Identification of Aquifer's Well‐Pumpage Histories

Abstract

A methodology is developed that estimates the parameters and the past pumping or recharge history of an aquifer simultaneously. A maximum likelihood procedure is used to estimate the aquifer's parameters in conjunction with a filtered stochastic description of the hyraulic‐head distribution in an aquifer to back‐calculate the well‐pumpage or recharge rates at known locations in the aquifer using a linear regression technique. The procedure is iterative so that updated well‐pumpage rates can be used to improve parameter estimates until the solution has converged. The methodology is tested using a hypothetical aquifer and two different cases. The first case involves the estimation of an aquifer's parameters and the pumping history of one well that was unknown; the second involves the estimation of an aquifer's parameters and the pumping history of two wells that were unknown. Results of these tests demonstrate the validity and usefulness of the proposed estimation technique.