ClimateData.ca has released a major update to its historical and climate change-scaled intensity-duration-frequency (IDF) rainfall datasets, providing users across Canada with expanded access to information about extreme rainfall under current and future climate conditions.
The update includes 37 new stations, new 1-in-200-year return period data for both historical and climate change-scaled datasets, an additional emissions scenario (SSP3-7.0), and updated climate change-scaled rainfall estimates based on the latest CanDCS-M6 climate projections. The updated datasets also now include additional percentile ranges to support the recently released CSA W231:25 standard, “Developing and interpreting intensity-duration-frequency (IDF) information under a changing climate.”
Together, these updates provide users with more flexibility and more up-to-date information for infrastructure design, engineering applications, stormwater management, and climate risk assessment.
ClimateData.ca provides both historical IDF rainfall data and climate change-scaled rainfall estimates based on Environment and Climate Change Canada’s recommended temperature scaling methodology for short-duration rainfall extremes. To learn more about how climate change is integrated into IDF rainfall data, including the temperature scaling approach used on ClimateData.ca, see the Learning Zone article IDF Data and Climate Change.
Since the last update, 37 additional stations have reached the minimum 10-year record length to be included in the IDF rainfall dataset. These stations are now available through the Download page along with the original 700+ stations, providing both historical and climate change-scaled IDF rainfall data.
Figure 1: Map of stations across Canada with IDF rainfall data.
Users can now access 1-in-200-year return period data across Canada in addition to 2-, 5-, 10-, 25-, 50-, and 100-year return periods for both historical and climate change-scaled IDF rainfall data. Larger return periods, like a 1-in-200-year event, offer additional insight into possible extreme rainfall events, allowing users to select the level most relevant to their risk tolerance, context and needs.
To align with the new “Developing and interpreting intensity-duration-frequency (IDF) information under a changing climate” standard released in February 2025 (CSA W231:25), the first quartile (25th percentile) and third quartile (75th percentile) values have been added to the IDF rainfall data. Originally, the data included the 10th and 90th percentiles as well as the 50th percentile (median) and the scaled 95% confidence limit. The addition of the first and third quartiles allows users to easily comply with the new standard, while still having access to the originally provided percentiles.
Uncertainty in the estimation of past extreme rainfall rates, arising from having a limited period from which to estimate the distribution of rainfall extremes, is captured in part by the 95% confidence limits, which refers to the statistical estimate of the most likely true values of the “fitted extreme value” (Gumbel) distribution parameters. Simply put, users can trust that rainfall intensity values for a given duration and return period are likely to fall within the confidence interval that brackets the central IDF curve line.
The SSP3-7.0 scenario is a ‘high’ emissions scenario, sitting between SSP5-8.5 (highest emissions) and SSP2-4.5 (‘middle-of-the-road’ scenario). When you download IDF rainfall data from ClimateData.ca, you will now see data for SSP3-7.0, in addition to SSP5-8.5, SSP2-4.5, and SSP1-2.6 (low emissions). The availability of both SSP5-8.5 and SSP3-7.0 allows users to select the high-end scenario that best aligns with their specific needs. Having a range of emissions scenarios available ensures users have access to more data to consider a range of future climates to better suit their specific needs and decision-making contexts.
To learn more about SSPs, check out this Learning Zone article.
M6, short for “Canadian Downscaled Climate Scenarios-Multivariate dataset for CMIP6 (CanDCS-M6)” offers several improvements and enhanced capabilities over the previous generation of downscaled data, referred to as CanDCS-U6 or just U6. This dataset improves on U6 by taking the relationship between temperature and precipitation into account when downscaling and using an improved target dataset for bias-adjustment, PCIC-Blend. Climate change-scaled IDF rainfall data has been updated using this dataset. While it is recommended to use the latest CMIP6 data for new work requiring future climate projections, IDF rainfall data for CMIP5 is still valid.
Learn more about the M6 dataset and CMIP6.
While the data has been updated, the method for accessing it has not changed. Users can navigate to the IDF rainfall data from the Variable menu or directly from the Download page. For each station, there are several different products in zipped folders. The zipped folders include CSV (Excel) files with:
Each of these files will now contain updated IDF rainfall data, with the features noted above.
Updated excel file format
Improvements have also been made to the data format that you see when you open the CSV files. The new format is intended to improve the readability of the data. Rows are organized by rainfall duration (i.e. a 5min, 10min, 30 min, 1 hour, etc. rainfall event) for each percentile (i.e., median, 10th percentile, 1st quartile, etc.). The columns (columns ‘C’ to ‘I’ in the image below) are the return periods, which represent the frequency of occurrence of a rainfall event. The values under each return period represent rainfall intensities, or the rate of rainfall given in units of mm/hr.
Note: ’…’ in the table above indicates rows have been hidden in the excel file. Data exists for each of the durations listed for the ‘median’ for each percentile provided. Rows were hidden for illustrative purposes.
ClimateData.ca has also recently released projected return level data . One of the variables included in this new data (Return Level for Maximum 1-day Precipitation) has similarities to the IDF rainfall data as both are indicators of extreme precipitation. They use the same statistical method (fitting a Gumbel distribution to the data) and the CanDCS-M6 future projections to arrive at the extreme values. However, they also differ in several ways, which are important to understand if you are choosing between the two datasets.
To help you navigate them, see the table below for key information on IDF rainfall and Return Level data.
| Parameter | IDF Rainfall Data | Return Level Data (Maximum 1-day Precipitation) |
|---|---|---|
| Station or gridded | Station-based | Gridded |
| Duration (timeframe) | 5 minutes to 1 day (very short events) | 1 day |
| Methodology | Extreme values are calculated using historical data. Then, this information is “scaled” using future temperature data. See IDF Data and Climate Change for more information. | All calculations are done directly on projected data. See Understanding climate extremes: Return Periods and Return Levels for more information. |
| Useful if… |
|
|
There are many resources available on ClimateData.ca to get you started or to further your understanding of IDF rainfall data and extreme precipitation. Explore the following resources to learn more:
This project was undertaken with the support of:
