| Abstract/Description or Keywords |
Given the importance of the Canadian portion of the Columbia River basin to climate change impacts throughout the basin (Hamlet 2003), and to local impacts in British Columbia, there is great interest in the hydrologic sensitivity of this region to climate change. Currently, two different large-scale hydrologic modeling studies (conducted by the Climate Impacts Group (CIG) at the University of Washington and the Pacific Climate Impacts Consortium (PCIC) at the University of Victoria respectively) have projected the impacts of regional climate change for the Canadian Columbia River basin. Both studies used the Variable Infiltration Capacity (VIC) hydrologic modeling package, implemented at 1/16th degree resolution, and an ensemble approach which provides a range of projected impacts. This range of responses from the two studies relates, in part, to fundamental scientific uncertainty in temperature and precipitation projections for specific decades in the future. Figure A, for example, shows a summary of hydrologic projections for natural flow in the Columbia River at Revelstoke Dam produced using the Variable Infiltration Capacity (VIC) Hydrologic model from a study by the Climate Impacts Group (CIG) at the University of Washington (Hamlet et al. 2010). In this case, the range of impacts for the projections (shown as pink bands in Figure A) relate directly to the uncertainty in the downscaled Global Climate Model (GCM) projected changes in temperature (T) and precipitation (P) for each emissions scenario and time period, because the statistical downscaling approach and hydrologic model used is identical for each ensemble member in the analysis. Within a particular study, changes in T and P are often key drivers of uncertainty (as above), however other sources of uncertainty come into play when comparing multiple studies. The potential differences in results between the CIG and PCIC investigations ultimately represent scientific uncertainties related to a) choice of emissions scenarios and GCM projections to include in the study, b) choice of statistical downscaling methods, c) different historical T and P driving data sets d) different hydrologic model parameter estimates for vegetation and soils e) different calibration strategies for the hydrologic models, and d) the inclusion of the effects of glaciers on the hydrologic cycle in the PCIC study. Although quantitative differences in the results in the two different modeling efforts are expected, a thorough investigation of the uncertainty resulting from these various assessment choices had not been previously undertaken. In this study, we carry out a detailed intercomparison of the hydrologic model projections from the existing CIG and PCIC hydrologic studies for the Canadian Columbia River basin. Key areas of consensus and divergence in these simulations will be identified using 18 hydroclimatic metrics (discussed below) in order to better understand information being used for decision making on water management (and other) issues in the Columbia Basin. |
