Severe fire seasons of the past decade in the western United States have spurred many government agencies to manage lands to reduce fire intensity and severity to ultimately protect human life and property.

However, seven decades of fire exclusion policies have resulted in the dense forest canopies, high surface fuel accumulations, and increased fuel continuity across large regions where fires were historically frequent. These abnormal fuel conditions may foster abnormally severe wildfires that are projected to increase with global warming. The western US has also experienced a marked increase in human development in areas surrounding public wildlands thereby creating and expanding a “wildland urban interface”. With this expansion comes an increased risk to human life and property as severe wildfires become increasingly common. In response, federal agencies have advocated fuels reduction treatments to mitigate the risk and hazard of severe wildfires, particularly in the wildland urban interface. With limited available funding and the cost of fuel treatments continually increasing, fire managers have been charged with developing a detailed methodology for identifying and prioritizing which federal lands are in the greatest need for fuels reduction treatments. A quantification of fire hazard and risk is critical for identifying and prioritizing areas for fuel fuels treatments, and comprehensive fire models are an important first step towards providing spatially explicit estimates of fire risk and hazard over a range of spatial and temporal scales.

This project involves the development of a research computer model called FIREHARM (FIRE HAzard and Risk Model) that computes common measures of fire behavior, fire danger, and fire effects over space to use as variables to portray fire hazard spatially, and then computes fire risk by simulating daily fuel moistures over 18 years to compute fire measures over time. The digital hazard and risk maps can then be used for fire management planning and real-time wildfire operations. Extensive validation of six FIREHARM output variables is being conducted to estimate model accuracy and precision to aid in the interpretation of results.

PRINCIPAL INVESTIGATOR

Robert E. Keane, Deputy Program Manager, Fire, Fuel, and Smoke Science (FFS); Research Ecologist; Director, Fire Modeling Institute (FMI)

Staff

Eva Karau, Ecologist; Stacy Drury, Ecologist; Jason Herynk, Research Technician (SEM)

Collaborators

Paul Hessburg and Keith Reynolds, Pacific Northwest Research Station, Wendel Hann, USDA Forest Service Fire and Aviation Management.

GOALS AND OBJECTIVES

The objective of this study is to develop methods of computing fire hazard and risk that minimize the limitations and drawbacks of previous efforts. To accomplish this we created the FIREHARM computer model that 1) increased consistency across hazard and risk layers, 2) standardized weather inputs, 3) employed a multiple scale analysis into its structure, 4) included some spatial effects, and 5) expanded the number of fire hazard and risk variables. The audience for this effort is managers and researchers interested in describing and evaluating fire hazard and risk across multiple spatial scales. This research may lead to new methods of prioritizing fuel treatments after major insect, disease, or fire events and it will provide important parameters and values for future efforts.

METHODS AND RESULTS

FIREHARM is a C++ program that computes landscape changes in fire characteristics over time by using a spatial daily climate database to simulate fuel moisture which is then used to calculate commonly used measures of fire behavior, danger, and effects . FIREHARM is more of a modeling platform than a fire model because it integrates previously developed fire simulation models into its structure and does not include any new fire behavior or effects simulation methods. The model assumes static fuel characteristics so it does not simulate vegetation development or fuel accumulation over time. Although FIREHARM input and output are spatial, the model is not spatially explicit because it does not simulate spatially explicit processes such as fire spread. Instead, the model assumes that every pixel or polygon experiences a head fire and then simulates the fire characteristics from antecedent weather. FIREHARM does not simulate crown fires directly but it does calculate crown fire intensity.

Currently, fuel hazard mapping for fire management is limited by four major factors: 1) computational resources available to fire management organizations, 2) high quality, spatially consistent, management-oriented spatial data layers at the appropriate scale and resolution, 3) lack of error and uncertainty estimates for the spatial data layers, and 4) improper spatial analysis techniques. This study is developing a method for generating spatially consistent spatial data appropriate for fire hazard analysis with the level of quality dependent on available input data, scale of analysis, and management objective. There are advantages and disadvantages of using FIREHARM hazard (event mode) or potential risk (temporal mode) maps. While hazard maps can be quickly created by assuming representative fuel moistures, they can be difficult to interpret because they do not incorporate the temporal frequency of the representative fuel moistures in the assessment. On the other hand, potential risk maps are difficult to create because FIREHARM 1) requires accurate estimations of site conditions (soil depth, texture, leaf area index), 2) must be linked to the very large DAYMET weather database, 3) must simulate fire characteristics for every day in the DAYMET record, and 4) must simulate daily ecosystem process (water budget) along with fire characteristics. FIREHARM risk maps may take days to create while hazard maps can be created in hours depending on the size and resolution of the landscape. We find that large, regional analysis can be successfully accomplished using the hazard maps, but fine scale project level analysis are appropriate for the potential risk maps. While we acknowledge that FIREHARM isn’t the perfect solution to quantifying fire hazard and risk across multiple scales, we feel it is a step in the right direction.

REFERENCES

Keane, R.E., Drury, S., Karau, E., Hessburg, Paul F., Reynolds, K. 2010. A method for mapping fire hazard and risk across multiple spatial scales and its application in fire management. Ecological Modelling 221:2-18

FUNDING SOURCES

This work is currently being funded by USDA Forest Service Washington Office of Fire and Aviation Management, Joint Fire Sciences Program, and the RMRS Fire Research and Development Project.

PROJECT STATUS

This is a 4 year study with model building and testing, results validation, and model refinement taking place from 2006 to 2010.