HR: 08:40h
AN: H51I-03 INVITED     [Abstracts]
TI: Wetter or Drier? Estimating regional precipitation uncertainties in the IPCC Fourth Assessment global warming scenarios using a GCM super ensemble approach
AU: * Hamlet, A F
EM: hamleaf@u.washington.edu
AF: CSES Climate Impacts Group Department of Civil and Environmental Engineering University of Washington, Box 352700, Seattle, WA 98195 United States
AU: Salathe, E
EM: salathe@washington.edu
AF: CSES Climate Impacts Group Department of Atmospheric Sciences University of Washington, Box 354235, Seattle, WA 98195 United States
AU: Peacock, C
EM: cyngoat@u.washington.edu
AF: CSES Climate Impacts Group University of Washington, Box 354235, Seattle, WA 98195 United States
AU: Markoff, M
EM: mmarkoff@u.washington.edu
AF: Department of Civil and Environmental Engineering University of Washington, Box 352700, Seattle, WA 98195 United States
AU: Lettenmaier, D P
EM: dennisl@u.washington.edu
AF: CSES Climate Impacts Group Department of Civil and Environmental Engineering University of Washington, Box 352700, Seattle, WA 98195 United States
AB: Studies examining the historic record of temperature and precipitation over the western U.S. have shown that temperature trends have been spatially homogeneous, and over the last thirty years or so have been monotonically increasing along with warming trends in the northern hemisphere. These changes are consistent with global climate model (GCM) scenarios which show unambiguous, monotonic increases in temperature accompanying increasing greenhouse gas concentrations. Precipitation uncertainties associated with global warming, however, have a completely different character. There is little evidence from the historic record of a robust or spatially homogeneous precipitation trend accompanying the consistent warming shown in the observations. Instead, decadal precipitation variability at the regional scale appears to be the primary driver of observed trends over difference parts of the historic record(s). These historic trends are also consistent with GCM precipitation scenarios, which show model-specific patterns of decadal precipitation variability, with little consensus between models. When GCM sample sizes are small (as in all current IPCC GCM experiments) applying conventional downscaling approaches to precipitation can result in the misinterpretation of decadal variability in a particular set of GCM transient runs as a systematic change in precipitation. Furthermore, GCM simulations may contain little defensible information about decadal precipitation variability at a particular time in the future, which exacerbates these problems. To avoid these difficulties, we take a multi-model, 'super-ensemble' approach to estimating precipitation uncertainties. In this approach, a set of bias-corrected GCM ensembles from different models are pooled to produce a larger sample of precipitation changes in each season. From these results a systematic change in precipitation can be estimated with higher confidence, and a range of future decadal scale variability can also be estimated. Precipitation scenarios for the Columbia River basin in the Pacific Northwest, using the IPCC fourth assessment scenarios, are presented as an example of the approach. Differences between precipitation projections in the IPCC third (2000) and fourth (2005) assessment scenarios are also discussed.
DE: 1616 Climate variability (1635, 3305, 3309, 4215, 4513)
DE: 1626 Global climate models (3337, 4928)
DE: 1630 Impacts of global change (1225)
DE: 1655 Water cycles (1836)
DE: 1854 Precipitation (3354)
SC: Hydrology [H]
MN: Fall Meeting 2005