Estimating the Routing Parameter of the Xin’anjiang Hydrological Model Based on Remote Sensing Data and Machine Learning

The parameters of hydrological models should be determined before applying those to estimate or predict processes. Xin’anjiang (XAJ) model is widely used throughout China. Since the prediction in ungauged basins (PUB) era, regionalization XAJ has been a subject intense focus; nevertheless, while many efforts have targeted related runoff yield using in-site data sets, classic regression predominantly applied. In this paper, we employed remotely sensed underlying surface and machine learning approach establish for estimating routing parameter, namely, CS, model. study was conducted on 114 catchments from Catchment Attributes MEteorology Large-sample Studies (CAMELS) set, relationships between CS various characteristics were explored by gradient-boosted tree (GBRT). results showed that drainage density, stream source density area catchment three major factors with most significant impact CS. best correlation coefficient (r), root mean square error (RMSE) absolute (MAE) GBRT-estimated calibrated 0.96, 0.06 0.04, respectively, verifying good performance GBRT Although bias noted simulations could still achieve comparable Further validations based two China confirmed overall robustness accuracy simulating processes Our confirm following hypotheses: (1) help large sample associated remote sensing data, ML-based can capture nonstationary nonlinear (2) estimated ML samples guarantee mode. This advances methodology quantitatively extended other models.