83rd Annual

Wednesday, 12 February 2003: 5:00 PM
Climate Change Streamflow Scenarios for Water Planning Studies in the Pacific Northwest
Alan F. Hamlet, University of Washington, Seattle, WA; and D. P. Lettenmaier and P. Mote
Potentially significant changes in regional climate over the time horizon for water resources planning may create challenges associated with risk assessment and the development of appropriate adaptation strategies. An important first step in assessing vulnerabilities to an altered climate (and ultimately creating response strategies) is understanding the implications of potential changes in hydrologic variability and seasonality. In the Pacific Northwest (PNW), for example, reductions in snowpack are likely to reduce summer and fall streamflows, and to increase flows in winter. While hydrologic models can be used to predict such changes, planning functions at many water management agencies tend to be linked to specific periods of the historic streamflow record and internally developed water management models. Incorporation of climate change into water management planning studies has therefore been somewhat hampered by the lack of methods that produce appropriate perturbations to the specific historic record(s) of interest. Given a set of adjustments to the historic streamflow record that reflect alternative future climates, there is little additional cost associated with including this information in water planning studies. The NOAA funded Center for Science in the Earth System at the University of Washington is developing methods to make such streamflow scenarios freely available to regional water management entities in the PNW. We describe the methods being developed, which are based on probability mapping from a regional hydrologic model to observed or naturalized streamflows for a set of locations within the Columbia River basin. The perturbation method creates a climate change streamflow scenario that is consistent with the observed record by mapping between the probability distributions for the simulated and observed streamflow record, and preserving other important aspects of the simulations such as the relationship between the annual mean streamflows in the hydrologic simulations. Test data and reservoir model simulations for the Columbia River basin are used to illustrate the procedure.

Supplementary URL: