Estimating Spatial Distribution of Hydrogeological Parameters from an Aquifer Pumping Test: The Value of Qualitative Prior Knowledge

An aquifer pumping test in a hypothetical groundwater system is evaluated by calibrating a three-dimensional groundwater model using drawdown data from a 72-hour pumping test followed by a recovery period. Two different approaches are undertaken in estimating hydrogeological parameters (hydraulic conductivity and specific storage): 1) a zonation approach where there is prior knowledge of the spatial distribution of hydraulic conductivity; and 2) there is no knowledge of how hydraulic conductivity varies spatially. Both approaches use an inverse parameter estimation modeling code. In the first case, a limited number of parameters are estimated using zones delineated on the basis of spatial trends in lithology. The second case involves using pilot points, where pilot points are distributed evenly and at regular distances across the three layers of the groundwater system with each pilot point representing a hydraulic conductivity or specific storage value. A total of 14 parameters were estimated using the zonal approach by fitting the pumping test drawdown and recovery data against the simulated values and parameters compared against known values. Using the pilot point approach, a total of 506 pilot points was employed. In both cases, the simulated drawdowns closely followed the observed drawdowns in all 10 wells. The pilot point approach provided a heterogeneous distribution of parameters with more confidence of the estimates within the radius of influence of the aquifer pumping test. Overall, parameter values estimated using zonal approach were more accurate than those estimated using the pilot point approach, which indicates the value of even qualitative prior knowledge.