Integration of sentinel-1a/landsat imageries and altimetry to monitor the Mekong River level

This study proposed an alternative approach that combined multiple spaceborne sensors to investigate water level anomaly along the Mekong River potentially caused by the two dams in the upper stream, Xiaowan and Jinghong Dams. We examined water level (WL) dynamics in multiple river sections before and after the impoundment of these dams. Satellite missions including radar altimetry (RA, Envisat and Jason-2), satellite laser altimetry satellite (ICESat), Landsat-5/-7/-8 optical imageries and Sentinel-1A synthetic aperture radar (SAR) imageries were used to construct composite WL time series with enhanced spatial resolutions and substantially extended WL data records. WL is considered as an important indicator of climatic impacts on water resources. Obtaining in-situ data could be expensive and time consuming for rivers/reservoirs in remote regions. As one of the most developed rivers in the world, Mekong River is in need of a monitoring system that could facilitate basin-scale management of water resources facing future climate change. In the study, a conversion function between altimetry WL and water extent from remote sensing images was first established for each dam and 6 checkpoints, and the combined long-term WL time series from Landsat/Sentinel-1A images were reconstructed for all checkpoints. 7 of 8 sites have R2 between altimetry WL and Landsat water area (WA) measurements larger than 0.85. Finally, the impact of hydrologic dynamics caused by the impoundment of the dams is assessed. The root-mean-square error (RMSE) between estimated WL and in-situ data is at 2-5 m level at the dams, and about 1 m at downstream checkpoints. We concluded that the downstream river level is affected by a combined operation of these two dams in 2009-2015, which has decreased WL by 0.20 m year^-1 in wet seasons and increased WL by 0.35 m year^-1 in dry seasons.