Evaluating different lightning parameterization schemes to simulate lightning flash counts over Maharashtra, India

Abstract

Thunderstorms source of lightning discharge is a major hazard to humans. In India, the loss of human life due to thunderstorms is high because of frequent lightning during pre-monsoon season (March-May). Therefore, simulation of lightning flash counts based on various lightning parameterization schemes and Lightning Potential Index (LPI) in Weather Research and Forecasting (WRF) model are evaluated over Maharashtra, India. The offline diagnostic methods for calculating lightning flashes are also evaluated from the model generated storm parameters. The observation from Maharashtra Lightning Detection Network (LDN) is used to validate the simulated total lightning flash for four events. The lightning flashes calculations by following Price and Rind (1992) based on cloud top height (PR92CTH) and vertically integrated ice water path (IWP) simulated the spatial pattern comparably well. The lightning parameterization based on cloud top height defined by the radar reflectivity factor threshold of 20 dBZ (DLP2) has performed better as compared to observation. The results show (i) better spatial pattern and frequency distribution of lightning flashes, (ii) accumulated rainfall, maximum reflectivity and time evolutions are in good agreement with flashes, (iii) correlation between simulated flash and hydrometeors are higher, (iv) the number of matching grid boxes due to randomness is also higher, 74.9%, 56.5%, 68.1% and 82.7% of matching grid boxes for the four cases and (v) the results based on LPI are also in consistent with the results of DLP2. The study highlights the robustness of DLP2 and indicates a promising future for the operational forecast of lightning prediction.