Effect of different soil data on hydrological process modeling in Weihe River basin of Northwest China

Hydrological process modeling depends on the soil data spatial resolution of the watershed. Especially, in a large-scale watershed, could a higher resolution of soil data contribute to a more accurate result? In this study, two soil datasets with different classification systems FAO (World Reference Base) and GSCC (the Genetic Soil Classification of China) were used as inputs for the SWAT model to study the effects of soil datasets on hydrological process modeling in Weihe River basin, China. Results show that the discharge simulated using FAO soil data was better than one simulated using GSCC soil data before model calibration, which indicates that FAO soil data needed less effort to calibrate. After model calibration, discharges were simulated better by both of FAO and GSCC soil data but statistical parameters demonstrate that we can make a relatively more accurate estimation of discharge using the GSCC rather than FAO soil data. Soil water content (SW) simulated using GSCC soil data was statistically significantly higher than those simulated using FAO soil data. However, variations in other hydrological components (surface runoff (SURQ), actual evapotranspiration (ET), and water yield (WYLD) were not statistically significant. This might be because SW is more sensitive to soil properties. For studies aiming to simulate or compare SW, merely calibrating and validating models using river discharge observations is not enough. The hydrological modelers need to identify the key hydrological components intrinsic to their study and weigh the advantages and disadvantages before selecting suitable soil data.