Verification and Analysis of Impact-Based Warnings: Assessing the Performance of Quantitative Hazard Forecasts in the Transition toward the FACETs Paradigm

Following the 2011 Joplin, MO tornado, the National Weather Service (NWS) began to incorporate impact-based threat information into severe weather warnings to provide guidance on the type and magnitude of severe weather hazards. Since their experimental deployment in 2012, Impact-Based Warnings (IBWs) and the accompanying hazard “tags” for maximum hail size, wind speed, and tornado damage are now issued by all forecast offices east of the Continental Divide. However, aside from a single study examining tornado IBWs for 2013, no substantial work has been published on impact tag accuracy by either the NWS or the literature.

The present research analyzes severe and tornado IBWs and Severe Weather Statements issued by Central and select Southern Region forecast offices over the period 2013-2015. Ground-truth data are used to quantitatively and categorically evaluate hazard tag accuracy and explore trends in biases and performance. The analysis also explores possible effects of other variables—including population density, whether the IBW was issued on a severe weather outbreak day, and the number of simultaneously valid warnings—on tag accuracy. Furthermore, particular cases are examined in detail with WSR-88D radar data to assess the tools currently available to the forecaster in issuing IBW tags. Results are expected to help validate the potential viability of IBWs, highlight areas for improvement in tag performance, and identify key implications for the warning process.