Linking satellite based lightning initiation nowcasts with WRF lightning threat forecasts to produce a combined nowcast of lightning onset and extent

Nowcasting 0-1 hour lightning initiation (LI) from geostationary satellite data is a recently developed concept. Harris et al. (2010) used the GOES-R Convective Initiation (CI) Tool (Walker et al. 2012) to quantify the behavior of ten Geostationary Operational Environment Satellite (GOES-12) infrared interest fields beginning 1-hour in advance of LI. Results indicated that an average lead time of 35 minutes prior to LI can be gained when geostationary satellites are used to identify locations where in-cloud charging is likely to be occurring, with a range in lead time from 15 minutes to 1 hour. Recent research by Iskenderian et al. (2013) demonstrated a probability of detection of ~76% (false alarm ration of ~30%) for nowcasting cloud-to-ground lightning when additional numerical weather prediction information is considered together with the GOES satellite infrared data of growing convective clouds. Current work is being done to integrate these LI nowcasting capabilities into the mature CI algorithm/tool, and to demonstrate its usefulness to forecasters. With the basis for satellite-based LI nowcasting in place, it is desired to incorporate nowcasts of lightning flash density (or lightning amounts) in order to achieve a complete CI, LI, and flash density nowcast algorithm/tool.

The Weather Research and Forecasting (WRF) model currently outputs lightning threat forecasts developed by McCaul et al. (2009). These 2-km to 4-km resolution forecasts are based on proxy fields of graupel flux at –15° C and vertically integrated ice. These forecasts offer a time and spatial coverage of lightning threat density. Research is being done to develop the methods for combining the WRF model lightning threat output with the GOES-R CI Tool. Data are being collected from the North Alabama Lightning Mapping Array (NALMA) to provide a ground truth of lightning density for numerous first-flash LI events during summers 2011-2013. WRF model output data from the corresponding days are being analyzed to find a correlation between the lightning threat product and actual lightning observed, both in extent of lightning, as well as location. Statistics from the WRF output are being calculated and various weighting methods are being applied to achieve this goal. Data from satellite interest fields from the GOES-R CI Tool leading up to LI, as seen from the NALMA, are also being collected for comparison with the flash density results. By integrating these unique data sets, the goal is to develop an algorithm that will provide a nowcast of LI along with the potential flash density associated with the future convection.

Our presentation will highlight recent results, provide the more complete methodology, and show example cases on how the LI–flash density algorithm performs, with a focus on convection in the Southeastern U.S.