Look for complete geospatial metadata in this layer's associated xml document available from the download link

* Metric Name: Annual Burn Probability  

* Tier: 1  

* Data Vintage: 08/2023. Includes disturbances through the end of 2022  

* Unit Of Measure: Probability, 0 to 1  

* Metric Definition and Relevance: Annual Burn Probability represents the likelihood of a wildfire of any intensity occurring at a given location (pixel) in a single fire season. In a complete assessment of wildfire hazard, wildfire occurrence and spread are simulated in order to characterize how temporal variability in weather and spatial variability in fuel, topography, and ignition density influence wildfire likelihood across a landscape. In such cases, the hazard assessment includes modeling of burn probability, which quantifies the likelihood that a wildfire will burn a given point (a single grid cell or pixel) during a specified period of time. Burn probability for fire management planning applications in this case is reported on an annual basis - the probability of burning during a single fire season.  

* Creation Method: Annual Burn Probability was produced by Pyrologix LLC, a wildfire threat assessment research firm, as part of a spatial wildfire hazard assessment across all land ownerships for the state of California. The ongoing work generally follows the framework outlined in Scott and Thompson (2013), with custom methods and significant improvements developed by Pyrologix. The project generally consists of three components: fuelscape calibration and updates, wildfire hazard assessment, and risk assessment. It utilizes a combination of wildfire models and custom tools, including the FSim large wildfire simulator (Finney et al., 2011). To date, this work has resulted in a wide variety of spatial data layers related to wildfire hazard and risk, including Annual Burn Probability, representing conditions prior to the 2023 fire season. Work to date has been funded by the USDA Forest Service Region 5, the California Energy Commission, and the USDI Bureau of Land Management with data contributions from CAL FIRE.

For this project, the USFS modeling system called FSim is used to quantify
annual wildfire likelihood across California. The model is parameterized using
spatial datasets of historical weather, fire occurrence, fuels, weather, and
topography in order to simulate thousands of fire-years on a landscape. Annual
Burn Probability is calculated from these simulations using a Monte Carlo
approach to make a spatially resolved estimate of the contemporary annual
likelihood of wildfire across the landscape. For more information on FSim or
the wildfire hazard modeling being performed by Pyrologix, please see Volger
et al., 2021.

  

* Credits: Pyrologix, LLC \--James Newman (California State BLM Office) jnewman@blm.gov