Information regarding the calibration of the Applicant's model of the South Park ground water system was subjected to additional analyses. The differences between potentiometric head elevations, predicted by the model as a result of its calibrations, and apparently corresponding observed values of potentiometric heads, referred to as residuals, were analyzed. The spatial analysis of residuals as depicted earlier in Figures 5-32 through 5-50 was refined. Such a refinement was achieved by using the identical kriging technique for spatial interpretations as employed before but restricting the spatial extent of the interpreted distribution of residuals to distances that lie within an envelope of one mile from each selected calibration target location. This refined spatial analysis of residuals is presented in the sequence from Figures 5-53 through Figures 5-71, which follow the earlier sequence and were prepared using the identical format.

Each figure in this sequence consists of five component figures each of which correspond to a given model layer within which the model calibration targets chosen by the Applicant lie. They illustrate the distribution of residuals within that area covered by the one-mile-distance envelope. The magnitude of these residuals and their spatial distribution indicates that the model cannot be viewed as satisfactorily calibrated, either under the so-called quasi-steady-state condition or under transient conditions. This entire sequence of figures also indicates that, within the selected one-mile-distance envelope, the spatial interpretation of residuals presented earlier in Figures 5-32 through 5-50 is virtually identical to that presented in Figures 5-53 through 5-71.

The analysis of residual composites, for the transient calibration period from 1980 through 1996, is presented in Figures 5-72. This set also consists of five component figures each of which correspond to a given model layer within which the chosen model calibration targets lie. They also illustrate the distribution of residuals within that area covered by the one-mile-distance envelope, but denote composite residuals for the seventeen-year transient calibration period. The magnitude of these composite residuals and their spatial distribution again indicate that the model cannot be viewed as satisfactorily calibrated under transient conditions.

Analysis of well pumping assignments made in the model was conducted and is presented in Figure 5-73. These assignments are identified in this figure by layer and by time spans covering a total period of 135 years from 1955 through 2090. The assigned pumping by model grid cell in 1996 during the last stress period of the Transient-2 model calibration run is exactly one-half that assigned for the NOCUP model run, with the exception of cell (58,29) in Layer 4 were the assigned pumping is the same in the NOCUP model as in stress period 17 of the Transient-2 run. Throughout the NOCUP and SPCUP model runs, the prescribed values of well pumping remain constant except those prescribed for the proposed SPCUP production wells. This assignment of pumping is clearly inconsistent with the calibration well pumping, rendering both the calibration and predictive results unreliable.