Enhancement of physical representation in a basin-scale hydrologic model, SWAT

Abstract

Along the development of hydrologic models, a numerous concepts and empirical equations related to hydrologic processes have been included into hydrologic numerical models. With calibration, these models generally give reasonable simulation results and can meet the practical requirement. However, those empirical equations in these models usually constrain the expansion of their application mainly due to the lack of representing related physical processes. In this paper, an improvement of updating hydrologic representation by using physical-based processes in a basin-scale hydrologic model, Soil and Water Assessment Tool (SWAT) (Neitsch et al., 2002), is presented. In SWAT, SCS (Soil Conservation Service) method is used to compute the direct (overland) runoff, and the steady-state response of groundwater to recharge is used to compute the baseflow. Additionally, the process of re-evaporation in SWAT is used to simulate the water movement from the shallow aquifer into the overlying unsaturated zone (Neitsch et al., 2002). With field experiments, hydrologists have observed that most overland flow in wet regions is generated from saturated areas over basins and a topographic-information based hydrologic model, TOPMODEL (Beven and Kirkby, 1979), was then developed. The concept of the TOPMODEL was also integrated into large-scale land surface models (e.g., Chen and Kumar (2001)). In this study, the salient features of TOPMODEL are used to enhance the physical representation of hydrologic processes in SWAT, and the overland flow, baseflow and the groundwater re-evaporation in SWAT are replaced. The updated SWAT with TOPMODEL features is applied over the East River basin in southern China, and the results are compared with the original SWAT simulations. It was revealed that the updated SWAT can provide more reasonable simulation. This indicates that the representation of hydrologic processes in numerical models should be carefully revisited and model physically-based hydrologic processes for improving their simulation capacity