See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/226896960

Understanding Aviation Meteorology and Weather Hazards with Ground-

Based Observations

Chapter · January 2010

DOI: 10.1007/978-3-642-12968-1_9

CITATIONS READS

6 25,169

1 author:

Christian Page
Centre Européen de Recherche et de Formation Avancée en Calcul Scientifique
88 PUBLICATIONS   1,509 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Climator View project

AGU2020 Session: Digital Transformation of Earth Science Cloud: Data-Oriented and Self-Optimizing Architectures and Workflows View project

All content following this page was uploaded by Christian Page on 30 March 2016.

The user has requested enhancement of the downloaded file.

Understanding aviation meteorology and
weather hazards with ground-based
observations

Improving aircraft safety in terminal area

Christian Pagé

CERFACS, 42 avenue G. Coriolis, 31057 Toulouse Cedex, France

+33 (0)5 61 19 30 11
[email protected]

Introduction

Weather hazards have a significant negative impact on aircraft safety. Based on

recent surveys, 20 to 30% of worldwide air accidents are due to adverse weather
conditions. In Europe, as much as 22% of air traffic delays are due to bad weather.
The problem is that the latest trends estimate that air traffic will increase by a
factor of three within the next 20 years. If the accident rates stay the same, the
number of accidents will be increasing significantly and this is considered
unacceptable by airline and aircraft companies, and government agencies. To
mitigate these risks, several actions have been taken in the world.
In Europe, the FLYSAFE consortium is developing a mock-up of a system to
uplink, in real-time, weather information in the cockpit and to the Air Traffic
Controllers (Pradier-Vabre et al., 2008). In FLYSAFE and in other projects,
weather systems are being developed to identify weather hazards using weather
data fusion. These systems are trying to combine weather data from different
sources and instruments, such as ground-based and space-based observation
platforms in combination with meteorological numerical models.

Main weather hazards for aviation

There are several weather hazards that are dangerous for aviation. Because of the
different scales involved in weather phenomena, the hazards have not the same
impact if the aircrafts encounter them En-Route (Regional and Global scales) or in
1
the airport terminal area (Local scale). Furthermore, some phenomena are only
present either at local scale or at high altitude when the aircraft is En-Route.
Consequently, the main weather hazards can be tentatively classified as follows:
1. Airports Terminal Area: Local Airport Weather
• Wake Vortex
• Thunderstorms: microbursts, hail, wind shear
• Icing (waiting aircraft stacks)
• Low ceilings and visibility
2. En-Route: Regional and Global scales Weather
• Thunderstorms (hail, turbulence)
• Clear Air Turbulence (CAT)
• Icing (regional flights at lower altitudes)
To assess, identify and predict these weather hazards, observations are needed
from different instruments and sources.

The needs of ground-based remote sensing

observations to support aviation activities

Now, let's focus on the local airport terminal area weather hazards. In this section,
we will review weather data that is available for this scale.
The main sources of weather observations at the airport are from the standard
meteorological observation system. These provide data appropriate for synoptic
scale weather systems, e.g. at the scale of standard high and low pressure weather
systems:
1. METAR which gives surface data (temperature, altimeter setting,
humidity, winds), cloud types and bases as well as significant weather
such as precipitation, thunderstorms, etc.
• Data taken at 30 min. to 1 hour interval
2. Radiosoundings, which gives a vertical profile of temperature, humidity
and winds.
• Data taken at 12 hour interval
The limitation of these instruments is mainly that the data frequency is quite low,
and that they are point data. Other types of observation systems have greater data

2
frequencies, such as space-based satellites, which are remote-sensing observation
systems.
These space-based satellites can be categorized into two main families,
respectively geostationary and polar-orbiting. The geostationary satellites have
the following main characteristics and limitations:
• Orbit at 35 000 km altitude;
• Limited spatial resolution: 1 to 6 km;
• Typical temporal resolution is 15 min;
• Information is limited in the vertical;
while the polar-orbiting satellites have the following ones:
• Orbit at 700 to 800 km altitude;
• Very high spatial resolution: 250 to 5000 meters;
• Limited temporal resolution: 2 passes per day;
• Information is also limited in the vertical.
One of the main advantages of most of these remote-sensing instruments is that
their measurements are quasi-continuous. The space-based instruments have the
advantage of covering a large part of the earth, at the expense of either low spatial
resolution or temporal resolution.
For local scale weather needs at the airport terminal scale, specific ground-based
instruments can be more appropriate. Some are sensor-based, such as
meteorological towers which have instruments at different heights, and taking
measurements at very high temporal resolution, e.g. at each second. However,
these instruments also have the following problem: they give only point data and
their vertical resolution is quite limited. Other instruments, which are rather based
on remote-sensing technology, are:
• 1D and 3D weather radars (several band frequencies)
• Vertical Wind Profilers (several band frequencies)
• Radio Acoustic Sounding System (RASS)
• Lidar
• Sodar
• Radiometer
• Ceilometer

3
Applications of ground based remote sensing
instruments to aviation

In the previous section, several weather observing instruments that can be used to
mitigate weather hazards were briefly introduced. One conclusion that can be set
is that the standard meteorological observation system is not always suitable
because of limited time and spatial scales, as well as limited 3D coverage. On the
other hand, ground-based remote sensing systems are helpful to provide additional
weather information. However, remote-sensing data is often difficult to use
directly: data post-processing and/or data fusion is needed. In the next sub-
sections, practical applications on how remote-sensing based instruments can be
used to mitigate aircraft delays and accidents are described.
San Francisco Marine Stratus Initiative: Burn-off forecasts
In the U.S., low-ceilings and visibility are responsible for 35% of all weather-
related accidents in the U.S. civil aviation sector. They are also a major cause of
flight delays. To mitigate these problems, a project to develop an integrated
automated system was initiated by the MIT/LL. The first U.S. airport test site that
has been chosen was San Francisco: this became the San Francisco Marine Stratus
Initiative (Wilson, 2004). The objective of the project was to develop accurate
forecasts of the time that Marine Stratus will clear in the approach to San
Francisco Airport (SFO). The integrated system that has been developed was for
direct use by forecasters. The display integrated all the weather data available in
the airport area along with a consensus forecast of the burn-off time using several
numerical meteorological models:
• Statistical Model;
• 1D Column Model;
• Boundary Layer Model.
The instruments available were high temporal resolution surface data, sodar and
ceilometer, infrared satellite data and radiometers. An array of ground-based
instruments was deployed in the San Francisco Bay Area, especially at SFO and
San Carlos (Figure 1).

4
Figure 1: Array of ground-based instruments in the San Francisco Bay area for the San
Francisco Marine Stratus Initiative.
The sodar provides the boundary layer inversion height using acoustic
backscatter. It uses the fact that at the top of stratus clouds, there is a sharp
temperature inversion that gives high backscatter energy picked up by the sodar
(Figure 2).

Figure 2: Sodar estimation of the boundary layer inversion height using acoustic
backscatter.
The ceilometer provides the cloud bases of stratus clouds using the reflectivity of
very small water droplets.
The data of these two instruments can thus be used in combination with surface
and radiosounding data along with a 1D column numerical model to provide burn-
off stratus forecasts. The observation data is used to build a vertical profile of
temperature and humidity to initialize properly the column model (Figure 3).

5
Figure 3: Reconstruction of a vertical profile of temperature and humidiy using informations
from a numerical column model and observations. (Figure courtesy of Robert Tardif).

In summary, an fully integrated system approach has been used in this project: it
shows the potential of direct and indirect use of ground-based remote sensing
data. The system combined the use of sodar and ceilometer data along with
numerical modeling and an array of ground-based observations. The important
aspect was also that an integrated display was designed for use directly by
forecasters and Air Traffic Controllers.
The technology of this system has been transferred in 2004 to the U.S. National
Weather Service (NWS), and it is now running operationally: a great success!

Terminal Ceiling & Visibility in the northeastern U.S.

The success of the SFO Marine Stratus Initiative has led to a project follow-up:
the Terminal Ceiling and Visibility Product Development Team in the
NorthEastern (NE) US (Clark, 2006). The main problematic of this project is that
Instruments-Flight-Rules conditions are quite frequent in the NE US during winter
(November through April). These conditions are generally the consequence of
transient synoptic-scale situations, combined with high traffic. The physical
phenomena responsible for these conditions are quite various.
The project objective was to identify independent forecast technologies able to
give some skill to forecast these IFR conditions, in a terminal-centric approach in
high impacted terminals. The project focused on tactical-scale forecasts (0-3h),

6
which involve tracking and trending techniques, using high-resolution
observations.
Some techniques were developed to use 3D ground-based radar to detect low-
visibilities. This technique involved correlating reflectivity and surface-based
observations, along with cell tracking (Figure 4). The calibration of such a
technique (radar-visibility relationship) showed to be quite tricky, because
visibilities are quite dependent on precipitation type (snow/rain) (Figure 5). The
results showed that forecasting low visibilities is a non-trivial task, and data-
fusion techniques must be used to improve the accuracy of the methodologies,
such as using Wind Profiler, Sodar and Lidar information combined with 3D
ground-based radar data.

Figure 4: Particle trajectory for wind profile with no wind shear (Dixon et al., 2005).
Technique used to detect low visibilities.

Figure 5: Variation in observed surface visibilities during transition from rain to snow
during winter event at Boston Logan International Airport (Clark D., 2006).

7
Aircraft In-Flight Icing
Another meteorological condition that can be quite hazardous for aircraft is in-
flight icing, especially for regional flights. It causes reduced lift, raise stall speed,
imbalance and increased fuel consumption. In the US, there were almost 600
accidents in the last 20 years related to in-flight icing. This phenomenon is caused
by super-cooled water droplets. The main problems are that it cannot be measured
by remote sensing and that it is very difficult to predict (fast transient
phenomena).
One of the techniques that can be used to infer in-flight icing regions is to identify
the melting layer (0 C altitude) along with temperature and humidity. However, in
some conditions, there is no melting layer while there is some icing, such as in
deep convection and super-cooled rain.
Another information that can be used to identify icing regions (and determine its
intensity) is precipitation occurrence, which causes aerosol scavenging and an
increase in super-cooled water potential. In fact, when there are fewer aerosols,
especially the ones that act as ice nuclei, the water vapor is forced to condense as
water and cannot freeze because there are not enough ice nuclei. It has been
shown that precipitation scavenging dominates the aerosol removal process: it is
as high as 80%.
Other promising methodologies to diagnose in-flight icing involve a data fusion
process. The idea is to combine the information from different data sources
according to their potential to detect icing related features. Such a system was
developed at Météo-France, called SIGMA (Le Bot, 2004). Its first version used
data from numerical models (temperature, humidity), satellite infrared
temperature, along with ground-based weather radar reflectivity (precipitation
occurrence). This type of diagnostic system can be run in near-realtime, providing
diagnostics of aircraft icing potential (Figure 6).

8
Figure 6: Severe icing risk diagnosed using data fusion techniques. Figure is showing severe
icing risk objects around Charles-de-Gaulle Paris Terminal Approach (Figure courtesy of
Sébastien Geindre, Météo-France).
This type of system can also be extended to provide forecasts using one of the
following techniques:
• Extrapolating radar and satellite data (experimental) and perform
data fusion with temperature and humidity fields from numerical
model forecasts.
• Use a numerical model approach only, without using any remote
sensing data, such as in the CIP algorithm from NCAR (Bernstein,
2005).

3D Wind field Nowcasting

The 3D wind structure is very important to aviation safety and performance. This
is why a good knowledge of the current and future wind field is crucial for aircraft
activities. In the terminal airport area, the wind can threaten aircrafts when there is
low-level wind shear and turbulence, wave vortex or convective cells.
In the Terminal Airport area, there are several ground-based remote sensing
instruments that can be used to measure the wind structure. One of these
instruments is the ground-based weather radar. This can be achieved using either a
network of Doppler radars (at least three) or a bistatic network. A bistatic network
was built in Montreal, Canada, using the McGill Weather Doppler Radar (Protat
and Zawadzki, 1999). There were two bistatic receivers installed at proper
locations to provide good coverage over the Dorval Airport (Figure 7).

9
Figure 7: Schematic of the bistatic radar network at McGill Radar in Montreal, Canada
(Protat et al. 1999).
The three radial wind measurements can then be combined using a variational
data assimilation approach to compute a 3D wind field. This methodology has
also been extended to perform thermodynamic retrievals of pressure and
temperature perturbations.
Smaller-scale weather phenomena, like wake vortex, can be very dangerous for
aircraft and can cause accidents. They limit airport capacity because these
vortexes take time to dissipate or be advected away from the runway, thus
imposing a minimum aircraft separation. However, these vortexes are sensible to
atmospheric conditions such as crosswinds and atmosphere stability, which can
cause them to take a longer or a shorter time to dissipate. So, in theory, it would
be possible for aircrafts to have a shorter separation between them to increase
airport capacity if we can measure or calculate the time it takes for the wake
vortex to dissipate or be advected away.
One of the methodologies used is to install a ground-based remote sensing LIDAR
network around the airport, such as in the WAKENET Europe project (Paris
CDG, Frankfurt, London Heathrow) (Gerz et al., 2005) or in Memphis (1994)
(Dasey et al., 1997) (see Figure 8) or Dallas (2000) (Dasey et al., 1998). Lidar,
along with other instruments, can be used to measure horizontal winds that can be
used to determine a wake avoidance strategy to reduce aircraft spacing.

10
Figure 8: Memphis (1994) Wake Vortex project instruments.
Low-Level wind shear and turbulence
Important problems can arise from another type of 3D wind level structures, such
as turbulence and low-level wind shear. Turbulence can cause many problems to
crew and passengers, while low-level wind shear can cause crashes as it produces
trajectory differences compared to calculated trajectory. Low-level wind shear can
be caused, among others, by lee-side mountain peaks, updrafts near coasts (sea
breeze), thunderstorm gust fronts and microbursts, and low-level jet streams
(Figure 9).

Figure 9: Vertical wind shear: problems for aircrafts (Figure courtesy of Hong-Kong
Weather Service).
To mitigate the risks caused by these phenomena, a system based on ground-
based weather radar has been designed: the Terminal Doppler Weather Radar
(TWDR). It can detect wind shear and also microbursts. This system is designed

11
specially for terminal area in a high-clutter environment, and it includes
sophisticated clean-up algorithms to eliminate clutter and small moving targets
such as insects, birds, and other aircrafts. The wavelength used is 5 cm. For the
vertical wind profile, SODAR, LIDAR and vertical wind profiler can also be
used. The TDWR has a major problem though: it cannot be used when there is no
precipitation. Thus, to complement the TDWR, a system using one or several
LIDARs has been successfully operated using a wavelength of 2 microns. The
LIDAR operates better in rain-free (or fine rain) conditions. An integrated system
was developed by the Hong-Kong Weather Service for windshear and turbulence
alerts over various runway corridors at Hong-Kong International Airports (HKIA)
(Tsui et al., 2000).
Thunderstorm avoidance
One of the main weather hazards are thunderstorms. These storms are often
accompanied by almost all of the aircraft weather hazards, such as severe icing
and turbulence, hail, low ceiling and visibility, microburst and downburst, wind
shear, strong surface winds, heavy rain and lightning. Identification and spotting
of these storms almost always involve the use of ground weather radars. Many
algorithms have been developed to detect hail, heavy rain (reflectivity) and
outflow (Doppler radial winds and reflectivity). Nowcasting thunderstorm
movements is also very important: several cell-tracking algorithms have been
developed using ground weather radar data, which are usually based on
reflectivity thresholds and time cross-correlation. More sophisticated cell-tracking
techniques have also been developed, such as the MAPLE McGill Radar system
(Turner et. al., 2004).

Conclusion

Weather hazards have a significant negative impact on aircraft safety. The use of
ground-based remote sensing instruments can help to mitigate these hazards by
observing the weather phenomena associated with these hazards. By using data
fusion from several instruments or algorithms, it is possible to better identify and
nowcast them to improve airport terminal area security. Several important
application projects have been pursued in the last decade to tackle with these
problems, and the use of ground-based remote sensing instruments was almost
always included. Among many projects, the San Francisco Marine Stratus

12
Initiative, the Terminal Ceiling & Visibility project, and the FLYSAFE project
were briefly discussed. These extensive projects showed that it is possible to
develop algorithm that post-process and merge data from different sources,
especially from quasi-continuous data sources such as remote sensing instruments.
By combining data from different instruments, one can capitalize on strengths of
each of these instruments to overcome their individual defaults. Results have
shown that ground-based remote sensing instruments is useful to improve security
in airports terminal area.

References

Bernstein, B.C., F. McDonough, M.K. Politovich, B.G. Brown, T.P. Ratvasky,

D.R. Miller, C.A. Wolff and G. Cunning, 2005: Current Icing Potential (CIP):
Algorithm description and comparison with aircraft observations. J. Appl.
Meteor., 44, 969-986.
Clark, D. A., 2006: Terminal ceiling and visibility product development for
northeast U.S. airports. Preprints, 12th Conf. on Aviation, Range, and Aerospace
Meteorology, Atlanta, GA, Amer. Meteor. Soc.
Dasey, T., Campbell, S., Heinrichs, R., Matthews, M., Freehart, R., Perras, G.,
Salamitou, P., 1997: A Comprehensive System for Measuring Wake Vortex
Behavior and Related Atmospheric Conditions at Memphis. Tennessee, Air
Traffic Control Quarterly, 5(1), 49-68.
Dasey, T. J., Cole, R. E., Heinrichs, R. M., Matthews, M. P., Perras, G. H., 1998:
Aircraft Vortex Spacing System (AVOSS) Initial 1997 System Deployment at
Dallas/Ft. Worth (DFW) Airport. Project Report NASA/L-3, MIT Lincoln
Laboratory, Lexington, MA.
Dixon, M. and R. Rasmussen, S. Landolt, and J. Simard, 2005: The
implementation of short-term forecasting of airport surface visibility using ASOS
and radar. Progress report submitted by NCAR to FAA, 31 June 2005.
Gerz, T., Holzäpfel, F., Bryant, W., Köpp, F., Frech, M., Tafferner, A.,
Winckelmans, G. 2005: Research towards a wake-vortex advisory system for
optimal aircraft spacing. Comptes Rendus Physique, 6 (4/5), Special issue on
Aircraft trailing vortices, 501-523.

13
 Le Bot, C., 2004: SIGMA: System of Icing Geographic identification in
Meteorology for Aviation. Book of abstracts of 11th Conf. on ARAM, AMS,
Hyannis, MA.
Pradier-Vabre, S., C. Forster, W. W. M. Heesbeen, C. Pagé, S. Sénési, A.
Tafferner, I. Bernard-Bouissières, O. Caumont, A. Drouin, V. Ducrocq, Y.
Guillou and P. Josse, 2008: Description of convective-scale numerical weather
simulation use in a flight simulator within the Flysafe project. Meteorol. Atmos.
Phys., Online First, DOI 10.1007/s00703-008-0317-4.
Protat, A. and I. Zawadzki, 1999: A semi-adjoint method for real time retrieval of
three-dimensional wind field from multiple-Doppler bistatic radar network data. J.
Atmos. Oceanic Tech., 16, 432-449.
Tsui, K. C., C.M. Cheng and B.Y. Lee, 2000: Remote-Sensing the Weather to
Support the Hong Kong Airport. The Third International Symposium on
Electronics in the Air Transport Industry, Hong Kong, China, 28-29 November
2000.
Turner, B., I. Zawadzki, and U. Germann, 2004: Scale dependence of the
predictability of precipitation from continental radar images. Part III: Operational
nowcasting implementation (MAPLE), J. Appl. Meteor., 43, 231-248.
Wilson, F.W. 2004: An operational Marine Stratus forecast system for San
Francisco International Airport. 11th Conference on Aviation, Range, and
Aerospace Meteorology, Hyannis, MA, Amer. Meteor. Soc.

14

View publication stats