Learn more about the seasonal forecasts coming to ClimateData.ca and find answers to common questions.
FAQ
Seasonal forecasts predict likely future climatic conditions for a given season. They represent anomalies or differences between the conditions that are predicted for a season compared to the average conditions normally experienced for that season, as represented by the 30-year average of observations from 1991 to 2020. These forecasts are generally available for three-month periods over the following 12 months.
In meteorology and climatology, seasons are based on the annual temperature cycle and typically last three months. Standard meteorological seasons are defined as follows:
Seasonal forecasts get their name from the concept of a season based on a three-month period. On ClimateData.ca, forecasts are provided for these four meteorological seasons as well as overlapping three-month periods (e.g., January to March, February to April, April to June, May to July, July to September, August to October, October to December, November to January). Over a 12-month span, there are ten distinct seasonal forecasts. Figure 1 shows an example of forecasts released on June 1st, with 10 forecasts covering three-month periods out to 12 months. For more information, consult What are Seasonal Forecasts?.
Figure 1: Example showing forecasts released on June 1st. Seasonal forecasts overlap, with 10 forecasts covering rolling 3-month periods out to 12 months.
The most commonly forecasted variables are temperature and precipitation. Other variables include:
Daily weather conditions are unpredictable on seasonal timescales. For example, determining a precise value for daily temperature or precipitation on a given day, months from now, is impossible. However, it is possible to predict how conditions for a given season are likely to compare to normal. Seasonal forecasts provide insight into average conditions over three-month periods. They are presented in the form of probabilities of conditions being above, near, or below normal. ‘Normal’ is represented by average conditions observed over the 30-year period 1991 to 2020.
Yes. A new set of seasonal forecasts is issued at the beginning of each month, covering the upcoming 12 months. These forecasts replace those that were issued in the previous month. For example, seasonal forecasts produced in April 2025 cover the period from April 2025 to March 2026. Those produced in May 2025 cover the period from May 2025 to April 2026, and so on.
In Canada, official seasonal forecasts are produced by Environment and Climate Change Canada using the Canadian Seasonal to Interannual Prediction System (CanSIPS). For more information, consult What are Seasonal Forecasts? and the CanSIPS technical documentation.
Yes, observations are used to provide the seasonal forecasting system with initial atmospheric, terrestrial, and oceanic conditions. In addition to the initial conditions for the model simulations, observations are used for model calibration (to improve forecast reliability in post-processing) and to assess the skill of the model. Average conditions observed over the 30-year period from 1991-2020 are used to define ‘normal’.
No. To produce seasonal forecasts, the system relies directly on initial conditions based on recent climate observations. During model execution, these conditions evolve over time. Future emissions, as defined in various Shared Socio-economic Pathways (SSPs), are not used since the differences between the scenarios are negligible on a seasonal to annual timescale.
Seasonal forecasts project the likelihood of upcoming summer temperatures being above, near, or below normal compared to historical temperatures observed over a 30-year period (1991 to 2020). This temperature forecast indicates the probability that more, the same, or less air-conditioning will be required during the upcoming summer as compared to the average summer temperatures experienced during the 1991 to 2020 period. The accuracy of forecasts generally increases as lead times decrease (i.e., the time between the production of the forecast and the time period of interest). For example, for the months of July to August, a seasonal forecast released in June will be more accurate than one released in March. Consulting the forecasts as they are updated each month will help you better plan your expenses.
Seasonal forecasts are updated monthly. Each month, updated forecasts replace previous ones. For instance, a forecast for the summer season of June to August 2024 was first issued in September 2023 and the forecast was updated monthly until June 2024. The closer the seasonal forecast is to the time period of interest, the greater the accuracy of the forecast.