Siqi Ke

Abstract

An accurate estimate of river discharge is vital to quantifying the global hydrological cycle and managing water resources. In view of the steadily deteriorating data provision from gauge networks, hydrological monitoring through spaceborne sensors becomes a necessity. The SWOT mission is the first satellite to conduct a global survey of the Earth's surface waters, measuring water surface elevation, river width, and surface slope for estimating discharge. Since SWOT can only sample mid-latitude locations approximately twice per its 21-day cycle, we develop a linear dynamic system for daily discharge estimation over a continuous single-branch river network. The linear dynamic system includes a process model based on a physically-based spatiotemporal correlation and observation equations utilizing SWOT discharge products. We solve this dynamic system through a Kalman filter, which is simultaneously executed in the time and space domain to obtain daily discharge. Since SWOT discharge products are currently inaccessible, we use a perturbed version of synthetic SWOT datasets obtained by Monte Carlo simulation to test the feasibility of our approach. The validation of the estimated discharge against true discharge in the synthetic datasets over all rivers leads to a median correlation as high as 0.95, a median NSE for residuals as high as 0.82, and a median relative bias as high as 5.22%, respectively. Our method delivers promising results and the hope of obtaining daily discharge once the required SWOT data is available.