Scholarlink

<p>Understanding the variability of surface water in the context of climate change is crucial for sustainable water management in semi-arid regions. This study utilizes a multi-index remote sensing approach via Google Earth Engine (GEE) to analyse the spatiotemporal dynamics of surface water bodies in Telangana, India, from 2000 to 2023. Five remote sensing-based water indices (NDWI, MNDWI, NDPI, MBWI, and AWEInsh) were evaluated using Otsu and simple-thresholding techniques to optimize thresholding and estimate changes in surface water area. Among these indices, AWEInsh demonstrated the highest visual accuracy, particularly in detecting ephemeral water bodies, while MNDWI and NDPI were identified as the most effective overall. To assess the applicability of ensemble and fusion of water indexes, a weighted average fusion index (Weighted Composite Water Index or WCWI) was developed combining all the indices used in this study, where the overall accuracy (OA) of each index was used for weightage selection. This OA-based weighting strategy ensures that indices with higher reliability contribute more significantly, while still leveraging the complementary strengths of lower-performing indices. . Correlation analysis further revealed a strong agreement among most indices and their fusion, emphasizing the reliability of multi-index fusion for regional water body mapping. Temporal trends reveal significant seasonal and inter-annual variability in surface water extent, closely associated with Indian Summer Monsoon rainfall patterns. Notably, drought years linked to El Niño events (e.g., 2004, 2014, 2015) showed sharp declines in surface water area, whereas high-rainfall years (e.g., 2005, 2013, 2016) corresponded to an increase in surface water coverage. Despite moderate rainfall, a declining trend in observed area after 2016 suggests anthropogenic influences such as land-use change and reduced catchment efficiency. This approach enhances classification accuracy and aids in water resource planning, drought preparedness, and agricultural decision-making. The findings underscore the value of remote sensing in shaping climate adaptation strategies in water-stressed, semi-arid regions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multi-indices assessment of spatiotemporal dynamics of climate-driven surface water variability for sustainable water management in semi-arid climate system

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年11月6日

查看全文
  • Jay Kumar Singh Chauhan,
  • Abhilash Kumar Paswan

摘要

Understanding the variability of surface water in the context of climate change is crucial for sustainable water management in semi-arid regions. This study utilizes a multi-index remote sensing approach via Google Earth Engine (GEE) to analyse the spatiotemporal dynamics of surface water bodies in Telangana, India, from 2000 to 2023. Five remote sensing-based water indices (NDWI, MNDWI, NDPI, MBWI, and AWEInsh) were evaluated using Otsu and simple-thresholding techniques to optimize thresholding and estimate changes in surface water area. Among these indices, AWEInsh demonstrated the highest visual accuracy, particularly in detecting ephemeral water bodies, while MNDWI and NDPI were identified as the most effective overall. To assess the applicability of ensemble and fusion of water indexes, a weighted average fusion index (Weighted Composite Water Index or WCWI) was developed combining all the indices used in this study, where the overall accuracy (OA) of each index was used for weightage selection. This OA-based weighting strategy ensures that indices with higher reliability contribute more significantly, while still leveraging the complementary strengths of lower-performing indices. . Correlation analysis further revealed a strong agreement among most indices and their fusion, emphasizing the reliability of multi-index fusion for regional water body mapping. Temporal trends reveal significant seasonal and inter-annual variability in surface water extent, closely associated with Indian Summer Monsoon rainfall patterns. Notably, drought years linked to El Niño events (e.g., 2004, 2014, 2015) showed sharp declines in surface water area, whereas high-rainfall years (e.g., 2005, 2013, 2016) corresponded to an increase in surface water coverage. Despite moderate rainfall, a declining trend in observed area after 2016 suggests anthropogenic influences such as land-use change and reduced catchment efficiency. This approach enhances classification accuracy and aids in water resource planning, drought preparedness, and agricultural decision-making. The findings underscore the value of remote sensing in shaping climate adaptation strategies in water-stressed, semi-arid regions.