Statistically downscaled multi-model ensembles of maximum temperature are available at a 10km spatial resolution for 1951-2100. Statistically downscaled ensembles are based on output from twenty-four Coupled Model Intercomparison Project Phase 5 (CMIP5) global climate models (GCM). Daily maximum temperature from GCM outputs were downscaled using the Bias Correction/Constructed Analogues with Quantile mapping version 2 (BCCAQv2). A historical gridded maximum temperature dataset of Canada (ANUSPLIN) was used as the downscaling target. The 5th, 25th, 50th, 75th and 95th percentiles of the monthly, seasonal and annual ensembles of downscaled maximum temperature (°C) are available for the historical time period, 1951-2005, and for emission scenarios, RCP2.6, RCP4.5 and RCP8.5, for 2006-2100. Note: Projections among climate models can vary because of differences in their underlying representation of earth system processes. Thus, the use of a multi-model ensemble approach has been demonstrated in recent scientific literature to likely provide better projected climate change information.

Statistically downscaled multi-model ensembles of mean temperature are available at a 10km spatial resolution for 1951-2100. Statistically downscaled ensembles are based on output from twenty-four Coupled Model Intercomparison Project Phase 5 (CMIP5) global climate models (GCM). Downscaled daily mean temperature was calculated by averaging downscaled daily minimum and maximum temperature. Daily minimum and maximum temperature from GCM outputs were downscaled using the Bias Correction/Constructed Analogues with Quantile mapping version 2 (BCCAQv2). Historical gridded minimum and maximum temperature datasets of Canada (ANUSPLIN) were used as the respective downscaling targets. The 5th, 25th, 50th, 75th and 95th percentiles of the monthly, seasonal and annual ensembles of downscaled mean temperature (°C) are available for the historical time period, 1951-2005, and for emission scenarios, RCP2.6, RCP4.5 and RCP8.5, for 2006-2100. Note: Projections among climate models can vary because of differences in their underlying representation of earth system processes. Thus, the use of a multi-model ensemble approach has been demonstrated in recent scientific literature to likely provide better projected climate change information.

Environment and Climate Change Canada’s (ECCC) Climate Research Division (CRD) and the Pacific Climate Impacts Consortium (PCIC) previously produced statistically downscaled climate scenarios based on simulations from climate models that participated in the Coupled Model Intercomparison Project phase 5 (CMIP5) in 2015. ECCC and PCIC have now updated the CMIP5-based downscaled scenarios with two new sets of downscaled scenarios based on the next generation of climate projections from the Coupled Model Intercomparison Project phase 6 (CMIP6). The scenarios are named Canadian Downscaled Climate Scenarios–Univariate method from CMIP6 (CanDCS-U6) and Canadian Downscaled Climate Scenarios–Multivariate method from CMIP6 (CanDCS-M6). CMIP6 climate projections are based on both updated global climate models and new emissions scenarios called “Shared Socioeconomic Pathways” (SSPs). Statistically downscaled datasets have been produced from 26 CMIP6 global climate models (GCMs) under three different emission scenarios (i.e., SSP1-2.6, SSP2-4.5, and SSP5-8.5), with PCIC later adding SSP3-7.0 to the CanDCS-M6 dataset. The CanDCS-U6 was downscaled using the Bias Correction/Constructed Analogues with Quantile mapping version 2 (BCCAQv2) procedure, and CanDCS-M6 was downscaled using the N-dimensional Multivariate Bias Correction (MBCn) method. The CanDCS-U6 dataset was produced using the same downscaling target data (NRCANmet) as the CMIP5-based downscaled scenarios, while the CanDCS-M6 dataset implements a new target dataset (ANUSPLIN and PNWNAmet blended dataset). Statistically downscaled individual model output and ensembles are available for download. Downscaled climate indices are available across Canada at 10km grid spatial resolution for the 1950-2014 historical period and for the 2015-2100 period following each of the three emission scenarios. Note: projected future changes by statistically downscaled products are not necessarily more credible than those by the underlying climate model outputs. In many cases, especially for absolute threshold-based indices, projections based on downscaled data have a smaller spread because of the removal of model biases. However, this is not the case for all indices. Downscaling from GCM resolution to the fine resolution needed for impacts assessment increases the level of spatial detail and temporal variability to better match observations. Since these adjustments are GCM dependent, the resulting indices could have a wider spread when computed from downscaled data as compared to those directly computed from GCM output. In the latter case, it is not the downscaling procedure that makes future projection more uncertain; rather, it is indicative of higher variability associated with finer spatial scale. Individual model datasets and all related derived products are subject to the terms of use (https://pcmdi.llnl.gov/CMIP6/TermsOfUse/TermsOfUse6-1.html) of the source organization.

Multi-model ensembles of mean temperature based on projections from twenty-nine Coupled Model Intercomparison Project Phase 5 (CMIP5) global climate models are available for 1901-2100. Specifically, the 5th, 25th, 50th, 75th and 95th percentiles of the monthly, seasonal and annual ensembles of mean temperature (°C) are available for the historical time period, 1901-2005, and for emission scenarios, RCP2.6, RCP4.5 and RCP8.5, for 2006-2100. Note: Projections among climate models can vary because of differences in their underlying representation of earth system processes. Thus, the use of a multi-model ensemble approach has been demonstrated in recent scientific literature to likely provide better projected climate change information.

Argo is a key component of the Global Ocean Observing System (GOOS) with an array of about 4,000 autonomous instruments reporting on ocean conditions. These floats collect data on ocean temperature and salinity, and in some cases, additional properties that characterize the ocean’s biological and chemical processes. Established in 1999, Argo represents an international collaboration involving contribution from more than 30 nations. Data from Argo floats are made publicly available within 24 hours of collection time, for free. The data provide valuable information on changes to the Earth's climate and hydrological cycle. They are used for a variety of purposes, such as assessing climate change, improving weather forecasts and developing ocean models. Argo Canada, led by Fisheries and Oceans Canada, has been a key contributor to the International Argo Program since its inception in 2001 . The program has been supported by contributions from Department of Environment and Climate Change Canada, Department of National Defense, Dalhousie University, University of Victoria and Ocean Networks Canada.

Seasonal and annual multi-model ensembles of projected change (also known as anomalies) in sea ice thickness, based on an ensemble of twenty-six Coupled Model Intercomparison Project Phase 5 (CMIP5) global climate models are available for 1900-2100. Projected change in sea ice thickness is with respect to the reference period of 1986-2005 and expressed as a percentage (%). The 5th, 25th, 50th, 75th and 95th percentiles of the ensemble of sea ice thickness change are available for the historical time period, 1900-2005, and for emission scenarios, RCP2.6, RCP4.5 and RCP8.5, for 2006-2100. Twenty-year average changes in sea ice thickness (%) for four time periods (2021-2040; 2041-2060; 2061-2080; 2081-2100), with respect to the reference period of 1986-2005, for RCP2.6, RCP4.5 and RCP8.5 are also available in a range of formats. The median projected change across the ensemble of CMIP5 climate models is provided. Note: Projections among climate models can vary because of differences in their underlying representation of earth system processes. Thus, the use of a multi-model ensemble approach has been demonstrated in recent scientific literature to likely provide better projected climate change information.

Environment and Climate Change Canada’s (ECCC) CMIP6 statistically downscaled agroclimatic indices are an updated version of the CMIP5 agroclimatic indices dataset making use of two different sets of downscaled scenarios created by the Pacific Climate Impacts Consortium (PCIC): 1. Canadian Downscaled Climate Scenarios–Univariate method from CMIP6 (CanDCS-U6), and 2. Canadian Downscaled Climate Scenarios–Multivariate method from CMIP6 (CanDCS-M6). To address the needs of different user groups in Canada, 49 indices, including agroclimatic indices, were proposed by the Canadian adaptation community through a series of consultations. Please see the definition list for the equations of each index. In 2025, PCIC expanded the CMIP6 agroclimatic indices, by adding CanDCS-M6, which includes SSP3-7.0 for most GCMs. Additionally, PCIC introduced 18 new indices to the previous 49. The 67 indices are available for both the CanDCS-U6 and CanDCS-M6. The range of impact-relevant climate indices available for download includes, indices representing counts of the number of days when temperature or precipitation exceeds (or is below) a threshold value; the episode length when a particular weather/climate condition occurs; and indices that accumulate temperature departures above or below a fixed threshold. The statistically downscaled climate indices are available for individual models and ensembles, historical simulations (1951-2014) and three emissions scenarios called “Shared Socioeconomic Pathways” (SSPs), SSP1-2.6, SSP2-4.5, and SSP5-8.5 (2015-2100), at a 10 x 10 km degree grid resolution. The CanDCS-M6 agroclimatic indices dataset also includes SSP3-7.0 results. Note: projected future changes by statistically downscaled products are not necessarily more credible than those by the underlying climate model outputs. In many cases, especially for absolute threshold-based indices, projections based on downscaled data have a smaller spread because of the removal of model biases. However, this is not the case for all indices. Downscaling from GCM resolution to the fine resolution needed for impact assessment increases the level of spatial detail and temporal variability to better match observations. Since these adjustments are GCM dependent, the resulting indices could have a wider spread when computed from downscaled data as compared to those directly computed from GCM output. In the latter case, it is not the downscaling procedure that makes future projection more uncertain; rather, it is indicative of higher variability associated with a finer spatial scale. Individual model datasets and all related derived products are subject to the terms of use (https://pcmdi.llnl.gov/CMIP6/TermsOfUse/TermsOfUse6-1.html) of the source organization.

Multi-model ensembles of surface wind speed based on projections from twenty-nine Coupled Model Intercomparison Project Phase 5 (CMIP5) global climate models are available for 1900-2100. Specifically, the 5th, 25th, 50th, 75th and 95th percentiles of the monthly, seasonal and annual ensembles of surface wind speed (m/s) are available for the historical time period, 1900-2005, and for emission scenarios, RCP2.6, RCP4.5 and RCP8.5, for 2006-2100. Note: Projections among climate models can vary because of differences in their underlying representation of earth system processes. Thus, the use of a multi-model ensemble approach has been demonstrated in recent scientific literature to likely provide better-projected climate change information.

The High Resolution Deterministic Land Prediction System (HRDLPS) produces high-resolution medium-range forecasts of land surface, subsurface variables, and of near-surface atmospheric variables (1.5 m temperature and dewpoint, 10 m wind). HRDLPS is initialized with analysis and trial fields provided by the Canadian Land Data Assimilation System of the National Surface and River Prediction System (CaLDAS-NSRPS). The system is then driven with atmospheric forecasts provided by the HRDPS over the first two days of integration and by the GDPS over the next four days. Predictions are performed twice a day. The system runs on a grid with a 2.5 km horizontal spacing covering Canada and part of the USA.

Multi-model ensembles for a suite of variables based on projections from Coupled Model Intercomparison Project Phase 6 (CMIP6) global climate models (GCMs) are available for 1850-2100 on a common 1x1 degree global grid. Climate projections vary across GCMs due to differences in the representation and approximation of earth systems and processes, and natural variability and uncertainty regarding future climate drivers. Thus, there is no single best climate model. Rather, using results from an ensemble of models (e.g., taking the average) is best practice, as an ensemble takes model uncertainty into account and provides more reliable climate projections. Provided on Canadian Climate Data and Scenarios (CCDS) are four types of products based on the CMIP6 multi-model ensembles: time series datasets and plots, maps and associated datasets, tabular datasets, and global gridded datasets. Monthly, seasonal, and annual ensembles are available for up to six Shared Socioeconomic Pathways (SSPs) (SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0, SSP4-6.0, and SSP5-8.5), four future periods (near-term (2021-2040), mid-term (2041-2060 and 2061-2080), end of century (2081-2100)), and up to five percentiles (5th, 25th, 50th (median), 75th, and 95th) of the CMIP6 ensemble distribution. The number of models in each ensemble differs according to model availability for each SSP and variable, see the model list resource for details on the models included in each ensemble. The majority of products show projected changes expressed as anomalies according to a historical reference period of 1995-2014. The products provided include global, national, and provincial/territorial datasets and graphics. For more information on the CMIP6 multi-model ensembles, see the technical documentation resource.