Modifications to the modular three-dimensional finite-difference ground-water flow model used for the Columbia Plateau Regional Aquifer-System Analysis, Washington, Oregon, and Idaho

Abstract

This report documents modifications to the U.S. Geological Survey's modular three-dimensional finite-difference groundwater flow model used for a regional aquifer-system analysis of the Columbia Plateau. The changes were needed because the aquifer system includes unconformities, deep canyons, and steeply dipping structural barriers that could not be represented realistically by the existing model. The modifications permit flow from a cell in any layer to a cell in any other adjacent layer; allow a cell that has been cut completely through by a canyon to remain in the model by setting the branch conductance to zero only on specified cell walls; simulate barriers to lateral groundwater flow by reducing the branch conductance only on specified cell walls; and allow use of a convergent grid. The concepts and mathematical basis for these modifications are described. The changes in the data-input items for this modified model, compared with those of the existing modular model, are described.

Simulation analysis of the ground-water flow system in the Portland Basin, Oregon and Washington

“…These problems stem from the inability of most models (including MODFLOW) to allow vertical flow across inactive model cells that must separate non-adjacent layers; thus, if a hydrogeologic unit has been truncated or has pinched-out, vertical flow cannot be simulated between the overlying and underlying units. This problem has historically been addressed by either leaving "pseudo-cells" in missing layers with very large vertical conductances and very low horizontal hydraulic conductivity (Morgan and Dettinger, 1994), or by modifying the model to allow vertical flow between non-adjacent layers (Hansen, 1993).…”

Hydrogeologic framework of the Puget Sound aquifer system, Washington and British Columbia

“…Ground-water flow was simulated using the finitedifference model MODFLOW (McDonald and Harbaugh, 1988) that was modified by Hansen (1993). The models were operated under steady-state conditions with the top active cell as a water table and all other cells fully convertible to a water table.…”

Simulation analysis of the ground-water flow system in the Portland Basin, Oregon and Washington

“…These problems stem from the inability of most models (including MODFLOW) to allow vertical flow across inactive model cells that must separate nonadjacent layers; thus, if a hydrogeologic unit has been truncated or has pinched out, vertical flow cannot be simulated between the overlying and underlying units. This problem has historically been addressed by either leaving "pseudo-cells" in missing layers with very large vertical conductances and very low horizontal hydraulic conductivity (Morgan and Dettinger, 1994) or by modifying the model to allow vertical flow between nonadjacent layers (Hansen, 1993).…”