川中紫色丘陵区径流泥沙SWAT模型的模拟应用分析

针对川中丘陵区紫色土坡耕地严重水土流失,选取典型代表李子溪流域为研究区,构建了其SWAT的模型数据库,包括地形、土壤、气象和土地利用数据库。并利用赵家祠水文站1970-1979年的实测径流和泥沙资料,对该流域的SWAT模型参数进行率定再采用1980-1986年的实测资料,对模型的适用性进行验证,同时用相对误差Re和Nash确定性系数Ens评价模拟效果。结果显示,径流和泥沙模拟相对误差均在±15%范围以内,Nash确定性系数均大于等于0.70,说明SWAT模型对李子溪流域年、月径流和年泥沙量的模拟精度较高。同时模拟值与实测值和降雨量的变化趋于一致。可见,用SWAT模型模拟和预测雨量较为丰沛、土壤侵蚀较严重的紫色丘陵地区的产流产沙是实用、可行的。

Runoff and Sediment Simulation in Purple Hilly Area Based on SWAT Model

Considering both various hydro-meteorological factors and different underlying surfaces in watersheds, a physically based and distributed hydrological model called SWAT (Soil and Water Assessment Tool) was developed to predict the runoff, sediment, and agricultural chemical yields in watersheds and large river basins over long periods of time. Lizixi Watershed is a typical representative with moderate erosion in Sichuan purple hilly area. Based on the serious purple soil erosion and water loss of slope farmlands in central Sichuan hilly areas, Lizixi Watershed is chosen as the study area to analyze the law and degree of soil and water losses in pur-ple hilly areas. Firstly, the SWAT model databases of Lizixi Watershed are constructed including the database of topography, soil, weather and land use. Then the Zhaojiaci Hydrometric Station's actual runoff and sediment data from 1970 to 1979 are used to calibrate the hydrological parameters of SWAT model, while the observed data from 1980 to 1986 are used to validate the model. The effects of simulation are evaluated by the relative error Re and Nash determinacy coefficient Ens. The results show that the relative errors of runoff and sediment simulation are within the scope of ±15% and their values of Nash determinacy coefficient are equal to or greater than 0.70. The values indicate that the simulation of annual and monthly runoff and annual sediment load is of high accuracy. And the tendency of the simulation value is consistent with the corresponding measured value and changes of rainfall. So it is feasible to make use of the SWAT model to simulate and predict the runoff and sediment yields in the purple hills with abundant rainfall and serious soil erosion. The use of the model can provide a reference for preventing soil erosion and making the control measures of water and soil conservation.