Carniato Luca, Schoups Gerrit, van de Giesen Nick, Seuntjens Piet, Bastiaens Leen, Sapion Hans
Department of Water Management, Delft University of Technology, PO Box 5048, 2600 GA, Delft, Netherlands.
Department of Water Management, Delft University of Technology, PO Box 5048, 2600 GA, Delft, Netherlands.
J Contam Hydrol. 2015 Feb;173:38-58. doi: 10.1016/j.jconhyd.2014.12.001. Epub 2014 Dec 8.
In this study a numerical groundwater reactive transport model of a shallow groundwater aquifer contaminated with volatile organic compounds is developed. In addition to advective-dispersive transport, the model includes contaminant release from source areas, natural attenuation, abiotic degradation by a permeable reactive barrier at the site, and dilution by infiltrating rain. Aquifer heterogeneity is parameterized using pilot points for hydraulic conductivity, specific yield and groundwater recharge. A methodology is developed and applied to estimate the large number of parameters from the limited data at the field site (groundwater levels, groundwater concentrations of multiple chemical species, point-scale measurements of soil hydraulic conductivity, and lab-scale derived information on chemical and biochemical reactions). The proposed methodology relies on pilot point parameterization of hydraulic parameters and groundwater recharge, a regularization procedure to reconcile the large number of spatially distributed model parameters with the limited field data, a step-wise approach for integrating the different data sets into the model, and high performance computing. The methodology was proven to be effective in reproducing multiple contaminant plumes and in reducing the prior parameter uncertainty of hydraulic conductivity and groundwater recharge. Our results further indicate that contaminant transport predictions are strongly affected by the choice of the groundwater recharge model and flow parameters should be identified using both head and concentration measurements.
在本研究中，开发了一个受挥发性有机化合物污染的浅层地下含水层的数值地下水反应运移模型。除平流 - 弥散运移外，该模型还包括源区污染物释放、自然衰减、场地中渗透反应屏障的非生物降解以及降雨入渗稀释。利用先导点对含水层非均质性进行参数化，以描述水力传导率、比出水量和地下水补给。开发并应用了一种方法，从现场有限的数据（地下水位、多种化学物质的地下水浓度、土壤水力传导率的点尺度测量以及化学和生化反应的实验室尺度推导信息）中估计大量参数。所提出的方法依赖于水力参数和地下水补给的先导点参数化、一种正则化程序以协调大量空间分布的模型参数与有限的现场数据、一种将不同数据集整合到模型中的逐步方法以及高性能计算。该方法被证明在再现多个污染物羽流以及降低水力传导率和地下水补给的先验参数不确定性方面是有效的。我们的结果进一步表明，污染物运移预测受地下水补给模型选择的强烈影响，并且应使用水头和浓度测量来确定水流参数。
