Monthly Monitoring of Inundated Areas and Water Storage Dynamics in China's Large Reservoirs Using Multisource Remote Sensing

作者：Yongzhe Chen, Yiming Wang, Luoqi Li, Yanhong Cui, Xingwu Duan, Di Long · 发表于：Water Resources Research · 年份：2024 · DOI：10.1029/2023wr036450 · 被引用次数：31 · 研究领域：Flood Risk Assessment and Management、Hydrology and Watershed Management Studies、Groundwater and Watershed Analysis

Abstract High‐frequency monitoring of reservoir inundation and water storage changes is crucial for reservoir functionality assessment and hydrological model calibration. Although the integration of optical data with synthetic aperture radar (SAR) backscattering coefficients (backscatters) offers an effective approach, conventional methods struggle to consistently provide accurate retrievals over diverse regions and seasons. In this study, we introduce reservoir‐ and monthly‐specific classification models to enhance the integration of Sentinel‐1 SAR backscatters with optical‐based water dynamics. Our method covers 721 reservoirs with a capacity greater than 0.1 km 3 in China during 2017–2021. Furthermore, we leverage multisource satellite altimetry records (e.g., ICESat‐2, CryoSat‐2, and GEDI) and digital elevation models to derive hypsometry relationship (i.e., water level–water area relationship) for reservoirs, enabling the transformation of inundated areas into monthly water storage changes for 662 reservoirs, representing 93% of the total storage capacity of large reservoirs. Validation against in‐situ measurements at 80 reservoirs reveals improved monthly inundated area monitoring compared to existing data sets. Additionally, our reservoir water storage change estimates exhibit an average R 2 of 0.79 and a mean relative root mean square error (rRMSE) of 21%. Our findings highlight reservoir water increases from May/June to November and declines in winter–spring in most ...