WMO Aircraft Observations: http://www.wmo.int/amdar/  ¦  WMO AMDAR News and Events: http://www.wmo.int/amdar-news-and-events

WMO AMDAR Panel Newsletter News Group: https://groups.google.com/a/wmo.int/forum/#!forum/wmo-aircraft-observations-news

RSS News: https://sites.google.com/a/wmo.int/amdar-news-and-events/latest-news/posts.xml
RSS Events: https://sites.google.com/a/wmo.int/amdar-news-and-events/events/posts.xml

Volume 4, October 2012

Message from the AMDAR Panel Newsletter Editor

Dear Reader,

Welcome to the fourth edition of AMDAR Panel Newsletter issued by the World Meteorological Organization AMDAR Panel.  I hope you find the following articles both interesting and informative.

As AMDAR matures and increases its role in the WMO Integrated Global Observing System (WIGOS), the impact of this data on forecast/watch/warning and NWP applications as well as airline operations will continue to grow.  This newsletter provides an outlet for informing the meteorological and aviation communities about the status of AMDAR programs worldwide, what plans are being made to expand data coverage and capabilities, and examples of current data usage.  By keeping AMDAR in the forefront, we hope to stimulate NMHSs to consider beginning their own national/regional programs.  

In this edition, you will read about:

  • The EUMETNET AMDAR Optimization System (E-ADOS)
  • The status and future of the Water Vapor Sensing System
  • The latest version of the ARINC 620 data reporting format
  • The growth in the number of AMDAR observations
  • The future of E-AMDAR
  • A trial of high resolution E-AMDAR observations
  • A look ahead to the 15th AMDAR Panel meeting and technical workshop on turbulence, and
A feature article by Dr. Ralph Petersen summarizing the results from recent studies on the impact AMDAR data have on NWP forecasts.

The AMDAR Panel, its Chairman, Vice Chairman and Technical Coordinator welcome any questions or comments on the global AMDAR program and, in particular, any interest in the development and coordination of new national programs.  The names and contact details can be found at the end of this Newsletter.


Carl Weiss, AMDAR Panel Vice-Chairman and Newsletter Editor

Aircraft Observations Output

The graphic at left shows the smoothed, monthly average of daily observations (single point measurements made by an aircraft in space and time) transmitted on the World Meteorological Organization (WMO) Global Telecommunications System since 2003 up to September 2012, as contributed by:

- All aircraft and all systems (blue);
- From the AMDAR Programme with reports submitted in binary format (BUFR, red);
- From the AMDAR Programme with reports submitted in text format (FM42, green); and,
- From ICAO data sources (AIREP and ADS, magenta).

It can be seen that in recent months, total aircraft-based meteorological observations levels remain up at around 300,000 observations per day.

WMO and the AMDAR Panel express gratitude to our aviation industry and airline partners for their continued contribution to the WMO Aircraft Observing System and the AMDAR Programme, that produce these data for utilization within meteorological applications and forecasts that benefit aviation operations and safety.

AMDAR Panel 15th Session and Technical Workshop on Turbulence

WMO AMDAR Panel 15th Session, 6 to 9 November 2012, Boulder, USA

Since its formation in 1998, the AMDAR Panel has been responsible for the development and management of the AMDAR Programme and is comprised of WMO Member National Meteorological and Hydrological Services operating under the Terms of Reference of the AMDAR Panel. For more information visit the WMO Aircraft Observations site.

The Panel has since met on an annual basis, often in countries and locations that have expressed an interest in developing an AMDAR Programme and might benefit from the interaction of local experts and aviation representatives with Panel Members. On several occasions, Panel meetings have also been held conjointly with technical workshops on AMDAR related systems and science.

The Fifteenth Session of the WMO AMDAR Panel will take place over the 6th to 9th November 2012 and will be kindly hosted by NOAA and the National Center for Atmospheric Research (NCAR) at the NCAR Center..

Technical Workshop on Turbulence, 5 November 2012

In conjunction with and immediately prior to the AMDAR Panel meeting, on the 5th of November, the AMDAR Panel, in cooperation with the Federal Aviation Administration (FAA) and the National Center for Atmospheric Research (NCAR) will hold a technical workshop on turbulence, with a particular focus on the monitoring and reporting of turbulence from the AMDAR system.

More details are available from the WMO Aircraft Observations Meeting site.

ARINC 620 Meteorological Report Updated to Version 5

The WMO AMDAR Panel, guided by the expertise of several Members and the WMO Secretariat have recently worked within the procedures of the ARINC Airlines Electronic Engineering Committee (AEEC) DataLink System Sub-Committee on the specification of Version 5 of the (AMDAR) Meteorological Report within Supplement 7 to  the Data Link Ground System Standard and Interface Specification (ARINC 620-7).

The Meteorological Report and the Meteorological Report Command Uplink (within Chapters 4 and 5 of ARINC 620-7 respectively) along with the programming notes and compression technique in Appendix G and H respectively, together provide the uplink and downlink data and message protocols, in the form of Aircraft Communications Addressing and Reporting System (ACARS) standards, that enables the automated reporting of meteorological data, which characterizes the AMDAR (Aircraft Meteorological DAta Relay) Programme.

Meteorological Report, Version 5 incorporates the following additional functionality or modifications to Version 4: 
  1. For the Ascent Data, the Series 1 duration (seconds), the allowable maximum range has been extended from 200 to 600 seconds and the Top of Climb allowable maximum range from 250 to 350 (ftx100). 
  2. For the Descent Data, the Descent Interval allowable minimum range has been extended from 20 to 10 seconds and the Top of Descent allowable maximum range has been extended from 250 to 350 (ftx100). 
  3. The Turbulence parameter has been added to both series of the Ascent Report. 
  4. The data width of the turbulence parameter, Eddy Dissipation Rate (EDR), has been extended from 4 characters to 8 characters. 
  5. The Time Message Assembled parameters in all reports have been extended to a resolution of seconds 
  6. The resolution of the Time of Observation parameter has been increased from minutes to seconds. 
  7. The resolution of reporting of Latitude and Longitude has been increased from tenths of minutes to 1 second for all reports in all flight phases. 
  8. Global Navigation Satellite System (GNSS) Altitude has been added as a parameter to be reported in all flight phases. 
  9. The resolution of the Water Vapor (WV) /Humidity parameter has been increased. 
  10. The parameter Icing has been added to report the status of ice accretion within the Meteorological Report. 
  11. The parameters, True Airspeed, True Heading, De-icing and Aircraft Configuration have been added to be optionally reported within two parameter groupings to facilitate enhanced quality control of other meteorological parameters. 
  12. The use of the Roll Angle Flag has been modified to provide an alternative mode in which the aircraft roll angle can be reported as lying within a series of range bins.
If you would like more information on the AMDAR Programme or the ARINC 620 Meteorological Report for ACARS, please contact one of the people via the details provided at the end of the Newsletter.

What is the future of the E-AMDAR?

The E-AMDAR is parented by the organization EUMETNET (www.eumetnet.eu) which is an Economical Interest Group – EIG – with 29 European National Weather Services as members. EUMETNET finances a number of programmes, some of them observational like E-AMDAR, E-ASAP (radiosonde launches from commercial ships) and SURFMAR (moored and drifting bouys as well as observations from voluntary ships).

The programmes are defined for periods of five years and each programme is run by a national metservice as coordinating member on behalf of all the EUMETNET members. The current programme period for the observation programmes  is coming to its end by the end of 2012. Therefore, during the past winter and spring the requirements for the next programme period 2013-2017 were defined and in the late spring EUMETNET published a document describing the requirements for all observation programmes. Member institutes were invited to submit bids by the end of August at the latest on the programme(s) they were interested in coordinating.

After this the bids have been evaluated by designated teams and the result will be discussed first week in October by the EUMETNET Scientific and Technological Committee and then the Assembly will decide at its meeting in late November which metservices will assume the responsibilities of coordinating programmes for the new period.

So, what will happen to E-AMDAR? One thing is certain, my institute the Swedish Meteorological and Hydrological Institute (SMHI) will not continue as coordinating member. SMHI has had the task for 10 years (first carried out by Ture Hovberg and then by me). As I will soon retire from active duty and not continue as Programme Manager, SMHI decided to not bid for a third period. And in late November we will know who is “the lucky winner.” Yes, I say lucky winner because this job is great!  It offers great personal/professional development and stimulation along with running and developing the programme itself! So many interesting things I have learnt and many interesting and helpful colleagues I have met in the AMDAR and aviation communities!

What we know so far is that the nearest future we will see E-AMDAR both accelerating and retarding at the same time.  In 2013 we will equip nine Lufthansa A319 with the humidity sensor WVSS-II v3. This has obviously to be carried out stepwise, but by the end of 2013 they will all be in operation! On the other hand the preliminary budget for 2013 indicates an 8 per cent reduction of funds for data acquisition, and some reductions also for infrastructure and management costs.

Stig R. Carlberg
E-AMDAR Programme Manager

The EUMETNET AMDAR Optimization Program (E-ADOS) - A Component of the E-AMDAR Programme

Aircraft Meteorological DAta Relay (AMDAR) has become the generic name for all systems of automatic meteorological reporting from aircraft.  Over the years, several AMDAR programs have been established to provide MET offices with upper-level meteorological observations.

The number of AMDAR observations being used operationally in global numerical prediction models has increased from fewer than 10,000 in 1994 to more than 200,000 per day in 2012 and this number still is growing. In order to reduce the number of duplicated and unnecessary AMDAR reports while maximizing data coverage at minimum cost, an optimization system for the selection of aircraft to report weather data, as defined in requirements set by national MET Services, has become an important element of the AMDAR system.

On behalf of EUMETNET AMDAR and the German MET Service (DWD), an optimization system for the AMDAR aircraft fleet was developed several years ago. This system is designed to improve the efficiency and cost-effectiveness of AMADR reporting and, although currently hosting Lufthansa, Finnair and KLM fleets, is available to all airlines.

The system selects the most recent flight which best meets the stated MET requirements at each requested airport. This airline-independent selection of the most suitable flight makes the AMDAR data production less prone to airline issues, e.g., flight delays or cancellations, labor interruptions, etc.  Such an optimization scheme preferably should not be used independently for each individual AMDAR airline, but several airlines can be optimized by the same system.  If more than one optimization system is in operation, it is necessary that the systems are compatible and are able to exchange information freely.

The highly sophisticated EUMETNET AMDAR system performs optimization following defined basic rules and options. The following list provides a brief overview of possible criteria and options:

  • Selection of the most suitable flight independent of the airlines by anonymous aircraft and flight identifiers (no airline preference)
  • Selection of priority of ascent and/or descent profile production for selected airports and time interval
  • Selection of en-route data production for short- and long-haul flights or specific city pairs, countries or ICAO regions
  • Definition of areas for target observations including separate aircraft configuration
  • Special aircraft configurations
  • Special aircraft configuration for each individual airport
  • Generation of the trigger requests and transmission to the airline data link system.
This AMDAR optimization system offers an opportunity for MET Offices to request targeted observations from airports as well as to define specific aircraft for special tasks e.g., humidity measurements.  Furthermore, the system maintains a list of aircraft which should be excluded from optimization due to inaccurate measurements.

A special en-route configuration is an option to reduce AMDAR reports on routes with high AMDAR traffic.

The optimization of the AMDAR fleet is based on the aircraft movement messages and together with some additional assumptions, e.g., taxi out time, the best fitting aircraft are selected.

The figure illustrates the effect of the optimization in 2001 when E-ADOS became operational.  At this time, the number of AMDAR reports from the optimized AMDAR fleet aircraft dropped significantly. Although the number of aircraft reporting AMDAR grew in the following months, the number of AMDAR reports was stable or decreased.

High Resolution E-AMDAR Data Trial

The E-AMDAR Programme was asked to carry out an increase to reporting resolution in the profile phase (ASC/DES) as part of a High Frequency Trial (HFT) – requested by the meteorological scientific community (E-SAT Group).

The HFT utilised the capabilities of the E-AMDAR Data Optimisation System (E-ADOS) which allows for reporting resolution to be increased at selected airports by remote web access.

Figure 1 (click to enlarge) shows the “normal” reporting resolution and Figure 2 shows the increased “high frequency” resolution.

Fig 1 – Normal E-AMDAR configuration settings in E-ADOS.

Fig 2 – High Frequency configuration settings in E-ADOS.

The HFT was carried out using a phased approach. Phase 1 consisted of one aircraft being configured for high frequency reporting on 6th May 2012. This was to ensure that there were no processing issues with the additional data generated.
No problems were encountered and Phase 2 was initiated. This consisted of six airports being configured for high frequency reporting by visiting aircraft during period 10:00UTC 17th May 10:00 to 14:00UTC 18th May 14:00, listed in Table 1.




Paris (CDG)



London (LHR)


United Kingdom













Table 1 – List of airports targeted for high frequency profile reporting.

During the trial period, 40 profiles with increased data were generated. The two graphs below show the comparison of data between a “normal” and “high frequency” profile.

The left hand graph shows an ASCENT profile using the “normal” reporting frequency as described in Fig 1 above, whereas the right hand graph shows a similar ASCENT profile but using the “high frequency” settings described in Figure 2. (Please note that the profiles are for different times).

On average the high frequency profile produced 110 reports (observations) compared with 36 for a normal profile. This resulted in a threefold cost to the profile.

The following figures show some more detail of the high frequency profiles.

Fig 3 – shows the distribution of observations during the ASCENT phase. Clearly shown is the increased resolution in the low levels (surface to 850hPa).The red line indicates the runway level based on METAR QNH pressures.

Fig 4 – shows the temperature profile. The increased resolution allows details of specific layers (inversions) to be better represented.

The results show that this additional data would be of use to the data user – be they forecaster or NWP modellers. 
The ASCENT profile – due to the onboard avionic software coding – provides highest resolution data.  An increase to the DESCENT profile produces a similar profile to a normal ASCENT.
The structure of an increased data profile may be more useful during the sunrise and sunset hours, where boundary layer conditions are forming or dissipating, with associated benefits to airline and airport operations.
The E-AMDAR Programme is ready to carry out further experimentation as required.

The Water Vapor Sensing Program: Present and Future

The First WVSS-II Equipped 737-700 Goes Operational

The initial Second Generation Water Vapor Sensing System (WVSS-II) equipped 737-700 aircraft has become operational at Southwest Airlines (SWA). This brings the total U.S. National Weather Service (NWS) WVSS-II network to 57 aircraft. During the first 10 days of operation, this SWA 737-700 aircraft reported 6,529 water vapor observations, from 100 vertical profiles across 50 flights. Water vapor data from this aircraft are within the normal performance parameters.  

At the present time the U.S. network produces over 25,000 water vapor observations in an average day. There are an additional 35 WVSS-II units scheduled to be installed on SWA 737-700 aircraft in the coming months. This will bring the total number of WVSS-II equipped aircraft in the U.S. network to 92. The U.S. WVSS-II Program Team is led by ARINC and consists of Southwest Airlines, United Parcel Service Airlines, and SpectraSensors, working together with NOAA/NWS. The team looks forward to continuing their support for the U.S. NWS, and growing the WVSS-II network to meet the expanding U.S. objectives.

The First 10 Days of WVSS-II Operations on a SWA 737-700
The inclusion of the 737-700 airframe as a WVSS-II platform, opens the door for approximately 600 aircraft among the fleets of current AMDAR participating airlines. With approximately 1,400 Boeing 737-700 aircraft worldwide, there are many opportunities for further expansion of the water vapor observation network via AMDAR.

Next Steps for the U.S. WVSS-II Network

ARINC has been awarded a third WVSS contract with NOAA to provide an additional 200 soundings/day. This will allow ARINC to capture soundings from all of the U.S., including the Pacific Northwest. The data will be collected during normal flight operations and transmitted via the ARINC GLOBALink data link service, and forwarded in near real-time to NOAA’s NWS. The improved picture of the atmosphere provided by these observations will benefit the entire weather enterprise. 

Our partners in the airline industry receive direct benefits from improved safety and operational efficiency as a consequence of accurate NWS aviation services such as Terminal Aerodrome Forecasts (TAFs), en route winds and temperatures (fuel planning), and SIGMET and AIRMET warnings. Better measurements of atmospheric water vapor will also support more accurate predictions of winter precipitation type, location, intensity, and timing, and improve severe weather predictions, such as thunderstorms and tornados. This supports not only aviation operations, but will also help local emergency managers and all our citizens better plan their activities.
Planned Expansion of the U.S. WVSS-II Network to Include Alaska and Hawaii

Global WVSS-II Activities

Three WVSS-II equipped A319-100 aircraft operating in Europe also are in the process of being upgraded to the latest Revision 3 of WVSS-II. Another 6 aircraft of the European fleet will be added to this group. This effort is being undertaken by Deutscher Wetterdienst (DWD) and Lufthansa, on behalf of the E-AMDAR community. The Australian Bureau of Meteorology has also initiated efforts towards equipping a number of aircraft in the Australian AMDAR fleet. Initial efforts will focus on the achieving the necessary WVSS-II certifications, and equipping between one and three aircraft. 

Feature Technical Article - Summary of Recent Studies on the Impact of AMDAR data in NWP Forecasts

Ralph A. Petersen

Cooperative Institute for Meteorological Satellite Studies (CIMSS)
University of Wisconsin – Madison
Madison, Wisconsin, USA

1 – Introduction  

For several decades, the World Meteorological Organization (WMO) has been sponsoring a series of scientific meetings to understand and assess the impact of various observing systems on Numerical Weather Prediction (NWP) skill.  The latest meeting was held in Sedona, Arizona in May 2012.  Participants included representative of all major global and regional Data Assimilation (DA) and NWP centers, as well as experts in various observation systems and forecast applications.  This paper presents a summary of the findings of this and previous workshops on the use and impact of AMDAR data on global scales.  Discussion of the impact on regional scales will be presented in a later newsletter.

2 – Earlier tests of the impact of AMDAR observations on Global Forecasts

Before reviewing recent findings, results of data compiled by Kelly et al. (2004), and expanded by Petersen (2004), will be presented as background.  These tests followed focused AMDAR evaluations by Andersson et al. (2005).  The tests by Kelly et al. were conducted by first running the European Centre for Medium Range Weather Forecasts (ECMWF) global analysis and forecast systems using all data sources and then repeatedly rerunning the systems with each major type of datar emoved (e.g., with all data expect AMDAR temperatures and winds were used in the ‘No AMDAR’ tests).  The skill of the various “data denial” tests where then compared with that of the forecasts using all data to determine a measure of the impact of each data source.

The impact of AMDAR temperature (left) and wind (right) data on 12 to 48 hours forecasts for the entire global and the more-data rich North American areais shown in Figs. 1 and 2 from Petersen (2004).  In this case, the impact is expressed as the percentage of error reduction when AMDAR data were used, using the formula:

Error Reduction (%) = ( Error(with AMDAR data) – Error(without AMDAR data) ) / Error(without AMDAR data)

Results for the Northern Hemisphere, where much of the data coverage was only at flight level (e.g., over ocean areas), showed that AMDAR observations:
  1. Had their largest impact during the first day of the forecasts, reducing 12-hour temperature forecasts in the upper troposphere by 15-20%,  
  2. Had their largest impact in the upper troposphere (e.g., greatest at 200 and 300 hPa near major flight levels, but also with notable impact at 500 hPa),
  3. Generally had greater influence on temperature forecasts than wind forecasts (especially during the first day of the forecasts), and 
  4. Had positive impact on forecasts out to 48 hours.

Figure 1:  Impact of AMDAR Temperature (left) and Wind (right) observations on 12, 24 36 and 48 hour forecasts at 500, 300, 200 and 100 hPa in the Northern Hemisphere.  Data provided by ECMWF.

Statistics were also compiled for the North American area, where the AMDAR data were more dense and available both at flight levels and from ascents/descents.  These results showed that AMDAR observations:
  1. Had larger impacts during the first day of the forecasts, reducing 12-hour temperature forecasts in the upper troposphere by more than 20% and improving wind forecasts by nearly 15%,
  2. Had greater impact in the upper troposphere(with negligible impact at 100 hPa), 
  3. Had greater influence on temperature forecasts than wind forecasts throughout the 2-day forecast period, and
  4. Retained greater positive impact on forecasts out to 48 hours than for the global forecasts.

Figure 2:  Same as figure 1, except for North America.

The impact of AMDAR data over other portions of the global is shown in Fig. 3, which compares wind forecast error improvements at four levels (500, 300, 200 and 100mb – corresponding to FL180, FL300, FL390 and FL 500) and four forecast lengths (12, 24, 36 and 48hrs).  The results not only reaffirm the previous findings but also shows that, even though the majority of AMDAR available at the time of this study were concentrated over the United States, Europe, and oceanic fight routes in the Northern Hemisphere, the benefits of AMDAR data extended to most regions the global, including the tropics and the Southern Hemisphere. In all but one instance, the forecasts use of aircraft data improved the wind forecasts.  On average, forecasts for the Northern Hemisphere are improved by more than 0.5 knots at FL300-FL390.  From the North Pacific, across North America and the North Atlantic and into Europe, improvements ranging from .6 to nearly 1 knot are typical for the 4 month averaged statistics.  For individual cases, the improvements can be many times larger than this – often exceeding 10 knots.  Smaller improvement is shown in the Southern Hemisphere, where fewer aircraft data are available.  In the tropics, the improvements are surprisingly large, with the greatest impact (4 month average improvements nearing 1 knot) in Arabia, Indonesia and the Central Pacific (not shown).  The degree of improvement decreased with forecast length, with reductions of from 25 to 75% shown at different levels and locations.  Similar improvements and trends were also noted in temperature forecasts for these same levels for each region.

Figure 3:  Improvement in Vector Wind Errors (m/s) at 500, 300, 200 and 100 hPa and 12, 24,36 and 48 hour ECMWF forecasts.  Verification areas shown on plot of typical 6-hr aircraft data distribution at time of study.  Figure taken from Petersen (2004).

3 – Recent tests of the impact of AMDAR observations on Global Forecasts

During the past several years, more advanced techniques have been developed to better understand and assess the impact of various observing systems relative to one another using more sophisticated DA systems.  Two studies, one performed by Gadnotiet al. (2010) at the ECMWF and another reported by Ohring (2011) from the NOAA/NASA Joint Center for Satellite Data Assimilation (JCSDA) demonstrate the importance of AMDAR data to the global observing system. 

The ECMWF results fromJune 2009 show that AMDAR data were the fourth most important data set needed to improve global 24 hours NWP forecasts, trailing only three data sets obtained from polar orbiting satellite temperature sounders - microwave (‘AMSUA’) and hyperspectral infrared (IR) observations (notes as ‘IASI’ and ‘AIRS’).  Unlike AMDAR data (marked as ‘AIREP’ in this figure), the satellite data sets provide full global data coverage, although at lower vertical and horizontal resolution than provided by AMDAR data.   The AMDAR data were shown to make a nearly 10% improvement to the forecasts for the month, more than either Global Positioning System (GPS) Radio-Occultation (GPS-RO), rawinsonde (TEMP) data or surface synoptic reports. 

Figure 4:  Impact of various different data sets on accuracy of 24-hour ECMWF forecasts (Based on Radnoti et al, 2010)

Tests from the JSCDA using the Global Modelingand Assimilation Office assimilation and forecast system presented in Fig. 5 showed similar findings.  In the JCSDA DA experiments, the AMDAR data (noted here as ‘Aircraft’) proved to be the third most important data set, in this instance behind only AMSUA and the combination of Rawinsondes and Dropwinsondes (noted as RaobDsnd).  In these evaluations, the AMDAR data contributed almost 15% of all data impacts and proved to be more important than hyperspectral IR satellite data.  Overall, more than half of the AMDAR reports had positive impacts.

Figure 5:  Fraction Error reduction by data type (right) and % of data providing useful information to the analyses (right).  From JCSDA Quarterly (2011)

The results shown here from the ECMWF and JCSDA were supported by similar results from the US Navy and the UK Met Office, among others.  Other studies showed that, although the availability of many different data sets is essential for sustaining improvements in NWP into the future, the AMDAR data were by far the most cost-effective.

4 – Summary

Tests have been conducted by numerous NWP centers over the past decade to assess the impact of various observing systems on the skill of global NWP systems.  Results show that most data sets provide important information for improving forecasts, either for individual events or long-term performance.  Although satellite microwave observations have the largest influence on 24 hours global forecastswith their global, all-weather coverage,AMDAR observations have become acritical component of operational NWP systems around the world, ranking in the top 5 in every test and contribution between 10-15% to improvements in NWP skill.  Although AMDAR data have their greatest influence in the areas and around the levels where they are most abundant, they impact on forecast forecasts extends to 48 hours and beyond.  The unique feature of providing not only temperature and wind data at individual locations and in ascent/descent but also information about the gradients of both parameters only flights paths may contribute further to the importance of AMDAR data relative to other, more costly data sets.

5 – Reference

Andersson, E., C. Cardinali, B. Truscott and T. Hovberg, 2005:  High frequency AMDAR data – a European aircraft data collection trial and impact assessment.  ECMWF Technical Memorandum # 457, 15 pp.

Kelly, G., T. McNally, J.-N. Thépaut and M. Szyndel, 2004: OSEs on all main data types in the ECMWF operation system. Proc. 3rd WMO Workshop on The Impact of Various Observing Systems on Numerical Weather Prediction, Alpbach, Austria, 9-12 March 2004, Eds. H. Böttger, P. Menzel and J. Pailleux. WMO/TD No. 1228, 63-94.

Ohring, G., ed., 2011:  Measuring the global impact of ASCAT observations.  JCSDA Quarterly. Joint Center for Satellite Data Assimilation, Camp Spring, MD, #35.

Petersen, R. A., 2004:  “Do automated meteorological data reports from commercial aircraft improve forecasts?”  ICAO Magazine, 6 pp.

Radnoti G., P. Bauer, A. McNally, C. Cardinali, S. Healy and P. de Rosnay, 2010:  ECMWF study on the impact of future developments of the space-based observing system on Numerical Weather Prediction.  ECMWF Technical Memorandum # 638, 115 pp.

Gelaro, R., 2011:  Summary article. Joint Center for Satellite Data Assimilation,Quarterly report, Available from Joint Center for Satellite Data Assimilation, Beltsville, MD.

WMO AMDAR Panel Contacts

WMO AMDAR Panel Chairman

Mr Frank Grooters
Prunuslaan 17
NL-3723 WC Bilthoven
The Netherlands
Tel : +31 30 229 3250
Mob : +31 6 1122 5867
Email : fgrooters@gmail.com
WMO AMDAR Panel Vice-chair

Mr Carl Weiss
NOAA/NWS National Weather Service
1325 East, West Highway
SILVER SPRING 20910-3283
United States of America
Tel: +1-301-7131726-149
Email: carl.weiss@noaa.gov
WMO Scientific Officer, Aircraft Observations

Mr Dean Lockett
World Meteorological Organization
7 bis, avenue de la Paix Case postale No. 2300
CH-1211 GENEVA 2
Tel: +41-22-7308323
Email: dlockett@wmo.int