Closing the gap between wildfire modeling and real-world response

Canada's 2023 wildfire season scorched over 18.5 million hectares of land - an area larger than Greece. As these destructive fires become more frequent and unpredictable, a team of researchers from the University of British Columbia Okanagan is examining how operations research can improve real-world responses. The review, published in Operations Research Forum, highlights how mathematical modeling could transform wildfire management.

However, the researchers found that current models struggle to transition from academic journals to practical applications on the ground. Undergraduates Paria Rostamian and Kibele Sebnem Yildirim, alongside professors Amin Ahmadi Digehsara and Amir Ardestani-Jaafari, compiled 177 peer-reviewed studies from 2000 to 2024, focusing on how operations research can support decisions related to wildfires.

The team notes that despite the power of these mathematical tools, data availability remains a major barrier. Poor data quality, such as unreliable satellite imagery and inconsistent data from active fire zones, hinders effective decision-making. Even when models are robust, they often remain theoretical, never reaching operational use.

The researchers suggest that future progress will likely come from blending operations research with artificial intelligence and machine learning. They also recommend investing in standardized wildfire data and encouraging closer collaboration between researchers and fire agencies to ensure models are validated in real-world operational settings.

Ultimately, the researchers emphasize the need for interpretable and transparent systems that can help frontline responders trust and act upon model recommendations.