A new wildfire risk map developed by researchers at the University of California, Irvine identifies the environmental and structural factors that make buildings most vulnerable to wildfire damage across California.
The tool combines more than a decade of fire inspection records with weather, landscape and building data to pinpoint areas at greatest risk.
The research analysed around 100,000 inspection records collected by the California Department of Forestry and Fire Protection between 2013 and 2024.
The findings revealed clear patterns behind wildfire destruction, allowing researchers to create a 100-metre-resolution map capable of forecasting where buildings are most likely to be damaged during future fires.
Published in Science Advances, the study provides an evidence-based framework that could help policymakers, emergency planners, insurers and homeowners prioritise wildfire mitigation efforts.
The model achieved prediction accuracy of up to 88%, offering a practical approach to improving resilience as California faces increasingly severe wildfire seasons.
Study reveals predictable wildfire damage patterns
Researchers examined nearly 55,000 buildings that were either destroyed or damaged during the 12-year study period.
The analysis found that 86% of affected structures were located within California’s wildland-urban interface (WUI), where residential development borders forests, grasslands and other natural vegetation.
Rather than finding random patterns of destruction, the team identified consistent relationships between weather conditions, surrounding landscapes and building characteristics that significantly influence whether a property survives a wildfire.
The resulting wildfire risk map integrates these variables into a machine learning framework that estimates the probability of building damage across the entire state.
Weather and landscape drive wildfire intensity
The study highlights several environmental conditions that sharply increase wildfire danger.
Low dew-point temperatures, which indicate exceptionally dry air, were identified as a major contributor to destructive fires. When combined with strong winds, these conditions accelerate fire spread and increase the likelihood of buildings being damaged.
California’s diverse landscapes also play an important role. Grasslands, mixed oak woodlands and mesic chaparral create pathways that allow fires to move rapidly from natural vegetation into developed communities.
Topography further amplifies the threat. Many wildfire-prone communities are located at middle elevations of approximately 500 metres, where sloping terrain and narrow canyons channel winds that intensify fire behaviour.
The devastating Palisades and Eaton fires in Southern California during January 2025 illustrated how powerful wind events, combined with steep terrain and proximity to vegetation, can rapidly escalate wildfire impacts, destroying more than 16,000 structures.
Expansion into wildfire-prone areas increases exposure
The researchers also point to ongoing development within the WUI as a major factor behind increasing wildfire losses.
Across the United States, construction within WUI zones expanded by around one-third between 1990 and 2010, adding more than 12 million new buildings. In California, over 80% of new homes constructed between 2020 and 2022 were built in areas already considered to face high wildfire risk.
As more communities expand into fire-prone landscapes, the number of properties exposed to increasingly frequent and intense wildfires continues to rise.
Building design can significantly reduce risk
Beyond environmental conditions, the research found that building design has a substantial influence on wildfire vulnerability.
Features such as wooden fences, combustible exterior cladding, exposed eaves and standard ventilation openings can allow wind-driven embers to ignite structures even when flames remain some distance away.
The study suggests that upgrading these features with more fire-resistant alternatives could substantially lower the probability of buildings being destroyed.
The researchers found that wildfire risk increases most when combustible building materials coincide with extreme weather, dry vegetation and highly flammable landscapes.
High-resolution wildfire risk map supports prevention
To translate these findings into a practical planning tool, the team created the Wildfire Building Damage Risk Index (WBDRI), a continuous wildfire risk map covering the whole of California at a resolution of 100 metres.
The index combines satellite-derived environmental information, fire-weather data and detailed building characteristics within a machine learning model capable of estimating damage probability for individual locations.
Researchers evaluated multiple versions of the model by progressively incorporating building flammability, weather variables and environmental exposure data, ultimately achieving prediction accuracy of up to 88%.
The study suggests the map could help authorities prioritise fuel reduction programmes, defensible space inspections and infrastructure improvements in areas where predicted damage probability exceeds 0.55, where mitigation measures are expected to deliver the greatest benefit.
Supporting long-term wildfire resilience
The researchers believe the wildfire risk map has applications far beyond emergency response.
It could support land-use planning, insurance risk assessment and investment in climate-resilient infrastructure while helping local authorities target prevention programmes where they are likely to have the greatest impact.
The findings also reinforce the importance of combining stronger building standards, careful development planning and active vegetation management, particularly within WUI communities.
As wildfire seasons become longer and more severe, data-driven tools such as the WBDRI could play an increasingly important role in reducing damage and protecting California’s growing at-risk communities.
