Unraveling the Role of Land Use Changes in Shaping Extreme Rainfall in South India

Abstract

This study examined how heterogeneous land-use and land-cover (LULC) changes associated with urbanization influence extreme rainfall event (ERE) characteristics across metropolitan and non-metropolitan cities in South India. Our observational analysis indicated a notable increase in the ERE frequency and intensity during 2000-2022 compared to the long-term record (1950–2022), particularly during the summer monsoon season. Concurrently, rapid urbanization across South India has driven substantial LULC transformations, necessitating an investigation into their impact on rainfall patterns. To assess the LULC impact on the ERE simulations, we employed the Weather Research and Forecasting (WRF) model at 1km resolution using different datasets: contemporary Indian Space Research Organization (ISRO) LULC data (2019) versus United States Geological Survey (USGS) data (1993). For each ERE, we conducted ensemble simulations with both datasets and evaluated outputs against observations using standard statistical metrics. Results show that changes in LULC associated with urbanization significantly influence the spatial distribution and intensity of EREs, and simulations using ISRO LULC data consistently outperform those using USGS data across all study regions. The ISRO-based simulations exhibit lower rainfall bias and remarkable performance improvements relative to USGS-based simulations: Hyderabad (75%), Palakkad (71%), and Bengaluru (60%), followed by Chennai (50%) and Tirupati (25%). Additionally, ISRO data simulations enhanced the forecast skill for surface temperature, latent heat flux, sensible heat flux, and soil moisture. The improved ERE representation in ISRO-based simulations stems from more realistic LULC changes that increase surface temperature, convective available potential energy, and elevated planetary boundary layer height while reducing convective inhibition, collectively enhancing convective processes. These findings highlight the critical importance of incorporating current, high-resolution LULC data in weather prediction models for improving the extreme rainfall forecasts over rapidly urbanizing regions, where accurate rainfall prediction is essential for disaster preparedness, land-use policy making, and urban planning.