1910013635134C
19NGI385
17NGI334, 19NGI385
8/1/2017
2020731 0:0:0
Completed
$237,918.00
Bayesian Merging of GLM Data with GroundBased Networks
Bitzer
Phillip
UAH
CHCE
NESDIS
Calculations for tropical cyclone intensification was sought by researchers using a product that merges Geostationary Lightning Mapper (GLM) data with groundbased observations in a Bayesian manner for a ratio of intracloud flashes to cloud to ground flashes. They trained and validated a random forest model to estimate the probability a flash is intracloud, based on optical attributes, and applied the model to tropical cyclone case studies for 2019 and 2020 to investigate its relationship to tropical cyclone intensification. An analysis of the case studies followed along with the development of a GLM beta data set with near real time capabilities that contain the probability a flash is intracloud. Their work suggests there is a signal in the cloud flash fraction that relates to tropical cyclone intensification. The cloud flash fraction (CFF) product will be developed further into the operational environment, including the exploration of suitable time steps in which to display the product to forecasters.
