Climate data and scenarios: synthesis of recent observation and modelling results, chapter 3

3. Future climate

The climate of the future will continue to experience natural variability, much as it has in the past. However, the background change in mean climate, already being driven by human activities, will continue at a rate that is determined primarily by current and future emissions of greenhouse gases and aerosols. Because future emissions are difficult to predict, it is necessary to use plausible scenarios, ranging from low to high emission pathways, to project future climate change. Global Earth System Models--which produce comprehensive computer simulations of the global climate system and the related carbon-cycle processes (see: Flato, 2011)--provide scientifically-based tools to make projections of future climate by simulating the response to atmospheric greenhouse gases and aerosols, land-use change, and other external forcings. Owing to uncertainties in the detailed representation of many complex climate processes, individual Earth System Models vary in their representation of these processes and will have biases of various kinds. Because of this, it is preferable to make use of a multi-model ensemble of projections for many applications. The average of a multi-model ensemble generally produces smaller historical errors than any individual model (Flato, et al., 2013) and the spread amongst models allows some quantification of uncertainty. The World Climate Research Programme,^Footnote1 (WCRP) coordinates multi-model climate projections via its Working Group on Coupled Modelling (WGCM) and the Coupled Model Intercomparison Project (CMIP^Footnote2). The results presented in the following sections are based on the CMIP5 results that were also featured in the Working Group I contribution to the IPCC Fifth Assessment Report (IPCC, 2013: see chapters 9, 11, and 12, and Annex I).

The CMIP5 projections make use of Representative Concentration Pathways (RCPs), which are designed to provide plausible future scenarios of anthropogenic forcing spanning a range from a low emission scenario characterized by active mitigation (RCP 2.6), through two intermediate scenarios (RCP 4.5 and RCP6), to a high emission scenario (RCP 8.5).^Footnote3 Figure 4 illustrates some of the assumptions underlying these scenarios. These scenarios make use of various combinations of projected population growth, economic activity, energy intensity, and socio-economic development. These, in turn, lead to calculations of energy consumption and related emissions and finally atmospheric concentrations of greenhouse gases and other climate forcings. These RCP scenarios serve as input to the Earth System Models, which simulate the climate system response and resulting climate conditions.

Figure 4 - Socioeconomic (top row), energy intensity (second row), greenhouse gas emission (third row), and ultimately greenhouse gas concentration (bottom row) assumptions underlying the representative concentration pathways (RCPs) used to drive future climate projections. From van Vuuren, et al., 2011, reproduced with permission.

A new feature of the IPCC Fifth Assessment Report (AR5) is the Atlas of Global and Regional Climate Projections (Annex 1--IPCC, 2013), which provides a synthesis of results from the CMIP5 multi-model ensemble. For application to Canadian impact studies and adaptation planning, the regional boundaries of the Atlas are less than optimal: western Canada is combined with the western United States and Alaska, and eastern Canada is combined with Greenland and Iceland (but separated from western Canada). We have therefore generated multi-model ensemble results specific to Canada, using output from 29 CMIP5 models from which results were available for historical simulations, RCP2.6, RCP4.5, and RCP8.5 (results for RCP6.0 are also available, but from fewer models; so this scenario is not illustrated here). Further details on the models used in this document are presented in Table 2.

The ensemble climate model results include output representing a broad range of climate variables. For example, model output includes temperature, precipitation, snow depth, ocean pH and salinity, soil moisture, downwelling solar radiation, and many other quantities. As an example, a full listing of results from the Canadian model (CanESM2) is available. In this document we focus on temperature and precipitation changes in the Canadian context.