Thunderstorm forecasting typically has proved to be one of the most difficult of the tasks encountered by meteorologists. Major problems in the forecasting of thunderstorms include both the evaluation of the importance of an timing the approach of weak upper-level features, and the monitoring of low-level mesoscale boundaries.
Both the methodology and technology for the forecasting of severe thunderstorms, however, have improved dramatically over the past several years. Improvements have occurred in the use of numerical prediction, microburst modeling, data profilers, Doppler radar data, surface mesonet data developed by satellites, the use of special soundings, and the application of real-time automated identification and tracking based on volume-scan weather radar data.
Severe thunderstorms are capable of producing some of the most violent and unpredictable of weather conditions. The ability to provide accurate and timely forecasts of severe thunderstorms, thus, is an issue of significant concern. This research examines recent developments in the methodology and technology employed in thunderstorm forecasting.
The data used in this examination consists of published studies of the forecasting of thunderstorms. Where appropriate and illustrative, quantitative data included in these published studies are reported and discussed in this examination.
Where feasible within the context of the purpose of this research, thunderstorm forecasting is considered within the context of the performance of this function in the Colorado area. Five published studies are examined. The first study dealt with the forecasting implications of a tornadic thunderstorm in Northeastern Colorado in the summer of 1985 (Szoke & Brady, 1989, pp. 1834-1860). A methodology for the numerical prediction of thunderstorms was reviewed in the second study (Lilly, 1990, pp. 779-798). Microburst simulation for use in thunderstorm prediction was descri...