Date of Award


Document Type


Degree Name

Master of Science


Department of Engineering Physics

First Advisor

Peter Saunders, PhD


A method of generating synthetic lightning through the use of a convective available potential energy (CAPE) times precipitation rate (P) proxy is applied over three distinct climatological zones of the world for a single warm season: central and southern AZ of the United States, central Cuba, and North Korea. Global Forecast System (GFS) 0.25° by 0.25° forecast data for June, July, and August of 2019 is used to provide 6-hourly CAPE and precipitation rate, while Global Lightning Dataset (GLD360) data for the period 2016 to 2020 is used to provide observed lightning strokes. A five-year lightning climatology study is conducted on each region to identify lightning density and energy patterns. A correlation of monthly and seasonal lightning to the proxy is then conducted for each region at a 0.25° grid resolution to obtain correlation coefficients (r), p-values, and establish validity of the proxy. Linear regression is then applied to the two datasets to determine an appropriate constant of proportionality, enabling the proxy to be used to explicitly forecast lightning for each region. The proxy is then used to forecast lightning at monthly and seasonal timescales and compared to observed lightning stroke densities at the same location and time to assess performance. The lightning climatology study shows that long-term variations in lightning patterns for each region are most influenced by warm-season convection, topography, and local effects. Correlation results showed best r values at 0.799 (AZ), 0.711 (Cuba), and 0.562 (North Korea), and poorest r values at 0.462 (AZ), 0.260 (Cuba), and 0.005 (North Korea), indicating that while moderate correlations exist at times over the monthly and seasonal timescales, a correlation over a longer time period encompassing several warm seasons would likely result in more accurate r values and constants of proportionality. Forecast results show that the CP proxy provides a fairly accurate depiction of lightning timing and location at monthly and seasonal timescales but performs poorly at capturing the correct strokes per area.

AFIT Designator


DTIC Accession Number