The effect of soil data resolution on hydrological processes modelling in a large humid watershed

Hydrological process modelling depends on the data characterizing the heterogeneity of the catchment. In particular, in a large-scale catchment, could a higher resolution of input data contribute to a more accurate result? In this study, surveyed soil data with two different spatial resolutions were used as input data for a Soil and Water Assessment Tool (SWAT) model simulation in a large humid watershed of Xinjiang River (15 535 km(2)) in Poyang Lake catchment in China. Simulations of stream flow, soil water storage and evapotranspiration (ET) generated using the two soil datasets were compared, and the impact of soil data resolution was analysed. The results indicate that the different resolutions of soil data have a great impact on the distribution of hydrological response units in the SWAT model, but do not lead to obvious differences in stream flow simulation. The finer resolution data produced a higher monthly soil water storage simulation than the lower-resolution data across the watershed, while the seasonal variation of soil water content (SW) shows a close relation to the local precipitation and temperature. Results also show that the soil ET calculation method in the SWAT model is not particularly sensitive to the different resolutions of the soil maps. The implications of this study are that improvement of the resolution of soil data does not necessarily contribute to a more accurate prediction of stream flow in large-scale watersheds. In practical studies, modellers need to select an appropriate resolution of soil data depending on the scale of watershed and the level of accuracy required, and also need to consider the principle of the model and the physical meaning of some key parameters to explain the simulation results. Copyright (C) 2010 John Wiley & Sons, Ltd.