A ceiling and visibility prediction system suitable for Antarctic flight operations

Aviation forecasts of ceiling and visibility are of vital importance to the aviation community worldwide, with pilots relying on these forecasts to inform them of the landing conditions expected at their destination. These forecasts become especially critical for flights into remote locations with rapidly changing weather conditions, primitive aircraft navigation systems, and few divert landing options (e.g. Antarctic research stations). A ceiling forecast gives the height of the cloud base above the runway and a visibility forecast gives the visual range along the runway. Both are key components of Terminal Aerodrome Forecasts (TAFs) issued four times per day by operational forecasters. U.S. Antarctic Program forecasters issue TAFs for research stations (e.g. McMurdo and South Pole) and a large number remote field camps by manually analyzing available data relevant to each TAF location. In addition to being critical information for pilots, these forecasts play a key role in the near term planning of flight operations.

To improve the operational efficiency of issuing TAFs and to improve the accuracy of TAFs, an objective aviation ceiling and visibility forecast system has been developed and validated to provide guidance to forecasters. The forecast system ingests available and relevant observational meteorological data, multi-spectral operational satellite data and numerical weather prediction model fields. Forecasts of 1, 3 and 6 hours for ceiling and visibility are produced using either multiple linear regression or fuzzy logic algorithms. Algorithms are customized for specific locations, with the algorithms trained on previous cases of low ceiling and low visibility at a particular location. This is necessary because of local climate differences, unique terrain factors, and widely varying data coverage and quality in the surrounding region. A description of how this forecast system could be implemented at an Antarctic research station is described. Included are an identification of relevant data available, a description of procedures used to determine algorithm coefficients, and the data assimilation requirements for efficient operational implementation of this forecast system.