Updating the Canadian snow course database

Updating the database has contributed to improving and validating climate tools and products.

Project details
Scientific program
2014-2019 programming
Theme(s) and priority(s)
Climate Scenarios and services
Start and duration
October 2016 • September 2017
Project Status
Principal(s) investigator(s)
Ross Brown


In 2000, the Ministry of the Environment and Climate Change Canada (ECCC) developed a national snow course database (DB) for Canada based on data from provincial and federal observing programs.

This included a digitization of the annual “Snow Cover Data” summaries published in hard copy form from 1955 to 1985. The DB has often been used for validating the outputs from hydroclimatic models used, among other things, to carry out climate change impact studies. However, the DB was last updated in 2004 and does not include historical Hydro-Québec and MELCC data.



  • Update the snow course database (snow depth, snow water equivalent (SWE) and density) covering Canada to 2016;

  • Collect and include the main datasets that were omitted in previous updates. Standardize data file formats;

  • Check plausibility and quality of data. Document the new database;

  • Make the new database available to partners of Ouranos for their research and application needs.


  • Obtain data and metadata of snow course observations by Canadian agencies;

  • Process data/metadata files in standard format (CSV);

  • Perform a quality control on the data; Group and sort the data;

  • Convert the data to standard NetCDF format;

  • Document the data and updating procedures; Write a scientific article describing the update and an analysis of the database.


The update expanded the spatial distribution of SWE observations in the Yukon, the Northwest Territories and southern Québec (Fig. 1).

Evaluation of SWE observation quality showed that less than 3% of manual observations failed the tests. In contrast, daily SWE observations by the British Columbia Ministry of Environment using snow pillows showed a higher error rate. There are several reasons for these numerous measurement errors, including the “blocking effect”, where the increase in the weight of the snow on the snow pillow is reduced or even blocked by a more resistant layer of snow.

Figure 1

Figure 1. Spatial distribution of data before and after the update.

The trend analysis of the data is complicated by the high temporal and spatial variability of the observations. The preliminary results for southwestern Québec, for which consistent observations have been available since 1965, show a significant decrease in SWE for the January-February period, but little change for the maximum accumulation period in March-April (Fig. 2).

Figure 2

Figure 2. Mean SWE over southwestern Québec (45.5-46.0°N, 74.0-75.0°O) for 1965-1989 and 1990-2015

The interpolation of the data on a 10 km grid by Brown et al. (2019) made it possible to analyze trends in water equivalent, snow depth and density over the 50-year period from 1967 to 2016. The results revealed large spatial variability in trend sign and strength, with a relatively small percentage of points showing statistically significant trends. Any significant SWE and snow depth trends tended to be negative. This is consistent with existing studies on snow cover changes in Canada.

The results show a latitudinal dependence in SWE trends (Fig. 3), with the largest negative trends occurring at lower latitudes, and a tendency for mainly positive trends in Arctic SWE. This is consistent with observations from Russia and climate model projections of the response of Arctic snow cover to climate change (Brown et al., 2019).

Figure 3

Figure 3. SWE trend on March 1 (1967-2016) versus latitude (°N). SWE trend is measured in percentage per decade relative to the mean SWE for 1967-2016. Brown et al. (2019)

The database was used to evaluate SWE products on the St-Maurice River watershed including the Canadian Regional Climate Model (Version 5), satellite data, reanalysis, and operational Canadian Meteorological Centre snow depth forecasts. The evaluation showed that most of the products underestimated SWE across the watershed because of a significant underestimate of solid precipitation (Brown et al., 2018, Evaluation of snow water equivalent datasets over the Saint-Maurice river basin, Hydrological Processes, 32: 2748–2764, https://doi.org/10.1002/hyp.13221).

Benefits for adaptation

Benefits for adaptation

Updating the database has contributed to improving and validating climate tools and products.

Snow cover and its water equivalent (SWE) have an influence for many sectors of economic activity in Canada, such as agriculture (frost penetration), winter tourism, snow removal operations and water reservoir management in spring.

Scientific publications

Document type
Update of Canadian historical snow survey dataset. Project report, Environment and Climate Change…
Fang, B.
Update of Canadian Historical Snow Survey Data and Analysis of Snow Water Equivalent Trends, 1967…
Brown, R.D., Fang, B. and Mudryk, L.


Other participants

  • Churchill Falls (Labrador) Corporation Limited

  • Environment and Parks, Alberta

  • Environment Yukon

  • Environment and Local Government, New Brunswick

  • Hydro-Québec

  • Ministère du Développement durable, Environnement et Lutte contre les changements climatiques (MDDELCC)

  • Ministry of Environment, BC Ministry of Natural Resources and Forestry, Ontario

  • Ontario Power Generation

Related projects

button back to top