Classic approaches to water resources planning have used cost-benefit analysis, where a single estimate of a project's costs and benefits are compared to determine if a project is funded. Recent studies have shown that aggregating multiple, conflicting objectives into a single objective cost function is problematic, because they penalize certain objectives in ways that are difficult to predict. Moreover, concerns about environmental change and the nonstationarity of hydrologic data motivate approaches that can move beyond the historical record for evaluating project alternatives. In this presentation we introduce the concept of many-objective analysis—combining heuristic evolutionary algorithm optimization, simulation models, and interactive visualizations to help aid robust, sustainable decision making. Two examples are shown, from water marketing in the Lower Rio Grande Valley in Texas and infrastructure planning for the Thames basin in the UK. We also suggest potential linkages between RiverWare modelling and evolutionary algorithms, model diagnostics, and visualizations.

Click HERE for a PDF version of the presentation slides.