Redrawing the wildfire evacuation map

The 2026 wildfire season is a reminder that communities everywhere are becoming more exposed to extreme fire risk. More than 5.5 million acres have burned across the United States, while the European Union has recorded over 500,000 hectares burned, exceeding even 2025's record pace.

The U.S. wildland–urban interface (WUI), the zone where houses meet or intermingle with wildland vegetation, grew by an area the size of Washington State between 1990 and 2020 and now shelters 44 million homes, roughly a third of all U.S. housing.

As that footprint expands into forests, foothills and coastal scrub, the evacuation map has become as critical as the fire line itself and the communities that fare best will be the ones that plan it with data, drills and a clear-eyed understanding of how people really behave under pressure.

For us, working in transport planning and movement strategy, these events are more than statistics. They are human stories about families displaced, communities fractured and economies damaged. Climate change has made this reality more acute. Hotter, drier conditions extend fire seasons, while extreme winds drive fire behaviour in ways that outpace traditional response models.

Research in Nature Sustainability found that 12.54 percent of global WUI areas - home to more than 10 million people - fall within 4.8 km of wildfire threat zones. This is a global issue. In Europe, Scotland now issues “extreme” wildfire alerts, while in Spain researchers map social vulnerability in fire-prone WUI areas. In Asia-Pacific, China accounts for 96 percent of worldwide WUI expansion due to rapid urbanisation, while in Australia, major cities like Sydney, Melbourne and Brisbane are fringed by high-risk WUI zones that threaten both dense suburbs and intermix developments. Evacuation planning must reflect not only fire physics but the human realities of population growth, infrastructure limits and social vulnerability.

Traditional evacuation models were designed with buildings in mind, using simple timelines that assumed people would leave before conditions became unsafe. Applied at the community scale, this approach overlooks realities such as:

• Pre-evacuation delay: When residents hesitate or take too much time to prepare before leaving
• Traffic congestion: As thousands rush to limited escape routes
• Communication gaps: Leading to slower, or ignored, responses
• Dynamic fire spread: Shifting unpredictably with wind and topography

Plans that rely on these oversimplifications often look good on paper but fail in practice.

This is where the WUI-NITY model comes in. Developed by a global research consortium, including our teams at GHD, WUI-NITY is a GIS-based platform built on the Unity 3D game engine. It integrates fire spread, human behaviour and traffic flows to simulate community-wide evacuation performance.

Unlike static models, WUI-NITY allows planners to run scenarios with or without fire simulation to test “what if” situations, such as:

• How evacuation time changes if a road or bridge is lost
• Whether certain neighbourhoods can reach safe zones within target windows
• The effect of zoning or development decisions on future risk

As an open-source tool, WUI-NITY is designed for collaboration. Government agencies, first responders, transport planners and community leaders can all use the same evidence base to shape their evacuation plans.

Models must be grounded in real scenarios. That is why community drills are invaluable.

In Roxborough Park, Colorado, residents staged their first wildfire evacuation drill in 2019. Analysis of the data, published in Fire Technology in 2024, revealed that pre-evacuation behaviour was the dominant factor influencing outcomes. In other words, the time people took to decide and prepare mattered more than traffic or distance.

A second drill in 2024 further refined the WUI-NITY model, demonstrating how local datasets - such as land use, density and response times - must be tailored for each community. The exercise also showed the social benefits of practice: trust grew between residents and authorities.

Evacuation planning should not be viewed in isolation. It is one layer of a broader strategy that combines physical design, behavioural insight and community trust. The most effective approach brings together five priorities:

1. Adopt data-driven methods: Planners, responders and policymakers must move beyond qualitative judgement and use evidence-based tools like WUI-NITY to test scenarios, quantify risks and design evacuation strategies that stand up to real-world complexity.
2. Invest in local drills: Practice reveals blind spots, validates models and builds social capital. Drills also create community “champions” who help sustain preparedness over time.
3. Design smarter places: Urban design and zoning decisions must avoid funnelling residents into high-risk corridors. Building standards should be strengthened to reduce vulnerability to ember attack, while landscape fire management can slow fire spread.
4. Communicate clearly and consistently: Residents need to understand what elevated fire danger ratings mean and how they should respond. Transparent, plain-language communication builds the trust needed for compliance during an actual evacuation.
5. Engage communities early: Evacuation is not just about cars and roads. It is about people - their decisions, behaviours and trust in institutions. Involving residents early ensures that strategies are realistic, accepted and actionable when it matters most.

Integrated in this way, evacuation becomes more than a desperate last resort. It is a planned, tested and trusted safeguard within an effective framework.

Wildfires are escalating, and WUI communities are on the frontline. The numbers tell the story - tens of millions of homes, trillions of dollars in assets, and entire regions of the world are at risk. Traditional evacuation models cannot keep up.

The answer is to combine simulation tools like WUI-NITY with local drills, make sure fire risk is embedded into urban planning and empower communities to act before disaster strikes.

There is no doubt that data saves lives when it is translated into action and trusted locally. The challenge is immense, but so is the opportunity to protect the people of wildfire-prone communities.

Listen to Dan and Kamryn’s conversation with Enrico Ronchi, Associate Professor in Evacuation Modelling at the Department of Fire Safety Engineering at Lund University in Sweden.