Comparisons and verification of an automated thunderstorm potential index output to manual products

Forecast centers supporting the U.S. civilian and military communities produce time-phased and point-in-time thunderstorm forecasts valid for specific synoptic and mesoscale regions. These forecasts are produced manually a routine number of times each day, with updates/amendments issued when necessary. Forecasters creating the products use a wide variety of data, visualization tools, and output from forecast algorithms as part of their production process. An automated Thunderstorm Potential Index (TPI) algorithm which relies on synoptic and/or mesoscale model data has been developed to provide a “first guess” product to assist forecasters producing these thunderstorm products. The TPI output is being compared to the time-phased 1- and 2-Day Convective Outlooks produced by the NWS Storm Prediction Center during the Mar-Sep 2005 period. Output is also being compared to thunderstorm forecasts produced by the Air Force's Operational Weather Squadrons supporting CONUS DoD installations. Finally, all products will be verified against lightning data and severe weather reports. The goal of the work is to determine whether or not the TPI provides a useful automated first-guess estimate of the final forecaster-derived products created at the forecast centers, and thus be a potential time- and task-saving tool for civilian and military thunderstorm forecasters.