Evaluating Lightning Forecasts in the HRRR Model: Comparing Lightning Diagnostics to Remotely Sensed Observations

Severe thunderstorms are almost always associated with lightning and frequently occur throughout the continental U.S. The threat of lightning increases due to the randomness of a strike, making forecasting lightning difficult, which is why this study examines a possibly more effective way to forecast an area of lightning occurring. In doing so, the High-Resolution Rapid Refresh (HRRR) model, provided by the National Oceanic and Atmospheric Administration (NOAA), will be used because of its well-known convective resolving details. The goal of this study is to determine which diagnostic in the HRRR model produces a better forecast for lightning. Other studies have done this verification using only satellite-based observations, which are unable to differ between cloud-to-ground and in-cloud lightning strikes. Others have also used older versions of the HRRR model, whereas, in this research, the newest fully operational and updated version (HRRR v4) is utilized. This study implements the use of not only satellite based, but also ground based observations, to help identify the strike type and a more accurate number of strikes observed to verify the forecast variables found in the HRRR. This content is aimed to provide more insight on the proficiency of the HRRR and which diagnostics should be paid attention to the most for forecasting lightning, but also to inspire new theories and ideas to other atmospheric science researchers.