Investigating the use of dual-polarization radar variables in forecasting lightning initiation

Lightning, often found in summer convection over the southeastern United States, has a large impact on outdoor activities. Lightning has caused on average 52 (35) fatalities per year during the 30 year period from 2003-2012 (10 year period from 1983-2012). While decreasing, the number of lightning fatalities is still unnecessarily high. Therefore, forecasting lightning initiation with radar could be an important tool for public safety by alerting the population and allowing time for precautions that could reduce lightning injury risk. Recent work shows that dual-polarization radar variables can be used to add skill and lead time to the often-used Doppler radar reflectivity thresholds for forecasting first lightning. Although results from recent research is promising, the small sample size of these studies leaves questions as to whether the skill increase are statistically significant. Therefore, the purpose of this study is to analyze additional cases, including multicellular convection, to evaluate the operational utility of dual-polarization radar variables in first flash (IC and CG) forecasting. This study will focus on using the most successful dual-polarization algorithms from recent work: the presence of graupel at -150C and the first occurrence of reflectivity of 40 dBZ or higher and ZDR of 1dB or greater at -100C. These dual-polarization approaches will be compared to various reflectivity-based algorithms (e.g., 40 dBZ at -150C). The data used in this project will be from Northern Alabama during the months of May and June of 2012. This data set will include the National Weather Service's Hytop, AL WSR-88D NEXRAD dual-polarization radar (KHTX), UAHuntsville's Advanced Radar for Meteorological and Operational Research (ARMOR) C-band dual-polarimetric radar, the North Alabama Lightning Mapping Array, and the National Lightning Detection Network.