A Study of Severe Thunderstorm Warnings Based on WTLN Total Lighting Data and the WRF Model

Severe thunderstorms are a serious hazard resulting in loss of life and damage to property. Accurate and advanced prediction of atmospheric conditions with higher lightning potential helps in protecting communities and infrastructure as well as responding more efficiently to emergency events. Numerical weather prediction systems at finer spatial resolution linked to historical thunderstorm databases have been used in many operational forecast centers to issue lightning warnings for lead times up to a few hours. Accuracy of lightning prediction depends on several factors including model resolution, a choice of physics parameterization schemes, geographic location. WeatherBug currently provides WeatherBug Dangerous Thunderstorm Alerts (WDTA) based on the WeatherBug Total Lightning Network (WTLN) data and proprietary lightning cell tracking algorithms. Relying on an extensive amount of lightning data collected through WTLN during several years and on the operational WRF model forecasts, relevance of simulated atmospheric conditions to the actual inter-cloud (IC) and ground-to-cloud (CG) lightning events, event duration, speed, coverage, and intensity for climatologically prone to storm areas is analyzed. A multi-physics ensemble based on various cloud microphysics and boundary layers schemes is used to account for model uncertainties. This study is focused on the sensitivity of indirect thunderstorm forecasts and on variations in simulated storm characteristics, vertical profiles, intensity and freezing level heights depending on model parameters. Discussion of the results and future work toward developing more accurate lightning prediction schemes for lead times up to a few hours based on the WRF model forecasts is included in this study.