Monitoring Inland Waterbody from Multiple Remote Sensing Satellites: A Case Study in Tsengwen Reservoir, Taiwan

Operational retrieval of water parameters over inland waterbodies usually requires solid infrastructure and costly maintenance. Owing to the increasing amount of satellite imageries with high spatial resolution, it becomes possible to obtain useful information for small waters in a regular manner. Tsengwen Reservoir, a major freshwater storage located in Chiayi county of Taiwan, is an important hydrological facility for irrigation and industrial water supply. However, due to the trends in weather extremes and uneven distribution of precipitation, Tsengwen Reservoir has experienced high variability of water storage between 1.2×10^8 m^3 and 5.1×10^8 m^3 in difference seasons. This study demonstrates a workflow to first reconstruct slope digital elevation model from a series of optical satellite images, including the Vegetation and Environment monitoring on a New Micro-Satellite (VENµS) mission, and then to track water level and water volume changes from a densified time series composed of multiple optical satellites. Our preliminary results show that an inclusion of VENµS images is able to increase water level accuracy from meter to decimeter level. The water volume estimates are in a range of 5%–10% when we add VENµS image sources in addition to other mid-resolution imageries, such as Landsat family and Sentinel-2 data. It is concluded that microsatellite missions are a network of supportive dataset which can help on decision making and planning in hydrological management.