Evaluation of a Width-Based Satellite Discharge Algorithm for Detecting Longitudinal Flow Changes in a Human-Regulated Continental River Basin

Abstract

This study investigated the capabilities and limitations of estimating spatially continuous river discharge using satellite-observed river width and the AMHG-based algorithm (BAM). We applied this method to 668 reaches along the Yellow River mainstem, examining whether width-derived discharge estimates can capture both natural and anthropogenic streamflow variations. Results show that discharge increases at tributary confluences and decreases in irrigated sections were successfully reproduced, indicating the potential to represent realistic spatial patterns from satellite observations. We further evaluated the impact of prior discharge, which serves as an initial guess in BAM; incorporating irrigation-corrected priors improved estimation accuracy, particularly in downstream reaches affected by human activities. However, large errors remained in levee-confined reaches, where poor co-variability between width and discharge limited performance. Our findings highlight that satellite-based width measurements, when supported by appropriate priors, offer a promising means to monitor discharge in ungauged basins, while also revealing key challenges in regulated rivers.