Joint clarification of contaminant plume and hydraulic transmissivity via a geostatistical approach using hydraulic head and contaminant concentration data

Takai, Shizuka ; Shimada, Taro ; Takeda, Seiji ; Koike, Katsuaki*

To enable proper remediation of accidental groundwater contamination, the contaminant plume evolution needs to be accurately estimated. In the estimation, uncertainties in both the contaminant source and hydrogeological structure should be considered, especially the temporal release history and hydraulic transmissivity. Although the release history can be estimated using geostatistical approaches, previous studies use the deterministic hydraulic property field. Geostatistical approaches can also effectively estimate an unknown heterogeneous transmissivity field via the joint data use, such as a combination of hydraulic head and tracer data. However, tracer tests implemented over a contaminated area necessarily disturb the in situ condition of the contamination. Conversely, measurements of the transient concentration data over an area are possible and can preserve the conditions. Accordingly, this study develops a geostatistical method for the joint clarification of contaminant plume and transmissivity distributions using both head and contaminant concentration data. The applicability and effectiveness of the proposed method are demonstrated through two numerical experiments assuming a two-dimensional heterogenous confined aquifer. The use of contaminant concentration data is key to accurate estimation of the transmissivity. The accuracy of the proposed method using both head and concentration data was verified achieving a high linear correlation coefficient of 0.97 between the true and estimated concentrations for both experiments, which was 0.67 or more than the results using only the head data. Furthermore, the uncertainty of the contaminant plume evolution was successfully evaluated by considering the uncertainties of both the initial plume and the transmissivity distributions, based on their conditional realizations.