Evaluation of the 29-km Eta Model. Part II: Subjective Verification over Florida

Abstract

This paper describes a subjective evaluation of the National Centers for Environmental Prediction 29-km (Meso) Eta Model during the 1996 warm (May?August) and cool (October?January) seasons. The companion paper by Nutter and Manobianco presents results from an objective evaluation of the Meso Eta Model at three selected locations during the 1996 and 1997 warm and cool seasons. The overall evaluation is designed to assess the utility of the model for operational weather forecasting by the U.S. Air Force 45th Weather Squadron, National Weather Service (NWS) Spaceflight Meteorology Group, and NWS Office in Melbourne, Florida. In the subjective verification, limited case studies are used to highlight model capabilities and limitations in forecasting convective activity, the location and movement of cold fronts, and the onset of sea breezes over regions including east-central Florida. In addition, contingency tables and categorical scores are used to verify the occurrence of these phenomena throughout the season. Results from the subjective verification demonstrate that model forecasts of developing weather events such as thunderstorms, sea breezes, and cold fronts are not always as accurate as might otherwise be implied by the seasonally averaged error statistics. Although the objective statistics do not indicate whether the model provides more accurate forecast guidance on average during either the warm or cool seasons, results from the subjective verification suggest that model forecasts over central Florida may be more useful during the cool season. This is because the Meso Eta Model resolution is not yet sufficient to resolve the small-scale details of sea and river/lake breeze circulations, thunderstorm outflow boundaries, and other phenomena, which play a dominant role in determining the short-term evolution of weather over east-central Florida during the warm season. Lessons learned from the subjective portion of the Meso Eta evaluation should apply equally as well to the recently upgraded ?early? Eta Model running with a similar 32-km horizontal resolution.