This Cooperative Project was designed to improve the ability to forecast mesoscale heavy precipitation events over the Middle Atlantic region of the eastern United States. The primary role of the university was in four areas: (1) develop, test and apply a mesoscale numerical model (the MM4 model) at 30-km resolution for real-time forecasting, (2) supply mesoscale model products and profiler winds to the forecast office in real time, (3) evaluate results of mesoscale numerical forecasts, and (4) provide training opportunities for forecast personnel in mesoscale meteorology. The contributions of the forecasting office were in two areas: (1) apply real-time mesoscale numerical products, in conjunction with conventional data and operational model guidance, in the creation of heavy-precipitation and flash-flood forecasts, and (2) evaluate the effectiveness of mesoscale real-time numerical products for improving forecasts of heavy precipitation events.
Numerical forecast products were produced on a regular basis for two years, but generally were limited to cases where significant precipitation was likely. Profiler data were made available on virtually a daily basis. During the first year of the project, PSU developed the capability to generate real-time MM4 forecasts and to deliver appropriate products to the Philadelphia forecast office. Over the next two years, more than 120 real-time forecasts were created and made available. During the second year, the delivery methodology was improved so that forecast office personnel could easily connect to the PSU Meteorology Department computer system via a high-speed modem and obtain a set of plotted figures depicting numerical model output fields. The pre-plotted figures were made available through a software package called MMF. Personnel at the Philadelphia forecast office used these forecasts successfully during heavy precipitation events to issue forecasts and warnings for its area of responsibility. In addition, the forecasters critiqued the skill of the numerical guidance and the types of presentation formats of products made available.
Forecast details were generally found to be more reliable during cold season events, as compared to summer events. As a result of the forecasters' comments, PSU was able to improve the quality of both the modeling system physics and the post-processed fields supplied to the forecasters. PSU also carried out its own evaluation of the precipitation forecasts made with MM4. This led to two Masters Theses and a paper describing the skill of the mesoscale precipitation forecasts.