As more than 8,000 wildfires rage across Canada every year — and the fire season grows ever longer and more intense — responding quickly and effectively with the right resources is more important than ever.
Containing fires within the first few hours, before they spiral out of control, is crucial, say experts. Yasser Zeinali, a doctoral candidate in the Alberta School of Business, is designing a rapid response tool to make initial attack operations faster and smoother.
Working with Alberta Wildfire, Zeinali is using predictive modelling — analyzing historical data with the help of AI — to forecast daily helicopter, air tanker and firefighter demands, as well as expected wildfire incident counts.
He is also using a “transient queueing model” to anticipate where bottlenecks or traffic jams might occur in the deployment of firefighting resources, as well as an “optimization model” to pre-position and redeploy resources across forest areas.
Together, these models will help wildfire managers know immediately after a forecast is issued how to allocate air tankers, helicopters and crews across regions efficiently while accounting for capacity limits, travel time and surge needs.
“The framework is designed to be operationally realistic and decision-ready, accounting for congestion and short planning horizons, and producing clear, actionable guidance for wildfire managers,” says Zeinali, adding that it can be applied across the country.