Accounting for the influence of aquifer heterogeneity on spatial propagation of pumping drawdown

Abstract

It has been previously observed that during a pumping test in heterogeneous media, drawdown data from different time periods collected at a single location produce different estimates of aquifer properties and that Theis type-curve inferences are more variable than late-time Cooper-Jacob inferences. In order to obtain estimates of aquifer properties from highly transient drawdown data using the Theis solution, it is necessary to account for this behavior. We present an approach that utilizes an exponential functional form to represent Theis parameter behavior resulting from the spatial propagation of a cone of depression. This approach allows the use of transient data consisting of early-time drawdown data to obtain late-time convergent Theis parameters consistent with Cooper-Jacob method inferences. We demonstrate the approach on a multi-year dataset consisting of multi-well transient water-level observations due to transient multi-well water-supply pumping. Based on previous research, transmissivities associated with each of the pumping wells are required to converge to a single value, while storativities are allowed to converge to distinct values.