Item

Abstract

Monitoring of river discharge is a critically important part of understanding a broad range of science questions focused on hydrology, hydraulics, biogeochemistry and water resource management. Despite its importance, the publicly available in situ river discharge database has been declining steadily (from over 8000 in 1970 to less than 1000 in 2020) over the past few years. Most gauging stations in the GRDC database do not provide updated data, either because they are inactive or because the data are supplied with a large latency. For such stations, we benefit from the legacy data to extend their discharge by developing an empirical relationship between spaceborne width or height and legacy in situ discharge. We evaluated the feasibility of extending discharge estimates of gauges in the GRDC dataset benefiting from river width estimates obtained from Landsat 4-8 mission images (1984–2020) and also river water level estimates obtained from satellite altimetry missions (2000--2020). Our analysis shows that from about 10000 stations, river discharge can be extended for 2032 GRDC stations. For the discharge extension, the empirical river water level- or width-discharge is developed through a nonparametric stochastic quantile mapping function. The algorithm employs a stochastic quantile mapping scheme by iteratively (1) generating realizations of discharge and height (width) time series using a Monte Carlo simulation, (2) obtaining a collection of quantile mapping functions by matching all possible permutations of simulated discharge and height (width) quantile functions, and (3) adjusting the measurement uncertainties according to the point cloud scatter.