Identification of groundwater basin shape and boundary using hydraulic tomography

Abstract

Shapes and boundary types of a groundwater basin play essential roles in the analysis of groundwater management and contaminant migration. Hydraulic tomography (HT), a recently developed new approach for high-resolution characterization of aquifers, is not only an inverse method but a logical strategy for collecting non-redundant hydraulic information. In this study, HT was applied to synthetic 2-D aquifers to investigate its feasibility to map the irregular shapes and types of the aquifer boundaries. We first used the forward model of VSAFT2 to simulate hydraulic responses due to HT surveys in the aquifer with irregular geometry and predetermined constant head conditions at some boundaries, and no-flow conditions at others. The SimSLE (Simultaneous Successive Linear Estimator) inverse model in VSAFT2 was then used to interpret the simulated HT data to estimate the spatial distribution of hydraulic properties of the aquifer using a domain with a wrong geometry surrounded by boundaries of a constant head condition. The inverse modeling experiment used steady-state and transient-states data from the HT forward simulations, and it used the same monitoring network as in the aquifer with irregular geometry to assess the ability of HT for detecting types and shapes of the boundary as well as heterogeneity in the aquifer. Results of the experiment show that no-flow boundaries, which were incorrectly treated as constant head boundaries in inverse models, were portrayed as low permeable zones of the aquifer near the boundaries. Overall, the results show that HT could delineate not only the irregular shape of the aquifer in general but also heterogeneity in the aquifer. Improvements of the estimation with prior information of transmissivity and storage coefficient was also investigated. The study shows that using homogeneous initial guess parameters resulted in a slightly better estimate than others. Moreover, this study employs Monte Carlo simulations to ensure statistically meaningful conclusions.