Anticipatory modeling for water supply sustainability in Phoenix, Arizona

City level water policy and management decisions are typically based on past experience and in-house estimates of future conditions. Naturally, this management approach relies on so-called “craft skills”—problem solving abilities that are acquired through long, place-based experience (Lach et al., 2005). Analyses using this approach depend on a limited number of socio-economic, water supply, and water demand projections, often only a single one. Increasingly, water planners are realizing that the high uncertainty associated with population projections and water use trends, as well as with future water supply estimates, greatly limit their ability to adequately predict a city's water future. We suggest that water governance at the municipal level could benefit markedly from water planning tools that generate and analyze a large ensemble of potential future scenarios, scenarios that address population growth dynamics, water supply and demand projections, and alternative governance possibilities. For this work, we adapted our existing provider-level water supply model to create a demand-based water planning and analysis tool that can be used to explore the effects of drought, climate change, population growth, and policy options on surface water supplies, water demand, and groundwater pumping for the Phoenix Metropolitan Area. Our decision support tool (DST) framework uses numerical simulations to project future water supply and use, enabling advanced scenario analysis; quantitative or qualitative methods can be used to reveal patterns and associations among metrics for a large ensemble of scenarios. From this analysis, a smaller set of heuristics, describing the complexity and uncertainty of the scenarios, provides a basis for guiding anticipatory planning. Our DST uses population estimates in conjunction with water use to estimate water demand, and legal rights, in combination with estimates of groundwater, stream flows, and reservoir operations to estimate water supply. Policy options permit evaluation of short- and long-term alternative governance strategies. We present two examples—one regional and water provider specific—of the types of scenario analyses that are possible using this model.