APPLICATION OF AN INVERSE ALGORITHM USING A TWO-DIMENSIONAL LABORATORY SANDBOX MODEL

Abstract

An iterative geostatistical inverse technique was developed by Liu and Yeh (1999) to use sets of head/discharge data to determine a detailed saturated hydraulic conductivity distribution within an aquifer. A two-dimensional laboratory flow cell that contains a heterogeneous soil configuration is used to evaluate the effectiveness of this numerical model. Two separate heterogeneous soil conditions are represented by laboratory "sandbox" models. Based on the known hydraulic conductivity of the soils used in the models, a synthetic data set is produced by solving the flow equation for a given pumping event. With this error-free data set, the inverse algorithm is implemented, and the effects of the inputs to the model, such as the quantity of data, the correlation structure, and the mean conductivity, are observed. Using this knowledge, the laboratory data is then supplied to the inverse algorithm, and the true conductivity field of the sandbox is successfully represented.