Fire Risk Methods Overview¶

This page explains the high-level methodology used to compute building-level wildfire risk scores for CONUS. It provides a conceptual overview with references to more detailed documentation pages.

Summary¶

OCR's fire risk model computes building-level wildfire risk scores by:

Taking baseline burn probability data from USFS and spreading it into developed areas using wind-informed blurring
Multiplying the wind-adjusted burn probability by USFS conditional risk values (cRPS)
Sampling the resulting risk surface at building locations

The model produces risk scores (RPS: Risk to Potential Structures) representing the expected net value change for a hypothetical structure at each location. The scores account for directional fire spread patterns driven by prevailing winds.

Conceptual Framework¶

Risk to Potential Structures (RPS)¶

The model calculates Risk to Potential Structures (RPS), defined as:

\[ \text{RPS} = \text{BP} \times \text{cRPS} \]

Where:

BP (Burn Probability): Annual likelihood that a given pixel burns, derived from USFS wildfire simulations
cRPS (Conditional Risk to Potential Structures): The conditional net value change for a hypothetical structure if it were to burn, from USFS data

RPS represents the expected net value change per year for a generic structure at each location. It combines both probability (how likely fire is) and consequence (how much damage would occur).

Key limitation

This approach models risk to a hypothetical "potential structure" rather than actual buildings with specific characteristics. Building-level attributes (materials, retrofits, defensible space) are not included.

Wind-Adjusted Fire Spread¶

A key innovation is incorporating wind-driven fire spread patterns into the burn probability:

Fire weather analysis: Identify wind directions during high fire weather conditions (99^th percentile FFWI) from CONUS404 data
Elliptical spread kernels: Apply oval-shaped blurring filters (inspired by Richards 1990) pointing in eight cardinal/ordinal directions
Upwind spreading: For each pixel, calculate zonal mean BP from upwind areas weighted by fire-weather wind direction frequencies
Iterate spreading: Repeat blurring 3 times to spread BP up to ~1.5 km into non-burnable (developed) areas

This differs from the uniform circular blurring in USFS methods. The wind-informed approach better represents how embers transport fire downwind from wildland into developed areas.

Input Data¶

Raster Datasets¶

USFS Burn Probability (BP): 270m resolution, present-day (circa 2011) and future (circa 2047) from Riley et al. (2025)
- Only available for wildland areas; non-burnable areas have BP = 0
USFS Conditional Risk to Potential Structures (cRPS): 30m resolution, present-day (circa 2023) from Scott et al. (2024)
- Represents conditional net value change if a hypothetical structure at that pixel were to burn
- Available for all of CONUS
CONUS404 Meteorological Data: 4km gridded hourly data (1979-2022) from Rasmussen et al. (2023)
- U and V wind components, temperature, relative humidity
- Used to calculate Fosberg Fire Weather Index (FFWI)

Vector Datasets¶

Building Footprints: Overture Maps
- Coverage: CONUS-wide
- Attributes: Building geometries, centroid coordinates for risk sampling

See Data Sources and Provenance for detailed information on data access, versions, and preprocessing.

Processing Workflow¶

The calculation follows these steps:

1. Fire Weather Wind Analysis¶

Calculate Fosberg Fire Weather Index (FFWI) for every hour 1979-2022
For each 4km pixel, identify 99^th percentile FFWI threshold
Extract wind directions for all hours exceeding that threshold ("fire-weather winds")
Bin fire-weather winds into 8 cardinal/ordinal directions
Create distribution of fire-weather wind directions for each pixel

2. Upscale and Prepare BP¶

Convert 270m USFS BP raster to 30m resolution
Identify "non-burnable" pixels (where BP = 0 in Riley et al. data)

3. Wind-Informed BP Spreading¶

For each 30m pixel:

Extract nearest-neighbor 4km fire-weather wind distribution
Create 8 oval-shaped blurring filters (elliptical wavelets) pointing in 8 directions
- Each filter represents wind coming FROM that direction (spreading BP downwind TO the pixel)
- Distance from pixel to far side of oval along major axis: 510m
For each direction, apply upwind filter to calculate zonal mean BP
Weight the 8 smeared BP values by fire-weather wind direction frequencies
Repeat this process 3 times → maximum spread of ~1.5 km into non-burnable areas

4. Calculate RPS¶

Multiply wind-adjusted BP by 30m cRPS raster
Result: 30m RPS (Risk to Potential Structures) for present and future
RPS = expected net value change per year for a hypothetical structure

5. Sample at Building Locations¶

Intersect 30m RPS raster with Overture Maps building footprints
Assign risk score to each structure based on value at building centroid

6. Convert to Categorical Scores¶

Convert continuous RPS values to categorical risk scores (1-10 scale)
Scores based on percentile bins of RPS across full CONUS domain

Spatial Processing Architecture¶

The model uses a spatial chunking system for efficient parallel processing:

CONUS is divided into 595 processing regions (30m resolution chunks)
Each region is processed independently using distributed compute (Coiled/Dask)
Outputs are stored in Icechunk (for rasters) and GeoParquet (for vectors)
Failed regions can be reprocessed without affecting completed work

See Horizontal Scaling via Spatial Chunking for details on the parallelization strategy.

Outputs¶

Building-Level Outputs¶

GeoParquet files containing:
- Building geometries and IDs
- Wind-adjusted fire risk scores
- Baseline USFS risk metrics for comparison
- Metadata (region ID, processing timestamp)

Regional Aggregations¶

PMTiles for interactive web visualization:
- 1, 15, and 30-year time horizons
- County and census tract aggregations
- Pre-rendered at multiple zoom levels
Summary statistics (GeoParquet, GeoJSON, CSV):
- Buildings at risk counts by jurisdiction
- Percentile distributions of risk scores
- Comparison metrics (wind-adjusted vs baseline)

Key Assumptions and Limitations¶

"Generic" structure assumption: Risk scores represent a hypothetical structure, not actual building characteristics
Static building inventory: Does not account for new construction or demolition post-data-collection
Historical wind climatology: Uses past wind patterns (1979-2022) as proxy for fire weather
No explicit fire suppression: Model does not account for firefighting efforts
Wildfire focus only: Does not include structure-to-structure fire spread in WUI
99^th percentile FFWI: Fire weather threshold may not perfectly capture conditions during largest fires
Point-specific wind data: Wind directions at pixel B determine spreading, regardless of wind at upwind pixel A

Factors Not Included¶

The risk scores described above represent risk to a "potential" or "generic" structure. They do NOT account for factors that drive actual risk at specific buildings up or down:

Factor	Risk Impact	Notes
Building retrofit (fire-resistant materials, ember-resistant vents)	Lower	Could significantly reduce actual risk
Community emergency response capabilities	Lower	Firefighting effectiveness not modeled
Previous fire / fuel reduction	Lower	Changes to burnable landscape not captured
Access limitations (road conditions, remote locations)	Higher	Evacuation and response challenges
Building-specific characteristics (materials, defensible space, vegetation management)	Variable	Generic structure assumption means individual building attributes ignored
Development impacts on BP/cRPS	Variable	Risk scores for undeveloped land assume sole structure; large developments alter fire conditions

Interpretation guidance: The risk score at a given address should be understood as risk to a generic building at that location, leaving it to users to assess how their actual building compares to that generic baseline.