March 10, 2026
By Brian Bastian, Head of Product
Wildfire risk is often framed as a landscape problem.
Models typically focus on fuels, vegetation, drought, weather, and topography. These factors unquestionably determine where fires start and how they spread across landscapes.
But when it comes to what actually burns, the story is more nuanced.
A recent paper published in Nature Communications examined structural losses across several major California wildfires, including the Camp Fire, Tubbs Fire, and Glass Fire. The researchers used post-fire damage assessments and machine learning models to understand which factors most strongly influence whether structures survive or are destroyed.
Their conclusion reinforces something many practitioners have long observed in the field:
Wildfire risk to buildings is not purely a landscape problem.
It is also a structure and neighborhood problem.

The Four Drivers of Structural Loss
The study identified four primary drivers of structural loss:
- Spacing between structures
- Fire exposure intensity
- Structure hardening
- Defensible space
Each of these elements contributes to whether a structure ignites. But the research highlights an important dynamic: these variables interact with each other.
For example:
- Homes located close together increase the likelihood of structure-to-structure fire spread.
- Even well-hardened buildings can be vulnerable if neighboring structures ignite.
- Defensible space helps reduce ignition risk, but its effectiveness depends on local exposure conditions.
This reinforces an emerging understanding in wildfire science: community layout and parcel-level conditions matter just as much as the surrounding wildland.
The Neighborhood Effect
One of the most striking insights from the research is the role of structure density and proximity.
Once a fire reaches developed areas, the primary driver of destruction often becomes house-to-house ignition.
This creates a cascade effect:
- Wildfire reaches the edge of a community.
- A small number of structures ignite.
- Radiant heat and ember showers ignite nearby buildings.
- Fire spreads through the built environment.
In this phase, wildfire behaves less like a vegetation fire and more like an urban conflagration problem.
This means that assessing wildfire risk purely at the regional or landscape scale misses an important piece of the puzzle.
Why Parcel-Level Risk Matters
For insurers, municipalities, and property owners, the relevant question is not simply: Is wildfire possible in this region?
Instead, the more actionable question is: What is the ignition risk for this specific structure?
That risk depends on a combination of factors, including:
- Construction materials
- Roof and vent design
- Vegetation immediately surrounding the structure
- Parcel configuration
- Distance to neighboring structures
- Local fuel patterns and historical fire activity
These factors exist at the parcel and structure level, not just the landscape level.
This is where high-resolution risk assessment becomes essential.

Moving Beyond Landscape-Only Models
Traditional wildfire hazard models are extremely valuable for understanding where fire is likely to occur.
But when the objective shifts to predicting structural loss, the modeling framework needs to evolve.
Effective structural risk assessment requires combining multiple layers of information, including:
- Hazard exposure
- Local fire history and recurrence
- Proximity to fuels
- Structure characteristics
- Parcel configuration
- Neighborhood structure density
The goal is not simply to map wildfire hazard, but to understand how that hazard interacts with the built environment.
The researchers estimate that combining structure hardening and defensible space could reduce structural losses by roughly 50% in similar wildfire scenarios.
What This Means for Mitigation
Perhaps the most encouraging result of the study is that mitigation measures matter.
The researchers estimate that combining structure hardening and defensible space could reduce structural losses by roughly 50% in similar wildfire scenarios.
This underscores an important point: Wildfire risk is not just something that must be endured. It is something that can be measured, managed, and reduced.
But doing so requires the right level of analysis.
The Path Forward
Wildfire risk management is increasingly moving toward structure-level intelligence.
Understanding wildfire behavior across landscapes will always be essential. But as communities expand into fire-prone regions, the next frontier of risk modeling lies in understanding how fire interacts with individual structures and neighborhoods.
Recent research continues to reinforce this shift.
And as data, modeling techniques, and parcel-level insights improve, the ability to identify actionable mitigation opportunities at the property level will continue to grow.
Because in the end, wildfire risk is not just about where the fire burns.
It is about what ignites when it arrives.

