CanDCS-U6

1. Overview of CanDCS-U6

  • Purpose: Provides statistically downscaled daily climate projections for Canada.

  • Developed by: Environment and Climate Change Canada (ECCC) and the Pacific Climate Impacts Consortium (PCIC).

  • Geographic Coverage: Canada (national scale, grid-based).

  • Time Coverage:

    • Historical data: 1950–2014

    • Future projections: 2015–2100

  • Resolution: 1/12° (~10 km) spatial grid

  • Emissions Scenarios (SSPs):

    • SSP1-2.6 (low emissions, strong mitigation)

    • SSP2-4.5 (moderate emissions stabilization)

    • SSP5-8.5 (high emissions, worst-case scenario)

Key Climate Variables Modeled

Variable

Unit

Projected Change Metrics

Mean Temperature

°C

Absolute temperature change

Maximum Temperature

°C

Highest daily temperature

Minimum Temperature

°C

Lowest daily temperature

Total Precipitation

mm/day

% change relative to 1971-2000 baseline

2. Downscaling Methodology

2.1 Statistical Downscaling Process

CanDCS-U6 applies Bias Correction/Constructed Analogues with Quantile Mapping (BCCAQv2), a hybrid statistical downscaling method.

  • Process:

    1. Bias Correction (BC): Adjusts GCM output to remove systematic model biases.

    2. Constructed Analogues (CA): Identifies historical weather patterns similar to the model output.

    3. Quantile Mapping (QM): Matches GCM-simulated climate distributions to observed distributions.

  • Training Data Used:

    • NRCANmet/ANUSPLIN (observational gridded climate data for Canada)

    • Calibration period: 1951–2010

  • Future Projections:

    • Each CMIP6 model projection is downscaled separately.

    • Downscaling applied to daily temperature and precipitation.

3. Climate Projections and Uncertainty Representation

3.1 Multi-Model Ensemble

  • CanDCS-U6 includes 26 CMIP6 climate models (see Table 2 for the full list).

  • Each model is run independently, and ensemble statistics (percentiles) are computed.

3.2 Uncertainty Representation

  • 10th, 50th (median), and 90th percentile projections provided.

  • Range of uncertainty reflects model spread but does not capture all possible outcomes.

3.3 Time Periods Analyzed

  • Historical Reference: 1971–2000

  • Future Projections:

    • 2021–2050 (near-term)

    • 2041–2070 (mid-century)

    • 2071–2100 (end-of-century)

4. Climate Models Used in CanDCS-U6

CanDCS-U6 includes 26 global climate models (GCMs) from CMIP6.

Institution

Model Name

CSIRO (Australia)

ACCESS-CM2, ACCESS-ESM1-5

Beijing Climate Center (China)

BCC-CSM2-MR

Canadian Centre for Climate Modelling and Analysis (Canada)

CanESM5

Met Office Hadley Centre (UK)

HadGEM3-GC31-LL, UKESM1-0-LL

NOAA-GFDL (USA)

GFDL-ESM4

IPSL (France)

IPSL-CM6A-LR

MPI (Germany)

MPI-ESM1-2-HR, MPI-ESM1-2-LR

JAMSTEC (Japan)

MIROC6, MIROC-ES2L

Norwegian Climate Center (Norway)

NorESM2-LM, NorESM2-MM

(Full list available in Table 2 of CanDCS-U6 documentation.)

5. Applications of CanDCS-U6

5.1 Scientific Research & Climate Impact Studies

  • Analyzing regional temperature and precipitation changes.

  • Studying extreme weather events (heatwaves, heavy precipitation).

  • Evaluating seasonal variability and long-term climate shifts.

5.2 Government Policy & Adaptation Planning

  • Supporting Canada’s climate adaptation strategies.

  • Providing climate projections for infrastructure planning.

  • Developing flood and wildfire risk mitigation measures.

5.3 Industry & Resource Management

  • Energy Sector: Forecasting changes in hydropower and energy demand.

  • Agriculture: Assessing growing season length and drought risk.

  • Insurance & Risk Assessment: Evaluating climate-related financial risks.

6. Best Practices for Using CanDCS-U6 Data

  • Use a multi-model ensemble approach to account for uncertainty.

  • Compare different emission scenarios (SSP1-2.6, SSP2-4.5, SSP5-8.5).

  • Downscaling improves resolution but does not eliminate model uncertainty.

  • Do not use single-model results in isolation—ensemble percentiles provide better insight.

Use Limitations

  • CanDCS-U6 only includes daily temperature and precipitation.

  • Does not include other climate variables (wind, humidity, soil moisture, etc.).

  • Statistical downscaling assumes historical relationships remain valid under future climate.

7. Future Developments & Transition to CMIP7

  • Future updates will refine downscaling techniques and expand variables (e.g., wind, humidity).

  • Integration with CMIP7 datasets is planned to improve long-term projections.

8. Data Access & References

8.1 Where to Access CanDCS-U6 Data

8.2 References

  • Bush, E., Lemmen, D.S. (2019). Canada’s Changing Climate Report. Government of Canada.

  • Cannon, A.J., Sobie, S.R., Murdock, T.Q. (2015). Bias Correction & Quantile Mapping in CMIP6.

  • Werner, A.T., Cannon, A.J. (2016). Hydrologic Extremes – A Comparison of Statistical Downscaling.

  • Riahi, K. et al. (2017). Shared Socioeconomic Pathways & Climate Change.

9. Conclusion

The Canadian Downscaled Climate Scenarios (CanDCS-U6) provides high-resolution (10 km) statistically downscaled climate projections for Canada, using 26 CMIP6 climate models and three SSP emissions scenarios.

With improved downscaling techniques (BCCAQv2) and daily climate variables, CanDCS-U6 enhances local-scale climate impact assessments, supporting scientific research, policy-making, and industry adaptation strategies.

As climate science advances, CanDCS-U6 will continue to evolve, incorporating higher-resolution CMIP7 models and additional climate variables to further improve regional climate predictions.

PreviousCMIP6NextSPEI

Last updated

Was this helpful?