A Lightning and Hail Storm prediction system using WRF at Cloud resolving scale

Thunderstorm associated with lightning, gusty wind, rainfall and hail is one of the catastrophic weather events that affect various part of India. In this present study, the probable area, time and threat of the severe storms are estimated based on model derived microphysical proxy parameters and model simulated lightning flashes. To evaluate the skill of prediction of lightning/hail storms, WRF-ARW (cloud resolving model at ~1km) are used in different regions of India. Cases like 6-7 February 2019, 12-13 December 2019 and 26 March 2018 thunderstorms associated with hail, lightning and gusty wind are chosen for detail analysis. First two widespread cases occurred mainly in north India and north-west India. Over the study area, presence of high instability, large vertical currents inside the clouds, high moisture content in the atmosphere is noticeable in both the cases. These winter hail/thunderstorms are mainly influenced by western disturbance. The characteristic behaviour like lower freezing level is also observed and conducive to formation of hailstorms. The third hailstorm case occurred in northeast India over the hilly regions of Assam, Meghalaya, Manipur, and Tripura. This is a case of pre-monsoon thunderstorm influenced by orography. Here, simulations of storms are performed using GFST1534 (12.5km) as initial and boundary conditions. For the first time, we have used two new indices Supercell Composite Parameter (SCP) and Hail Environmental Index (HEI) in addition to the classical thermodynamical indices for estimating the strength of the storm. Freezing level, wind shear, maximum wind speed is realistically captured by the model. The probable region and timing of severe hailstorm is well represented by analysing the hourly evaluation of Lightning Potential Index (LPI), SCP and HEI. Prediction of the zone of maximum hail size with gusty wind and maximum updraft velocity is nearly accurate, though there is a need of improvement. This study represents an important baseline for the forecast of hail storm at short term to nowcast time scale. Fidelity of the model in simulating lightning flashes is found realistic as compared with observed lightning flashes obtained from the Lightning Location Network. The importance of accurate representation of physical processes in models is emphasized.