Development and application of an integrated surface runoff and groundwater flow model for a catchment of Lake Taihu watershed, China

Abstract

Severe pollution in Lake Taihu in the lower reach of the Yangtze River has led to the need to evaluate transport pathways and the overall water balance of the catchment. This paper describes the development of a grid-based spatially distributed hydrological model for integrated simulations of surface runoff and groundwater flow at catchment scales. Surface water flow is modeled based on rainfall-runoff transformation and stream flow routing, while the saturated groundwater is modeled by MODFLOW-2005. Surface water flow is coupled to the groundwater flow regime through the soil layer. Relatively simple equations are used to describe the soil storage and soil water percolation. These equations are computationally inexpensive and overcome some of the limitations in existing models for coupled simulation of surface and subsurface domains, such as either multiple soil layer data are needed in the vertical direction or numerical difficulties are encountered in dealing with the high non-linearity of the equations representing the variably-saturated subsurface. The model was verified against the V-catchment benchmark problem and was compared to existing models to demonstrate its accuracy and capability. The model was applied to Xitiaoxi catchment of Lake Taihu watershed. The accuracy of the model was satisfactory with daily and monthly efficiencies of 0.80 and 0.92, respectively for stream discharge comparison, and an absolute error of 0.144 m for groundwater level comparison. The model indicated that for Xitiaoxi catchment, a percentage of 37.4% of rainfall transferred to surface overland flow, and a percentage of 16% infiltrated as groundwater recharge, of which 81% returns to rivers as base flow and the rest is lost to the system via evapotranspiration. The model can be used as an alternative to more complex methods to assess impacts of land use or/and climate changes on water resources availability, and the interactions between surface and subsurface flow domains.