An Overview of Hail Detection Techniques Using SHAVE Hail Reports

The Severe Hazards Analysis and Verification Experiment (SHAVE) was run by CIMMS and NSSL from 2006 through 2015. The decade-long verification project ran primarily during the summer months to remotely collect hail reports at high spatial density in the wake of thunderstorms across the CONUS by cold-calling the public and completing a short survey on the hail fall. These reports included confirmations of no hail fall near the thunderstorm path and smaller hail diameters not typically reported in Storm Data. Additionally, great effort was put forth to ensure hail diameters were reported to the nearest 6 mm (or one quarter inch). The project made 255,489 phone calls collecting a total of 73,877 reports of which 54,299 were hail. Hail reports within 175 km range of a dual-polarization WSR-88D numbered 19,363. Recently, analysis of single-polarization and multi-radar, multi-sensor hail detection techniques have been completed; compilation of a dual-polarization data set, and the accompanying MRMS data, is currently being completed. The single-polarization data set runs through 2012 and contains 21,564 reports across 389 events; of these events, 130 were used for manual radar inspection, yielding 1,417 individual volumes, for evaluation of different single-polarization hail detection and sizing techniques. When completed the dual-polarization data set should contain between 11,000 and 12,000 hail reports across 300+ cases. Hail detection techniques range from manual inspection of radar data to simple relations that can be derived from manual inspection to automated algorithms running on both a pixel-by-pixel and volumetric basis. Previous work in verifying these techniques and algorithms, using Storm Data, would simply use a maximum distance and minimum thresholds to pair radar data and hail reports. The high spatial density of SHAVE reports allows for a new matching technique to be used to correlate the radar data to different hail diameters, which allows for tracking trends in the radar data and hail size concurrently. This presentation will focus on summarizing SHAVE operations, results from analyses already completed, and preliminary analyses of the dual-polarization data set. Evaluations and comparisons of the skill of each technique to diagnose hail fall will be included. Future needs for observational hail data, both surface hail observations and radar observations, will also be discussed.