet
the brown dwarfs at 13 Jupiter masses and red dwarfs at around
75 Jupiter masses. Brown dwarfs make an interesting case and,
yes, weather, has been observed on some of these by examining
changes in the intensity and wavelength of radiation emitted.
Such objects are important test-beds for understanding the overall
structure of Jupiter-like worlds. All of these begin life hot. The only
real difference between an infant Jupiter and a brown dwarf is the
strength of its gravitational field. Both are born as hot, incandes-
cent objects with broadly the same diameter. The increase in mass
is broadly off-set by greater gravitational compaction. It is only
when the core becomes hot enough to ignite does the diameter of
the object increase again. Jupiter would cool more rapidly, having
less internal energy to spare than a brown dwarf. The lower grav-
ity in Jupiter primarily serves to puff-up its atmosphere, relative
to a typical brown dwarf. The temperature gradient is somewhat
less steep as a result, but otherwise Jupiter makes a pretty good
model for an old and cold brown dwarf. Consequently, a young
brown dwarf will have clouds of metal oxides (primarily titanium
and vanadium oxide). As the brown dwarf cools it will go through
the same series of changes we expect for Jupiter-mass planets at
increasing distances from their host star. Clouds of oxides will
retreat deeper inside the brown dwarf and become replaced by
metal halide clouds. Over time the carbon monoxide gas, form-
ing the background, will morph into methane as temperatures fall
below 1000 C. Such brown dwarfs are called T-class objects in an
extension of the Hertzsprung-Russell diagram that is used to dis-
play stellar classification.
Finally, as temperatures fall lower, first water clouds, then
ammonia ice clouds will form the uppermost cloud decks and the
object will take on the veneer of our Jupiter and Saturn. Indeed,
some very cool brown dwarfs have been found. The coolest brown
dwarf (WISE J085510.83-071442.5 (W0855)) is colder than Mars
with a cloud-top temperature of (48 C to 13 C (225260 K).
Spectra taken and analyzed by Jacqueline Faherty (Carnegie
Institute of Washington) suggest the planet has water clouds across
half of its face. It still is far from an Earth-like atmosphere: aside

from hydrogen and helium, the most abundant gas is methane,
making it uninhabitable for humans.
Many other, hotter brown dwarfs also show indications
of clouds of silicates driftingor rather powering through their
super-charged atmospheres. Brown dwarfs form a fascinating link
between stars and planets, exhibiting weather as complex as any
planet. This is infant territory at present, but the identification of
growing numbers of nearby brown dwarfs is opening up another
window on our universes weather. The main difference between
brown dwarfs and hot (or at least lukewarm) Jupiters is that the
former are heated internally, while the latter are heated (primar-
ily) by an external source. This should lead to differences in the
dynamics of their atmospheres: brown dwarfs have convection-
driven weather, while hot Jupiters have advection-driven weather,
because they are hottest highest up.
This is an evolving picturequite literally. Imagine the Solar
System of 67 billion years hence. Jupiter will be heated to a few
hundred degrees by the expanding Sun. Aside from steadily vapor-
izing its icy moons, the upper atmosphere will become too hot
to allow first ammonia, ammonium hydrogen sulfide and finally
water clouds. Heat will slowly percolate downwards from the
upper atmosphere steadily evaporating Jupiters cloud decks. After
several tens of millions of years the upper atmosphere will play
host to clouds of metal halides rather than ammonia. Deeper lay-
ers of silicate dust might be discernible. Further out, the methane
clouds of Uranus and Neptune will also evaporate leaving deeper
clouds of ammonia, ammonium hydrogen sulfide and water: our
view of the Solar Systems meteorology will be transformed. In
essence if you want to see inside Jupiter, look at an alien hot
Jupiter and this will reveal what lurks within its deeper layers.
After the Sun expires and becomes a white dwarf, Jupiter will
slowly cool off once more. Eventually, the planet will play host to
methane clouds, before these, too freeze out and retreat to deeper,
warmer layers. Only when the planet cools below the freezing
point of hydrogen will the final changes occur. Jupiter and the
other giants will truly liquefy throughout, before solidifying into
immense icy balls. That will be a long, long way off into our uni-
verses distant future.
Some time before our planets irrevocably alter, HD 209458bs

will suffer the same fate. First its halide layer will be boiled away
revealing deeper clouds of silicates or perovskite. This will be a
brief transition before the planet as a whole is vaporized inside the
outer layers of its expanding Sun.
There is a limit to how deep you can go as below some par-
ticular depth some of the materials that could form clouds will
forever be too hot to do so. Moreover, below a critical depth the
dominant hydrogen and helium gases morph into an ocean of liq-
uid. At this point, meteorology gives way to oceanography and our
experiment with a planets atmosphere comes to an end.
What about the underlying driving force for the weather on
giant planets? Well, this is a bit more controversial. For hot Jupiters,
with temperatures above 500 C, and an energy input measured in
the tens of thousands of Watts per meter squared, it is likely that
most of the energy needed to drive their weather comes from their
parent star. Once the amount of energy falls to a few watts per
meter squared then internal energy sources begin to play a more
dominant role. On the Earth the surface heat flux released from
the interior is around 0.087 W per meter squared, which is far less
than the 240 W per meter squared it receives from the Sun (prin-
cipally in the tropics). Jupiter releases considerably more energy
from its steady contraction. This amounts to roughly 2.5 times the
14 W per meter squared captured by its planets atmosphere. Here,
the internal energy is more than sufficient to drive convection all
the way from the planets core to the top of the atmosphere.
Out at Neptune this world receives a dismal 0.7 W per square
meter and the internal energy sources again dominate, driving
convection. A hot Jupiter has an atmosphere which is hottest at
its topan extreme manifestation of our thermosphere and heat
is readily conducted downwards making the atmosphere com-
pletely stratified to considerable deptharound 100 bars or ten
times the depth of the water clouds on Jupiter. While convection
ultimately delivers energy from the interior of our giant worlds,
this is impossible in the atmosphere of a hot Jupiter where more
than 100 times the energy available from the interior is delivered
by its star. Winds are driven by the horizontal temperature con-
trasts between the day and night hemispheres, not by differences
in temperature at height.
Finally, while Jupiter spins once every 10 h or so and Neptune

once every 16 h, a tidally locked hot Jupiter will take at least twice
this long to complete one revolutionand more typically it will
take around 510 times as long. Consequently, the weaker Coriolis
Effect these planets experience has a significant impact on how
the atmospheres of hot Jupiter and hot Earth-like worlds work, as
we shall now see.
The Mystery of Super-Rotation

The atmospheres of several bodies within the solar system show
super-rotation: the movement of atmospheric gases in the same
general direction as the planet or satellite rotates, but at a faster
rate (Chaps. 5, 7 and 9). Super-rotation is a problem as how it isnt
immediately obvious how you can power up a planets winds so
that they move faster than the underlying surface that drives their
rotation around the planets axis. The solution to this problem has
come from various sources, including: observations of hot Jupiters;
experiments based in over-sized round bathtubs; and observations
of the atmosphere of our planet, particularly our tropical atmo-
sphere. Within the tropics, rather out-of-sight of ground-based
observations, are different forms of atmospheric wave. One of
these, the Kelvin Wave is particularly relevant (Chap. 1). These
Kelvin waves are the key to understanding how an otherwise slug-
gish easterly flow can be turned into a powerful westerly current
that has profound impacts on the climate of tidally-locked exo-
planets. For it is in the tropics that our atmosphere is heated most
strongly and experiences the weakest Coriolis Effect, making it
particularly analogous to the atmospheres of extrasolar worlds
where the planet is tidally-locked to its star.
Before we examine how super-rotation is achieved (at least
in some cases), we need to look a little more closely at how hot
Jupiter planets operate.
How Do the Atmospheres of Tidally-Locked Worlds Move?
One might, naively assume that if a planet is tidally-locked to

its star then air will be heated strongly on the star-facing side
and simply rise by convection. This air will then blandly flow
at altitude to the opposing, dark hemisphere where it will cool,
descend and flow back across the surface to the star-lit side. This
may be true for an unnaturally static planetone that does not
rotate on its axis. Radio and infrared observations of hot Jupiter
worlds show that winds blow strongly around the equators of
these worlds in sharp contradiction to any expectation based on
the strength of their Coriolis Effect. How is this possible? We now
look at how seemingly insignificant waves can rev up the atmo-
sphere of a planet and drive super-rotation.
Kelvin waves are convection-driven eastward moving waves
that move along the equator of the Earthand obviously any other
similar exoplanet. These waves are broadly symmetrical around
the equator and have large wavelengths of 3090 of longitude.
Kelvin waves move at around 1020 m per second somewhat
faster than the Madden-Julian Oscillation (MJO) waves that were
also discussed in Chap. 1. MJO waves can have wavelengths as
long as the planets circumference. Why mention these, aside from
their importance in transporting energy in our tropics? Well, the
odd property of these waves appears to lead, ultimately to the phe-
nomenon of super-rotation seen in hot exoplanets and probably
all planets that are tidally-locked to their stars. For while the pre-
dominant wind direction at the Earths equator is easterly, Kelvin
waves move from west to east: it is this that ultimately allows a
hot Jupiter world to experience winds that are faster around the
equator than the planet rotates.
How does this work? A few facts, first. Kelvin Waves, named
after their theorist, Lord Kelvin, are alternating regions of low and
high pressure that string along the equator. Where pressure is low
there is enhanced convection (indeed, convection causes the pres-
sure to drop). Air rising in these areas of low pressure ultimately
cools, becomes denser and then falls back to the surface forming
rear-flanking areas of higher pressure. Because of the underlying
structure of these low and high pressure areas there is more conver-
gence (coming together) of air at the eastern or leading edge of the
low pressure than along its western edge. Winds and evaporation are
stronger along the eastern edge, as well. Like the longer wavelength
MJO waves (Chaps. 1 and 2), this leads to movement of the entire
low pressure area (and rear-flanking high pressure area) towards the
region of maximum uplift, pulling the wave from west to east.
Experiments done by Dargan Frierson (University of Chicago)

confirm that it is the pattern of evaporation and wind that is the
dominant driver of these waves. All of this is important because it
sets up a pattern of west-to-east motion of mass and energy through
the planets tropical belt. On the Earth, this effect is largely, but
not completely subordinate to the overall easterly trades so that
although these waves move from west to east on monthly times-
cales, the predominant wind is still from northeast to southwest
north of the equator, and south east to northwest to the equators
south (Chap. 1). On occasion these Kelvin waves do drive or at least
contribute to a reversal of the prevailing trade winds, generating
terrestrial superrotation which we know as an El Nio (Chap. 2).
How do Kelvin waves lead to the reversal of the prevailing
winds and the development of super-rotation on tidally-locked
planets? The key to solving this conundrum came from further
modeling work by Adam Showman (University of Arizona) and
Lorenzo Polvani (Columbia University). Together, they showed
that this west-to-east motion can be amplified by disturbances
generated by Rossby Waves moving further towards the Poles of
these planets. This happens despite the contrary motion of the
Rossby waves towards the west.
Where there is a strong difference in the temperature of
the planet on its day and night side, this sets up large amplitude
Rossby Waves in the middle and upper atmosphere (Chap. 1). For
a hot Jupiter, or tidally-locked planet with one side permanently
heated and the other side in permanent darkness, a large dome of
warm high pressure is locked on either side of the equator around
the point where most sunlight is received.2 Mid-latitude westerly
winds then flow upwards and around the feature. On the dark,
anti-solar, side pressure is lower because the air is colder (Chap. 1).
The westerly winds then flow back towards the equator (Fig. 10.2).
This pattern of heating and cooling establishes fixed Rossby Waves
in the planets atmosphere around these fixed low and high pres-
sure areas.
2
You can see the same effect during an El Nio where two areas of high pressure ank the equato-
rial low over the central Pacic. On Earth, the surface low is anchored not by a xed Sun, but by the
pool of warm water that lies beneath it (Chap. 2).
a b
L R Momentum Transfer
H L H L H K L
Jet
H L H L H K L
Momentum Transfer
L R
FIG. 10.2 The emergence of super-rotation in the atmosphere of hot

Jupiters and super-Earth planets. Eastward propagating pressure waves
form on either side of the equator (a). These features are known as Kelvin
waves (K, red circles and arrows). North and south of these are westward
propagating Rossby Waves (R, Chap. 1). The largest amplitude Rossby
and Kelvin waves are then fixed by very uneven heating and by the equa-
tor, leaving shorter wavelength disturbances to move with the mean flow
from west to east. Despite atmospheric momentum being lowest fur-
thest away from the equator, this set up causes the combined Rossby
and Kelvin waves to develop a northwest-southeast tilt north of the
equator; with a south west-northeast tilt to its south (b). This allows the
atmosphere to transport angular momentum from the poles towards the
equatorial regions by directing more and more energy into an equatorial
westerly jet stream. Once sufficient momentum has been transferred the
Kelvin waves are modified by the developing westerly jet which prevents
further acceleration. This leaves a self-sustaining westerly jet stream
above the equator and light westerlies beneath it. In our Solar System,
Venus (Chap. 5), Jupiter and Saturn (Chap. 7), Uranus (Chap. 8) and Titan
(Chap. 9) all show super-rotating jets above their equators. Venus has an
easterly flow; the others westerly. Super-rotation in the atmospheres of
Venus and Titan is best explained by this mechanism; while Jupiter and
Saturn, by different mechanisms involving convection and rotation
Normally this means that air flowing nearest the equator has
the highest momentum and shouldnt be able to accept any more
from air moving closer to the Poles, but the westerly (eastward-
moving) Kelvin Wave motion to pick up momentum from the
Rossby Waves further north and south. In their models, two
sets of waves combine, forming tilted, chevron-like structures
(Fig. 10.2). Although the eastward moving Rossby Waves would be
expected to cancel out the westerly Kelvin waves, instead eddies

and other disturbances generated by the Rossby waves can trans-
port energy and momentum into the Kelvin waves lying at the
equator and to quote the authors, pump-up the tropical west-
erlies. The chevron-like waves naturally direct momentum from
the higher latitudes towards the east at the equator. Showman and
Polvani show that even when you start with a static atmosphere,
the interaction between the westward-moving Rossby Waves and
eastward moving Kelvin Waves causes an acceleration of the equa-
torial westerly motion: in no time at all, super-rotation begins
with a strong westerly jet stream overlying the equator (Fig. 10.2).
This odd, and rather counterintuitive, phenomenon may explain
super-rotation in the atmosphere of slowly-rotating Venus (Chap.
5), but probably not Jupiter and Saturn (Chap. 7), Uranus (Chap.
8) or Titan (Chap. 9). Here, other factors appear to be at work.
At Venus momentum is transferred from the upper to the lower
atmosphere, which undoubtedly also affects super-rotation on hot
Jupiters (Chap. 5); while at Titan, Saturns gravitational pull may
affect or drive super-rotation through the action of the tidal bulge
it creates. Tapio Schneider and Junjun Liu (Caltech) suggest that
superrotation is driven by convection at the equator. This gener-
ates eddies and Rossby Waves (Chap. 1) on either side of the equa-
tor. This allows momentum to flow into the equatorial regions,
accelerating winds into two prograde jets (Chap. 7). All of this is
speculation at present.
Super-rotation ensures that on hot Jupiters the hottest con-
ditions are found to the east of the region that is most strongly
heated. Winds blow the hottest air towards the day-night termi-
nator. In general, planets that rotate most slowly show the weak-
est displacement of the hot zone from the sub-stellar point. This
means that for all tidally-locked worlds those that are located clos-
est to their stars are most likely to show hot spots in their atmo-
spheres that are located furthest from the sub-stellar point, while
those furthest away, and rotate the slowest, will show maximum
atmospheric temperatures in a lobster-claw-like pattern around
the SST. Compare the rotation periods. For a typical hot Jupiter
the planet takes at most a few days to rotate on its axis as it orbits
its host star. Venus takes 243 days to complete the same trick,
thus Venus experiences a far more leisurely Coriolis Effect than
a typical hot Jupiter. These have a 1-6 day day-long orbit, while a
super-Earth or Neptune orbiting a red dwarf in its habitable zone
will have a 2040 day-long orbit.
All of these models show that the maximum effects are
found in the troposphere (1 bar) through to the lower stratosphere
(100 millibar), while at higher altitudes (around the 1020 milli-
bar level) the hottest conditions are likely to be found where you
expect them to be: under the SST. This is simply because the thin-
ner air at this altitude is less able to hold and transport energy.
Here, air flows in a thermal wind, much like in the middle atmo-
sphere of Venus, in a relatively simple flow from day-lit side to
night side. Unlike Venus, the illuminated and dark hemispheres
are a permanent feature and do not rotate across the face of the
planet: tidal locking prevents this.
In general, all tidally-locked worlds show the same, overall,
pattern of atmospheric circulation. With one side heated and the
other side not, air is forced into a fairly simple, generic pattern
of airflow from warm to cool side. What is perhaps unexpected
is that this is not accomplished by a single Hadley cell running
from warm to cold. Even the modest rotation tidally-locked plan-
ets experience is sufficient to modify air movement through the
Coriolis Effect (Chap. 1). Thus although there is a broad ascent of
air at the sub-stellar point, towards the anti-solar point at height,
this is modified by the presence of a super-rotating westerly
jet along the equator. Likewise, the return flow is diverted by the
jet stream and by the large hemispheric lows and highs caused by
uneven heating and planetary-scale Rossby waves (Fig. 10.2).
Hot Jupiters
What does this mean for the weather on a hot Jupiter? Well, there
are two obvious effects that are illustrated in Fig. 10.3. The strong
westerly flow around the equator brings cooler air from the east
towards the sub-stellar point (SSP) where there is maximum stellar
heating, and takes warm air eastwards over the daynight termina-
tor towards the antistellar point (ASP). Depending on the strength
of the wind and the amount of heating this displaces the hottest
Day Hemisphere
90
Low High
45
Low
Latitude
High Low High Low High 0
Low
-45
Low High
-180 -120 -60 0 +60 +120 -90
Longitude
FIG. 10.3 The weather on a tidally-locked hot Jupiter. Models predict that
on hot Jupiters the hottest conditions are found 60120 east of the sub-
stellar point, or SSP for short (shown by a red and orange star on the
day hemisphere). A strong westerly jet stream (equatorial black arrows)
blows air around the equator and displaces the hottest regions down-
wind of the sub-stellar point. Where winds are particularly strong, the
hottest regions, perhaps split by the jet, may be displaced all the way
across the terminator into the night hemisphere (pink circles). Spitzer
observations of HD 189733b, HD 209458b and Tau Botes b appear to
confirm this prediction. To the north and south of the equator at 30
east of the SSP, in the mid-to high latitudes, warm-cored areas of high
pressure (may) bring more settled conditionsor form large Great Red
Spot-like storms. In the night hemisphere, downwind of the anti-stellar
point (or ASP, black and grey star), air descends and warms, generating
a second, weaker warm spot (or spots) that are again split into two by
the intruding westerly jet stream (dark pink ovals). Large areas of low
pressure form on either side of the ASP and, again, these could be broad
cyclonic storms. Across the entire planet, winds are thought to be pre-
dominantly westerly. These patterns are for the 0.1 bar level and below.
Higher in the stratosphere, winds are weaker with the hottest conditions
found near the SSP. However, thermal winds blowing from day to night
side still generate a weak warm spot where the air descends at either side
of the ASP. Kelvin waves generate weak areas of higher and lower pres-
sure along the equator. Equatorial winds move at hundreds of meters per
second, dwarfing even the intensity of Neptunes
conditions eastwards, potentially over the terminator and into the

dark hemisphere. Although this also may sound unlikely, VLTI,
Spitzer and Hubble observations of various hot Jupiters, including
HD 187933b and Tau Botes Ab appear to confirm this pattern,
with the maximum infrared emission occurring in a hot region far
to the east of the area that is most strongly heated. Moreover, the
observations confirm that the hottest region is shaped more like
a lobsters claw than a sphere around the SSP because cooler air
is being brought into the region by the westerly jet. There is no
surface low pressure at, or near, the SSP simply because there is
no surface. Air that is strongly heated has a high pressure and sim-
ply moves outward and away in a thermal wind towards the dark
and cooler hemisphere. Here, pressure is lower because the air is
colder. The overall pattern is then controlled by large, atmospheric
Rossby and Kelvin waves, which produce super-rotation with
strong west-to-east flow.
In the dark hemisphere, there is predicted to exist, a second
zone where temperatures are relatively high. East of the ASP is a
warm zone (or strictly speaking zones where air flowing from the
sunlit hemisphere converges and descends into the planetary inte-
rior. This zone is again displaced eastwards by the action of the
westerly jet. Observations of exoplanetary atmospheres are cur-
rently not sensitive enough to confirm or refute the existence of
this region but will likely do so in the near future.
In general, planets that rotate most slowly show the weak-
est displacement of the hot zone from the sub-stellar point. This
means that for all tidally-locked worlds those that are located clos-
est to their stars are most likely to show hot spots in their atmo-
spheres that are located furthest from the sub-stellar point, while
those furthest away, and rotate the slowest, will show maximum
atmospheric temperatures in a lobster-claw-like pattern around
the SST. Compare the rotation periods. For a typical hot Jupiter
the planet takes at most a few days to rotate on its axis as it orbits
its host star. Venus takes 243 days to complete the same trick,
thus Venus experiences a far more leisurely Coriolis Effect than
a typical hot Jupiteror even a super-Earth or Neptune orbiting a
red dwarf in its habitable zone, with 16 day and 2040 day-long
orbits, respectively.
All of these models show that the maximum effects are found
in the troposphere (1 bar level) and lower stratosphere (100 mil-
libar level), while at higher altitudes the hottest conditions are
likely to be found where you expect them to be: under the SST
(1 millibar) with little displacement to the east.
Observations of HD 209458b, discovered in 1999, are particu-

larly illuminating. This hot Jupiter orbits its Sun-like star every
3.5 days. Its mass is roughly 0.69 Jupiters but it is over twice
Jupiters volume. This contradictory pair of observations points
to the trauma HD 209458b is experiencing. In its 0.7 million kilo-
meter radius orbit the star is cooked by an intensity of radiation
from its star over 1000 times that experienced by the Jupiter. Its
troposphere has a searing 1000 C heat that steadily increases
with height. Spectral observations suggest that the stratosphere
is relatively cloud free for most of its height. Here, measurements
by Canadas MOST (Microvariability and Oscillations of STars
telescope) show that the stratosphere grows hotter with altitude,
reaching 2200 C at the 33 millibar level. MOST also showed
that the planet was very dark: its albedo was as low as 4 % mean-
ing that 96 % of the visible radiation reaching it was absorbed.
Compare that to Jupiter, which has an albedo of 52 %. This tells
Radius: 1.29 Jupiter Radii to 1 bar level 970oC. Hot silicate clouds likely below this level
Stratosphere: 1.32 Jupiter radii; 1,200oC but rising to 2,200oC 100-25mb contains dark cloud layer observed by HST
Exosphere: 3.1 Jupiter Radii, temp rising to 9,700oC 25mb-1b

Hot hydrogen coma containing escaping gases that include carbon and oxygen
FIG. 10.4 The overall structure of the atmosphere of HD 209458b. The

planet is embedded in a large, dispersed cloud of hydrogen and other gases
that is being driven away from the planet by a combination of intense
heating and a strong stellar wind. This is the coma and is analogous to
a comets outer cloud. Beneath this, within 3.1 Jupiter Radii lies the
exosphere that constantly supplies gases to the expanding coma. Below
this will lie a poorly defined thermosphere. Sarah Seagers group identi-
fied sodium in this layer, while Korey Haynes identified a layer of metal
oxide clouds at 25 mb (stratosphere). Temperatures fall with depth until
the troposphere, which is yet to be clearly observed but likely contains
clouds of silicate dust from which a rain of molten rock likely descends.
Layers not to drawn to scale
us that either HD 209458b has dark clouds or is relatively cloud-

free and light is Rayleigh scattered by deep, translucent layers
of hydrogen, much like the oceans of early Earth may have done
(Chap. 2) (Fig. 10.4).
Moving upwards into the thermosphere, temperatures even-
tually reach over 9700 C at its lid at the 25 millibar base of the exo-
sphere. In 2001 Sarah Seager (currently at MIT) predicted that the
atmosphere of this and other hot Jupiters would contain sodium.
Subsequent spectral measurements made by Hubble detected this
in the thermosphere, at pressures from 50 millibars to 1 micro-
bar (one millionth of a bar). Separate work, carried out by Jeremy
Richardson of NASAs Goddard Space Flight Center, showed that
the lower atmosphere contained surprisingly little water vapor,
which contradicted models. It might well be that the upper atmo-
sphere has had all of its water broken down to oxygen and hydro-
gen through the action of the intense ultraviolet light the planet
experiences in such a tortuous orbit. In 2009 JPL carried out fur-
ther spectral analysis and this time identified water along with
carbon dioxide and methane. This is an odd assemblage of incom-
patible gases: methane wouldnt normally be expected to coexist
with carbon dioxide as the two will tend to react with another at
high temperatures. By implication, there is a lot of dynamic atmo-
spheric chemistry going on around HD 209458b that is caused by
the rather hellish conditions.
What else do we know about HD 209458b? There are indi-
cations that at, or near to, the 25 millibar level there is a layer
of dusty clouds resting above the clear hydrogen layer. Given the
temperatures of 2000 C (with a range of 260), these are clearly
not water. Instead a froth of vanadium oxide and titanium oxide
seems likely. This makes HD 209458b rather similar to a cool red
dwarf star at least in terms of its appearance. Titanium oxide also
acts as a strong absorber of ultraviolet radiation, much like ozone
in our atmosphere or titanium dioxide, which is used in commer-
cial sun-creams. Like a sun cream, this layer absorbs energy which
causes heating of the stratosphere. This partly explains why the
temperature of this layer exceeds that of the troposphere, beneath.
Such a hot, absorbing layer is also seen in the hot Jupiter Wasp
33b. Hubble spied on this world and showed its stratosphere broils
at 6000 C. Research by Korey Haynes (George Mason University)
and Drake Deming (University of Maryland) again made spectral

measurements of Wasp 33b at different wavelengths, this time
using water vapor as a tracer. In this planets atmosphere, the lower
stratosphere was a cool 1700 C, but the layer immediately above
it was twice as hot. Again, absorption of ultraviolet radiation by a
titanium oxide sunscreen appeared to the driving force in gener-
ating this temperature inversion.
Returning to HD 209458b, deeper down, in the cooler tropo-
sphere, absorption patterns in the spectra suggest the presence of
silicate clouds, which might bequeath HD 209458b with show-
ers of molten rock. Imagine basalt falling out of the sky: perhaps
Peles hair drifting downwards through the hot, but otherwise
fairly clear sky.
Yet, this is more than a little misleading. For measurements
of carbon monoxide indicate that the atmosphere would hardly
allow drifting in a leisurely sense. Carbon monoxide, again
somewhat surprisingly, is one of the atmospheres dominant gases
after hydrogen and helium. Although only contributing a small
percentage, overall, this gas is a particularly useful tracer of atmo-
spheric motion. Not only does carbon monoxide emit and absorb
radiation at useful wavelengths, revealing that the hottest part of
the atmosphere lies eastwards of the sub-stellar point, in agree-
ment with predictions, carbon monoxide emission also clearly
shows that the atmosphere is in extreme motion. Doppler mea-
surements made by the CRIRES spectrograph attached to ESOs
Very Large Telescope show that at least at some levels, winds blow
at nearly 7000 km per hour from the sunlit side to the perma-
nently dark hemisphere. This intense wind is simply a response
to the extreme difference in heating. Were HD 209458b to lack an
atmosphere its dark face would freeze at less than 170 C, while
its sunlit side was over 1000 C. With such an extreme contrast
in temperatures, the fluid, dense atmosphere does what it can to
even out the temperatures. The result is winds that blast at more
than five times those of our windiest planet, Neptune (Chap. 8).
Therefore, drifting might be something of a misnomer. Instead
imagine a blizzard of molten basalt blasting downwards through
the increasingly dense atmosphere, before it vaporizes deeper in
the planets gassy interior.
In general most hot Jupiters are thought to have the hottest
conditions highest up. The atmospheres are stable and stratiform
in structure above the cloud tops because more and more radiation
from their star is available to absorb. Various observations con-
firm this is true for most hot Jupiters. But the keyword is most.
Like any good theory (and set of observations) there is always an
odd one out. In this case it is Tau Botes Ab. Observations by
the VLTI show that the upper atmosphere cools with height. The
only way that this is possible is if the lower stratosphere contains
materials that absorb energy extremely effectively. This would
allow this layer to warm more than more transparent layers that
were higher up. This set-up should allow the efficient transport of
energy by convection. However, in the other hot Jupiters that have
been observed, the upward transport of energy (if it occurs) will be
through atmospheric gravity waves that we shall come to later in
this chapter.
While the troposphere and stratosphere reveal unimaginably
violent winds, HD 209458bs exosphere is a marvel of planetary
destruction. The ellipsoidal exosphere extends over three Jupiter-
radii from the planets center (over 200,000 km) and reveals the
scale of the torment this planet experiences. Somewhere between
100 million and 500 million kilograms of hydrogen gas is stream-
ing away from this planet every second. Within this outward
moving flow are heavier carbon and oxygen atoms that are being
dragged away from the planet by the out-flowing hydrogen gas.
This hydrodynamic drag helps strip away much of the heavier
atoms that contribute to the mass of the atmosphere. This is the
same process that is happening today above Titan, albeit at a far
lower rate (Chap. 9). Mars (Chap. 6) also experienced this early in
its history, with out-flowing hydrogen pulling away some of the
planets carbon and oxygen. For both of these small worlds, the
amount of gas that was shed is easily dwarfed by the vast outflow
from this gas giant. Compared to the Earth, which loses around
1000 kg per second, HD 209458b is losing mass at a rate 100,000
500,000 times faster. Even with this prodigious rate, the massive
planet will survive up until its parent star swallows it at the end of
its main sequence life. However, it will survive in a much reduced
state. These discoveries are described in Fig. 10.4.
How is HD 29458b losing so much mass? Obviously the
extreme day-side heating is the underlying cause, but look again at
the planets dimensions. The mass is less than three quarters that
of Jupiter but its volume more than twice as much. This shows
that the planets mass is far more dispersed than Jupiter, or even
our fluffy Saturn. While Venus only gets hot enough to drive a
thermal wind from day side to night side and this is sufficient to
cool off the planets atmosphere, this process is grossly insufficient
to cool HD 209458b. Here, the planet is so strongly heated that it
generates high pressures within the atmosphere and upper layers
of the planet that blow the atmosphere off into space at hundreds
of kilometers per second. The figure of 100 million to 500 million
kilograms per second might well be conservative. Such hydrody-
namic escape was discussed in Chaps. 5 and 6 but has little cur-
rent impact on the atmospheres of our terrestrial worlds: Titan
(Chap. 9) is a likely exception. In the case of HD 209458b it is
stripping sizable quantities of the planets mass over the course
of a few billion years. Overall, if the current rate of mass loss is
maintained, HD 209458b will lose only another 7 % of its mass
before its host star becomes a red giant and destroys the planet.
This adds to the similar amount the planet will have already lost
since its formation. One might expect this rate to increase as the
planet loses mass and its gravitational pull on its gases decreases.
Conversely, the planet has a decent magnetic field which may help
it retain gasat least within its magnetosphere and thus slow the
overall rate of gas loss. Is HD 209458b alone in suffering this state
of affairs? Absolutely not: other hot Jupiters, such as Tau Botes
Ab, HD 198733b and 55 Cancri b also show gas escape revealing
that this is a general phenomenon with hot Jupiter worlds.
Tidally-Locked Earths and Super-Earths: General Ideas
We can lump the hot worlds together: the hot Jupiters, hot Neptunes
and hot Earth-like worlds. These all share the same principle atmo-
spheric motion that is driven by strong heating on one side and
cooling on the other. The only difference will be that the smaller
super-Earths will have less voluminous atmospheres and thus be
less efficient at transferring energy. Unless they have atmospheres
much less massive than that of the Earth they should distribute
heat in pretty much the same way. This will give rise to similar
weather. The only exception to this rule will be where a smaller
planet has a thinner atmosphere. This will allow the star to heat
the surface as well as the atmosphere. Here, convection might

dominate transport of energy from the day-lit to the night sides.
39 light years from Earth lies the red dwarf GJ 1132. In 2015
Berta-Thompson and his colleagues (MEarth-South Observatory)
published their discovered of a planet orbiting this star. This
Harvard University-led array of eight 40-cm-wide robotic tele-
scopes is located in the mountains of Chile and is dedicated to
scanning the heavens for Earth-like planets that orbit these little
stars. GJ 1132b is a relatively small world, only marginally wider
than the Earth. Consequently, with its density estimated to be
similar to the Earth it is likely to be rocky in composition.
As the orbit of GJ 1132b takes only 1.6 Earth-days to com-
plete it is well inside the region where temperatures exceed the
boiling point of water. With its fairly substantial mass GJ 1132
should host a dense atmosphere. With cloud-top temperatures
likely to lie around 230250 C, this world should resemble Venus,
but spinning at over 151 times the speed of Venus it must have a
much more substantial Coriolis Effect. Indeed, its Coriolis Effect
is similar in strength to ours, so we can expect some unusual and
distinctly non-Venusian weather. While Venus does show super-
rotation higher in its atmosphere (Chap. 5), near the surface there
is very little movement of air at the surface. By contrast, GJ 1132
would be expected to show a large-scale circulation that is domi-
nated by a westerly equatorial jet and broad convection to the east
of the SSP. With appropriate optics we might just be able to deter-
mine whether this is true in the very near future.
Continuing the trend from Hot Jupiter to tidally-locked
Venus, what if we take our imaginary world to a location where
the temperatures drop to the point at which liquid water becomes
feasible? These tidally-locked worlds are only possible around
red dwarf stars. For only these stars have a habitable zone, where
temperatures are modest enough for liquid water in the region in
which tidal locking will occur. What happens to the atmosphere
of these worlds and how would this affect their weather and their
habitability? As each of these worlds must orbit proportionately
further from its star in order to be habitableand this is only a
few million kilometersthe orbital period is anything from 15 to
40 days. This depends on the mass of the red dwarf: larger periods
are obviously associated with more massive and more luminous
red dwarfs. Consequently, the Coriolis Effect is weaker for habit-

able Earth-like planets around red dwarfs than it is for hot Jupiters
and other worlds orbiting Sun-like stars. Models show that there
is a transition between broadly static and super-rotating atmo-
spheres that occurs at different rotation rates, which is depen-
dent on the size of the planet. With an Earth-sized world, once
the rotation period increases beyond 5 days super-rotation slows
and soon stops. At this critical point Rossby waves are ineffective
at pumping energy into the tropics through eddies in the mid-lat-
itude atmosphere. For planets with radii above that of the Earth,
this transition moves to greater periods. A planet with twice the
radius of the Earth will cease super-rotation at around 10 day
orbital periods (and hence 10 Earth-day long rotations). Therefore,
the more massive the red dwarf, and the more extended its habit-
able zone, the less effective super-rotation will be at redistributing
heat around the planet.
Further refinements to this trend were made to these models by
Ludmila Carone (KU Leuven, Belgium). In planets where the orbital
(and rotation) period was greater than 45 days the equatorial super-
rotation was steadily replaced by westerly jets at higher latitudes
(around 60o north and south of the equator). Overall, there was a
more direct flow of air from the sunlit hemisphere to the dark hemi-
sphere at height. Indeed, as the rotation period was slowed further
to 100 days (so, for habitable planets tidally-locked to the most mas-
sive red dwarf stars) the flow of air showed the least super-rotation
and the greatest direct flow. In essence what is happening is that
there is more and more flow from all directions into the thermal low
pressure area at the surface. As this air rises and diverges it blocks
the westerly flow around the equator at the tropopause. These plan-
ets experience the weakest Coriolis Effect so the weakest east-west
drive with latitude. However, equatorial super-rotation became the
dominant means of transporting air in planets with periods shorter
than 25 days. Shorten the rotationand hence move the habitable
planet closer to its starand the super-rotation increases.
When you continue the trend to even shorter orbital periods
some odd things happen. Planets with periods less than 6 days
oddly re-develop the same higher latitude westerly jets, but this
time accompanied by equatorial super-rotation. When the rotation
period is reduced to less than 4 days the mid-latitude westerly jets
finally vanish and the equatorial jet dominates. On these worlds
super-rotation is so strong that it disrupts the uplift of air east of

the SSP and presumably cloud formation with it. On such plan-
ets the strength of the Coriolis Effect is close to, or the same as,
the Earths, but the difference in the pattern of heating drives the
observed difference in the circulation.
In each case the super-Earth experiences the change from one
regime to another at greater orbital periods than smaller, Earth-
like worlds. This is primarily due to the greater radius, and hence
circumference, of the planet. Larger worlds have to rotate faster
than smaller ones in any give periodand Rossby waves can fit
across the surface in increasing numbers or with greater wave-
length with increasing surface area.
The only issue I have with these is that in the shortest period
planets they used models where the planets had an orbital period
measured in a few days, but insolation equivalent to the Earth.
Now, in reality this can never occur. Any planet orbiting a red
dwarf with such a short orbital period must be far closer to its
star than the habitable zone and thus receive far greater amounts
of radiation than the Earth doesunless it orbits an even dimmer
brown dwarf, of course. That this aside, these models do suggest
that the circulation on tidally-locked super-Earth like worlds might
be much more complex and variable than we might first imagine.
Model 1: A Smooth, Unrealistic Planet with No Oceans
Now, lets take this information and make some worlds. If we first
imagine a perfectly smooth, dry world (i.e., a completely boring
and unrealistic one) then the atmosphere will move much like that
around a hot Jupiter. Youll, therefore, find a lobster-claw shaped
hot zone around the SSP (sub-stellar point) but extending to the
east. A westerly jet stream will extend its influence all the way
to the planets surface driving warm air eastwards from the SSP
across the day-night terminator. Above the SSP air will rise and
blow towards the ASP (anti-solar point) predominantly from west
to east. Around the SSP are two large areas of high pressure flank-
ing the equator, while two areas of low pressure flank the ASP.
The SSP will have low pressure where air is rising and the ASP will
have high pressure where air is descending.
Lets spice it up with an ocean; then well add land and moun-
tains to see what happens.
Model 2: Oceans on Super-Earths and Earth-Like Worlds
Life must have wateror at least some other solvent to transport

heat, substances and to promote the chemical reactions necessary
for life. If we stick with the universes most abundant solvent,
water, then we can begin to model our world. When we cover the
speculative planet with oceans some interesting things start to
happen that may ultimately have a bearing on the habitability of
the planet.
When the first models of atmospheres were done around
habitable, but tidally locked planets, it was thought that the only
region in which the oceans would remain liquid was under the
SSP. For here and only here would there be sufficient energy avail-
able to warm and liquefy H2O. This led to "eyeball" like surfaces
with open, ice free ocean only under the SSP. The rest of the globe
would be frozen over, except that these models neglected one
important thing: water is a fluid and will move. More recent mod-
els of potentially habitable super-Earths and Earth-like planets
take into account both the efficient heat transport by water and
the effect of planetary rotation on its motion. When you put all the
information together you get oceans that move pretty much like
the atmosphere above. Warm ocean currents move from west to
east, carrying energy from the SSP towards the dark hemisphere.
Concurrently, cold ocean currents can flow westwards around
to the day side of the planet, cooling the atmosphere under the
SSP. These motions both strengthen and are strengthened by cir-
culation in the overlying atmosphere, which is predominantly
west to east (Fig. 10.5). Therefore, conditions on a tidally-locked
habitable planet resemble closely those in a hot Jupiteronly
more moderate in terms of their temperature. Instead of an eye-
ball, open water will exist, at least in a lobster-claw-like pattern,
reflecting the motion of the ocean around the equator.
For the ocean, this circulation has two effects. Firstly, the
ocean remains ice-free over a greater proportion of its surface as
heat is more effectively transported across the planet. Secondly,
there will exist two overlapping circulations within the ocean
that should serve to transport nutrients effectively around the
planet. This is vital if the oceans are to remain vibrant and poten-
tially habitable. Detailed modeling work by Yongyun Hu and Jun
90
High Low
45
Latitude
Low High 0
-45
High Low
0 60 120 180 240 270 300 360 -90
90
High Low
45
Latitude
Low
High Low 0
Low
-45
High
Longitude Low
0 60 120 180 240 270 300 360
-90
FIG. 10.5 Winds on tidally locked habitable planets generally blow from
west to east, particularly near to the equator and at height. In these mod-
els of Yang et al. above slowly rotating planets air rises to the east of the
SSP and generally descends elsewhere. Winds are very sluggish except
west of the SSP. Nearest the surface winds are more strongly affected by
heating so that winds converge on the SSP (180 Longitude at the equator)
and diverge from the ASP. However, strong is a relative term. In the slow-
est rotating planets, the westerlies are 0.51 m per second (or 14 km per
hour)a very modest breeze. Top: 355 ppm CO2; bottom, 200,000 ppm
CO2. In Timothy Merliss models (not shown) a pair of relatively strong
super-rotating westerly jets emerge at mid-high latitudes, above a region
of westerly winds at the surface as planetary rotation is stepped up to 1
day. Meanwhile, at the equator the crescent shaped area of lowest pres-
sure becomes more pronounced with the lowest pressure displaced 90 to
the east by strengthening super-rotating winds
Yang (Peking University) suggests that oceans with depths greater

than 2 km will have two circulations. In the top-most layer, above
2 km, warm waters move from the equator towards each pole.
This circulation is driven by the zone of equatorial westerly winds

that drive water downwards at the equator then outwards to each
pole, where the waters rise up once more. Beneath these hemi-
spheric cells, if the ocean is deep enough, is a thermohaline-driven
circulation. In this thermohaline cell, water moves downwards
from the Poles towards the equator. The driving mechanism lies
with the formation of ice. As sea water freezes, denser, salty water
is displaced from the growing ice crystals because salt does not
fit readily into the structure of water ice. This cold, salty water
descends under its own weight towards the ocean floor. This gen-
erates a flow from the poles towards the equator: i.e., in the oppo-
site direction to water driven by the westerlies in the atmosphere
nearer the equator. If, however, the oceans are relatively shallow,
this arm of the global ocean circuit will be broken.
What are the consequences for the climate, and how does
changing the concentration of carbon dioxide, or the amount of
stellar radiation affect the climate? Some of the effects confound
expectations. In general increasing the concentration of carbon
dioxide increases the area of open water. This isnt because the
planet as a whole warms up, but rather because the atmosphere
can hold more heat energy and thus transport it more efficiently
from lit to dark hemispheres. While the day side barely warms at
all, increasing the concentration of greenhouse gases, particularly
carbon dioxide to around 2 % from an early-industrial, terrestrial
0.035 %, causes the dark hemisphere to warm substantially.
Indeed, it causes the ice to melt across the entire globe. In terms
of the atmosphere, the strength of the super-rotating westerly jet
increases substantially and extends downwards towards the ocean
beneath. In the atmosphere with early-industrial levels the atmo-
spheric motion at the surface is only weakly from west to east,
except closest to the western edge of the SSP. Once the carbon
dioxide concentration is increased the increasingly deep overly-
ing westerly jet stream drives a very noticeable westerly airflow
around the equator, and ultimately it is this that causes the dark
hemisphere to warm up.
If we then think back to the end-Permian mass extinction
(Chap. 3), ocean circulation appears to have played a pivotal role in
exterminating most of the planets complex life. An already warm
ocean became super-heated and stagnant, with microbial life filling
its depths with methane and hydrogen sulfide. Ultimately, it was

the hydrogen sulfide that killed off most of the ocean and possibly
land-based life. For a tidally-locked planet, the active circulation
described here, should minimize the risk of ocean anoxia (lack of
oxygen) and euxinia (the build-up of hydrogen sulfide), which would
likely cause the death of most aerobic (oxygen-loving) life forms.
Increasing carbon dioxide concentrations has other effects
which might be surprising. You might expect more carbon diox-
ide and a warmer world, overall, might mean more convection
and hence more clouds. Yongyun Hu and Jun Yangs work sug-
gests instead that as carbon dioxide increases it is the effect of
the increasingly strong westerly jet stream and ocean current that
matters most. Instead of enhancing convection the strengthening
westerly jet and ocean currents mix cooler air more thoroughly
around the planet, along with cooler water. So, as was mentioned
above, global temperatures are more even throughout. With less
of a contrast in temperature convection is actually suppressed
by increasing carbon dioxide. This is despite a higher tropopause
(Chap. 5) which might favor taller convective clouds. In the end
the planet is more cloud-free than if it has less carbon dioxide.
What about habitability? There are some contradictory out-
comes here. You might imagine that a larger area of open water
would allow more habitability. That might be true for a planet
right in the middle of the habitable zone, but overall, more open
water decreases the planetary albedo: the planet becomes darker.
This was suggested for the early Earth as a mechanism to warm
the early Earth with a dimmer Sun (Chap. 2). Now, the authors
conclude that the end result is that with a darker planet, it will
absorb more energy from its star and this, in turn, will cause the
inner edge of the stellar habitable zone to move outwards away
from the star. After all, a darker planet will be hotter at any given
amount of stellar heating and this will lead to the planet overheat-
ing and experiencing a thermal runaway, sooner (Chap. 5). This
narrows the habitable zone.
It must be stressed that this conclusion is still open to inter-
pretation. Look for example at another piece of work by Jun Yang
in the preceding year. Yang and colleagues concluded that convec-
tion under the SSP will produce enough cloud to cause negative
feedback. What this means is that the extra cloud reflects incoming
radiation and thus reduces the likelihood that the planet will over-
heat. This prevents the dreaded thermal runaway and keeps the
planet habitable. Therefore, in this work convection broadens the
habitable zone because the clouds it creates are reflective. This
work also moves the warmest area of the planet back towards the
SSP despite the pervasive westerly jet stream around the equator.
Now, in this set of published models oceans were included but
they were not truly global in extent, with continents partly block-
ing the movement of water around the globe. Thus, it is only on
true water-worlds that the habitable zone will be affected.
Turn again to another study and something slightly differ-
ent comes out. In the work of Jrmy Leconte and colleagues that
was described in Chap. 5, thermal runaway is averted because of
negative feedback as the planet warms. In their model, as tem-
peratures rise the Hadley Cells strengthen. Although this leads
to increased cloud cover near the equator, it causes a reduction in
cloud cover elsewhere, where air descends, warms and can hold
more water vapor.
In Yangs work reducing cloud cover is thought to have a
destabilizing effectreducing planetary albedo because oceans are
darkerhowever, Yangs work suggests this is dependent on the
type of planet. If you increase the amount of radiation a tidally-
locked planet receives, then the amount of convection increases
and the planet gets cloudier. This makes it more reflective. This
has two effects. More cloud cover makes the dayside more reflec-
tive and the night-side less: this evens out temperature contrasts,
increasing habitability.
By contrast a non-tidally-locked planet, such as the Earth,
experiences the opposite effect. Stronger heating, in Yangs model,
means less of a temperature gradient and weaker Hadley cells.
This causes greater warming as fewer clouds mean reduced reflec-
tion of incoming radiation and more absorption of radiation by
the planets surface. Lecontes work suggests that on an Earth-like
world greater warming leads to stronger Hadley circulation and
(overall) fewer clouds, but those changes to albedo are off-set by
increases in the ability of air to hold moisture. With less water
escaping to the stratosphere, a runaway greenhouse is held off until
the amount of radiation the planet receives increases to 375 W per
meter squaredor approximately 100 W per meter squared more
than the Earth currently receives from the Sun. This, broadly cor-
responds to an average surface temperature of 6570 C for an
Earth-like planet. Lecontes work also shows that over the Horse
Latitudes, where the air is far from saturated because it is warming
and drying out, the atmosphere acts like a giant radiator and cools
the planet down. Thus in the cloudless areas energy is lost and the
planet can absorb more energy before it experiences a thermal run-
away. More recent work by James Kasting, which was discussed in
Chap. 5, also reaches the same conclusion. While Lecontes work
only looked at the troposphere, Kastings included more water
vapor and examined what happened in the stratosphere. At similar
temperatures to those seen by Leconte, the stratosphere fills with
water vapor and the planet dries out. These two pieces of work
suggest that planets can remain habitable for longer, with greater
amounts of heating before they expire.
Moreover, Lecontes work reaches the opposite conclusion of
Yang: increasing cloud cover in their model increases the green-
house effect and leads to a faster runaway. Here, the water droplets
and ice crystals in clouds absorb more energy than they reflect and,
therefore, increase warming. How come this is so different? Well,
its down to the type of clouds in the models. Yang has mostly
cumulus and cumulonimbus convective clouds, while Leconte
favors an increase in cirrus and cirrostratus because they suggest
under hotter conditions more high level clouds form, while lower
clouds are evaporated. Cirrus and cirrostratus are dominantly
absorbing rather than reflecting, meaning that as they increase in
abundance, the greenhouse effect increases in sync. Think back
to Chap. 2 and the 9/11 attacks in the US. When planes were
grounded, there was a 1.8 increase in the diurnal (daynight) tem-
perature rangeprimarily because more radiation was escaping at
night, with fewer plane contrails to absorb it. Despite less reflec-
tion in the daytime, the dominant effect was absorption of outgo-
ing radiation at night, not more reflection during the day.
All of this is rather confusing. It comes down the role of clouds.
If their dominant effect is to reflect energy as Yang concludes,
then habitable zones are extended where tidally-locked planets
have limited ocean circulation. If Leconte is correct then increas-
ing cloud will have the opposite effectbut that this depends on
the type of clouds a tidally-locked planet will form. In Lecontes
models decreasing cloud will stabilize the planetary temperature

and limit the chances of a thermal runaway because, overall, there
will be less cirrus and cirrostratus-like clouds. At the moment, the
best we can do is speculate (Figs. 10.6 and 10.7).
a 90
High High 45
Latitude
Low 0
-45
High High
0 60 120 180 240 270 300 360
-90
Longitude
90
High
High 45
Latitude
Low
Low 0
High High -45
0 60 120 180 240 270 300 360

-90
Longitude
FIG. 10.6 A different pair of models by Jun Yang that show the effect of
slowing the Earths rotation to once every 197 days (top) and moving it
to the current location of Venus (bottom). With a slow rotating Earth, in
roughly its current orbit, a large area of low pressure tracks the slowly
moving Sun across the face of the planet, flanked by areas of high pres-
sure to it rear (0/360, approximately). When the amount of radiation
is increased to that of modern day Venus, in Wangs models, the Earth
remains habitable because enhanced convection, linked to the larger sur-
face low pressure area, produces more clouds and these reflect more sun-
light. Where Yang simply moves the current Earth, with its 24 h rotation
period, the Earth becomes uninhabitable as there are no longer enough
clouds to reflect the higher level of solar energy causing the Earth to
overheat. Leconte, on the other hand would have an uninhabitable Earth.
See text for details
FIG. 10.7 The types of clouds hot planets manufacture appears critical to
their fate. If they produce large, convective, cumulus and cumulonim-
bus clouds, like those above left, they will reflect much of the incom-
ing radiation and remain cool and habitable for longer. However, should
they produce mostly high, ice clouds, such as this unusual example of
cirrus (above, right) or the cirrostratus ice cloud above it, then they will
preferentially trap more heat at night time than they release during the
day. This will cause the planet to overheat more quickly than if it had no
cloud at all. Photographs by author
In another set of models, Yang puts another effect into play:

rotation (Fig 10.6). In their models where the Earth is made to
rotate slowly (roughly similar to modern Venus, with a 197 day
rotation period) the Earth takes on an aspect similar to a tidally-
locked planet. There is a strong area of uplift, low pressure and
convective clouds tracking the Sun as it slowly moves across the
globe, much like the SSP on tidally-locked worlds. If you then
move this slowly rotating Earth to the location of Venus the Earth
surprisingly remains habitable. This is again down to how you
view the role of clouds. In Yangs models, clouds are reflective
and moving the Earth towards the Sun makes it, naturally hotter.
This enhances the formation of convective clouds which causes a
negative feedback. The Earth is then shielded from the additional
radiation, keeping it habitable. Yang goes further and runs another
simulation but this time with an Earth that rotates as fast as it
currently does. In this model the Earth soon overheats as it cannot
produce sufficient cloud cover to keep the Earth cool. With a fast
rotating Earth (as it is now) convective clouds remain fixed to a
narrow equatorial belt and unlike Lecontes models, cannot emit
or reflect enough energy to keep cool. A thermal runaway ensues.
This is all to do with the types of clouds planets form at high tem-
peratures. Where icy cirrus and cirrostratus clouds dominate, the
planet is destined to boil dry. Large, fluffy cumulus, filled with

relatively large water droplets are found, lead to planets that will
reflect more energy and stay cool for longer.
Until we can accurately model the type and distribution of
clouds on tidally-locked worlds, we really wont understand how
they will affect climate. How this will combine with the effect
of cloud cover on ice cover and ice cover on evaporation from
oceans remains in the realms of speculation, but at least we are
beginning to see how tidally-locked planets will begin to function
and hence make some accurate models of weather and climate on
these worlds.
One outcome from Yangs models of fast and slow-rotating
Earths is we can again look at why Venus overheated. Yang sug-
gests that if the early Venus rotated quickly, this would have led to
an early thermal runaway, making Venus uninhabitable early on.
However, if it had rotated slowly then it could have been cooler
for longer and thus, remained habitable even as the Sun slowly
brightened. One must consider the influence of the Suns gravity
on Venus here. Although this might have slowed Venus early on, it
seems more likely that Venus only slowed to its current, sluggish
(retrograde) rotation after its atmosphere overheated and filled
with sufficient carbon dioxide that the Sun could slow it down.
All these models show a few things. For one they are still not
completely consistent. More than this, they indicate how a wealth
of different factors will likely influence how a planets climate
adapts to distinctly unearthly conditions, such as slow rotation or
increased stellar radiation.
Tidally-Locked Earths and Super-Earths:

Towards A Realistic Planetary Model
Lets begin to put all of the above information together and com-
bine it with what we know about our Solar Systems worlds. Weve
got some issues with the effect of clouds, but we can model the
rest of the atmosphere and oceans with a fair degree of accuracy
(or likely accuracy). Now, well take a lead from Yangs 2013 paper,
which modeled the circulation with varying amounts of carbon
dioxide and the presence or absence of land-masses (Fig. 10.8).
Temperature Inversion
SSP ASP
FIG. 10.8 The general pattern of air flow across a tidally-locked (or very
slowly rotating) planet. Air rises at the SSP then flows at height to the
dark hemisphere. There is uplift within 15 of the SSP and general, slug-
gish descent elsewhere. This leads to warming and the formation of a
temperature inversion across most of the planet. In turn, this limits how
high clouds can rise by convection (left) so that only at the SSP is convec-
tion strong enough to produce thick enough clouds for precipitation. Bear
in mind this model is for a perfectly smooth planet. This pattern of circu-
lation is superimposed on the general west to east flow around the planet
Here, as weve seen, without land-masses, air on a tidally-

locked Earth-like planet, or super-Earth, will super-rotate around
the equator with Hadley Cells that flank a large area of warm
low pressure east of the region that is most strongly illuminated.
Warm air flows from this equatorial region into each Hadley cell
and moves north and south at height. Air also flows to the coldest
region, east of the ASP, predominantly eastwards, again at height.
Air then descends and returns, across the Polar Regions, from the
ASP to the region east of the SSP. Air also flows in the surface
westerly, super-rotating flow from the ASP towards the SSP. This
completes the global circuit in the troposphere. The whole process
is driven by heating at the SSP and cooling at the ASP. Put land
in the way and the land disrupts the flow in two ways. First of
all, land exerts greater friction than water and cannot be shifted
by it. Therefore, the surface westerlies are broken up. Moreover,
land heats and cools more rapidly than seawater (it has a lower
heat capacity). Land also dries out meaning that it contributes
less moisture to the atmosphere above it, which also allows it to
heat up more rapidly than sea water. This is the driving mecha-
nism for the Monsoon (Chap. 2). Consequently, air tends to have a
lower pressure over sun (star)-heated land than sea water. The con-
verse is true where land cools down. Here, land cools the fastest.
This generates higher surface pressure. What does this mean for
the climate? Well, for a tidally-locked planet the arrangement of
continents (if there are any) is crucial to whether the planet will
be habitable.
On Earth the continents are both formed and shuffled around
by the process of plate tectonics. If an Earth-like world is less than
six billion years old there is a good chance that this process will
operate. On larger, super-Earths plate tectonics may or may not
work. Studies suggest that anything up to ten times the mass of
the Earth will operate this process, but that as the planet grows
larger plate tectonics may gum up faster, once the mass exceeds
23 times that of the Earth. The reason is continental crust. Larger
planets will be hotter inside and will brew out their granite faster
(see also Chap. 5). Granite, the stuff of continents, is too light
to subduct and gradually builds up on the planets surface. This
eventually chokes off the recycling of the crust which is needed to
operate the process.
Where plate tectonics operates the arrangement and position
of the continents will change over periods of tens to hundreds of
millions of years. This, as we saw in Chaps. 2 and 3, changes the
movement of the atmosphere and hence planetary climate. In
extreme cases plate tectonics can cause the climate to shift radi-
cally across the entire globe, potentially turning a balmy planet
into a freezer. There is yet another way in which plate tectonics
can conspire with the fundamental properties of a planet to alter
its climate: this is angular momentum. On a rotating globe equa-
torial regions spin the fastest around the center of mass, while
Polar Regions spin the least. In this situation bulky, massive con-
tinents are arranged in their most stable configuration when they
lie closest to the equator. Move them to the Poles and its like
putting a toy car on a spinning top: it will topple over. On Earth,
this process has been identified and is known as true polar wander
(or TPW for short; Chap. 2). Instances of TPW are relatively rare
as the process requires a particular arrangement of continents
and a suitable mass of them. On planets with few continents this
process may not happen at all. Increase the mass of these and the
planets spin may become unstable. On its own, this process can
accelerate a continent over tens of degrees of latitude in only a few
million yearsaround ten times faster than normal plate motion.
FIG. 10.9 An early attempt at reconciling different pieces of information

on the likely circulation on tidally-locked exoplanets that was used in
Under a Crimson Sun. Some features such as the low pressure near the
SSP and the super-rotating westerly flow are correct. However, the other
flanking lows and highs are probably not. Nice, imaginative try, however.
As the next figures show the circulation for aquaplanets is likely much
simpler, although changing the parameters used in models has some sig-
nificant changes. Despite the inherent simplicity of the circulation above
water worlds, the introduction of landmasses can have significant effects
making these planets more like the one proposed here
For a tidally-locked planet this process is likely to become

rather cyclical because the likely most stable configurations of the
continents will be either grouped facing their star or grouped facing
away from their star. This is a reflection of the current situation
with our Moon. The Maria face us while the highlands, predom-
inantly, face away. This, as it turns out is the most stable con-
figuration with mass centered facing towards or away. But, why
might the process be cyclical? Well, think about plate tectonics.
This will periodically assemble supercontinents and these will

experience the largest pull as they are the largest collection of
mass. However, super-continents are ephemeral: neighboring sub-
duction zones pull on the edges of these continental beasts, while
hot spots undermine their interiors. Ultimately, this causes them
to fall apart, before closing ocean basins bring them back together
once more. When assembled in random locations, true polar wan-
der will bring them back in line with their star; then disassembly
will scatter them once more, only for the process to repeat.
Taking all this into account, lets assemble most of the con-
tinents nearest to the SSP and then see what a likely consequence
for the planets climate is. Figure 10.10 illustrates this, taking into
account what we already know about super-rotation and the effect
of the fixed heating. Ive placed the SSP over something that looks
awfully like Africa. Indeed, for the sake of argument I used the
Earth of the future: Amasia (Chap. 2). This way you can relate
some of the climatic features of this model to those of the Earth.
The same arrangement of continents will be used in the subse-
quent scenarios.
What of the climate? To keep things straightforward equato-
rial Africa was chosen as its a relatively small continental mass
compared to the assemblage further north and will heat very dra-
matically under the intense illumination at the SSP. This will draw
in moisture from surrounding oceans generating persistent rainfall
and a wet, tropical climate, particularly along its windward coasts.
To the north of this region, the current Sahara desert conditions
might persist, as air flowing into this region predominantly comes
across a long fetch of land and will have dried out. On the con-
tinents north-western shores onshore westerly winds will bring
precipitation leading to something more like a Mediterranean cli-
mate. Along the eastern Atlantic shoreline a cooler, maritime cli-
mate, much like the present day one would be found. Australia
would have a cold climate, but it may be kept temperate by super-
rotating westerly winds. If these winds are blocked or sufficiently
subdued by the arrangement of the continents Australasia will
freeze over; most likely with an ice cap developing as weak west-
erlies deliver snowfall from the neighboring warmer oceans.
Turning our attention to North America and South America,
the Amasian organization places most of these across the day-night
ASP
b
High Low Tundra
f North
Pole
High
Katabatic High a
Winds SSP
Low
Low
Desert
d Rainforest
Katabatic High 45oN
Winds
f
Low c
e
FIG. 10.10 Transporting Amasia to a tidally-locked Earth around an

M-class dwarf. The SSP has been placed over Africa and the ASP over
the future Panthalassa. Rotation is slowed to once every 30 days (720 h),
which is appropriate for an Earth-like planet in an orbit around a mid-M-
class red dwarf, were it to be habitable. Features: The large thermal low
nearest the SSP is distorted by the continent which warms faster than
neighboring ocean (a); pressure is highest east of the ASP (b); weak equa-
torial westerly winds carry cool air to the SSP from the ASP and warm air
away (c); high pressure zones for over mountains where westerlies cross
them (d) and low pressure in their lee (e). Mountain ice caps are white
(f). Light black arrows indicate regional winds caused by surface features
terminator so that they are cold, dark deserts. On the north-

eastern side of North and South America some mountain belts
are retained to add some flavor to the climate. These mountains
are produced by subduction and continental collision (Chap. 2).
As these have predominantly cold climates, but are adjacent to
potentially warmer oceans, some ice caps that are fueled by mois-
ture from the adjacent seas have been added. What is interesting
about dong this is that the ice caps will likely shed very cold, dry
winds onto the neighboring oceans. This is a result of both the
prevailing light westerly winds which are likely at these latitudes,
but also because the very cold air is denser than surrounding air
and will sink rapidly under its own weight. Thus locally, particu-
larly where glaciated or other valleys empty out into the ocean,
there will be particularly strong cold winds called katabatic winds.
These local or regional features will lower temperatures along the
affected coastlines and also generate a lot of shower activity as
cold air pushes out over warmer water. In Europe one such valley
wind is called the Mistral, which blows up to hurricane strength
down the Rhone valley when conditions are suitable, particularly
during the early spring.
Mountains have another important effect on the overall cir-
culation of the atmosphere. They divert westerly winds north and
south. As winds approach the windward face they are compressed
and develop a northward (cyclonic) spin. At the summit of the
mountains, winds diverge (spread out) and develop the opposite
(anticyclonic) spin. This leads to the formation of a semi-permanent
ridge of high pressure aloft, with airflow bending to the south in
the mountains leeward down-slope. Further east, as the air moves
back down to the leeward side of the range, it develops the oppo-
site (cyclonic) spin to that on the ridge, bending to the north. This
forms a semi-permanent trough in the atmosphere. As we saw in
Chap. 1 this is a Rossby wave that generates further waves down-
stream. Since these waves develop in the predominantly light
westerly or north-westerly mid-latitude winds, at this longitude,
they will then move towards the south-east and interact with the
northern fringes of the large, warm cored (or thermal low) over
Central Africa.
In terms of the climate, on our M-class Amasia, will likely
then have a low pressure area anchored to the east of the moun-
tains that fringe North and South America (Fig. 10.9). In combina-
tion with cold air pouring off the continents onto the ocean, one
would expect storms to develop here, which would likely drift east
towards the European side of our super-continent. At the surface
these might manifest as frontal systems (Chap. 1), bringing steady
precipitation to the European flank of our tidally-locked planet.

If they were to continue further east or south, they might stir up
dust storms over drier areas, much as they do on Mars (Chap. 6).
How these incursions of colder air would affect the low pressure area
over Africa is hard to say. They might well peter-out over southern
Europe or they may, like Martian storms intensify bringing severe
dust storms into the tropics or drive the formation of more persis-
tent storm cells over the equatorial regions east of the SSP.
Overall, the climate is likely to be extremely varied on such
a world. Yes, one hemisphere is likely to be a tad chilly, with tem-
peratures permanently below freezing, but a rich variety of geo-
graphical locales will produce a wide variety of weather. All that
can be said is once you have that style of weather it is fairly likely
to persist. For without a proper day-night cycle and a fixed angle of
illumination and heating the weather is not likely to change much
from day to day. Your best chance for weather is on the windward
slopes of the mid-latitudes, where onshore winds and possible
frontal systems would likely bring day-to-day, or week-to-week
variation. The temperature will depend on your geometric dis-
tance from the SSP and whether you are on its western or eastern
side of the SSP. East of the SSP is where the warmest conditions
will be foundbut not necessarily the driest for converging moist
air will generate copious convective showers and storms. Had I
chosen to put the SSP in the heart of Eurasia or North America,
the overall climate would have been much drier. Rainfall would
have been on the ocean-facing coastlines as warm, moist air con-
verged across these. The interior of the continents would likely be
dry, hot desert by comparison, depending on the strength of heat-
ing and the distance from the SSP to the nearest shoreline.
The End of Weather: Atmospheric Collapse and Atmospheric Erosion
We conclude our tour of the tidally-locked planet with two prob-

lems that could make habitability rather precarious: atmospheric
collapse or erosion. Collapse occurs when the temperature of the
atmosphere falls below that need to condense (or re-sublimate) its
major component. In the case of the terrestrial planets the most
important gases would be water and carbon dioxide, for without
these greenhouse gases global temperatures would fall lower still.

The critical temperature for our world and Mars is around 80 C
for a fairly unrealistic concentration of carbon dioxide. This is
an average temperature and far below the average temperature
achieved even on Mars (Chap. 6). Thus while carbon dioxide can
freeze out over the Martian Poles it never becomes cold enough
for the entire atmospheres mass to freeze out: instead on Mars, it
circulates between polar stores.
Atmospheric collapse was considered to eliminate habit-
ability on most tidally-locked planets (or our tipped-over world)
where one hemisphere would, potentially, become cold enough for
water vapor and carbon dioxide to freeze out. It turns out that for
a planet in its stars habitable zone in most cases the atmosphere
around a potentially habitable planet needs to be rather thin (and
therefore not very habitable) before collapse happens. How thin?
In the late 1990s work by Joshi suggested it could be as low as
30 millibaror three times the average surface pressure on Mars.
Robin Wordsworth took another look at this and found it could
happen for pressures as high as 140 millibaror one tenth the sur-
face pressure on Earth. Low by our standards, rather a lot of organ-
isms on Earth would survive this pressure.
Wordsworths results made the atmospheres of tidally-locked
planets appear less stable than previous work. This was a result of
changing the effectiveness of carbon dioxide as a greenhouse gas
at low pressure. At such low pressures carbon dioxide has broader
windows (narrower absorption bands) which allow more energy
to escape the planet over its dark hemisphere (Chap. 3). At pres-
sures of less than 0.14 bar, carbon dioxide leaks enough energy to
allow collapse on planets that receive the same stellar heating as
the Earth does at present. Therefore, although we shouldnt worry
that relatively large red dwarf planets, such as GJ 581d, might
suddenly lose their habitability, it does pose a threat to Mars-
like planets. With a thinner or less stable atmosphere these could
lose some critical support when gases on the permanent night-
side become severely chilled. Even when we consider our Mars,
the atmospheric pressure has decreased at a fairly high rate over
geological time, thanks to various escape mechanisms (Chap. 6).
On a tidally-locked planet, the atmosphere might suddenly change
from habitable to inhabitable over a few million years. This would
happen once the overall pressure had fallen below the cut-off; and
this would be despite the presence of a sizable amount of green-
house gases that were still present.
While, atmospheric collapse is not a significant problem for
most planets in orbit around red dwarfs, atmospheric erosion is.
This is all a question of time. In our Solar System the only planets
and moons that show significant loss of atmospheres are Mars and
Titan. The latter has large enough reserves for this not to matter at
present (Chap. 8). What about a planet orbiting a red dwarf? Such
planets are only a few million kilometers from their Sun and are
subject to a greater bombardment of charged particles from their
stars stellar wind. Although red dwarfs typically sport limited
stellar winds (less than one hundred trillionth of a solar mass of gas
per year) when they are young this rate is perhaps a thousand times
greater. Calculations by Jess Zendejas (Max Planck Institute for
Astrophysics) suggest that if this persists for more than one billion
years then any planet without a sizable magnetic field, in its red
dwarfs habitable zone, could lose much, or all, of its atmosphere.
What this means depends on a number of factors. Do tidally-
locked planets lack magnetic fields? That is hard to say, but we
can look at Mercury, which has a low mass and rotates very slowly
(much more slowly than most habitable red dwarf planets). Clearly,
Mercury still sports a decent field. Now, although most red dwarf
worlds might well retain a protective shieldor an atmosphere
like Titan that is replenished from deeper storesthese stars have
lifetimes measured in trillions of years. It seems fairly likely that
over time any planet residing in the habitable zone of a red dwarf
will, ultimately lose its atmosphere to space long before its parent
star dies.
The biggest threat to such worlds will probably happen after
six billion or so years. For it is here that plate tectonics is likely
to fail, ether because the planet has become too cold or because
the surface has become clogged with granite and is too buoyant
to support the subduction needed to stir convection within its
planets core. Here, the loss of the magnetic field will most likely
instigate a very slow process of attrition that whittles the planets
atmosphere down over tens of billions or (less likely) hundreds of
billions of years. Gas giants and ice giants will most likely retain
most of their mass because they have a higher gravitational pull
and because they have more mass to lose. However, Earth-like

worlds, or more massive super-Earths will likely lose most of
their atmosphere long before their central star dies. Life, it seems,
is not eternal for any world. Remember, once the density has fallen
to suitably low levels, atmospheric collapse might complete the
job of converting a habitable planet into an icy desert like Mars.
Earth-Like but not Tidally-Locked
Instead of placing our planet near an M-class red dwarf star, were
going to put it around a slightly more massive orange K-class
star. If it is to be habitable, then it must lie further from its more
luminous host. This places it outside the region around the star
where gravitational forces will be strong enough to cause tidal
locking. Its year is still relatively short at 130 Earth days (in an
orbit of 0.4 AU). The length of year is based on some reasonable
assumptions about planetary mass, tidal-locking radius and stel-
lar mass.3 The planet rotates relatively slowly (3 Earth days) so
that the Coriolis Effect is more like that experienced by Mars than
our Earth. Weve also given it a tilt of 23.5 so that it has seasons.
Whats the weather like?
On this world the atmosphere circulates in a manner that is
somewhat like that seen on our planet. With slower rotation the
tropical Hadley cells reach further across the globe, bringing a trop-
ical climate closer to the Poles. Hadley cells span roughly 30 on
either side of the equator on our planet. On Mars, which has a com-
parable Coriolis Effect to the planet weve made here, Hadley cells
extend to 60 north and south of the equator. Smaller Ferrel cells
are induced on the Polar side and extend almost the full way to each
pole. Finally, like the atmosphere of Venus at each Pole there is a
quieter area where cold air descends towards the planetary surface.
During the winter months the arrangement of continents
around the northern polar regions will lead to strong internal cool-
ing as land has a lower heat capacity than water (Chap. 2). This leads
to strong outward blowing winter monsoon winds to the south of
these land masses (Chap. 2). Over what is the North Atlantic, an
3
See Jrmy Lecontes 2015 paper in the journal Science for more details. This is listed in the
references.
High
Low
Low
High
Low North
High Pole
Low
Low
High
High
FIG. 10.11 The surface circulation during the northern winter on an

Earth-like world with a Mars-like Coriolis Force. On this map the sub-
tropical jet stream (orange) and polar-front jet stream (purple) are indi-
cated. Large areas of cold-cored high pressure dominate the continental
interiors driving cold, dry winter monsoon winds out-wards. The Fer-
rel cell over the northern oceans is strengthened because of the larger
temperature contrast between the cold air and the ocean underneath.
Summer monsoon winds blow southwards over Africa towards its south-
ern half (black arrows). The Weaker Coriolis Effect means that air blows
more strongly across pressure lines (isobars) than it does on our Earth
invigorated polar front jet stream extends stormy weather across

our alien Europe. Similar storm belts lie along the southern mar-
gins of the other continents bringing seasonal wind and rain.
This is shown by cold and warm fronts anchored to deep, cold low
pressure areas (Fig. 10.11). An anchored Rossby Wave in the jet
stream generates storm systems east of what is South America.
A second branch of the Polar front jet brings seasonal storms across
what is our alien Southern Europe and northern India. Meanwhile,
the weather inland is predominantly cold, windy but dry.
Zooming in on the northern side of our alien South America
is a small narrow in the mountains. Here, we might see gap winds.
One of the finest examples of these is the Chivela Pass winds in
Central Mexico. These occur in the winter and early spring when
cold fronts penetrate far to the south of the United States south-
ern shores. These can arc around the Gulf of Mexico, frequently
south of a strong blocking anticyclone which is situated over
North America (Chap. 1). When these cold fronts curve back west-
wards, their progress is blocked by the Sierra Madre Mountains.
However, at Chivela Pass there is a clear route east towards the
Pacific basin. The cold front carries denser air to its rear, which
ponds behind the mountain barrierexcept at Chivela Pass. After
filling the Gulf of Mexico to a certain depth, the cold air pours like
a waterfall, through Chivela Pass and out into the Pacific. This
generates fierce easterly winds along the western coast of Mexico
that fan outwards for over 150 km into the Pacific. The whole
process makes an aneurysm-like blob on the otherwise blocked
surface cold-front. This aneurysm balloons westwards through the
Chivela Pass and out into the Pacific, carrying with it the other-
wise stalled cold-front. On our world, the gap winds can advance
eastwards into our future Atlantic-like basin and generate both
storm-force winds and vigorous convection where they push over
warmer waters.
The summer circulation (Fig. 10.12) is dominated by large
warm-cored thermal lows over the continents that bring strong
summer monsoon climate to the shores and mountains of our
world. Again, because of the weaker Coriolis Effect and the arrange-
ment of the landmasses, the Monsoon regime extends much fur-
ther towards the North Pole. All of the alien Americas, North
Africa and Asia are under its influence, bringing warm, summer
rains. To get an idea of climate think of Central China or the South
Western States. Cold and relatively dry winters followed by a short
mild spring, then a hot and fairly wet summer. Weaker circulation
around the North Pole generates frontal storms that afflict mainly
the most northerly latitudes. The polar front jet stream organizes
the polar circulation with Rossby waves partly anchored by the
Low Low
High
Low
High Low
Low
Low
North
Low Low Pole
Low
Low
Low
Low
High High
High
Low 45oN
Low
High
FIG. 10.12 The surface circulation during the summer on an Earth-like

world with a Mars-like Coriolis Force. Tropical Hadley Cells extend fur-
ther towards the north of the equator than on Earth. The northern extent
is shown by areas of high pressure centered at 5060 N. Large areas of
continent north of the equator heat strongly during the summer and gen-
erate broad areas of surface low pressure. These will be overlaid with
easterly jet streams. These lows drag in strong monsoon winds (black
arrows). A small Ferrel Cell is indicated nearer to the North Pole, indi-
cated by the presence of surface fronts. Tropical storms are indicated by
whirlpool-like symbols along with their likely tracks (red, arrows). Polar
Front jet stream (purple); sub-tropical jet (orange)
arrangement of mountains and oceans (Figs. 10.11 and 10.12). This

will lead to storm formation along what is northern Canada (and
indicated by a cold front on Fig. 10.12).
Despite the grouping of the continents hot deserts are likely
limited in extent due to the much stronger monsoon circulation
that brings moisture further inland. Much like the Miocene climate
of Earth, areas that are currently desert will most likely be open
savannah with summer rains (Chap. 2).
The southern hemisphere is largely open-ocean on our world.
The only land area (southern-central Africa) would experience
a monsoon climate, with hot, wet summers and cool winters.
However, unlike the northern continents, greater exposure to
the mid-latitude westerly winds in the winter would likely bring
some rain, much as it does to South Africa today. Over the south-
ern ocean, unimpeded westerly winds will bring a fairly steady
mild and wet climate, with relatively little seasonal variation.
Overall, the climate would most resemble the Eocene climate of
Earth with limited Polar chill and a broad tropical climate that
was mostly fairly wet (Chap. 3).
Tipped Over Worlds
Now lets go a bit crazy and tip our planet over, so that its rotation
pole is analogous to Uranuss. What would an Earth-like planet
experience now? Sadly, this arrangement isnt stable and tidal
forces would soon align the planets equator with that of its star,
making it much like the planet in Fig. 10.10. However, well take
the planet and place it around another K-class star at a suitable
distance and let it Uranian climatic sensibilities get to work. What
will the weather be like?
To keep things in perspective, weve stuck to our uber-
Earth with its super-continent organized around the North Pole.
This makes it possible to make a comparison with the other model
planets and with contemporary Earth. The orbit is still 130 days
long in an orbit four tenths the radius of Earths, just as it was for
our planet in the section above. The spin is kept at a third of ours
to give it a weaker, but still substantial Coriolis Effect. The main
difference is heating over the North Pole for roughly 42 days (the
northern summer); heating over equatorial regions for its 42-day
long spring and autumn; and heating over the South Pole during
the northern winter (southern summer).
Overall, strong heating on one hemisphere should lead to the
formation of a strong, surface thermal low over the North Pole or
South Pole during their summers (Figs. 10.13 and 10.14). However,
the relatively short seasons may prevent them from becoming as
a b
FIG. 10.13 The likely, overall circulation within the atmosphere of a

tipped Earth-like planet. Such a world has winds that blow from hemi-
sphere to hemisphere for half of the year. During the northern summer,
a strong area of low pressure will dominate the North Pole with winds
blowing from southwest to north east towards it. These winds begin
their life south of the equator at the opposite end of a large Hadley cell,
bending from south-easterly to south-westerly as they cross the equator.
Above the equator should lie a strong easterly jet stream, which persists
all year round. Observations of Uranus suggest a westerly jet stream is
likely over the mid-high latitudes in the southern hemisphere A similar
jet is found over the mid-high latitudes in the northern hemisphere dur-
ing its summerat least on Uranus. This flow completely reverse dur-
ing the northern winter. During spring and autumn (b). Sunlight strikes
equatorial regions most strongly giving rise to a circulation like that
found on Earth during its spring or autumn. The weaker Coriolis Effect
on our imaginary world might limit the number of cells to two per hemi-
sphere, rather than the three shown here. Any land will alter this flow
both by warming and cooling at a faster rate than neighboring ocean and
because of greater friction. Vertical extent of cell is exaggerated
strong as I suggest in the figures. That aside, this band of story

weather will migrate towards the equator from the south during
the autumn and away from it to the north during the hemispheres
spring. Hadley cells will dominate the overall circulation, but this
will be complicated by the spin of the planet, which adds a sig-
nificant Coriolis Effect. Winds would, therefore, be south easterly
to the south of the equator, then bend to form a south westerly
flow north of the equator. Winds would converge on the surface
thermal low over the northern continents that were closest to the
North Pole. For here, heating would be strongest and air would
High
High
Low
LowNorth
Low
Pole
Low
Low
45oN
High
High
FIG. 10.14 Summer circulation in the northern hemisphere of a tipped-

over planet? Strong surface heating generates a large area of low pressure
will develop over the northern continents, with a focus or foci on elevated
landmasses. Winds will circulate clockwise around this region but cross
lines of equal pressure more strongly than on Earth. Winds are south east-
erly south of the equator and south westerly to its north because of the
Coriolis Effect (black arrows). Above the equator an easterly jet stream
will persist all year round. As autumn comes on pressure will rise over
the poles and the zone of low pressure retreat equator-ward. Although this
model seems reasonable it doesnt match Uranus More details in text
be the most buoyant. With such a strong temperature contrast

between the sunlit and dark hemispheres and the modest Coriolis
Effect, winds might be rather strong. However, remember that this
is not true for tidally-locked planets where winds are expected to
peak at less than 50 km per hour: a modest gale. Moreover, surface
features such as coastlines, mountains or forests would further
weaken surface winds due to friction.
Thus, in the model, illustrated in Fig. 10.14, there is a rela-

tively simple circulation with winds bringing warm air from the
southern to northern hemisphere. There are two problems with
this. Firstly, the best match for such a world would be Uranus.
However, Uranus has a very different pattern of circulation (Fig.
8.2). Yes, it has a weak easterly jet over the equator, but winds are
westerly around each Pole. There are two possible explanations:
for one Uranus has a much stronger Coriolis Effect, with a rotation
of around 17 h; secondly, Uranuss weather is probably driven by
heat from its interior, rather than from the Sun. The first factor is
definitely the case, while the second is still uncertain. However,
a lack of solid surface which would heat up under its Sun and the
two factors suggested here, might allow for the difference.
Secondly, it isnt clear how the weaker, but still significant
Coriolis effect would influence the circulation on such a topsy-
turvy planet: in principle there should be 23 different cells per
hemisphere with the spin weve given the planet, here, much like
the Mars shown in Fig. 6.5. Circulation in the cold hemisphere
might, therefore, be broken up into multiple cells (as is suggested
in Fig. 10.13) rather than simply stream from Pole to Pole.
Weather-wise, obviously all seasons would be severe.
Summers in the northern hemisphere during its summer would
be brief but hot, and likely very wet along coastlines and over
mountains. The continental interiors would be hot and dry. The
transition to winter would be dramatic as winter itself. The sur-
face low pressure would peak in strength roughly a month after
the Pole received its maximum heating due to thermal inertia: the
lag in heating and cooling caused by the structure of the material.
However, not long after the low pressure reached its strongest, the
retreating Sun would drag the northern margin of the Hadley cell
back towards the equator. Reaching it 33 days later, the Sun would
drag the region of maximum precipitation back across the equa-
toragain with a lag of a couple of weeks. Meanwhile the north-
ern Hadley cell would uncouple from the North Pole, with its high
pressure edge also retreating south towards, but perhaps not mov-
ing beyond 60 N. However, while the southern Hadley cell was
shoved further south and eventually eliminated, the northern cell
would extend across the equator.
An increasingly large Hadley cell would extend southwards
towards the South Pole as this warmed up. To the north of the
Hadley cell, over the North Pole a large, cold cored low pressure
area might form, bringing first rain then snowfall on western
coasts and some distance into the interior. Finally, this too might
then be displaced south to around 7080 N while a small cold-
cored high pressure area takes up residence over the North Pole,
persisting until spring arrived 33 days later. However, equally, dur-
ing the northern winter, the pressure may just be predominantly
high across the entire hemisphere.
Higher up, as on Earth a westerly jet stream would likely
develop above the northern edge of the Hadley Cellthe analogy
of the sub-tropical jet stream. Meanwhile, as the zone of maximum
heating moves south towards the South Pole, pressure would rap-
idly fall until the southern thermal low appeared bringing its dose
of summer rainfall.
One interesting feature that we mustnt forget is the absence
of a continent over the South Pole. Weve lost Antarctica from
our alien Earth leaving open ocean. This will have profound
implications for our Uranian (or terran) mirror-world, for oceans
heat and cool more slowly than land. The brevity of each season
(33 days) will likely preclude the formation of polar ice caps,
although snowfall is a certainty over the planets northern ter-
ritories. Open ocean means a vigorous circumpolar circulation,
probably at all seasons, and even with strong heating in the sum-
mer. Thermal inertia in the oceansthe resistance of a material
to change its temperaturemay even prevent a warm-cored sum-
mer low from forming over the southern hemisphere. This would
have radical implications for the movement of air north and south
of the equator during the southern summer. Moreover, pressure
may remain stubbornly low in a belt around the South Pole with
westerly winds circulating around it. There is no contemporary
analogy that we know of.
What of the equatorial regions? While the Poles experience
nights and days 65 Earth-days long, the equator will have an illu-
minated part of the day that is never shorter than half the spin
period: 72 h. Therefore, the shortest day will have approxi-
mately 36 h of daylight. However, within the 130 Earth-day-long
year that our planet has, the Sun-lit portion of the day increases
twice to a maximum value of approximately 144 h, when the
Sun is overhead each Pole. The equatorial climate is the most
stable, with two rainy seasons per year that are associated with
the biannual crossing of the doldrums. This is the same as on
Earth, except that on this topsy-turvy world the doldrums
move Pole to Pole rather than Tropic to Tropic with the overhead
Sun (Chap. 1). Roughly overhead, but in the summer to the south
of the North Polar region will lie an easterly jet stream: easterly,
because the coldest air is going to lie to the south of the Pole.
This jet will retreat equator-wards during the autumn but likely
never reach the Pole (and disperse) because of inertia: the air
wont have time to heat strongly at height before the brief sum-
mer comes to an end. The jet will be strongest along the northern
march of the Sun; then weaken when warmer conditions lie to
its north in the early autumn. Shortly after the autumn equinox
the easterly jet will die out as it approaches the equator, with
little temperature contrast on either side of this geographical
divide. Later, as autumn continues in the north, a westerly jet
stream may emerge during the remainder of the southern spring.
This will follow the Suns continued march south towards the
southern Pole. With the coldest air to the jets north, but with
the jet over the southern hemisphere, this should also be of the
easterly type.
Atmospheric Gravity Waves

As a final thought, we will look at the effects of mountains on the
transport of energy and on the weather within a planets atmo-
sphere. Figure 10.10 showed one impact of mountains. Where
they are very cold, either all year round or seasonally, cold air can
drain down valleys and onto adjacent warmer surfaces. This not
only generates locally strong and potentially damaging winds,
but it will lead to the development of showers and so-called polar
lows where underlying warmer water triggers convection in the
overriding cold air. In Chap. 1, and here, we saw how mountains
can also trigger horizontal Rossby Waves within the atmosphere
that have a critical influence over the weather and climate. Such
impediments can also generate vertical waves in the atmosphere,
as can storms, which have an equally critical role in the formation
of atmospheric weather.
There are other waves which move in the vertical plane which
have regional influence, as well as acting as agents that deliver
energy from the troposphere to the lower ionosphere. These are
atmospheric gravity waves. Although the name may conjure up
images of colliding black holes and neutron stars, these gravity
waves are simply vertical motions within the atmosphere that
propagate like ocean waves. Atmospheric gravity waves are initi-
ated in three, broad settings: motion across mountains; motion
initiated by thunderstorms; or finally frontal motion. Figure 10.15
illustrates one of these. This atmospheric gravity wave had a
wavelength of several kilometers and moved, broadly, from south
FIG. 10.15 The passage of a long wavelength gravity wave across the Cor-
nish coast. A series of troughs and crests passed from south to north,
roughly perpendicular to an advancing cold front out to the west (photo-
graphs (a) through (c)). Small, blue arrows indicate air movement through
the waves. Bands of thick altocumulus castellanus gave thundery down-
pours interspersed with warm sunshine, as successive wave-fronts moved
overhead in an upper layer of unstable air. The lowest layer in the tropo-
sphere was stabilized by the cool ocean, beneath. Single-headed arrows
indicate direction of motion of the waves, while the double-headed arrow
indicates approximate wavelength. Photograph (c) is a panorama of three
photographs and is slightly distorted in the middle by camera motion.
Photographs by author
to north ahead of the cold front that likely initiated it. In this
wave an upper layer of warm unstable air, undulates up and down
over a layer of drier, more stable air. Clouds were thickest in the
troughs of the wave. Air is blowing in the general direction of the
wave, but the wave is propagating more rapidly through this layer
than the wind is blowing. This takes energy horizontally along the
boundary as well as vertically through the atmosphere above and
below the boundary that separates both layers.
Two additional, well-characterized forms of atmospheric
waves are associated with mountains and weather fronts: these
are illustrated in Fig. 10.16. In the top half of the figure, waves
develop as faster, less dense and more humid air advances over
slower moving, cooler, denser and drier air. Instabilities at the
interface cause ripples, much like the wind blowing over water.
This forms lines of altocumulus clouds perpendicular to the
Warm Front
cirrus
cirrostratus
altostratus
nimbostratus
stratus
b Mountain Winds Altocumulus Lenticularus

Wind Direction (flying saucer) clouds
Temperature Inversion
Rotor Clouds (cumulus)
Foehn/Chinook Wind
FIG. 10.16 Gravity wave clouds caused by air masses moving at differ-
ent speeds at different altitudes. In (a), warm air advances over colder
air ahead of a warm front and produces lines of altocumulus cloud. In
(b) air moves across mountains and undulates. In each case waves have
a wavelength measured in tens to hundreds of meters. In (a) undulating
air above the temperature inversion forms clouds, and in (b), roll clouds
form beneath it with altocumulus above. Photographs by author
direction of air flow. In the lower half of the figure, air crossing
mountains is forced to ascend rapidly. Where there exists a tem-
perature inversion, the air on either side of the inversion is driven
into a series of waves. These waves form roll clouds, where cooler,
but moist air periodically rises above its dew point causing con-
densation against the inversion. A few thousand meters above
the inversion, in the warmer air, altocumulus clouds can form at
higher elevations. Each of these lies above the lower roll clouds,
with cirrus and cirrocumulus forming higher still if the waves
extend high enough into the troposphere. Thus the entire depth of
the wave is illustrated by clouds.
Toshitaka Tsuda (Kyoto University) analyzed how such waves
propagate through the atmosphere. Until the 1970s it was thought
that west-east winds would strengthen with height all the way
into the lower thermosphere (Chap. 1). However, observations
showed that winds not only decreased in strength above 70 km but
that they often reversed direction. It has been successfully shown
that this attenuation and reversal of these mid-atmospheric winds
on Earth is due to the breaking of gravity waves. Such waves are
launched by motion across mountains or by storms. By the time
they reach the mesosphere they have wavelengths measured in
tens to hundreds of kilometers. Tsuda used ground-based radar to
monitor the propagation of such gravity waves that were associ-
ated with frontal storms. The waves measured by Tsuda penetrated
the polar front jet stream and continued unabated into the meso-
sphere. Waves that began with a vertical wavelength of 12 km in
the upper troposphere grew until they were 1015 km in size at
6070 km up in the mesosphere. At this height the waves began
to break apart as a combination of wind-shear and convection tore
the waves apart. Convection is caused by the steady drop in tem-
perature in the mesosphere with height that makes this region of
the atmosphere unstable.
In the stratosphere, winds blow towards the west during
the summer (an easterly jet stream) and to the east in the winter
(a westerly jet). Gravity waves, on the other hand move in all direc-
tions. However, this summertime easterly flow captures westward
propagating gravity waves and prevents them rising higher into the
atmosphere (Fig. 10.17). However, eastward expanding waves can
80
70
Mesosphere
60
Height (kilometers)
50
b
40 Polar Night Jet Stratosphere
(Winter only) a
Tropical
30
Easterly Jet
Sub-Tropical (summer)
Jet
20 Polar
Front Jet Tropopause
Arctic Jet (Seasonal)
10 Hadley Cell
Ferrel Cell Trade Winds
Horse Latitudes
90oN 60oN 30oN 0oN
Latitude
FIG. 10.17 Atmospheric gravity waves deliver energy to the mesosphere

from the troposphere. Atmospheric gravity waves (a) are launched by
thunder or frontal storms, or by motion across high mountains, where
air is forced to rise rapidly through stable upper layers. In the troposphere
and lower stratosphere waves can spread evenly east and west. Gravity
waves deliver energy through the lower stratosphere, growing in ampli-
tude as they rise. Stratospheric jet streams can block upward motion in
the direction opposing the motion of the jet (b), but they continue to
around 80 km elsewhere
rise higher until they break in the mesosphere. Here, these breaking
waves deposit energy into the layer at an altitude of 6090 km and
cause it to accelerate from west to east, driving the overall circula-
tion. Measurements showed an acceleration of 713 m per second
where gravity waves hit this layer. In the winter the opposite effect
occurs with westward propagating waves reaching the mesosphere
causing the winds to reverse direction. Thus, phenomena that are
seemingly insignificant on a planetary scale, such as tropospheric
storms or mountains that barely penetrate the depth of the tropo-
sphere, can in fact alter the motion of the atmosphere upwards to
the edge of space.
If all of this still sounds rather academic, then look at Saturn.

Measurements by Cassini show the clear signature of atmospheric
waves affecting the planets equatorial winds and stratospheric tem-
peratures. A clear downward propagating pattern of temperature
changes and alterations to the strength of the planets jet streams
is best explained by upward moving waves from the troposphere.
As these interact with easterly and westerly jets they cause the jets
to move downwards, towards the tropopause. Indeed, at present
Saturn is the only planet we know of, other than the Earth, that
shows this pattern. It would be interesting to find out is this kind
of energy transfer lies behind super-rotation of Saturns equatorial
atmosphere (Chap. 7). It certainly seems to tie with observations
of Venus, where the transport of momentum from the top to the
bottom of the atmosphere seems to lie behind super-rotation, here
(Chap. 5). Our understanding of this phenomenon will undoubt-
edly change once Juno starts its mission at the King of worlds;
and when better measurements come in from extrasolar planets.
As with the thunderstorms we looked at in Chap. 4, there exist
a plethora of connections between events on the ground and the
phenomena we observe at the top of our atmosphere and this must
be true for all planets. Weather is a phenomena not bound to the
clouds below.
Atmospheric gravity waves and Earthquakes
After the March 11th 2011 magnitude 9 earthquake, instruments

at Japans National Institute of Information and Communication
Technology detected concentric waves radiating outwards 300 km
above the Tohoku quake epicentre. The magnitude 9 quake and
accompanying tsunami generated atmospheric gravity waves that
extended upwards to the Ionosphere, where they altered its struc-
ture and conductivity. This observation opens up an interesting
avenue for researching geology on planets such as Venus, where
the surface cannot be observed. Strong ground shaking should be
evident through the waves this generates hundreds of kilometres
overhead.
Weather, Unbound: Brown Dwarfs, Stars,

Galaxies and Galaxy Clusters
When we think of weather we naturally think of planets.
Essentially this is true, but a number of meteorological phenom-
ena have found themselves applied to more esoteric objects some
hundreds of thousands of light years across. Here we find the idea
of weather reach its somewhat illogical conclusions.
Thinking first of brown dwarfs, the idea of weather is fairly
understandable. For although these objects begin their lives as hot
as the coolest, red dwarf stars, they soon cool down, once their sup-
ply of deuterium fuel has been exhausted. For most of these objects
this happens in less than ten million years, leaving a cooling orb.
Like the hot, young Jupiters, these objects go through a succession
of orderly changes in their atmospheres as they cool. Initially, with
temperatures above 2000 C, the atmosphere is filled with metal
oxides, drifting in an atmosphere of carbon monoxide, hydrogen and
helium. As temperatures fall, these compounds fall out until they
have precipitated to deeper, hotter levels where they can become
gaseous once more. This weather falls deeper and deeper within
the brown dwarf with time. As it does so other gases reach their
condensation points and emerge. First, compounds like perovskite,
then metal halides. When temperatures fall below 1000 C the car-
bon monoxide and hydrogen react to form water and methane and
the Brown dwarf morphs from an L to a T-class object.
Beyond 100 million years of age the brown dwarf cools down
to a few hundred degrees and its red hue vanishes beneath a sea of
cloud. Initially, these are composed of metal halides: compounds
such as sodium and cesium chloride. But as temperatures fall below
300 C or so, water vapor clouds take prominence and the brown
dwarf finally takes on the aspect of Jupiter. Perhaps one billion
years old, this dwarf has bands of water clouds in a sea of methane,
hydrogen and helium. Much older still and the water vapor cools
until it can form ice clouds. This is the status of the coldest brown
dwarf known (see earlier in this chapter). Over time, the water
vapor and ice clouds will be replaced by ammonia ice clouds the
methane ice clouds and aside from the stronger gravity, will leave
an object indistinguishable from our Jovian world.
What of stars and weather? In a loose sense, yes, they do have

it. Stars obviously have atmospheres that are affected by the energy
generated beneath their torrid surfaces. Cooler stars (red dwarfs,
orange dwarfs and yellow dwarfs) have fairly strong magnetic fields
that are generated within their outer layers. This leads to a variety
of interesting phenomena such as prominences and flares. Now,
although these are not standard weather the Sun does occasion-
ally generate something we might recognize on Earth. For example
in September 2015 the Sun generated a tornadoa rather large one
more five times higher than the diameter of the Earth. Now, as
tornadoes are generated from the topdown and this was clearly
generated from the solar photosphere, the term fire devil might be
more appropriate (Chap. 6). Convection, generated from the solar
surface combined with a generous twirl from twisted solar mag-
netic fields, whisked very hot plasma into the characteristic shape
of a tornado. The temperature, based on its spectrum, was a rather
fearsome 2.8 million degrees Kelvin.
The whole sequence lasted 2 days and delivered another batch
of energy from the solar interior to the corona above. How the Sun
accomplished this process has been something of a conundrum for
astronomers: energy is apparently transferred from the cooler pho-
tosphere to the hotter corona above. However, the reality is a little
more complex: energy is in fact transferred from the solar inte-
rior to the corona through a combination of magnetic waves and
shock waves. The tornado was simply another manifestation
of this energy transfer process, with its rapidly spiraling structure
carrying energy from the interior to the hot corona. Similar twist-
ers had been seen before but on this occasion the Solar Dynamic
Observatory (SDO) had a beautifully clear observation of the event
over its entire lifetime, giving astronomers the opportunity to see
how such storms develop and deliver their cargo of energy to the
suns outer atmosphere.
Solar Wind and Stellar Gales
The Sun is obviously hot and all hot objects radiate energy. Liquid
and gaseous objects also release particles because there will always
be a population of particles that have enough energy to overcome
the force of gravity. In the case of the Sun, the outflow of gas
from its surface blows at hundreds of thousands of miles an hour

(greater than or equal than 400 km per second). The solar wind
is divided into two components: the first is known as the slow
wind, with slow being very much a relative term. Wind speeds
are 400 km per second with a temperature of 1.4 million to 1.6
million Kelvin. This wind blows primarily from the Polar Regions
and from regions called coronal holes, where magnetic field lines
extend from the solar surface into interplanetary space. The gases
have a composition matching the corona and essentially represent
gases boiling away from it. Gases are fed into the coronal holes by
small-scale magnetic fields anchored in the Suns outer convect-
ing region. When these field lines reconnect they launch particles
which can then flow directly away from the Sun.
The second component of the solar wind is far more vari-
able in nature and blows at much higher speeds from the Suns
equatorial regions (within 35 of the equator). These winds carry
material from the photosphere and include gigantic bursts called
coronal mass ejections. Although wind speeds are higherup to
750 km per secondand their temperature is lower, at around
800,000 K. These fast winds are twice as dense as the slow wind
and have a chemistry more broadly matching the photosphere.
Because these winds tend to owe their existence to violent explo-
sions, from prominences, through flares to coronal mass ejections,
they are very turbulent and carry a lot more momentum than the
steadier slow winds.
Would you then feel the Solar wind blowing gently (or
fiercely) on your face, assuming you wouldnt die from exposure
to the vacuum of space? Not really. The term vacuum is impor-
tant. Now, obviously you cant have a wind in a vacuum: thats
an oxymoron. However, the gas is so rarified that even close to
the Sun where the wind is densest, the pressure the wind exerts is
only 1-6 billionths of a Pascal. Recall the Earths atmospheric pres-
sure is around 100,000 Pas at the Earths surface. Thats a differ-
ence of 15 orders of magnitude. Hence, the term vacuum is often
applied. Yes, strictly speaking no part of the volume of space is a
true vacuum, but with a handful of particles per cubic centimeter
youd be hard-pressed to think of it as anything else.
As for other stars, as all stars are by definition hot, they all
generate a wind. The red dwarfs generate the weakest winds, with
outflows similar or weaker to those of the Sun. More massive

stars generate proportionately stronger winds. These winds blow
fasterperhaps ten times faster, and carry up to a million times
more mass. Stars also vary the amount of mass they lose over time.
Very young stars tend to be more magnetically active than older
stars and various processes happening within their rapidly rotating
magnetic fields tend to launch stronger winds. So a red dwarf that
releases less than a trillionth of its mass every year when a few bil-
lion years old may have released 1000 times this amount per year
when it was first formed. We encountered the problems this may
cause for any planet that was orbiting closely earlier in the chap-
ter. Likewise, the Sun, and indeed all stars, expand when they age.
Red dwarfs expand the least and the smallest red dwarfs do this in
the middle of their lives. More massive stars expand greatlyper-
haps a few hundred times their original size. With their surfaces
far removed from their center of mass the pull of gravity is less-
ened. Moreover, older stars are brighter and the higher luminosity
blows gases more effectively. Therefore, both effects mean that
aging red giant stars can blow vast amounts of mass away from
their surfaces. Now, although the speed is reduced to only a few
tens of kilometers per second, the much greater mass (up to a few
tens of millions of times that of the Sun) have considerable effects.
The star is effectively blown away leaving only its core.
When the Sun dies and its wind finally subsides, will mark
the beginning of the end of interplanetary weather. For then, the
remaining planets will cool and their atmospheres progressively
rain out into their interiors. Trillions of years from now, each
planet that remains orbiting the cold, dark Sun will see one gas
after another condense then freeze. The process of atmospheric
collapse will mark the ultimate end of weather in our solar system
and presages the end of weather in the universe as a whole.
For now, the Suns hot breath impacts us in two ways. For
one it slowly erodes the Earths atmosphere but also contributes
to geomagnetic storms. For when those coronal mass ejections
hit our planets magnetosphere, the change in the Earths protec-
tive field causes part of it to collapse into our ionosphere. Thus,
the continual attrition of our atmosphere parallels the steady loss
of mass from the Sun. Everything loses out in the end. However,
each moment of loss is marked with aurorae that light up the
atmospheres of all the planets in the solar system bar Mercury.

Pluto? Well, its thin atmosphere might just glow green-blue when
the solar wind blows more strongly. We will need to wait for the
continual download of New Horizons data to complete to see if
this is apparent above its hazy blue firmament.
So, when your power supply fails in a geomagnetic storm, just
look at what the planet and its Sun lay on for compensation. We
may love our artificial sodium-neon glow; however, the blue-green
fluorescence of aurora has a far greater majesty than anything we
can create.
Galactic Winds
Galaxies that contain active star formation exhibit galactic

winds. In our galaxy these winds are modest and only blow gas
from the disc into the halo. This gas is expelled from the disc
by a combination of strong stellar winds generated by massive
stars and through the actions of supernovae. These winds form as
hot bubbles within the cooler gas of the disc. As this gas expands
upwards and downwards away from the galactic plane, it pushes
some cooler disc gas ahead of it.
As this gas reaches the halo of the galaxy it cools and rains
back into the disc. In the loosest sense, this movement of hot and
cold gas might be thought of as weather. However, if you are dis-
satisfied with this analogy then look at starburst galaxies, such
as M82. Here, furious rates of star formation generate winds hun-
dreds of times stronger than our galaxies puny breeze. Such winds
propel gas out of the galaxy and much of it into intergalactic space.
Chandra x-ray observations, coupled to Hubble observations reveal
cool streamers of gas shredding and blasting outwards away from
the galactic core.
Elsewhere in the universe, giant elliptical galaxies hide a
dark secret: a central super massive black hole. These dark lords
reveal their presence through their eating habits. The galaxies
they inhabit reside in a deep well of hot gas (below). This gas can
cool and stream into the core of the central cluster galaxy. As it
plunges in towards the central black hole, it gathers into a disc
around the central behemoth. Although 99 % of the gas is ulti-
mately consumed, magnetic fields within the disc can propel gas
outwards in long, thin jets at 90o to the disc: these jets point along
the magnetic poles of the black hole and its disc. This gas pum-
mels outwards at velocities between 60 and 99 % the speed of light
where it slams into and heats gas in the galaxy clusters surround-
ing medium. This turmoil generates regions where the temper-
ature and pressure within the gas changes over relatively short
distances. Perhaps we could think of these as fronts?
Cold Fronts in Galaxy Clusters
Taking this analogy somewhat further and explicitly using the

term front are meteorological features that are seen in the
centers of entire clusters of galaxies. Its been known for a couple
of decades that galaxy clusters contain very hot x-ray emitting gas
and that in some of these clusters there appear to be regions, hun-
dreds of thousands of light years long, where the temperature of the
gas changes by millions of degrees over relatively short distances
(hundreds of light years). We might well be pushing the boundaries
of what could reasonably be called weather, but these structures
do bear some resemblance to terrestrial cold fronts. For example,
as we cross the front temperature and pressure falls and there
is a large difference in the direction gas is flowing. Dont expect
thunderstorms or even a slight drizzle. Instead (if you wait long
enough a few tens of millions of years) there will be a reduction in
the amount of x-radiation and cosmic rays that you are exposed to.
How do such cold fronts form? Simon Ghizzardi (Istituto di
Astrofisica Spaziale e Fisica Cosmica) and colleagues analyzed a
large number of galactic clusters, containing up to a few thousand
galaxies, each orbiting the clusters mutual center of gravity. As
they sweep through the cluster they interact with very hot gas that
permeates the cluster in a number of ways. Some gas-rich galaxies
violently collide with the hot cluster gas, generating shockwaves
that speed through it. Gas that is compressed by the advancing
shock cools as its density increases, simply because the denser gas
can radiate heat more efficiently than the remaining rarified mate-
rial. These are the cold fronts in most cases. The vast majority of
clusters that show cold fronts have gas-rich galaxies plowing into
and merging with the cluster.
A few clusters dont show collisions but still show the pres-
ence of active cold fronts near the cluster core. In these clusters,
the galaxies are rather more sedate and not interacting as strongly
with one another. Here it appears that the fifty percent or so of
clusters with cold fronts have a very steep temperature gradient
from the outside to the center of the cluster. The hottest gas lies
furthest from the core of the cluster where gas is least dense. Near
the center gas is colder and denser: cold still means temperatures of
a few million degrees, but this is much less than the 2030 million
Kelvin gas found further away. The cold fronts form near the edge of
the coolest gas. Not precisely determined, it looks like cold fronts
form here when small galaxies approach the dense, cold core. As
they pass their gravity pulls on the gas causing it to slosh back and
forth near the edge of the hotter, surrounding gases. The advancing
boundary of this sloshing gas is a cold front. Does that mean there
is a warm front where the cold gas is retreating back towards the
cluster core? Although this use of front is somewhat unconven-
tional, it is certainly not unreasonable and pushes our concept of
what weather is to scales tens of times larger than an entire galaxy.
Conclusions
Understanding the atmospheres of exoplanets is perhaps the most
exciting area of meteorology there is. For not only does there now
exist a vast array of complex and highly diverse planets to exam-
ine, but there are also clear impacts on the understanding of our
weather. Exoplanets are the testing beds for our ideas on how
our planets atmosphere works. Even the most extreme planets,
the hot Jupiters, have a direct connection to the atmosphere and
weather on our world. Their super-rotating atmospheres mirror,
in the extreme, the tropics. In each case eastward moving Kelvin
waves drive super-rotation. On earth this is a very modest affair,
with bursts of westerly winds moving against the prevailing east-
erly trades, or on occasion assisting in the development of El Nio
patterns. On Hot Jupiters, with a far stronger push, Kelvin waves
drive super-rotation measured in hundreds of kilometers per hour.
More modestly, the effects of atmospheric tides on hot Jupiters
and other less strongly heated worlds are reflections of atmo-
spheric processes on Earth, Venus and Mars. Although our current
telescopic resolution is insufficient to probe frontal boundaries on
planets hundreds of light years away, soon that may be possible.
Then we can get a comprehensive picture of how these planets

atmospheres work and truly be able to give an accurate weather
forecast for a planet we formerly could merely dream of. For exam-
ple, the aging Kepler 452b paints a view of our planet in a billion
or so years time. Although we cannot quite probe its weather yet,
its age and location suggest its atmosphere is likely evolving from
one able to support life to one that is as hot and as torrid as mod-
ern day Venus. As we obtain more and more extensive libraries of
worlds we can map weather and climate in space and time.
While we extend what weather we can see, we also expand
the definition of what might reasonably be called weather. We are
more than happy to talk about the weather on Jupiter, but what
about a brown dwarf? After all these take on the aspect of Jupiter
with time, so where will you draw the line? Happy with brown
dwarfs having weather, then what about stars, groups of stars in
clusters or entire galaxies and groups of galaxies? If weather is
simply the motion of gases, then the universe has weather on all
scales from objects the size of Pluto through to clusters of galaxies
that extend over tens of millions of light years.
This chapter gives only the simplest introduction to what
is becoming an increasingly comprehensive picture of exoplan-
etary weather. With a bit of dreaming and a lot of hard graft plan-
etary scientists are developing increasingly sophisticated views of
worlds hundreds or thousands of light years beyond our cosmic
shore. And while the atmospheric scientists dream of worlds still
out of sight, cosmologists see weather on scales that we can only
imagine. This is an amazing time.
References
1. Visscher, et al. (2010). Atmospheric chemistry in giant planets, brown dwarfs, and low-mass dwarf
stars III. Iron, Magnesium, and Silicon. Preprint available at: arXiv:1001.3639 [astro-ph.EP].
2. Heath, M. J., Doyle, L. R., Joshi, M. M., & Haberle, R. M. (1998). Habitability of planets
around red dwarf stars. Preprint, available at: http://www.as.utexas.edu/astronomy/education/
spring02/scalo/heath.pdf.
3. Marley, M. S., Ackerman, A. S., Cuzzi, J. N., & Kitzmann, D. (2013). Clouds and hazes in
exoplanet atmospheres clouds and hazes in exoplanet atmospheres. Preprint, available at:
http://arxiv.org/pdf/1301.5627v1.pdf.
4. Loadders, K., & Fegley, B. (2006). Chemistry of low mass substellar objects. Preprint, avail-
able at: http://arxiv.org/ftp/astro-ph/papers/0601/0601381.pdf.
5. Hu, Y., & Yang, J. (2014). Role of ocean heat transport in climates of tidally locked exoplanets
around M dwarf stars. PNAS, 111(2), 629634.
6. Holmstrm, M., Ekenbck, A., Selsis, F., Penz, T., Lammer, H., & Wurz, P. (2008). Energetic
neutral atoms as the explanation for the high-velocity hydrogen around HD 209458b. Nature,
451, 970972.
7. Showman, A. P., Cho J. Y-K. & Menou, K. Atmospheric circulation of exoplanets. Retrieved
from http://www.lpl.arizona.edu/~showman/publications/showman-etal-exoplanets-review-
revised.pdf.
8. Liu, B., & Showman, A. P. (2012). Atmospheric circulation of hot Jupiters: Insensitivity to
initial conditions. Preprint, available at: http://arxiv.org/pdf/1208.0126v2.pdf.
9. Showman, A. P., Menou, K. & Cho, J. Y-K. Atmospheric circulation of Hot Jupiters: A review
of current understanding. Retrieved from http://www.lpl.arizona.edu/~showman/publications/
showman-etal-santorini-paper-submitted.pdf.
10. Holmstrm, A., Ekenbck, F., Selsis, F., Penz, T., Lammer, H., & Wurz, P. (2008). Energetic
neutral atoms as the explanation for the high-velocity hydrogen around HD 209458b. Nature,
451, 970972.
11. Swain, M. R., Vasisht, G., & Tinetti, G. (2008). The presence of methane in the atmosphere of
an extrasolar planet. Nature, 452, 329331.
12. Schinder, P. J., Flasar, F. M., Marouf, E. A., French, R. G., McGhee, C. A., Kliore, A. J.,
Rappaport, N. J., Barbinis, E., Fleischman, D. & Anabtawi, A.(2011). Saturns equatorial oscil-
lation: Evidence of descending thermal structure from Cassini radio occultations. Geophysical
Research Letters, 38, Article Number: L08205. http://onlinelibrary.wiley.com/
doi/10.1029/2011GL047191/full.
threshold for runaway greenhouse processes on Earth-like planets. Nature, 504, 268271.
14. Yang, J., Bou, G., Fabrycky, D. C., & Abbot, D. S. (2014). Strong dependence of the inner
edge of the habitable zone on planetary rotation rate. Preprint, available at: http://arxiv.org/
pdf/1404.4992v1.pdf.
15. Yang, J., Cowan, N. B., & Abbot, D. S. (2013). Stabilizing cloud feedback dramatically expands the
habitable zone of tidally locked planets. Preprint, available at: http://arxiv.org/pdf/1307.0515.pdf.
16. Tsuda, T. (2014). Characteristics of atmospheric gravity waves observed using the MU (Middle
and Upper atmosphere) radar and GPS (Global Positioning System) radio occultation.
Proceedings of the Japanese Academy, Series B, 90, 1227. Preprint available at: http://www.
ncbi.nlm.nih.gov/pubmed/24492645.
17. Faherty, J. K., Tinney, C. G., Skemer, A. & Monson, A. J. (2014). Indications of water clouds
in the coldest known brown dwarf. Preprint available at: http: //arXiv:1408.4671v2.pdf.
18. Wordsworth, R. (2014). Atmospheric heat redistribution and collapse on tidally locked rocky
planets. Preprint, available at: http://arxiv.org/pdf/1412.5575v1.pdf.
19. Zendejas, J., Segura, A. & Raga, A. C. (2010). Atmospheric mass loss by stellar wind from plan-
ets around main sequence M stars. Preprint available at: http://arXiv.org/pdf/1006.0021v1.pdf.
20. Hummocky terrain and volcano collapses. Retrieved from http://volcanoes.usgs.gov/observato-
ries/cvo/cascade_debris_avalanche.html.
21. New Horizons public access data. Retrieved from https://www.nasa.gov/mission_pages/
newhorizons/main/index.html.
22. Carone, L., Keppens, R., & Decin, L. (2015). Connecting the dots II: Phase changes in the
climate dynamics of tidally locked terrestrial exoplanets. Preprint, available at: http://arxiv.org/
pdf/1508.00419v1.pdf.
23. Leconte, J., Wu, H., Menou, K., & Murray, N. (2015). Asynchronous rotation of Earth-mass
planets in the habitable zone of lower-mass stars. Preprint, available at: http://arxiv.org/
pdf/1502.01952v2.pdf.
24. Ghizzardi, S., Rossetti, M., & Molendi, S. (2010). Cold fronts in galaxy clusters. Astronomy
and Astrophysics, 516, A32. Preprint available at: http://arxiv.org/pdf/1003.1051v1.pdf.
25. Tsugawa, T., Saito, A., Otsuka, Y., Nishioka, M., Maruyama, T., Kato, H., et al. (2011).
Ionospheric disturbances detected by GPS total electron content observation after the 2011 off
the Pacic coast of Tohoku Earthquake. Journal of Earth Planets and Space, 63, 875879.
Glossary
A
Acetonitrile A simple compound of carbon nitrogen, oxygen
and hydrogen, produced through the action ultraviolet light on
nitrogen, methane and water or carbon monoxide.
Advection The horizontal transport of energy by wind.
Aerosols Small particles that are produced by varying processes
in planetary atmospheres or in interstellar space. These include
spray from oceans or falling water feldspar dust from dry areas
and chemicals released by plants or plankton in the oceans.
Terrestrial aerosols also include bacteria viruses and other bio-
logical materials.
Altocumulus A type of mid-level (approximately 30005000 m
in altitude) cumuliform cloud composed of very cold water
droplets. These may be arranged in rows or streets lying in the
direction of the wind or at 90 to it.
Altocumulus castellanus A type of altocumulus cloud associated
with unstable mid- level air masses. These may produce showers
or develop into high base storms.
Altostratus Flat stratiform clouds composed of water droplets
that typically obscure but do not completely block out the sun.
These often form ahead of warm or occluded fronts.
Ammonia A compound of one nitrogen and three hydrogen atoms
found abundantly in the outer solar systems planetary atmo-
spheres and in interstellar clouds and comets.
Ammonium Hydrogen Sulfide A noxious compound of ammonia
and hydrogen sulfide found within the atmospheres of the giant
planets. Ammonium hydrogen sulfide smells of rotten eggs.

DOI 10.1007/978-3-319-25679-5
432 Glossary
Anthropocene The terms given to the latest geological period

beginning around 1700 and characterized by our propensity to
alter the atmospheric chemistry, temperature and climate, as
well as significantly alter surface and deep-level geology through
industrial processes.
Anthropogenic Global Warming The term given to the outcome
of putting additional greenhouse gases into a planets atmo-
sphere - causing more outgoing energy to become trapped and
the temperature to rise.
Anticyclone A region where air flows in a clockwise direction
in the northern hemisphere and anticlockwise in the southern
hemisphere. Anticyclones are characterized by calm conditions
on Earth with higher surface pressure. They are formed when
air descends from greater altitudes. They may have a warm or
cold core. Those with warm cores extend through the entire
troposphere while cold -cored high pressure areas eventually
give way to regions of lower than normal pressure at height.
Atmospheric Tide The bulk motion of a large portion of a planet or
its satellites atmosphere caused by gravitational pulling (Titan)
or day-night differences in heating (Venus Earth and Mars).
Aurora (aurorae/auroras) Luminous displays associated with the
ionization of gases in the high atmosphere. Charged particles
from the Sun arrive in waves during solar storms and these cause
realignments in the magnetic field around the Earth (or other
worlds). These reconnection events cause a cascade of particles
that sweep down into the upper atmosphere blasting electrons
off neutral atoms. As these fall back onto their charged particles
they emit light. The blue-green color is predominantly caused by
oxygen and nitrogen ions collecting electrons and emitting light.
B
Bacteriophage A type of virus that infects bacteria.
Ball lightning A rare form of lightning that involves some form
of organized and roughly spherical ball of plasma that is gen-
erated (most likely) by some mechanism involving lightning
discharges.
Glossary 433
Ballooning (spiders) The process by which (mostly) small spiders

can take off using small lengths of silk to catch the wind and/or
pick up electrical charge which causes them to repel from the
underlying surface.
Bergen frontal model The original model of frontal storm sys-
tems with an advancing warm front, closing cold front and
intervening warm sector. Developed in the wake of the First
World War this model successfully describes most frontal low
pressure areas that are found in mid to high-latitudes.
Blocking (blocking anticyclone) A block is an area of warm-cored,
high pressure that often becomes stranded on the cold side of the
polar front jet stream. Such high pressure areas then block the
normal west-east progression of frontal low pressure areas and
give rise to periods of steady and predominantly dry weather.
Blue jets (associated with lightning) Upward directed jets of
luminous gas that connect thunderstorms with the mid to
upper stratosphere.
C
Charge exchange (escape mechanism) An atmospheric escape
mechanism, whereby a fast moving ion exchanges its charge
with a less massive neutral atom, such as hydrogen or oxygen.
In the process the ion becomes electrically neutral and is accel-
erated. With enough kinetic energy it then escapes into inter-
planetary space. This escape mechanism is important in the
upper atmosphere of Venus, for example.
Chinook (Foehn/Fhn) A warm wind that blows down the lee of
mountains and (in winter and Spring) readily melts lying snow.
The Chinook is formed when warm, moist air cools on the wind-
ward side of mountains and latent heat is released as moisture
condenses and precipitates. As this, now dry, air descends in the
lee of the mountains, it warms at a faster rate than it cooled on
the windward side, because it is moisture-free. It is also warmer,
through the earlier release of latent heat. It thus reaches lower ele-
vations considerably warmer than it would otherwise have done.
Cirrus A type of high (typically above 5000 m), fibrous ice cloud.
434 Glossary
Cirrostratus A fibrous ice cloud that forms translucent sheets.

Cold front A dividing line at the Earths surface where cold air
advances into and displaces upwards warmer air. Cold fronts
fall into two classesKata and Anawith the former domi-
nated by descending, warming air and the latter by ascending
and cooling air.
Cold high An area of high pressure typically produced by air
cooling and rapidly descending. These occupy only the lowest
15,000 ft (4000 m) or so of the troposphere. Above, they gradu-
ally morph into areas of cold low pressure.
Cold low A mid-latitude or sub-tropical area of low pressure that
is cold throughout its depth and has low pressure throughout.
Conduction The process of heat transport that that involves
the exchange of particle energy through collisions or the direct
exchange of electrons between atoms that lie close to one
another.
Convection The transport of energy by the bulk motion of mat-
ter in a liquid or a gas that is driven by density differences.
Coriolis effect The outcome of any fluid (including gases) mov-
ing on a rotating surface. In the context of planets with atmo-
spheres, particles that move away from the equator towards
either pole are moving faster than the ground underneath and
must curve to the east. Particles moving from either pole to
the equator are moving slower than the underlying surface and
curve towards the west. The Coriolis Effect is thus dependent
on two factors the speed (or angular velocity) of the planets
rotation and its size. Small planet and those which rotate most
slowly experience the weakest effect.
Cryovolcanism The release of icy fluids and gases from vents or
fissures on the surface of a cold, icy body. These can be made
of ammonia, water, methane or any other moderately abundant
substance that is above its melting point.
Culminate Whereby a star or other celestial body reaches
the highest altitude relative to the observer when it crosses
Glossary 435
the imaginary line between the observers southern horizon

and the point immediately above his or her head. (Observers
northern horizon if they are in the southern hemisphere.)
Cumulus A bulbous cloud that is produced by convection and is
composed mostly of water droplets.
Cumulonimbus The tallest type of cloud with a typically low
base. The cloud is composed of water droplets through the lower
half of its bulk, but ice particles in its middle to upper levels.
Such clouds give rise to hail, heavy showers or thunderstorms
on all planets we observe them. On Earth and warm to hot plan-
ets, cumulonimbus clouds may have high bases where the air
feeding them is relatively dry.
Cyanide A compound of carbon and nitrogen (with one other ele-
ment), which is highly toxic to aerobic organisms but is a key
component in the non-biological formation of parts of DNA and
RNA. Found in Titans atmosphere.
D
Derecho Literally horizontal wind these are violent, gale to
hurricane force winds that blow outwards from the base of vio-
lent thunderstorms. These storms are largely self- sustaining,
once they are initiated, as long as there is a continued supply of
unstable air to drive the formation of thunderstorms.
Dry line A type of atmospheric frontal boundary between warm,
moist air and warm, dry air that forms in a few geographical
locations on Earth. In each there is a source of warm dry air
(preferably at some elevation) to the west of the source of moist
air. Consequently, dry lines most commonly form in places like
SW USA, eastern Spain and Argentina. They are central to the
formation of tornadoes.
Dust devil A type of swirling, convection-driven storm where
the air is dry and the column of swirling dust is generated from
the ground upwards.
436 Glossary
E
Earthquake lights (EQLs) A blanket term given to lights of
various sorts which appear to be associated with earthquakes.
However, although some lightssuch as those erupting from
the groundare clearly linked to tectonic strain, others have
more tenuous associations which may be coincidental.
Electromagnetic pulse (EMP) A burst of radio waves associated
with strong electrical dischargesor nuclear explosions.
El Nio A pattern of weather that emerges once or twice per
decade and is associated with the warming of the eastern and
central Pacific Ocean surface waters. Changes to the flow of air
in the tropical Pacific are followed by a realignment of mid-lat-
itude weather patterns. In most (but not all) years El Nio tend
to develop in the autumn and persist into the following spring.
El Nio Southern Oscillation (ENSO) The blanket term cover-
ing the movement of weather patterns from normal trade-wind
dominated tropical airflow to El Nio conditions and then
(often) to El Nios opposite pattern, the La Nia.
Elves (associated with lightning) A very faint, outward propagat-
ing pulse of light and other electromagnetic radiation associated
with an EMP (above) which accompanies a Sprite (below).
Eocene The geological period from 55 to 34 million years ago
that was associated with predominantly high temperatures and
a higher than present atmospheric level of carbon dioxide gas.
Euxinia Stagnant and largely anaerobic conditions, usually with
an abundance of hydrogen sulfide.
Exoplanet Any planet outside the solar system.
Exosphere The region at the top of a planets atmosphere where
gases can escape through different mechanisms.
Equatorial Rossby Wave (ERW) A pairing of alternating high
and low pressure areas which move from east to west on either
side of the equator. Unlike eastward moving Kelvin waves that
appear as fast moving changes in wind direction, on either side
Glossary 437
of a region of enhanced rainfall, ERWs appear as pairs of high

pressure leading pairs of low pressure with rainfall enhanced at
the leading edge of the low pressure regions.
Fall-streak (see Virga).

Ferrel cell A pattern of tropospheric circulation in the mid to high
latitudes of the Earth and Mars (but also apparent in the atmo-
sphere of Venus). This motion is induced by Hadley cell circula-
tion above the tropics. Air descending in the cool limb of each
Hadley cell displaces air along its poleward side causing it to move
in a counter flow towards the pole, along the planets surface.
Fire Devil A nasty incarnation of a dust devil, where strong heat-
ing above the ground, within a large fire, induces the formation
of a swirling column of plasma. This can ignite fires elsewhere.
Fire Tornado Fortunately, a very rare phenomenon on Earth, where
very large fires generate first cumulonimbus clouds, which then
generate tornadoes. These vortices then entrain fire and other
hot debris within their circulation, causing significant additional
damage and posing a major threat to fire-fighters. The most
closely documented example struck parts of Canberra in 2003.
Foehn (Fhn) wind (see Chinook).
Front A generic term that arose in the light of the trenches in the First
World War. Fronts mark the dividing line between air masses with
different properties, such as temperature, pressure and humidity.
Gap wind A strong wind that blows through mountain valleys.

In a classic case a front becomes trapped behind mountains,
but some of the air it leads can flow through valleys. A classic
example occurs in the Chivela Pass in Mexico where cold fronts
438 Glossary
moving south across the Gulf of Mexico come up against the

Sierra Madre mountains. Cold air eventually floods through a
gapthe Chivela Passand out into the Pacific Ocean.
Gigantic jets As the name suggested these enlarged versions
of the blue jets (above) reach the mesosphere from the top of
thunderclouds, rather than petering out in the stratosphere. The
exact composition is unknown but they are believed to share
some similarities with lightning.
H
Hadley cell The largest circulatory cell within the troposphere of
terrestrial planets. This cell (or rather pair of cells, in most cases)
is a convection cell driven by solar heating near the equator. Air
warms, becomes less dense and rises. From here it expands out
towards either pole. The Coriolis Effect (and increasing density)
eventually limits the poleward advance. Air then descends to
the surface and either returns to the region of rising air or blows
out across the surface towards the poles.
High pressure A region of higher than normal pressure caused by
subsidence of air from aloft.
Horse latitudes A belt of high pressure that forms the poleward
limit of the Hadley Cells. Here, air is relatively calm and the
weather cloud free. On Earth, the Horse Latitudes are advanc-
ing towards the poles at a rate of around 60 km per decade,
in a process almost certainly caused by anthropogenic global
warming.
Hydrodynamic escape (including hydrodynamic drag) A dynamic
process of atmospheric escape. Gases are heated strongly within
the atmosphere of a planet, either by stellar heating or tidal heat-
ing. This creates sufficient pressure to drive the gases out into
space. Typically, light gases such as hydrogen are most strongly
affected. However, these can drag other, heavier gases, such as
nitrogen or oxygen, with them. This is hydrodynamic drag.
Glossary 439
I
Impact erosion The removal of part or all of a planets atmosphere
following a collision or collision with asteroids or comets.
Insolation The term given to the total amount of radiation a
planet receives from its host star.
Isobar An imaginary line connecting points of equal pressure.
Isotope Isotopes are versions of elements that have differing
numbers of neutrons.
J
Jeans escape The name given to the process whereby particles
have a range of kinetic energies. Those with sufficient energy
can overcome the pull of gravity and escape the atmosphere.
Jeans escape is therefore proportional to the mass of the atom
or molecule and its kinetic energy.
Jellyfish clouds (see Virga) Fall-streak or Virga is a precipitation
of ice, typically from cirrocumulus or altocumulus clouds that
evaporates before reaching the ground.
Jets (associated with lightning) See blue jets and gigantic jets,
above.
Jet stream The generic term given to relatively narrow bands of
high wind speed within the atmosphere of a planet. These may
lie at varying altitudes and move at speeds varying from 10 to
several hundred meters per second. They are associated with
sharp differences in pressure and temperature.
K
Katabatic wind A fast moving type of mountain wind that
descends rapidly under its own weight from very cold upland
regions. These winds pour into valleys and can attain storm-
force. If they cross warmer ocean waters they can pick up mois-
ture and develop squalls.
440 Glossary
Kelvin wave A fast-moving westerly wave in the otherwise east-

erly trade winds. Easterly winds are fastest on its leading edge,
while strong westerly winds dominate to the rear of the advanc-
ing wave and rainfall band. These waves and Equatorial Rossby
Waves (above) are crucial drivers of super-rotation on hot, tid-
ally locked planets. Kelvin waves are also found in the oceans
and have similar roles in directing or reversing the direction of
ocean circulation.
L
La Nia The colder sibling of the El Nio, where waters across
the eastern Pacific become colder than normal as a result of
stronger trade winds. While the eastern and central Pacific
Ocean are colder than normal, the western Pacific is warmer
than normal.
Little Ice Age (LIA) A period of climatic variability extending
from 1200 AD to 1900 AD. In Western Europe the peak of the
period coincided with the frequent freezing of the Thames in
London. However, despite claims to the contrary the LIA was a
regional phenomenon with different regions experiencing cold
while others, simultaneously, experienced enhanced warmth.
Loess Fine, wind-borne silt-like deposits that are produced by
glaciation and then blown away from their source by regional
winds. Loess deposits can be hundreds of meters thick. Perhaps
the most extensive are found in central China where they con-
tribute to the coloration of the Yellow river.
M
Madden-Julian oscillation A periodic, long-wavelength atmo-
spheric disturbance that is found in the tropics. These waves
move from west to east, crossing the Indian then Pacific oceans.
As they do so they first enhance then reduce convective rainfall
over periods 6090 days long.
Glossary 441
Maunder minimum The somewhat infamous period during the

1600s where there appear to have been few significant and pos-
sibly no sunspots. Many pieces of work suggest a link to the
Little Ice Age; a link that falls when the global temperatures are
looked at, rather than focusing on those of Western Europe.
Medieval Warm Period (MWP) Another regional climatic event
where temperature in the North Atlantic basin, and southern
Greenland in particular, rose by 13 C compared to the 1960
1990 average. This excursion in temperatures appears to be
linked to a prolonged La Nia in the Pacific, which simultane-
ously cooled and dried out Meso-America. The tail end of this
period is associated with Chinas coldest temperatures of the
last 1000 years.
Mesosphere The region in a planets atmosphere above its strato-
sphere, where temperatures fall again, with height. Above the
Earth the mesosphere lies at 6090 km above the surface and is
associated with a variety of unusual meteorological effects such
as noctilucent clouds, Elves, Sprites and aurora.
Metallic hydrogen When hydrogen is compressed to 3 or more
million atmospheres -300 billion Pascals (and this is still not
confirmed) it should behave like an alkali metal. At present it is
assumed to be present in the interiors of Jupiter and Saturn but
the conditions needed to produce it have yet to be reproduced
on Earth.
Mei-yu/Baiu front A seasonal frontal band that marks the inter-
section of the mid-latitude westerly winds over Asia and the
advancing or retreating edge of the summer monsoon winds that
are blowing in from the Pacific. The northern side of the front
has cool, drizzly weather, while thunderstorms and showers pre-
dominate along its southern edge. Frontal lows may develop and
move northeastwards into the Pacific Ocean. A similar frontal
feature is occasionally visible along the eastern seaboard of the
United States where tropical air moves north along the coast in
the summer and impinges upon cooler westerly winds blowing
from the continental interior.
442 Glossary
Milankovi cycle A series of overlapping cycles that affect how

much solar radiation reaches the planet at different latitudes
over time. Cycles are related to the shape of the Earths orbit;
where the Earth is in its orbit during the northern winter and
the amount of tilt in the axis of the planet. Each of these cycles
varies in a predictable manner and evidence for its influence
in climate extends at least as far back as the Eocene. Here,
Milankovi cycles may have driven the catastrophic release of
greenhouse gases causing massive global warming. More con-
temporaneously they are largely responsible for the cycles of
glaciation and thaw we call ice ages. There is some uncertainty
over which cycles drive glaciation and why this appears to have
changed in the last million years.
Momentum A physical property of moving matter which is the
product of mass and velocity. Momentum is always conserved.
Angular momentum is a slight variant which takes into account
the motion of a mass around a center of rotation, with the distance
the mass lies from the center of rotation of equal importance.
Monsoon Seasonal winds that blow from cooler to warmer
regions. On Earth the summer monsoon blows warm, moist air
into the heart of Asia, bringing life-sustaining rains. On Mars
the southern summer monsoon brings dust storms.
N
Night-glow A pervasive but dim blue-green glow that is detect-
able in the visible and ultraviolet portions of the spectrum. It
is produced by the recombination of oxygen (and nitrogen) ions
with electrons high in the atmospheres of the terrestrial planets
(except Mercury, for obvious reasons).
Noctilucent clouds Rare, high (6080 km) ice clouds that form
when the mesosphere is at its coldest.
Non-thermal escape mechanisms A group of mechanisms
whereby particles obtain sufficient energy to escape the gravi-
tational clutches of their planet. These may involve chemical
reactions, collisions between ions and atoms, or the collision of
Glossary 443
cosmic ray particles with gas atoms and molecules high in the
thermosphere and exosphere. These mechanisms are "non-ther-
mal" because the particles involved do not have a continuum of
energies that would be found in a warm gas.
O
Ozone A molecule consisting of three oxygen atoms. Electrons
are shared between them in an unstable organization that
makes the molecule particularly suitable for absorbing a wide
range of ultraviolet wavelengths that might otherwise harm liv-
ing organisms on the ground beneath. Ozone, however, is toxic
if inhaled and is a major pollutant that is produced by combus-
tion engines.
P
Paleocene-Eocene Thermal Maximum (PETM) A period of sev-
eral tens of thousands of years in duration where the tempera-
ture of the Earth rose by 68 C on average, but perhaps 10 C at
the poles. The underlying cause was the release of large quan-
tities of carbon dioxide and methane - two potent greenhouse
gases. The underlying reason for the release may have been vol-
canic activity in the North Atlantic or the action of Milankovi
cycles.
Paleoclimate Any prehistoric climate.
Permian The geological period lasting from 298 to 252 million
years ago.
Permian mass extinction An abrupt period lasting 60,000 years
or so at the end of the Permian where 90 % of complex spe-
cies became extinct. The underlying cause was a period of
volcanism-induced global warming. Temperatures likely rose
1015C. The effect of heat was almost certainly exacerbated
by the release of copious hydrogen sulfide from warm, anoxic
ocean waters.
444 Glossary
Permian-Triassic boundary The narrow divide between the

Permian and Triassic periods where the mass extinction of life
occurred and its subsequent recovery began.
Polar cell The portion of the Earths or another planets atmo-
sphere on the polar side of the Ferrel cell.
Polar front jet stream The semi-continuous ribbon of high veloc-
ity air that streams in an easterly direction. The polar front jet
stream is the most visible manifestation of the Earths and Mars
Rossby waves that buckle and twist it as it courses around the
globe. The jet marks the interface of the mid-latitude Ferrel and
Polar cells.
Prograde In the context of this book prograde motion is the rota-
tion of a planet in the same direction as the planet orbits its star
(or a satellite orbits its planet). Thus the Earth and most other
planets orbit the Sun in a prograde manner (same direction as
the Sun rotates on its axis) and rotate on their axis in this man-
ner as well. (See, also, retrograde.)
R
Radial velocity method (of planet finding) As a planet orbits its
star it pulls on it. This the star moves back and forth in its orbit.
This, in turn, causes its spectral lines to Doppler shift back-
wards and forwards in manner proportional to the mass and
orbital period of the planet.
Retrograde A planet that orbits its star in the opposite direction
to the rotation of the star. Or a planet that rotates on its axis in
the opposite direction to its direction of motion around its star.
In terms of atmospheres these would blow around their planets
axis in the opposite direction to the planet rotated. Except in
specific regions or at specific seasons this is not seen.
Glossary 445
Rossby wave A series of intertwined waves of differing wave-

lengths which form in different locations around the earth and
other worlds. Rossby waves may be equatorial (above) or they
may form very visible structures in the Polar Front Jet Stream
(above). Rossby waves have eastwards and westward directed
components, with the waves that have the shortest wavelengths
moving in an easterly direction and bearing frontal storms in
the mid-latitudes.
S
Shapiro-Keyser model A revised model applying to some frontal
storms whereby the normal undercutting cold front is ripped in
two by a descending jet of cold air. The more equatorial portion
of the front then slices into the warm sector at speed, while the
central low is isolated on the polar side of this feature. The warm
front wraps around the core of the central low and warm air is rap-
idly bulldozed to altitude, causing the low to explosively deepen.
Several notable historical storms, such as the 1979 Fastnet storm
in the UK, or the 1993 superstorm in the US were of this type.
Solar wind stripping A non-thermal mechanism of atmospheric
loss whereby the atmosphere is progressively removed through
the action of particles in the solar wind colliding with and
removing them in the planetary exosphere.
Sprites (associated with lightning) Transient, red luminous phe-
nomena seen above the anvils of some thunderclouds following
the discharge of positively charged lightning bolts to the ground.
Sputtering Another non-thermal process, whereby energetic
cosmic ray particles collide with atoms and molecules within
a planets atmosphere. These may transfer sufficient kinetic
energy to overcome their planets gravitational pull.
Sub-tropical jet stream The ribbon of high velocity westerly
winds that marks the temperature belt edge of the Hadley cells
in the terrestrial atmosphere. The sub-tropical jets overlie the
Horse Latitude high pressure belts.
446 Glossary
Stratosphere The region in a planets atmosphere immediately

overlying the troposphere where temperatures rise with alti-
tude. On Earth the rising temperatures are caused by the absorp-
tion of ultraviolet radiation by ozone. This process reverses at
greater height where the density of gases is much lower. Here
the stratosphere gives way to the mesosphere.
Sumatra (storm) A type of tropical line squall that forms near east-
ern Sumatra during the months of April through to November,
during the southwesterly monsoon. Sumatras bring relatively
narrow bands of thunderstorms and strong winds as they move
north eastwards towards the coastal Malaysia.
Superbolt (lightning) A rare (five in ten million strikes) version
of lightning that comes from the anvil portion of a cumulonim-
bus cloud. Such strikes last ten times longer than conventional
negative strikes and may be associated with diverse electrical
phenomena such as ball lightning, sprites and elves
Supercell A type of long-lasting rotating thunderstorm that may
generate tornadoes, strong horizontal winds and extensive hail.
These storms usually form in the spring and summer in a few
restricted locations such as Tornado Alley in the US, eastern
and central Europe, Argentina and southern Australia.
Super-rotation The phenomenon whereby part of the atmosphere
moves faster in the same direction (has a greater velocity) than
the underlying surface or liquid interior. The phenomenon is
still fairly poorly understood and more than one mechanism
may underlie it. However, super-rotation is yielding to improved
modeling and its origin may well be understood on hot planets
that are tidally-locked to their parent star.
T
Temperature A measure of the average kinetic energy of parti-
cles within a substance. The greater the average temperature,
the greater the objects temperature. However, temperature is a
measure of the overall particle energy and many particles in an
object may have higher or lower kinetic energy. This explains
Glossary 447
how gases can sublimate from solids when the temperature is

far lower than the boiling point.
Thermal escape mechanisms A group of mechanism through
which gases can escape a planets clutches that are linked to their
overall temperature. The higher the temperature the greater the
overall kinetic energy of the particles and hence the greater like-
lihood that the particles will have a velocity in excess of the
planets escape velocity.
Thermal (warm) low A region of lower than average pressure at
a planets surface, where air is rising because it is warm and less
dense than surrounding air. Thermal lows morph into upper
level high pressure areas because the air is warm and the parti-
cles moving quickly. Therefore, they exert greater pressure than
the surrounding air.
Thermal wind Also known as a thermal tide, this is the flow of
air from the warm to the cold side of a planet caused by uneven
heating. Such flow is detectable in all of the terrestrial planets
that have an atmosphere and is particularly significant in the
hot atmosphere of Venus and the tenuous atmosphere of Mars.
Although detectable as pressure waves within the Earths atmo-
sphere, the thermal tide has a limited role compared with other
atmospheric processes.
Thermosphere The region near the top of all planetary atmo-
spheres where cosmic rays and stellar radiation heat the very
tenuous gases to high temperatures. The density of gas is so low
that the concept of hot is largely lost, despite temperatures mea-
surable in hundreds or thousands of degrees Celsius.
Tornado A swirling column of air that is generated by converging
and spinning air within a thunderstorm. In most cases the storm
is a rotating supercell. The tornado is visible in part because the
column contains a focus of low pressure which lowers the con-
densation point and also because such vortices kick up debris
from the ground.
Trade winds Belts of easterly winds that span the bulk of the trop-
ics on the earth, Venus and Mars. These winds move towards
the equator from areas of high pressure and blow towards the
448 Glossary
region of lowest pressure underlying the overhead Sun. Trade

winds should be present on all planets with moderate or greater
spin and a modest input of heat from their star.
Transit-timing variation method (of planet finding) If a planetary
system hosts more than one planet the gravitational pull of each
planet affects the orbital period of each. If one of these planets
is visible in its orbit by a transit, then the presence of a second
or third unseen planet (one that does not transit) can be inferred
from the effect it has on the time the observed planet takes to
transit the star.
Tropics An arbitrary name given to the region on a planets sur-
face that lies between the most northerly and southerly extent
of the overhead Sun. This is dependent on the tilt of the planet
so that, for example, on Jupiter the tropics are from 3 north and
south of its equator. This means that the cloud tops of Uranus
are all at some point within its tropics over the course of its 84
year orbit.
Tropical (equatorial) jet stream The tropical (easterly) jet stream
is a seasonal feature formed over Asia and Africa when the land
north of the jet is hotter than the air to its south during the
northern summer.
Tropopause The narrow transition region between the tropo-
sphere where temperatures fall with altitude and the strato-
sphere where they rise with height.
Troposphere A lowest region within a planets atmosphere,
where temperatures fall with increasing altitude. The tropo-
sphere tends to show the most dramatic weather, because it is
here where gases are densest and transport heat energy most
efficiently.
Turbulence Short wavelength disturbances that transport energy
rather chaotically through a combination of advection and con-
vection. Turbulence can occur when warm air overlies cold air
or when the temperature gradient in a gas or liquid is too great
for organized convection to occur. Turbulence will also appear
where air is forced over or around obstacles.
Glossary 449
V
Virga (See fall-streak)
W
Warm high A region of higher than surrounding pressure where
the air is relatively warm. These high pressure areas extend
throughout the depth of the troposphere and may block the
motion of other pressure features in the polar front jet stream.
By definition tropical high pressure areas are all warm cored.
Warm low (see Thermal low)
Z

Index
A F
Amasia, 150154, 225, 400402 Ferrel cells, 47, 199, 238, 239, 244, 247, 255,
Anticyclones, 3, 27, 51, 52, 75, 199, 200, 283, 406, 407, 409
286288, 301, 408 Foehn. See Chinook
Fronts, 410, 1519, 29, 37, 38, 45, 48, 57, 84,
162, 170, 199, 240, 241, 244246, 295,
C 338, 417, 426, 427
Chinook, 19
Cirrus, 79, 320, 333, 393395, 418
Cold fronts, 15, 17, 21, 37, 41, 43, 44, 48, 199, G
243246, 294, 296, 298, 408, 409, 416, Great Red Spot (GRS), 279, 280, 283, 285,
417, 426427 287, 289, 290, 300, 310, 322, 323
Cumulonimbus, 17, 39, 44, 162, 179, 181, 208,
210, 241, 243, 277, 288, 339, 393, 395
Cumulus, 79, 393, 395, 396 H
Hadley cells, 9, 23, 47, 67, 72, 78, 87, 119,
196, 197, 199, 225, 237, 238, 244247,
D 282, 332, 333, 335337, 377, 392, 397,
Derechos, 290302, 310, 338 406, 409, 411, 413
Dragon Storm, 300, 301 High pressures, 3, 24, 26, 2931, 40, 45, 47,
Dry line, 17, 20 73, 74, 76, 97, 263, 287, 323, 373, 374,
378, 379, 384, 387, 394, 401, 402, 407,
413, 414
E Hurricanes, 3, 4, 3638, 41, 42, 4649,
El Nio, 4, 29, 45, 55, 70, 71, 7678, 105, 107, 118119, 199, 287, 296, 402
374, 427
Eocene, 62, 63, 66, 82, 91105, 116, 117, 123,
124, 126, 127, 130, 131, 149, 157, 410 I
Eocene hyperthermals, 95, 124, 134 Ionopause, 200, 202, 203, 217
Equatorial easterly jet stream, 411, 412 Ionosphere, 13, 181, 184, 200, 202204, 252,
Exosphere, 14, 202, 252, 308, 350, 380, 381, 264, 307, 308, 315, 319, 326, 331, 416,
383 424

DOI 10.1007/978-3-319-25679-5
452 Index
J Polar front jet stream, 25, 27, 29, 33, 52, 58,
Jet stream, 11, 12, 1924, 26, 28, 30, 32, 35, 74, 76, 121, 280, 289, 310, 407, 409,
3947, 5153, 57, 58, 70, 74, 75, 81, 418
240, 284, 287, 289, 299, 301303, 312,
316, 349
R
Rossby waves, 2534, 39, 40, 4547, 51, 76,
K 237, 244246, 289, 302, 303, 310, 319,
Kelvin waves, 45, 46, 372, 373, 375, 378, 363, 374, 375, 377, 379, 386, 387, 402,
379, 427 407, 408, 415
L S
La Nia, 4, 70, 71, 7678, 105, 121 Saturns Great Spring Storm, 290, 295, 296,
Low pressure, 3, 15, 2729, 3941, 46, 52, 57, 299, 310
58, 69, 71, 7376, 97, 136, 246, 289, Stratosphere, 7, 11, 14, 23, 25, 84, 87, 179,
310, 373, 378, 379, 386, 387, 394, 395, 181, 182, 184, 198, 202, 204, 214, 215,
397, 399, 401, 402, 404, 407, 412414 220, 221, 232, 275, 315317, 319321,
330332, 338, 342, 377379, 381, 383,
392, 418, 419
M Sub-tropical jet streams, 24, 238, 407, 414
Madden-Julian oscillation (MJO), 45, 50, 52, Super-rotation, 28, 46, 196, 206, 284, 332,
78, 373 372396, 400, 420, 427
Magnetosphere, 14, 184, 209, 249, 305,
307309, 311, 324, 325, 344, 349, 384,
424 T
Mei-yu (Baiu) front, 37, 50, 51 Tethys, 62, 63, 6669, 96, 151
Mesosphere, 1214, 180, 181, 196198, 200, Thermosphere, 1214, 180, 181, 197, 198,
201, 204, 232, 254, 330, 331, 341, 342, 209, 233, 275, 307, 315, 319, 330, 331,
418, 419 342, 349, 350, 371, 380, 381, 418
Monsoons, 30, 3337, 39, 46, 49, 51, 52, 55, Tornado, 49, 164, 242, 422
5763, 6671, 75, 7985, 94, 119, 123, Tropopause, 8, 9, 11, 14, 23, 26, 34, 39, 47,
236, 237, 239, 244, 295, 299, 363, 397, 198, 199, 220, 221, 232, 268, 275,
406409 279, 286, 288, 290, 305, 314, 315,
317, 319, 331, 341, 349, 367, 386,
391, 420
N Troposphere, 4, 1014, 17, 23, 25, 29, 41, 57,
Nacreous clouds, 200 81, 84, 87, 113, 184, 195, 199, 200,
Noctilucent clouds, 12, 206 220, 221, 232, 233, 265, 275, 276, 288,
293, 314317, 320, 322, 330, 331,
335338, 342, 349, 350, 367, 377,
O 379381, 383, 393, 397, 416,
Occluded front (occlusion), 16, 18 418420
Ozone, 11, 101, 114, 161, 198, 201, 215, 221,
232, 255256, 269, 330, 381
U
Upper level front, 17, 19, 23
P
Pangea, 69
Panthalassa, 96, 97, 153, 401 W
Permian, 68, 69, 91105, 117, 123, 131, 149, Warm fronts, 15, 16, 18, 41, 297, 407, 417,
151 427

3319256777

Uploaded by

Copyright:

Available Formats

3319256777

Uploaded by

Document Information

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

3319256777

Uploaded by

Copyright:

Available Formats

David S.

More information about this series at http://www.springer.com/series/6960

The Exo-Weather Report

ISSN 1614-659X ISSN 2197-6651 (electronic)

Library of Congress Control Number: 2016943436

Springer International Publishing Switzerland 2016

Printed on acid-free paper

This Springer imprint is published by Springer Nature

My father fought in both World Wars. In 1915, aged 16 he joined

We are at a point that we have gathered a lot of data on the

that simple observations can provide some very useful truths. In

Nottingham, UK David S. Stevenson

1. What We Know About the Weather on Earth ................. 1

2. Climate Oscillations in Space and Time ......................... 55

Milankovi Cycles: Astronomical Influences

3. Tales of Mass Destruction ................................................ 91

4. Weird Weather ................................................................... 159

5. Venus ................................................................................. 191

6. The Wispy Weather of Mars ............................................. 231

Dust Devils, Tornadoes and the Other Whirling

7. The Gas Giants ................................................................. 273

8. The Ice Giants................................................................... 311

9. Ice Dwarves: Titan, Triton and Pluto .............................. 329

10. Tales of Other Worlds ....................................................... 363

Galactic Winds............................................................... 424

Glossary ..................................................................................... 431

Index .......................................................................................... 451

Why Do We Have Weather?

Springer International Publishing Switzerland 2016 1

A: Light striking a planets

Closer to the poles, the same

B: The depth of air energy

in part by absorbing and retaining some of the Suns energy, but

our planet. On the Earth, convection dominates the transport of

The Highs and Lows of Meteorology

Wind Direction: Waterwheels and a Suspect

the volume of gas increases as the temperature increases. And

because the ground underneath it is moving more swiftly towards

Unfortunately, after some tweaking of the trajectory, Paris

of energy: the ability of a force to move a mass over a distance.

The Vertical Structure of the Earths

as a jet stream (returned to later in this chapter). Thus, where as

a sign that rising global temperatures (Chap. 2) are driving more

able to experience it: the concentration of gases is so low that you

The Language of War-Fronts

During the First World War, and in it immediate aftermath, meteo-

Ana Warm Front cirrus

altostratus Ana Cold Front

Kata Cold Front

The frequently strong differences in temperature, humidity

Advancing cold front occlusion

Less cold air

Advancing warm front occlusion

Less cold air Coldest air

The air mass concept has its limitations, however. In particu-

arcing back northwards, the air is generally warmer than the

The Jet Stream

Between 7 and 16 km above the surface of the planet lie narrow

equatorial smoke band. The significance of this band of rapidly

with some of the coldest, most snowy weather in the winter

Stratosphere Sub- Tropical

oN oN oN oN