Analyzing Patterns of Home Destruction in California Wildfires

Insights from Elevation, Building Materials, and Temporal Analysis

Author

Tony Fraser

Published

April 6, 2025

1 The Climate-Fire Crisis

California’s wildfire landscape is undergoing dramatic transformation, with eight of the ten largest wildfires in state history occurring between 2017-2022, destroying over 40,000 structures and causing more than $40 billion in property damages. Climate change has fundamentally altered the fire risk landscape through three critical mechanisms:

Intensified wind patterns (particularly Santa Ana and Diablo winds)
Elevated temperatures leading to drier vegetation
Reduced rainfall extending fire seasons

As wind patterns break historical records and fire behavior becomes increasingly unpredictable, traditional risk assessment methods based on historical data no longer provide adequate predictive capabilities.

The crisis is compounded by the insurance industry’s retreat from high-risk areas in California, with many companies limiting coverage or exiting markets entirely despite rising premiums. Homeowners in wildfire-prone regions face a growing challenge: maintaining adequate insurance coverage while understanding how to effectively reduce their vulnerability.

2 Literature Review

Spatial Factors

Structural Factors

Environmental Factors

Economic Factors

Spatial

Housing arrangement and location determine structure vulnerability more significantly than previously recognized. Homes are most vulnerable when "located in small, isolated housing clusters, built at low to intermediate housing density, situated near development edges."

Syphard et al., 2012

Spatial

Variables describing the wider landscape—vegetation connectivity, topography, and building arrangement—proved more predictive of structure loss than specific vegetation types, challenging conventional fuel-focused risk assessment.

Alexandre et al., 2016

Environmental

Significant increases in surface easterly downslope winds have been detected on western slopes of the Sierra Nevada Mountains, especially during autumn. The 95th percentile wind speeds increased approximately 3.1% over a 41-year period.

Thompson et al., 2019

Environmental

Hazardous Wind Events (strong, dry northeasterly winds lasting at least 6 hours) have become more frequent since the early 2000s in California, particularly in the Sierra Nevada region and Southern California.

Thompson et al., 2019

Structural

Window characteristics emerged as surprisingly important in structure survival—more so than roofing or exterior wall materials. Structures with dual-pane windows and vinyl frames showed significantly higher survival rates during wildfire exposure.

Syphard, Brennan, & Keeley, 2017

Structural

Building age proved to be a significant factor in structure survival, with newer construction (post-1997) generally demonstrating better fire resistance. This corresponds with the implementation of improved building codes in many California counties.

Syphard, Brennan, & Keeley, 2017

Spatial

Creating defensible space adjacent to homes was found to be equally important as building construction, with the most critical aspects being vegetation immediately adjacent to, touching, or overhanging the structure.

Syphard, Brennan, & Keeley, 2014

Economic

Insurance market dynamics show asymmetric risk classification, where insurers with better risk information can identify and offer lower prices to lower-risk properties, while less-informed firms face adverse selection and often exit high-risk markets.

Boomhower et al., 2024

3 The Elevation Paradox

Probability of Structure Destruction by Elevation Range

Plot

Structures at 500-1,000 feet elevation face a 68% probability of destruction when exposed to wildfire—significantly higher than those at elevations above 3,000 feet (47-52%).

Key Finding: Contrary to conventional assumptions that higher elevations with steeper slopes face greater fire danger, our analysis identified elevation as the single strongest predictor of structure vulnerability (10.8% importance in Random Forest modeling).

This counterintuitive finding reveals that topographic position relative to wind corridors is more important than absolute elevation in determining fire risk. The elevation effect results from multiple converging factors:

Lower elevations often coincide with wind acceleration zones during Santa Ana and Diablo events
Development at canyon bottoms creates natural wind tunnels that channel fire behavior
Vegetation patterns at different elevations support varying fire spread rates

Regional variations exist across California. Southern California coastal regions show a U-shaped relationship with elevation, where mid-elevations demonstrate higher survival rates than lower elevations or mountain areas, aligning with Santa Ana wind corridor dynamics. In contrast, Sierra Nevada foothills show decreasing destruction probability as elevation increases from 1,000 to 4,000 feet.

This finding has significant implications for development planning, as it suggests that particular caution should be applied to new construction in the 500-1,000 foot elevation band.

4 Building Materials & Construction Era

Structure Destruction Rates by Building Characteristics

Plot

Building materials and construction era significantly impact wildfire survival. Dual-pane windows provide a 22% reduction in destruction probability compared to single-pane windows.

Key Finding: Building materials and design features significantly influence structure survival during wildfires, with window characteristics and construction era emerging as surprisingly important factors.

Our analysis revealed several critical insights about building vulnerability:

Window Impact: Dual-pane windows are associated with a 22% reduction in destruction probability compared to single-pane windows. This effect amplifies during high-wind events, where dual-pane windows provide a 27% reduction in destruction probability when wind speeds exceed 15 m/s.
Siding Materials: Exterior siding materials show a strong relationship with destruction probability. Structures with wood siding show a 71% destruction rate when exposed to fire, compared to 49% for stucco and brick/stone.
Construction Era Effect: Structures built after 1997 show destruction rates 15-18% lower than pre-1997 buildings. This aligns with the implementation of more stringent wildfire-related building codes beginning in 1997 in many California counties.
Material-Elevation Interaction: The protective effect of fire-resistant materials increases with elevation. At elevations above 3,000 feet, structures with fire-resistant siding show a 32% lower destruction rate compared to combustible siding, whereas this difference is only 18% at elevations below 1,000 feet.

These findings provide clear guidance for prioritizing retrofit investments, with window upgrades and exterior siding improvements emerging as particularly cost-effective interventions based on their protective value.

5 Beyond Traditional Fire Seasons

Structures Destroyed by Calendar Month

Plot

While fall remains the peak destruction period, the unexpected spike in January challenges traditional assumptions about "fire season" and highlights increasing year-round vulnerability.

Key Finding: Traditional fire season definitions are becoming increasingly inadequate as climate change alters historical weather patterns, with winter months showing unexpectedly high destruction rates when fires occur.

Our temporal analysis revealed several important trends that challenge conventional wisdom about wildfire timing:

Peak Destruction Period: Fall (September-November) remains the period of highest vulnerability, accounting for 61% of all destroyed structures. This reflects the convergence of dry vegetation following summer drought, onset of Santa Ana and Diablo wind patterns, and delayed rainy season in recent years due to climate shifts.
Winter Anomalies: Winter (December-February) shows the second-highest destruction rate when structures are exposed to fire—a counter-intuitive finding that challenges assumptions about winter as a “safe” season. This appears driven by the catastrophic Eaton and Palisades fires in January 2025, which together destroyed over 18,000 structures under unusually dry winter conditions.
Increasing Year-Round Risk: Annual trends show an overall increase in both the number of structures affected by wildfire and the destruction probability when structures are exposed to fire. The years 2017-2018 and 2024-2025 stand out as particularly destructive periods, with destruction probabilities exceeding 65% for exposed structures, compared to an average of 53% during other years.

These patterns suggest a fundamental shift in California’s fire risk landscape, requiring year-round preparedness and challenging the notion of a limited “fire season” for planning and resource allocation.

6 Stakeholder Applications

For Homeowners

Prioritize window upgrades to dual-pane systems, especially in wind-exposed locations
Consider exterior siding replacement from wood to non-combustible materials, particularly at higher elevations
Recognize that elevation impacts risk profiles and adapt mitigation strategies accordingly
Maintain year-round preparedness, not just during traditional fire seasons

For Insurers

Develop more nuanced risk assessment models that incorporate elevation, building materials, and their interactions
Consider recalibrating policy renewal cycles to account for changing temporal risk patterns
Design incentive programs that reward resilient construction through premium discounts
Use the quantified risk reduction associated with specific building features to actuarially calibrate incentives

For Community Planners

Implement more restrictive development policies in high-vulnerability elevation zones (500-1,000 ft)
Maintain and strengthen wildfire-related building codes, given their demonstrated effectiveness
Prioritize community-scale interventions like improved access roads and strategic fuel breaks for older developments
Revise infrastructure design standards to account for climate change projections

For Fire Management Agencies

Consider pre-positioning resources near development in the 500-1,000 foot elevation band during high-risk weather conditions
Target prevention education efforts to specific combinations of building types and locations
Plan for continued increases in both fire frequency and intensity, with corresponding infrastructure investments
Develop regionally tailored response strategies that reflect geographic variation in risk patterns

7 Conclusion & Future Directions

This research provides empirical evidence that home destruction patterns during California wildfires follow distinct spatial, temporal, and structural patterns that challenge conventional wisdom about wildfire vulnerability.

Contrary to traditional assumptions that higher elevations with steeper slopes face greater fire danger, elevation emerged as the strongest predictor of structure vulnerability, with buildings between 500-1,000 feet showing the highest destruction rates (68%).

Equally unexpected was the critical importance of window characteristics (22% reduction in destruction probability with dual-pane windows) and the strong protective effect of post-1997 construction (15-18% lower destruction rates), highlighting how building codes have measurably improved structure survival.

The evolving temporal patterns identified—particularly the high destruction rates during winter months and increasing year-round fire risk—underscore the need for adaptive management approaches that respond to changing climate conditions rather than relying on traditional fire season definitions.

By integrating these insights into risk assessment models, building codes, community planning, and fire management strategies, we can work toward more resilient communities despite the increasing challenges posed by climate-driven wildfire activity.

7.1 Future Research

Fine-scale Wind Modeling: Computational fluid dynamics models of wind flow around structures could reveal complex wind-structure interactions and better explain the relationship between topography and fire behavior
Community-Scale Integration: Expanding analysis to include community-scale variables like road networks and housing density patterns
Climate Projection Integration: Combining empirical findings with climate model projections to generate forward-looking risk forecasts