Development of Numerically-Based and Expert System Approaches for Airfield Nowcasting/Very Short Range Forecasting

摘要

Among the new capabilities that most USAF weather stations will acquire are NEXRAD Doppler weather radar, DMSP and GOES satellite imagery and soundings, conventional observations and model-generated analysis and forecast gridded fields via a system called AWDS (Automated Weather Distribution System), automated airfield observations and, in time, wind and thermodynamic vertical profiles on an hourly-updated basis. Each of these systems will comprise individual computer-based processing and display systems, none of which will be able, in their initial configuration, to effectively communicate or share data with other systems in the weather station. An R&D program at AFGL was recently undertaken to seek effective ways to process these data from diverse sources into an integrated database. The program called AMPS (for Advanced Meteorological Processing System), has the objective to develop, test and evaluate both numerically-based and expert system procedures for the assessment, analysis and prediction of weather events critical to safe and efficient aviation activities on and around air bases. Current emphasis is on (1) The development of procedures to nowcast NEXRAD reflectivity and doppler wind fields via statistically based extrapolation methods and to (at a later date) link these methods with a limited area numerical forecast model designed to account for convective initiation and development and (2) Expert system approaches to terminal area fog prediction. Research progress in these areas and plans regarding the establishment of an AMPS Test Facility involving prototype versions of the system capabilities cited above are discussed. Reprints. (FR)