Abstract Manager

Wildfires represent a significant threat not only to natural landscapes and urban areas but also to highway network. The increasing frequency and intensity of wildfires, driven by climate change and other environmental factors, have posed a heightened risk to highway transportation. The 2023 wildfire season in Canada, marked as the most destructive in recorded history, has led to closures of multiple major highways across both eastern and western regions of the country.

In this presentation, we’ll introduce a wildfire risk prediction model specifically tailored to Canada’s highway network. Based on integrating diverse data sources, including the Fire Weather Index, the developed model has been applied to predict highway segments that could be threaten by wildfires in the coming week. The prediction results are compared with the active wildfire locations on the highway network extracted from VIIRS (Visible Infrared Imaging Radiometer Suite) satellite data. The result shows that the developed model can generate higher recall and less false alarms, for wildfires on the highway network, than the prediction based on the Fire Weather Index. The further improved model can be used by road users and regulators for mitigating wildfire risk.