Evaluation of unconfined aquifer parameters using a successive line relaxation finite difference model.

Abstract

A finite difference model was developed specifically for analyzing the Grand Island, Nebraska aquifer test. Time-drawdown data for the aquifer test were fitted by least squares to an exponential type equation. To facilitate calibration of the model, interpolated distance-drawdown profiles also were fitted to an exponential type equation. The treatment of aquifer boundaries and the assumption of isotropic aquifer conditions affected the model computed water table profile. The effect was significant enough as to defy making accurate estimates of saturated hydraulic conductivity and specific yield. When the analysis was extended to long time periods of discharge, problems with the boundaries, particularly the distance to the lateral constant head boundary, led to unrealistic estimates of pumping level. The finite difference technique has its greatest application as a research method for analyzing short-duration aquifer tests provided that the aquifer conditions are well defined, measurements of pumping level are available and drawdown measurements have been secured for at least two observation wells within close proximity of the discharge well. Because of difficulties in maintaining convergence and model stability, the finite difference model reviewed in this study is too cumbersome to be considered a practical, field method for the analysis of unconfined aquifer parameters.