Spatial analysis of thunderstorms and lightning casualties in Bangladesh

Abstract

Considering the adverse outcomes of thunderstorm‐mediated lightning in recent years, this study aimed to identify the most thunderstorm‐and‐casualty prone regions and seasons in Bangladesh, via geospatial mapping. We attempted to forecast the number of yearly thunderstorm (TS) days for each meteorological station and district‐level lightning casualties by using TS days as a proxy variable. Data on TS days and lightning casualties were collected from Bangladesh Meteorological Department and Network for Information, Response And Preparedness Activities on Disaster respectively. This study analysed 629 fatalities and 232 injuries. The Integrated Nested Laplace Approximation with Matern covariance function was employed to assess the spatial pattern of TS days. Polynomial regressions were used to forecast the number of TS days and the single clustered Generalized Estimating Equations method was employed to explore the relationship between the number of TS days and lightning casualties. The results indicated that the mean number of TS days per station increased in recent years, particularly in February, March, April, and September. We found the months of April, May, June, and September to be the most threatening months due to TS events and related casualties. The northeast region was identified as the most TS‐prone region with the highest number of casualties in Bangladesh. Our analysis suggests that it might remain the most hazardous region in coming years.