Most wildfire risk services in the market today are flawed, only evaluating the area immediately surrounding the home. However, history demonstrates even well landscaped properties can be at significant risk from adjacent high-risk areas.
RedZone’s cutting edge fire science provides a superior model. RZRisk not only evaluates the individual parcel, it also assesses the “big picture” in a single easy-to-use score. The RZRisk comprehensive approach provides a thorough analysis including the proximity to areas of high wildfire threat, susceptibility to dangerous ember showers, and the frequency of catastrophic wildfire events.
Hazard Score Methodology
RedZone RZRisk service makes use of state-of-the-art software modeling to create a nationwide map of areas most at-risk due to wildfire.
Fire Behavior Modeling
In the development of the Fire Behavior module, we use the same industry standard wildfire software/techniques as land management agencies, fire management teams, and fire scientists. LANDFIRE (http://landfire.gov/) forms the basis of the ecological and terrain related datasets. LANDFIRE is the best source for these datasets since it is national in scope, updated frequently, and sourced using a consistent methodology across all geographic regions in the United States. LANDFIRE provides the slope, aspect, elevation, fuel type (using 40 Scott and Burgan Fire Behavior Fuel Model), and numerous characteristics of the fuel composition and arrangement (height, base height, density and crown bulk). The LANDFIRE layers are fed into the Fire Behavior Module to calculate the intermediate scores for fireline intensity, rates of spread, crown fire potential, and heat output. These intermediate scores are then combined to calculate the Fire Severity Score.
One unique component of RZRisk is the use of firesheds – a minimum mapping unit comprised of an area that will exhibit common fire behavior. For example: a slope, fuel type, and density, as shown in the left hand image below, has been calculated as a fireshed since fire spread and intensity will behave in a fairly consistent manner through this area. Each fireshed appropriately accounts for the influence of the surrounding area, as shown in the right hand image below.
Traditional fire behavior models have been designed with a primary purpose of resource management – not insurance. This is especially important in the Wildland Urban Interface (WUI) where the presence of homes creates a unique ignition source and changes the dynamic of how wildfires burn. Additionally, homes at risk to losses due to ember or structure conflagration would not typically be captured in a traditional fire behavior model.
RZRisk uses three separate models in its wildfire risk calculations to account for the diverse characteristics of different geographic areas based on many factors including ecosystem, vegetation type, fire behavior, and urbanization. RZRisk uses this unique combination of models to divide the landscape based on the relative amounts of built environment (structures, roads, and other infrastructure) versus wildland fuels. The rationale for this distinction is that wildland fires behave differently when burning in pure wildland fuel beds than when burning through fuel interrupted by structures and roads. Similarly, suppression of wildland fires is conducted differently, and with varying degrees of success, when in remote areas compared with densely populated areas. These differences are captured in RZRisk by categorizing the landscape into three separate wildfire risk types, each of which is modeled with its own individual set of inputs and associated methodology. The three risk types are divided into the following modules of the model: Wildland, Intermix, and Interface.
Wildland: Characterized as traditional vegetation, forested areas, and areas with minimal structure density. The Wildland model can be thought of as areas represented by relatively continuous fuel with limited presence of structures, roads, and other human-caused disturbances. Relatively few people live in these areas, which limits one type of ignition source (anthropogenic). Any structures that are located in these areas are relatively surrounded by fuel.
Intermix: Characterized as areas with moderate to high structure density with combustible vegetation present, the typical Wildland Urban Interface (WUI). This model takes into account the presence of homes as a possible fuel source along with an algorithm to model the complexity of road networks, home density, and distances to water sources and fire stations which impact the firefighting resources ability to fight fires in this environment.
Interface: Characterized as areas that cannot sustain wildfire due to lack of wildland fuel but are at risk for losses due to heavy ember showers from adjacent wildland areas or structure conflagration. Interface scores are heavily influenced by the fireline intensity score of the adjacent firesheds and a distance calculation of the nearest fireshed that would sustain a crown fire.
Model Type is identified in the generated report and generally used for reference only.
Fire History and Future Fire Simulation
Severity is an important component to any wildfire score; however, it must be evaluated along with a Frequency score in order to accurately represent the risk. Wildfire frequency varies widely for many reasons including elevation, occurrences of lightning and other ignitions, and typical weather patterns.
It is important to note that many agencies, regions, and states do not require fire reporting and thus the historical record can be unreliable and misleading. Records for federal land are generally more complete.
Fire Frequency is an often-overlooked component of the wildland fire challenge. The RZRisk service uses a 3-component approach to model Wildfire Frequency: Fire History, Fire Ecological Characteristics including Fire Return Interval, and Simulated Fire Occurrence Probability.
The Fire History model uses records of historical wildfires tracked by the US Forest Service and other government agencies. Actual past fire occurrences are an important measure of the expected future fire frequency. In some geographic regions, reliable records of wildfires on state and private land do not exist. A statistical smoothing technique was used to
extrapolate average wildfire occurrences on private/state lands given similar ecological and climatic conditions.
Fires in wildland vegetation display a range of fire behaviors and fire characteristics that depend on factors such as the vegetation composition and fuel structure, stage of succession after previous fires or other disturbances, types of past management, climate and weather patterns, terrain, and landscape patterns. As an example, an area dominated by a grass or shrub fuel model would often expect to experience wildfire occurrence more often than a high alpine coniferous landscape. The Fire Return Interval represents these factors and is incorporated into RZRisk’s Ecological Characteristics component.
Stochastic simulations of wildfire occurrence and growth are an integral part in the assessment of wildfire frequency. RZRisk uses the FSPro (http://www.fs.fed.us/rm/pubs_other/rmrs_2010_finney_m002.pdf) simulation system which was originally implemented in the Wildland Fire Decision Support System (WFDSS).
The Overall Score, ranging from 0 to 50, is a summary of all the sub-scores including Fire Severity, Fire Frequency, Fire Station Proximity, Past Fire Score, and Loss Type. This score represents the best single rating encompassing all inputs and intermediate scores.
Wildfire Severity is an estimate of how severe fire behavior would be in the event of an ignition. Factored into this estimate is the topography (slope, aspect and elevation), the prevailing weather patterns in the area (based on weather reading at stations located nationwide), and the fuel type present (40 different subsets of grass, shrub, and timber vegetation types). Values range from 0 (least severe) to 50 (most severe).
Wildfire Frequency is a normalized score relating the frequency value indicating the likelihood of damaging future wildfires based on their historical occurrence in the area and the modeled future frequency value using fire growth simulations. Fires are simulated on the landscape based on topography, weather, and fuel. The model incorporates the known historical behavior, volume, and location of ignitions to start simulated fires in specific locations across the United States. These simulated fires are then allowed to burn using the computer model and the locations where they burned (within the simulated perimeters) are recorded. Values range from 0 (least frequent) to 50 (most frequent).
Description: The Frequency Description field contains five descriptive text classifications of the Wildfire Frequency Score.
Fire Station Proximity
Fire Station Proximity is a normalized relative score indicating the distance from the risk to the nearest fire station. Structures located nearer to fire stations generally have a greater probability of a successful wildfire suppression or structure protection effort. (0 is closer to a fire station and 49 is farther away).
RedZone Wildfire Risk Scores can be ordered through three channels.
- Online Store
We make it easy to order a report. Visit our online store (http://secureaws.redzone.co/rzrisk- store) to make individual purchases and download a custom PDF with embedded mapping.
- Web Service
RedZone offers a secure REST web service for ordering wildfire risk scores. A user account is required.
The RZRisk API can be accessed through https://secure.redzone.co/api/rzrisk/api.html (Requires a login. Contact us for more information.)
- Batch Processing
Batch processing is available by contacting our sales team at email@example.com or 303-386- 3955.