College of Forestry

In 2021 and 2023, the Oregon legislature passed a series of bipartisan bills to help people living in Oregon improve their wildfire preparedness. As a trusted, non-biased source of wildfire risk information, OSU was directed to develop a map of wildfire hazard for each property across the state using four criteria: climate, weather, topography, and vegetation. 

An updated draft version of the statewide wildfire hazard map was released on July 18, 2024, via Oregon Explorer, which serves as a place for individuals, agencies, Tribes, and organizations to access mapping tools and resources relevant to natural resources decision making. Since 2007, the Oregon Explorer, a program of the Institute of Natural Resources at Oregon State University and OSU Libraries and Press, has been a public source of geospatial data and wildfire risk information used in state, regional and local risk management applications. 

Oregon is fortunate to have an abundance of wildfire risk-related planning data and assessments, but it’s important to remember that each assessment and dataset was developed to meet a specific objective. In other words, all risk assessments are a little different and the differences matter.

Oregon’s statewide wildfire hazard map was specifically designed to meet the requirements and needs described in Senate Bill 762 (2021) and Senate Bill 80 (2023). These bills provide direction and investment to multiple state agencies to help improve wildfire preparedness across the state. As part of Senate Bill 762, Oregon State University was directed to create a wildfire risk map based on rule adopted by the Board of Forestry and four environmental criteria: climate, weather, topography, and vegetation. In 2023, Senate Bill 80 adjusted the direction on some of the statewide wildfire mitigation measures in SB 762, including the development of a statewide wildfire hazard map (rather than a wildfire risk map) with revised hazard zone classifications. While Senate Bill 762 described the development of a wildfire risk map, in reality the map has always technically been a wildfire hazard map because of the four criteria required to make it. Senate Bill 80 made this correction in wording.

Oregon state agencies will refer to the statewide wildfire hazard map and the WUI map when fulfilling their responsibilities in Senate Bill 80. The data and methods used by OSU scientists were carefully selected to meet the needs of state agencies based on current available data, peer-reviewed science, and guidance from Rulemaking Advisory Committees (RACs) organized by Oregon Department of Forestry (ODF). Other hazard maps that exist were likely developed for different objectives and therefore used different data and methods. In most cases, it would not be appropriate to directly compare results from one hazard map to another unless the same specific methods were used in both.

Wildfire hazard is the combination of how likely a wildfire is to occur in a specific location (burn probability) and how much heat energy the fire gives off ("fire intensity"). For more details on wildfire hazard, visit Understand the Map.

Burn probability represents the annual probability of a wildfire occurring at a specific location, and, along with fire intensity, burn probability is included in the wildfire hazard calculation.

Burn probability was evaluated using best available science, state of the art wildfire simulation models and expert knowledge from local fire and fuel professionals. The simulation models incorporate spatial data about climate, weather, topography, vegetation, and characteristics of historical fire occurrence. A simulation model is a computer representation of objects, systems, or events that can be used as a tool to help explain or predict the behavior of real systems. Scientists ran 10,000 simulations of plausible fire seasons to capture the enormous variation in where and when fires could ignite and how they could spread across the landscape. Burn probability was calculated by counting the number of times a given location was impacted by simulated fires and dividing that count by the total number of simulated seasons (i.e. 10,000).

Burn probabilities are represented as decimals. For instance, a burn probability of 0.005 is equivalent to saying that in any given year there is a 0.5% chance of wildfire occurring in that location. Burn probabilities are not forecasts of where wildfires will occur in any specific year but are instead long-term average probabilities. Those long-term averages help scientists and managers prioritize mitigation work in landscapes and communities that are most likely to experience a wildfire.

Wildfire intensity represents the amount of heat produced by a wildfire and, along with burn probability, is included in the wildfire hazard calculation. Wildfire intensity is an essential component of hazard because varying intensities can lead to different impacts to structures and therefore different levels of hazard.

Similar to burn probability, wildfire intensity is evaluated using best available science, state of the art wildfire simulation models, and expert knowledge from local fire and fuel professionals. The models incorporate spatial data describing climate, weather, topography, vegetation, and characteristics of historical fire occurrence. A simulation model is a computer representation of objects, systems, or events that can be used as a tool to help explain or predict the behavior of real systems. Scientists simulated wildfire behavior under 200 different weather scenarios to capture the range of fire intensities and their relative likelihood at all locations in Oregon. Any location could experience varying fire intensities because all locations experience a range weather severity throughout fire season. So, fire intensity at any given location is calculated as the average fire intensity experienced at that location across all simulations.

Fire intensity is represented as flame lengths. The greater the flame lengths, the greater the intensity and the more opportunity for damage to structures.

Burn probability and fire intensity were both modeled using best available science and data describing climate, weather, topography, and vegetation. To model burn probability and fire intensity, researchers used a collection of spatial datasets that characterize Oregon’s vegetation during the 2022 fire season and how those vegetation characteristics might affect fire behavior. The models consider the amount and type (e.g. timber, grass, shrub) of vegetation as well relevant characteristics like the amount of litter, canopy density, canopy height and other factors that influence fire behavior. These spatial datasets are generally available from LANDFIRE, but were modified based on the input of more than 50 fire and fuels specialists from across Oregon to more accurately reflect conditions on the ground.
 

Within the simulation models, scientists used observed weather data (2007 – 2021) collected from Remote Automated Weather Stations around Oregon.

Recent wildfires will have variable impacts on property-level hazard values according to the proximity of the fires to the property in question, how long ago they happened, in what kind of vegetation the fires burned and where in the state they occurred. On one hand, a recent wildfire can reduce the amount of fuel available to future fires in the same area, thereby reducing intensity. Reduced intensity can lead to a lower hazard rating. Similarly, where a previous fire burned (called a “fire footprint”) can affect fire spread and may result in fewer fires reaching the property in question, again resulting in a lower modeled risk. On the other hand, as a burned area re-vegetates and recovers the new vegetation may actually be more flammable, leading to more fires, faster spread, and higher intensity, with an overall increase in hazard.

Importantly, the modeled reduction in fire intensity and fire spread in recently burned areas does not last forever. The hazard map will be updated at least every five years, and each time the landscape will be updated to account for changes in vegetation, including vegetation that may have regrown in historic fire footprints.

Wildfire hazard is the combination of how likely a wildfire is to occur in a specific location (burn probability) and how much heat energy the fire gives off ("fire intensity"). Researchers at OSU worked with the nation’s leading fire modeling experts as well as local fire experts to model and map burn probability and fire intensity all across Oregon. A simulation model is a computer representation of objects, systems, or events that can be used as a tool to help explain or predict the behavior of real systems. Burn probability and fire intensity are first modeled on 30-meter rasters across all of Oregon. Rasters are a way of representing spatial data in a uniform grid. In this case, all of Oregon is represented by a grid comprised of cells (called “pixels”) that are each 30 x 30 meters. Burn probability and fire intensity were calculated at each pixel and then combined to calculate hazard at each pixel. To calculate property-level hazard values, researchers averaged all the pixel-level hazard values from pixels within each property. Based on the property-level average hazard value, each property was assigned to one of three zones: Low, Moderate or High Hazard.
 

For details on how wildfire hazard was mapped, visit Understand the Map.

Property hazard values might differ among neighbors for many reasons, some of which are observable on the ground. Differences in the amount and type of combustible vegetation on and around the property can affect fire hazard calculations. For instance, the fuels adjacent to one neighbor’s structure may be more timbered than the fuels adjacent to another neighbor’s structure, resulting in higher modeled fire intensity and therefore higher hazard. Or, neighboring tax lots might be in different hazard classes because one tax lot is situated closer to an adjacent large area of flammable vegetation, elevating the burn probability compared to the other tax lot.

However, property hazard values are also affected by landscape-level processes which are not observable when looking at the property itself. For instance, landscape features like ridges and hillsides might shelter one property more than another from local winds – and therefore when fires occur they are pushed away from the sheltered property and towards its neighbor more often than not. These kinds of features are not always visually obvious, but over the course of 10,000 simulations the result is that the sheltered property has a lower burn probability than its neighbor and could have a lower hazard value as well. Topography, landscape features (e.g. water bodies), and vegetation can all interact in ways that are not immediately visible from a single property, but which create local conditions (i.e. microclimates) that either raise or lower burn probability and fire intensity. 

Defensible space is the buffer that can be created between a structure and the grass, trees, shrubs, or any wildland area that surrounds it. The statewide wildfire hazard map does not incorporate any work that has been done on the property when determining the property-level hazard rating. This is because Senate Bill 80 specifically directs OSU to consider only four criteria: climate, weather, topography, and vegetation. There is no adequate statewide spatial dataset that documents the unique defensible space characteristics of each and every structure in Oregon.

The statewide wildfire hazard map does not consider fire hardening characteristics when determining the property-level hazard rating. Fire hardening refers to using building materials and practices that can reduce the hazard of a structure becoming ignited by embers from wildfires. The primary reason that pre-existing fire hardening is not considered when determining property-level hazard is that Senate Bill 80 specifically directs OSU to consider only four criteria: climate, weather topography and vegetation. Fire hardening conditions do not fit into those four criteria. Further, there is no adequate statewide spatial dataset which might be used to consider the unique fire hardened characteristics of each and every structure in Oregon.

No. Working with the hazard input data and calculating property-level hazard values requires access to and knowledge of how to use specialized spatial analysis software.  

Although the Oregon Explorer presents property-level average hazard values, hazard can vary across a single tax lot. In fact, depending on the size of the property, where it’s located and how the property is managed, hazard can vary tremendously from one part of the property to another. That’s why researchers at OSU calculate hazard at a very fine scale using a 100’ by 100’ grid across all of Oregon before determining property-level average hazard values. In other words, all of Oregon is divided into uniform cells (“pixels”) that are each approximately 100’ by 100’. Researchers generate estimates of burn probability and fire intensity for each pixel and then calculate hazard for each pixel. Property-level hazard values represent the average of all the pixel-level hazard values from pixels within the property boundaries.  

Depending on the size of the property, there could be hundreds or thousands of pixels representing variation in hazard across a single property. Reproducing property-level hazard values requires that those many pixel-level hazard values be averaged within a specific tax lot boundary using spatial analysis software.  

A draft version of the Oregon Statewide Wildfire Hazard Map is now available for public comment on the Oregon Explorer. All public comments should be directed to the Oregon Department of Forestry via hazardmap@odf.oregon.gov. The public comment period closes on Aug. 18. After the public comment period and the Oregon Board of Forestry’s September meeting, a finalized map will be released and property owners who are both within the wildland-urban interface and have a “high” hazard classification will be notified by mail. More information will be available this fall. 

The wildland-urban interface is the geographic area where structures and other human development meet or intermingle with forests, rangelands and other vegetation. This area is the transition zone between wildlands and human communities.

You can think of the WUI as a scale. On one side of the WUI, in the wildlands, fires are less likely to damage buildings because there are too few of them. On the other side of the WUI, in the developed core of a community, there is not enough vegetation to support wildfires, and therefore the wildfire hazard is lower. Between those two zones, the developed core and wildlands, is the WUI where there is enough vegetation to support a wildfire AND there is enough development that wildfires could result in significant damage to homes, critical infrastructure and human lives.

The WUI can be defined and mapped in different ways. In Oregon, the WUI is defined in state statute (OAR 629-044-1005 (m)) as “a geographical area where structures and other human development meets or intermingles with wildland or vegetative fuels.” The criteria for mapping that geographical area are defined in OAR 629-044-1011 and described on the Understand the Map webpage.

The wildland urban interface map is separate from the wildfire hazard map, though the two are designed to be used together by state agencies. To identify properties where defensible space and fire hardening codes might apply, agencies will use both maps to identify properties that are in the WUI and classified as high hazard. Only these properties meeting both criteria are potentially subject to defensible space and fire hardening rules described in Senate Bill 762 and Senate Bill 80.

The Wildland-Urban Interface (WUI) can be defined and mapped in different ways depending on how the information is to be used. Oregon’s Senate Bill 80 (2023) requires that the statewide WUI map be used together with the statewide wildfire hazard map to identify structures that are both in the WUI and located on properties classified as high hazard. Structures meeting both those criteria could be subject to further defensible space and/or fire hardening rules. Accordingly, OSU scientists mapped the statewide WUI by focusing primarily on structure locations which, compared to other WUI maps in Oregon, resulted in much less of the state being identified as WUI.

Oregon's Senate Bill 80 specifically requires a map of the "wildland-urban interface" (WUI) because that term has been adopted by international wildfire, natural hazard, emergency, and land use professionals. In the context of the WUI, the word "urban" does not refer to cities or highly developed areas. The WUI refers to all places on a landscape where vegetation and structures meet or intermingle (see OAR 629-044-1005 (m)), thereby increasing the potential for wildfires to damage or destroy structures. See Understand the Map for more information.

The WUI map was developed using an objective set of rules and analytical techniques. WUI classifications were determined by the location of buildings within each tax lot, the density and proximity of adjacent buildings, and the density and proximity of vegetation. The unique setting of each building informs whether or not it is in the WUI and its estimated wildfire risk.