Surface Runoff Modeling For Xitiaoxi Catchment Of The Taihu Region

Drainage basin——lake has formed a complete system, and they both connect with eachother through network of waterways. When the runoff pouring into the lake, the nutrients in the basin taken in the lake can cause the pollution for the lake. With the developing of economic and community in Taihu region, Taihu lake now faces increasingly severe eutrophication, pollution from lakeside drainage basin contributes greatly to eutrophication of water body. Xitiaoxi catchment, which locates on upper reaches of Taihu basin, investigation on surface runoff and associated mass transport for Xitiaoxi catchment is of significant importance as it is one of the major catchments of lake Taihu basin.Based on watershed base data, surface runoff and water budget for Xitiaoxi catchment are carried on compuer simulation in this study, supported by GIS(Geographic Information System) and remote sensing technology, using SWAT(Soil and Water Assessment Tool) model, a distributed hydrological model, developed by ARS(Agricultural Research Service) of USDA(United States Department of Agriculture). And then the main factors that influenced the water budget in this region are analysised.Firstly, the consistence of rainfall-runoff relationship for Xitiaoxi catchment of Taihu region is analyzed for a period of 34 years' data. The analyses indicate that the rainfall-runoff was in a consistent relationship for the period of 1968-2001. It is concluded that the increase of rainfall intensity and rainfall volume is the major reason for the change of runoff coefficients. This finding is helpful in understanding the runoff process for Xitiaoxi catchment and in constructing and verifying rainfall-runoff models as a further work.Secondly, We collect base maps which include the DEM data, landuse map, soil class map, watershed limits and weather data. Database structure of map mentioned above are modified to fit the request of model. So the spatial databases and attributive databases are built-up to provide input data to SWAT. The databases not only can strongly support the study of hydrological numerical simulation for Xitiaoxi catchment, but also is a strongly support for other studies in this area.47 subbasins are delineated by using the autodelineat tools of AVSWAT. Based on land use/soil class combinations and distributions, 180 hydrologic response unites(HRUs) are generated in the watershed. HRUs data and all attributive databases are used by SWAT to simulate monthly ,daily stream flow for 14 years in Xitiaoxi catchment. After ascertaining the parameters, satisfactory monthly simulation results are obtained with the accuracy of 0.848 for linearity regression coefficient and 0. 84 for Nash-Sutcliffe coefficient, the daily results are 0.8321 for R² and 0.908 for Ens.Then through validating , comparing simulated surfacerunoff to the observed ones, a series of good results for month are gained with the accuracy of 0.8945 for R² and 0.88 for Ens.the daily results are 0.8109 for R² and 0.856 for Ens. The results show that SWAT model can be successfully used to model this area.Then, every subentry of water balance is modeled in this region by SWAT from the year 1988 to 2001, the model is well calibrated and the results fit the reguest of water budget. The three most sensitive factors, including ESCO (Soil evaporation compensation factor) , CN₂(SCS runoff curve number for moisture condition II) and A WC (Available water capacity of the soil layer) are studied using single-dimensinal sensitivity analysis. The result indicates that: runoff from the basin will increase with the increase of ESCO and CN₂ and decrease with the increase of AWC, and that runoff is more sensitive to variation in CN2.In this study, surface runoff is simulated using SWAT(Soil and Water Assessment Tool) model, a distributed hydrological model. It is concluded that SWAT model can be successfully used to model this area. The finding is helpful for constructing nutrients transport model as the next step.