Study region: Tsengwen Reservoir, Taiwan. Study focus: Water level (WL) and water volume (WV) are important indicators for analyzing surface water resources. Satellite remote sensing enables continuous monitoring of inland water bodies in human-inaccessible areas. We integrate Landsat imagery and satellite altimetry to derive long-term (2003–2020) WL and WV variations of Tsengwen Reservoir. First, water area (WA) was extracted from Landsat imagery by Modified Normalized Difference Water Index method and a second-order regression model is proposed to recover the entire WA from cloud-covered images to enhance the data usage. Then, WAs and WLs provided from satellite altimetry are utilized to build a linear regression model which is used to transfer WA into WL. Finally, WV was computed based on the WA and WL. New hydrological insights for the region: Results showed that the usage rate of Landsat-8 imagery utilized for conversion from WA to WL can be increased from 23% to 43%. Moreover, the root-mean-square error of the difference of WLs between the estimates and a local gauge is 2.95–5.56 m, with correlation coefficients (CC) of 0.93–0.99. In addition, the derived WV variations and ground truth showed a good agreement with CC in 0.88–0.97. The results indicated that the integration of multi-source remote sensing technologies can effectively provide long-term hydrological parameters to assist administrative agencies with an appropriate plan for water resources management.
|期刊||Journal of Hydrology: Regional Studies|
|出版狀態||已出版 - 12月 2022|