定量评估参数不确定性传递对径流模拟的影响

水文模型是模拟流域水文过程的重要工具,其参数的不确定性降低了模拟精度。为了定量探究参数不确定性对径流模拟的影响,文章基于构建的子午河流域SWAT(soil and water assessment tool)模型,采用析因分析方差分解方法评估其影响。结果表明:不同流量下参数不确定性对径流模拟影响差异明显,低流量时参数不确定性影响大,高流量时其影响小;对水文响应最敏感的参数依次为主河道传输损失流量进入深层含水层的比例系数、主河道河床有效导水率和土壤饱和导水。主河道传输损失流量进入深层含水层的比例系数与土壤饱和导水率之间的交互作用,主河道河床有效导水率与土壤饱和导水率之间的交互作用对径流模拟影响显著。结果可为降低水文模型参数不确定性的影响提供参考。

Quantitative assessment on influence of parameter uncertainties on runoff simulations

Hydrological modeling is a key tool for analyzing the hydrological process of a river basin, but uncertainties in its parameters reduce the accuracy and applicability. This paper examines the influence of parameter uncertainties on runoff simulations, using a variance decomposition method of factor analysis along with a case study of the runoff calculated with the Soil and Water Assessment Tool (SWAT) model. The results show that under different floods, the influence of parameter uncertainties on the simulations is obviously different and it is larger in the case of lower flows than higher flows. From more to less sensitive to hydrological responses, the parameters can be listed as: fraction of main channel transmission losses partitioned to the deep aquifer, effective hydraulic conductivity of the channel bed, and saturated hydraulic conductivity of soil. Interaction of the first or second parameters with the third also produces a significant influence on the simulations. This study would help reduce the impact of the parameter uncertainties in hydrological modeling.