The Groundwater-Management Process for MODFLOW: Building on MODFLOW's Strengths

The Groundwater-Management (GWM) Process for MODFLOW consists of a set of procedures to solve an optimization problem that minimizes or maximizes a specified objective function subject to a set of constraints. The primary decision variables of each optimization problem are managed flow rates at a set of candidate pumping or recharge wells. The current implementation of GWM uses a numerical solution method that is based on construction of a response matrix and its use in a linear program. The response matrix is comprised of response coefficients, which are changes in system state (head, streamflow, or storage volume) at constraint locations that result from a perturbation in the flow rate at each candidate well. GWM requires repeated runs of the MODFLOW Groundwater Flow (GWF) Process to calculate response coefficients for each candidate well pumping independently of the others. For groundwater systems that respond linearly to external stresses, a single set of calls is made to the GWF Process to calculate responses for the candidate wells; for systems with nonlinear responses, a second level of calls is required to iteratively converge to the optimal solution. MODFLOW’s modular structure has facilitated the initial integration of the GWM and GWF Processes and subsequent enhancements to include local grid refinement and an expanded set of groundwater-management options.

In developing GWM we have tried to adhere to the MODFLOW design goals established by Arlen Harbaugh and Michael McDonald—that the program is easy to understand, use, enhance, and modify. We are challenged to balance code simplicity and robustness with a long-term goal of expanding GWM’s capabilities to address an increasingly wider range of groundwater-management problems. Our inclination is to favor simplicity and robustness, and to add complexity slowly with time to benefit an expanding user base.