地下水数值模拟分析中降水入渗补给强度及渗透系数不确定性评价

Uncertainty Evaluation of Precipitation Infiltration Rate and Hydraulic Conductivity in Groundwater Numerical Simulation Analysis

由众多因素造成的地下水数值模拟的不确定性,在很大程度上影响了模拟结果的可靠性,因此分析地下水数值模拟的不确定性十分必要。以假想地下水流问题为例,选择降水入渗补给强度和渗透系数为代表参数,采用蒙特卡洛模拟方法,基于拉丁超立方抽样和替代模型,定量分析了源汇项和水文地质参数不确定性对地下水数值模拟结果的综合影响。结果表明,地下水数值模拟源汇项及水文地质参数的不确定性对模拟结果有较大影响,基于概率统计方法对模拟结果做进一步分析,可有效提高结果的可靠程度;模拟结果的不确定性特征既取决于计算点的空间位置,又与模拟时间有关;借助替代模型进行基于蒙特卡洛模拟的不确定性分析,有效节省了计算成本,具有较大的发展潜力。

The uncertainty of groundwater numerical simulation caused by some factors has a great influence on its results＇ reliability.Therefore,the uncertainty analysis of groundwater numerical simulation is of great necessary.In this paper,the precipitation infiltration rate and hydraulic conductivity were chosen as the typical variables depend on a hypothetical example.Based on the surrogate model and Latin hypercube sampling（LHS）technology,the Monte Carlo（MC）approach is used to research the impacts of the uncertainty of source and sink terms and hydrogeological parameters on the simulation results quantitatively.It indicates that the uncertainty of source and sink terms and hydrogeological parameters of groundwater numerical simulation has a great influence on its results.The confidence of the model results can be improved effectively by analyzing the results based on the probability and statistics methods.The uncertainty characters of the results are not only affected by the position of calculated points,but also affected by the periods of simulation.Analyzing the uncertainty by Monte Carlo method based on the surrogate model has a great potential to be applied,which can save the calculation of time-consuming.