Long-term performance metrics for National Weather Service tornado warnings: an archeological investigation
Tornado warnings are one of the flagship products of the U.S. National Weather Service. We will update the time series of various metrics of performance in order to provide baselines over the 1986-2016 period. Lead time, probability of detection, false alarm ratio, and warning size/duration will be shown. We look at the warnings as artifacts left by a culture and attempt to ascertain the values of that culture that led to the performance. In order to do this, we have used metrics that work in a consistent way across the official changes in policy for warning issuance, as well as across points in time when unofficial changes took place. We discovered that relatively small performance differences are seen when official changes in policy for warning issuance took place, with larger performance differences when no official changes in policy for issuance took place. Our analysis is based, in large part, on signal detection theory, which separates the quality of the warning system from the threshold for issuing warnings. Threshold changes lead to trade-offs between false alarms and missed detections.
Dr. Brooks majored in physics and mathematics at William Jewell College, graduating in 1982, with a year at the University of Cambridge studying archaeology and anthropology His M.S. and Ph.D. degrees in Atmospheric Sciences are from Columbia University and University of Illinois, respectively. In his work at NSSL since 1992 he focused on why, when, and where severe thunderstorms occur and what their effects are.
Dr. Correia received his degrees in Meteorology from SUNY-Albany (undergrad), Florida State (M.S.), and Iowa State (Ph.D.) Universities. He is a Research Scientist with the NOAA/SPC and CIMMS, and Liaison to Warn on Forecast, Hazardous Weather Testbed, and Social Science. His research interests are in storm-scale modeling, data post-processing, verification, and mining.