WSP09428 - Laserfiche WebLink

Home Browse Search

-- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , Essentially all of the aquifer perimeter boundaries are assigned to a "No Flow" condition to represent the fact that each of the aquifer units terminates around the basin boundaries. This includes the "basement geology" for the model, in that the entire basin is underlain by over 4,000 feet of low permeability materials associated with the Pierre shale. · Aquifer Recharge Recharge is applied to the uppermost unit (uppermost active gridblock) at each x,y location in the model. For this model, it was one of the parameters adjusted during the calibration process. To begin, the same recharge distribution developed for the SB-74 model was used, and overlay zones were utilized to locally adjust the recharge value to improve model fit. Near the end of the transient calibration phase, some difficulty was encountered in obtaining an optimal solution for both the steady-state model and the transient model. In an effort to resolve this problem, an analytical solution was used for the finite- difference form of the steady flow equation to compute the recharge for each upper-most grid cell. This equation used as input the current best-fit hydraulic conductivity and vertical conductance fields together with the interpolated pristine head fields for each layer (provided in the SEO's SB-74 model). The recharge calculation procedure involved recursively solving the equation (for the mass imbalance) one grid cell at a time beginning with the deepest layer, which yielded an array of mass imbalances for each active grid cell in that layer. Next, the array of mass imbalances was routed to the overlying layer where the procedure was repeated. The algorithm continued up to the uppermost active grid cell, for which the mass imbalance was the calculated recharge. The resulting calculated recharge field, with only minor local adjustments, quickly led to our best-fit optimal solution (for both steady-state and transient calibrations). The final best-fit values used in the South Metro model are described in Section 3.2.2, Model Calibration. . Alluvial aquifer - bedrock aquifer interface boundaries As in the SB-74 model, interaction between the shallow alluvial aquifer with the underlying bedrock aquifers was treated using the River Package. The MODFLOW River Package simulates surface water - ground water interaction by specifYing a constant head representative of the river's water surface elevation and computes flows between the river and underlying ground water system as the head difference (river head minus gridblock head) times a streambed conductance. Conceptually, use of the River Package to simulate the alluvial aquifer means that heads in the alluvial aquifer are generally unaffected by transfer of water with the underlying bedrock system. For alluvial aquifers characterized by large saturated thicknesses (compared to typical head variability over the simulation period) and relatively small streambed conductance values, this is probably a reasonable way to treat these interactions. Again, simulations were started using the SB-74 values for the alluvial aquifer heads and streambed conductances, and the streambed conductance values subsequently were adjusted during the model calibration period to improve model fit to observations (Section 3.2.2). As in the SB- 74 model, a separate streambed conductance value was used for each model layer. 3.2.1.4 Internal Boundary Conditions: Aquifer Stress - Pumping with Time - In ground water modeling, there is a tendency to assume the uncertainty in the model is associated with uncertainties in the hydrological parameters of the geologic media, especially the transmissivity. However, in most models the magnitude of the stress and the nature of boundaries have more impact on the results than the uncertainty in the parameters. For example, in most instances there Page 3-9