ILUES算法融合ERT数据反演污染源参数与渗透系数场

Estimation of Contaminant Source Parameters and Hydraulic Conductivity Field by Assimilating Electrical Resistance Tomography Data with Iterative Local Updating Ensemble Smoother Algorithm

摘要:

利用MT3DMS程序构建地下水污染物运移模型，采用迭代局部更新集合平滑（iterative local updating ensemble smoother， ILUES） 算法作为数据同化方法并引入（Karhunen-Loeve， KL）展开对渗透系数场进行降维，通过融合高密度电阻率法（electrical resistance tomography， ERT）测量数据和传统观测数据（水头与浓度）实现地下水污染溯源和渗透系数场的联合反演。结果表明，融合ERT数据的ILUES数据同化方法反演效率更高，模拟结果也更精确；对场地非均质渗透系数场的刻画，融合ERT数据的ILUES数据同化方法明显优于用水头与浓度数据作为观测值的情形，融合ERT等环境地球物理数据的数据同化框架能够有效地提高参数反演效果。

Abstract:

Based on the MT3DMS program， a numerical model which characterized the migration of groundwater contaminants was constructed. Iterative local updating ensemble smoother （ILUES） algorithm is used as the data assimilation method and Karhunen-Loeve （KL） expansion was introduced to implement the principal component analysis（PCA） dimensionality reduction of hydraulic conductivity field. The joint estimation of contaminant source parameters and hydraulic conductivity field was realized by assimilating measurement data of electrical resistivity tomography （ERT） and conventional observation data （concentration and hydraulic head）. The results show that the ILUES data assimilation method in which ERT data are used as the observation has better estimation efficiency and more accurate simulation results. As for the characterization of the site’s heterogeneous hydraulic conductivity field， the results also demonstrate the obvious advantage of using ERT data as the observation. In summary， the data assimilation framework in which environmental geophysical data such as ERT are used as the observation can effectively improve the effect of parameter estimation.