- Restrictions on Access:
- Unclassified, Unlimited, Publicly available.
- Current work on the lightning jump algorithm to be used in GOESR Geostationary Lightning Mapper (GLM)'s data stream is multifaceted due to the intricate interplay between the storm tracking, GLM proxy data, and the performance of the lightning jump itself. This work outlines the progress of the last year, where analysis and performance of the lightning jump algorithm with automated storm tracking and GLM proxy data were assessed using over 700 storms from North Alabama. The cases analyzed coincide with previous semiobjective work performed using total lightning mapping array (LMA) measurements in Schultz et al. (2011). Analysis shows that key components of the algorithm (flash rate and sigma thresholds) have the greatest influence on the performance of the algorithm when validating using severe storm reports. Automated objective analysis using the GLM proxy data has shown probability of detection (POD) values around 60% with false alarm rates (FAR) around 73% using similar methodology to Schultz et al. (2011). However, when applying verification methods similar to those employed by the National Weather Service, POD values increase slightly (69%) and FAR values decrease (63%). The relationship between storm tracking and lightning jump has also been tested in a realtime framework at NSSL. This system includes fully automated tracking by radar alone, realtime LMA and radar observations and the lightning jump. Results indicate that the POD is strong at 65%. However, the FAR is significantly higher than in Schultz et al. (2011) (5080% depending on various tracking/lightning jump parameters) when using storm reports for verification. Given known issues with Storm Data, the performance of the realtime jump algorithm is also being tested with high density radar and surface observations from the NSSL Severe Hazards Analysis & Verification Experiment (SHAVE).
- NASA Technical Reports Server (NTRS) Collection.
- Document ID: 20140008582.
NOAA Satellite Science Week Virtual Meeting; 10-14 Mar. 2014; Huntsville, AL; United States.
- No Copyright.
View MARC record | catkey: 15436038