地质统计学法在地下水污染溯源中的应用及参数敏感性分析

为了确定地下水污染物来源,将地质统计学方法应用于追溯一维均质含水层中污染源的排放过程,利用数值实验分析了观测点布设位置、污染物浓度观测误差、水动力弥散参数和地下水流速对地质统计学法溯源效果的影响。结果表明,观测点提供的浓度信息能准确反映观测时刻污染物浓度的分布时,根据这些浓度信息可得到好的溯源效果。反演出的污染物排放过程与真实排放过程相似度较高时,污染物浓度测量误差的数量级介于10-6~10-3.5之间,水动力弥散系数的变化幅度在-10%~5%内,实际平均流速的变化幅度在±2%间。在本次数值实验条件下,地质统计学法给出了较好的溯源效果,与观测浓度误差相比,水动力弥散参数和地下水流速对溯源效果影响更显著。

Application and sensitivity analysis of geostatistical approach to groundwater pollution source identification

To identifying the groundwater pollution source, a geostatistical approach is applied on the groundwater pollution identification (PSI) in a 1D homogeneous aquifer. Numerical test is employed to ana-lyze the impact of observation point configuration on the PSI and the sensitivity analysis of pollution concen-tration observation error and model parameters. It is found out that if concentration observation data could precisely describe the real concentration distribution at the observed time point, a nice identification of the pollution discharge process could be obtained. If the calculated pollution discharge process is similar with the real discharge process,the dimension of the observation error is between 10^-6 and 10^-3.5,the dispersion coefficient varies between -10% and 5%,and the actual mean velocity varies between ±2%. In the nu-merical case, the geostatistical approach shows a good performance for identifying the contaminant release process. Comparing to the observation error of concentration,the influence of dispersion coefficient and actu-al mean velocity are more outstanding in this case.