The volume of meteorological data from commercial aircraft that are available at the operational numerical weather prediction centers has undergone a tremendous increase over the past decade, primarily a result of increases in ACARS (Aircraft Communications, Addressing, and Reporting System) automated reports. Benjamin et al. (1991) cited less than 4000 ACARS reports per day for the contiguous U.S. in 1988. Recent counts at FNMOC (Fleet Numerical Meteorology and Oceanography Center) peg ACARS reports at roughly 40,000 per day. Not only do these aircraft observations depict the state of the atmosphere at flight level, but they also can depict the vertical structure of the atmosphere during ascents and descents. Although the error rate in automated aircraft reports is much lower than that for manual aircraft reports, significant errors do exist in ACARS data. In addition, significant overlap exists between reports broadcast in the ACARS data stream and in the AIREP data stream. Results from quality control (QC) checks for a two-month period in 1996 and for a recent two-week period will be presented in order to portray the overall characteristics of all operationally available aircraft data [ACARS/MDCARS, AIREP, PIREP (1996 only) and AMDAR (1998 only)], error rates for each type of aircraft data, and characteristic errors. The QC code used in this work was developed at NRL based on the ACARS QC checks developed by Dr. William Moninger (NOAA-FSL) but expanded to examine all types of aircraft data together, eliminate duplicates among types as well as within each type, sort the data into tracks following individual aircraft, and perform quality control prior to passing the data to the data assimilation system
The 8th Conference on Aviation, Range, and Aerospace Meteorology