Việt Nam has more than 6700 reservoirs, including natural and artificial ones, with a total volume of up to 70 billion m3. Surprisingly, the water volume is only 440 m3/person, ranked lowest in the world. Reservoirs have always played important roles in regulating and balancing water resources. For decades, reservoirs have been functioning in flood control for many river catchments of various scales, as well as hydropower generation. Due to its multi-purpose character, it is hard to prevent conflicts and disputes in operating reservoirs, particularly in the flood seasons. In general, information on water levels provides input data to estimate the amount of water available. Diagrams of level-area-volume or depth-area-volume are often established in advance to use during the operation of any reservoir. It is important to know reservoir bathymetry and to estimate its storage because of strong impacts on both hydrological processes and management of water resource. Many methods determining bathymetry very often use data derived from field measurements, which would possibly restrict their application. Recently, Remote Sensing (RS) inversion has been applied more and more to observe reservoirs on large-scale simultaneously and periodicity. Featured by a short revisit, SAR Sentinel-1 images would properly capture changes to monitor and reliably estimate water surface extent, waterline, and water volume of reservoirs. Therefore, this paper will search to detect and extract the waterline using SAR data given in polarization. We simply hypothesize that when one or more surrounding pixels is/are not water, the considered point is a waterline point. Accordingly, one may define the waterline of a reservoir by extracting all points that satisfy the predefined criterion.