Analysis of Leaky Aquifer Pumping Test Data: An Automated Numerical Solution Using Sensitivity Analysis

The Kansas Geological Survey is pursuing an effort to automate some of the more common methods of aquifer pumping‐test analysis. This paper discusses the results of work done on the leaky artesian aquifer as defined by Hantush and Jacob (1955). The paper covers the basic theory of the aquifer type, the numerical solution of the leaky artesian‐well function, and the methodology of achieving the “best fit” parameters in the least squares' sense. Several data sets are used to demonstrate the applicability of the proposed technique. These examples indicate the generally satisfactory results produced by the automated analysis documented here. The algorithm has good convergence properties. Initial estimates for the aquifer parameters may vary by about three orders of magnitude above or below the correct values. For typical data sets the rms fitting error should be less than a few tenths of a foot. If this is not the case, one is probably not dealing with a simple leaky aquifer. This method of pumping‐test analysis does not eliminate the role of an experienced hydrologist to define the local hydrogeology and aquifer type. However, once the decision is made as to which aquifer configuration is being observed, this program will, in a quick and unbiased fashion, give an accurate assessment of the leaky‐aquifer parameters within the limits of the theoretical approximations and the data quality.