In recent years, the threat of wildfires has intensified, a trend likely to continue due to climate change. Anticipating this risk is crucial for effective management. This post demonstrate how Fire Weather Index projections and satellite data can be used to assess and mitigate wildfire risks, focusing on hazardous industrial sites in France.
Fire Weather: A Growing Threat for Industry and Infrastructures
In the aftermath of the devastating Los Angeles fires of January 2025, awareness of wildfire risks has surged among homeowners, urban planners, and insurers. However, residential and commercial buildings are not the only structures vulnerable to these threats. Industrial facilities, utilities and large infrastructures also face significant risks from wildfires, with potentially catastrophic consequences.
In March 2022, a fire in South Korea was halted just meters away from one of the world's largest nuclear plants; in 2021, several coal plants in Turkey were threatened by encroaching flames; and in 2016, a wildfire swept into the industrial zone of Fos-sur-Mer in southern France, destroying warehouses and coal yards.
These incidents underscore the growing challenges posed by wildfires near hazardous industrial sites. Such events can cause massive economic losses, long-lasting disruptions and severe environmental damage. They can also limit firefighters' ability to combat the blaze, as they must prioritize protecting these critical facilities.
This string of recent incidents raise a critical question: Are these isolated, unlucky events, or are industrial sites particularly exposed to wildfire? And how will this risk evolve with climate change?
Combining FWI Projections and Satellite Data to Detect Facilities at Risk
In this case study, we aim to assess the wildfire risk for hazardous factories and industrial sites in France. We focus on the 1,247 facilities identified under the European Seveso Directive for containing large quantities of dangerous substances. Our objective is to pinpoint installations potentially exposed to wildfire risk and evaluate the severity of that risk in the near future.
As discussed in a previous post, the fire weather index (FWI) is a widely used indicator for predicting wildfire risks based on weather conditions including temperature, humidity, rainfall and wind. It is valuable for both short-term forecasting and anticipating long-term trends influenced by climate change.
However, atmospheric conditions alone do not determine fire risk for a location: the physical environment plays a crucial role. Even with highly favorable weather for fires, the risk remains limited if there is no combustible material nearby.
Therefore, our team developed and implemented a methodology to systematically identify facilities situated in wildfire-prone environments based on land cover data derived from satellite observations. This automated approach allowed us to calculate the proportion of forest or shrubland within a radius of 1 and 5 kilometers for each site and select those with bordering large wooded areas.
Using Fire Weather Index Projections to Evaluate Current and Future Wildfire Risks
After identifying 316 hazardous industrial plants located near wooded areas and potentially exposed to forest fires, our next step was to determine whether the weather conditions near these locations are conducive to wildfires.
Satellite data analysis revealed that approximately one-fourth of France's dangerous industrial plants are located in areas at risk from wildfires. But are these sites also exposed to fire weather conditions?
To answer this question, we computed the daily Fire Weather Index values from 1986 to 2100 using detailed climate projections for temperature, humidity, wind and rainfall, sourced from the CMIP5/EuroCORDEX datasets. This dataset provides high-resolution climate information for Europe allowing for precise risk assessments. Similar data are available for other parts of the world.
Our analysis focused on calculating the average number of days per year with significant fire risk, defined as days with an FWI above 30. This threshold indicates conditions where wildfires are likely to ignite and spread rapidly. We performed this analysis for the mid-century period (2036-2065) and compared the results to a historical reference period (1986-2015).
In line with scientific best practices, we performed those calculation with multiple climate models to account for uncertainties and variability in future climate projections and used the multimodel median as the most likely outcome. We also considered three different emissions scenarios—RCP2.6, RCP4.5 and RCP8.5—to explore a range of possible future climate pathways, from optimistic to pessimistic outcomes.
The Changing Landscape of Wildfire Risks for Hazardous Industrial Sites
The findings from this analysis reveal a stark shift in wildfire risks in mainland France. More than 300 Seveso-classified industrial installations are located near forests, representing about a quarter of French hazardous industrial sites. In the climate of the 2000s, only one-third of these installations faced fire weather risks conditions for more than 10 days per year on average. However, this proportion is expected to rise significantly with climate change:
By 2050, under a median emissions scenario (RCP4.5), three-quarters of the hazardous industrial facilities located in wooded areas will be exposed to fire weather for more than 10 days per year on average.
The average annual period of risk is projected to increase by 50% between 2000 and 2050. Under the high emissions scenario (RCP8.5), this upward trend is projected to continue into the latter half of the century.
Geographically, while the southeast of France remains the region most exposed to wildfire, significant risk increases are expected in central and southwestern area. Risks are also projected to appear across the northern half of the country, including around Paris, where wildfires have historically been negligible. This shift raises the possibility of fires occurring near facilities that are unprepared for such events, as they have never faced this threat in the past.
To mitigate the risk of industrial accidents, plant operators, emergency services and the public authorities must anticipate these changes, particularly by implementing preventive measures, updating emergency plans and scaling response capabilities. Fire weather projections are invaluable in this context, providing the foresight needed to identify emerging risk before they materialize.
More broadly, future climate risks—such as wildfires, heatwaves, water stress, flooding or coastal submersion—can be anticipated through the use of climate projections and other relevant data. This approach allows for the integration of accurate risk assessments into the planning and investment strategies for industrial sites, critical infrastructure and other long-term projects.
By leveraging these tools, industrial operators, financial institutions, insurance providers or certification bodies can better anticipate the impacts of climate-related hazards and ensure resilience and sustainability in the face of climate change.
If you need to anticipate and manage future climate risks, Callendar can help. Contact us to explore how our expertise and data-driven solutions can support your planning and investment decisions.