Environment and Climate Change Canada announced on Thursday its intention to enhance the accuracy of its weather forecasts by incorporating artificial intelligence technology. The department is set to introduce a hybrid model in the upcoming spring season, merging AI capabilities with traditional forecasting methods to deliver more precise predictions.
According to a news release, the new hybrid model combines AI algorithms to forecast future weather conditions and traditional physics-based models to account for local factors like wind, temperature, and precipitation. By leveraging AI’s ability to swiftly analyze extensive historical data spanning entire continents, the department highlighted that AI models can establish correlations between temperature, wind, and pressure to anticipate future weather phenomena, particularly for significant events such as heatwaves and hurricanes.
The hybrid model is particularly adept at predicting extreme weather occurrences like strong winds and heatwaves by complementing the AI’s broad-scale analysis with the traditional model’s focus on intricate details often overlooked by AI systems, as stated in the release.
Environment Canada emphasized that the hybrid model will elevate the accuracy of its six-day forecasts to the level of its previous five-day forecasts, marking a significant advancement that typically required years of research and development. Additionally, the department noted that the hybrid system will expedite the prediction of major weather systems like winter storms, heatwaves, and atmospheric rivers.
Over the past year, Environment Canada conducted extensive testing on the hybrid model, running it parallel to the traditional model to assess its efficacy in predicting Canadian weather conditions. The department affirmed its continued reliance on meteorologists, underscoring their critical role in interpreting forecast results and effectively communicating them to the public.
Meteorologist Cindy Day, with over 40 years of experience and based in Halifax, lauded the swift analysis of vast climate data facilitated by AI technology, expressing excitement over the potential benefits for public safety in early storm identification. However, she raised concerns about the applicability of historical data in light of climate change’s rapid effects, questioning its impact on forecasting accuracy for the upcoming days.