Sensitivity study of WRF model in capturing extreme rainfall associated to an MCS over Central India on 15–17 September 2023

Abstract

Accurate representation of extreme rainfall events (EREs) associated with Mesoscale Convective Systems (MCSs) is a serious challenge for weather prediction models at convection-permitting resolutions. Central India, a part of the Monsoon Core Zone (MCZ), is one of the regions where MCS events are active, particularly during the Indian summer monsoon. Between 15 and 17 September 2023, multiple EREs and heavy rainfall events (HREs) occurred that caused the highest rainfall over Madhya Pradesh in the last few decades and led to floods and flood-like situations. To study this event, using various combinations of physical parameterization schemes, 18 simulations were carried out in a triple-nested domain, with the third domain having a resolution of 2 km over Madhya Pradesh. For this event, Set 8 (Thompson MP + KF CU+ YSU PBL) was found to exhibit the best performance. The model was found to capture the MCS features, although it was weaker than that in the Indian Monsoon Data Assimilation and Analysis (IMDAA) reanalysis dataset. The system is identified to primarily be low-to-mid level that extended up to 17.5 km during peak rainfall. It is also observed, in contrary to the common practice, the use of Cumulus Parameterization in the convection permitting resolution improve the result significantly. These findings underscore the importance of scheme selection in simulating MCSs over Central India and contribute to improving forecast skills for high-impact weather events in the MCZ. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2026.