AVN Based Statistical Forecasts of Total Cloud Amount and Ceiling Height

Mitchell Weiss
RS Information Systems, Inc.
McLean, Virginia

J. Paul Dallavalle
Techniques Development Laboratory
Office of Systems Development
National Weather Service/NOAA
Silver Spring, Maryland

Cloud amount and ceiling height are difficult to predict, either objectively or subjectively, yet skillful forecasts of both elements are important to aviation forecasts. Cloud amount forecasts are obviously important to the public weather forecasts. Since 1992, the Techniques Development Laboratory (TDL) of the National Weather Service (NWS) has produced objective forecasts of opaque cloud amount and ceiling height by using the Model Output Statistics (MOS) approach applied to the Nested Grid Model (NGM). The MOS equations predict the probability of certain categories of cloud amount and ceiling height. Subsequently, categorical forecasts of cloud amount and ceiling height are obtained from the categorical probabilities by using probability thresholds. These probability thresholds were obtained from the developmental sample and were designed to produce as many forecasts of each category of cloud amount or ceiling height as actually occurred in the sample. The NGM-based MOS guidance for opaque cloud was developed from observations at approximately 250 stations in the continental U.S. and Alaska; the ceiling height guidance was developed from observations at over 400 stations in the continental U.S. and Alaska.

TDL is now developing a new system to predict the total cloud amount and the ceiling height from the Aviation (AVN) run of the National Center of Environmental Prediction's (NCEP) Global Spectral Model (GSM). The AVN-based MOS equations use the same category definitions and geographic regional boundaries as the NGM-based MOS guidance. Changes from the NGM-based MOS equations however, include: forecasting total cloud rather than opaque cloud amount, increasing to approximately 800 the number of stations used for development, simultaneously developing total cloud and ceiling height equations, and extending the forecast projections from 60 to 72 hours. In addition, since cloud amount is observable only up to 12,000 feet at Automated Surface Observing System (ASOS) sites, we complement the cloud observations with Satellite Cloud Product (SCP) estimated cloud coverage. In this talk we outline the method used to generate the AVN-based cool season (September - April) MOS equations. A summary of the predictors making up the equations and the goodness of fit of the equations for the projections 6 to 72 hours after the 00 and 12 UTC AVN model runs are presented. Finally, we compare NGM-based and AVN-based MOS forecasts for a matched, independent sample of data. Preliminary comparisons have shown improvement of AVN-based MOS guidance over NGM-based MOS forecasts for cloud amount based on warm season (April - September) results. TDL is aiming to implement AVN-based total cloud and ceiling height guidance in late 1999.