Protect communities with data-driven wildfire modelling and evacuation

How prepared is your community for a wildfire? If a wildfire cannot be contained, are you equipped to evacuate residents within a short period of time? GHD’s Kamryn Kubose, Transport Planner, and Dan Stockdale, Senior Advisor, Movement Strategies, unpack the growing threat of wildfires and the importance of data-driven planning and evacuation in at-risk communities. They also discuss WUI-NITY, a new modelling tool that simulates evacuation scenarios to inform decision-making in the development of emergency plans and procedures.

Destructive wildfire events are leading to more burned acreage, damaged properties, displaced residents, fatalities and massive economic losses. Record-high temperatures and prolonged dry spells due to climate change have made environments more susceptible to burning and blazes harder to extinguish. Variable and extreme winds also play a significant factor in extreme fire behaviour events.

Around the world, there are many vulnerable communities that are located in wildland-urban interfaces (WUI), the zone where wildland and human led development meet. The increased frequency and intensity of wildfires, combined with human settlement expansion, make WUI zones more vulnerable. 

Research published in Nature Sustainability that analyses 2001-2020 wildfire burned area maps against population datasets reveals that “globally, 12.54 percent of WUI areas housing 10.11 million people are within a 4,800 m buffer zone of wildfire threat.” 

• The US: The US Fire Administration reports that the WUI area continues to grow by approximately 2 million acres per year due to new housing developments, and more than 60,000 communities are at risk of WUI fires. To illustrate the magnitude of risk, over 46 million homes, valued at approximately USD 1.3 trillion, are situated within the WUI and could impact 99 million people, approaching one-third of the US population. In North America, the National Interagency Fire Center’s forecasts for Northern California indicate a shift from cooler, wetter weather in March to warmer, drier conditions by early summer, potentially increasing wildfire risk as an increased amount of vegetation dries out.
• Europe: While specific forecasts for Europe’s summer wildfire activity remain limited, ‘extreme’ wildfire risk alerts in Scotland emphasise the need for preparedness as climate change drives warmer, drier conditions. In Spain, studies have analysed the spatial patterns of social vulnerability in relation to wildfire risk in WUI areas, underscoring the need for targeted interventions.
• Asia-Pacific: The growth of WUI areas is particularly notable in China, accounting for nearly 96 percent of the total increase worldwide. The country’s rapid urbanisation has led to densely populated areas and more communities being situated in high-risk wildfire zones. In Australia, WUI has been adapted from the US and European approaches to provide a nationally consistent definition. Interface WUI are high density developments adjacent to wildland areas. Intermix WUI are lower density housing intermingled with wildland vegetation. Major city outskirts in New South Wales, Victoria, Queensland and South Australia pose risks.

With increasing threats, it’s critical to continue advancing wildfire management and evacuation research to develop solutions for at-risk communities. Often, it’s a community-derived grassroots interest in preparedness, usually because they have been affected by fires in the past and they recognise the opportunity to develop a plan.

In the event of a wildfire, the evacuation of a community is a complex set of activities, and it varies by location and threat severity. Consider the example of a simplified community evacuation timeline: a community evacuation timeline may start with messaging about preparedness and the daily fire danger rating, as is the case in Australia. If ignition does occur, incident detection follows, then firefighter assessment, and hopefully containment. However, if the fire cannot be sufficiently suppressed, emergency management and the jurisdiction's communication personnel must make a series of decisions, such as whether an evacuation warning or mandatory order should be sent out and to which communities, and how emergency management should focus their efforts in any given area - firefighting or live saving.

Community officials and planners are increasingly using simulation models to inform evacuation plans for WUI communities. 

The challenge in quantifying community movement and behaviour in relation to wildfire conditions is that historically, a simple timeline has been used to characterise the time for people to get to safety similar to the approach taken to assess a building evacuation scenario and the fire conditions in the building at the point when it becomes untenable. While this enables a performance-based assessment — to determine whether the building facilitates safety under the scenario conditions examined, the model is too simplistic to provide guidance for the evacuation of a WUI area. 

Another challenge is that plans are also typically focused only on fire movement and spread, which do not provide information on community vulnerability and the effectiveness of different measures. Traditional models do not represent the evolving conditions and the likely variability of evacuees’ responses or behaviours, making it hard to assess overall community evacuation performance. 

To add to the complexity, traditional evacuation tools do not interact with fire modelling tools, making it difficult to interpret outputs. Altogether, these challenges make it hard to derive insights to inform emergency response and evacuation plans at the community scale. 

GHD is part of a global research consortium developing the WUI-NITY model, a GIS-based tool built on the Unity 3D game engine, which simulates and visualises wildfire spread and human and traffic behaviour during evacuation of communities in WUI area. We collaborate with academic and government partners to develop this platform. 

Roxborough Park, a community in a WUI area in Colorado, organised their first community wildfire evacuation drill in 2019. This resulted in a paper published in 2024 in the Fire Technology Journal, titled Roxborough Park Community Wildfire Evacuation Drill: Data Collection and Model Benchmarking. Results showed that pre-evacuation behaviour and time is the dominant factor in rural wildfire evacuation scenarios in a community with a small number of households. 

At the time, the researchers found that both the model and data impact each other, and it is not enough to use the same dataset to configure different models; for example, land use, population density, and response times vary for each specific locality. Another evacuation drill was conducted in June 2024 to calibrate the WUI-NITY model. The outcomes of this drill highlighted the value of a data-driven and evidence-based approach to planning and modelling. Additionally, understanding that comes from this data and modelling may help to increase understanding, influence and benefit city planning/zoning, design considerations and transport authority network planning. 

WUI-NITY quantifies community evacuation performance and provides data for projections, which ultimately support the development of robust plans to save lives. It has a built-in wildfire model, but its evacuation model can can be run without simulating wildfire spread to conduct rapid assessment of plans, quantify evacuation time and vulnerability of different communities, and demonstrate the impact of community design or plan changes. 

The WUI-NITY model can be run without the fire behaviour by assuming a given impact of the fire e.g. the loss of a specific route, or availability of each refuge location, among others. It also calculates evacuee travel times, routes and destinations. Users can set other parameters such as evacuation goals, fire characteristics and smoke spread. WUI-NITY is open source because the intent is to equip communities with the tools to help them develop robust evacuation plans, whilst involving the relevant stakeholders, including first responders and residents. 

A community’s ability to respond to and prepare for the increasing threat of wildfire events is heavily dependent on the information and resources available. Consider these recommendations to safeguard your community:

• To develop effective wildfire evacuation strategies, officials must supplement traditional methods and adopt data driven methodologies; WUI-NITY helps community planners, emergency responders and policymakers to design and test comprehensive evacuation strategies. 
• WUI communities need more and better data to understand their vulnerability, test “what if” scenarios, aid in real-time emergency management and mitigate the negative consequences of wildfires. It is important to note that evacuation planning is just one element in a suite of wildfire protection measures that need to be in place. This includes planning and development, with consideration of fire risks, construction to identify attack levels, landscape fire management to create a mosaic of fuels across the area to modify fire behaviour and communication for preparedness on elevated fire danger days.
• Community-wide evacuation drills are key to assessing scenarios, evolving conditions and residents’ responses. They form part of a robust emergency plan to improve preparedness and minimise the impact of wildfires on homes, businesses and the environment. Having a community champion or group of champions on the ground is also key to reinforcing the importance of evacuation drills.

Wildfires are not only a problem for fire departments but for entire communities as they adversely affect human health and safety as well as critical infrastructure. It’s time local governments and communities adopt a data driven approach in response to the increasing threat of wildfire and to underpin emergency plans.