Linking a MODFLOW Groundwater Flow Model with an Urban Water Policy and Management Model

City level water policy and management decisions in the Phoenix Metropolitan Area, like most metropolitan cities, are based on water supply and demand projections. Provider-specific complexity and uncertainty in the dynamics of supply and demand for many communities requires the use of decision support systems to help water resource managers make educated decisions about water management. Water managers often use a groundwater flow model, such as MODFLOW, to evaluate the natural performance of their aquifers under various constrained scenarios within the context of the legal rights granted by either a provider designation or a certificate of groundwater assurance. Often, hydrological consultants are engaged to develop pumping and recharge scenarios, but months or years may pass before scenarios are complete and results are available for evaluation of alternative growth and management options. To date, these analyses are burdened by the typical, methodological challenges associated with standard MODFLOW simulations (i.e., convergence issues in general, large pre-process data sets, post-processing logistics, etc.). The Decision Center for a Desert City has been developing a data management interface to link the Arizona Department of Water Resources’ Salt River Valley (SRV) groundwater flow model (SRV-GFM; based on MODFLOW 2005) to our urban water policy and management model, WaterSim 5. WaterSim 5 considers exogenous and endogenous effects of climate change, population growth, per capita water use, and governance (policy levers) within a water use network for 33 major water providers. Our C sharp wrapper combines libraries, a Fortran dll (from WaterSim 5), and the FORTRAN exe DOS box process to run the SRV-GFM. Using visualization tools (e.g., dot spatial), and an SQL database, we have a dynamic run-time data management and output display environment. The current SRV-GFM model uses MODFLOW-NWT to provide robust, efficient simulations and, when using the multi-node package, a pumping-rectified analysis for often depleted cells typical in the SRV. When completed, our framework will provide automated and dynamic georeferenced controls (for a 15,700 cell grid) over data management objectives for pumping and recharge definitions and scenarios, and control over run-time events that could nominally interrupt simulation convergence. This coupled model permits a solution space for advanced scenario planning and analyses in the pursuit of surface and groundwater anticipatory water governance.