Spatiotemporal Analysis of Total Lightning Data for Assessing their Utility in Warning of Pulse-Type Thunderstorm Severity

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Wednesday, 5 February 2014
Hall C3 (The Georgia World Congress Center )
Andrew W. Ellis, Virginia Polytechnic Institute and State University, Blacksburg, VA; and P. W. Miller and S. J. Keighton

Research over the past 25 years has established the utility of total lightning as an indicator of thunderstorm intensity, focusing on rapid increases in flash rates as evidence of storm intensification to a severe level. The potential utility of total lightning as a diagnostic and prognostic tool for assessing thunderstorm severity will increase with the establishment of the Global Lightning Mapper to be launched onboard NOAA's GOES-R satellite. Data intensive, the analyses to-date have been confined to a few specific areas of the United States and generally small samples of multiple storm types aggregated into single sample groups. The utility of total lightning as an indicator and predictor of storm severity for meteorological environments different from those of previous studies is unknown. Presented here is a spatiotemporal analysis of total lightning data for the 2012 summer thunderstorm season of the central Appalachian Mountains region of the eastern United States in support of improving severity forecasts of pulse-type thunderstorms common to the region.

A subset of total lightning data from the Earth Networks Total Lightning Network were analyzed for the daily timeframe of 2-9 pm from May 1 through August 31, 2012 within a spatial domain that approximates the County Warning Area of the Blacksburg, Virginia National Weather Service Forecast Office (NWSFO) (75,630 km2). The two-dimensional (x-y) location and the time of occurrence of each lightning strike were used to identify clusters of strikes in space and time using a single linkage clustering technique, with each cluster representing a discrete thunderstorm cell or multi-cell complex. Each lightning-defined thunderstorm was characterized in terms of its duration, areal extent, rate of motion, and two-dimensional geometric shape, with the idea that pulse-type thunderstorms are relatively short-lived, of small spatial extent, slow-moving, and non-linear. Numerical values for these variables were used as inputs to create a "Pulse Storm Index" (PSI). The PSI was generated using an analytic hierarchical process approach that was informed through a survey of forecaster perspectives regarding the relative importance of the various characteristics of pulse storms when operationally evaluating convective mode. The PSI was then used to identify the thunderstorm days during the 2012 study period that were most typified by individual storms that appeared and behaved like pulse storms. Atmospheric sounding data were then used to cross-validate these days, focusing on the strength of low-level wind shear as an additional indicator of a pulse storm environment. The time series of total lightning associated with each individual thunderstorm identified as a likely pulse-type storm was analyzed using an established “lightning-jump” algorithm. Lightning jumps, or rapid increases in flash rates, were analyzed as indicators of subsequent storm severity, as evidenced by local storm reports, to assess the viability of using the tool to help forecast severe pulse-type thunderstorms specifically.