Support for Automated Observations from U.S. Commercial Aircraft

The following statement(s) have expired and are here for historical purposes and do not represent statements of the AMS that are “in force” at this time.

(Adopted by AMS Council on 9 February 2003)
Bull. Amer. Met. Soc., 84, 515—517



The United States Meteorological Data Collection and Reporting System (MDCRS) has become an extremely useful asynoptic data source allowing for the measurement and dissemination of real-time wind, temperature, humidity, and turbulence data from commercial aircraft and enabling a significantly better understanding of atmospheric conditions. There is a wealth of documented evidence that this dataset, along with its international equivalent, Aircraft Meteorological Data Relay (AMDAR), has led to improved weather forecasts that benefit the public at large and contribute to improved safety and efficiency of flight operations.

The AMS strongly supports a well-focused national and international effort that ensures continued viability and growth of this cost-effective data source. Innovative cooperation between major airlines and the federal government is essential for this purpose. The AMS recommends that the federal government and the major airlines take formal steps now to cement this partnership for the long term, thereby ensuring the continued availability of these data. They should also explore ways to engage regional carriers for increased coverage, add capability to observe moisture on more aircraft, and optimize the strategy for collecting data from scheduled flights.


1. Introduction. Improving weather services to the aviation community continues to be a top priority for both the public and private sectors. The key to effective service is the ability to observe and forecast the detailed state of the atmosphere. The advent of numerical modeling with powerful computers has revolutionized forecasting, especially of temperature, winds, and storms. However, even with improved model physics, an accurate numerical forecast depends upon an accurate description of the initial state of the atmosphere.

For decades following World War II, radiosondes provided the bulk of sounding data assimilated by the models. However, the 12-hour frequency and spatial distribution of radiosonde observations have never been adequate to support the mesoscale analyses and predictions so important for aviation weather services. Accordingly, other data sources have been developed to provide either direct or proxy information on the vertical distribution of temperature, wind, moisture, and cloudiness: the MDCRS, the NOAA wind profiling network in the central United States, the WSR-88D Doppler radar network, and geosynchronous [the Geostationary Operational Environmental Satellite (GOES)] and polar-orbiting satellites, whose greatest utility is over the oceans, which are otherwise poorly observed.

Of these sources, MDCRS is probably the most cost-effective in supplying high-resolution wind and temperature data over the United States, including the soundings obtained during ascent and descent. As of early autumn 2002, the airlines supplied approximately 700,000 wind and temperature reports per week over the Aircraft Communication, Addressing and Reporting System (ACARS) in various formats and at various reporting frequencies. The airlines support the cost of communicating these observations to the ground. The principal drawbacks of the MDCRS data are that 1) bad weather can curtail flights (airports occasionally close and flights are cancelled) or restrict airspace (e.g., causing long detours around squall lines); 2) no moisture data are available except experimentally on a few aircraft; 3) the U.S. coverage depends upon flight schedules; it is somewhat less at night than during the day and considerably less on Friday and Saturday nights when package carriers are less active; and 4) the MDCRS data source is not guaranteed; there are no long-term agreements with the major carriers for the continued provision of data.

There is no shortage of documentation on the value of MDCRS data in numerical weather prediction and local forecasting. In a recent data denial experiment (Cardinali et al. 2002), the European Centre for Medium-Range Weather Forecasts (ECMWF) withheld all automated aircraft reports below 350 hPa from their data assimilation system, thereby effectively removing all ascent/descent data. Comparisons of forecasts including and excluding the ascent/descent data were striking. The ascent/descent data led to a reduction in geopotential height errors over North America, the North Atlantic Ocean, and Europe. The signal propagated eastward with forecast time and is clearly visible out to day five of the forecast and beyond. The atmospheric profiles from aircraft appear to have a significant impact on the four-dimensional variational data assimilation, resulting in improvements of the short- and medium-range forecast over North America, the North Atlantic Ocean, and Europe. Most impressive is that the 500-hPa anomaly correlation for the entire Northern Hemisphere showed an almost one-half day improvement at days eight and nine. These results support the expansion of the coverage of aircraft observations, especially those taken during ascent and descent, to other parts of the globe.

Other data impact tests consistently show the value of en route data in supporting significantly more accurate wind forecasts in the high troposphere. Short-range forecasts benefit similarly from MDCRS data (Moninger et al. 2003). For example, the Rapid Update Cycle (RUC) model was developed to serve U.S. aviation interests with hourly analyses and frequent short-range forecasts, out to 12 hours. More than any other data source, MDCRS sustains the RUC with high-frequency, high-density observations, especially in the high troposphere (Martin et al. 1993).

Accurate and detailed forecasts of temperature, wind, clouds, and storms have a direct bearing on aviation safety. The timing, location, and intensity of aviation weather hazards are central to the selection of air routes that will support expected traffic volumes with minimal delays or diversions. The domain of interest is primarily the National Airspace System (NAS), but also the International Airspace System (IAS).

Individual forecaster use of the ACARS data has proven to be successful. The subjective applications of ACARS data include the following: 1) tracking diurnal evolution of the lower-tropospheric wind and temperature profiles as an indicator of the potential for convection (Mamrosh 1998); 2) tracking the evolution of the temperature profile when the phase of precipitation reaching the ground is in question (rain, freezing rain, sleet, or snow); 3) tracking the shear profile as an indicator of turbulence or differential advection; and 4) tracking the freezing level because of its relevance to aircraft icing. In hundreds of examples, MDCRS data, especially ascent/descent data, have led to improved forecasts in critical situations.

Dozens of commercial aircraft carry special software on board designed to convert aircraft motions associated with turbulence into an aircraft independent measure called eddy dissipation rate (EDR). This quantity is carried in numerical prediction models and could conceivably be directly assimilated. Aircraft-independent measures of turbulence are valuable in their own right in that they characterize atmospheric turbulence, not the effect of that turbulence upon a particular air frame. It is too early to say how valuable EDR will be either in model prediction or verification, but this is another example of information that the MDCRS data format accommodates.

Opportunities for enhancing the MDCRS data source have not been fully exploited (WMO 2002). They include 1) equipping more aircraft with MDCRS capability; 2) making MDCRS systems more affordable and more flexible through onboard software to meet operational needs of aviation meteorology; 3) avionics manufacturers providing increased support for MDCRS on aircraft installations; 4) equipping low-flying regional aircraft in order to obtain more profiles at smaller airports and hence better coverage in currently data-sparse regions. This will increase the number of both en route observations in the 15,000-20,000-ft altitude range and the number of low- and midtropospheric profiles; 5) introducing humidity sensors; 6) designing a suitable system to measure and report icing conditions; and 7) optimizing the density, timing, and locations of automated aircraft reports for numerical weather prediction, field forecasting, and public sector uses.

With respect to 4), NASA is supporting the Tropospheric Airborne Meteorological Data Reporting (TAMDAR) program, designed to collect meteorological measurements from scheduled short-hop carriers that fly at lower altitudes, schedule more takeoffs and landings, and service smaller airports than the major airlines. With respect to 5), a second-generation Water Vapor Sensing System (WVSS-II), based upon diode laser technology, is under development with support from NOAA and the University Corporation for Atmospheric Research.

At the present time, six U.S. airlines provide MDCRS observations from standard avionics at an estimated annual, incremental communications cost to themselves exceeding $400,000. Capital equipment, software, and operating costs are the responsibility of agencies such as the federal government if special equipment and/or sensors are placed on the aircraft. An example is the sensor for moisture. On a year-to-year basis, NOAA and the Federal Aviation Administration jointly fund ($150,000 each) the collection and decoding of observations from participating aircraft, their conversion to Binary Universal Form for the Representation (BUFR) of meteorological data code format, and their transmission to the NWS Telecommunications Gateway. Canada and several European countries operate programs similar to the MDCRS with varying funding arrangements, including direct payment to the airlines.

The international "umbrella" for all these efforts is the AMDAR program, for which the WMO has primary responsibility. In Europe, the Network of European Meteorological Services (EUMETNET) has just completed a three-year trial program in support of AMDAR.

Continued availability of the U.S. MDCRS data is by no means certain. Even less certain is whether resources will be available to increase the numbers and types of observations. The AMS believes that evidence of the value of the data to meteorology is so strong that a viable arrangement for continued access to the data is both critical and urgent to achieve. The AMS recommends that the federal government and the major airlines take formal steps now to cement their partnership for the long term, thereby ensuring the continued availability of these data. They should also explore ways to engage regional carriers for increased coverage, add capability to observe moisture on more aircraft, and optimize the strategy for collecting data from scheduled flights. Consideration should also be given to making MDCRS data openly available to the public at large.

2. Summary. The MDCRS is a rich source of meteorological data providing critical input to numerical forecasts and to meteorological analyses. Studies have shown that MDCRS and AMDAR observations improve both the initial analyses and model forecasts of the upper air, which contribute to safer and more efficient aircraft operations. The future success of the MDCRS and AMDAR programs depends on a partnership between the airline operators and the federal government, and shared responsibility for continued viability. This important national resource must be nurtured so as to improve the safety and comfort of passengers, and provide economic benefit to operators though fuel savings and more efficient flight routing when bad weather is anticipated.


Cardinali, C., L. Isaksen, and E. Andersson, 2002: Use and impact of automated aircraft data in 4D-Var. ECMWF Tech. Memo 371, European Centre for Medium Range Weather Forecasts, 43 pp.

Mamrosh, R. D., 1998: The use of high-frequency ACARS soundings in forecasting convective storms. Preprints, 16th Conf. on Weather Analysis and Forecasting, Phoenix, AZ, Amer. Meteor. Soc., 106-108.

Martin, R. C., M. M. Wolfson, and R. G. Hallowell, 1993: MDCRS: Aircraft observations collection and uses. Fifth Conf. on Aviation Weather Systems, Vienna, VA, Amer. Meteor. Soc., 317-321.

Moninger, W. R., R. D. Mamrosh, and P. M. Pauley, 2003: Automated meteorological reports from commercial aircraft, Bull. Amer. Meteor. Soc., 84, 203-216.

WMO, 2002: Expert team on observational data requirements and redesign of the global observing system. Annex VI, Final Report, Open Programme Area Group on Integrated Observing Systems, Commission for Basic Systems, World Meteorological Organization, Reduced Session, Oxford, United Kingdom, 9 pp. + 9 annexes.