Date of Award
Master of Science
Department of Engineering Physics
Cecelia A. Miner, PhD
This thesis creates a new algorithm to replace the Neumann-Pfeffer Thunderstorm Index (NPTI). The 45th Weather Squadron at Patrick AFB, Florida, uses the NPTI as an objective means of determining the probability of thunderstorm occurrence at Cape Canaveral. The probability is used for mission planning and resource protection, and increasing the accuracy of NPTI can potentially save billions of dollars for the United States space program. Stratified logistical regressions are performed and probability equations are derived for May through September using upper air data and surface observations for Cape Canaveral. A logistical regression of NPTI was also performed. Variables include combinations of the climatological frequency of thunderstorms, the conditional probability of thunderstorms, the u and v components of the 850-mb, 700-mb, and 600-mb winds, the 800-mb to 600-mb mean relative humidity, the K index, and the Thompson index. The two resulting algorithms are compared to NPTI and persistence, and are evaluated based on their ability to forecast thunderstorms correctly. The primary performance metrics used to evaluate the algorithms are hit rate, threat score, probability of detection, false alarm rate, Brier skill score, and ratio skill score. The investigation results indicate that the new algorithms are suitable for use by the 45th Weather Squadron and are an improvement on NPTI. The results show that the best algorithm, Stratified Logistic Thunderstorm Index (STLI), has a 57% better hit rate, a 51 % better threat score, and a 68% better probability of detection than NPTI. In Addition STLI shows a 59% lower false alarm rate than NPTI. Because of the significant improvement over NPTI, the algorithm should be prepared for operational use at Patrick AFB.
DTIC Accession Number
Everitt, James A., "An Improved Thunderstorm Forecast Index for Cape Canaveral, Florida" (1999). Theses and Dissertations. 5236.
The author's Vita page is omitted.