Explore Resources
Home Articles Expert Commentary Wildfire Modeler Perspectives on Insurance
Catastrophe Risk Management

Wildfire Modeler Perspectives on Insurance

Scott Ritter | October 30, 2026

On This Page
flames in retail buildings near a forest wildfire as viewed from a window across the street

Wind-driven urban conflagrations, million-acre megafires, overwhelmed fire suppression resources, and unprecedented insured losses—these are the headlines of the last decade of wildfire activity across the globe. History shows the events of the present (and the probable future) resemble the past, only with increasing frequency and severity. The impacts of damaging wildfires have been felt in the familiar territory of North America and Australia, but also in Mediterranean Europe, South Korea, and Chile.

Though the events and losses have been unprecedented in their scale, the types of fires and the physical mechanisms driving them are understood, and historical analogues for these events exist. Major loss events share a common thread: high winds and extreme rates of spread leading to the transition into an urban conflagration—be they driven by downslope wind events like the Diablo winds that drove the 2017 Tubbs Fire (5,636 structures lost) or the Santa Ana winds that drove fires like the 2018 Woolsey Fire (1,643 structures lost), 2025 Eaton Fire (9,418 structures lost), or 2025 Palisades Fire (6,837 structures lost), or fire-induced winds like the 2018 Carr Fire (1,614 structures lost), whose convective plume formed a pyrocumulonimbus strong enough to generate a vortex that reached an altitude of 17,000 feet and speeds in excess of 140 mph, which is on par with an EF-3 tornado.

As extreme and damaging as these events were, fast-spreading fires that lead to urban conflagrations are not new phenomena—one can look back to 1923 in Berkeley, California, where an urban conflagration began with a wind-driven fire in the dry grass and chaparral-covered hills east of town. The fire overwhelmed the limited resources of the Berkeley Fire Department and destroyed 584 homes before the dry northeasterly winds were replaced by a humid afternoon onshore breeze, stopping the conflagration.

Also look to 1967 in Hobart, Tasmania, Australia, where multiple fires swept into the city and surrounding communities during extreme fire weather. In just 5 hours, the fires killed 62 people and destroyed approximately 1,400 buildings. The 2003 Canberra, Australia, fires provide a more recent analog when fires west of the city accelerated under extreme weather, overwhelmed suppression resources, and entered southwestern suburbs. The firestorm killed 4 people, destroyed 487 homes, and produced a pyrocumulonimbus linked to a tornado. Canberra, therefore, provides a direct precedent for both the urban spread and fire-generated vortex observed during the Carr Fire 15 years later.

The escalation of extreme wildfire loss is clear not only in North America but around the globe. Together, the 2017 Portugal fires, 2023 Greek fires, 2024 Chilean fires, and South Korean 2025 fires claimed more than 300 lives. The Chilean fires alone destroyed at least 14,000 homes. The human disruption continued in 2026, when fires in France and Spain forced more than 300,000 people to evacuate. These events demonstrate how extreme fire behavior and exposed communities can produce catastrophic losses across very different climates, vegetation types, and building stocks.

Identifying Wildfire Causes and Frequency

Although causal attribution for any single event is often difficult, a confluence of factors has increased the probability of an extreme-loss event. These include continued development in fire-prone ecosystems; hotter, drier, and longer fire seasons; and, in some fire-adapted forests, the accumulation of vegetation as a result of successful fire exclusion. The importance of each factor varies among regions and individual fires, but extreme wind is the common proximate driver that can transform a wildland fire into an urban disaster.

On the frequency side of the equation, the evidence points toward more extreme fire-spread events under warmer, drier conditions. In 2022, I contributed to research that found that these extreme spread events accounted for 70 percent of the total area burned across the Western US over the past 2 decades. We also found that the annual number of extreme events was strongly associated with fire-season aridity, including vapor-pressure deficit, and our models projected that their frequency would more than double under 2°C of warming. 1

Other subsequent research connected rapid-fire spread directly to destructive outcomes across the continental US, finding that the fastest-growing fires accounted for 78 percent of the structures destroyed from 2001 through 2020. 2 Over the same period, maximum daily fire growth increased by 249 percent across the Western US and by 398 percent in California. Together, these studies show that the frequency of exceptionally fast-growing fires is increasing, and these are the very fires that shape the potential for catastrophic loss.

The severity of destructive fires has also increased, and research from 2025 shows that a larger share of the buildings exposed to fire are being lost. 3 This research found that between 2002 and 2022 not only has the number of buildings destroyed each year increased nearly sevenfold, from an average of 905 buildings per year in 2002–2012 to 6,239 per year in 2013–2022, but also the proportion of exposed buildings that were destroyed increased from 10 percent to 32 percent. This reveals that, in more recent years, structure survival rates are lower and fires are converting exposure into loss with increasing efficiency. When combined with increased development in fire-prone areas, escalating losses are nothing but inevitable.

Lessons Learned

The biggest lessons that fire scientists have learned from the past decade are that model underprediction has real consequences as the probability of fast-spreading tail events increases and that loss cannot be inferred from acres burned or final fire size. Catastrophic losses are concentrated in wind-driven, fast-spreading events that overwhelm suppression resources and penetrate developed areas.

Once structures ignite, the event can become self-propagating: Burning homes generate embers and radiant and convective heat that ignite neighboring buildings. Wildland vegetation may initiate the transition of fire into the community, but after that transition, the community itself becomes the driving fuel.

What Insurers Can Do

For an insurance industry under stress, there are a number of pathways to build sustainable books of business and steer the world toward a more resilient future in the face of high temperatures, drier conditions, and increased potential for fast-spreading wildfires.

  • Model the fires that create loss. Wildfire models need to explicitly represent the rapid-growth events that most stress suppression resources and lead to the transition from wildland fire to urban conflagration. A model calibrated based on annual frequency or final acreage can reproduce the overall average of historical fire activity while missing the extreme events that truly drive portfolio loss.
  • Underwrite the likely ignition pathway. Standards should respond to spatial context. Ember attack, direct flame contact, and structure-to-structure spread create different demands on a property. Rather than assessing each property based on a fixed list of requirements, underwriting should consider the expected exposure pathway when determining eligibility.
  • Price the exposure environment. Pricing should reflect the reality of the exposure environment and not overly index on features like siding type or parcel-level vegetation. Parcel- and structure-level characteristics matter, but they should not be allowed to obscure structure spacing, development pattern, landscape position, and the potential for correlated loss.
  • Use eligibility to influence new development. Insurers can encourage responsible development through eligibility rules that consider community design as well as building design. This includes the community's physical location and design factors like structure separation, road and evacuation capacity, reliable water, and maintained green space.
  • Treat wildfire loss as a climate and development problem. Extreme wildfires are above all else a weather phenomenon, and insured loss occurs where fire weather meets human geography. Individual homeowner action remains necessary, but it cannot change regional aridity, the frequency of high wind events, subdivision layout, or emergency-system capacity.
  • Recognize the boundaries of mitigation. Some risks cannot be mitigated enough to support a realistic price because those prices may not be affordable for many property owners. A functioning market needs validated risk signals, targeted public policy, and a clear definition of which residual risks belong in private insurance and which are being transferred to public backstops.

Better wildfire science increases insurability by supporting underwriting and pricing at the resolution wildfire demands. Rather than completely pulling out of geographies, insurers can offer affordable policies to properties in the areas where modeled risk supports them and identify the highest-risk properties that have been underpriced or must be transferred to public backstops. The trends of the last decade are likely to continue, and without accurate modeling and pricing, the property insurance industry will face escalating wildfire losses.

Opinions expressed in Expert Commentary articles are those of the author and are not necessarily held by the author's employer or IRMI. Expert Commentary articles and other IRMI Online content do not purport to provide legal, accounting, or other professional advice or opinion. If such advice is needed, consult with your attorney, accountant, or other qualified adviser.

Footnotes

1 Jonathan D. Coop, et al., "Extreme Fire Spread Events and Area Burned Under Recent and Future Climate in the Western USA," Global Ecology and Biogeography, April 6, 2022.
2 Jennifer K. Balch, et al., "The Fastest-Growing and Most Destructive Fires in the US (2001 to 2020),"Science, October 24, 2024.
3 Amanda R. Carlson, et al., "Rising Rates of Wildfire Building Destruction in the Conterminous United States," Proceedings of the National Academy of Sciences, December 15, 2025.