Assessing the impact of climate and land use change on flood vulnerability: a machine learning approach in coastal region of Tamil Nadu, India

作者：Devanantham Abijith, Subbarayan Saravanan, Parthasarathy Kulithalai Shiyam Sundar, Nagireddy Masthan Reddy, Janardhanam Niraimathi, Ahmed Ali Bindajam, Javed Mallick, Maged Muteb Alharbi, Hazem Ghassan Abdo · 发表于：Geoscience Letters · 年份：2025 · DOI：10.1186/s40562-025-00377-7 · 被引用次数：31 · 研究领域：Flood Risk Assessment and Management、Hydrology and Watershed Management Studies、Hydrology and Drought Analysis

Abstract Flooding and other natural disasters threaten human life and property worldwide. They can cause significant damage to infrastructure and disrupt economies. Tamil Nadu coast is severely prone to flooding due to land use and climate changes. This research applies geospatial tools and machine learning to improve flood susceptibility mapping across the Tamil Nadu coast in India, using projections of Land Use and Land Cover (LULC) changes under current and future climate change scenarios. To identify flooded areas, the study utilised Google Earth Engine (GEE), Sentinel-1 data, and 12 geospatial datasets from multiple sources. A random forest algorithm was used for LULC change and flood susceptibility mapping. The LULC data are classified for the years 2000, 2010, and 2020, and from the classified data, the LULC for years 2030, 2040, and 2050 are projected for the study. Four future climate scenarios (SSP 126, 245, 370, and 585) were used for the average annual precipitation from the Coupled Model Intercomparison Project 6 (CMIP6). The results showed that the random forest model performed better in classifying LULC and identifying flood-prone areas. From the results, it has been depicted that the risk of flooding will increase across all scenarios over the period of 2000–2100, with some decadal fluctuations. A significant outcome indicates that the percentage of the area transitioning to moderate and very high flood risk consistently rises across all future projections. Th...