Environment Canada to use AI in new weather forecasting model
listen to this article
estimated 3 minutes
The audio version of this article has been generated by AI-based technology. There may be mispronunciations. We are working with our partners to continually review and improve results.
Environment and Climate Change Canada will use artificial intelligence to make its weather forecasts more accurate, the federal department said Thursday.
It plans to launch a hybrid model this spring that uses both AI and traditional forecasting and says combining the two will result in more accurate predictions.
“The new hybrid model relies on AI to better predict future weather conditions, as well as traditional physics-based models to bring together our knowledge of unique local factors such as wind, temperature and precipitation,” it said in a news release.
AI can analyze decades of historical data covering an entire continent in minutes, the department said, adding that AI models “identify relationships between temperature, wind and pressure and use those learned patterns to predict the future climate, especially for major weather events like heat waves, or to track hurricanes.”
The release said the hybrid model is better at predicting extreme weather, such as strong winds or heat waves, because the traditional model “captures small-scale details that AI models miss.”
Experts offer safety tips to stay safe and protect your property, food, equipment and vehicles during adverse weather events in Quebec.
With the new model, Environment Canada said its six-day forecast will be as accurate as its five-day forecast. It added that this is “particularly important because previous improvements in forecasting were only possible after many years of research and development.”
It said the hybrid system would also be faster at predicting major systems such as winter storms, heat waves and atmospheric rivers.
Environment Canada said that since last year, its “scientists and meteorologists have been conducting extensive testing on the hybrid model, running it in parallel with our traditional models to evaluate its performance for predicting weather conditions in Canada.”
The department said it would continue to rely on its meteorologists, whose judgment is “critical” in interpreting the results and communicating them to the public.
“The fact that so much climate data can be analyzed so quickly and incorporated into a product is something we can use,” said Halifax-based Cindy Day, who has been a meteorologist for more than 40 years.
Day said being able to identify systems earlier will benefit public safety by giving people more warning of major storms.
But he questioned how helpful historical data can be in looking at the impacts of climate change.
“The rate at which our temperatures are changing and our climate is changing is significant. And so I’m not sure that analyzing so much data from so long ago so quickly is going to make a big difference in producing forecasts for the next five, six, seven days,” Day said.
