基于原位土壤观测的SWAT关键参数及模拟优化分析

Optimization of Key Parameters for SWAT Model Based on Field Soil Observation

利用现场观测土壤属性数据,对基于传统数据库的率定值和监测数据的参数计算值进行模拟结果对比分析。以三江平原阿布胶河流域的现场观测数据,通过物理机制中公式或模型,计算得到SWAT模型关键土壤属性参数的值,再进行两次SWAT模型的模拟。先运用基于土壤数据库参数率定值模拟,再用观测数据计算的参数值替换掉部分率定的参数值;与已经率定好的模型进行对比,评价参数计算的结果。结果表明,SOL_AWC、SOL_K、SOL_BD和USLE_K等4个参数在基于数据库的率定值和观测数据计算值之间有显著性差异,而SOL_CRK的率定值和基于观测数据的计算值之间无显著性差异。两种模拟结果相比,发现基于观测数据的农业面源负荷在数值上不存在显著性差异,且在年尺度和月尺度上具有相同的变化规律。结果表明基于观测数据的SWAT模型能够很好地模拟农业面源负荷在时间尺度上的变化规律。

In this study, onsite observed soil data were used to calculate the values of five key soil parameters by physically based mathematical formula and model in the Abujiao River basin in Northeastern China. Then two simulations of agricultural diffuse pollution were performed with SWAT（Soil and Water Assessment Tool）model. One model simulation was based on calibrated parameter values, while the other simulation was based on calculated parameter values. There were considerable differences between calibrated parameter values and calculated ones for SOL_AWC, SOL_K, SOL_BD and USLE_K, but no significant difference found for SOL_CRK. However, two simulations showed no significant difference in the agricultural diffuse loads. They also had similar dynamics both on yearly and monthly scales. Therefore, SWAT model based on observed soil data could be a good simulation on agricultural diffuse pollution, especially in watersheds without monitoring data on hydrology and water quality.