This project used a regional climate model, the Canadian Regional Climate Model (CRCM), to estimate future hydrologic conditions under the influence of climate variability and change. It had the following specific objectives:
- to validate the modelled water balance of the CRCM in selected BC watersheds
- to simulate future climate and hydrologic conditions for the 2050s (2041-2070)
- to contribute to estimates of future streamflow in the chosen watersheds
The selected watersheds for this project included the Upper Columbia and Upper Peace rivers in British Columbia. Results for the smaller Campbell watershed were also reviewed, but were not considered reliable due to its small size relative to the spatial resolution used by the CRCM. Results for other watersheds, the Upper Fraser and the entire Columbia Basin, were computed as a reference.
Methods
The SRES A2 emissions scenario was used to drive several versions of a Canadian coupled global climate model (CGCM) developed by Environment Canada. Those results were used subsequently to drive several different versions of the Canadian Regional Climate Model (CRCM) at 45 km horizontal resolution. The CRCM employs an implicit Land Surface Scheme (LSS) to resolve hydrologic components like surface evapotranspiration, snow accumulation, and surface runoff.
Modelling results were compared to historical records from the period 1961-1990 to determine model bias, the effects of internal (and natural) variability, and to test for structural uncertainty due to different parameterizations of physical processes. These results were used to estimate the overall uncertainty of hydro-climatic projections in the future (2041-2070). The experimental setup allowed the estimation of several sources of uncertainty associated with hydro-climatic projections. A Results Matrix was developed as an analysis tool for systematically assessing hydrologic impacts and how they vary from climate model results.
Results
Monthly mean time series of historical and future conditions were examined for bias, internal variability, and climate change signal. Based on annual average results from the integrated response of both the global and regional climate models for the selected watersheds the following observations can be made:
- The CGCM has a significant cold bias that is accompanied by excess snow accumulation during winter. However, this bias may be removed by considering the differences between the future and historical projections taken from the same model.
- Most of the internal variability comes from the global climate model, which is the source of our best estimate of the natural variability in the real climate system. The influence of internal variability on the projection of future hydrological change for the selected watersheds is estimated to be less than 6% for both snow accumulation and runoff.
- Climate change is projected to decrease annual average snow accumulation in the Upper Peace by 2%, in the Upper Columbia by 6% and almost 20% for the entire Columbia Basin in the 2050s, relative to the baseline period 1961-1990 (Figure 2). The peak period in surface runoff is projected to occur earlier in the spring and runoff is estimated to increase by 17-18% for the Upper Peace and Fraser watersheds, 9% for the Upper Columbia, and 7% for the entire Columbia Basin.
Figure 1 – Map showing CRCM projected changes and associated uncertainties in water budget components from the historical period (1961-1990) compared to the 2050s (2041-2070) over the Upper Peace, Fraser, Campbell and Columbia watersheds, using the A2 emissions scenario. (B.Music, Ouranos)
Selected Recommendations
A higher resolution model of regional climate is needed and a new 15 km Pacific Grid has been defined for performing additional computations. Finally, the results of this project on Climate Model Diagnostics will be compared to the traditional methods of hydrologic modelling, culminating in a synthesis report to be published in April 2011.
Additional Information
Rodenhuis, D., B. Music, M. Braun, and D. Caya, 2011. Climate Diagnostics of Future Water Resources in BC Watersheds. Pacific Climate Impacts Consortium, University of Victoria. 74 pp.
Acknowledgements
Dr. Daniel Caya contributed to the design of this project and Dr. Biljana Music of Ouranos and Dr. Marco Braun, UQAM, analyzed the model output. Arelia Werner and Katrina Bennett of PCIC’s Hydrologic Impacts group provided advice and defined the target watersheds examined in the project.