On the Prediction of Cloud-to-Ground Lightning Occurrences over the Indian Region using Different Initial and Boundary Conditions

Atmospheric lightning, one of the deadliest natural disasters globally, poses significant risks, making research in this area crucial for risk reduction. This study evaluates the performance of the Weather Research and Forecasting (WRF)-Elec model for forecasting lightning occurrences over India during September 2023, utilizing initial and boundary conditions from the Global Forecast System (GFS) and the one from the modified GFS from the National Centre for Medium Range Weather Forecasting (NCMRWF), viz. NGFS. The WRF model outputs are compared with data from the National Remote Sensing Centre’s Lightning Detection Sensor Network (NRSC-LDSN). Results indicate that NGFS provides better forecasting accuracy compared to GFS, as reflected by higher Probability of Detection (POD) of 0.79 and lower False Alarm Ratio (FAR). We suggest that The NGFS data’s integration of advanced assimilation techniques and comprehensive observational data improves the model performance, emphasizing the importance of localized and enhanced inputs for accurate lightning forecasting, which is crucial for mitigating lightning-related risks.