1. Global Climate Model (GCM) Scenarios for Climate Change Impacts Research Trevor Murdock, M.Sc. tmurdock@uvic.ca tmurdock@uvic.ca Canadian Institute for Climate Studies www.cics.uvic.ca www.cics.uvic.ca 14 Sep 2004

2. Outline 1.Context & definitions 2.Global Climate Models (GCMs) ScenariosScenarios UncertaintyUncertainty EmissionsEmissions VegetationVegetation 3.Applying scenarios to impacts studies Downscaling: RCMs & other approachesDownscaling: RCMs & other approaches Climate normalsClimate normals 4.Scenarios for BC 5.Summary

3. Context and definitions Identify Vulnerabilities - what aspects of climate change is a community / region susceptible to?Identify Vulnerabilities - what aspects of climate change is a community / region susceptible to? Study potential Impacts of climate change; projections of future climate are required; use a range of Scenarios from Global Climate Models (GCMs) to deal with uncertaintyStudy potential Impacts of climate change; projections of future climate are required; use a range of Scenarios from Global Climate Models (GCMs) to deal with uncertainty Adaptation strategies involve managing for current and potential future climate (change and variability) impactsAdaptation strategies involve managing for current and potential future climate (change and variability) impacts

4. What are Global Climate Models? GCMs compute global weather patterns several times per day projected over the next century GCMs are the “…only credible tools currently available for simulating the physical processes that determine global climate...” [IPCC]

5. Sources of Uncertainty [Source: Hadley Centre for Climate Prediction and Research, UK Met. Office]

6. Emissions Scenarios –Emissions scenarios based on different assumptions about how the global economy will evolve and emit fossil fuels over the next century –SRES recommended –IS92a gg & ga also available –Naming convention also denotes members of ensembles with numbers and ensemble averages with x (i.e. CGCM2 A11, A12, A13, A1x)

7. Emissions & Concentrations

8. GCMs GCMs Each GCM has different parameterizations of physics of the Atmosphere, Ocean, Cryosphere & Biosphere CGCM1 Canadian Centre for Climate Modelling and Analysis Global Coupled Model 1 CGCM2 Canadian Centre for Climate Modelling and Analysis Global Coupled Model 2 HadCM2 Hadley Centre Coupled Model 2 HadCM3 Hadley Centre Coupled Model 3 GFDLR15 Geophysical Fluid Dynamics Laboratory R15 GFDLR30 Geophysical Fluid Dynamics Laboratory R30 ECHAM4 European Centre/Hamburg Model 4 CSIROMk2b Commonwealth Scientific Industrial Research Organization Mk2b CCSRNIES Center for Climate Research - National Institute for Environmental Studies NCARPCM National Center for Atmospheric Research

9. Vegetation in GCMs GCMs typically ignore climate/vegetation feedbacks:GCMs typically ignore climate/vegetation feedbacks: CCCMa (Canada)CCCMa (Canada) –CGCM2 - no vegetation –CGCM3 - simple CLASS scheme –CGCM4 - will include more sophisticated vegetation and biophysical processes Hadley Center (UK)Hadley Center (UK) –HadCM3 includes representation of freezing and melting of soil moisture and evaporation includes the stomatal resistance on temperature, vapour pressure, and CO 2 concentration (Cox et al., 1998)

10. Applying Scenarios: Downscaling GCM scenarios coarse resolution (100s of kms / monthly) Dynamic methods retain internal physical consistency high resolution AGCMs, Regional Climate Models (RCMs) Statistical methods less costly/less complicated Weather generators – LARS-WG, Multiple linear regression - SDSM Best solution often not to downscale at all interpolate (introduces false geographical precision) apply change fields from larger spatial scale to working scale

11. Regional Climate Models 1. 1.Account for sub-grid scale forcings such as topography and land cover in a physically-based way 2. 2.Note: more physics can mean more uncertainty

12. Applying Scenarios: Climate Normals - “Baseline Climates” - Usually 1961-1990 - Good baseline data needed for 2 reasons: -GCM scenario differences need to be applied to an observed baseline (to remove model bias). -Impacts assessment should include analysis of recent climate - Types of observational baselines: -Individual station data (raw, homogenized) -Gridded station data (interpolated) -Gridded satellite data -Gridded reanalysis data (statistical/dynamical/modeling) -PRISM 4km x 4km ( or resampled to 2km x 2 km ) grid – Temperature, Rainfall, Snowfall

13. CCIS – Custom Regions http://www.cics.uvic.ca/scenarios/data/select.cgi Predefined regions or create by clicking on mapPredefined regions or create by clicking on map Control panel interface (rather than steps)Control panel interface (rather than steps) Dynamic map creation allows for user customization of many features (legend, decimal places, grid, etc.)Dynamic map creation allows for user customization of many features (legend, decimal places, grid, etc.) Meta-information about full map and region (min, max, median, area-weighted mean, stddev)Meta-information about full map and region (min, max, median, area-weighted mean, stddev)

14. http://www.cics.uvic.ca/scenarios/data/select.cgi

17. Summary Impacts work most effective in the context of Vulnerability, Impacts, and AdaptationImpacts work most effective in the context of Vulnerability, Impacts, and Adaptation GCM based scenarios used to represent range of plausible future climates for impacts studiesGCM based scenarios used to represent range of plausible future climates for impacts studies Vegetation beginning to be included explicitly in GCMsVegetation beginning to be included explicitly in GCMs Downscaling may be used to overcome differences in scale between GCMs and impacts analysisDownscaling may be used to overcome differences in scale between GCMs and impacts analysis GCM change fields are applied to climate normalsGCM change fields are applied to climate normals Scenarios for BC – see www.cics.uvic.ca/scenarios for more tools, scenarios, and dataScenarios for BC – see www.cics.uvic.ca/scenarios for more tools, scenarios, and datawww.cics.uvic.ca/scenarios