Assessment of Future Streamflow in the Colorado River Basin With Some Implications for Water Management

Long-range water supply planning in many river basins, such as the Colorado, requires assessing plausible future streamflow scenarios. Given nature's randomness and climate change, it is important to evaluate water resources system behavior under a wide range of plausible conditions. The research described in this dissertation examined various future streamflow scenarios in the Colorado River Basin to provide essential inputs needed for planning in the uncertain future. First, the research examined historical severe droughts in the basin and created three sets, or ensembles, of plausible future streamflow time series by resampling these past droughts. These ensembles are thus consistent with the idea that if it has happened in the past, it can happen again and should be planned for. These drought ensembles serve as plausible stress tests to support system simulations and test various operation and management strategies. These ensembles added to many existing ensembles produced by others for planning in the Colorado River Basin. Second, a broad range of metrics was assembled to assess the existing ensembles and classify them. These metrics help decide if a streamflow ensemble is suitable for specific planning purposes. Third, the research developed storylines of plausible future hydrologic conditions in the basin and then identified the most representative ensembles for each storyline. Since no matching ensemble was found for one of the defined storylines, a new streamflow ensemble was generated to fill this gap. The findings of this research could help inform choices on adequate ensembles for various planning purposes. Although the study focused on the Colorado River Basin, the methods are general and can be applied in other areas.