Doing hydrology when no in-situ data exists: Surrogate River discharge Model (SRM)

Abstract

The surrogate river discharge model (SRM) uses remote sensing surrogates of river discharge (SR) to estimate streamflow in ungauged basins. Integrating SR derived from L-band microwave data with climate inputs of rainfall and potential evapotranspiration, the model operates within a hydrological framework. While SR is strongly correlated with streamflow, it is unitless and requires calibration for physical coherence. Calibration translates SR into an actual discharge value using the average or mean discharge (QM) derived from the Budyko framework. A novel likelihood approach employing SR and QM eliminates reliance on direct discharge observations. Validation across three Australian catchments demonstrates satisfactory performance, with NSE >0.6 and KGE >0.6, highlighting its applicability in data-scarce regions. The SRM software includes tools for L-band microwave data acquisition, SR generation, and hydrological model calibration, enabling global application in river discharge estimation.