How Operations Research can manage wildfire risk

Wildfires have dominated news headlines this summer, both globally and within the UK.  

In Canada, over 3 million hectares have burned in 2026, with major air quality impacts from the fires being reported across North America. In Greece, around 8,000 people were evacuated from areas on Crete, and several firefighters tragically lost their lives during efforts to contain the fires. Hospital units were evacuated and entire villages displaced in Turkey during July, and in California, over 4,000 wildfire incidents have been recorded to date this year.  

Closer to home, the UK experienced wildfires across the country, from devastating blazes in the Cairngorms National Park in the Scottish Highlands, to heathland fires in Oxfordshire and on the Isle of Wight. 

With predictions that heatwaves are to become even hotter, the risk of wildfires is expected to increase. While the immediate destruction caused by wildfires is often the most visible consequence, their effects are more widespread and long-lasting.  

Wildfires can devastate ecosystems, destroy wildlife habitats, and threaten biodiversity. They can also have serious consequences for human health and wellbeing. People may be exposed to harmful smoke and pollutants, and, in some cases, fires can contaminate water supplies and generate toxic waste that persists long after the flames have been extinguished. 

As the risk of wildfires increases, emergency services and local authorities are also facing budget and resource constraints. Effective planning, prevention and response strategies are needed to reduce wildfire risk and minimise their impacts when they occur.  

Wildlife risk management typically involves five interconnected stages: 

• Prediction e.g. where and when a wildfire is likely to occur and how it is likely to spread 

• Prevention e.g. prescribed burning, fire breaks and fuel management   

• Preparation e.g. resource planning and evacuation planning 

• Response e.g. fire suppression, evacuating people and/or asset protection 

• Recovery e.g. assess damages and restore affected areas. 

In the UK, the Home Office’s Wildfire Action Plan aims to address several of these areas through improvements to risk assessment, emergency response capabilities, and land management practices to reduce wildfire impacts. 

Using Operations Research to support wildfire management 

All five of these actions require effective planning, resource management and coordination between multiple stakeholders. Operations Research models and methods – like those we use in the Centre for Technology, Operations and Supply Chain Research (TOSCA) – can help with this planning and decision making. They include: 

• Facility Location Problems: Identifying the most suitable locations for fire stations, information points and emergency response hubs. 

• Vehicle Routing Problems: Determining the most effective order and routes for deploying firefighting resources, helping responders decide which fires to tackle first, and how to move between affected locations efficiently. 

• Network Flow Problems: Designing the most effective evacuation routes to move people safely and minimise traffic congestion during wildfire emergencies. 

• Resource Allocation Problems: Deciding how best to allocate firefighters, equipment and budgets across fire stations and affected areas to maximise response effectiveness. 

• Scheduling Problems: Planning the timing of preventive activities, such as prescribed burning, as well as co-ordinating the deployment and movement of firefighting personnel and resources. 

• Inventory Planning: Managing stocks of critical resources, such as fire retardants and emergency equipment, including decisions about how much to store and when to replenish supplies. 

• Queuing Theory: Helping emergency services manage wildfire-related incidents alongside routine calls, ensuring urgent requests can be prioritised and handled efficiently. 

• Game Theory: Supporting decision-making between multiple stakeholders, including emergency services, local authorities, healthcare providers and environmental agencies, to identify solutions that balance competing priorities. 

• Stochastic and Robust Optimisation: Enabling planners to prepare for uncertain factors such as weather conditions, fire severity and fire spread, helping them develop strategies that will be effective across a range of scenarios. 

As an example, I previously led a British Academy-funded project exploring “revolutionising forest fire risk management: harnessing AI and new technologies for enhanced safety and resilience.” Building on this experience, researchers in TOSCA and I are interested in working with organisations, public bodies and other stakeholders to develop future research and practical solutions for wildfire risk management. If you are interested in exploring a potential collaboration on this topic, please get in touch with me at M.Bashiri@leeds.ac.uk.