5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

1. Is It Time To Say Goodbye to Fire Rotations? Cecil Frost, Landscape Fire Ecologist, University of North Carolina Abstract: The idea of fire rotations goes back to widespread dawning, in the 1960s, of the realization that we had to get fire back into the woods. This was a part of a ponderous midstream turnaround that began with a few small voices such as H.H. Chapman, speaking against the unresistable tide of fire suppression of the early 20th century. Movement in the new direction toward restoration of fire gained way with early publications in the Proceedings of the Tall Timbers Fire Ecology Conferences from 1960- 1966. The idea of fire rotations at its simplest was to take a natural area, say a 500,000 acre national forest, and if you had the resources to burn 5000 acres a year, start in one place, burn the next 5000 acres the next year and so on until you had reached the beginning. 500,000/5000 = 100 years to complete one rotation. This, we now can see, is a long term recipe for disaster. What we see from the most recent iterations of LANDFIRE—from Florida, through the central grasslands to the Palouse prairie of eastern Oregon and Washington—is that most of the U.S. was a non-seral landscape, naturally stabilized by frequent fire—the more frequent, the more stable. Within this fire system, recent maps show that local fire frequency at fine scales was far more complex than previously perceived. Fire frequency depends upon landscape and the natural variation in terrain in most U.S. landscapes creates hotspots and fire sheltered spots, each with its suite of fire dependent or fire refugial species. Subjecting naturally complex areas to fire rotations means eliminating all of the most fire related species and conversion to low diversity, fire refugial vegetation—the fire-excluded thickets seen all across the continent today. The alternative to fire rotations is knowledge-based focused burning. The critical needs of fire dependent species and communities (and there are more than we ever thought) demand that we burn the habitats they need, as often as they need, and leave the rest to wildfires. Keywords: Fire rotations, Focused Fire, Prescribed Fire Presenter Bio: Cecil Frost is a landscape fire ecologist with a specialty in mapping Pre-European fire regimes of the U.S.

2. Restoration of xeric oak forests in south-central United State with prescribed fire Stephen Hallgren, Assoc. Prof. Forest Ecology, Oklahoma State University Abstract: Xeric oak forests of south-central United States were regularly burned by Native Americans over thousands of years prior to Euro-American settlement. Fire suppression beginning in the mid- Twentieth Century led to increased stand density and encroachment of fire-intolerant mesophytic species. Prescribed fire has recently been rediscovered as a tool for managing these forests for biological diversity, fuel management, wildlife habitat and municipal watershed protection. Research was conducted to determine effects of long-term prescribed burning on stand composition and structure, litter decomposition, coarse woody debris, and soil nitrogen and carbon. Prescribed burning had little effect on dominant oak species and strongly reduced fire-intolerant woody species. Herbaceous vegetation showed large increases with increasing frequency of burning. Prescribed burning did not appear to strongly affect litter decomposition, perhaps because decomposition was relatively rapid in the humid subtropical climate. Likewise coarse woody debris did not show strong effects of prescribed burning even at high burn frequencies. Soil nitrogen and carbon showed only minor effects of prescribed burning. Burn frequencies similar to those used by Native Americans of one to two burns per decade produced dramatic reductions in understory woody vegetation and increases in herbaceous vegetation without strong impacts on soils. Keywords: post oak, blackjack oak, decomposition, soil carbon Presenter Bio: Steve Hallgren teaches and researches fire ecology of forest ecosystems at Oklahoma State University. In addition, he is the OSU Fulbright Program Adviser.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

3. Post-fire tree mortality model assessment following prescribed burning treatments in National Park units of the western U.S. Jeffrey Kane, Assistant Professor, Humboldt State University Additional Authors: Phil van Mantgem, Research Ecologist, United States Geological Survey Laura Lalemand, Research Technician, United States Geological Survey Abstract: A common way to assess the effectiveness of prescribed fires is through monitoring tree mortality. Managers require accurate models to predict tree mortality to maximize the effectiveness and benefits of prescribed burns. Here we assessed the performance of a commonly used post-fire tree mortality model with a geographically robust monitoring dataset of 18 tree species (12 gymnosperms, 4 angiosperms) from 16 National Park Service units in the western U.S. Model performance was generally strong with mortality predictions for 12 of the 18 species within 20% of the observed values, however, correctly classified dead trees ranged between -5 and 43%. Variation in model accuracy among species across all sites was not related to sample size (r2 < 0.08, p > 0.24), bark thickness (r2 < 0.17, p > 0.09), or percent observed mortality (r2 < 0.09, p > 0.22). Although we did find a significant relationship between bark thickness and the percentage of correctly classified live trees, where thicker bark species had more live trees correctly classified (r2 = 0.62, P = 0.0001). These results indicate that the commonly used post-fire mortality model generally performs well; however, thinned barked species and angiosperm tree species would benefit from future studies that provided model improvements. Keywords: fire ecology, fire effects, modeling, model validation, prescribed fire, tree mortality Presenter Bio: Jeff is an Assistant Professor of Fire Ecology and Fuels Management at Humboldt State University. He is also the Director of the HSU Wildland Fire Lab where he conducts research on the effects of prescribed fire and wildfire on improving forest health and biodiversity.

4. 2015 National Prescribed Fire Use Survey Pete Lahm, Air Resource Specialist, USFS Additional Authors: Mark Melvin, Conservation Management/Education Technician, Jones Ecological Research Center Abstract: In 2012 the Coalition of Prescribed Fire Councils (CPFC) and the National Association of State Foresters (NASF) partnered to produce the first-ever report that investigated national prescribed fire activity. Understanding how much and where prescribed fire occurs is important to the entire fire community; and the report has been used as an informative tool to help guide decision making, research, and policy by wildland fire agencies, prescribed fire councils, academia, and air quality agencies. In an effort to garner and maintain the most current data, the NASF and CPFC have partnered once again to produce a sister 2015 report. This 2015 survey is a national evaluation focusing on the scale of prescribed fire use, state-level supporting programs, and factors that limit more use of prescribed fire. It is based on 2014 prescribed fire activity (agricultural, rangeland and forestry burning) and utilizes the 2012 report as a comparison to illustrate and describe areas of stabilization and trends. This report uses the best information available from state forestry agencies, and 100% participation from states was achieved. The report is intended to inform the fire community by identifying challenges at a national scale, as well as defining key regional and state differences and similarities. Keywords: Prescribed Fire Use, National Presenter Bio: Air Resource Specialist for USDA Forest Service, Fire and Aviation Management Program in the Washington Office. Chair of the NWCG Smoke Committee and Forest Service Liaison to the Coalition of Prescribed Fire Councils. Smoke watcher for 25 years.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

5. The Smoke-wise Community and the Path to More Fire Peter Lahm, Air Resource Specialist, USDA-Forest Service Abstract: As we collectively strive to create more fire-adapted communities and restore fire dependent ecosystems, there is a need to progressively address the challenge of fire effects and specifically smoke. This is a challenge that many see as the key obstacle to more use of prescribed fire and use of some wildfire. The immutable reality is that smoke negatively affects many in this country who have respiratory issues, in fact one in three households has someone with these issues. So as we continue in the important path of recognizing the role of fire in our landscapes with Firewise Communities and Community Wildfire Protection Plans, the time has come to address smoke in a more earnest manner. The EPA has signaled a clear recognition of the role of fire in ecosystems and prescribed fire as an important tool to combat catastrophic wildfire in many recent rules. Efforts are underway to warn the public of hazardous air quality down-wind of significant wildfires with the Wildland Fire Air Quality Response Program. Air quality matters to the American public and where there is fire there is always smoke. Smoke which will affect more people directly and potentially just as fatally as a raging wildfire. The time has come to understand our preparation as a fire community to address smoke as part of wildland fire. Not only is this challenge for those in the public who are sensitive, but also anyone who cares about their health including firefighters. How the fire and land management communities address smoke will have a substantial impact on the outcome of our collective cohesive strategy for restoring fire-adapted ecosystems. Preparation of Smoke-wise Communities will be needed for the efforts to succeed, not only for air quality impacts from wildfire but to maintain and increase the use of prescribed fire. New partnerships for fire use will need to be developed, innovative collaboration and exploration of flexibility in air quality rules will need to be discussed and acted upon. The fire and land management community’s training and experience will need improvement to meet this important challenge. Keywords: Smoke; Community Protection; Air Quality; Restoration Presenter Bio: Air Resource Specialist for USDA Forest Service, Fire and Aviation Management Program in the Washington Office. Chair of the NWCG Smoke Committee and Forest Service Liaison to the Coalition of Prescribed Fire Councils. Smoke watcher for 25 years.

6. Multiphase CFD Model of Wildland Fire Initiation and Spread Vladimir Agranat, President & Senior CFD Consultant, Applied Computational Fluid Dynamics Analysis Additional Authors: Valery A. Perminov, Professor, National Research Tomsk Polytechnic University, Tomsk, Russian Federation Abstract: The aim of this paper is to create a user-friendly computational tool to be used by students, researchers and fire management consultants for the analyses of wildland fire propagation and its behavior in Wildland-Urban-Interface (WUI). A physics-based multiphase Computational Fluid Dynamics (CFD) model of wildfire initiation and spread has been developed and incorporated into the multi- purpose CFD software, PHOENICS. The model is based on previous works of Grishin with co-workers and Porterie with co-workers and it accounts for all the important physicochemical processes: drying, pyrolysis, char combustion, turbulent combustion of gaseous products of pyrolysis, exchange of mass, momentum and energy between gas and solid phase, turbulent wind flow and convective, conductive and radiative heat transfer. Turbulence is modeled by using a RNG k-ε model and the radiative heat transfer is approached with a model similar to P-1 radiation model. The Arrhenius-type kinetics is used for heterogeneous reactions (drying, pyrolysis and char combustion) and the eddy-breakup approach is applied for modeling the gaseous combustion. The CFD model has been validated using the experimental data of Mendes-Lopes and co-workers on surface fire propagation in a bed of Pinus pinaster needles studied in a wind tunnel. The predicted rate of spread (ROS) is well agreed with 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

experimental values obtained at various wind speeds (from 1 to 3 m/s). The model is being further validated using the data on large forest fires including crown fires. Keywords: Multiphase Model, CFD, Wildland Fire, Rate of Spread Presenter Bio: President and Senior CFD Consultant of Applied Computational Fluid Dynamics Analysis (ACFDA) since 1998. Published more than 90 research papers. Obtained PhD in Fluid Dynamics in 1978. Worked in various R&D organizations and engineering consulting firms, including Tomsk State University in Russia, Weizmann Institute of Science in Israel, the University of Toronto in Canada, Atomic Energy of Canada Ltd., Hydrogenics Corporation and ACFDA. CFD applications have covered fire propagation and prevention, nuclear and power engineering, chemical engineering, environmental protection, hydrogen generation, gas-liquid separation, etc.

7. Data-driven Forecasting Paradigms for Wildland Fires using the CAWFE modeling system and Fire Detection Data Janice Coen, Project Scientist, National Center for Atmospheric Research Additional Authors: Wilfrid Schroeder, Research Professor, University of Maryland Patricia Oliva Pavon, Postdoctoral Researcher, University of Maryland Abstract: Coupled weather-wildland fire models tie numerical weather prediction models to wildland fire behavior modules to simulate the impact of a fire on the atmosphere and the feedback of these fire- induced winds on fire behavior. Case studies retrospectively applied the CAWFE® coupled weather- wildland fire modeling system to landscape-scale wildland fires, demonstrating its capability to model the unfolding of large fire events over a wide range of conditions, reproducing rate and direction of spread and expanding perimeter shape; events such as locations of sudden acceleration, flank runs up canyons, and bifurcations of a fire into two heads; and locations favorable to formation of phenomena such as fire whirls and horizontal roll vortices. Applied as a forecast, such systems must consider the decrease in weather forecast model skill with time, effects such as firefighting that interfere with natural fire growth, and processes such as spotting that may be beyond the ability of deterministic models to predict. Prior work applied forecasting approaches such as cycling, in which a sequence of CAWFE simulations initialize the fire 'in progress' with VIIRS data and updated atmospheric analyses to overcome such forecasting issues and allow good 12-36 hour forecasts of fire weather and behavior from first detection until extinction. Here, we adapt and apply this and additional forecasting techniques in the context of recent wildfire outbreaks. Gridded output from large-scale numerical weather prediction models provide the initial conditions and later boundary conditions for CAWFE forecasts. Forecast evaluation, with verification data provided by other atmospheric and fire data, will be discussed. Keywords: forecast, active fire detection data, error growth, weather Presenter Bio: Dr. Janice Coen is a Project Scientist at the National Center for Atmospheric Research in Boulder, Colorado. She received a B.S. in Engineering Physics from Grove City College and an M.S. and Ph.D. from the Department of Geophysical Sciences at the University of Chicago. She has served as a member of the Board of Directors of the International Association of Wildland Fire and is currently an Associate Editor for the International Journal of Wildland Fire. She investigates wildland fire behavior and its interaction with weather using coupled weather- wildland fire computer simulation models and by analyzing infrared imagery of wildfires and prescribed fires.

8. GridFire: A Fast Raster-Based Fire Spread and Severity Model Gary Johnson, Senior Scientist, Spatial Informatics Group, LLC Additional Authors: David Saah, PhD, Managing Principal, Spatial Informatics Group 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Max Moritz, PhD, Associate Cooperative Extension Specialist, UC Berkeley Abstract: GridFire is a fast, raster-based fire spread and severity model implemented in the Clojure programming language and released as open source software under the GPLv3. Designed to be driven from the command line, GridFire is straightforward to run for one-shot or comparative analyses but can also easily be scripted into large scale Monte Carlo simulation pipelines. In static mode, the model can produce maps of fire type (i.e., non-burnable, surface, passive crown, active crown), maximum rate and direction of spread, flame length, fire line intensity, and wind adjustment factor given static weather conditions and a raster stack of topography and fuels data (e.g., LANDFIRE). In its dynamic mode, individual ignition events may be simulated for a fixed burn time to give the fire size as well as all of the static metrics listed previously in the burned area. To enable relatively inexpensive horizontal scaling (i.e., increasing computing capacity by using many commodity computers rather than one large supercomputer), GridFire is designed to work with the open source geospatial database system Postgresql + PostGIS. All input raster data, such as the LANDFIRE stack or climatology data, are first loaded into a PostGIS-enabled database where they may then be served to any number of computers running the GridFire application. Using this system architecture, Spatial Informatics Group (SIG) was able to run over 100 million 1-hour fire simulations under extreme historical weather conditions for the state of California using a cluster of relatively inexpensive computers (most with 16 cores or less) in approximately three days. This presentation will discuss the details of the GridFire model, including its software architecture and the published fire spread and severity algorithms it implements. It will also present some of the results of recent work in creating statewide fire risk maps for the state of California using a large scale Monte Carlo simulation approach. Future plans for the software and avenues for collaboration will also be discussed. Keywords: Fire modeling, raster-based models, open source software, scalable architectures, Monte Carlo simulation, Clojure, PostGIS Presenter Bio: Dr. Gary Johnson is a computational scientist with over eight years of experience modeling the spatial flow of services from ecosystems to people and changes to these services by different land management, natural hazard, and climate change scenarios. In addition to ecosystem services, his research has focused on wildfire modeling, extreme weather statistics, decision support systems, risk assessment and uncertainty modeling, data mining, and machine learning. As a senior scientist with Spatial Informatics Group (SIG), his passion has become finding new ways to use advanced computational techniques to solve meaningful environmental problems.

9. Towards an integrated fire-atmosphere prediction system with data assimilation Sher Schranz, CSU/CIRA Associate Director, NOAA ESRL Global Systems Division Additional Authors: Adam Kochanski, Research Assistant Professor, Atmospheric Sciences Department, University of Utah Mary Ann Jenkins, Associate Professor, Department of Earth and Space Science and Engineering Lassonde School of Engineering York University Jan Mandel, Professor and Chair, Department of Mathematical and Statistical Sciences University of Colorado Denver Martin Vejmelka, Research Assistant Professor, Department of Mathematical and Statistical Sciences University of Colorado Denver Abstract: We present an integrated fire modeling system for wildfire simulation, based on a multi-scale weather forecasting model, coupled with a semi-empirical fire spread model and a prognostic dead fuel moisture model and chemical transport to model the plume behavior and for smoke forecasts. The system operates on a high-resolution atmospheric grid (i.e., horizontal grids 10s to 100s m) to explicitly resolve fire-induced convection and to capture the effects of complex topography and land use 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

characteristics on the local weather. Fire behavior is therefore driven by realistic meteorological forcing, while heat and moisture fluxes released at the fire line are fed back into the model atmosphere, altering air temperature, humidity, and flow. Although fire-induced circulations are considered a local meteorological condition with respect to weather prediction, this two-way coupling may have a significant impact on the larger-scale weather prediction. The modeling system includes a predictive fuel moisture model, with meteorological components, providing detailed temporal and spatial evolution of fuel moisture. This enables simulation of diurnal changes in fire activity associated with wind speed variations and nighttime fuel moisture recovery. No assumptions are necessary regarding the diurnal or spatial fuel moisture variability, as the model simulates it based on its forecasted weather conditions. The dead fuel moisture model assimilates 10h fuel moisture measurements from RAWS stations, while NDWI from satellites can be used to adjust live fuel moisture. The moisture data assimilation significantly increases the skill of the fire spread forecast. The modeling system’s fire component is designed to assimilate satellite fire detections from VIIRS and MODIS, and fire spread forecasts are corrected as new satellite data become available. The model also supports the input of fire perimeter data, while maintaining the consistency of the fire with the state of the atmosphere. This assimilation of fire detection data permits computational simulations of large, mature fires starting from an observed perimeter rather than assumed point ignitions, which are typically not available early in the fire, or even simulations ingesting only automated data. The results of the simulations can be visualized online or delivered in the NOAA’s FX-Net (FX-CAVE) AWIPS client or the NOAA Earth Information System (NEIS). Keywords: WRF, WRF-Sfire, fire modeling, fire data assimilation, FX-CAVE, FX-Net Presenter Bio: Sher Schranz, CSU/CIRA Associate Director at NOAA ESRL Global Systems Division

10. High Fidelity Reduced Order Models for Wildland Fires Alan Lattimer, Vice President R&D, Jensen Hughes Additional Authors: Brian Lattimer, Ph.D., Vice President R&D, Jensen Hughes Serkan Gugercin, Ph.D., Professor, Mathematics Department, Virginia Tech Jeff Borggaard, Ph.D., Professor, Mathematics Department, Virginia Tech Abstract: The use of high fidelity models with detailed physics to predict wildland fires is in part limited by the computational expense of performing these simulations. This paper provides new results in employing reduced order model (ROM) techniques to decrease the computational expense of fire models. The technique was demonstrated through developing ROMs for detailed computational fluid dynamics (CFD) models of fire plumes as well as the fire spread model of Mandel et al. (2008). These ROMs reduce the overall size of the computation by 90% while retaining the physics of the full model, resulting in a computational efficient model that produces output detail similar to that of the full model. ROMs were developed using proper orthogonal decomposition produced from snapshots of results from the full order model. Nonlinear terms in the governing equations were treated separately using discrete empirical interpolation method (Chaturantabut and Sorensen, 2010) to further increase computational efficiency. For the fire plume predictions, the CFD model Fire Dynamics Simulator (FDS) was the full order model used to generate the snapshots for creating the ROMs. Fire spread full order model predictions were performed using the phenomenological fire spread model by Mandel et al. These ROMs provide a reduction in computational time of 2-3 orders of magnitude, which 1-2 orders of magnitude higher than reported by other researchers that have reported using proper orthogonal decomposition alone on wildland fires. The significant decrease in computational time for the work reported in this paper was attributed to the technique employed in this work for treating the nonlinear term. ROMs were also found to maintain accuracy to within 2% when compared with the full order model solutions. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

References Mandel, J., Bennethum, L. S., Beezley, J. D., Coen, J. L., Douglas, C. C., Kim, M., & Vodacek, A. (2007). “A wildland fire model with data assimilation,” Computational Science – ICCS 2007. S. Chaturantabut and D. C. Sorensen. (2010) “Nonlinear model reduction via discrete empirical interpolation,” SIAM Journal on Scientific Computing, 32(5):2737–2764. Keywords: Efficient Model, Fire Plume, Fire Spread Presenter Bio: Alan M. Lattimer, Ph.D. Candidate, Mathematics Department, Virginia Tech

11. Field-scale testing of detailed physics-based fire behavior models Eric Mueller, PhD Student, University of Edinburgh Additional Authors: William Mell, Research Combustion Engineer, USDA Forest Service Nicholas Skowronski, Research Forester, USDA Forest Service Kenneth Clark, Research Forester, USDA Forest Service Michael Gallagher, Research Technician, USDA Forest Service Rory Hadden, Lecturer, University of Edinburgh Albert Simeoni, Senior Manager, Exponent Abstract: Detailed physics-based fire behavior models have great potential for studying the driving mechanisms of wildland fire. These models are built around the fundamental chemical and physical phenomena (fluid flow, heat transfer, solid fuel degradation, and combustion) and consequently offer advantages in application scope over empirical or simplified physical models. With the appropriate input parameters, they can be applied in a wide range of ecosystems and environments. Not only can they be used to investigate specific problems, but they can lead to the development of a suite of more refined, simplified models to be used by the community at large. The complexity of the detailed physics-based approach means that these models must be tested rigorously against experimental datasets. This can be particularly problematic for field-scale applications, where providing measurements of sufficient quality and resolution is difficult. This work presents a methodology developed over three years to investigate experimental measurements of fire behavior made in the field and to model the observed behavior using the Wildland-urban interface Fire Dynamics Simulator (WFDS). To the authors’ knowledge, this work represents an unprecedented effort to apply a model of this nature to a field-scale fire in a forested environment with complex features (e.g. transitional behavior between surface and canopy fuels) with a detailed experimental dataset (including remote sensing and point-based measurements). In order to evaluate the performance of the model and highlight areas for improvement, comparisons are made between the experiments and model predictions for global fire behavior and local measurements. This includes spread rate and local temperature and heat flux measurements. It is demonstrated that the fidelity of global predictions, such as the spread rate, can be better understood through an investigation of local phenomena. This project has created clear insights as to the type and resolution of measurement required to enable the identification of controlling mechanisms in subsequent experimentation. By adopting this approach, the authors hope to begin creating a consistent and comparable ensemble of datasets through future research efforts in the community. Such coherent work is necessary if detailed physics-based models are to be developed and used to their full potential. Keywords: fire behavior, field-scale, experiments, modeling, physics-based Presenter Bio: Eric Mueller is a PhD student at the University of Edinburgh in Scotland. Having completed a B.S. in engineering physics and a M.S. in fire protection engineering, with a focus on wildland fires, he joined the program in 2013. His research is focused around the testing and improvement of physics-based numerical models of fire behavior. His additional research interests 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

include the development of accurate and robust techniques for in-situ experimental measurement of fire environments, as well as methods for quantifying the generation of firebrands.

12. A National Wildfire Risk Assessment for U.S. Forest Service Lands Greg Dillon, Spatial Fire Analyst, USDA Forest Service, Rocky Mountain Research Station, Fire Modeling Institute Additional Authors: Joe Scott, Owner, Pyrologix LLC Abstract: Research over the past decade on improved decision-making in U.S. wildfire management has resulted in a robust and scalable framework for assessing wildfire risk across broad landscapes. Building on advances in wildfire simulation modeling that enabled creation of national datasets depicting the likelihood and potential intensities of wildfires, the U.S. Forest Service published a report in 2010 titled, “Wildfire Risk and Hazard: Procedures for the First Approximation” [Calkin, Ager, and Gilbertson-Day, editors. 2010. RMRS-GTR-235]. This risk assessment became known as the “first approximation,” and laid the groundwork for assessing wildfire risk to specific resources and assets (i.e., “values at risk”) across any size landscape. In the years since, the risk assessment process has been repeated many times at regional or local scales, and refined as a result. Improvements have been made to the wildfire simulation modeling used to generate burn probability and conditional fire intensity data. Spatial data for fuels and other inputs to the modeling process have been updated. Spatial data depicting various resources and assets ranging from infrastructure to wildland urban interface (WUI) to watersheds have been improved. In addition, there is an increasing oversight and regulatory need for the Forest Service and other land management agencies to provide up-to-date information on landscapes at risk of wildfire, and to use assessments of risk as the basis for decisions about budget allocations and fuels management priorities. All of these factors together point to the need for a new national-scale assessment of wildfire risk on National Forest System lands – a “second approximation.” This new assessment has been underway since 2013, and in this presentation I will provide an overview of the assessment and an update on its current status. Keywords: risk assessment, strategic wildfire management, United States, Forest Service Presenter Bio: Greg Dillon is a Spatial Fire Analyst with the Fire Modeling Institute, part of the USDA Forest Service’s Fire Sciences Lab in Missoula, Montana. His work generally involves geospatial and statistical analyses of large spatial datasets related to fire and fuels management on public lands. Among other things he is currently involved in spatial wildfire risk assessments at national and local scales. Greg’s previous work includes potential vegetation mapping for the LANDFIRE project and analysis of satellite-derived burn severity data. He has a BS in Geography from James Madison University and an MA in Geography from The University of Wyoming.

13. Perception and Management of Sociopolitical Risks on Large Fires Armando González-Cabán, Research Economist, USDA Forest Service, Pacific Southwest Research Station Additional Authors: Donald G. MacGregor, Principal Investigator, MacGregor Bates, Inc. Abstract: This work examines the perceived impact of sociopolitical factors on large fire decision making. The study is based on a set of 74 large fires in USDA Forest Service Regions 5 and 6 for the years 2009-2013. We interviewed 178 fire managers, including administrators of units where incidents occurred and staff members of incident management teams. A protocol was implemented to support a combination of information collection approaches, including interviews, surveys, and encoding of information from incident documentation sources. The information was collected at different times after incidents occurred. Participants were asked whether there was direct involvement from influential 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

individuals or groups in the incident management process. Their combined response suggests that about 50% of the time they were aware of direct involvement by influential individuals and influential groups. When queried whether or not they personally saw, heard or read media coverage associated to an incident at the time of the incident, 63.3% reported either that they had not or could not recall. Overall, respondents were somewhat aware of media reporting of incidents at the time of the incidents, and their knowledge of media reporting types covered a broad range of media pathways, including the Internet. When asked about the potential influence of media on incident decisions both in general terms and specific to the incident in question, most participants (57%) did not believe that media reporting influenced incident decisions. When asked specifically about what kind of actions, if any, they have taken to manage sociopolitical risks and objectives only 7.5% of them indicated taking some actions for management strategies, 17% for operational tactics and only 6.4% for incident objectives. In terms of the values at risk, sociopolitical factors may play a key role in fire management decisions. However, protection of values at risk at any given time and place on an incident can depend on the perceptions of sociopolitical actors who may have special interests in incident decisions and who as a result may influence those decisions. The present research suggests that such a potential does exist, though it may not be always and everywhere present. Keywords: Decision making, media influence, values at risk, wildfire management Presenter Bio: Dr. Armando González-Cabán is a Research Economist with the USDA Forest Service, Pacific Southwest Research Station, Riverside, CA since 1980. His areas of research interest includes fire economics, fire managers decision making process, valuation of market and nonmarket goods and services affected by wildfires, and climate change effects on wildfire risks. His work has been published several journals including the International Journal of Wildland Fire, Forest Science, Ecological Economics, Land Economics, Forest Policy and Economics, Journal of Environmental Management, and Journal of Forest Economics among others. He has traveled extensively throughout Latin America and Europe providing technical assistance and training.

14. Investigating temporal trends in wildfire hazard Jessica Haas, Rocky Mountain Research Station Additional Authors: Donald Long, Fire Ecologist, Rocky Mountain Research Station Abstract: Large scale disturbances alter the likelihood and intensity of wildland fires. These changes in wildfire hazard in turn alter the risk profile of a given area. While many wildfire risk assessments have focused on a snapshot of current wildfire risk across a landscape, none have looked at the temporal changes in risk through time. As a necessary first step in quantifying temporal changes in risk, the changes in wildfire hazard must be identified. I use a time series of landscape level fuel characteristics from 2000 – 2012 to investigate changes in wildfire hazard for the Southern Rocky Mountains in Colorado. This area has seen multiple large scale disturbances in this time period including prolonged drought, extensive insect related tree die-off and large wildland fires. By comparing estimates of wildfire likelihood and intensity across the time series, the temporal trends in the hazard profile is obtained and the drivers of this change can be identified. This information is critical for assessing expected future changes in wildfire hazard due to large scale disturbances such as climate change. Keywords: temporal trends; wildfire hazard assessment; Landfire; Disturbance Ecology Presenter Bio: Jessica R Haas is an ecologist with the Rocky Mountain Research Station. Her research focuses on the development of decision support tools for analyzing risk of various hazards to human communities and watersheds. Her work has been used by government, research and non-government organizations in a variety of wildfire and post-fire debris flow risk assessments, ranging from large national scale assessment, to local level project planning. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

15. Wildfire threat to residential structures in the Island Park Sustainable Fire Community Joe Scott, Senior Wildfire Consultant, Pyrologix Additional Authors: Matthew P. Thompson, Research Forester, USDA Forest Service Rocky Mountain Research Station Julie Gilbertson-Day, Spatial Wildfire Analyst, Pyrologix LLC Donald J. Helmbrecht, Wildland Fire Analyst, USDA Forest Service TEAMS Enterprise UnitTEAMS Abstract: We combined a comprehensive stochastic wildfire simulation (using FSim) with individual- building information on residential structure location and susceptibility to assess wildfire risk to residential structures in the Island Park Sustainable Fire Community in Idaho, USA. We calculated conditional and expected loss for each residential structure in the project area, and show how this information can be displayed in a simple graphical guide for stakeholders. For the project-area as a whole, we estimated the exceedance probability curve for the number of structures exposed to fire. We also generated risk-source maps, delineated the biophysical fireshed surrounding the community, and tabulated the number of structures exposed and affected as a function of distance from structure to ignition. This information can be used to identify which structures are most at risk, which components (e.g., probability, intensity or susceptibility) contribute the most to risk, and which mitigation activities might be most appropriate. This type of detailed analysis can help inform collaborative, local development of solutions to mitigate wildfire risk to human communities. Keywords: WUI, risk assessment, fireshed, exceedance probability Presenter Bio: Joe has led projects related to surface and canopy fuel characteristics, wildfire behavior modeling, crown fire hazard assessment, and wildfire risk assessment. Mr. Scott is the lead developer of NEXUS; lead developer of FireWords; co-developer of FuelCalc; the lead developer of a set of standard fire behavior fuel models for national application; and the lead author of a report detailing a quantitative wildfire risk assessment framework. Mr. Scott's current work focuses on the application of Monte Carlo wildfire simulations to wildfire risk assessment and land management planning for private enterprises and local, state and federal government agencies.

16. Impact Oriented Fire Paths Joaquín Ramirez, Technosylva Additional Authors: Santiago Monedero Abstract: The threat of a wildfire depends on the expected economic and environmental impact that it may cause to nearby valuable assets, the risk of human lives, and the probability of reaching areas where the fire may get out of suppression control. Fire planning should balance these spatially distributed threats in order to prioritize suppression actions to a fire. In this context, it is desirable to be able to spatially visualize the contribution of different paths of the fire (paths perpendicular to the isochrones) to certain impact or risk variables. We present a way to do this by adding a filtering condition to the well known output Minimum Travel Time Fire and using a Montecarlo approach to take into account small deviations in weather data. The resulting output may be used to represent the paths leading to a greater economic or environmental impact, but could also be used for identifying the paths leading to regions where the fire is expected to be out of suppression capacity, crown fire transition is likely, or regions where there is a Campbell alignment of forces. Keywords: Fire Paths, Impact, Risk Presenter Bio: Dr. Joaquín Ramirez is the Chief Technology Officer and President of Technosylva. Joaquin is well respected as a leading fire scientist and software architect in Europe and North America, and is the chief designer of the Wildfire Analyst™ and fiRESPONSE™ software products.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

17. Trends and thresholds in fire behavior across Yellowstone’s young lodgepole pine forests Kellen N Nelson, PhD Candidate, Program in Ecology and Department of Botany, University of Wyoming Additional Authors: Monica G. Turner, Professor, University of Wisconsin William H. Romme, Professor-emeritus, Colorado State University Daniel B. Tinker, Associate Professor, University of Wyoming Abstract: The extent of early-seral forests is increasing across western landscapes due to a recent escalation in fire occurrence, severity, and extent—a trend that is projected to continue over the next century. Young, post-fire lodgepole pine stands cover a large proportion of the Greater Yellowstone Ecosystem as a result of the extensive 1988 Yellowstone Fires. Observed fire behavior in this vegetation type by fire monitoring programs indicate that rates of spread and intensity are frequently low, but can become extreme under high winds and low fuel moisture. We address uncertainty in the conditions and thresholds that regulate this shift in fire behavior by integrating fire observations, fuel measurements collected in post-fire stands across Yellowstone National Park, and parameterizing and calibrating fire models. We aggregated 83 fuel profiles collected across 24-year-old post-fire stands into groups of similar fuel loads using a cluster analysis—two fine surface fuel groups, two 1000-hr fuel groups, and two canopy fuel groups resulted. Fine surface fuels are used to estimate rate of spread using the BehavePlus software program and results are calibrated with ~30 fire observations provided by land managers in the greater Yellowstone region. 1000-hr fuel loads are incorporated into fire intensity estimates using Byram’s 1959 model. Model parameters—fuel consumption and rate of spread—are estimated by incorporating 1000-hr fuels with fine fuels in the First Order Fire Effects Model and using the calibrated BehavePlus rate of spread results, respectively. The conditions under which passive, active, and independent crown fires may occur are estimated using van Wagner’s 1977 and Rothermel’s 1991 crown fire models. Fire weather is systematically varied between 70th and 98th percentile weather conditions to identify thresholds in fire behavior for each fuel group combination. Using the results from this modeling exercise, an evaluation of fire behavior and fire weather thresholds in young lodgepole pine forests will be presented. Results will include a calibrated surface fire rate of spread model, fire intensity estimates, and crown fire potentials across a range of fire weather conditions. Keywords: fire behavior, young forests, 1988 Yellowstone Fires, lodgepole pine, climate change Presenter Bio: Kellen Nelson is a forest disturbance ecologist and PhD candidate at the University of Wyoming. His research focuses on how disturbance processes interaction with vegetation and the environment. Current investigations include the successional dynamics of fuel loads and fire behavior in post-fire, lodgepole pine stands and the spatial and temporal patterns of fuels and fuel moisture content in contrasting young, mature, and mountain pine beetle-affected lodgepole pine stands.

18. Fuels and Fire Behaviour in New Zealand Wilding Conifers Tara Strand, Atmospheric Scientist, Scion, New Zealand Crown Research Institute Additional Authors: Grant Pearce, Fire Scientist, Scion, NZ CRI Veronica Clifford, Fire Scientist, Scion, NZ CRI Marwan Katurji, Lecturer in Meteorology (Mechanics and Thermodynamics), University of Canterbury Daisuke Seto, PhD Student, University of Canterbury Richard Parker, Rural Fire Research Leader, Scion NZ CRI Abstract: “Wildings” is a term used in New Zealand for the unintended spread of introduced trees across the landscape. Many introduced plantation conifer species grow well in New Zealand, which has no native conifers of its own, and provide a seed source for wilding trees. Wildings have reached “weed” 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

status, changing the fire hazard in remote rural areas from a New Zealand fuel based fire hazard to an exotic fuel fire hazard. In addition, fuel loads and distribution are altered by the chemical control of wilding pines which results first in reddening of the needles as the trees die, followed by needle drop and eventually accumulation of downed large woody debris. Recent wildfires have shown that existing New Zealand forest fire behaviour models do not accurately predict fuel loads or fire and smoke behaviour in these wilding fuel types. To address this issue, research burn plots were established in young wilding Pinus contorta (lodgepole pine) trees in blocks of chemically treated and untreated wildings. These experimental burns were the most heavily instrumented ever in New Zealand – with measurements of fire spread rates, in-fire temperatures, fire turbulence and smoke, fuels, weather and upper atmosphere conditions – in many cases, using newly developed equipment and techniques. Data collected provide information on wilding fuel loadings (treated and untreated), fire turbulence, smoke behaviour, and fire spread rates for this mix of high country native bunchgrass-wilding lodgepole pine fuel type and will be used to develop improved fire behaviour models and prediction tools for rural fire managers. This presentation will outline the background and objectives of this research and describe measurement methods, results to date on fuel loads, fire and smoke behaviour, and turbulence. The presentation will also describe future plans for further experiments in more mature wilding conifers. Keywords: smoke, turbulence, wilding pines, chemically treated pine Presenter Bio: Dr Tara Strand specialises in measuring turbulence in complex environments and modelling the dispersion of gas and particles in the atmosphere, such as smoke plumes near a fire, aerosol drift over a forest canopy, and pollutant plumes in the urban environment. She wears two hats, she is the Pest Management Research Leader at Scion, New Zealand's Crown Research Institute and she is a member of Scion’s Rural Fire Research Group, where she studies in-fire turbulence and works to create smoke modelling tools applicable to the New Zealand Fire environment.

19. Using McArthur Model To Predict Bushfire Prone Areas In New South Wales Liran Sun, UNSW Australia Additional Authors: John Trinder, Professor, UNSW Australia Chris Rizos, Professor, UNSW Australia Abstract: Fire intensity is a key component of bushfire regimes and its assessment has been widely used for mapping of bushfire prone areas. This paper will examine the potential of a spatial model of fire intensity, which is based on McArthur forest fire behaviour model and Byram’s fire-line intensity, to detect and map state-wide bushfire prone areas in the state of New South Wales. The potential fire-line intensity was calculated from spatial inputs of total fuel load, the McArthur Forest Fire Danger Index (FFDI) and terrain slope. The “Overall fuel hazard Assessment Guideline” was used to estimate the potential fuel loads for each vegetation hazard class for a state wide vegetation map showing the distributions of vegetation classes. FFDI, as an index to represent a range of weather variables, is also an important variable affecting the estimation of fire intensity. In this research the weather data from 16 stations across New South Wales covering the period June 1972 to December 2009 were used to calculate the spatial distribution of FFDI. The indicator of fire weather conditions was assessed by calculating the return period (1, 5, 20, 50 and 100 years, respectively) of extreme values through fitting extreme value distributions to records of FFDI. Slope mapping was created from 30 m Smoothed Digital Elevation Mode (DEMs) of SRTM V2, which has been smoothed to reduce noise to improve the representation of surface shape compared with the original version. The maximum slope was used to simulate the possible influence of fire spread rate on fire intensity. However, the normal approaches to calculating maximum slope are based on Zevenbergen & Thorne’s formula, which is not suitable for 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

estimating local conditions. Therefore a GDAL code was developed to calculate the maximum slope between each pixel and its eight neighbouring pixels. To assess the reliability of predicted bushfire prone areas, bushfire hazard based on the potential fire intensity was classified and compared with the high fire severity class from historical observations of burn data with a resulting overall accuracy of 72%. The reliability is highly correlated with the interpretation of vegetation mapping. This research should be useful for understanding bushfire behaviour and for creating policies for bushfire risk management. Keywords: fire intensity, McArthur model, FFDI, fuel load, maximum slope Presenter Bio: Liran Sun is currently a master student in school of civil and environmental engineering at the University of New South Wales Australia. She received her bachelor’s and master’s degree in Geography Information System from Beijing Normal University. She served at CNIC CAS as a remote sensing engineer from 2009 to 2014. Her research interests include remote sensing application, bushfire behaviour research and bushfire risk assessment.

20. An experimental study of the stochastic nature of firebrand flight Ali Tohidi, PhD Candidate, Clemson University Additional Authors: Nigel Berkeley Kaye, Associate Professor of Civil Engineering, Clemson University Abstract: Detrimental effects of climate change such as global warming will increase the frequency of occurrence of large wildfires. This escalates the risk of fire spotting to people, properties, and infrastructure, particularly in wild-land urban interfaces (WUI). Through fire spotting, firebrands are lofted into the atmosphere by the fire plume, and then transported downwind by the atmospheric boundary layer and land where they may ignite a spot fire. In order to alleviate the risks from fire spotting, it is crucial to understand the behavior of firebrands in flight. However, due to the stochastic nature of the problem and its large scale, there is virtually no experimental verification for existing models. A series of firebrand free-fall experiments in a no-wind environment are conducted to quantitatively evaluate performance of the current firebrand transport models. Radial travel distances are recorded and results are compared with corresponding Mont-Carlo type free-fall simulations. Comparison of PDFs and CDFs of the results indicate that a deterministic six degrees of freedom transport model with steady aerodynamic force and moment coefficients is capable of predicting the statistical properties of firebrand flight. Further wind tunnel experiments of firebrand lofting and transport processes are conducted in which non-combusting polyurethane rod-like firebrand models with different aspect ratios are released multiple times through various velocity fields induced by the interaction of the boundary layer and an air jet that models the fire plume behavior. Firebrand models are released from the center of the jet with zero initial velocity and with randomly varied initial release angles at least 400 times per aspect ratio / velocity field combination. Flight trajectories of firebrands are captured digitally. We present maximum lofting height, and maximum downwind transport distance of the firebrand models for each experimental set. Probability distribution functions of longitudinal and lateral firebrand dispersion are presented in terms of the relative distance from the release point. This work is in progress with plans for further validating firebrand flight models through corresponding Monte-Carlo simulations of the wind tunnel experiments and a parametric study of the fire spotting phenomenon using our validated model. Keywords: Firebrand, Firebrand Transport, Fire Spotting, Wind Tunnel, Experiments, Monte-Carlo Simulations Presenter Bio: Ali Tohidi is a PhD candidate of Civil Engineering at Clemson University. He received his B.S. from IAUCTB, Tehran, IRAN in 2009. He continued his studies at Sharif University of Technology where he obtained his M.Sc. in Civil Engineering at Hydraulic/Structures division on 2011. He’s worked on wind induced gravity currents in aquatic canopy zones. He, then, moved to Clemson, SC and his current research focus is on wildfire spread, firebrand properties & flight characterization, and fire 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

plume behavior. His corresponding research thrusts are Environmental Fluid Dynamics, CFD, Turbulence, Finite Element Modeling, Image/Signal Processing, Particle Tracking, Data-Driven Modeling, and Spectral Analysis.

21. The Frequency in the Flames: Acoustic Impulse Events Generated by Wildland Fire Fuels Kara Yedinak, Postdoctoral Fellow, University of Idaho Additional Authors: Additional Authors: Michael Anderson, Professor of Mechanical Engineering, University of Idaho Kent Apostol, University of Arizona Leslie Fowler, Undergraduate Research Assistant, University of Idaho Sally Mei, Undergraduate Research Assistant, University of Idaho Alistair Smith, Associate Professor of Forest, Rangeland, and Fire Sciences, University of Idaho Abstract: Accurately characterizing the physical properties of wildland fires is essential to quantifying their impacts on atmospheric chemistry and physics as well as human health and livelihoods. However, due to the highly variable nature of combustion dynamics within wildland fires, there are significant challenges in quantifying fire behavior metrics of these phenomena. Current quantification methods rely heavily on temperature measurements and observations in the visible and infrared regions of the electromagnetic spectrum. An energetic source of information that has been largely untapped is the sound, or acoustics, emanating from fires. Acoustic impulse events have long been used as diagnostics for discrete phenomena in the natural world, including the detection of meteor impacts and volcanic eruptions. Wildland fires display an array of such acoustic impulse events including, but not limited to crackling noises or “pops”. Our research has shown that the nature of “pops”, i.e., acoustic impulse events during combustion, is affected by, among other possible factors, the species and moisture content of the vegetation being burned. We present the results of acoustics experiments involving fires fueled by dead pine needles and live conifer seedlings burned in a laboratory setting. The seedlings were of two species, and were subjected to differing levels of water stress during their growth. Individual “pops”, attributed to the combustion of these seedlings were acoustically isolated for analysis. A statistical analysis of the acoustic spectra of individual pops revealed a surprising degree of correlation in the spectra related to species and water stress level in the frequency range of 11-16 kHz. Our results open the door to the development of digital acoustic filters able to distinguish specific vegetation properties of fuels involved in wildland fire events. Keywords: acoustics, wildland fire behavior, acoustic impulse events, combustion acoustics Presenter Bio: Kara Yedinak is a postdoctoral research fellow in the Department of Forests, Rangelands, and Fire Sciences at the University of Idaho. Kara received her BS in Physics from Pacific University (2002) where she studied low dimensional chaos and fluid flow. She worked as a fire behavior science research technician at the Fire Sciences Laboratory in Missoula, Montana (2004 to 2007). In 2013, Kara completed her PhD at the Laboratory for Atmospheric Research at Washington State University studying coupled atmosphere-fire behavior interactions using simulations and field observations. Currently Kara is focused on investigating concepts surrounding wildland fire propagation theory and acoustics.

22. Exploratory analysis of interactions of patchy/clumpy fuel configurations on fire behavior with a physics-based fire model Francois Pimont, National Institute for Agricultural Research (France) Additional Authors: Russell Parsons, Research Ecologist, USFS Fire Sciences Lab Rod Linn, Los Alamos National Laboratory Jeremy Sauer, Los Alamos National Laboratory 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Judy Wintercamp, Los Alamos National Laboratory Chad Hoffman, Colorado State University Abstract: Fuels management strategies increasingly suggest that incorporation of spatial heterogeneity in fuel treatments can lead to more resilient, diverse and healthy forest ecosystems. However, current operational fire models are limited in their capability to represent fuel heterogeneity, particularly at fine scales, or to adequately capture how such heterogeneity may affect fire behavior over a range of wind conditions. Consequently, managers face uncertainty regarding how well patchy/clumpy fuel configurations will affect fire behavior. In this study, we used the physics-based fire behavior model, FIRETEC, to explore interactions between different levels of canopy fuel aggregation, canopy cover, and wind speed. As little is yet known regarding how these factors interact, we chose an exploratory approach, spanning each factor broadly. We found that patchy/clumpy fuel configurations significantly altered fire behavior, but these effects varied greatly with different wind speeds and with canopy cover. Higher canopy cover reduced windspeed profiles due to drag effects, but had higher rates of spread, intensity, and lateral spread rates. While larger clumps produced the highest variability in fire behavior,this variability decreased with increasing windspeed. Clump effects were most pronounced at moderate wind speeds. Our results suggest patchy/clumpy fuel patterns produce complex interactions between the fire, fuel and atmosphere. Keywords: Fire behavior, spatial fuel patterns, patchy, CFD, FIRETEC Presenter Bio: François Pimont holds an Engineer-Scientist position in the Mediterranean Forest Research Unit (URFM, Avignon) of the National Institute for Agricultural Research (INRA, France). He is an engineer from Ecole Polytechnique and AgroParisTech and obtained his PhD thesis of Environmental Sciences in 2008. His domains of experience and expertise are wildfire physics and wildfire fire behaviour modelling, as well as fuel modelling and fuel remote sensing. Since 2005, he has developed a close collaboration with LANL on fire modelling and more recently with USFS (RMRS) on fuel modelling. He obtained the Silver Medal of French Academy of Agriculture in 2011.

23. Managing Fire in the Only EPA Declared Public Health Emergency in America Nikia Hernandez, District Fire Management Officer, USDA Forest Service Abstract: The Libby Superfund in Northwest Montana is the first and only public health emergency declared by EPA under CERCLA law in American history. Vermiculite was mined from a large open-pit mine 5 miles northeast of Libby, MT. The vermiculite was contaminated by a minerologically unique amphibole asbestos referred to as Libby Amphibole (LA). Historic mining, processing, and community utilization of vermiculite are known to have resulted in the release of LA into the environment causing asbestos related diseases to sicken 1000s of people and directly contributing to 400+ deaths. In 2000, EPA began emergency removal actions to clean-up residential and commercial properties. The contaminated area around the mine and local communities consists of about 80,000 acres and was listed on the Superfund National Priority List in 2002. Through data collection and activity based sampling EPA determined the forested area surrounding the former vermiculite mine had LA contamination in the tree bark, duff, and soil. Thus, the Forest Service designated an area of about 30,000 acres surrounding the mine as a unique fire management unit on the Kootenai National Forest. The Forest Service has primary wildland fire protection responsibilities of the area. In order to provide for firefighter safety, no ground firefighting resources are allowed in the area without powered air purifying respirators and without going through decontamination after firefighting activities. This is the only known circumstances within the US that wildland firefighters are required to wear respirators when engaging in wildland firefighting operations. Due to a lack of data there is significant concern by EPA and Forest Service that a large wildland fire could pose a human health risk to the public and firefighters by transporting asbestos in the smoke causing asbestos exposure through the ambient air and redeposition of asbestos to areas already cleaned by EPA. The Kootenai National Forest is managing this unique 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

situation by utilizing aviation resources for aggressive initial attack. Ground wildland firefighters operate on a volunteer basis for duty in the contaminated area. Agencies continue to assess levels of contamination and potential exposures to firefighters and the public to address this long-term fire management issue. Keywords: Libby, Superfund, Fire Management, Asbestos, Public and Firefighter Safety Presenter Bio: Nikia Hernandez is the District Fire Management Officer on the Libby District of the Kootenai National Forest in Northwest Montana. He began his wildland firefighting career with the Montana Department of Natural Resources. He smokejumped in Missoula, Montana while attending the University. After completing his Bachelor of Science in Forestry he joined the Kootenai National Forest as a Fuels Specialist prior to becoming the District Fire Management Officer. He is active as a Division Group Supervisor on a Type 2 Incident Management Team and has fought fire from Alaska to Florida over his 23-year career in fire management.

24. When there’s Fire there’s Smoke: Linking Wildfire to Distant Urban Airsheds. A 5 Year Health Economic Assessment of the Western US, 2010-2014. Benjamin Jones, PhD Candidate, University of New Mexico Additional Authors: Robert P. Berrens, Professor of Economics, University of New Mexico Abstract: There is concern that wildfire smoke is increasingly affecting urban airsheds distant from the flame zone, particularly in the Western US. Climate change, drought, and continued fuels build-up are expected to increase the magnitude and severity of Western wildfire smoke events. However, little is known about the aggregate health costs of smoke across this geographically expansive and increasingly, highly urbanized, region. This is particularly troublesome in light of recent and ongoing discussions surrounding wildland fire management, broadly defined, and mitigation budgets, in which a complete picture of the full costs of wildfire is presently lacking.

To address this knowledge gap, this study provides the first regional time series estimates of morbidity incidence and associated costs of wildfire smoke across the 18 largest metropolitan areas (populations greater than 750,000) in the Western US over 2010-2014. A benefit transfer methodology is applied using the US EPA Benefits Mapping and Analysis Program – Community Edition (BenMAP-CE). Wildfire- specific incidence functions and willingness to pay measures are used to reduce potential bias. Results suggest that the health costs of wildfire smoke in the Western US are substantial. There is significant spatial and temporal variability in incidence and costs across the landscape that is only marginally associated with the size of the burn area. Costs are increasing at a rate of 7-15% annually, on average, though are highly variable depending on the severity and duration of the smoke event. This research establishes the first trend and reference point of the magnitude of economic costs of health impacts associated with large-scale airshed events precipitated by wildfire smoke. Given the ongoing implementation challenge of finding sustainable mitigation funding mechanisms, a policy implication of this research is to expand the scope of potential funding sources to communities, distant from the flame zone, who nonetheless can be significantly impacted by wildfire. Individuals in Western communities, consistent with their economic preferences, may be willing to pay for distant mitigation efforts in order to reduce the variability and severity of disruptions to their airsheds caused by wildfire smoke. Keywords: wildfire smoke; health impacts; economic costs; Western US Presenter Bio: Benjamin Jones is a PhD Candidate in the Department of Economics at the University of New Mexico and Doctoral Fellow of the Robert Wood Johnson Center for Health Policy. His research is at the nexus of environmental and health economics, where he studies the impacts of shocks to the natural environment. Jones' recent research has looked at wildfire smoke health impacts and costs, 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

health costs of uranium mining, optimal management of invasive species in the presence of health externalities, and non-market values surrounding re-purposing of coupled human and natural systems.

25. Smoke in the City: How Often and Where Does Smoke Impact Summertime Ozone in the United States? Steven Brey, Colorado State University Additional Authors: Emily V. Fischer, Assistant Professor, Colorado State University Abstract: The mechanisms and magnitude of the contribution of fires to ozone (O3) production is poorly understood. In this work we investigate the influence of fire on O3 abundances over the contiguous United States. Using co-located observations of particulate matter (PM2.5) and the National Weather Service Hazard Mapping System smoke data, we identify summertime days between 2005 and 2014 that Environmental Protection Agency (EPA) Air Quality System (AQS) O3 monitors are influenced by smoke. We compare O3 mixing ratio distributions for smoke-free and smoke-impacted days for each monitor, while controlling for temperature. We observe that the mean O3 abundance measured on smoke- impacted days is higher than on smoke-free days. The magnitude of the effect varies by location with a range of 0 to 37 parts per billion by volume (ppbv). We find that smoke is present on a non-negligible proportion of days when the 8-hour average O3 mixing ratio exceeds the EPA limit of 75 ppbv in regions and locations with significant O3 issues, including the Northeast urban corridor, Dallas, Houston, Atlanta, Birmingham, and Kansas City. Most U.S. cities maintain a similar proportion of smoke-impacted exceedance days when they are held against the new more stringent limit of 70 ppbv for an 8-hour average. We show that smoke-impacted O3 mixing ratios are most elevated in U.S. cities with the highest emissions of nitrogen oxides (NOx). Keywords: Ozone, PM2.5, smoke, EPA, AQS, ozone production Presenter Bio: Steven Brey is a graduate student studying atmospheric chemistry at Colorado State University. His primary research interests are the impacts wildfire smoke has on air quality and tropospheric composition. Before starting work at Colorado State, Steven worked for Mazama Science, a small consulting company in Seattle.

26. Impact of wildfires on regional air pollution Alexandra Larsen, PhD Student, North Carolina State University, Department of Statistics Additional Authors: Ana Rappold, NCSU Brian Reich, Associate Professor, Department of Statistics, NCSU Mark Ruminki, NOAA Abstract: We examine the impact of wildfires and agricultural/prescribed burning on regional air pollution and Air Quality Index (AQI) between 2006 and 2013. We define daily regional air pollution using monitoring sites for ozone, PM2.5 collected by Federal Reference Method, and constituents of PM2.5 from the Interagency Monitoring of PROtected Visual Environment (IMPROVE) network and use satellite image analysis from the NOAA Hazard Mapping System (HMS) to determine days on which visible smoke plumes are detected in the vertical column of the monitoring site. To examine the impact of smoke from these fires on regional air pollution we use a two stage approach, accounting for within site (1st stage) and between site (2nd stage) variations. At the first stage we estimate a monitor-specific plume day effect describing the relative change in pollutant concentrations on the days impacted by smoke plume while accounting for confounding effects of season and temperature. At the second stage we combine monitor-specific plume day effects with a Bayesian hierarchical model and estimate a pooled nationally-averaged effect. HMS visible smoke plumes were detected on 6% of ozone, 8% of 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

PM2.5 and 6% of IMPROVE network monitoring days. Our preliminary results indicate that the long range transport of air pollutants from wildfires and planned burns increase ozone concentration by 11% and PM2.5 mass by 34%. On all of the days where monitoring sites were AQI code Green for ozone, 6% of those days experienced smoke plume cover. We observed 18% plume coverage on code Yellow (‘moderate’) days, 26% on code Orange (“unhealthy for sensitive groups”) days, 30% on code Red (“unhealthy”) days and 29% on code Purple (“very unhealthy”) days. Similarly, for PM2.5, we observed plume coverage on 4% of code Green days, 11% on code Yellow days, 16% on code Orange days, 16% on code Red days and 50% on code Purple days. Our preliminary results suggest that smoke from wildfires and prescribed burns has a substantial effect on regional air quality and accounts for a large percentage of days with unhealthy AQI levels. This work does not necessarily represent EPA views or policy. Keywords: Air Quality, Ozone, PM2.5 Presenter Bio: Alexandra Larsen is a 3rd year PhD student in the North Carolina State University Department of Statistics. Her current research interests are in Bayesian and spatial statistics with applications to environmental areas such as air pollution. Her dissertation focuses on the impacts of wildfires on air quality.

27. Sensor Messaging: Guidance for Interpretation of Short-Term Concentration Readings Susan Stone, Senior Environmental Health Scientist, U.S. Environmental Protection Agency Additional Authors: Kristen Benedict, Physical Scientist, U.S. EPA Elizabeth Mannshardt, Statistician, U.S. EPA Michael Stewart, Environmental Protection Specialist, U.S. EPA Abstract: Emerging technologies such as air sensors are providing individuals and communities with real-time, short-term (e.g. 1-minute) concentration readings of various pollutants. Instantaneous access to local pollutant information from instruments of varying or unknown quality is presenting a real challenge in the interpretation and communication of the data. Because health effects studies generally do not exist to support interpretation of short term exposures, EPA is developing guidance to help individuals understand how to use these readings to inform outdoor activities. This presentation focuses on the analysis and communication of 1-minute ozone (O3) and fine particulate (PM2.5) sensor data collected from four EPA Village Green benches. The data is examined to understand how real-time data compares to nearby Federal Reference or Equivalent monitors that average data over longer time periods consistent with National Ambient Air Quality Standards. The findings are used to pilot new behavioral messaging for O3 and PM2.5 - distinct from the Air Quality Index (AQI) health effects messaging. EPA hopes this guidance new communication tools will provide manufacturers of emerging devices with a consistent way of presenting the data to the public. This is also important and useful in understanding concentrations and health effects from smoke events, particularly given that the concentrations from these events are usually high and highly variable in time and space. The purpose of this presentation is to inform members of the wildfire community about this guidance, and to solicit feedback to improve its usefulness in reducing exposure and risk in wildfire situations. Keywords: Air quality sensors, Short-term (1-minute) air quality data, Behavioral responses and guidance, Air Quality Index Presenter Bio: Susan Stone is a Senior Environmental Health Scientist with EPA, and is the leader of the team reviewing the national ambient air quality standard for ozone. Ms. Stone is also the Air Quality Index (AQI) team leader, has coauthored many of EPA’s public information documents about the AQI, the health effects of criteria pollutants, and the multi-agency document Wildfire: A Guide for Public Health Officials. She is an author of two studies of the health impacts of smoke from a fire in Eastern 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

North Carolina and one wood smoke controlled human exposure study. Susan Stone has an M.S. from the School of Public Health at the University of North Carolina at Chapel Hill.

28. Landscaping with Ornamental Trees and Exterior Structure Features using EcoSmart Fire Model Mark Dietenberger, Research General Engineer, USDA Forest Service - Forest Products Lab Additional Authors: Charles Boardman, Physical Science Engineer, USDA Forest Service - Forest Products Lab Abstract: EcoSmartFire is a Windows program that models heat damage and piloted ignition of structures from radiant exposure to discrete landscaped tree fires. It calculates the radiant heat transfer from cylindrical shaped fires to the walls and roof of the structure while accounting for radiation shadowing, attenuation, and ground reflections. After calibrating the model by comparison with selected fire tests in a previous work, the PC version was used to examine fire and damage risk predictions for sensitivity to relevant parameters and geometric configurations, which was limiting for the on-line version. The parameters that have been varied are structure cladding, ground reflection coefficient, ornamental vegetation moisture content, and weather. The geometric configurations that have been varied are structural features such as presence of decking, corner walls, and roofing shapes and vegetation features such as a line of trees or tree heights, particularly for testing the radiation shadowing and flaming attenuation algorithms. This exercise of the model has increased its robustness and appears to be reasonable. Despite this level of testing, model upgrades and further experimental comparisons are recommended. Keywords: Clearance Zones, Home Ignitability, Structure Hardening Presenter Bio: Dr. Dietenberger is a research general engineer researching in wood fire safety and building moisture with an emphasis in computer modeling development and properties evaluation in specialized tests.

29. Setting Wildfire Evacuation Triggers by Coupling Fire and Traffic Simulation Models Dapeng Li, Research Assistant, University of Utah Additional Authors: Thomas J. Cova, Professor, University of Utah Philip E. Dennison, Professor, University of Utah Abstract: In the American West, wildfires pose a significant risk to the fire-prone communities in the Wildland-Urban Interface (WUI). When a fire approaches a community and becomes a threat, the incident commander (IC) may need to evacuate some residents to ensure their safety. In wildfire evacuation practices, prominent geographic features such as ridges, roads, and rivers are common trigger points, and when a fire crosses a feature, an evacuation order will be issued to the residents at risk. This work improves current approaches to trigger modeling by coupling fire and traffic simulation models to set evacuation triggers. From a coupled human-environmental system (CHES) perspective, fire spread is an environmental system, while the evacuation to the threatened population is a human system. We use traffic simulation to model the evacuation traffic and estimate the evacuation time, which can be used as the input for fire simulation to create a trigger buffer around the community. These estimated evacuation times are aggregated to create probability-based trigger buffers. In this way, trigger modeling takes into account both predicted fire spread and evacuation traffic to better support IC decision making during wildfire evacuations. A case study in Julian, California was conducted to evaluate the proposed method, and the results reveal that the method effectively couples the two systems and could improve the applicability of trigger modeling. Keywords: wildfire evacuation, coupled human-environmental system, trigger modeling, wildfire simulation, traffic simulation, GIS, uncertainty

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Presenter Bio: Dapeng Li is a Ph.D. candidate in the Department of Geography at the University of Utah. He earned a Bachelor's Degree in Geographic Information Systems (GIS) from China University of Geosciences (Beijing) and a Master's Degree in Cartography and GIS from Peking University in China. His research interests include GIS, wildfire spread and evacuation modeling, emergency management, and transportation. His dissertation research focuses on employing fire spread modeling, trigger modeling, reverse geocoding, and traffic simulation to model wildfire evacuation as a coupled human- environmental system.

30. Coupling the human and biophysical dimensions of wildfire to better understand wildfire risk and risk mitigation Max Nielsen-Pincus, Assistant Professor, Portland State University Additional Authors: Alan Ager, Research Forester, Rocky Mountain Research Station, USDA Forest Service Travis Paveglio, Assistant Professor, University of Idaho Abstract: The United States’ (US) National Cohesive Wildland Fire Management Strategy (Cohesive Strategy) recognizes creating fire-adapted communities and maintaining resilient landscapes as critical 21st century challenges for community leaders and fuels managers alike. Increasingly, human decisions about where to live and play intersect with ecological processes like wildfire, straining government capacities protect homes and other infrastructure while simultaneously increasing the need for effective planning to mitigate wildfire risks. Although, the Cohesive Strategy recognizes this intersection, integrating social and ecological understanding of wildfire into risk planning and management at a meaningful scale is more challenging. For example, the 2013 West Wide Wildfire Risk Assessment concluded that an analysis of community wildfire risk was not possible because of the lack of definition of community. In contrast, locally crafted community wildfire protection plans often acknowledge local social capacity to address wildfire risk, but can lack the scale and complexity at which wildfire events occur. We hypothesize that a critical function of fire-adapted communities is the ability to couple social and ecological processes together and adapt that understanding to the local community context. In this talk, we report on a project that coupled the social and ecological dimensions of wildfire in 50 western US communities to better understand wildfire risk and the potential for risk mitigation. We used data from structured interviews with key informants (n=114) and from wildfire simulation modeling to examine the drivers of community exposure to wildfire and community capacity to manage wildfire risk. Using regression modeling and ensemble methods for classification trees, we found that community social dimensions and landscape composition were important predictors of wildfire exposure across our study communities. Furthermore, we found that key informant perceptions of risk mitigation effectiveness were predicted by a number of social and landscape factors. The results demonstrate the importance of representing both social and ecological dimensions in furthering wildfire policy and management in the western US and elsewhere. Keywords: Communities, Key Informants, Adaptive Capacity, Simulation Modeling, Landscape Context, Housing Exposure Presenter Bio: Dr. Nielsen-Pincus is an assistant professor of environmental science and management at Portland State University’s School of the Environment. His research on social-ecological systems includes a focus on wildfire social science. He has contributed to and led projects related to landowner fuels management behaviors, community-based wildfire risk mitigation planning, and the effects of large wildfires on local labor markets. Dr. Nielsen-Pincus teaches graduate and undergraduate students in the Department of Environmental Science and Management about environmental management and policy. Prior to joining the faculty at Portland State University, Dr. Nielsen-Pincus was a faculty research associate at the University of Oregon.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

31. Wildland/Urban Interface: U.S. Fire Department Wildfire Preparedness and Readiness Capabilities Michele Steinberg, Division Manager for Wildland Fire Operations,National Fire Protection Assocation (NFPA) Additional Authors: Angela Garcia, Associate Professor, Bentley University Rachel Madsen, Ph.D. Candidate, Brandeis University Abstract: Understanding local fire department levels of preparedness and readiness in the event of a wildfire is very important. Many of the 30, 052 local fire departments inventoried in NFPA’s survey serve the 72,681 communities at risk to wildfire in the United States. Local fire departments are normally the initial responders to a wildfire and are often overwhelmed when a wildfire goes beyond the initial attack period. Due to limited resources, local fire departments are beginning to address the problem differently by focusing on the susceptibility of structures to the inevitability of wildfire exposure. Learn about the results of NFPA research on fire departments located in the wildland/urban interface in the United States. Keywords: Public safety, wildfire response, community risk reduction, relationship building, cooperation, training, preparedness, mitigation Presenter Bio: Michele Steinberg is the Division Manager for Wildland Fire Operations at the National Fire Protection Association (NFPA), where she leads a team dedicated to wildfire safety education, advocacy and outreach. NFPA is a global nonprofit organization established in 1896 and devoted to eliminating death, injury, property and economic loss due to fire, electrical and related hazards. Michele has worked in disaster safety arena for more than 25 years. Current NFPA wildfire safety initiatives include the Firewise Communities/USA® Recognition Program, home wildfire risk evaluation seminars, Wildfire Community Preparedness Day, and TakeAction, focused on preparedness for youth and families.

32. The effect of static stability on the atmospheric response to a wildland fire Joseph Charney, Research Meteorologist, U.S. Forest Service Additional Authors: Daniel Keyser, Professor, University at Albany Brian Potter, Research Meteorologist, U.S. Forest Service Alan Srock, Assistant Professor, St. Cloud State University Julia Ruthford, NOAA/NWS/WFO Anchorage, AK Scott Goodrick, Research Meteorologist, U.S. Forest Service Warren Heilman, Research Meteorologist, U.S. Forest Service Abstract: It is widely assumed by the fire management community that the static stability profile above a wildland fire influences the strength of the updraft above the fire, which in turn affects the intensity of the fire, the horizontal winds in the vicinity of the fire, and the fire spread. However, little work has been done to test these widely held assumptions. This study employs an idealized two-dimensional numerical model to assess the effect of static stability on the atmospheric response to a wildland fire. Static stability is specified in the model by fixing the profile of potential temperature, while the fire is specified by applying a time-invariant heat flux at the ground. Our goal is to determine whether altering the profile of potential temperature will modify the resultant updraft above the fire and the horizontal winds in the vicinity of the fire such that those modifications could affect the intensity of the fire and the fire spread, were the specified fire allowed to react to changes in the updraft and horizontal winds.

We have identified two types of atmospheric response on which we will focus our investigation. The first type of atmospheric response occurs when altering the profile of potential temperature modifies 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

the vertical velocity (w-wind hereafter) in the atmosphere above the fire. The w-wind is hypothesized to be the primary factor in our simulations that could affect the intensity of the fire. The second type of atmospheric response occurs when altering the profile of potential temperature modifies the near- surface horizontal velocity (u-wind hereafter) in the vicinity of the fire. The u-wind is hypothesized to be the primary factor in our simulations that could affect the fire spread. The results of this study will establish a benchmark for the atmospheric response to a wildland fire in an idealized two-dimensional numerical model when the fire is specified by applying a time-invariant heat flux at the ground. This benchmark will be used to inform a more general investigation that uses a coupled fire-atmosphere model with a fire parameterization that permits two-way interactions between the fire and the atmosphere. Keywords: stability, fire behavior, fire modeling Presenter Bio: Dr. Joseph Charney is a research meteorologist with the U.S. Forest Service in Lansing, MI. He uses numerical models to study fire-atmosphere interactions at multiples scales and works to develop tools and theories for use by fire weather forecasters, fire managers, and research scientists.

33. A Study of the Influence of Vertical Canopy Structure on Fire-Atmosphere Interactions Michael Kiefer, Assistant Professor, Michigan State University Additional Authors: Warren E. Heilman, Research Meteorologist, USDA Forest Service Shiyuan Zhong, Professor, Michigan State University Joseph J. Charney, Research Meteorologist, USDA Forest Service Xindi Bian, Meteorologist, USDA Forest Service Abstract: In this study, we use the ARPS-CANOPY model to explore the sensitivity of fire-perturbed variables within and above a forest canopy (e.g., mean wind speed, turbulent kinetic energy) to vertical canopy structure. The ARPS-CANOPY model, a version of the Advanced Regional Prediction System (ARPS) with a canopy sub-model, is applied in this study with 10 m (2 m) horizontal (vertical) grid spacing in order to partially resolve turbulence in the forest canopy. A series of numerical experiments with different vertical canopy structures are performed with a 50-m wide parameterized low-intensity fireline. For each of the cases considered, an additional simulation is performed without the fire to facilitate comparison of the fire-perturbed and background atmospheric conditions. A total of five canopy profiles are considered, including thick overstory / thin understory canopies (e.g., Loblolly Pine), thin overstory / thick understory canopies (e.g., bare deciduous with evergreen understory), and more uniformly distributed forest canopies. Analyses of both mean and instantaneous wind velocity, turbulent kinetic energy, air temperature, and turbulent mixing of heat are presented in order to examine the fire- perturbed atmosphere on multiple time scales. A brief discussion of the potential applicability of results to fire behavior, tree mortality, and smoke transport and dispersion is included. Presenter Bio: Michael is an Assistant Professor in the Geography Department at Michigan State University specializing in numerical modeling of meso- to micro-scale atmospheric phenomena. He has worked on topics as varied as simulating mean and turbulent flow downstream of wildland fires to examining nocturnal cooling processes inside basins and valleys. Michael came to MSU as a post- doctoral research associate in 2009, and became an Assistant Professor in 2013. He holds a BS in Atmospheric Science from the University at Albany and MS and PhD degrees in Atmospheric Science from North Carolina State University.

34. Ignition from fire perimeter and assimilation into a coupled fire-atmosphere model Adam Kochanski, Research Assistant Professor, Atmospheric Sciences Department, University of Utah

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Additional Authors: Mary Ann Jenkins, Associate Professor, Department of Earth and Space Science and Engineering Lassonde School of Engineering York University Sher Schranz, Senior Project Manager, Global Systems Division, NOAA ESRL Jan Mandel, Professor and Chair, Department of Mathematical and Statistical Sciences University of Colorado Denver Volodymyr Kondratenko, Research Assistent, Department of Mathematical and Statistical Sciences University of Colorado Denver Martin Vejmelka, Research Assistent Professor, Department of Mathematical and Statistical Sciences University of Colorado Denver Abstract: Assimilation of fire perimeter data into a numerical coupled atmosphere wildfire-fire prediction model is key to simulations of existing wildfires where the ignition point is not known. Also, long-lived fires require simulated fire spread to be corrected by cyclically assimilating newly available perimeter observations. In existing operational wildfire spread models without two-way coupling with the atmosphere, the state of the fire does not affect the state of the atmosphere; the existing fire perimeter is simply specified by prescribing the burnt area. However, in a coupled numerical fire-atmosphere prediction framework, the whole burn area cannot be ignited in a single time step. Excessive heat would be released instantaneously, resulting in numerical instability and/or non-physical plume rise and atmospheric circulation. Using fuel data reflecting prior fire history can alleviate this problem, as the already burnt fuel limits fire ignition to the fire perimeter only. Unfortunately, the initial atmospheric state would still be out of sync with the fire; the atmospheric circulation at start time would be as if there was no prior fire. The model would require some spin-up time before the fire-atmosphere coupling establishes the fire-atmosphere induced circulation. We present a fire perimeter assimilation technique that enables starting from an observed fire perimeter, while preserving the fuel balance and the fire-atmosphere equilibrium, so that the model can seamlessly continue simulating fire progression from an observed perimeter, while maintaining realistic plume properties and fire-induced circulations. The method is based on running the fire propagation method from a given fire perimeter backward in time. The resulting map of fire arrival time is used prior to the perimeter time instead of the fire propagation model to release the heat and moisture forcing from the fire into the atmosphere gradually, so that atmospheric circulation consistent with the fire is established. The coupled model then continues from the perimeter time. We will show how this technique can be applied to large multi-day fires, and how fire progression and plume rise benefit from a cyclical assimilation of fire perimeter observations. Keywords: WRF, WRF-Sfire, fire modeling, fire data assimilation Presenter Bio: Adam Kochanski is a modeler interested in simulating wildland fires in coupled fire- atmosphere frameworks. He is one of the authors of the WRF-SFIRE with over 7 years of experience in coupled fire-atmosphere modeling.

35. Developments in the BlueSky smoke modeling framework and related smoke tools Narasimhan Larkin, Research Physical Climatologist and Team Leader, U.S. Forest Service Additional Authors: Susan O'Neill, U.S. Forest Service Robert Solomon, University of Washington Joel Dubowy, University of Washington Jonathan Callahan, Mazama Science Shih-Ming Huang, Sonoma Technology Pete Lahm, U.S. Forest Service 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Abstract: The BlueSky smoke modeling framework is used for emissions inventory and smoke modeling work for both research and operational decision support. A variety of smoke tools have been developed that use BlueSky including daily smoke model predictions across the U.S. and Canada.. We present here the latest updates on development work occurring in BlueSky and related tools, such as a new smoke model viewer and analysis tool suite. Many of these have been built in partnership with the Air Resource Advisor community that serves as the incident specialists for smoke during major wildfire incidents. The new BlueSky Framework v4.0 will be announced which is freely available and includes a complete redesign to make it faster, more portable, simpler to maintain, and more robust. Additionally the new BlueSky Playground v3 web tool will be shown with various user-requested features including reports and a redesigned interface. Other related systems to be discussed include air quality monitoring report features, a smoke complexity analysis tool, and additional modeling capabilities. Keywords: Smoke modeling, decision support tools Presenter Bio: Dr. Larkin is a senior scientist and team leader for the U.S. Forest Service AirFire Team in Seattle, Washington. He has led development of numerous smoke related decision support tools in use today.

36. The Effect of Forest Gaps on the Transport and Dispersion of Smoke Plumes from Low-Intensity Wildland Fires Jovanka Nikolic, Michigan State University, Department of Geography Additional Authors: Jovanka Nikolic, PhD candidate, Department of Geography, Michigan State University, East Lansing, MI Michael T. Kiefer, Assistant Professor, Department of Geography, Michigan State University, East Lansing, MI Shiyuan Zhong, Professor, Department of Geography, Michigan State University, East Lansing, MI Warren E. Heilman, Research Meteorologict, USDA Forest Service Northern Research Station, Lansing MI Joseph J. Charney, Research Meteorologist, USDA Forest Service Northern Research Station, Lansing MI Xindi Bian, Research Meteorologist, USDA Forest Service Northern Research Station, Lansing MI Abstract: Smoke emissions from wildland fires affect local and regional air quality. However, relatively little is known about the role played by the forest canopy structure and especially gaps in the forest canopy in the transport and dispersion of smoke from wildland fires in forest environments. In this study, we examine the effects of forest gaps on the dynamic behavior of smoke plumes from low- intensity surface fires. This is accomplished by tracking smoke particles using the Pacific Northwest National Laboratory’s (PNNL) Integrated Lagrangian Transport (PILT) model, with background and fire- perturbed meteorological inputs provided by the ARPS-CANOPY model [a version of the Advanced Regional Prediction System (ARPS) model with a canopy submodel]. Several numerical experiments are conducted wherein gaps in an otherwise homogeneous forest canopy are placed in different positions relative to a stationary low-intensity fire line. The sensitivity of the smoke plume characteristics to differences in the fire-perturbed local temperature, wind velocity and turbulent kinetic energy fields are examined to evaluate the impacts of the relative positions of the fire line and forest gaps on smoke plume behavior. Keywords: wildland fires, smoke dispersion, ARPS-CANOPY, PILT Presenter Bio: Jovanka is a PhD candidate in the Geography Department at Michigan State University. Her work is focused on the transport and dispersion of smoke from forest fires, especially the impact of canopy structure on smoke behavior. Jovanka uses a particle dispersion model to better explain the impact of canopy and fire-modified atmospheric conditions on smoke spreading in order to assist fire managers when planning prescribed burns, as well as impact of prescribed and wildfires on local air 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

quality. She has a BS in Meteorology from the Faculty of Physics at the University of Belgrade, and MS from Michigan State University.

37. Fire weather drives the population collapse of obligate-seeder forests David Bowman, Professor, University of Tasmania Additional Authors: Grant J. Williamson, Dr, University of Tasmania Lynda D. Prior, Dr, University of Tasmania Brett P. Murphy, Dr, Charles Darwin University Abstract: Frequent landscape fires can cause the abrupt switch from forest to non-forest state with substantial knock on effects on carbon storage and biodiversity. Particularly vulnerable are forests that regenerate exclusively from seed following high-severity fire. Frequent fires can cause local extinction of such ‘obligate-seeders’ because there is insufficient time for regenerating plants to reach sexual maturity. Climate change has been identified as a potentially powerful driver of land cover change through a variety of mechanisms including reduced growth and increased time to maturity, decreased fecundity and seedling establishment, while simultaneously increasing the frequency of high-severity fires via worsening fire weather conditions, a process described as ‘interval squeeze’. Another potential driver of state change is the increased risk of subsequent fires in regenerating stands which can result in a positive feedback, described as a ‘landscape trap’. Here we argue that high-severity fires are primarily driven by fire weather, with a negligible effect of stand age, highlighting the vulnerability of many obligate-seeder forests to worsening fire season under climate change. Keywords: Climate change, fire weather, Australia, fire frequency, regeneration, assisted migration Presenter Bio: Professor David Bowman holds a research chair in Environmental Change Biology in the School of Biological Sciences at the University of Tasmania at Hobart Australia. After completing his PhD in forest ecology and silviculture at the University of Tasmania in 1984, he spent two decades in the undertaking full time research throughout northern Australia working closely with Aboriginal people. He has received travelling fellowships from the Australian Academy of Science, Harvard, Kyoto, Leeds and Arizona universities. His research is collaborative, transdisciplinary with the following themes: understanding how fire influences the Earth system, how humans alter ecologies with fire, how fire shapes landscapes, and the quest for sustainable fire management. Collectively he describes his research program as pyrogeography.

38. Contributions to a megafire: Fire-induced winds, drought, and fuel buildup due to fire suppression Janice Coen, Project Scientist, Mesocale and Microscale Meteorology Laboratory, NCAR, Boulder, CO Additional Authors: E. Natasha Stavros, California Institute of Technology, Jet Propulsion Laboratory, Pasadena, CA Josephine A. Fites-Kaufman, Pacific Southwest Region, USDA Forest Service, Vallejo, CA David Schimel, California Institute of Technology, Jet Propulsion Laboratory, Pasadena, CA Abstract: The 2014 King fire garnered societal concern and scientific interest because of its size, cost, and long-term ecological and hydrological impacts. It exhibited extreme and unanticipated behavior as it was driven by complex mountain airflows beneath the resolution of standard forecast models and periods of growth apparently driven primarily by fire-induced winds. It spread in an area characterized by complex fuel beds shaped by stressors such as drought, land management practices (e.g. forest cultivation and harvesting, fire suppression, and fuel mitigation), and burn scars from prior fires. Here, the CAWFE® coupled weather-fire modeling system was used to simulate three periods during the King fire’s evolution, including an unanticipated surge up the Rubicon Valley. Control simulations use LANDFIRE data to prescribe spatial fuel model variability, with National Infrared Operations and Visible and Infrared Imaging Radiometer Suite active fire detection data to initialize and validate fire extent. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Sensitivity experiments test the impact of fuel moisture and fuel load on fire growth. Another experiment analyzes the fire-induced winds, that is, the winds creates by heat release from the fire itself. Results showed the King Fire’s extreme behavior and unanticipated size resulted from complex mountain airflows beneath the resolution of forecast models and periods of growth apparently driven by fire-induced winds, both captured by the model. Fuel moisture had relatively little impact on fire extent or rate of spread, while fuel buildup impacted rate of spread primarily on sloped terrain, suggesting fuel treatments aimed at mitigating megafire behavior might be most effective there. Keywords: fire, climate, megafire, case study Presenter Bio: Dr. Janice Coen is a Project Scientist at the National Center for Atmospheric Research in Boulder, Colorado. She received a B.S. in Engineering Physics from Grove City College and an M.S. and Ph.D. from the Department of Geophysical Sciences at the University of Chicago. She has served as a member of the Board of Directors of the International Association of Wildland Fire and is currently an Associate Editor for the International Journal of Wildland Fire. She investigates wildland fire behavior and its interaction with weather using coupled weather- wildland fire computer simulation models and by analyzing infrared imagery of wildfires and prescribed fires.

39. Climate-induced variations in global wildfire danger from 1979 to 2013 W. Matt Jolly, Research Ecologist, USFS, RMRS, Fire Sciences Laboratory Additional Authors: Mark A. Cochrane, Professor, South Dakota State University Patrick H. Freeborn, Physical Scientist, Missoula Fire Sciences Laboratory Zachary A. Holden, Remote Sensing Analyst, USFS R1 Timothy J. Brown, Director, Western Region Climate Center Grant J. Williamson, Research Associate, University of Tasmania David MJS Bowman, Professor, University of Tasmania Abstract: Climate strongly influences global wildfire activity, and recent wildfire surges may signal fire weather-induced pyrogeographic shifts. Here we use three daily global climate data sets and three fire danger indices to develop a simple annual metric of fire weather season length, and map spatio- temporal trends from 1979 to 2013. We show that fire weather seasons have lengthened across 29.6 million km2 (25.3%) of the Earth’s vegetated surface, resulting in an 18.7% increase in global mean fire weather season length. We also show a doubling (108.1% increase) of global burnable area affected by long fire weather seasons (41.0 s above the historical mean) and an increased global frequency of long fire weather seasons across 62.4 million km2 (53.4%) during the second half of the study period. If these fire weather changes are coupled with ignition sources and available fuel, they could markedly impact global ecosystems, societies, economies and climate. Keywords: fire danger, climate change Presenter Bio: Dr. Matt Jolly is a Research Ecologist in the Fire, Fuel and Smoke Science Program of the US Forest Service, Fire Sciences Laboratory in Missoula, MT. He received a BA in Environmental Science from the University of Virginia and a PhD in Forestry from the University of Montana. His main research interest is to improve our understanding of the roles that live and dead fuels play in wildland fires and to use this improved understanding to develop or improve predictive tools that can help support fire management decisions.

40. Exploring interactions among multiple disturbance agents and future climates in forest landscapes Robert Keane, Research Ecologist, US Forest Service Rocky Mountain Research Station Additional Authors: 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Rachel Loehman, Research Landscape Ecologist US Geological Survey, Alaska Science Center 4210 University Drive, Anchorage, AK 99508 Carol Miller, USDA Forest Service, Rocky Mountain Research Station, Aldo Leopold Wilderness Research Institute, 790 East Beckwith Ave, Missoula MT 59801 Erica Smithwick, The Pennsylvania State University, 318 Walker Building, State College, PA Abstract: Interactions among disturbance, climate, and vegetation determine landscape patterns and influence ecosystem processes. Dynamic and reciprocal interactions among disturbances can also temporarily or persistently alter landscape trajectories, especially in new climate regimes. Ecological models are used routinely to explore ecological dynamics across heterogeneous landscapes, but few models are able to simulate effects of multiple interacting disturbance events. Projecting how multiple disturbance interactions might result in novel and emergent landscape behaviors is critical for addressing climate change impacts and designing land management strategies that are appropriate for future climates. In this chapter, we demonstrate the importance of interacting disturbances using an example from fire-dominated, pine forested ecosystems of the northern Rocky Mountains, USA, where mountain pine beetle (Dendroctonus ponderosae), white pine blister rust (Cronartium ribicola), and wildland fire interact with the vegetation and climate to create unique landscape behaviors. First, we review the literature on the effects of these three disturbances and their interactions in the northern Rockies forests. Then we used the mechanistic landscape process model FireBGCv2 to simulate effects of multiple disturbance interactions on vegetation composition and basal area for two landscapes under current and projected future climates. Our findings are that (1) multiple disturbance interactions influence landscape patterns more than single or no disturbances; (2) disturbance responses are typically indirect feedbacks mediated through changes in vegetation and fuels; (3) disturbance interactions may overwhelm direct effects of climate changes or effects of a single disturbance on ecosystems, and (4) exploring disturbance interactions demands a mechanistic simulation approach to fully represent those important ecological processes that are directly and indirectly affected by disturbances and their interactions. Disturbances and their interactions must be addressed to properly assess future landscape changes under projected climate regimes Keywords: Landscape modeling, climate change, FireBGCv2, simulation Presenter Bio: Keane's most recent research includes 1) developing ecological computer simulation models for the exploring landscape, fire, and climate dynamics, 2) conducting field research on the sampling, describing, modeling, and mapping of fuel characteristics, and 3) investigating the ecology and restoration of whitebark pine. He received his B.S. degree in forest engineering from the University of Maine, Orono; his M.S. degree in forest ecology from the University of Montana, Missoula; and his Ph.D. degree in forest ecology from the University of Idaho, Moscow

41. Projected impacts of climate change on vegetation and fire in the Huachuca Mountains of Arizona Christopher O'Connor, Ecologist, USFS RMRS Additional Authors: Gregg Garfin, Associate Professor, University of Arizona SNRE Donald Falk, Associate Professor, University of Arizona SNRE Abstract: The American Southwest is expected to experience temperature increases and reductions in winter precipitation over the next several decades. These rapid climatic changes will influence forest species distributions and fire activity; with implications for human safety and sustainability of ecosystems. To assess the influence of projected climate on vegetation and fire dynamics in this sensitive region, we conducted a landscape simulation of the Huachuca Mountains of Southeastern Arizona. We used a regionally downscaled ensemble of three Global Climate Models to drive simulations of vegetation and fire dynamics in the forest process model FireBGCv2. Results suggest a significant reduction in plant biomass in forests of the Huachuca Mountains. This reduction includes the 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

loss of large old pine, Douglas- fir, and aspen forests from much of the upper elevation ecosystem. Smaller diameter Madrean oak woodland and shrubland species are likely to expand into formerly conifer-dominated forests. Landscape fire simulations suggest that changing climate is likely to increase the risk of high-severity fire in the short term, however the loss of biomass associated with high-severity fire and increasing occurrence of persistent drought are likely to function as negative feedbacks on fire spread, and may result in a net reduction in fire frequency over the next 50 years. Proactive fuel reduction treatments have potential to reduce the risk of high severity fire in and around protected Mexican Spotted Owl (MSO) breeding sites in the short term, but this protective effect is lost within 20 years without additional fuel treatments. Simulation of a second series of fuel reduction treatments 20 years into the simulation suggest that additional thinning treatments may be able to instill further protection from high-severity fire for additional decades. While fire management in the form of direct fire suppression or fuel reduction treatments does not appear to slow the rate of biomass loss under the specific climate change scenario used, fuel modification in conjunction with fire use may promote forest tree species diversity and longevity while providing some degree of protection for sensitive wildlife habitat. Keywords: Climate-vegetation interactions, downscaled climate projection, drought, biomass change Presenter Bio: Kit O’Connor is an Ecologist with the USFS Rocky Mountain Research Station in Missoula, Montana working on the human dimensions of fire and forest changes resulting from land management and changing climate. His current projects involve support for the National Forest Planning process and critical examination of the incident command structure to assist in development of scalable, long-term forest and fire planning decision systems. Kit’s previous work was centered on ecological reconstruction of landscape changes, response to climate and human impacts, and projected changes to forest and disturbance dynamics in response to projected future climate conditions.

42. ForestFireFOAM: A Numerical Tool For Investigating The Burning Dynamics Of Wildland Fuels Mohamad El Houssami, PhD Student, University of Edinburgh Additional Authors: Aymeric Lamorlette, Aix-Marseille Universite Dominique Morvan, Aix-Marseille Universite Rory Hadden, University of Edinburgh Albert Simeoni, Exponent Inc Abstract: Understanding the behavior of litter fuels is essential in developing a complete understanding of wildfire spread. The challenge of understanding the fire behavior of such fuels arises from the porous nature and the resulting coupling of the physico-chemical complexities of the fuel and the surrounding environment, which strongly influences the burning dynamics. In this work, a method is presented to accurately understand the processes, which control the burning behavior of a porous wildland fuel layers using numerical simulations coupled with laboratory experiments. Simulations are undertaken using ForestFireFOAM – a modification of FireFOAM – that uses a Large Eddy Simulation solver to represent porous fuel by implementing a multiphase formulation to conservation equations (mass, momentum, and energy). This approach allows the fire-induced behavior of a porous, reactive and radiative medium to be simulated. Conservation equations are solved in an averaged in a control volume at a scale sufficient to contain both coexisting gas and solid phases, considering strong coupling between the phases. Processes such as drying, pyrolysis, and char combustion are described through temperature-dependent interaction between the solid and gas phases. Different sub-models for heat transfer, pyrolysis, gas combustion, and smouldering have been implemented and tested to allow better representation of these combustion processes. Numerical simulations are compared to experiments undertaken in a controlled environment using the FM Global Fire Propagation Apparatus. Pine needle beds of varying density and surface to volume ratio 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

were subject to radiative heat fluxes between 20 and 50 kW/m2 to interrogate the ignition and combustion behavior. The key flammability parameters of mass loss rates, heat release rates, gas emissions and temperature fields that agree well with experimental observations. It is anticipated that this approach, with the appropriate modifications, will enable better understanding of the fundamental combustion behavior of complex wildland fuels to support larger-scale model development and optimization of large-scale CFD simulations. Keywords: Simulation, combustion, multiphase, pine needle Presenter Bio: Mohamad is a PhD student at the University of Edinburgh. He is working on characterising the different dynamics of forest fuels, by developing numerical models and conducting controlled experiments at laboratory scale. He received his Masters degree from the University of Aix- Marseille in Mechanical Engineering and his Bachelor degree in Applied Physics at the University of Strasbourg. His background is in physics, fluid mechanics, and heat transfer. He has also worked on large scale fire metrology, firebrand generation, impact on structural ignition, and fire investigation.

43. A Fundamental Exploration of Flame Structure in Wildland Fires Colin Miller, PhD Candidate, University of Maryland, College Park Additional Authors: Wei Tang, PhD Candidate, University of Maryland, College Park Salman Verma, PhD Candidate, University of Maryland, College Park Arnaud Trouve, Professor, University of Maryland, College Park Jason Forthofer, Mechanical Engineer, Missoula Fire Sciences Laboratory Mark Finney, Research Forester, Missoula Fire Sciences Laboratory Michael Gollner, Assistant Professor, University of Maryland, College Park Sara McAllister, Research Mechanical Engineer, Missoula Fire Sciences Laboratory Jack Cohen, Research Physical Scientist, Missoula Fire Sciences Laboratory Abstract: In the wildland environment, many fundamental features of flame spread remain unexplained. This work reviews several recent projects, including laboratory experiments and numerical simulations, which have begun to shed light on the flame spread process. An understanding of the three-dimensional structure of turbulent flames becomes very important in wildland fire spread, which is driven by convective ignition of fine fuels. Knowledge of the intermittent behavior of wind-blown flames may help us to properly determine the influence of these processes on wildland fire spread. In order to develop a thorough statistical analysis of flame structure, experiments and simulations on stationary heated plates and gas burners have been performed. Flame geometry and pulsating behaviors of both the visible flame and the incident heat flux have been examined. Additionally, dominant coherent structures in the form of streamwise streaks and spanwise waves are consistently observed along the flame front. Thermal and inertial quantities of these instabilities, which may drive the formation of the flame, have been quantified and scaled. The coupled roles of buoyancy, shear forces, and the boundary layer have been investigated in an effort to determine the mechanisms driving the development of a wind-driven flame. Keywords: fire spread, flame structure Presenter Bio: Colin Miller is a PhD candidate in Mechanical Engineering at the University of Maryland, College Park. He recently completed his M.S. and B.S. in Fire Protection Engineering. His work now focuses on wildland fire spread and he works closely with the Missoula Fire Sciences Laboratory.

44. Forward Heating in Wind-Driven Fire Spread Wei Tang, PhD Candidate, University of Maryland Additional Authors: Colin Miller, PhD Candidate, University of Maryland, College Park 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Mark A. Finney, Research Forester, Missoula Fire Sciences Laboratory Michael J. Gollner, Assistant Professor, University of Maryland, College Park

Abstract: Flame spread is known to be governed by downstream heating from flames to unburned fuels, however the process by which this heat is transferred still remains a subject of debate, especially under high wind or steep slope conditions. An experimental apparatus was designed to produce stationary gas- burner generated flames to study the effect of wind on the shape of these flames and the downstream heat flux to an inert surface, simulating unburned fuels ahead of the flame front. A variety of forced flow rates (0.87 – 2.45 m/s) and fire sizes (4.2 to 9.5 kW) were employed here to study different fire scenarios. Local total heat fluxes were measured for different downstream locations from the trailing edge of the burner. It was found that there is a steep change in downstream heat fluxes where the flame begins to lift off the downstream surface. This balance of inertial versus buoyant phenomena was well correlated with the Richardson number, a measure of natural versus forced convection over a surface. This transition occurred close to the location where the Richardson number equals 1. This was similarly found when analyzing flame attachment by side-view video and local temperature measurements. The results from this study suggested that local heat flux values reach a maximum when the flame makes a transition from wind-dominated to buoyantly-dominated fire behavior. This parameter was then used to estimate flame attachment length and correlate heat fluxes, which will be useful in future development of fire spread models. Keywords: fire spread, heat fluxes Presenter Bio: Wei Tang is a PhD Candidate in the Department of Fire Protection Engineering studying wildland fire spread. He hold a B.S. and M.S. in Fire Protection Engineering from the University of Science and Technology, China.

45. Laboratory Studies on the Generation of Firebrands and Ignition of Structural Components Raquel Hakes, University of Maryland Additional Authors: Sara Caton, Graduate Research Assistant, University of Maryland Matthew Weston-Dawkes, Undergraduate Researcher, University of Maryland Nelson Bryner, Division Chief Fire Research Division, NIST Dr. Jiann Yang, WUI Group Leader Fire Research Division, NIST Dr. Michael J. Gollner, Assistant Professor, University of Maryland Abstract: The increasing frequency and severity of destructive wildland-urban interface (WUI) fires has become a recurring threat to lives and property. One of the primary drivers of fire spread through WUI communities are firebrands, small burning embers generated by surrounding wildland fires or nearby structures. Despite recognition of their impact, it has proved incredibly difficult to quantify the contribution of firebrands to fire spread. While many approaches exist to determine forward transport of firebrands, very little work addresses how firebrands are generated from vegetative fuels or the propensity of structural materials to ignite based on a specific flux of firebrands. This talk will present recent work at the University of Maryland addressing both problems through small-scale, laboratory experiments, where significant into the physical processes occurring can be gained. Firebrand generation experiments were conducted using dried, dead wooden dowels to determine the characteristics responsible for the breakage of firebrands off of vegetative fuel. Different levels of heating and combustion were applied to sticks of different diameters and species, followed by observation of their mass and size after extinguishment. These properties were compared with critical fracture stresses, measured with a three-point bending test. Based on the fracture stress under different levels of heating, the relationship between ambient and fire-generated winds and breakage and eventual lofting of brands can be accomplished. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Additional experiments were conducted to determine how firebrands ignite structural components, such as decking, which often causes subsequent ignition of the home itself. Smoldering firebrands were deposited onto flat and L-shaped wooden samples which mimic deck materials to observe the conditions required to transition from smoldering to flaming ignition. Rather than individual firebrands, piles of firebrands were deposited, similar to what has been observed in wind tunnel experiments of firebrand generators. The influence of self-heating, wind, geometry and firebrand loading have been explored and related to the probabilities of transition to flaming. Keywords: Firebrands, Generation, Ignition, WUI Presenter Bio: Sara Caton and Raquel Hakes are graduate students in the Department of Fire Protection Engineering at the University of Maryland, College Park pursuing their Masters of Science in Engineering. Sara Caton is also a Pathways intern at the National Institute for Standards and Technology. Both were part of the NFPA-sponsored report on Pathways to Fire Spread in the Wildland Urban Interface and are now focusing their research on firebrands in the WUI.

46. Experimental Study on the Surface Spread of Smouldering Peat Fires Xinyan Huang, University of California Berkeley Additional Authors: Francesco Restuccia, Imperial College London Michela Gramola, University of Cambridge Guillermo Rein, Imperial College London Abstract: Smouldering combustion is the driving phenomenon of wildfires in peatlands, and is responsible for large amounts of carbon emissions and haze episodes in Southeast Asia and Northeast Europe. Compared to flaming fires, smouldering is slow, low-temperature, flameless, and most persistent, yet it is poorly understood. Peat, as a typical organic soil, is a porous and charring natural fuel, thus prone to smouldering. The spread of smouldering peat fire is a volumetric phenomenon, including two components: in-depth vertical and surface spread. In this study, we investigate the surface spread of peat fire under various moisture and wind conditions. Visual and infrared cameras as well as a thermocouple array are used to measure the temperature profile and the lateral spread rate. For the first time the overhang, where smouldering spreads beneath the free surface, is observed in the laboratory, which helps understand the interaction between oxygen supply and heat loss in peat fires. The periodic formation and collapse of overhang is observed, and the overhang thickness is found to increase with peat moisture content and wind speed. Experimental results also show that the surface spread rate decreases with peat moisture content, while increases with the wind speed. The proposed simple analysis on the heat loss and oxygen supply shows that the oxygen supply is the dominant mechanism for dry peat samples, so the spread rate decreases with depth. As the moisture content increases, the spread rate becomes less sensitive to the depth and the wind speed because peat MC becomes the dominant mechanism. The analysis further suggests that for surface peat fire spread, formation of overhang is caused by the spread rate difference between the top and lower peat layers. Keywords: wildfire, organic soil, spread profile, moisture, wind Presenter Bio: Xinyan is a final year PhD student in the Department of Mechanical Engineering at Imperial College, funded by the Exceptional Overseas Scholarship. He is working with Dr Guillermo Rein in the area of Smouldering Combustion and Fire Science. His PhD thesis focuses both experimental and computational studies on smouldering fires of peat, the largest fire on the earth, including chemical kinetics, fire dynamics, and emissions.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

47. Differential respiratory health effects from the 2008 northern California wildfires: a spatiotemporal approach Colleen Reid, Robert Wood Johnson Health and Society Scholar, Harvard T.H. Chan School of Public Health Additional Authors: Michael Jerrett, Professor and Chair, UCLA Fielding School of Public Health Ira Tager, Professor Emeritus, UC Berkeley School of Public Health Maya Petersen, Associate Professor, UC Berkeley School of Public Health Jennifer Mann, Project Scientist, UC Berkeley School of Public Health John Balmes, Professor, UC Berkeley School of Public Health Abstract: Background: We investigated health effects associated with fine particulate matter during a long-lived, large wildfire complex in northern California in the summer of 2008. Methods: We estimated exposure to PM2.5 for each day using an optimal exposure prediction model created through data- adaptive machine learning methods from a large set of spatiotemporal data sets. We then used Poisson generalized estimating equations to calculate the effect of exposure to PM2.5 on cardiovascular and respiratory hospitalizations and emergency department (ED) visits. We further assessed effect modification by sex, age, and area-level socioeconomic status. Results: We observed a linear increase in risk for asthma hospitalizations and ED visits with increasing PM2.5 during the wildfires. The association of PM2.5 with asthma health care utilization was not present in the periods before and after the fires. The effect of PM2.5 during the wildfire period was more pronounced in women compared to men and adults ages 20-64 compared to children and adults 65 or older. chronic obstructive pulmonary disease (COPD) ED visits were associated with PM2.5 during the fires, but this effect was not significantly different from that found before or after the fires. We did not find consistent effects of wildfire smoke on other health outcomes. Conclusion: Using a novel spatiotemporal exposure model, we found that hospitalizations and ED visits for asthma were significantly associated with PM2.5 from the 2008 northern California wildfires. Keywords: health effects; air pollution; wildland fire; spatiotemporal modeling Presenter Bio: Colleen Reid work focuses on the health impacts of exposures influenced by global climatic changes and society’s responses to those changes. Colleen received her Ph.D. and M.P.H. in Environmental Health Sciences from the University of California, Berkeley and a BS from Brown University. She is currently a Robert Wood Johnson Health and Society post-doctoral scholar. Colleen is researcher on a Joint Fire Science Program grant titled “Health effects from wildfire air pollution: a spatiotemporal modeling approach”.

48. Montana Idaho Airshed Group Smoke Management Decision Support Erin Law, Smoke Management Program Coordinator, Montana/Idaho Airshed Group Additional Authors: Matt Mavko, Air Sciences Inc Thomas Dzomba, MSPH, Asst. Director Air Quality and Smoke Management US Forest Service Region 1 Abstract: In 1978, the major open burners in Montana and their colleagues at the Department of Environmental Quality organized to develop a plan that addressed the need to perform prescribed burning in the face of increasing air quality regulations. They decided a meteorologist and a forester should work together to gather daily burn information, weather forecasts, and air quality data to make prescribed burning “go” or “no-go” recommendations on a daily basis. The success of the Montana Airshed Group attracted burners in North Idaho to join in 1990, with burners in South Idaho joining in 1999. Today the Montana/Idaho Airshed Group has 22 members and over 200 active burners in the federal, state, private, and tribal sectors, and its Smoke Management Unit provides a burn-by-burn recommendation for almost 6,000 burn requests on average each year across two states. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

The heart of the process is the Airshed Management System website and database, upgraded and currently maintained by Air Sciences Inc. of Portland, Ore. Decades of feedback across the diverse group membership led to a sophisticated refinement of fuels and location inputs, and communication protocols for coordination, conflict resolution, and decision notification. The database stores burn fuel and location information, and the website provides daily-updated dispersion meteorology forecasts and outlooks, along with a map of daily proposed, approved, and completed burns. In this way, instant access to in-depth burn information is available to county, tribal, and regional airshed coordinators and air quality regulators, with the opportunity to discuss individual and geographic area burns with respect to vulnerable receptors, commensurate activity, and forecast meteorology in real-time. This allows for issue resolution that balances risk with opportunity from all smoke sources, including agricultural and residential, within a timeframe before “no-go” becomes a foregone conclusion for forest burners due to wildland logistics. This presentation will discuss how a comprehensive coordination process—cutting- edge technology and old-fashioned civil discourse—allows Montana/Idaho Airshed Group burners the safest, most aggressive, and most successful short- and long-term smoke management decision support available. Keywords: smoke management, prescribed burning, air quality, real-time decision support Presenter Bio: Erin started her career in fire after she graduated from Hellgate High in Missoula, Montana many years ago. She worked on various suppression crews for the BLM and Forest Service in Colorado and Montana while going to school, and in 1992, she moved to Grangeville, Idaho to work on the Moose Creek Ranger District in the Selway-Bitterroot Wilderness. It was a great piece of country so she stayed for 15 years and was eventually promoted to District Fire Management Officer. In 2007, she accepted her current position as smoke coordinator for the Montana/Idaho Airshed Group.

49. A Flexible Decision Support Framework for Smoke Management: 3 Case Studies Matthew Mavko, Air Quality Scientist, Air Sciences Inc. Additional Authors: Jonny Malloy, Meteorologist/Smoke Management Coordinator, Arizona Department of Environmental Quality Dan Washington, Utah Interagency Smoke Coordinator, Division of Air Quality Abstract: Smoke management is a key component of maintaining air quality for many states and tribes in the United States. Section 309 of the Regional Haze Rule mandates smoke management programs (SMPs) for some states. For other states where burning is a primary land management tool, proactive smoke management may be used for public health and safety. Proactive smoke management requires extensive coordination and communication between stakeholders on a daily basis to maximize burn opportunities while preventing smoke intrusions in surrounding communities. One of the largest SMPs in the United States, the Montana/Idaho Airshed Group, uses a web-based tool called the Airshed Management System (AMS) to facilitate this process. Originally conceived by the University of Montana Fire Center, the AMS provides burn-by-burn recommendations for 6,000 burn requests per year. In the past two years, the AMS framework has been adapted to support SMPs in Arizona and Utah. This presentation will discuss how the AMS framework supports the work of these three SMPs, and its flexibility to adapt to programs with unique regulations and procedures. Key features of the framework will be described. In addition, the presentation will examine the recent transition from the previous (mostly paper-based) systems in Arizona and Utah to the AMS framework, along with positive and negative outcomes and whether the framework has improved decision support in each state. Keywords: smoke management, coordination,decision support, web applications Presenter Bio: Mr. Mavko is a senior scientist at Air Sciences Inc., focusing on regional air quality impacts of fires in the United States as well as dust mitigation. He has expertise in fire emissions characterization, ambient air monitoring, spatial database development, and data-driven web 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

applications. Prior to working for Air Sciences, he worked for a research group studying low-cost ambient monitoring techniques, and was a research assistant with the Remote Sensing Group performing vicarious calibration of remote sensing platforms. Mr. Mavko has a B.S. in Physics from the University of Arizona and an M.S. in Environmental Science from Portland State University.

50. Understanding Smoke Transport from Prescribed Burning in the Wildland Urban Interface of Bend, Oregon Susan O'Neill, Air Quality Scientist, USDA Forest Service Additional Authors: Miriam Rorig, Research Meteorologist, USDA Forest Service Colton Miller, Research Assistant, University of Washington Rick Graw, Air Program Program Manager, USDA Forest Service Narasimhan ("Sim") Larking, Research Climatologist, USDA Forest Service Abstract: Prescribed burning (Rx) is carried out to improve forest health, create a diversity of plant and wildlife habitat and to protect communities and quality of life. Federal land management policies are particularly focused on treating acres in the Wildland Urban Interface (WUI) to protect communities from the devastating effects of wildfire. These treatments include prescribed burning. Simultaneously, State and Federal regulations and policies prohibit allowing smoke to be carried into nearby populated areas. The Bend, Oregon area is a nexus of where these policies meet. Therefore, the Forest Service (FS) Region 6 Air Program and the FS Pacific Northwest (PNW) Research Station AirFire Team are collaborating with land managers on the Deschutes National Forest to gain a better understanding of meteorological conditions that can transport smoke into Bend, Oregon. The goal is to increase the pace and scale of forest restoration burning while protecting air quality in Bend and surrounding communities. We conducted two field campaigns deploying up to twelve weather stations (which measure wind speed and direction, temperature, and RH), and three portable air quality particulate monitors (E-Samplers, which measure PM2.5 or particulate matter less than 2.5 micrometers in diameter, and wind speed and direction) from Sisters, Oregon down through Bend to Sunriver, Oregon. Two automated cameras were also deployed at wildfire lookout points. Instruments were deployed for the Fall 2014 prescribed burn season, five weather stations remained in-place over the Winter when pile burning occurred, and the remaining instruments (E-Samplers and cameras) were re-deployed for the Spring 2015 Rx burn season. We also simulated dispersion from prescribed fires at spatial resolutions of 4km and 1km. During the Spring of 2015, a number of smoke intrusions occurred in Bend and were measured at the PM2.5 samplers. Hourly PM2.5 concentrations reached 356 micrograms/m3 on May 29, 2015, most likely due to overnight smoldering of large woody fuels and basal accumulations. We present a detailed analysis of the measured wind conditions and spatial distribution of the smoke for the Spring 2015 smoke intrusions into Bend, Oregon. Keywords: Smoke, PM2.5, Prescribed Burning, WUI, Meteorology, Deschutes Presenter Bio: Susan O’Neill is an Air Quality Scientist with the USDA Forest Service Pacific Northwest Research Station, AirFire Team, and has a Ph.D. from the Laboratory for Atmospheric Research at Washington State University. She began her career developing the BlueSky smoke modeling framework, continued on to the Natural Resources Conservation Service as part of their Air Quality and Atmospheric Change Team, and recently re-joined the AirFire Team. Her research interests extend to all aspects of modeling fire emissions, and modeling smoke dispersion, transport, and chemical transformation.

51. Wildland Fire Smoke: A Hazard for Health Disaster Management Darlene Oshanski, Special Projects Coordinator, Manitoba Health, Healthy Living and Seniors - Office of Disaster Management

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Abstract: As smoke due to wildland fire events is becoming an ever increasing population health concern, involving major risks for the health of the people and the environment, Health Canada and Manitoba Health’s Office of Disaster Management have partnered in an initiative to undertake a number of important components within this issue. Wildland fires occur regularly throughout much of Manitoba during the months of May to October, and as they are primarily caused by human activity (accidental or intentional) or by lightning strikes, they are often difficult to predict and prevent. The decision to evacuate residents of a community due to smoke from a wildland fire can be incredibly complex. Before an evacuation can occur, it is critical to assess health risks. As evacuations are inherently disruptive and costly, one should only occur when health benefits outweigh all risks. Unfortunately however, scientific evidence regarding the short-term and long-term effects of wildland fire smoke on human health is still limited at this time. Manitoba Health and Health Canada have also partnered with Environment Canada, to purchase smoke monitors which will assist with the validation process of smoke forecasts emanating from BlueSky (a forecast system that provides wildland fire location, animations of hourly smoke plume trajectories, and ground level concentrations of particulate matter). Data collected by the smoke monitors will be spatially displayed in Manitoba Health’s Common Operating Picture (COP). The COP is an advanced technology tool consisting of geospatial displays and acting as a common repository of information for decision makers. With more accurate wildland fire smoke forecasting and with the development of a number of smoke related tools, Manitoba Health and Health Canada will target the applicable audiences affected by wildland fire smoke to meet the following objectives; A) Develop specific smoke event health messaging B) Develop First Nation messaging for smoke events C) Examine health outcomes and appropriate messaging for situations when the Air Quality Health Index rises above 10 D) Provide nationally applicable guidance for decision-makers to protect people from wildland fire smoke E) Examine the combined risks of smoke and heat and develop appropriate messaging for a combination smoke/heat event By amalgamating these various tools, and aligning them with the above objectives, Manitoba Health and Health Canada is striving to enhance the capabilities for effective risk management and decision making when it comes to population health and wildland fire smoke in Canada. Keywords: smoke; public health safety; early warning; evacuation; messaging; forecasting Presenter Bio: Darlene Oshanski has worked in the field of Disaster Management for the last 12 years and is currently a Special Projects Coordinator at Manitoba Health's Office of Disaster Management. Her present focus surrounds the human health impacts from wildland fire smoke, how emergency managers, health practitioners, and Manitobans respond to the hazard. Darlene holds a Bachelor of Arts in Applied Disaster and Emergency Studies and Geography.

52. Smoke Monitoring in the Field: Understanding Equipment and the Value of Particulate Matter Data in Making Smoke Management Decisions Don Schweizer, University of California at Merced Additional Authors: Ricardo Cisneros, Professor, University of California Merced Abstract: There are multiple sources of data that can be used to make smoke management decisions during a wildland fire. These data sources include remote sensing, modeling, and ground based air quality monitoring. While all these data streams can help understand the intersection of wildland fire 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

emissions and health impacts, knowing the limitations and strengths of your data is an often overlooked component of effective smoke management. Model or remote sensing data using high quality presentation is widely accessible to smoke managers and can easily be consumed by an awaiting public, but is it accurately reflecting what is being experienced on the ground? Additionally, with multiple types of monitoring devices being simultaneously deployed, with what certainty is your ground based measurement being used appropriately? Accuracy and precision of fine particulate matter (PM2.5) data collection using permanent and temporary beta-attenuation monitors is discussed along with limitations when using these monitors in comparative analysis. Remote sensing data is compared to ground based monitoring data to explore its use as an indicator of population exposure to smoke. We also discuss the accuracy of public health advisories and the potential implications to the representation of health risk. Keywords: Smoke, particulate matter, monitoring Presenter Bio: Don Schweizer is a researcher with UC Merced studying smoke impacts from wildland fire. He is involved in air pollution research and smoke monitoring throughout the Sierra Nevada mountains of California and has worked as an Air Resource Specialist with the U.S. Forest Service and the National Park Service.

53. Lessons Learned from an Unexpected Spread Event on a Large Fire in a Remote Mountain Park Kelsy Gibos MSc., Wildfire Management Specialist, Edson Wildfire Management Area, Forest & Emergency Response Division, Alberta Environment & Sustainable Resource Development Dave Finn, Provincial Fire Behaviour Specialist, Alberta Agriculture and Forestry Abstract: While multiple large fires burned elsewhere in dry 2015 spring conditions in the boreal forest across the province of Alberta, a rare lightning strike ignited a fire in a remote mountain park in western Alberta. Given the infrequency of large fires in this region and the mild, only recently thawed conditions, fire managers did not expect the fire to erupt, taking an unanticipated 7.5 mile (12 kilometer) run in under four hours that blackened forest indiscriminately from treeline to treeline. In three days, it burned over 30,000 acres (12,000 hectares) of montane forest, challenging traditional suppression tactics with intense fire behaviour and steep, inaccessible terrain. Conventional wisdom and experience would suggest that this was an anomalous event related to specific unseasonable weather conditions and that large mountain fires in the spring are not normal. However, in post-fire reflection, we have uncovered a collection of reoccurring and not uncommon phenomena that suggest we may need to re-examine the clues we use to anticipate fire in an infrequent fire regime. Using our own experiences as local fire practitioners, this presentation will guide you through a case study that investigates the environmental and human factors that we feel contributed to the ‘surprise’ we experienced as Fire Behaviour Analysts (FBANs) on the Rockslide Creek Wildfire in Alberta’s Willmore Wilderness Park. We will step through the fire environment including: antecedent weather conditions; contributing decadent fuel dynamics; and terrain-induced weather effects. We will address the complacency that sets in when fire in your management area is relatively infrequent, along with our confidence in our ability to remotely sense what is actually going on in out-of-the-way environments. This presentation provides a collection of ‘watch-outs’ for those who manage infrequent fire regimes and difficult to access montane fire environments. By watching for our identified cues and thinking about their cumulative effects on potential fire behaviour we hope to teach others to be more open- minded and better prepared for fire in remote wilderness locations, especially when you least expect it. Keywords: fire behaviour, case studies, lessons learned, fire danger, fire weather, firefighter safety Presenter Bio: Kelsy Gibos is a fire behaviour specialist who has observed and studied fire behaviour in Canada, New Zealand and Australia. Her experiences abroad have ignited an interest in sharing lessons learned especially with regards to the application of science at the fireground. Her focus is in finding a way to translate complex, peer-reviewed scientific information into practical, on-the-ground feet-in-the- 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

ash applications. She currently resides in Edson, Alberta, Canada with her partner Travis and their two Australian cattle dogs.

54. Developing and Implementing Geospatial Data Collection of Fuel Treatments, Lessons Learned. Justin Shedd, Research Associate/GIS Specialist, North Carolina State University - Center for Geospatial Analytics Additional Authors: Dan Hurlbert; GIS Specialist, NPS - Shenandoah National Park Caroline Noble; Lead Fire Application Specialist, Wildland Fire Management Research Development and Application, US Forest Service Ross Meentemeyer; Director Center for Geospatial Analytics, North Carolina State University Skip Edel; Geospatial Fire Analyst, NPS - National Interagency Fire Center Abstract: The National Park Service Fire and Aviation Management (NPS) has mandated the collection of geospatial data representing fuels treatments to facilitate effective Fire Management Strategies. Making this declaration was easy; educating fire staff with the knowledge and tools to collect the identified data was a challenge. This presentation will discuss challenges faced, issues resolved and offers best management practice recommendations. Capturing geospatial data of fuel treatments enables NPS Fire and Resource Managers to understand ecological effects, enables accurate analysis that aides in wildfire response, and thus enhances park management. Furthermore, the ability to communicate the approximate location of planned treatments internally to park staff and to external stakeholders like park cooperators and adjacent homeowners, fosters a collaborative environment to achieve a successful fire program. Treatment data when coupled with ‘values’ data, provides a clear understanding of why a treatment was needed and helps explain the Park’s overall Fire Management Plan. Given the need for geospatial data of fuels treatments, we worked with the three eastern Regions (NCR, NER, and SER) of the National Park Service to identify an iterative approach of research and development to capture the required data followed by training and technology transfer. Local input into system development, especially when involving new technology is vital for successful adoption. Commonalities (management goals and personnel arrangements) across multiple park units were identified by interviewing “boots on the ground”. These commonalities were hen grouped into different ‘audiences’ with a unique workflow developed for each. Development of a workflow that combined existing conditions while leveraging web-based GIS technologies was paramount. Maturing web-based GIS technologies has removed many of the bottlenecks associated with the capture of geospatial data. However, to benefit from these technologies Park Fire staff had to be empowered. Our pedagogical approach involved real world examples facilitating a complete understanding of the need for geospatial data, thus encouraging ownership of the data. Implementation of the workflows developed has been successfully achieved at all three eastern Regions. The National Fire and Aviation Management Office of the National Park Service is adopting the developed workflows, enhancing them where changes in technology have occurred. Keywords: Fuels Treatment, Geospatial, Management Practices, Lessons Learned Presenter Bio: Justin Shedd is a Research Associate/Fire GIS Specialist at North Carolina State University’s Center for Geospatial Analytics. Since 2005 he has leveraged his proximity to the vast spectrum of geospatial students, scientists, and researchers to promote all aspects of geospatial analysis, modeling, and technologies to Fire Managers in the Eastern Regions of the National Park Service.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

55. A 72-day Probabilistic Fire Growth Simulation as a Decision Support Tool on a Large Mountain Fire in Alberta, Canada Kelsy Gibos MSc., Wildfire Management Specialist, Edson Wildfire Management Area, Forest & Emergency Response Division, Alberta Environment & Sustainable Resource Development Additional Authors: Neal McLoughlin, Wildfire Management Unit Lead, Alberta Agriculture and Forestry Abstract: Lightning ignited the Spreading Creek Fire on 3 July 2014 in the Rocky Mountain front ranges of west-central Alberta. The ignition location fell within an approved forest management plan which specified fire as the desired disturbance regime, making it an interesting fire management dilemma. The management strategy needed to strike a balance between doing nothing which may lead to a very large fire and sending in full suppression resources which may have been damaging ecologically and dangerous to firefighting staff. The fire started within a prescribed burn unit with a burn containment strategy that was dependent on wet, dense stands of Engelmann spruce on north-facing slopes to act as natural barriers to fire spread. Enthusiasm to let this fire run its course was subdued by climatology information that suggested critically dry fuel conditions would persist for several weeks. Fire behavior observations validated that these forest stands were no longer barriers to fire spread with large fire runs on July 4th and 10th, each moving approximately 5 kilometers across the landscape. Suppression efforts were ineffective during these spread events due to fire intensities in excess of 10,000 kW/m and 0.5 kilometer spotting. Three Fire Behaviour Analysts (FBANs) were deployed to the fire, one of whom prepared a 72-day probabilistic fire growth simulation that became an important decision-support tool in how the Spreading Creek Fire was managed over the remainder of the fire season. It helped incident managers decide on the best plan of action by quantifying the risk of an additional fire spread day through the next 60 days of summer. The authors of this presentation discuss their methodology and explore how this modelling approach could have influenced initial fire response decisions had it been completed on the first day of the fire. Keywords: decision support systems, fire behaviour, fire weather, firefighter safety, risk assessment, wildland fire management Presenter Bio: Kelsy Gibos is a fire behaviour specialist who has observed and studied fire behaviour in Canada, New Zealand and Australia. Her experiences abroad have ignited an interest in sharing lessons learned especially with regards to the application of science at the fireground. Her focus is in finding a way to translate complex, peer-reviewed scientific information into practical, on-the-ground feet-in-the- ash applications. She currently resides in Edson, Alberta, Canada with her partner Travis and their two Australian cattle dogs.

56. Something Wicked This Way Burns: A Wicked Fire Problem in a Coastal Oregon Town Ron Steffens, Professor, Green Mountain College Abstract: The town of Bandon, Oregon earned one tragic line in the NFPA’s “Key Dates in Fire History.” On September 26, 1936 — “Conflagration destroys 386 buildings and kills 13.” Today, one cause of this historic “wicked problem” — logging slash — is closely monitored and managed. But the key cause of the fire’s rapid spread — the widespread growth of gorse, a flammable invasive shrub, remains a fuel problem on the Oregon Coast. A third element is the periodic summer droughts resulting in landscape-scale fires in western Oregon. But when Oregon’s drought returned in 2014-15, it was likely intensified by our wickedest problem — climate change. Oregon’s gorse and global climate change are problems we’ve caused ourselves. In Bandon, it was the town’s founder who introduced gorse, his favorite Irish shrub.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

This case study compares a historic missed planning opportunity after the 1936 fire — one which might have led to managed fuels and a planned community — with with a near miss: 80 years after the town burned, a 50-acre fire north of Bandon and near a world-renowned golf resort has energized planning and fire management efforts. The fire was controlled in part by the luck of its location: it burned into irrigated turf grass on the links of Bandon Dunes. While this gorse-control initiative is in its early stages, there is a commitment to map, monitor and manage the gorse. Managing the fuel and fire behavior risk of gorse is a large challenge for this community but small compared to the massively wicked problem of climate change. Yet the actions being taken in Bandon — collaborative planning, regulations that support long-term and cost-managed solutions, plus innovative fuel management techniques such as wood vinegar production from gorse — are not so different from the large-scale changes that are being asked by climate-change adaptation initiatives. The lessons learned at a community scale might thus offer guidance as we scale up fuel-management solutions in the climate change era. Keywords: wicked problem, fuels management, collaborative solutions, climate change Presenter Bio: Ron Steffens is a professor of communications at Green Mountain College (winters) and incident commander and fire analyst (summers) at Grand Teton National Park. He's a long-time resident of Bandon, Oregon.

57. The effectiveness of large air tankers for containing wildfire ignitions Hari Katuwal, Post-doctoral researcher, University of Montana Additional Authors: Michael Hand David E. Calkin Abstract: A better understanding of effectiveness of suppression resources employed for initial attack (IA) of wildfire ignitions, such as large air tankers (LAT), is essential for effective IA and cost efficient wildfire management. Effectiveness of IA resources, especially LAT has not been empirically examined because of lack of adequate and reliable data. LAT retardant drops are intended to assist ground crew in suppressing fires while they are still small and increase the likelihood of IA success. However, it is difficult to empirically demonstrate the effects of LAT on IA success rates because the use of LATs on new ignitions is not random; LATs may be used for ignitions with the greatest growth potential. Traditional regression models could be inadequate in such circumstances. In this study, recursive bivariate probit and propensity score matching models are used to examine the effectiveness of LAT drops to contain fires while they are still small (<300 acres) by controlling for the growth potential and other fire and weather related variables. results provide some evidence that lat drops are effective in containing fires while fires are still small. however, result indicates that lat drops are not equally effective for all fires. more specifically, results suggests that lat drops are effective in containing ignitions with moderate growth potential, but less effective for containing fires with the greatest growth potential. the result also suggests that the effectiveness of the lat drops depends on the time duration from detection of the fire to the first lat drop. this analysis and results may have implications for the effective use of lat drops. Keywords: Large air tanker, wildfire suppression, initial attack effectiveness Presenter Bio: Hari Katuwal is a post-doctoral Researcher in Wildfire Economics and Non-market Valuation at The University of Montana. His areas of specialization lie at the intersection of Econometrics and applied Microeconomics with a specific focus on Natural Resource and Environmental Economics. Hari’s research is focused on providing information to support public forestland management, particularly in wildfire economics and non-market valuation of natural resources. Hari’s current research focuses on examining wildfire suppression effectiveness and understanding public preferences and effectiveness of wildfire management program. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

58. Providing Information about Uncertainty Using Probability Distributions: USDA Forest Service Wildfire Suppression Expenditure Forecasting Charlotte Ham, Postdoctoral Researcher, North Carolina State University & USDA Forest Service Southern Research Station Additional Authors: Karen L. Abt, Research Economist, USDA Forest Service Southern Research Station Forest Policy and Analysis Unit Jeffrey P. Prestemon, Research Forester, USDA Forest Service Southern Research Station Forest Policy and Analysis Unit Abstract: The federal budget for wildland fire suppression assumes a ten-year moving average process when estimating funds needed in the future, mostly due to tradition and the ease of understanding across agents with varying backgrounds. However, for the USDA Forest Service, this method authorized fewer funds than were needed in 11 of 16 years since FY 2000 and borrowing funds from other programs to cover these fire suppression expenditures result in inefficiencies and reduced effectiveness agency-wide. With the passage of the Federal Land Assistance Management and Enhancement Act (FLAME) of 2009, scientific forecasts of suppression expenditures based on human behavioral and climatic factors are produced prior to and throughout the fiscal year to provide fire managers the best available information. However, the big question is which forecasted value to select when one number is needed: the mean, the median, the upper 90% confidence level, or other depending on the consequences. FLAME forecast reports include model and parameter uncertainty using Monte Carlo simulations to produce probability distributions that provide management with the possible range. In this presentation/paper, a comparison of the forecast probability density distributions and actual suppression expenditures will be presented to demonstrate how well the forecasts perform. Keywords: forecasting suppression expenditures Presenter Bio: Charlotte Ham is a Postdoctoral Scholar/Economist with the Department of Forestry and Environmental Resources at North Carolina State University. She is collaborating with the USDA Forest Service Forest Economics and Policy Research Unit at the Southern Research Station to improve wildland fire suppression cost forecasts for the USDA Forest Service and the Department of the Interior. Her doctorate is from Colorado State University’s Graduate Degree Program in Ecology through an interdisciplinary program in Resource Economics. Her dissertation is a geospatial exploration of how people value properties in close proximity to different land uses for consideration in planning and policy analysis.

59.. What Does It Mean to Have a High Initial Attack Success Rate in Wildland Firefighting? Karen Short, Research Ecologist, USDA Forest Service, Rocky Mountain Research Station Additional Authors: Dave Calkin, Research Forester, USDA Forest Service, Rocky Mountain Research Station Mark Finney, Research Forester, USDA Forest Service, Rocky Mountain Research Station Isaac Grenfell, Statistician, RTL Networks Michael Hand, Research Forester, USDA Forest Service, Rocky Mountain Research Station Abstract: Initial attack (IA) success has long been one of the primary performance measures used by agencies with wildland firefighting responsibility in the United States (US) and elsewhere. The US federal agencies currently state that (1) they credit an IA success when an 'unwanted' wildfire is suppressed before it expands beyond 100 acres (40 ha) of forest or 300 acres (120 ha) of grass or brush, and (2) the US Forest Service and Department of Interior strive to achieve 98 percent and 95 percent initial attack success rates, respectively. Achievement of these levels is often hailed as extraordinary success, but, as we demonstrate via historical analyses and geospatial fire modeling, environmental factors alone (e.g., 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

weather, fuels, terrain) will tend to constrain the vast majority of wildfires to <300 acres, regardless of suppression activities. thus, a size-based ia success metric is a poor proxy for actual firefighting effectiveness. it is also subject to biases associated with fire detection and reporting, as we describe here. given these limitations and the fact that an emphasis on high ia success rates may be counterproductive from long-term ecological and fire-management perspectives, the federal agencies have been exploring use of other performance measures that better align with risk-based management principles. we discuss several of the proposed metrics and offer some additional evaluations and performance measures that should fit well within a risk-based wildland fire management framework. Keywords: wildland firefighting, fire suppression, initial attack, performance measures Presenter Bio: Karen Short is a Research Ecologist with the USDA Forest Service, Rocky Mountain Research Station, Human Dimensions program. She received her BSc in Wildlife and Fisheries Science from the University of Arizona and her PhD in Organismal Biology and Ecology from the University of Montana. Her work has included fire-effects research in southwestern ponderosa pine forests; development and maintenance of spatial datasets on vegetation, fuels, and fire-occurrence for several national applications; and mapping of wildfire hazard for risk assessment and other applications. One product of this work is a spatial database of wildfires in the US, now spanning 1992-2013: http://www.earth-syst-sci-data.net/6/1/2014/essd-6-1-2014.html.

60. Beyond ICS: Propositions on Managing Complex Fire Events Branda Nowell, Associate Professor, North Carolina State University Additional Authors: Toddi Steelman - Professor, University of Saskatchewan Abstract: Complex wildfire events in the wildland urban interface necessitate the coordinated involvement of multiple jurisdictions and operational areas in order to respond to a dynamic set of conditions under high levels of uncertainty. The transboundary nature of these incidents means that no host unit has the jurisdictional authority, legitimacy, or resource/technical capability to effectively assume hierarchical command and control over all aspects of the incident. The Incident Command System (ICS) provides an invaluable set of tools and concepts for managing complex incidents but it has limitations in managing horizontal coordination. Previously, we have argued that it is useful to think about incident response as a network of interdependent organizations, agencies, and individuals active in the field of operations. We refer to this network as an incident response network (Nowell and Steelman, 2013). A critical question in adopting a network perspective is “How should this network be structured in order to optimize performance across the whole network?”. This presentation will discuss the findings from a network study of the effective network structure for managing complex fire events based on the collective insights of 25 Type I incident/area commanders. Results of this research suggest that a network structure referred to in the literature as a core-periphery network holds insight for improving our ability to manage effectively across jurisdictions and discrete command operations. Keywords: incident response, social networks, inter-agency coordination Presenter Bio: Branda Nowell, Associate Professor, is an organizational-community psychologist specializing in inter-organizational relationships, social networks, incident response, and community resilience. She teaches courses in organizational behavior, change management, organizational theory, and program evaluation. She currently co-leads a research team (firechasers.ncsu.edu) focused on advancing the science of adaptive capacity toward more disaster resilient communities. Dr. Nowell’s research has received awards from both the APA Society for Community Action and Research as well as the Academy of Management Public and Nonprofit Division. Her published work appears in the American Journal of Community Psychology, Journal of Public Administration. Research and Theory, Evaluation and Program Planning, and Conflict Resolution Quarterly.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

61. Writing Incident Objectives in WFDSS: What we Know, How we can do Better Tim Sexton, Program Manager, USDA/DOI Wildland Fire Management RD&A Additional Authors: Tonja Opperman, Lead Fire Applications Specialist, Forest Service, WFMRD&A Lisa Elenz, Deputy Program Manager, WFM RD&A Tami Parkinson, Fire Applications Specialist, WFM RD&A Marlena Hovorka, Fire Technology Transfer Specialist, WFM RD&A Diane Rau, Fire Technology Transfer Specialist, WFM RD&A Abstract: For each fire incident managed beyond the initial action, federal fire management agencies must publish a decision in WFDSS (Wildland Fire Decision Support System) outlining the strategic direction for managing the fire. One of the first tasks is to assess the location and potential of the fire in order to develop “Incident Objectives” that will guide actions. Writing objectives is often done under tight timeframes with limited information. Often, the goal is to avoid constraining the incoming incident management team’s future tactics. In a perfect world, alignment would occur from the Land and Resource Management Plan to the WFDSS Incident Objectives, to the Incident Action Plan’s objectives, to the direction given to the firefighter on the ground for each operational period. Such alignment would assure line officers that tactical actions taken on the fireline are necessary to achieving strategic land management objectives. We assessed WFDSS Incident Objectives on real fires and found that they often lack specificity, are contradictory, are unclear, or do not exist. Additionally, WFDSS incident objectives do not always appear to be in alignment with the Incident Action Plan. This is of concern because it may lead to unnecessary and increased firefighter exposure, aircraft use, resource damage, or costs. We present a “What, Where, Why” method to help fire managers write clear incident objectives, with the intention of providing strategic direction to incident commanders while avoiding tactical specificity. Keywords: WFDSS, objectives, incident management, decision support Presenter Bio: Tim manages the Wildland Fire Research Development & Applications Program. Past work experience includes: District Ranger, National Fire Use Program Manager, National Fire Ecologist, FMO and Hotshot Superintendent. His assignments have taken him to California, Washington, Idaho, Oregon, Colorado, and Minnesota.

62. What is the Strategy? A Comparison of WFDSS and ICS 209. Tim Sexton, Program Manager, USDA/DOI Wildland Fire Management RD&A Additional Authors: Morgan Pence, Fire Application Specialist, USFS Abstract: Federal Agency Administrators (AAs) provide direction to fire managers regarding wildfire strategy via the Course of Action in a published Decision using the Wildland Fire Decision Support System (WFDSS). Beginning in 2015, AAs could indicate directed strategy based on the Land and Resource Management Plan for a given area using an optional “Slider Bar” with a scale from Monitor to Suppression. Fire managers report the fire suppression strategy for an incident using the Incident Command System (ICS) Incident Status Summary form 209, which was modified in 2014 to include a “Fire Suppression Strategy” breakout. Suppression strategy within the 209 can be reported by percentage in the following categories: Full Suppression, Point Zone Protection, Confine, and Monitor. Essentially the same question is posed to two different audiences (AAs and fire managers) in two different ways; by doing so does this inherently increase opportunities for misunderstanding and miscommunication about how to manage the wildfire? This presentation will highlight quantitative and qualitative comparisons between strategy as directed in WFDSS to the strategy reported in the ICS 209 for fires with WFDSS Decisions in 2015. Data will be summarized nationally, by geographic area, and incident level to convey key findings with regard to alignment of management strategies on wildfires. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Keywords: Wildfire Strategy Presenter Bio: Tim manages the Wildland Fire Research Development & Applications Program. Past work experience includes: District Ranger, National Fire Use Program Manager, National Fire Ecologist, FMO and Hotshot Superintendent. His assignments have taken him to California, Washington, Idaho, Oregon, Colorado, and Minnesota. Morgan earned her B.S. in Ecology and Conservation Biology and her Masters in Natural Resources from the University of Idaho. She first joined the Forest Service as a seasonal firefighter in 2000 on the Salmon-Challis N.F. and has worked on engines, helitack, and handcrews for multiple forests and regions. She joined the Wildland Fire Management RD&A team in the fall of 2009 and supports fire decisions and documentation with fire behavior and risk assessment tools. Morgan lives in Salem, Oregon with her husband and 2 children.

63. Introducing and Validating a New Fire Weather Index: The Hot-Dry-Windy (HDW) Index Alan Srock, Assistant Professor, St. Cloud State University Additional Authors: Joseph Charney, Research Meteorologist, U.S. Forest Service Scott Goodrick, Research Meteorologist, U.S. Forest Service Brian Potter, Research Meteorologist, U.S. Forest Service Abstract: A fire-weather index is beneficial to the fire community if two criteria are met: 1) the index must show diagnostic and prognostic value, and 2) the index must be based on sound scientific reasoning. Whether complex or simple, an effective index should be able to “pick out” days or times in which extreme or erratic fire behavior is more likely to occur. For the fire community, an index that maximizes on days when fires are likely to be dangerous, but remains relatively low before and after the main blow-up, is especially useful. Current fire-weather indices have significant limitations with one or both of the key criteria above. In this presentation, we will present a new fire-weather index that shows diagnostic and predictive potential and is based on the science of how weather affects fires. This index employs the most fundamental and basic fire-weather paradigm – hot, dry, and windy – and has therefore been named the Hot-Dry-Windy (HDW) Index. By going back to the basic principles of how the atmosphere affects a fire (e.g., heat and moisture transport, drying of fuels), we have developed a relatively simple calculation which has shown encouraging results for identifying “bad” fire days. During this presentation, we will first show the design and implementation of the HDW Index, and then follow with an analysis of a few recent major fires using the HDW Index. Keywords: fire weather index Presenter Bio: Dr. Alan Srock is an Assistant Professor at St. Cloud State University in St. Cloud, MN.

64. Testing the Hot-Dry-Windy Index for the 2015 Fire Season in the Pacific Northwest Brian Potter, Research Meteorologist, Pacific Wildland Fire Sciences Lab, USDA Forest Service Additional Authors: Scott Goodrick, Research Meteorologist, USDA Forest Service Southern Research Station Alan Srock, Assistant Professor of Meteorology, Saint Cloud State University Joseph J. Charney, Research Meteorologist, USDA Forest Service Northern Research Station Abstract: The Hot-Dry-Windy Index (HDW) is evaluated for 9 wildfires in the northwestern United States, during the summer of 2015. Four formulations of HDW are compared with daily fire growth on each fire, and the correlation of each formulation is compared to the correlation of the input meteorological properties (in the form of vapor pressure deficit, wind speed, temperature, and relative humidity.) The variations for the index and the values of the input properties examined included: surface values; average values over a layer of fixed depth; average values over the local thermodynamic equilibrium 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

height; and maximum value of each input variable within a layer of fixed depth. Only an index that shows a higher correlation with growth than the basic input variables can be considered useful. Keywords: hot;dry;windy;vapor pressure deficit; fire weather index Presenter Bio: Brian Potter studies fire-atmosphere interactions from the perspective of atmospheric dynamics. His research interests span spatial scales from the fire front (a few meters) to synoptic weather patterns affecting fire behavior and fire danger (hundreds of kilometers.)

65. Daily Relationships Between Fire Danger and Satellite-Derived Metrics of Fire Activity Across CONUS Patrick Freeborn, Research Physical Scientist, USDA RMRS Fire Sciences Laboratory Additional Authors: W. Matt Jolly, Research Ecologist, USDA RMRS Fire Sciences Laboratory Mark A. Cochrane, Professor, Geospatial Sciences Center of Excellence (GSCE) South Dakota State University Abstract: Recent analysis of fire danger indices between 1979 and 2013 reveals that much of the area across CONUS has experienced a steady increase in fire weather season lengths and/or an increase in the frequency of anomalously long fire weather seasons. Climate induced variations in fire weather season lengths will have different consequences across CONUS depending on the local relationship between fire danger, fire activity, and fire effects. Here we develop these local relationships by associating 12-years of daily gridded fire danger indices with MODIS-derived metrics of fire activity, including (i) daily counts of active fire pixels, (ii) daily sums of fire radiative power, FRP, and (iii) daily burned area. Results reveal biome level differences in the amount of FRP and burned area detected by MODIS within a given interval of fire danger. After accounting for the number of days experiencing a particular fire danger condition, these inter-biome differences are attributed to functional differences in the relationships between fire danger, the probability of fire occurrence, and the average daily extent and intensity of burning. Regardless of regional differences, MODIS is more likely to detect larger fire growth, greater FRP, and more FRP per burned area at higher fire danger. Overall these results demonstrate two pathways for achieving the same burned area: either several days of low fire growth at low fire danger, or a few days of high fire growth at high fire danger. Each of these burned area pathways will require a different fire management response, and will induce dramatically different fire effects. Moreover, if FRP is used as a proxy for fuel consumption, then these results also demonstrate the dual impact that fire danger has on pyrogenic carbon budgets: first, by expanding the opportunity for large fire growth, and second, by enabling greater fuel consumption completeness. Keywords: Energy Release Component, Daily Burned Area, Daily Fuel Consumption Presenter Bio: Patrick began his career as a wildland firefighter in 2002, and has spent three summers on a fire use module, mainly in the wilderness. Since then Patrick has received an MS in Forestry from the University of Montana, and a PhD in Geography from King’s College London. Patrick is a Research Physical Scientist at the Missoula Fire Sciences Laboratory where he is currently studying historical relationships between fire danger and fire activity.

66. Alaska Fire and Fuels System Joe Young, University of Utah Additional Authors: Robert Ziel, Alaska Interagency Coordination Center Joseph Young, University of Utah John Horel, University of Utah

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Abstract: Wildland fire managers use assessment of current fuel conditions and expected weather to anticipate potential fire growth and control difficulty and to make decisions about suppression readiness and response, tactical choices, public outreach, and resource allocation. These assessments need to consider both spatial and temporal variability in observed conditions and forecast scenarios. The Canadian Forest Fire Danger Rating System’s Fire Weather Index system is used in Alaska to integrate this information into quantitative evaluations of fire potential. Software has been developed to provide retrospective, current and forecast information regarding weather, fuel and fire behavior conditions (see http://mesowest.org/akff). The software was extensively tested during the 2015 fire season and a summary of its capabilities will be presented. Keywords: Alaska, fire, fuels, CFFDRS, FWI Presenter Bio: Joe Young has been a developer at MesoWest, SynopticLabs, and Synoptic Data since 2013. Joe's work focuses on improving MesoWest infrastructure as well as building and maintaining standalone applications using our data resources. These applications include several of our wildland fire and fuels management projects.

67. Comparison of temperature and relative humidity values from Sling Psychrometers and Electronic Weather Meters in an Controlled Environment Charles McHugh, Fire Spatial Analyst, Forest Service Rocky Mountain Research Station. Fire, Fuel, and Smoke Science Program Additional Authors: Larry S. Bradshaw, Meteorologist, Forest Service Rocky Mountain Research Station. Fire, Fuel, and Smoke Science Program Abstract: Belt weather kits for obtaining weather information on prescribed fire and wildfires have been in use since the 1960’s. The use of a sling psychrometer from these kits is standard practice for the determination of dry and wet bulb temperatures to calculate relative humidity from a corresponding set of tables. Electronic based meters for obtaining weather information could replace belt weather kits, streamline the process, and eliminate many of the errors often associated with the use of the sling psychrometer. However, anecdotal evidence from fire managers suggest relative humidity values of 5- 20% lower from electronic meters compared to a sling psychrometer. Thus contributing to a disbelief and lack of confidence in the data electronic meters could provide field personnel. We will present results of a study comparing measured temperature and relative humidity values from electronic meters -such as the Kestral®- to temperature and calculated relative humidity using a sling psychrometer. All measurements were taken in an environmental chamber allowing us to set the values across a known range of temperature and relative humidity. This study comparing values from these two types of measuring devices in a controlled environment is the first we are aware of. Keywords: weather, relative humidity, Kestrel, weather meters Presenter Bio: Charles (Chuck) W. McHugh has been a Fire Spatial Analyst for the Rocky Mountain Research Station, Missoula Fire Sciences Laboratory, since June 2002. His current research involves spatial data analysis, large air tanker use and effectiveness, geospatial fire modeling, historical burn probability analysis, and wildfire/fuel treatment interaction and effectiveness. During the fire season, he operates as a Long-term Fire Analyst (LTAN) in support of wildland fire management at the national, regional, and local levels often working with Type-1 and Type-2 Incident Management Teams in the development of long-term plans and fire analysis.

68. Benefits and Incentives for Fuels Treatment in the Mokelumne Basin Mark Buckley, Senior Economist, ECONorthwest

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Abstract: This presentation summarizes a series of related studies to understand the magnitude and distribution of costs and benefits associated with fuels treatment in the Mokelumne River Basin of the Sierra Nevada. The purpose is to identify socially efficient fuel treatments and understand needs and beneficiaries sufficient to develop a funding program for the treatments. The study, commissioned and directed by the U.S. Forest Service, Sierra Nevada Conservancy (State of California) and the Nature Conservancy involved several teams of experts to model the effects of fuel treatment on wildfire occurrence and quantify a wide range of biophysical and socioeconomic effects on valuable resources. This included estimation of effects on erosion, debris flows, sedimentation for water and electricity supply and storage, carbon, public and private structures, suppression effort, and several others. The project involved extensive participation from key stakeholders, including private timber and water and electricity utilities. During September 2015, the Butte Fire occurred within the region modeled and intended for fuel treatment as part of the study. The presentation will include updates comparing modeling results and estimates with actual occurrence of fire and associated costs identified to-date. Of particular interest is distribution of costs and liabilities across landowners and responsible entities, the driving force behind success and failure of the suite of targeted fuel treatments. Keywords: economics benefits erosion carbon structures Presenter Bio: Mark Buckley leads the natural resource practice at ECONorthwest, an economics consulting firm based in Portland. His work focuses on valuation and decision-making involving policy and projects, particularly concerning forests, water, land use and restoration.

69. Minority Households Willingness-to-Pay for Public and Private Wildfire Risk Reduction in Florida: A Latent Class Analysis Jose Sanchez, Research Statistician, USDA Forest Service Pacific Southwest Research Station Additional Authors: Gonzalez-Caban, Armando Abstract: The purpose of this work is to estimate Willingness-to-Pay (WTP) for minority (Latino and African American) homeowners in Florida for private and public wildfire risk reduction programs. A latent class (LC) model allowed us to determine whether WTP is higher for public or private actions for wildfire risk reduction, and whether households with personal experience and perceiving living in higher risk areas have significantly higher WTP. These results may help fire managers optimize allocation of scarce cost sharing funds for public vs private actions. Private actions were defensible space measures for homeowners to take on their own properties (e.g., removing flammable vegetation around homes). Public programs were described in terms of prescribed burning and mechanical removal of vegetation around the neighborhood. Data was collected using a random digit dialed telephone contact followed by a mailed survey to minority respondents providing an address. Of 1000 mailed we obtained 319 completed surveys. Results show that about 36% of respondents were classified in Class 1 (Less Experience) and 64% in Class 2 (More Experience). The Class 1 parameter estimate on risk is not significant, while loss is significant, suggesting that respondents focus their attention on losses. Further, these respondents prefer the status quo (do nothing) alternative. In contrast, respondents in Class 2 living in a low and moderate risk area have a positive WTP for 10-year public and private programs. However, Class 2 respondents living in areas classified as high-risk prefer the do nothing alternative. Surprisingly, Class 2 respondents that perceive they are living in a high-risk area have a negative WTP sign for both the public and private program; implying they would have to be compensated to participate in the programs. A plausible explanation for these findings is that the majority of respondents that perceive they live in a high-risk area are low-income households (78% have less than $37,500 annual household income). Our results suggest fire managers may need to increase outreach efforts (possibly including Spanish language 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

information) to better inform low-income minority households of the potential wildfire losses associated with living in a high-risk area. Keywords: African American households, Fire economics, fuel management, Latinos households, surveys Presenter Bio: Dr. José Sánchez has a bachelor’s degree in Economics from the University of California, Irvine and a master’s degree in Statistics from Washington State University. He graduated in June 2014 from University of California, Riverside with a PhD in Environmental Sciences. Dr. Sánchez is currently working with USDA Forest Service Pacific Southwest Research Station as a Research Statistician. His research focuses on nonmarket valuation of ecosystem services and evaluating the economic impact of wildfires on natural resources.

70. Hedonic Models for Homes Vulnerable to Wildfire David Rossi Additional Authors: Anne Byrne, Graduate student, Department of Agricultural & Resource Economics at Colorado State University Abstract: Federal investments in hazardous fuels management on public lands is suspected to be an environmental amenity accruing to nearby homeowners which counteracts the dis-amenity of possibly damaging unplanned wildfire. This research seeks to determine if the implied value of fire protection services are important to homeowner’s housing choices in wildland-urban interface communities. Fire protection can be naturally provided by a forest landscape where natural vegetative conditions limit high levels of wildfire activity, or artificially provided where human interventions have decreased the resulting intensity and spread of unplanned fire activity through the use of prescribed burning and mechanical thinning of overgrown forest fuels. We use temporally and spatial explicit data on housing transactions across Boulder County, Colorado from 2008-2015 to determine if there is a significant relationship between fire protection amenities and observed home prices. The previous 5 years of wildfires and fuel treatment activity occurring in the vicinity of each housing transaction is aggregated and used in a hedonic price model to determine if the premium on less fire-risky real estate is statistically significant over the course of the sample frame. We use this model to glean insight into the preferences homebuyers have for (a) forested land with relatively sparse development (b) a property’s proximity to historical fire activity and (c) a property’s proximity to nearby fuels management efforts. Spatially weighted specifications are compared to ordinary least squares regression results and evidence of homeowners’ risk aversion is explored. Presenter Bio: David Rossi is a graduate student specializing in Forest Policy and Economics at Colorado State University. His research has focused on optimal strategies for wildland fire management and patterns of household mitigation behavior in wildland-urban interface regions of the western U.S.

71. Systematic Investigation of Wildfire Damage and Risks on Property Values Qiuhua Ma, Economics Department, University of New Mexico Additional Authors: Brady P. Horn, Assistant Professor, University of New Mexico Jennifer A. Thacher, Associate Professor, University of New Mexico Robert P. Berrens, Professor, University of New Mexico Abstract: Wildfires have burned increasing acreage in the western United States over the past few decades. With the rapid expansion of housing in the Wildland-Urban Interface (WUI), wildfire risk exposure has increased significantly. Since most mitigation and suppression costs are borne by taxpayers, policy analysts seek both market (e.g., protection and suppression cost) and non-market cost estimates of wildfires. As one tool, the hedonic pricing method (HPM) is commonly used to investigate the relationship between housing values and wildfire characteristics. There are a variety of HPM studies 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

investigating wildfire, with mixed and/or inconsistent results. This is complicated by a variety of data availability issues as well as choices of the analyst in modeling both ex ante risk and ex post damage effects of wildfire. This analysis applies spatial econometrics modeling strategies in an HPM framework to examine the joint effect of both wildfire damage events and current risk on property values in Santa Fe County, New Mexico. The objective is to systematically investigate wildfire effects on housing values using a variety of modeling approaches, including varying or alternative measures for both damage and risk, functional forms, spatial econometric techniques and time frames. We examine how the effect of wildfire varies with data availability and choices of analyst through a comparison of alternative models. The analysis is focused on Santa Fe County, an area that is large enough to include both WUI and Non- WUI areas with multiple wildfire event and risk data available, and collected at different geographical scales. The systematic investigation includes: two dependent variables (estimated sale price and assessed property value); two commonly-used hedonic functional forms (semi-log and double-log); five spatial dependency approaches (independent, spatial lag, spatial error, spatial durbin and general); three risk measures (GIS-based composite hazard and risk assessment, WUI risk assessment and individual-level house risk assessment); and two damage measures (the nearest fire and aggregate fire damage), each with two time frames (seven and fifteen years). To structure the analysis, we test the sensitivity of econometric results for three crucial hypotheses about the effects of wildfire. Keywords: wildfire damage, wildfire risk, hedonic price model, spatial econometrics Presenter Bio: Qiuhua Ma is a PhD candidate in the Department of Economics at the University of New Mexico. She is an applied macroeconomist specializing in environmental and natural resource economics with a focus on water and wildfire. She systematically studies the effect of wildfire damage and risk on property values using a variety of modeling approaches, including varying or alternative measures for both damage and risk, functional forms, spatial econometric techniques.

72. The Effect of Wildfires on Recreation Visitation: A Historical Analysis of the National Park Service’s Intermountain Region Kara Walter, PhD Candidate, University of New Mexico Abstract: With the incidence of wildfires increasing annually, the ability to accurately estimate the impact of fires is becoming increasingly important. One of the potential impacts is on the use of wildlands. While the extant literature has estimated the willingness to pay for outdoor activities (primarily hiking) following a wildfire, few studies consider the aggregate impacts of wildfire on visitation to natural areas. We develop a study utilizing visitation data from National Parks Service’s Intermountain Region from 1980-2013. With data from 80 parks in 8 states, we examine the effects of wildfires on visitation, including how visitation varies over time. That is, are the effects long-lasting? Combining visitation data with fire occurrence, location (relative to the park), and size, we estimate the impact of fire on visitation. Our panel data allows us to control for lag effects of fire, as well as climatic impacts, and other economic and location variables. Preliminary analysis suggests the impact of wildfire is park specific, and thus the economic impact on a region will depend on a number of factors. This research adds to the body of knowledge of economic impacts of wildfires, which can help inform management and policy decisions. Keywords: economics, wildfire, recreation, NPS Presenter Bio: Kara Walter is a Ph.D. candidate at the University of New Mexico studying economics. She is currently completing her dissertation on the economics of wildfires and intends to graduate Spring 2017. She is an applied microeconomist studying is environment and natural resource economics. In 2013 she earned her masters in economics from UNM and in 2011 graduated cum laude from the University of Southern Indiana with a degree in economics and finance.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

73. Assimilation of satellite active fires detection into a coupled weather-fire model Jan Mandel, Professor, Department of Mathematical and Statistical Sciences, University of Colorado Denver, Denver, CO, USA Additional Authors: Mary Ann Jenkins, Associate Professor, York University, Toronto, ON, Canada Adam K. Kochanski, Research Assistant Professor, University of Utah, Salt Lake City, UT, USA Sher Schranz, Associate Director, Cooperative Institute for Research in the Atmosphere, Colorado State University and NOAA, Boulder, CO, USA Martin Vejmelka, Scientist, Institute of Computer Science, The Czech Academy of Sciences, Prague, Czech Republic Abstract: Currently available active fires detection products from the VIIRS and MODIS instruments on polar-orbiting satellites provide a data source suitable for wildland fire simulations. However, there is no detection under cloud cover, false negatives are common, and the resolution is much coarser than in a fire behavior model. Consequently, the data should be used to improve the fire spread in a statistical sense only, rather than as an input to start the simulation from. An additional challenge in a coupled atmosphere-fire model is to keep the states of the fire and the atmosphere consistent, because the atmospheric circulation due to the fire forcing develops over time only. We describe a data assimilation method for active fire detection from satellites. The state of the fire spread in WRF-SFIRE is encoded as a map of the fire arrival time on the simulation domain. The analysis step minimizes the change in the fire arrival time, treated as a spatial random field, and maximizes the likelihood of the fire detection data. The data likelihood takes into consideration the level of confidence of the detection as well as the cloud mask, which indicates where no fire detection data are available. The minimization problem has multiple solutions, essentially due to the fact the likelihood of fire detection at a location can be low either because the fire has not arrived there yet, or it was there already and is no longer detectable. A preconditioned descent method is devised that makes a spatially smooth correction to the fire arrival time, which avoids getting trapped in local minima and overfitting the data. The method converges very fast and few descent iterations are generally sufficient. After the new fire arrival time is found, the atmospheric model is rerun from a time in the past to the time, using the new fire arrival time to generate the fire forcing and to spin up the atmospheric model into a consistent state. The coupled fire-atmosphere simulation then continues from the state modified by the analysis step. The performance of the algorithm will be demonstrated on real fire simulations and satellite data. Keywords: MODIS, VIIRS, active fires detection, data assimilation, WRF-SFIRE Presenter Bio: Jan Mandel is a computational scientist and mathematician working on the design and rigorous mathematical analysis of fast numerical solution methods for problems in science and engineering on parallel supercomputers. His methods were used in the design of the Swedish Grippen jet fighter and, more recently, in the analysis of the earthquake damage to Fukushima nuclear reactors. Since the mid-2000s, he has been interested the design and probabilistic analysis of data assimilation methods and in the numerical simulation of wildland fires. He has contributed to WRF-SFIRE, and he has been coordinating the software development of SFIRE.

74. Evaluation and improvement of an advanced regional modeling framework, addressing effects of wildfire emissions on modeled air quality for the Pacific Northwest Vikram Ravi, PhD student, Washington State University Additional Authors: Serena Chung, Research Associate Professor, Washington State University Farren Herron-Thorpe, Emissions Inventory Specialist, Washington Department of Ecology Joseph Vaughan, Research Associate Professor, Washington State University 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Brian Lamb, Regents Professor, Washington State University Kelley Barsanti, Assistant Professor, University of California Riverside Lindsay Hatch, Postdoctoral Research Associate, University of California Riverside Abstract: Emissions from forest fires can cause significant deterioration in air quality far from the source through production of fine particulate matter (PM2.5) and ozone (O3). Both PM2.5 and O3 are formed by oxidation of pollutants emitted from fires; PM2.5 is also directly emitted. In general, performance of regional air quality modeling systems is worse for PM2.5 compared to O3. Improved forecasting of air quality downwind of wildland fires is needed for informing the regulatory agencies, health services and the public. Model improvement through evolving science and better emission inventories improves forecasting. Here, reliability of models is judged through various types of evaluation. AIRPACT-5, a WRF-SMOKE-CMAQ system with BlueSky fire emission modeling framework, is an updated version of the AIRPACT-4 modeling system providing air quality forecasts for the Pacific Northwest. Recent evaluation of AIRPACT-5 for O3 episodes in July 2012 (no fire period) and August 2012 (fire period) using observational data from Environmental Protection Agency’s Air Quality System (AQS) shows improved performance for O3 compared to AIRPACT-4. Results from further evaluation of AIRPACT-5 including PM2.5 will be presented for the fire season of 2013. Because previous analysis suggests that BlueSky with the Fuel Characteristic Classification System (FCCS) fuel map under-predicts PM2.5 emissions, further evaluation will include sensitivity simulations with an updated fire emission inventory using modified fuel loadings. Updated speciation profiles for non-methane organic compounds based on the Fourth Fire Lab at Missoula Experiment (FLAME-IV) will be applied to quantify potential missing sources of SOA in the current model. In addition to IMPROVE and AQS data, observational data from the Biomass Burning Observation Project (BBOP) aircraft campaign will be used for evaluating model performance for secondary organic aerosols (SOA) formation rate. Keywords: Bluesky, Wildfire, air quality modelling, model evaluation, secondary organic aerosol Presenter Bio: Vikram Ravi is a PhD student in Engineering Science with Environmental Engineering focus at Washington State University. He uses regional models to study the air quality impacts of prescribed fire emissions, aviation biofuel industry, regional haze modeling, also works on model performance evaluation.

75. The importance of biomass burning feedbacks: Focus on CALIOP-based estimates of smoke plume injection height Amber Soja, Associate Research Fellow, NIA / NASA Additional Authors: Dr. Hyun-Deok Choi, National Institute of Aerospace, NASA LaRC Dr. Thomas Duncan Fairlie , NASA Langley Research Center Dr. Mark Vaughan, NASA Langley Research Center Dr. David Winker, NASA Langley Research Center Dr. Charles Trepte, NASA Langley Research Center Dr. George Pouliot, Environmental Protection Agency Dr. James J. Szykman, Environmental Protection Agency Abstract: There is a significant connection between biomass burning (BB) emissions, the terrestrial environment and the atmosphere, which has strong implications for feedbacks to the climate system and Air Quality. BB has the potential to alter numerous land and atmospheric processes that, in turn, feedback to and interact with the climate system (i.e. black carbon on spring Arctic ice, land-vegetation cover change alters albedo). Specifically, the heights to which BB smoke is injected governs short- or long-range transport, which influences surface pollution, cloud interaction (altered albedo), and modifies patterns of precipitation (cloud condensation nuclei). We are working with the Cloud-Aerosol Lidar and Infrared Pathfinder 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Satellite Observation (CALIPSO) science team and other Applied Science partners, primarily the Environmental Protection Agency and regional partners, to generate BB plume injection heights using multiple platforms, sensors and models (CALIOP, MODIS, NOAA HMS, Langley Trajectory Model) that will provide value to national and international scientific and air quality communities, the CALIPSO science and algorithm teams, and to public land, fire, and air quality management and regulations communities. Specifically, we have: (1) developed a methodology that links BB injection height and CALIOP air parcels to specific fires; (2) connected the variables that control these dynamics, which include ecosystems, fire- specific and meteorological variables; and (3) defined the daily evolution of smoke plumes for specific fires. Statistics that link fire behavior and weather to plume rise are crucial for verifying and enhancing plume rise parameterization in local-, regional- and global-scale models used for air quality, chemical transport and climate. Finally, we will compare the daily CALIPSO-derived injection height estimates to CMAQ modeled injection height, in an effort to improve smoke plume height injection estimates. Keywords: smoke injection, remote sensing, aerosols, climate, fire feedbacks Presenter Bio: Amber’s research interests focus on connections between fire regimes, the atmosphere and biosphere, and feedbacks to and from the climate system. She has two decades of research experience, where she has taken part in and led numerous national and international teams of research scientists. Specifically, she uses Geographic Information Systems (GIS) and satellite-derived data as tools to explore these dynamic relationships. Dr. Soja is currently an Associate Research Fellow at the National Institute of Aerospace, resident in Climate Sciences at NASA LaRC. She has recently taken a part-time Associate Program Manager position in the NASA Applied Sciences Program, Wildland Fire.

76. Field-Scale Validation of Data-Driven Wildland Fire Spread Simulations Cong Zhang, graduate student, University of Maryland, College Park Additional Authors: Mélanie Rochoux, Research Scientist, CERFACS, SUC-URA1875, CNRS, 42 Avenue G. Coriolis, 31057 Toulouse, France Wei Tang, PhD student, Department of Fire Protection Engineering, University of Maryland, College Park, MD, 20742, USA Maria Theodori, MS student, Department of Fire Protection Engineering, University of Maryland, College Park, MD, 20742, USA Michael Gollner, Assistant Professor, Department of Fire Protection Engineering, University of Maryland, College Park, MD, 20742, USA Arnaud Trouvé, Professor, Department of Fire Protection Engineering, University of Maryland, College Park, MD, 20742, USA Abstract: The objective of this project is to demonstrate the benefits of data-driven wildfire spread modeling using ensemble-based data assimilation methods combined with sensor observations of the position of the fireline in order to provide a wildfire spread forecasting capability. The present data- driven modeling approach is similar to that used for weather forecasting applications. Our wildfire spread simulation prototype called FIREFLY-EnKF features the following components: a front-tracking fireline solver that adopts a regional-scale viewpoint, treats wildfires as propagating fronts, and uses a description of the local rate of spread (ROS) of the fire as a function of biomass fuel, topographical and meteorological properties (based on the Rothermel model); a series of observations of the fire front position; and a data assimilation algorithm based on an ensemble Kalman filter (EnKF) that can correct ROS parameters in order to decrease the distance between simulated and observed fire front positions. In this work, FIREFLY-EnKF is extended to a spatially-distributed parameter estimation approach to cope with the spatial variations of environmental conditions (in particular near-surface wind conditions). The potential of FIREFLY-EnKF for dynamically estimating the interactions between the near-surface wind 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

and the fire is evaluated in simulations of two field-scale prescribed-burn experiments: a 30-hectare fire experiment called FireFlux I (Clements et al. 2007), and a 2-hectare fire experiment called RxCADRE S5 (O’Brien et al. 2015). In both cases, data assimilation is shown to improve greatly the accuracy of the fire model and in particular its forecasting capability. References Clements CB, Zhong S, Goodrick S, Li J, Potter BE, Bian X, Heilman WE, Charney JJ, Perna R, Jang M, Lee D, Patel M, Street S, Aumann G (2007) Observing the dynamics of wildland grass fires: FireFlux - A field validation experiment. Bulletin of the American Meteorological Society 88, 1369–1382. doi:10.1175/BAMS-88-9-1369. O’Brien JJ, Loudermilk EL, Hornsby B, Hudak AT, Bright BC, Dickinson MB, Hiers JK, Teske C, Ottmar RD (2015) High-resolution infrared thermography for capturing wildland fire behaviour: RxCADRE 2012. International Journal of Wildland Fire, doi:10.1071/WF14165 Keywords: wildfire spread simulation, data assimilation, wind/fire interactions Presenter Bio: Cong Zhang is currently a PhD student in the Department of Fire Protection Engineering at University of Maryland, College Park. His research interests include wildfire spread, data assimilation, and fire-atmosphere interactions.

77. Effects of Forest Canopy on Atmospheric Turbulence During Wildland Fires Warren Heilman, Research Meteorologist, USDA Forest Service, Northern Research Station Additional Authors: Xindi Bian, Meteorologist, USDA Forest Service, Northern Research Station, Lansing, MI Kenneth L. Clark, Research Forester, USDA Forest Service, Northern Research Station, New Lisbon, NJ Nicholas S. Skowronski, Research Forester, USDA Forest Service, Northern Research Station, Morgantown, WV Daisuke Seto, PhD Student, Geography Department, University of Canterbury, Christchurch, New Zealand Craig B. Clements, Associate Professor, Department of Meteorology and Climate Science, San Jose State University, San Jose, CA John L. Hom USDA Forest Service, Northern Research Station, Newtown Square, PA Michael R. Gallagher, Forestry Technician, USDA Forest Service, Northern Research Station, New Lisbon, NJ Abstract: The atmospheric environment surrounding wildland fires is highly turbulent, with fire-fuel- atmosphere interactions producing turbulent circulations that can affect fire spread, heat transfer, and smoke dispersion. The presence of forest overstory vegetation surrounding and in the vicinity of wildland fires further complicates atmospheric turbulence regimes that develop in response to those fires. Through a number of recent field experiments involving in situ monitoring of high-frequency temperature and wind speed fluctuations during wildland fires occurring in forested environments, we have gained new insight into how fires and forest overstory vegetation can concurrently affect turbulent circulations and their role in heat transfer and the directional mixing of smoke plumes. This study focuses on the atmospheric turbulence regimes that developed during prescribed backing- and head-fire experiments conducted in forested environments in the New Jersey Pinelands National Reserve. Using sonic anemometers and thermocouples mounted at multiple levels on towers situated within 7-107 ha burn plots in the Pinelands, wind and temperature measurements were carried out as prescribed fire lines spread through the plots and through the tower locations. The effects of forest overstory vegetation and fire intensity on turbulent kinetic energy variability, turbulence anisotropy, the spectral characteristics of the three-dimensional turbulent wind fields, and horizontal and vertical turbulent heat fluxes were analyzed. Key results from the analyses suggest that maximum fire-induced turbulence during wildland fires may occur in an atmospheric layer above the forest canopy instead of near the 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

surface, forest overstory vegetation affects the vertical variation in turbulence anisotropy, and the horizontal component of the total turbulent kinetic energy dominates the vertical component primarily at large turbulent eddy sizes. These findings can facilitate the development of appropriate parameterizations of atmospheric turbulence effects on fire spread and smoke dispersion in operational fire behavior and air-quality prediction systems. Keywords: atmospheric turbulence, forest canopy, wildland fires, fire behavior, smoke dispersion Presenter Bio: Dr. Warren E. Heilman is a Research Meteorologist with the USDA Forest Service - Northern Research Station in Lansing, MI. He graduated from South Dakota State University and Iowa State University with undergraduate and graduate degrees in physics and meteorology, respectively. His current research is focused on examining fire-fuel-atmosphere interactions, turbulence regimes, and local dispersion of smoke during wildland fire events.

78. Black Carbon Production and Storage as a Result of Differing Fire Frequencies in Longleaf Pine Forests Adam Coates, Doctoral Candidate, Clemson University Additional Authors: Dr. Alex Chow, Assistant Professor, Clemson University Abstract: Black carbon is broadly defined as a continuum of byproducts resulting from the incomplete combustion of biomass and fossil fuels.. It is currently unclear how prescribed fire frequency affects the production and storage of black carbon in forest stands. To investigate this issue, a study has been implemented in Georgetown, South Carolina to investigate the production and storage of black carbon as a result of differing fire frequencies over the last 12 years in a longleaf pine (Pinus palustris) forest. These frequencies are as follows: unburned, burned 4 times, burned 6 times, and burned 8 times with 1- 3 year intervals between each fire. To evaluate black carbon production, litter samples were collected prior to burning in 2015. Dormant season burns were conducted in two of these forests in 2015 (6 times burned and 8 times burned). In these areas, ash samples were collected after burning. These samples will be evaluated using analytical pyrolysis/GC-MS to characterize the constituents within each sample, with a particular interest in polycyclic aromatic hydrocarbons (PAHs) as they tend to be most associated with black carbon. To evaluate storage, soil samples were taken at two depths: 0-10 cm and 10-20 cm prior to and immediately after burning. Soil organic matter from these samples will be extracted and fractionated into humic and fulvic acids to be freeze-dried and subjected to analytical pyrolysis/GC-MS, similarly to the litter and ash samples. Based upon our observations, we hypothesize that: 1. Increasing fire temperatures will result in an increased production of PAHs. 2. Due to an increase in fuel and thus higher burning temperatures, production will be highest in forests that have been burned 6 times in the last 12 years as opposed to forests that have been burned 8 times in the last 12 years. 3. Storage of PAHs will be greatest in soils that have been burned most frequently (8 times in the last 12 years) at both soil depths. These results, and others that will be presented, provide important information related to global carbon cycling and sequestration. Keywords: black carbon, prescribed fire, soils, longleaf pine Presenter Bio: Adam Coates is a Doctoral Candidate at Clemson University's Baruch Institute of Coastal Ecology and Forest Science in Georgetown, South Carolina. His research has focused on the effects of forest restoration practices on ecosystem processes, properties, and functions in the southeastern Piedmont, southern Appalachian Mountains, and coastal Longleaf Pine forests. His current work is focused on black carbon production and storage as a result of prescribed fire in coastal forests.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

79. Recovering Lost Ground: Effects of Soil Burn Intensity on Nutrients and Ectomycorrhiza Communities of Ponderosa Pine Seedlings Ariel D. Cowan, Oregon State University

Additional Authors: Jane E. Smith, Research Botanist, USFS Pacific Northwest Research Station Stephen Fitzgerald, Extension Silviculture Specialist, Oregon State University Abstract: Fuel accumulation and climate shifts are predicted to increase the frequency of high-severity fires in ponderosa pine (Pinus ponderosa Dougl. ex Laws) forests of central Oregon. The combustion of fuels containing large downed wood can result in intense soil heating, alteration of soil properties, and mortality of microbes. Previous studies show ectomycorrhizal fungi (EMF) improve ponderosa seedling establishment after fire but did not compare EMF communities at different levels of soil burn intensity in a field setting. For this study, soil burn intensity effects on nutrients and EMF communities were compared at Pringle Falls Experimental Forest, La Pine, Oregon. Twelve replicate sites were used, each with three treatments: high intensity soil burn from large downed wood combustion (HB), low intensity soil burn (LB), and unburned control (UB). Temperatures lethal to fungi were recorded at 0-cm, 5-cm, and 10-cm depths in HB soils and 0-cm depth in LB soils. Ponderosa pine seedlings planted post-burn were harvested after four months for EMF root tip analysis. We found: a) greater differences in soil properties and nutrients in HB soils compared to LB and UB soils; b) no differences in EMF richness and diversity; c) weak differences in community composition based on relative abundance between UB and either burn treatments; and d) EMF composition in HB and LB treatments correlated with low carbon and organic matter contents. These results confirm the combustion of large downed wood can alter the soil environment beneath it. However, an EMF community similar to LB soils recolonized HB soils within one growing season. We theorize that quick initiation of EMF recolonization is possible depending on the size of high burn patches, proximity of low and unburned soil, and survival of nearby hosts. The importance of incorporating mixed fire effects in fuel management practices will help to provide EMF refugia for ponderosa pine forest regeneration. Presenter Bio: Hailing from Brooklyn, NY, Ariel Cowan received her master's in Forest Ecosystems & Society from the College of Forestry at Oregon State University. Her research interests include ectomycorrhizal fungi and fire ecology. She has worked in California oak woodlands, coastal chaparral, sagebrush steppe, pinyon-juniper woodlands, as well as dry inland ponderosa forests.

80. Basal duff smoldering beneath old pines: a distinctive pattern of ground combustion Jesse Kreye, Postdoctoral Research Associate, Forest and Wildlife Research Center, Mississippi State University Abstract: Fires that burn with prolonged smoldering combustion often occur in long-unburned forests that develop a deep forest floor. These so-called “duff fires” that burn through lower forest floor horizons (Oe and Oa “duff”) are often implicated in post-fire stress and mortality in large conifers, long- duration soil heating, and noxious emissions in many temperate forests. Concentrated smoldering and subsequent fuel consumption near the bases of large trees is common across many fire-excluded coniferous ecosystems. We present data on duff ignition and consumption patterns from small-scale and operational-scale burn studies in long-unburned Pinus palustris forests in northern Florida, USA. In small-scale burns, ignition probability and duff consumption were greatest near tree bases, despite moisture condition or presence of ignition vectors. Similar patterns were observed during operational- scale prescribed burns. Oe horizons dried throughout the 45 days of small-scale experiments, while Oa moisture tended to be more constant. Duff adjacent to tree boles, however, was deeper and typically drier than beyond duff mounds. Oa horizons were least dense at the base of trees, whereas Oe bulk 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

density did not differ with distance to trees. Heterogeneous patterns of duff moisture and bulk densities may play a role, but their relative importance in the ignition patterns observed here are unknown. Keywords: ecosystem restoration, fire exclusion, forest floor, Pinus palustris, prescribed fire, smoldering combustion.

Presenter Bio: Jesse Kreye is a Postdoctoral Research Associate at the Forest and Wildlife Research Center at Mississippi State University. His experience in fire has included fire suppression, prescribed burning, as well as teaching and research in fire and fuels management and fire ecology. He has worked for the Minnesota Department of Natural Resources, the US Forest Service (California), Humboldt State University, and the University of Florida.

81. Quantifying Emission Factors from Smouldering Peat Fires: a Laboratory Study Rory Hadden, Rushbrook Lecturer in Fire Investigation, University of Edinburgh Additional Authors: Paolo Pironi, Postdoctoral Research Associate, University of Edinburgh Simon Santamaria, PhD Student, University of Edinburgh Guillermo Rein, Reader in Thermal Energy, Imperial College London Abstract: Smouldering wildfires are the most persistent fires on earth. These fires occur in large deposits of peat across the globe in boreal and temperate regions and consume large quantities of biomass which can take centuries or longer to regenerate. At the time of writing, there are ongoing large peat fires in Indonesia which have been acknowledged as causing significant health issues across a large geographic area while causing the loss of a valuable ecosystem. Similar events have occurred in Russia (2010) and Indonesia (1997). During the latter event, an estimated 13.7Gt of carbon was estimated to have been emitted to the atmosphere. Globally, the carbon stored in peatlands is greater than that stored in vegetation and is similar to that stored in the atmosphere. When peatlands become dry (through drainage or climatic conditions) the become susceptible to smouldering fires. Smouldering can be ignited easily in peat and the resulting fire can persist for extended periods of time (often many weeks or months) as the peat is consumed laterally and in depth. The loss of carbon from peatlands has been acknowledged as a significant contributor to the global carbon balance. This paper presents an overview of the lab-scale measurements of the carbon dioxide, carbon monoxide and methane flux from samples of smouldering sphagnum moss peat. This material is used as it has been previously studied extensively in the laboratory and numerically. By using the FM Global Fire Propagation Apparatus, a repeatable experimental condition can be defined in which the flux of CO, CO2 and methane can be linked to the smouldering fire dynamics. The apparatus allows simulation of the fire behavior and measurement of emissions in deep (i.e. low oxygen availability) and shallow (high oxygen and high heat loss) smouldering. Conditioning of the peat allows for consideration of the effect of fuel moisture content on the emissions. Consequently, this work presents the emission factors from smouldering peat across a broad range of conditions that may be experienced in real peat fires. This provides valuable data for fire-climate interaction models and the potential for improved estimates of carbon release during smouldering wildfires. Keywords: smouldering, carbon, emission factors, peat fires Presenter Bio: Rory Hadden is the Rushbrook Lecturer in Fire Investigation at the BRE Centre for Fire Safety Engineering, University of Edinburgh. He received his PhD from the University of Edinburgh in 2011 and continued to postdoctoral position at The University of Western Ontario, (Canada) and Imperial College London (UK). His research interests include solid-phase fire phenomena including, ignition of fuels, smouldering combustion and flame spread and he is active in developing improved 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

instrumentation techniques for large-scale fire experiments with particular application to model input data.

82. Flammability of North America Pines J. Morgan Varner, Assistant Professor, Department of Forest Resources & Environmental Conservation, Virginia Tech Additional Authors: Erin M. Banwell, Bureau of Land Management, Roseburg, OR Jeffrey M. Kane, Department of Forestry & Wildland Resources, Humboldt State University Erik S. Jules, Department of Botany, Humboldt State University Abstract: Pinus species are important components of fire-prone and fire-driven ecosystems throughout the northern hemisphere and their litter is a driver of surface, ground, and canopy fire behavior. In spite of the assumptions related to pine flammability, little data are available that quantify among species variation and no substantive work addresses the underlying mechanisms of these differences. We evaluated burning characteristics and physical leaf characteristics of 33 pine species collected throughout North America. For each species, seven 15 g replicates were burned under controlled laboratory conditions. Average fuelbed height (cm), maximum flame height (cm), flaming duration (s), smoldering duration (s), and residual mass (g) were all measured. These metrics were combined in a Principal Components Analysis that explained ca. 80% of the variation in flammability. At the time of each burn, moisture content of the litter was less than 3%. Leaf length, leaf thickness, and weight of fascicle were measured for each pine species collected. Among species burned, P. serotina, P. rigida, and P. washoensis, burned with greatest intensity, generating longer average maximum flame heights (mean = 87.2, 85.6, and 84.9, respectively). P. balfouriana, P. longaeva, and P. monophylla burned with the shortest average maximum flame heights (mean = 20.4, 25.7, and 28.1 cm, respectively). Among the physical needle traits measured, only needle length was related to flammability in a piece-wise relationship (R2=0.65). Unlike similar research in other species, the best predictors of flammability were linked to phylogeny, with clades having similar flammability regardless of physical traits. Keywords: flammability, laboratory fire behavior, litter, phylogeny, pines. Presenter Bio: Morgan Varner is Assistant Professor of Fire Ecology & Management at Virginia Tech. His research and teaching focus on fire ecology, fuels management, and natural history of fire-prone ecosystems. He serves as Chair of the Coalition of Prescribed Fire Councils and focuses outreach efforts on reducing impediments to prescribed fire.

83. Can Wildfire Restore Conifer-encroached California Black Oak Woodlands? Deborah Nemens, Virginia Tech Additional Authors: J. Morgan Varner, Assistant Professor, Department of Forest Resources & Environmental Conservation, Virginia Tech Kathryn R. Kidd, Post-doctoral Associate, Department of Forest Resources & Environmental Conservation, Virginia Tech Brian Wing, Research Forester, USDA Forest Service Pacific Northwest Research Station Jesse K. Kreye, Post-doctoral Researcher, Department of Forestry, Mississippi State University Nicole Vaillant, USDA Forest Service Pacific Southwest Research Station Abstract: Oak communities in the western United States have been in decline since the advent of large- scale fire exclusion in the early 20th century. In the absence of fire, woodlands that formerly supported oaks and the wildlife dependent on them are increasingly dominated by shade-tolerant conifers that 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

encroach on remnant oaks, reducing oak vigor. One potential avenue for restoration of California black oak (Quercus kelloggii) in mixed-conifer forests is via resprouting following high severity fires. We examined California black oak response across a spectrum of fire severities following two mixed-severity wildfires that burned over approximately the same landscape in 2000 (Storrie Fire) and again in 2012 (Chips Fire) in the Lassen National Forest in northern California. Ninety-three plots were established across the landscape burned by both wildfires. Plots were established in each of 16 Storrie-Chips fire severity combinations, ranging from unburned to high-severity. Over ninety-nine percent of oaks that sprouted following the Storrie fire and were top-killed in the Chips fire later resprouted. Sprout vigor was greatest in the moderate and high severity strata and diminished in sites that burned with the lowest severity. Overstory composition shifted from conifer to oak dominance with increasing fire severity. Our results will enable managers to prioritize restoration actions after wildfires in mixed conifer-oak communities. Keywords: Cascades, ecological restoration, oak woodlands, sprouting, wildfire Presenter Bio: Deborah Nemens is currently pursuing a Master’s degree in forestry with a focus in fire ecology at Virginia Tech. She hold a bachelor’s degree in plant ecology from Haverford College in Pennsylvania. Over the course of her career, she has in the fields of habitat restoration, rare species conservation, invasive species management, sustainable rural development, and prescribed fire management. Her research interests include the effects of fire on vegetation, conservation of fire- dependent species and habitats, and the use of prescribed fire as a restoration tool in degraded ecosystems.

84. The MesoWest/Synoptic Web Service: A Tool for Accessing Fire Weather Data Joshua Clark, Research Associate, MesoWest, University of Utah Abstract: MesoWest software to access, archive, and disseminate environmental information relevant to fire professionals in the United States and Canada has been extensively updated recently (see http://mesowest.org). MesoWest has been providing access to weather information for fire weather applications for nearly two decades. Over 40 million observations are added and archived each day from over 40,000 locations, including observations from permanent and temporary deployments of Remote Automated Weather Stations (RAWS). While the legacy map, graphical, and tabular interface software (MesoWest, http://mesowest.utah.edu, and ROMAN, http://raws.wrh.noaa.gov) continue to be used extensively by fire weather professionals, these were designed by necessity as “one size fits all” tools to meet common needs of operational, commercial, academic, and public users. The MesoWest development team at the University of Utah is now collaborating with software developers at Synoptic Data to expand access to environmental information around the world. To simplify access to both recent and archived data, the MesoWest/Synoptic Application Programming Interface (API) is now available to allow users to develop their own customized queries to obtain the environmental information of interest to them. Fire professionals may now use to the API to access observations in the vicinity of specific wildfires, obtain alerts when conditions change in selected areas, or design their own fire weather monitoring tools. Keywords: fire weather, fire climatology, MesoWest, weather observations, wildfire monitoring Presenter Bio: Joshua Clark is a developer of the MesoWest/Synoptic API. His previous work at the University Center for Atmospheric Research as a software engineer intern led him to create the MesoPy library to support python developers needing quick access to weather data based on the API. Additionally, Josh has worked as a fire meteorologist intern for the Bureau of Land Management in the Rocky Mountain GACC where he developed an iPhone application for local fire managers and provided weather support to seven large fires during the 2014 fire season while obtaining his Firefighter II certification.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

85. A Novel Wildfire Prediction Tool Utilizing Fire Weather and Machine Learning Methods Leo Deng, Portland State University Additional Authors: John Saltenberger, Fire Weather Program Manager, Northwest Interagency Coordination Center Marek Perkowski, Professor, Portland State University

Abstract: A novel Forest Fire Prediction Tool was developed based on historical meteorological data in the Pacific Northwest United States from the past 32 years. These data were collected from local sensors including multiple fire weather attributes, ten of which were strategically screened out to be utilized for the tool. The tool was constructed by testing six different Machine Learning methods. Three of them, a Disjunctive Normal Form rule based method, Decision Trees, and Naïve Bayes, are based on multiple-valued logic. The other three, Support Vector Machine, along with the radial basis and polynomial kernel functions, are based on continuous representation. To demonstrate the accuracy of prediction, a portion of the historical data were randomly selected and compared with the prediction results. The results showed that when testing learners multiple times on randomly selected data, the Disjunctive Normal Form method yields the highest accuracy of 98%. This paper demonstrates that the tool achieved a level of accuracy that will enable the wildfire management agencies to utilize the results for planning prior to fire seasons with high confidence. The weather data used in this tool can be collected ahead of time with relative ease and low costs. Keywords: Fire Weather Data, Proactive Wildfire Management, Machine Learning Presenter Bio: A sophomore at Jesuit High School and Student Board Member of the National Museum of Education, Leo began researching wildfire prediction in 8th grade. His science fair project on wildfires ultimately made him one of the top 30 National Broadcom MASTER Finalists, invited to meet the president, have a planet named after him, and undertake an Environmental Studies internship at the University of Portland. He is a member of the soccer team, chess team, and Model United Nations Club at Jesuit. Although he is mostly undecided on future careers, he hopes to make a big difference in the Environmental world.

86. Modeling of Thunderstorm-Induced Wind Shifts Scott Goodrick, Research Project Leader, USDA Forest Service Additional Authors: Gary L. Achtemeier, retired, USDA Forest Service Abstract: A serious threat to the safety of fire crews is wind shifts caused by thunderstorm downdrafts. Slow-spreading low-intensity flanking fires can be transformed, in a matter of seconds into a high- intensity head fire. Thunderstorm outflows can travel 50-100 miles from source regions to strike fire sites with few signs of changing weather. We describe a thunderstorm wind model that allows for nowcasting of outflow evolution which could provide fire crews with wind shift information 30-60 minutes before wind shifts arrive at a fire site. The models link real-time operational radar precipitation data with ambient temperature and relative humidity to map locations and fields of outflow wind velocities as they evolve relative to local terrain throughout the day. We demonstrate the potential of this thunderstorm wind model through a “proof-of-concept” study of weather conditions that contributed to the deaths of 19 fire fighters at Yarnell Hill on 30 June 2013. The model was set up for the terrain of Arizona surrounding the Yarnell region with a resolution of 900 m. Radar data for 30 June 2013 were supplied by the National Weather Service radar located south of Flagstaff, AZ. Temperature and surface winds were supplied from Peeples Vally (1209-1509 LST) and from Phoenix from 1509-1709 LST. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Around noon thunderstorms formed along the Mogollon Rim approximately 150 km (90 mi) ENE of Yarnell and outflows from these storms pushed southwestward to trigger new storms along a parallel mountain range beginning at 1330 LST. These outflows pushed southwestward until progress was blocked by a mountain range just east of Yarnell at 1500 LST. Meanwhile, outflows from a second cluster of storms pushed into the valley moving toward Yarnell and merged with Mogollon Rim outflows. The combined gust flow reached the southern perimeter of the Yarnell Hill fire between 1630-1645 LST (arrival reported at 1630 LST). The results show the potential for thunderstorm outflow wind models to inform fire fighters of dangers pending from sudden thunderstorm-induced wind shifts, giving them time to assess the risk of changes in weather, and take action to move to safety before the outflows arrive at a fire site. Presenter Bio: Scott Goodrick is project leader for the US Forest Service Southern Research Station's Center for Forest Disturbance Science. Scott's main areas of research are fire weather and fire- atmosphere interactions.

87. How Do Very Large Fires Get to be Very Large Fires? Harry Podschwit, Graduate Student, University of Washington Additional Authors: Brian Potter, Research Meteorologist, USFS Brian Stocks, Wildfire Investigations, Ltd. Narasimhan Larkin, Research Climatologist, USDA Forest Service Mike Wotton, Research Scientist, Canadian Forest Service Abstract: The largest fires each year consume a disproportionately large share of the fire fighting budget. Because of their temporal and spatial extents, they can also contribute to smoke impacts disproportionately. Earlier work as part of the Joint Fire Science Program’s megafire impacts project showed that very large fires’ (VLFs’) evolution differs in some ways from that of other fires. For example, half of all VLFs burned 90% of the final burned area within 14 days of discovery, whereas half of other fires did so within 5 days. Cursory examination of growth curves from 21 identified VLFs shows that most of the fires experienced one or two days when they grew dramatically. This pattern is consistent with the concept of critical weather patterns, or the more general idea that large fires become large because of one big growth day. Examining acres burned per day is relevant for determining final size, but does not accurately convey the rapidity of growth. A small fire growing at a given rate of spread will add less acreage than a larger fire growing at the same rate. We look further into the growth characteristics of these fires in this study. Specifically, we look at: whether the growth patterns measured in acreage persist when fires are examined using alternative measures of growth; the time distribution of large growth days within a fire’s duration; the prevalence of “complexes” in the very large fire population. Keywords: growth; large fires; weather Presenter Bio: Harry Podschwit is a third year Graduate Student in the Quantitative Ecology and Resource Management (QERM) program at the University of Washington. His research involves describing wildfire development and the effects of weather using simple statistical models. His undergraduate was in Applied and Computational Mathematical Sciences at the University of Washington.

88. Defining fire season length using daily climatic, satellite, and documentary fire records Karin Riley, Research Ecologist, US Forest Service, Rocky Mountain Research Station Additional Authors: Karen C. Short, Research Ecologist, US Forest Service 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Patrick Freeborn, Research Physical Scientist, US Forest Service Tim Brown, Research Professor, Desert Research Institute Matt Jolly, Research Ecologist, US Forest Service Nick Nauslar, Research Scientist, Desert Research Institue

Abstract: It may seem obvious what “fire season” means, however, there are many different ways to define it. Perhaps it is the time of year when conditions are conducive to the growth of large fires, a purely climatic definition. Perhaps it is the period between the first and last detections of the start of large fires, or the period between the first detection and last control dates of large fires. The first definition has the strength of being consistent and able to identify and predict changes in fire season length due to climate change; however, it does not capture the presence of fires nor indicate when significant fire activity actually takes place. The latter definitions are extremely sensitive to the occurrence of a single fire, as well as idiosyncrasies of fire reporting and changes in the ways that fires have been managed. Past estimates of the fire-activity season have been based on documentary fire records from the final fire reporting systems of federal agencies. Here, we leverage alternative sources to obtain the fire-activity information at a near-daily timestep and from a spatial domain not limited to federal lands. These sources include the MODIS Active Fire products and historical records in the form of ICS-209 (large-fire) reports and the daily area burned tallies from the national Situation Reporting system. Each of these sources has strengths and weaknesses. However, when taken together, the daily climate data and the satellite-based and documentary fire records can be used to generate more robust estimates of fire season length. We suggest that fire season length be carefully defined relative to analysis objectives. We examine ways in which human activity affects fire season length; for example, acreage from non-prescribed human-ignited fires begins to accrue earlier in the year than acreage from lightning-ignited fires. Careful examination of these issues is critical not only to a robust definition of fire season and estimates of its length, but also to analyses of how fire season length might be affected by climate change, and how different estimates of fire season length have different implications for fire policy and fire management decisions. Keywords: fire season length; MODIS; documentary fire records Presenter Bio: Dr. Karin Riley is a Research Ecologist with the Forestry Sciences Laboratory of the US Forest Service in Missoula, Montana. Her work focuses on the relationship of wildfire with climate and drought, and landscape-scale fire modeling for risk assessment.

89. Modeling alternative fire response policies: proof-of-concept and preliminary results Karin Riley, Research Ecologist, US Forest Service, Rocky Mountain Research Station Additional Authors: Matt P. Thompson, Research Forester, US Forest Service, Missoula, Montana Joe Scott, Fire Analyst, Pyrologix LLC, Missoula, Montana Abstract: Land and fire management agencies are increasingly recognizing the need to expand the footprint of managed fire on the landscape, under the right conditions and in the right locations, to achieve ecological benefits and reduce hazard. In turn, risk assessment tools and products are increasingly used to characterize spatial patterns of loss and benefit, and this information can guide strategic pre-fire response planning. Fire modeling plays a fundamental role in these assessment efforts. However, while the use of fire modeling to examine current conditions or hypothetically treated landscapes is relatively mature, how to model alternative response policies and subsequent spatial patterns of hazard and risk is less well understood. Here, we develop a framework and describe alternative mechanisms with which to simulate various response policies, and use the Large Fire Simulator (FSim) to demonstrate proof-of-concept. As a case study landscape we focus on the Sierra National Forest in California, USA, leveraging information and outputs from recent risk assessment and 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

spatial fire planning efforts on the Forest. To compare alternative policies we summarize metrics including annual burn probability, fire intensity, and fire size distributions. Keywords: fire management; suppression policy; FSim; fire modeling; fire risk

Presenter Bio: Dr. Karin Riley is a Research Ecologist with the Forestry Sciences Laboratory of the US Forest Service in Missoula, Montana. Her work focuses on the relationship of wildfire with climate and drought, and landscape-scale fire modeling for risk assessment.

90. Analyzing tradeoffs among socioeconomic and ecological restoration goals on the national forests of the Pacific Northwest. Kevin Vogler, Faculty Research Assistant, Oregon State University Additional Authors: Alan Ager, Research Forester, Missoula Fire Science Laboratory Michelle Day, Faculty Research Assistant, Oregon State University John Bailey, Associate Professor, Oregon State University Abstract: Despite both the large geographic scope and substantial investments in US Federal restoration and fuels projects, a quantitative decision support system framework to locate optimal project areas and examine tradeoffs among alternative management strategies is lacking. Here, we demonstrate a new prioritization approach for restoration and fuels planning using spatial optimization and production possibility frontiers (PPF), and describe three case studies on national forests in the Pacific Northwest. We selected a wide range of socioecological restoration objectives based on national forest planning documents, including ecological departure, insect risk, wildfire hazard, economics, and wildfire transmission to the WUI. We then used optimization methods to locate and design future projects that maximized the treatment of one or more restoration objectives. The result was a sequencing of projects with specific stands to treat, and the associated attainment for each restoration objective. The results indicated substantial benefits from the spatial optimization in terms locating projects and treatment areas to meet restoration objectives. However, substantial tradeoffs were generally associated with a particular restoration emphasis due to poor spatial concordance within the national forests. For instance, allocating restoration projects to treat wildfire transmission to the WUI led to a sharp drop in addressing ecological departure and generating economic outputs. Optimal projects for addressing ecological departure and wildfire transmission to the WUI generally had negative financial values, versus positive values when thinning volume was emphasized. We plotted a sample of 20 current restoration projects on the PPF’s and found that the majority are suboptimal, perhaps the result of local planning constraints on project designs. The tradeoffs highlighted the challenges associated with meeting both economic and fire protection goals. The methods have a number of potential application in restoration planning including: 1) Quantifying and mapping conflicts between ecological, wildfire protection and socioeconomic objectives as part of restoration programs; 2) Measuring efficiencies of ongoing fuels and restoration projects compared to optimal landscape designs; 3) Illustrating tradeoffs and finding socially optimal mixes of restoration goals as part of collaborative planning process; and 4) Quantifying the potential economic viability of accelerated restoration programs. Keywords: restoration prioritization, production possibility frontiers, fuels treatment planning Presenter Bio: Kevin is faculty research assistant at Oregon State University. He received his Masters in Forest Resources from OSU in 2013. His thesis investigated the role that woody biomass generated from fuel reduction treatments can play in aiding project economics. His research interests include: wildfire risk and hazard assessment, fuels treatment planning and timber valuation.

91. Using Natural Ignitions to Accomplish Land Management Objectives Kelly Martin, Chief of Fire and Aviation Management, Federal Government - Yosemite National Park 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Additional Authors: Kent vanWagtendonk, Fire Geographer, Yosemite National Park

Abstract: Agency Administrators and Fire Managers will need to effectively use wildland fire as a tool to accomplishment resource and protection objectives for many years to come. Defining these areas and understanding and mitigating risks associated with using wildfire to meet resource objectives is all too often dismissed due to uncertain outcomes. What is even more unnerving to agency Administrators is the potential for an escaped prescribed fire. There is little reward for AAs and FMOs to implement the discretionary use of fire (wildfires = put it out; prescribed fire = small units or No-Go). A multi-faceted approach is needed. 1 - Raising awareness with our noteworthy stakeholders; local businesses, neighboring communities and adjacent land management agencies – listen to their concerns. Most people want to know they are being heard and when practical make adjustments based on their input. 2 – Define your goals as a land management agency (NEPA; ); develop a strategy to achieve those goals (Fire Management Plan); list the objectives (Prescribed Fire Plan or WFDSS) and finally; Implement at a large scale (develop IAP and accomplish with Prescribed Fire Teams). Our national suppression organization is not likely to be downsized anytime soon. Can we upsize a Prescribed (Applied) Fire Program Nationally? Do we accept the inevitability of suppression and watershed restoration post fire, OR we can support an interagency, proactive approach to applied fire? The National Fire Plan was laudable at the time to increase the scale of prescribed fire. Game changers include climate change, population density and human health impacts from smoke. Small local units continue to operate with autonomy. Land Management Agencies that were once a Carbon sink are now a source of carbon due to severe fire, insect and drought stress that are killing all size classes of trees. Explore how Yosemite National Park uses science, technology and human factors to identify, decide and implement wildfires and prescribed fires in drought years and how this same technique can be used to time prescribed fire windows in the spring and fall and even throughout the winter. Keywords: Decision making; Risk Management; Social and Political Risk Management; Presenter Bio: Kelly Martin is the Chief of Fire and Aviation Management at Yosemite National Park. Her fire career spans 16 years with the National Park Service and 17 years with the USFS. Her Fire Management Team at Yosemite strives to change the fire suppression culture by demonstrating the safe and effective use of wildland fire as a viable landscape management tool. Kelly is a Fire Behavior Analyst; Chair for the National Fire Management Leadership course presented at National Advanced Fire and Resource Institute; and currently works an Operations Branch Director on a California Type 1 Incident Management Team.

92. Restoring Fire to North American Wildlands - A Call to Action Tim Sexton, Program Manager, USFS, Rocky Mountain Research Station Abstract: Fear of undesired outcomes, economic constraints, and opposition from some state and local government agencies has led to a reduction in using wildfires to achieve resource benefits on federal lands in the last few years. Hazardous fuel has continued to accumulate, ecosystems have continued to degrade, and costs of federal suppression programs have skyrocketed. Climate change will continue to exacerbate the impacts of fire exclusion on North American wildlands by creating episodic fire danger conditions that contribute to fire behavior that defies control efforts and burns at extreme levels of severity. Current practices by federal agencies closely mimic the twentieth century “ten am ten acre” policy and plan. Most wildfires occurring on federal lands are extinguished at less than 1 acre during the first operational period after discovery. It appears that the “ten am, ten acre” policy of aggressive suppression on almost all wildfires has not been abandoned. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Aggressive suppression on fires that may threaten lives, property and important cultural and natural resources must continue. However, we need to stop suppressing those fires that, with careful management help achieve land management objectives while posing little threat to values.

Keywords: ten am ten acre policy, restoration , fire exclusion Presenter Bio: Tim manages the Wildland Fire Research Development & Applications Program. Past work experience includes: District Ranger, National Fire Use Program Manager, National Fire Ecologist, FMO and Hotshot Superintendent. His assignments have taken him to California, Washington, Idaho, Oregon, Colorado, and Minnesota. He has served as a Type I IC in the Great Basin and as a Type II IC in the Rocky Mountain Region. He also served on Fire Use Management Teams as IC, PSC, OSC, and Fire Use Manager. Tim has a BA in History from Boise State University and an MS in Fire Ecology from Oregon State University.

93. Planning for a future of more fire, safer fire, and better fire Christopher OConnor, Ecologist, US Forest Service Rocky Mountain Research Station Additional Authors: Dr. David Calkin, Research Forester, USFS RMRS Dr. Matthew Thompson, Research Forester, USFS RMRS Abstract: The system of wildfire management in the United States has remained relatively unchanged for more than half a century while the scale of the fire management problem, the annual cost of fighting fire, and the loss of human and natural resource values has increased year after year. Federal fire management infrastructure is nearing a tipping point where continuing to increase budgets without addressing the underlying causes of ever mounting risks to human lives, infrastructure, and natural system function will only make escape from the fire suppression cycle more difficult. With the recent release of the National Cohesive Wildland Fire Management Strategy and call for revision of national forest plans across the country, there is an opportunity to intervene that may be able to shift fire management expenditures away from independent management of individual incidents, toward managing incidents and longer-term fire planning that meets pre-determined goals that serve human and natural systems and that promote functional and sustainable landscapes. In this study we introduce a series of changes to current command and control systems and new scientific approaches to fire planning and risk management, informed by spatial analyses, that have potential to help break the fire suppression cycle and to move federal forest management back to a holistic model of long-term landscape sustainability while protecting human resource values. Keywords: Fire planning, risk management, fire suppression cycle, sustainability Presenter Bio: Kit O’Connor is an Ecologist with the USFS Rocky Mountain Research Station in Missoula, Montana working on the human dimensions of fire management and changes to the fire-dynamics of managed landscapes. His current projects involve spatial analysis for fire management planning, and a series of studies designed to improve national forest planning systems and processes to promote resilience of natural and human-managed systems.

94. Emerging Communication Technologies for Wildland Firefighting Ed Mills, Firefighter/Public Safety Broadband Network (FirstNet) Manager, Evergreen Fire Rescue/State of Colorado Governor's Office of Information Technology Abstract: In this fast paced and dynamic session we will explore emerging LTE (Long Term Evolution) technology that has the potential to give Wildland Firefighting Operations new ways to locate and communicate with each other, command and the public in general. Using a widely adopted technology like Long Term Evolution (LTE) will give the Wildland Firefighters multiple tools that are rapidly 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

deployable, cover a greater area, are less impacted by the terrain, with increased flexibility and much more. In this session we will also discuss FirstNet and its potential to assist in Wildland fire operations. Keywords: Wildland Fire, Communications, LTE, FirstNet, National Public Safety Broadband Network

Presenter Bio: Ed Mills is the FirstNet Outreach & Education Manager in the Colorado Governor's Office of Information Technology. He and his team are responsible for coordinating efforts across the state of Colorado to support FirstNet’s mission to give first responders a “super highway” for critical communications. Check out more on FirstNet in Colorado at www.FirstNetColorado.org. Ed is a native of Colorado and has worked in the Public Safety sector nationally and internationally for the past 8 years. He is a domain specialist in counter terrorism technologies such as StreetLab Mobile (Raman), Mobile Trace (ITMS) and acoustic wave. He has worked alongside many of the major cities’ law enforcement and fire departments like Seattle Fire, Massachusetts Hazardous Materials Unit, Los Angeles Harbor Police and Bomb Squad. Ed has also worked with the DEA in Quantico, ICE, ATF, and the Coast Guard. He assisted Mexican Customs, Federal Police and the Mexican Army to help them detect drugs and explosives. Ed was the project lead on GE Security’s 3-B laser product - a device that detects chemical as well as biological threats. Ed also enjoys the role of educator. He is a firearms instructor in Colorado as well as a certified instructor with the Utah Bureau of Criminal Investigations. For the past 20 years he has also worked as a teacher and administrator in Missouri, Hawaii and Colorado, publishing books and developing curriculum along the way. Ed has always felt passionately about service and giving back to the community. He is currently a Firefighter, Wildland Firefighter with a current Red Card, Emergency Medical Responder, swift water rescue swimmer, with Evergreen Fire Rescue, a small mountain community in the Rocky Mountains. Mr. Mills serves on the high-risk extrication team in the Denver metro area and is a Fire Information Duty Officer for the mountain area. He also served as a military policeman in the United States Army for 6 years.

95. Synergistic Use of New NASA Technologies for Pre-, Active, and Post-Fire Applications E. Natasha Stavros, Post-Doctoral Research Scientist, Jet Propulsion Laboratory, California Institute of Technology Additional Authors: Michael Gunson, Program Manager and Project Scientist, Jet Propulsion Laboratory, California Institute of Technology David S. Schimel, Group Supervisor, Jet Propulsion Laboratory, California Institute of Technology Abstract: Wildfire size, frequency, and severity are increasing with a changing climate. Furthermore, their impact on society is increasing as the wildland urban interface continues to grow. This creates a need for new technological approaches to study and understand what causes, drives, and extinguishes wildfire a well as the resulting effects of fire on the landscape. Over the last three years, the Jet Propulsion Laboratory (JPL) has canvased the western United States using the latest technologies flown on aircraft to test their applications. The canvasing campaigns vary based on the technology being flown, but coordinate efforts among airborne campaigns have resulted in unprecedented spatial and spectral resolution mapping across large areas, particularly in California, using multiple technologies. These technologies include Light Detection and Ranging (LiDAR), hyper spectral visual to shortwave infrared Airborne Visual/Infrared Imaging Spectrometer (AVIRIS), high spatial resolution multi-band thermal infrared imaging (MASTER), coupled spectrometer and LiDAR on the Airborne Snow Observatory (ASO), and the Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR). These technologies provide different information that complements one another, especially with respect to fire ecology. LiDAR provides 3-dimensional forest structure and biomass. AVIRIS provides information on forest composition, chemistry, condition, and quantification of fire severity. MASTER provides very high- 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

resolution fire intensity, evapotranspiration and water-use efficiency maps. ASO provides snow depth and pack, which inform snow accumulation and melt useful in water resource management. Lastly, UAVSAR can provide estimates of canopy structure, inundation, and soil moisture. In November 2015, JPL hosted a study with the US Forest Service to develop informational data products useful for pre-, active, and post-fire applications. Because the technologies are either not yet available from satellite, or with limited extent on satellites, research is being conducted to assess their full utility for fire ecology including fire behavior modeling calibration and validation, and mapping of fuel models, soil absorption and stability, and drought stress. Fortunately, some of these technologies were recently competitively selected by NASA Earth Venture Instrument (EVI) Program are being developed for deployment on the International Space Station, thus providing a platform for much wider application beyond the extent of these experimental campaigns. Keywords: remote sensing; fire behavior; fuels management; decisions support; Presenter Bio: Dr. E. Natasha Stavros received her Ph.D. from the University of Washington, Seattle in Forest and Fire Ecology. Her dissertation focused on the relationship of megafire occurrence and climate and how megafire occurrence might change under different climate scenarios. She currently works at the Jet Propulsion Laboratory exploring multi-instrument remote sensing approaches to quantify and understand fire ecology. She is particularly interested in measuring using remote sensing and synthesizing observations to improve understanding of the role of fire in the Earth system including conditions leading up to fire, active fire behavior and air quality degradation, and post-fire effects.

96. Efforts to Enhance the Emergency Fire Shelter: A Collaboration between the U. S. Forest Service, NASA, and the University of Alberta Tony Petrilli, Equipment Specialist, US Forest Service - Missoula Technology and Development Center Additional Authors: Tony Petrilli, Fire Shelter Project Leader, US Forest Service - Missoula Technology and Development Center Josh Fody, Thermal Analyst, NASA - Langely Research Center Mark Ackerman, Adjunct Professor, University of Alberta Abstract: The Forest Service and other public lands agencies respond to an average of 73,000 wildfires per year, and responding firefighters are required to carry a number of safety gear items, including the M2002 emergency fire shelter. The emergency fire shelter is intended to serve as a last resort means of protection in case a firefighter’s escape route has been compromised in the face of an approaching flame front. Firefighters in the field work for extended periods under demanding conditions; as a result, the M2002 has been designed to provide thermal protection with minimal added weight and bulk to a firefighter’s gear pack. Accordingly, the M2002 is constructed from a lightweight reinforced aluminum laminate less than half of a millimeter in thickness. This material provides good protection against thermal radiation, but there is limited resistance in the event that energy is absorbed by direct contact with flames or hot gasses. As a result, firefighters are trained to deploy their shelters in sites clear of combustible materials in order to reduce the exposure to such convective heating; however, clear sites are not always available and personnel have been forced to deploy shelters in fuel directly in the path of flames at cost to human life. After the tragic loss of 19 Fire Fighters in Yarnell Hill, AZ in 2013 the Forest Service decided to expedite the next redesign cycle of the fire shelter in order to improve its performance. This presentation will give an overview of current efforts lead by the US Forest Service Missoula Technology and Development Center to search for technologies which seek to improve the ability of the fire shelter to stand up to direct flame heating, and the challenges of maintaining an acceptable weight, bulk, and other design requirements. Keywords: fire shelter 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Presenter Bio: Tony Petrilli is an equipment specialist for the US Forest Service Missoula Technology and Development Center (MTDC). He holds a bachelor’s degree in education from Western Montana College. Tony began working for the Forest Service in 1982 as a firefighter for the Lewis and Clark National Forest. He became a smokejumper in Missoula in 1989. In 1992 he began working part time at MTDC; he then joined the Center full time in 2000. He has been the fire shelter and firefighter clothing project leader since 2005. Tony has served on more than 30 fire entrapment safety review or investigation teams as a PPE specialist.

97. Work to Improve the Emergency Fire Shelter Using NASA Space Technology: “Convective Heating Improvement for Emergency Fire Shelter (CHIEFS)” Josh Fody, Thermal Engineer, NASA Langley Research Center Additional Authors: Tony Petrilli, Equipment Specialist, US Forest Service (MTDC) Mark Ackerman, Adjust Professor, University of Alberta Abstract: Engineers at NASA Langley Research Center have spent the better part of a decade developing thin flexible thermal materials for use in inflatable heat shields, and have demonstrated their performance in space flight test programs. After learning of the tragedy at Yarnell Hill, Arizona in 2013, NASA engineers wanted to help improve the fire shelter using experience and technology developed for these thermal protection systems. They launched the CHIEFS project in 2014, and reached out to the U.S. Forest Service Missoula Technology and Development Center to offer assistance. The overall objective of the CHIEFS project is to work with the Forest Service to improve the convective heating performance of the shelter as a result of direct flame contact, while maintaining the good radiant performance of the current shelter design. This presentation will give an overview of the research, challenges, and lessons learned in NASA’s CHIEFS project. Keywords: Fire Shelter Convective Heating NASA MTDC Presenter Bio: Josh was born and raised in Colorado. After serving as an enlisted sailor in the US Navy for 4 years working on the flight deck and in the jet shop performing intermediate level maintenance on aircraft power-plants, he came back home and attended CU Boulder for his undergraduate degree. He moved to Washington, DC where he later pursued a Master’s Degree in Mechanical Engineering at the University of Maryland. His research primarily focused on enhanced evaporator concepts for alternate energy generation. After defending his thesis, he took a job at NASA Langley where he has been working in the Structural and Thermal Systems Branch for the past 2 years as a thermal analyst.

98. Evaluating the Quality of a Wildfire Defensible Space with Airborne LiDAR and GIS Jason Harshman, Senior GIS Analyst, Penn State University Abstract: California State law requires property owners in the Urban-Wildland Interface to maintain vegetation within 100 feet from the edge of a structure in order to minimize its threat from wildfire. This boundary is commonly referred to as a defensible space and studies have shown that structures with high quality defensible spaces have a lower risk of damage from a wildfire compared to structures with poor quality defensible spaces. The defensible space is typically broken into two zones with specific fuel reduction requirements. The first 30 feet should be clear of most vegetation with the exceptions of plants with low flammability and vegetation critical to the surrounding landscape. From 30 to 100 feet, vegetation shall be maintained to remove ladder fuels and direct paths for fire. Based on these requirements, I demonstrate a methodology for evaluating the quality of structure’s defensible space using GIS analysis of LiDAR collected in Sonoma County, CA and San Luis Obispo County, CA. The analysis generates a value for each structure between 0.0 and 1.0, values closer to 0 indicate a structure with a higher quality defensible space than structures with values near 1. Preliminary analysis indicate values correlate to the amount of vegetation surrounding a structure, suggesting the methodology is sufficient 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

in quantifying a structure’s defensible space. Applying the methodology to analyze structures across a region may help stakeholders allocate resources to areas with lower quality defensible spaces as a way to increase awareness and compliance. The results could be combined with other wildfire metrics to provide a greater understanding of potential wildfire risk and hazard.

Keywords: defensible space, LiDAR, GIS, wildfire risk management Presenter Bio: Jason Harshman is a MGIS student at Penn State University and is a certified GIS Professional. He currently works in San Diego, CA at Cardno as a Senior GIS Analyst. Jason received his BA in Geography at Sonoma State University in 2006. He is a member of the Association of American Geographers and the Cartography and Geographic Information Society.

99. Detection of Forest Fires Impact with Remote Sensing Data, ALSAT, In Semi-arid Zones, Algeria Zegrar Ahmed, Research team leader, Centre of Spaces Techniques Abstract: The Algerians forests are some ecosystem dynamics witch very rarely reaches a state of balance. The strengths which govern these ecosystems are numerous and understand the succession among others, the senescence as well as the disruptions to the bugs, to the herbivorous and mainly to the fire and to the atmospheric phenomenon’s. So the Algerian forests are characterized by a particularly flammable material and fuel. The wind, the relief and the slope facilitates the propagation of fire; particularly hanging the dry period, of the end of June to September with a maximum in the month of August. The gait in this survey consists in localized, detect the of forest fires and measure their aftermath. In order to know the impact of these fires on the natural middle, it is necessary to determine their spatial and temporal distribution. For it make some satellite data ALSAT and LANDSAT to several dates of taking of seeing (2012 - 2014) was used and underwent some operations of specific treatments notably some geometric corrections and some classifications. These treatments permitted a thematic detailed analysis some forests ecosystems. The utilization of the indication of vegetation and of the indication of verdure permitted the determination of the pockets of fires and of characterizing the surface of fuel. We used a system of Geographical Information (SIG), for the cartographic management, the determination of the zones of vulnerability and finally for zonings of the fires. These last could be used in the global gait of forest harnessing. This approach allows showing the contribution of the data of Algerian satellite ALSAT in the detection and the well attended some forest fires in Algeria. Keywords: Forest fires, remote sensing, ALSAT, forest ecosystem, harnessing Presenter Bio: Zegrar Ahmed, Research team leader, Centre of Spaces Techniques

100. Impacts of Post-fire Salvage Harvesting on Early-seral Ecosystems in Western Oregon John Bailey, Associate Professor, Oregon State University Additional Authors: Chris Dunn, Faculty Research Associate, Oregon State University Abstract: Spatial extent and fire behavior of wildland fires is unprecedented recently in western Oregon, as with much of the West, and likely to continue or worsen into the near future with climate change. This trend has fueled higher percentages of stand-replacing fire within eventual fire perimeters, prompting interests in salvage harvests particularly on private lands. We synthesize existing literature on the effects of salvage harvesting with new data from intensively-measured western Oregon field plots collected from a chronosequence of past fires. Results are presented for dead wood structure (snags and coarse wood), carbon pools, and long-term tree regeneration and other early-seral vegetation communities. Salvage has immediate and predictable negative impacts on coarse wood abundance, whether lightly or heavily salvaged – effects that may persist for decades. Similarly, aboveground carbon pools are diminished. Effects on early-seral vegetation are typically insignificant in terms of abundance and composition; we present proposed mechanisms for potential positive and 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

negative impacts relative to unsalvaged areas. Viewed together, our intent is to present an integrated conceptual model of post-fire salvage impacts on forest productivity and sustainability that can be used as guidelines for forest management.

Keywords: Salvage, Early-seral communities, Coarse Wood, Regeneration Presenter Bio: John received his BS and MF from Virginia Tech (as a native Virginian) and then worked for 6+ years with the EPA in Corvallis, Oregon on forest responses to climate. He returned to school for a Ph.D. (Silviculture) at Oregon State University, then joined the faculty at Northern Arizona University for nine years working in semi-arid silviculture, fuels treatment and restoration treatments, then returned to Oregon State in 2006 to continue work on fuels/fire management and ecosystem restoration in drier forest types, as well as resume research on multi-story management in wetter forest types. It is Oregon, after all.

101. Mapping Severe Fire Potential in the Contiguous United States Brett Davis, Ecologist, USDA Forest Service Additional Authors: Matthew H. Panunto, Ecologist, USDA Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory, Missoula, MT Gregory K. Dillon, Spatial Fire Analyst, USDA Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory, Missoula, MT Robert E. Keane, Research Ecologist, USDA Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory, Missoula, MT Donovan S. Birch, Research Support Scientist, University of Idaho, Department of Forest, Rangeland, and Fire Sciences, Moscow, ID Penelope Morgan, Professor, University of Idaho, Department of Forest, Rangeland, and Fire Sciences, Moscow, ID Abstract: The Fire Severity Mapping System (FIRESEV) project is an effort to provide critical information and tools to fire managers that enhance their ability to assess potential ecological effects of wildland fire. A major component of the FIRESEV project is the development of the Severe Fire Potential (SFP) map, a geographic dataset covering the contiguous United States (CONUS) describing the potential for wildland fires to burn with high severity should they occur. To create the SFP map, we developed a series of statistical models, each relating a suite of independent geospatial variables to 30 years of fire severity inferred from satellite imagery by the Monitoring Trends in Burn Severity (MTBS) project. We partitioned continuous measures of burn severity into a binary dataset of ‘higher severity’ vs. ‘lower severity’. We then produced statistical models to determine the relationship between the two severity classes and a number of independent geospatial variables including pre-fire vegetation, pre-fire fuel moisture and a series of topographic variables. We developed separate statistical models for forest and woodland vs. non-forest settings in each of 25 distinct ecological regions across CONUS. We then applied these models to contemporary landscapes comprised of current vegetation data, 90th- percentile 1000-hr fuel moisture conditions and the static topographic variables to generate the30-m resolution SFP map. When coupled with information regarding current landscape conditions, the FIRESEV SFP map can assist managers in identifying areas where fire may help restore fire-adapted ecosystems and where it might have less favorable impacts. This map product is intended to be incorporated into existing decision support frameworks such as the Wildland Fire Decision Support System (WFDSS). The SFP map for the western US, not including Alaska, is currently available online (http://www.frames.gov/firesev) and will soon be available for the eastern US. Development of the Severe Fire Potential map has provided an opportunity to enhance our understanding of environmental influences on burn severity and has provided a new resource to support fire management decisions. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Keywords: Fire severity, MTBS Presenter Bio: Brett Davis is an Ecologist and Fire Modeler at the Fire Sciences Lab in Missoula, MT. Before joining the Fire Sciences Lab, Brett worked for eight years at the Aldo Leopold Wilderness Research Institute in Missoula as a GIS Specialist and Fire Modeler. His work at the Leopold Institute focused on wildland fire behavior and ecological impacts with a focus on the restoration of fire as a critical ecological process. He holds a B.A. from the University of Colorado, Boulder in Environmental, Population and Organismic Biology and an M.S. in Forestry from Colorado State University, Fort Collins.

102. Spatial Analysis of the Influence of Fire Severity on Forest Structure on the North Rim of Grand Canyon National Park Valentijn Hoff, GIS Analyst, FireCenter, The University of Montana Additional Authors: Eric Rowell, Image Analyst/Programmer, FireCenter, The University of Montana Tim Wallace, Remote Sensing Analyst, FireCenter, The University of Montana LLoyd Queen, Professor of Remote Sensing, FireCenter, The University of Montana Eric Gdula, GIS Specialist, NPS Windy Bunn, Fire Ecologist, NPS Casey Teske, Fire Ecologist, Grand Canyon National Park, NPS Abstract: Wildfires and prescribed fires occur frequently on the North Rim of the Grand Canyon in Arizona. Fire managers at Grand Canyon National Park (GRCA) are interested in the interaction between fire severity and forest structure. Fire severity is an import measure of change in an ecosystem, and the amount of change often varies greatly over the extent of a fire. The spatial distribution of structure also often varies greatly. To assess these relationships we used the Monitoring Trends in Burn Severity data for 58 fires that occurred between 2000 and 2013 on the North Rim of the Grand Canyon. In 2012, aerial Light Detection and Ranging (LiDAR) data were collected across the North Rim area of GRCA. The acquisition of these data provided an opportunity to merge LiDAR-derived forest structure data with the park’s CBI-calibrated, remote sensed fire severity data, to understand how time since fire and past fire severity influence forest structure. We produced 9 structure metrics, both vertical and horizontal, for the North Rim area. We also included vegetation type, to find relationships between structure and fire severity of both fires that burned prior and post the LiDAR acquisition. Our work-to-date indicates that the fire history of the Ponderosa pine forests on the plateaus close to the canyon rim shows frequent, low severity fire. The majority of these forests have a moderately open canopy with an uneven vertical structure. Of the forest that did not burn recently, 91% has a more uneven vertical structure profile. This indicates multi strata development and a lower crown-base-height of the overstory. These forests are also likely to have more closed canopy. When moderate-high and high fire severity have occurred in the last decade, a large proportion of the canopy is open with little vertical heterogeneity. We found that time-since-fire is an import factor in forest structure development. The results of this research will help managers decide if a current forest structure is likely to support a fire or severity sequence that results in a desirable forest structure. We present results of these analyses and discuss the potential practical implications for fire managers. Keywords: LIDAR, fire severity, forest structure, Grand Canyon National Park Presenter Bio: Valentijn Hoff is a GIS Analyst at the FireCenter at the College of Forestry and Conservation, The University of Montana, in Missoula, Montana. He enjoys using spatial technologies in helping natural resource managers make more informed decisions. During fire season he can be found embedded with Fire Management Modules, helping Incident Management Teams with GIS, or installing remote networks for fire management. Valentijn has Master of Science degree in Forestry from the University of Montana.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

103. Fire Moss as a Tool for Post-Wildfire Ecosystem Restoration Chris Ives, Master of Science in Forestry Candidate, Northern Arizona University, School of Forestry Additional Authors: Matthew Bowker PH.D., Assistant Professor, Northern Arizona University School of Forestry Anita Antoninka PH.D., Postdoctoral Scholar, Northern Arizona University School of Forestry Abstract: Increasingly large and severe fires across the western United States are creating difficult restoration challenges for land managers. Despite the wide usage of current fire restoration techniques, many studies have shown little to no benefit when compared with a no action alternative and some studies have even shown negative impacts, such as the spread of invasive species. The use of select disturbance colonizing mosses or “fire moss” is a promising alternative restoration tool that has never been investigated for use in post-fire environments. “Fire moss” refer to the moss species found colonizing post-fire environments, worldwide, within several months to years following a fire. Fire mosses possess traits that make them ideal candidates for restoration purposes such as: universal distribution, high desiccation tolerance, high water holding capacity, and soil aggregation ability. Fire moss has also been observed to have no negative effects on the return of native vascular plants. Harnessing their restoration power, finding ways to bring them to additional critical post-fire sites, and hastening their arrival on scene could provide a valuable restoration service not currently being utilized. Our research addresses the basic questions surrounding the effectiveness of fire moss as a restoration tool. Greenhouse experiments were conducted to find optimal greenhouse fire moss growth conditions to supply field application treatments. Results show greatest moss growth under five and seven day per week watering regimes. Additionally, the fire moss Bryum argenteum constituted a larger part of overall moss growth under a five day per week watering regime than under a seven day per week watering regime, suggesting that this moss would be the best candidate for use in marginal fire moss habitat (lower elevation, drier, and more exposed sites). Finally, mosses took longer to colonize sample units that contained an ash addition than those without an ash addition. This suggests that fire moss colonizes high severity burned areas via spore growth as opposed to clonal growth from moss fragments. Therefore, future studies should explore the use of a spore-based inoculum for restoration of high severity burned areas. Field experiments are in progress to identify appropriate field application techniques. Keywords: Fire Moss, Restoration, BAER Presenter Bio: As a serial land manager, Chris has been working in western US ecosystems since 2005. From leading conservation corps crews to working in small-scale private forest products industries and serving as a wildland firefighter, he has sought to understand and make a positive impact on natural environments and the people who enjoy them. In 2011, while assisting with a burn severity survey on the Wallow fire, he came across prolific moss growth in high severity burned areas. Chris soon enrolled at Northern Arizona University’s School of Forestry where he began studying fire moss.

104. Disentangling the Drivers of Wildfire Severity in a Multi-Owner Forest Landscape Harold Zald, Faculty Research Associate, College of Forestry, Oregon State University Additional Authors: Christopher J. Dunn, Faculty Research Associate, College of Forestry, Oregon State University Abstract: Fuels are the only component of the fire triangle that forest and fire managers can alter to change fire behavior. There have been numerous studies examining how fuel reduction treatments and salvage logging alter fire behavior, severity, and its’ ecological impacts. However, less attention has been paid to how different forest management objectives may influence fire severity in multi-owner landscapes, despite costly and politically contentious suppression of wildfires that do not acknowledge ownership boundaries. In 2013, the Douglas Complex burned over 30,000 ha of Oregon & California Railroad (O&C) lands in Southwestern Oregon, USA. The O&C lands are a geographic checkerboard of 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

private industrial and federal forest land with fundamentally different management objectives, subsequent forest conditions, and perceived fire risks, providing a unique opportunity to quantify the effects of forest management practices on wildfire severity. We bring together geospatial data on fire progression, fire weather, topography, pre-fire forest conditions, and past management activities to represent the different factors that influence fire behavior. Using ensemble machine learning and spatial autoregressive modelling techniques, we disentangled the relative importance of these factors on fire severity (relative differenced normalized burn ratio, RdNBR) as calculated from Landsat imagery. While daily fire weather strongly influenced fire extent (area burned), ownership was the most important driver of fire severity, with younger and structurally homogeneous stands on private industrial forests displaying higher fire severity compared to older and more structural complex forests on federal lands. Keywords: Fire severity, RdNBR, forest management, landscape ecology Presenter Bio: Harold Zald is currently a research associate in the College of Forestry at Oregon State University. Dr. Zald received his Ph.D. in forest science from Oregon State University in 2010. Dr. Zald’s research interests focus on forest responses to disturbance agents in complex mountain terrain, fuel- reduction and restoration treatments in dry conifer forests of western North America, integrating field data with remotely sensed imagery to support inventory and monitoring of forest ecosystems, and quantifying the landscape patterns and environmental drivers of tree invasion in mountain meadows.

105. Estimating Fire Induced Basal Area Mortality with Multi-temporal LiDAR Michael Hoe, Biometrician, Oregon State University Additional Authors: Christopher Dunn, Faculty Research Associate, Oregon State University Temesgen Hailemariam, Professor in Forest Biometrics & Measurements, Oregon State University Abstract: Fire severity is often estimated through the use of satellite imagery and observed conditions. While these techniques work well for differentiating horizontal variation across a landscape, they have difficulty estimating vertical components. Michael Wulder and Tian Han have shown that LIDAR metrics can help fill this gap by assisting in a more complete model of the variation across a large landscape. Light detection and ranging (LIDAR) has accurately estimated forest attributes such as canopy structure, biomass, volume, mean tree height, and canopy bulk density. Canopy bulk density (CBD) is the main parameter utilized by wildland fire managers to simulate fire behavior over a landscape. Furthermore, knowing the distribution of CBD across the landscape provides key information about the forest canopy and ladder fuels which can be used to asses risk and prioritize fuel treatments. Pre- and post-fire LiDAR images of the Douglas Complex and Big Windy fires are being compared in order to detect changes in pixel density, intensity, and color. Pixel density is currently the primary LiDAR variable of interest as it relates directly to estimates of CBD. Intensity and color values are still being examined to determine if additional information can be obtained. This presentation will describe the current methods, study area, results to date, and anticipated next steps. Keywords: Mixed-severity fire, basal area mortality, LiDAR Presenter Bio: Michael Hoe is a graduate student in Biometrics at Oregon State University. He obtained his Bachelors of Science in Forest Management in 2009 and has over 5 years’ experience as a timber cruiser. He has worked for companies such as: Mason, Bruce, & Girard, the Bureau of Land Management, Integrated Resource Management, and Terra Verde. He is currently working on area based estimates of fire induced tree mortality and anticipates moving to single tree segmentation along with single tree mortality predictions.

106. Automating Fuel Model Assignment and Spatial Alignment for Fire Spread Modeling in Roaded Areas 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Casey Teske, Fire Ecologist, Grand Canyon National Park Additional Authors: Amy M. Burzynski, GIS Technician, Wildland Fire Program at Center for Environmental Management of Military Lands Andrew M. Beavers, Wildland Fire Program Manager, Center for Environmental Management of Military Lands Jena A. Ferrarese, Spatial Analyst, Wildland Fire Program at Center for Environmental Management of Military Lands Abstract: Road depiction in the fuels layer of a fire spread simulation (e.g. FARSITE and FSim) strongly impacts modeled fire spread. Accurate risk analyses require coordinated ignition locations in the ignition probability file and road locations in the fuels layer (e.g. unburnable fuel models, typically represented by either 91 or 99 in LANDFIRE data) in order to minimize artificially influencing fire spread modeling outputs. Representing high-probability ignitions in areas immediately adjacent to roads demands high levels of accuracy and precision in the spatial alignment and width of roads in the two input data layers. Any misalignment risks obscuring high-probability ignition pixels by overlaying them with unburnable fuels pixels (e.g., pavement) and artificially decreasing fire probability associated with road-based ignition sources. In many instances, the locations of roads (vector or raster) do not coincide with those locations that LANDFIRE classifies as unburnable fuels in roaded areas. Manually editing the fuels layer to represent spatially accurate roads, and to create 'carrier pathways' to model fire spread across roads, is a tedious task that requires an expert analyst. We developed an automated process to address each of these issues by (a) spatially resolving locations of roads in LANDFIRE and the ignitions probability layer using precise vector data, b) systematically reducing 'hard barrier' unburnable LANDFIRE pixels by reclassifying misaligned unburnable pixels while classifying roads categorically into several custom fuel models and (c) facilitating targeted editing to reduce manual editing efforts. Keywords: Fuel Models, Wildfire Spread, Raster Processing Presenter Bio: Casey has been involved in wildfire and GIS / remote sensing of wildfires for the past 20 years. Although now at Grand Canyon National Park, at the time of this work, she was working in Colorado on wildfire risk analyses for military installations. Prior to that, she spent 15 years in Missoula, MT, working in fire research at the FireCenter, where she was able to hone her GIS and Remote Sensing skills. Her interests include bringing technological advances to the forefront of wildfire management.

107. Educating the Future Fire Workforce to Respond to Increasingly Complex Challenges Leda Kobziar, Assistant Professor, University of Idaho Additional Authors: Penelope Morgan, Professor, University of Idaho Heather Heward, University of Idaho Abstract: Managing fire is increasingly complex and high stakes, calling for advanced professional skills developed through training, experience and education. Access to degree programs which integrate education and experience is critical for current and future fire managers, both in the on-campus and online environments. Assignments that create deliverables useful to fire managers can ensure that education has clear, practical applications: burn plans for local tribes and agencies; short and long-term fire behavior assessments for regional ranger districts; and even monitoring programs to assess vegetation response to salvage logging and burn severity. Students can be challenged to work in interdisciplinary groups to conduct interactive debates about an array of issues related to natural resources as well as specifically living with fire. Leadership and collaboration can be emphasized and evaluated, encouraging students to think independently and as a team. Finally, online classes provide graduate and undergraduate education to fire professionals from around the world, helping them meet the needs of their jobs and advance in their careers by drawing on science and experience. In this 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

presentation, we use examples from the University of Idaho to demonstrate how the learning environment can meet the critical need for experiential learning in a higher education. Keywords: fire training, firefighter, education, experiential learning, workforce

Presenter Bio: Dr. Leda N. Kobziar is dedicated to improving wildland fire management through an increased understanding of fire ecology. As a faculty member in the Natural Resources and Society Department of the University of Idaho’s College of Natural Resources, Dr. Kobziar’s research and teaching explores how fire and fuels management affect forest conservation and restoration across diverse ecosystems. She draws from her experience working with land managers and conducting research in the western and eastern US, Europe, and Canada, and as a Florida Certified Prescribed Burn Manager, to promote the restoration of natural fire regimes for ecological and social benefits. She has served as the Principle Investigator for more than 20 federally-funded fire ecology research grants, and has had the honor of advising over 35 graduate students since 2006. As the President of the Association for Fire Ecology, Dr. Kobziar is actively engaged in improving the knowledge and use of fire in land management through science and education. She currently teaches Fuels Inventory and Management, Fire Ecology, Restoration Ecology, and GIS Applications in Fire Ecology and Management, and is the lead for UI’s Fire Master of Natural Resources online degree program (Fire MNR).

108. Burning for Blooms, Birds, and Butterflies: Partnerships and Pyrodiversity in the Willamette Valley Amanda Stamper, Fire Management Officer, The Nature Conservancy Abstract: Controlled ecological burning has been on an upward trend over the past 20+ years in fire dependent Willamette Valley wet and upland prairies, oak savannas, and woodlands, managed by The Nature Conservancy, US Fish and Wildlife Service, Bureau of Land Management, US Army Corps of Engineers, and private lands managed under the USFWS Partners for Fish and Wildlife Program and NRCS Wetland Reserve Program. Over 900-1000 acres are now being burned annually to restore and maintain habitat and other ecological conditions necessary for the conservation of culturally important, rare and native, threatened and endangered, plants and animals including Streaked horned lark, Golden paintbrush, Bradshaw’s lomatium, Common camas, Willamette Valley daisy, Fender’s blue butterfly, Vesper’s sparrow, and Kincaid’s lupine, among others. The biodiversity of these fire dependent plants, animals, and ecosystems begins to diminish after just a few years without fire. Partnerships and their critical role in providing the expertise and capacity to implement these burns are evident in the diversity represented by those attending briefing each burn day, ready to put more good fire on the ground. Partners include The Nature Conservancy’s Oregon Chapter and Colorado Chapter’s Southern Rockies Wildland Fire Module, USFWS Willamette Valley and Mid-Columbia Refuge Complexes, USFS Siuslaw and Willamette National Forests and Pacific Northwest Regional Office, Confederated Tribes of Grand Ronde, Lane County Parks and Recreation, Eugene and Salem Bureau of Land Management, Oregon Department of Forestry, USACE, City of Eugene Fire Department, Stayton Fire Department, West Valley Rural Fire Department, Friends of Mt. Pisgah and Buford Park, Lane Regional Air Protection Agency, Oregon Department of Agriculture, Oregon Department of Transportation, and the National Weather Service. Finding the best ways to get more good fire on the ground often hinges upon how many qualified resources you have on the same days as you have the right conditions to burn. Diversity in partnerships leads to increased resource capacity for burn implementation, enabled through agreements between agencies, cost-sharing, relationship building, training together, and the all-hands, all-lands approach to burning for the highest collective goals in conservation. Keywords: oak prairie ecological burning 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Presenter Bio: Fire program manager for The Nature Conservancy’s Oregon Chapter and burn boss for Willamette Valley program. Masters in Natural Resources, Fire Ecology and Management from the University of Idaho, Post-Bacc from Oregon State University in Natural Resources, and B.A. in Philosophy from the University of Oregon. Founder and former Chair of the Oregon Prescribed Fire Council. US Forest Service and BLM for 14 years on engines, hotshots, and handcrews, and as a prescribed fire technician, assistant fire management officer in fuels, post-fire rehabilitation coordinator, and invasives program manager.

109. Aerial Firefighting with Helicopters Jim Rankin,CEO, Columbia Helicopters Abstract: Helicopters are critical tools used in the management of wildland and urban fires. Firefighters and ground crews rely on aircraft to deliver equipment and supplies, deploy rappellers and specialized crews, transport firefighters, provide reconnaissance of fire behavior, locations, and movement, drop retardant or water and in some cases perform aerial ignition operations in support of prescribed fires. This presentation will focus on the following: 1. Current National Fleet of Helicopters and their capabilities a. Type 1, Type II and Type III 2. Methods for aerial fire suppression and the advantages/disadvantages of each method a. Bucket b. Tank 3. Commercial Operator limitations and safety standards and their relationship to Federal and State standards a. FAA / USFS / DOI / State 4. Overview of aerial firefighting resources and new technology in the helicopter industry.

This presentation will leave attendees with a general knowledge of National helicopter resources and their capabilities, as well as a thorough understand of compliance requirement for helicopters. Presenter Bio: Jim Rankin,CEO, Columbia Helicopters

110. UWSP Fire Crew Approaching Tomorrow’s Problems With Today’s Education and Training Jacob Livingston, Fire Ecology Undergraduate, University of Wisconsin Stevens Point Fire Crew Additional Authors: Paul Priestley Ethan Robers Abstract: The purpose of our presentation will be to showcase the importance of continual growth in fire ecology in the fire community as a whole by focusing on UWSP Fire Crew’s structure and opportunities offered to students. Fire Crew provides multiple opportunities such as burning with private land owners and the Wisconsin DNR, and in multiple different fuel types such as Oklahoma and Florida, but mostly the Great Lakes Region. They also offer the opportunity to study the ecological purposes of burning through sending around a dozen students to conferences around the country each year. We provide opportunities on and off the fire line for ecological monitoring such as pre and post burn vegetation surveys and seed bank analysis and fire behavior monitoring including fuels measurements, fire intensity, flame lengths, and smoke impacts on fire fighters. We have a mutual agreement with the Wisconsin DNR that provides us with federally certified fire instructors to enhance our training as students while they are receiving a local firefighting resource to assist them with wildfire suppression and prescribe burn operations. We provide basic fire certification training twice a year with student lead sections at our outdoor and hands on training sessions. We also offer higher certifications such as S-131 Look Up Look Down Look Around, S-230 Crew Boss Training, S-219 Firing Operations, and 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

many others which allows students to have a competitive advantage in the field before they graduate. We not only have an emphasis on technical training but scientific training by bringing the classroom to the field through hands on experiences. We bring the field to the classroom by inviting professionals from the fire science community to present new job opportunities and new ideas. We are on the precipice of a dynamic future that will need fire ecologists as much if not more than firefighters and here at UWSP we are developing students of fire to approach upcoming issues in new and inventive ways. Keywords: Students, opportunities, research, training, issues, classroom, certifications Presenter Bio: Jacob Livingston is Fire Ecology and Forest Ecosystem Restoration undergraduate student at University of Wisconsin Stevens Point. He has worked one season on Coronado National Forest as part of IA Type 2 Handcrew, as a Fire Ecology Research Technician at UWSP Treehaven, and as a prairie restoration technician. He is currently working on research on the effects of biochar on germination in prairie seed banks in jack pine barrens. He will be graduating in December of 2016

111. New frontiers in fuel sampling: new techniques for measuring fuels for fire management in the US Robert Keane, Research Ecologist, US Forest Service Rocky Mountain Research Station Abstract: Wildland fuels are important to fire managers because they can be directly manipulated to achieve management goals, such as restoring ecosystems, decreasing fire intensity, minimizing plant mortality, and reducing erosion. However, it can be difficult to accurately measure wildland fuels because of the great variability of fuel types and distribution of those characteristics that describe the wildland fuelbed over space and time. Several conventional sampling methods are available for quantifying fuel loading, but recent research has shown that these methods have scale and procedural problems. This presentation first details results from a set of studies that investigated the sampling and classification of a number of fuel characteristics of northern Rocky Mountain forested stands. Results from these studies show that planar intercept techniques are inferior to fixed area plots methods primarily because fuels vary at different time and space scales, and these scales must be represented in fuel sampling efforts. The problem is though that fixed area plot sampling methods are difficult to learn and may take more time than planar intercept techniques. Therefore, we present some new research results that show possible ways to speed up fixed area plot sampling using subsampling, truncated measurements, and visual estimates. Results from these studies have profound implications for fuel applications such as fuel sampling design, fuelbed classification, and fuel mapping. We need to explore a new frontier of wildland fuel sampling to develop new methods that are both popular with managers and accurate for loading estimates. Keywords: planar intercept, fixed area plot, fuel loading, variability Presenter Bio: Keane's most recent research includes 1) developing ecological computer simulation models for the exploring landscape, fire, and climate dynamics, 2) conducting field research on the sampling, describing, modeling, and mapping of fuel characteristics, and 3) investigating the ecology and restoration of whitebark pine. He received his B.S. degree in forest engineering from the University of Maine, Orono; his M.S. degree in forest ecology from the University of Montana, Missoula; and his Ph.D. degree in forest ecology from the University of Idaho, Moscow

112. Modeling fuels and fire effects in 3D with FuelManager and STANDFIRE François Pimont, Engineer-Scientist. Mediterranean Forest Research Unit (URFM, Avignon), National Institute for Agricultural Research (INRA) Additional Authors: Russel Parsons, Research Ecologist, USFS RMRS Jean-Luc Dupuy, Research Scientist, INRA URFM 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Eric Rigolot, Engineer Scientist, INRA URFM François de Coligny, Engineer Scientist, INRA AMAP Abstract: Scientists and managers need ways to assess how fuel treatments alter fire behavior, yet few tools currently exist for this purpose. FIRETEC and WFDS recently emerged as promising models in this context since they explicitly account for 3D fuel structure, but there is a need for a tool to facilitate the development of heterogeneous landscape-scale fuel beds based on available data, to mimic natural change over time and disturbances and to manipulate the fuel beds based on fuel treatments. Here, we present a spatially explicit fuel modeling system designed to interact with physics-based models. The flexible approach facilitates modeling fuels across a wide range of detail. Large trees or shrubs with specific coordinates are modeled as individual “Plants”, while understory plants are modeled as collections of plants called “LayerSets”. Both Plants and LayerSets contain various fuel particles (leaves, twigs) with various properties including surface-area-to-volume-ratio. A wide range of vegetation and treatments can be modeled, analyzed quantitatively and visualized in a 3D viewer. This framework is implemented in the “Fire” library of the CAPSIS platform, an open source collaborative Java-based software that hosts a wide range of forestry-related models and provides common functions, visualization tools, shared libraries and source code. FuelManager and STANDFIRE are two distinct modules of the CAPSIS platform that rely on this “Fire” library. They thus have the same capabilities to model fuel treatments, export to physics-based models and fire effects. However, FuelManager implements its own growth, succession models and empirical tree distributions, whereas STANDFIRE uses vegetation growth from the Forest Vegetation Simulator (FVS) and information from FFE-FVS. Using STANDFIRE, FVS stands can be converted to input data for FIRETEC or WFDS. More details are available in the online documentation (http://capsis.cirad.fr/capsis/help/fireparadox and http://capsis.cirad.fr/capsis/help_en/standfire). FuelManager and STANDFIRE facilitate rapid development of 3D data sets required for physics-based fire models, eliminating significant fractions of the effort required to study the impact of fuel beds on fire, by linking fuel, fire-behavior and fire-effects models. They also improve the replicability of fire modeling studies. In this presentation, the modeling system will be overviewed, and several applications will be demonstrated. Keywords: Fuel Modelling; Physics-based models; Fuel treatments; FIRETEC; WFDS; FFE-FVS; CAPSIS. Presenter Bio: François Pimont holds an Engineer-Scientist position in the Mediterranean Forest Research Unit (URFM, Avignon) of the National Institute for Agricultural Research (INRA, France). He is an engineer from Ecole Polytechnique and AgroParisTech and obtained his PhD thesis of Environmental Sciences in 2008. His domains of experience and expertise are wildfire physics and wildfire fire behaviour modelling, as well as fuel modelling and fuel remote sensing. Since 2005, he has developed a close collaboration with LANL on fire modelling and more recently with USFS (RMRS) on fuel modelling. He obtained the Silver Medal of French Academy of Agriculture in 2011.

113. Next-Generation Fuels Mapping at Regional Scales: accounting for uncertainty and spatial variability Susan Prichard, Research Scientist, University of Washington Additional Authors: Nancy H. French, Senior Research Scientist, Michigan Technological University Donald McKenzie, Research Ecologist, USDA Forest Service Maureen C. Kennedy, Assistant Professor, University of Washington Michael G. Billmire, Research Scientist, Michigan Technological University Eric S. Kasischke, Professor, University of Maryland 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Abstract: At regional to national scales, the structure and composition of wildland fuels vary in both space and time, but current geospatial fuel datasets do not account for this variability. This makes inventories of smoke and emissions challenging to produce, hard to replicate, and difficult to defend both scientifically and for their diverse applications to policy and management. We are developing a comprehensive geospatial database of fuel loadings for the Conterminous United States (CONUS) and Alaska (AK) that will incorporate variability by fuel strata and enable quantitative accounting of fire emissions with a realistic assessment of uncertainty. The project has three objectives: (1) develop methods to map and validate wildland fuel loadings for emissions modeling that incorporate spatial and temporal variation at multiple scales; (2) create and distribute geospatial fuels and emissions products for regional and national smoke management and emissions inventories; and (3) evaluate sources of uncertainty and data gaps for emissions estimates using a sensitivity analysis of emissions models to fuel variability informed by distributions of fuel characteristics. Here we present an analysis that sets the stage for the rest of the project: an in-depth accounting of spatial variability in fuel loadings across the CONUS, which will provide the raw material for maps, uncertainty estimates, and emissions calculations. Keywords: wildland fuels, emissions, smoke, mapping Presenter Bio: Susan Prichard is a fire ecologist with the University of Washington and works through a cooperative agreement with the US Forest Service Pacific Northwest Research Station. Her main interests are in the effects of fire and other disturbances on forest dynamics, climatic change on forest ecosystems, and fuel reduction treatments. Her current projects with include fuel consumption and emissions modeling; scientific development of the Fuel Characteristic Classification System, Consume, and FEPS; and wildland fire interactions with past treatments and wildfires.

114. Changes of masticated fuelbed properties over time in the western US Pamela Sikkink, Biological Scientist, U S Forest Service Rocky Mountain Research Station Additional Authors: Robert Keane, Research Ecologist, U S Forest Service Rocky Mountain Research Station Theresa Jain, Research Forester, U S Forest Service Rocky Mountain Research Station Abstract: Mastication, or the physical breakup of trees and shrubs by machinery, is becoming a common forest treatment to reduce wildfire hazard, especially where prescribed burning is risky. When fuels from the chopping and chunking of trees and shrubs by the masticator are left on the ground to decompose, however, several significant changes occur that can affect their ability to burn. In this study, we document some physical and chemical changes that occurred when masticated fuels are left on the ground in mixed-conifer masticated forests of the western U.S. We investigated how the physical fuel particles are affected by age (0-10 years), climate (mesic and xeric), and mastication methods (rotating head, horizontal drum, chipper, and mower). We also investigated how age affects moisture dynamics of the multi-aged mastication beds and whether age affects burning character using paired (young and old) masticated fuel beds. We found that most fuel particle characteristics examined in this study changed little between older masticated beds and those treated more recently. The proportion of shapes was fairly constant over time as were size measurements of the particles and density measures. Fuel bed depths were highly variable in the field but they were statistically significant only by mastication method (Kendall’s tau = - 0.559, p = 0.011) and not with age. The characteristics that were significantly correlated with age included (1) bulk density of the fuel layers (Kendall tau = 0.118, p = 0.048), (2) percent nitrogen (Kendall’s tau =0.310, p = 4.743 e-05), and (3) carbon-to-nitrogen ratio (Kendall’s tau = -0.194, p = 1.831 e-10). Water absorption/desorption properties also correlated with age. The burning experiments conducted in this study revealed that all of the masticated fuels burned, regardless of age, but there was not a strong flaming front for the entire length of the bed. In general, 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

the fuels were more likely to creep and smolder than to flame, but smoldering and spotting are the more major concerns for management. Keywords: fuel treatment, mastication, experimental burns Presenter Bio: Pamela G. Sikkink is a postdoctoral biological scientist with the U.S. Forest Service, Rocky Mountain Research Station. She is employed with the Forest and Woodlands Ecosystem Program in Moscow, Idaho, but stationed with the Missoula Fire Sciences Laboratory in Missoula, MT. Her most recent research includes (1) sampling and describing masticated fuels, (2) evaluating methods of describing fire and burn severity, (3) sampling and describing grassland fuels, and (4) testing sampling methods for long-term vegetation monitoring in Yellowstone National Park . She received her PhD in Forestry from The University of Montana.

115. Estimating Litterfall Rates Following Stand-replacement Disturbance in Northern Rocky Mountain Ecosystems Christine Stalling, RMRS Missoula Fire Sciences Lab Additional Authors: Robert Keane, Research Ecologist, RMRS Missoula Fire Sciences Lab Abstract: It is generally assumed that insect and disease epidemics, such as those caused by the mountain pine beetle, predispose damaged forests to high fire danger by creating highly flammable fuel conditions. While this may certainly be true in some forests, these dangerous fuel conditions may only occur for a short time when evaluated at a landscape level. Others feel these epidemics may cause high surface fuel loadings when the dead material from dead trees falls to the ground; these high fuel loadings may result in abnormally severe fires. This study evaluates, through intensive field collections and simulation modeling, the effect that exogenous disturbance events, namely fire and beetles, have on future fire hazard and risk. We measured surface fuel deposition rates for a number of forest types after stand-replacement wildfire events and after beetle outbreaks to quantitatively describe fuel dynamics to ultimately estimate resultant fire behavior in heavy mortality stands for up to 10 years after the disturbance. Fuel deposition was measured using semi-annual collections of fallen biomass sorted into six fuel components (fallen foliage, twigs, branches, large branches, logs, and all other material). This litterfall was collected using a network of seven, one meter square litter traps installed on plots established on seven sites across the northern Rocky Mountains USA. We also measured stand and surface fuel characteristics of the plot using FIREMON techniques at the beginning, and every year till the end of the study. Results indicate that after the initial pulse of needlefall 2-3 years after disturbance, few fine woody fuels are actually deposited over the next 10 years. Keywords: fuel deposition, fuel characteristics, litterfall, stand-replacement, disturbance, fire severity Presenter Bio: Christine Stalling is a biologist with the RMRS Fire, Fuel and Smoke Program at the Missoula Fire Science Lab. She has been involved with research focusing on landscape response to disturbance over much of her career. Christine is currently involved in research focused on fuels, fuels dynamics, and fire ecology.

116. Post Treatment Fuel Loading Differential in Two Logged Areas of Banff National Park Erin Tassell, Fire Technician, Parks Canada Agency Abstract: The town of Banff in Banff National Park has an extensive network of fuel breaks adjacent to the townsite designed to break up fuel continuity. However, there is still threat of large wildfires moving towards Banff and threatening outlying infrastructure due to prevailing winds and continuous stands within the Bow Valley. In order to manage and mitigate these risks, fuel loading must be reduced. In an 85 ha area upwind of the Banff townsite, a landscape level fuel break has been developed in order to manage fuel loading, tying it into existing shrub and meadow features. In order to maintain this area, future implementation of prescribed fire is also necessary. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

The fuel reduction unit was divided into two areas where different methods of mechanical removal were used. The area dominated by pine was completed using complete stem removal and skidded offsite to landings for processing. The second block included a higher spruce component which was cut into shortwood and processed on site, leaving more course woody debris (CWD). Data was collected post logging and CWD data was divided among three fuel moisture classes from the Canadian Forest Fire Danger Rating System (CFFDRS)-Fine Fuel Moisture Code (FFMC), Duff Moisture Code (DMC), and Drought Code (DC). In order to represent total fuel available to burn, the DMC and DC were combined to a value known as the Build Up Index (BUI). The analysis showed that there was significant difference between the fuel treatment blocks. This information will be used for future prescribed fire implementation and fuel break construction. Keywords: Fuel Loading, Logging, Prescribed Fire Presenter Bio: Erin Tassell began her career in fire management with Alberta Sustainable Resource Development as a helitack firefighter. After four years, she took a position as an initial attack firefighter with Parks Canada and has been with Parks Canada Agency for seven years. She recently became the fire technician for Banff National Park. She has also worked two seasons as a firefighter in Western Australia for the Department of Parks and Wildlife and holds a Bachelor of Science degree in Natural Resource Management from the University of Northern British Columbia.

117. Simulating the Joint Impacts of Wildfires and Fuel Management on Landscape Resiliency in Central Oregon USA. Ana Barros, Oregon State University Additional Authors: Alan Ager, Research Forester at the Fire Sciences Lab in Missoula Montana. Michelle Day, Faculty Research Assistant at the College of Forestry, Oregon State University in Corvallis, Oregon Haiganoush Preisler, Statistical Scientist, Pacific Southwest Research Station, Albany, California Abstract: In this study we used the agent-based landscape model Envision to examine predicted changes in burned area and fire severity under increasing levels of fuel management and wildfire over a 50 year simulation period. The study area was a 2 million ha fire prone region of central Oregon, USA containing two national forests. Envision simulates the concurrent temporal dynamics of wildfire, vegetation, and forest management in a highly detailed landscape at annual time steps. Alternative forest management scenarios corresponded to different treatment intensities - scenarios designed to treat 2%, 3% and 5% of the landscape per year with a combination of mechanical thinning, prescribed fire and fuels mastication. Treatments were applied according to ecological and administrative preferences as specified in national forest plans. Wildfire was simulated at four scenarios such that 0.5%, 1%, and 2% of the landscape is burned per year. We simulated all combinations of management and wildfire activity in a factorial design and then tested for temporal trends in future burned area and high severity fire. We also quantified the reduction in future area burned as a function of both cumulative area treated and area burned. The analysis identified the point of diminishing returns in terms of forest management and how wildfire feedbacks can complement fuel reduction activities to reduce burned area. Our results contribute to a better understanding of how feedbacks from wildfire and management interact in space and time, and can inform forest restoration activities underway on western US national forests. Keywords: Simulation, forest and fuel management, wildfire, national forests Presenter Bio: Ana Barros is Postdoc Scholar at the College of Forestry, Oregon State University in Corvallis, Oregon. She received hes PhD from the University of Lisbon. Her research interests include forest landscape and wildfire modeling, risk analysis, and the study of coupled social-ecological systems.

5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

118. Forest fuels and potential fire behavior twelve years after variable-retention harvest in lodgepole pine Justin Crotteau, University of Montana Additional Authors: Christopher R. Keyes, Research Professor, University of Montana Elaine K. Sutherland, Research Biologist, USFS Rocky Mountain Research Station David K. Wright, Ecologist, USFS Rocky Mountain Research Station Joel M. Egan, Entomologist, USFS Forest Health and Protection

Abstract: Variable-retention harvesting (VRH) in lodgepole pine offers an alternative to conventional, even-aged management. This harvesting technique promotes structural complexity and age-class diversity in residual stands and promotes resilience to disturbance. We examined fuel loads and potential fire behavior 12 years after two modes of VRH (dispersed and aggregated retention patterns) crossed by post-harvest prescribed fire (burned or unburned) in central Montana. Results characterize 12-year post-treatment fuel loads. We found greater fuel load reduction in treated than untreated stands, namely in the 10- and 100-hr classes (p=0.002 and p=0.049, respectively). Reductions in 1-hr (p<0.001), 10-hr (p=0.008), and 1000-hr (p=0.014) classes were greater in magnitude for unburned than burned treatments. fire behavior modeling incorporated the regenerating seedling cohort into the surface fuel complex. our analysis indicates greater surface fireline intensity in treated than untreated stands (p<0.001), and in unburned over burned stands (p= 0.001) in dry, windy weather. although potential fire behavior in treated stands is predicted to be more erratic, within-stand structural variability reduces probability of crown fire spread. overall, results illustrate tradeoffs between potential fire attributes that should be acknowledged with vrh. Keywords: northern Rocky Mountains; Tenderfoot Creek Experimental Forest; Little Belt Mountains; multiaged silviculture; fuels accumulation; custom fire behavior fuel models Presenter Bio: Justin studies silviculture at the University of Montana. His research focuses on the effects that fire and mechanical treatment have on residual stand structure and forest change.

119. Multi-dimensional cost-effectiveness of fuel treatments in dry mixed conifer forests: an inventory originated analysis Jeremy Fried, Research Forester, U.S. Forest Service Pacific Northwest Research Station Additional Authors: Theresa B. Jain, Research Forester, Rocky Mountain Research Station Sara Loreno, Research Associate, Portland State University Conor Bell, Graduate Research Assistant, University of Idaho Robert Keefe, Assistant Professor, University of Idaho Abstract: The BIOSUM (Bioregional Inventory Originated Simulation Under Management) framework was applied to over 5000 forest inventory plots in five western U.S. states to test the short and long- term effectiveness and net costs of 10 alternative silvicultural prescriptions, commonly implemented in this region, that are designed to achieve stand improvement, commercial thinning or precommercial thinning, depending on stand structure and management objectives. Because these plots form a representative sample of the entire forest, findings derived by analyzing these plots can be extrapolated to the larger landscape in which they are placed, and reflect variation in the costs, revenues and products extracted by virtue of the distribution of initial stand conditions, topographic settings and road accessibility. We score effectiveness on 4 dimensions describing the level of hazard posed by ladder and crown fuels (fire resistance), and by stand resilience as defined by basal area proportion in resilient species and larger size classes. Costs are estimated using the OpCost meta-model that matches terrain and prescription tailored harvest systems, and the libraries of harvest cost equations that represent 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

them, to each modeled stand, and offset, to varying degrees, by assumed sales of merchantable and energy wood generated by fuel treatment. Short and long-term single and multi-dimensional cost effectiveness is estimated and related to stand structure and context, policy constraints such as diameter caps, and other management objectives such as climate benefits produced by forest carbon management. Results offer guidance on when stand structure and context leads to fuel hazard reduction that is compatible with climate benefits and when cost reduction via enhanced production simultaneously enhances treatment effectiveness.

Keywords: fuel treatment cost effectiveness longevity carbon implications Presenter Bio: Fried is a research forester with the USFS PNW Research Station, applying systems analysis and geographic information science to contemporary natural resource management issues. He studies economic feasibility of landscape-scale fuel treatments, forest carbon accounting, initial attack simulation, climate change effects on wildland fire, and builds inventory based models of fire effects. He previously served on the forestry faculties at Michigan State and Helsinki Universities and has a Ph.D. in Forest Management and Economics from UC Berkeley and an M.S. in Forest Ecology and Soils from Oregon State University, where he is currently a courtesy faculty in the Department of Forest Engineering, Resources and Management.

120. The effects of a long-term, landscape-scale, fuel management program on three-dimensional fuel loading and distribution. Nicholas Skowronski, Research Forester, USDA Forest Service Additional Authors: Michael Gallagher, PhD Candidate, Rutgers University Kenneth Clark, Research Forester, USDA Forest Service Inga La Puma, Visiting Scholar, Rutgers University Albert Simeoni, Senior Manager, Exponent Inc Abstract: Prescribed fire (RxB) has been used as a management tool in the 445,000 ha Pinelands National Reserve in New Jersey, USA since the early 1900s. To estimate the effects that this burn program has had on fuel loading and distribution at the landscape-scale, we intersected a wall to wall airborne laser scanning (ALS) dataset, a geodatabase of historic burn treatments, and a landcover map. The ALS data was collected during leaf-off conditions in 2013 with a pulse density of 8 pulses m-2. These data were pre-processed and models were developed to predict vertical profiles of canopy bulk density that could accurately detect ladder fuels and understory vegetation at 20 x 20m resolution from estimates of CBD at spatially coincident field plots. Preliminary results indicate that these models account for more than 80% of the variability in CBD in field evaluation plots. The models were then applied to the raw LiDAR outputs to develop wall to wall rasters of canopy fuel loading for the study area. The historic RxB data was rasterized to represent both the time since the last fire treatment and the frequency of treatments since 1979 for each pixel on the landscape. Each pixel was then assigned to a cover classification per the landcover map. Preliminary results indicate that burn frequency is positively correlated with a reduction in fuels in the understory while time since last burn has a weaker relationship. Our research assists wildland fire managers evaluate the effectiveness of fuel reduction treatments, and the configuration of fuel breaks at the landscape scale. Keywords: Fuel treatments, LiDAR, Landscape-scale effects Presenter Bio: Nick Skowronski has been a Research Forester with the USDA Forest Service Northern Research Station (NRS) since 2007. His research is focused on the impacts that land management activities have on fuel loading, carbon cycling, and habitat quality. He is particularly interested in estimating canopy fuels in three-dimensional space using lasers on the ground and in aircraft. Having worked formerly in fire management, Nick is extremely interested in developing better relationships 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

between scientists and land managers so that research can become more integrated with the needs in the field.

121. Driving fire behaviour models with forest inventory data in Canada Dan Thompson, Forest Fire Research Scientist, Canadian Forest Service Additional Authors: Brian Simpson, Forest Analyst and Modeller, Canadian Forest Service Bill de Groot, Research Scientist, Canadian Forest Service Chelene Krezek-Hanes, Forest Fire Analyst, Canadian Forest Service Steve Taylor, Research Scientist, Canadian Forest Service Mike Wotton, Research Scientist, Canadian Forest Service Abstract: The development of crown fire models that utilize stand structure information necessitates a new approach to the description and categorization of fuels. In Canada, the current Fire Behaviour Prediction (FBP) system uses 16 fixed fuel types grouped largely by species; these 16 types feature a small number of fixed descriptors including crown base height and crown fuel load, although, with one exception, these properties do not vary in the fire behavior models Emerging semi-empirical crown fire models allow for variable stand structural attributes such as crown base height, surface fuel load, and crown bulk density. Operationalizing this next generation of crown fire models requires new tools in order to generate the required inputs at the landscape scale. Forest inventory ground plots (both silvicultural and ecological) in Canada provide a near-complete description of a local forest stand and are suitable to drive fire boreal crown fire models with little modification, though are spatially sparse. These ground plots also allow for a quantitative examination of the regional and within-type variability of fuel loading. Recently, hybrid models of forest inventory ground plot and remote sensing data is allowing, for the first time, the creation of continuous rasters of forest structure across Canada's 270 Mha of boreal forest. While these forest structure maps provide only a fraction of the information required for a structure-driven crown fire model, more detailed information from ground plots can be used to supplement the less detailed spatial models. Moreover, these continuous maps of forest structure can be used to categorize fuels according to the existing 16 FBP fuel types based on their structure, rather than by leading species alone. These structure-driven, “species agnostic” fuel type maps are especially useful in areas with a limited number of leading species and high phenotypic plasticity, such as is found in the boreal spruces. Taken together, these approaches offer a more holistic and continuous approach to fuel description for fire behaviour models that is inclusive to the wide diversity of forest structure found within the boreal. Keywords: Fuels, forest inventory, boreal, Canada Presenter Bio: Dan Thompson is a forest fire research scientist with the Canadian Forest Service. Based out of the Northern Forestry Centre in Edmonton, Alberta, Dan is a member of the Canadian Forest Fire Danger Rating Group, and focuses on a process-based understanding of fuel moisture as well as the characterization of northern fuels. Dan is also part of a project team utilizing the FIRETEC model to examine FireSmart fuel treatment effectiveness in boreal fuels.

122. Utilizing drought science and information in wildfire management decision context Timothy Brown, CEFA, Desert Research Institute Additional Authors: Tamara Wall, Assistant Research Professor, CEFA, Desert Research Institute Abstract: Dryness enables fire. The drought and wildfire nexus encompasses a plethora of human and physical impacts of which most are negative but not all. During recent years drought in the West has been enhanced by warmer temperatures, further exacerbating depletion of soil and vegetation moisture. Trees stress and become susceptible to beetle kill. Given the right slope of the terrain or 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

strong enough winds, the dryness exhibits itself through extreme burning conditions that rapidly spread, are difficult to control, and burn with a severity damaging to the ecosystem and are costly to suppress. Drought limits effective fuels management through much needed controlled burning on the landscape. It inhibits the reduction of hazardous fuels and the restoration of landscapes to improve ecosystem health. This feeds back in the fire-climate system, further increasing the potential for high impact fire. On the positive side, drought encourages management flexibility and adaptability, and increases interagency cooperation. Some ecosystems are adapted to drought in a way that allows for fire to initiate natural succession and regrowth.

The National Integrated Drought Information System (NIDIS), a multi-institution partnership established by congress in 2006 with roots in the National Drought Policy Act of 1998, has been establishing Drought Early Warning Systems (DEWS) with emphasis on monitoring, forecasting, and outlook tools; developing management plans and appropriate response actions; and communicating information about drought conditions and anticipated impacts. The science of drought and wildfire are beginning to allow these objectives to become realized in wildland fire management. In this presentation, we summarize the state of science in drought and fire, highlight a range of impacts of drought on the coupled physical fire- human system, and discuss future directions for utilizing drought science and information in wildfire management decision context. Presenter Bio: Dr. Brown is a Research Professor at the Desert Research Institute (DRI) in Reno, Nevada. His primary academic interests include analysis of wildland fire-climate and fire-weather connections; the fire environment; applications development for wildland fire management planning, decision- making and policy; and deliberate co-production of knowledge. Dr. Brown is Director of the Western Regional Climate Center, and established and directs the Program for Climate, Ecosystem and Fire Applications (CEFA). He is graduate faculty in the University of Nevada, Reno Atmospheric Sciences Program, and an Adjunct in School of Earth, Atmosphere and Environment, Science Faculty, Monash University, Melbourne, Australia.

123. Successful Stewardship Begins with Trust: The Southern Blues Restoration Coalition Dana Skelly, Assistant Fire Staff Abstract: The Malheur National Forest, at the southern edge of the Blues Mountains of Eastern Oregon, is at the center of one of the US Forest Service’s 23 Collaborative Forest Restoration Landscape Projects. The project is sponsored by the Southern Blues Restoration Coalition, which represents a number of governmental, non-profit, educational, business, and other community interests from around that region. At the heart of the project are two collaborative groups, Blue Mountain Forest Partners and the Harney County Restoration Collaborative. These groups were formalized in 2006 and 2008 and include representatives from all across the spectrum, including members of the environmental community as well as the timber industry. The relationships that have grown from investment in these collaboratives have directly resulted in a number of increased efficiencies across the forest. The forest has not been litigated since 2007. Environmental analyses which used to take multiple years and cover only a few thousand acres are now completed in less than two years and cover entire sub-watersheds. The forest has tripled its commercial output in the last three years, resulting in increased jobs locally and literally saving the last area mill from closure. The accomplishments continue, and this example of the benefits of such an investment—in the broadest sense of community to steward a people's forest—provide an example that will reinvigorate land managers and stakeholders from all sides of the conservation arena. In this presentation we give an overview of the collaborative process in practice and share lessons that will benefit others. Keywords: Collaborative, Fuels Treatments, Restoration, Resiliant Landscapes, Commercial Harvest 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Presenter Bio: Dana Skelly is the Assistant Fire Staff and Fuels Program Manager for the Malheur National Forest, in Eastern Oregon.

124. A Framework for Collaborative Learning: Forest Fuels and Vegetation Monitoring in the Southern Blue Mountains Becky Miller, Lead Research Assistant, Oregon State University Additional Authors: James Johnston, PhD Candidate, Oregon State University John Bailey, Associate Professor, Oregon State University Abstract: The Canyon Creek Complex burned more than 80,000 acres on the Malheur National Forest (MNF) in eastern Oregon in 2015. The fire burned during the third year of an ambitious 10-year landscape-scale forest restoration program undertaken with funds from the Collaborative Forest Landscape Restoration Program (CFLRP). CFLRP is a Congressionally authorized program supporting collaborative, science-based ecosystem restoration. Mechanical thinning and prescribed fire are tools being used to create resilient forest landscapes across approximately one million acres of Forest Service lands. CFLRP funds also support an exemplary multi-party, interdisciplinary monitoring program investigating the socio-economic and ecological effects of restoration treatments. This novel framework facilitates community participation and provides scientific data to aid the decision making process. Collaborators, including the Blue Mountains Forest Partners, MNF, and Oregon State University, have completed two years of data collection that informs adaptive management, generates baseline data to monitor long-term change, and provides information directly applicable to emerging questions. We will describe our forest collaborative framework, monitoring protocols being used and provide preliminary answers to a variety of questions including: 1. How do treatments influence fuels and fire behavior? 2. How do treatments affect forest structure, composition and function? We will also share preliminary information about the effects of the Canyon Creek Complex fire and its impact on present and future CFLRP efforts. Keywords: Collaborative Learning, Fuels Treatments, Forest Restoration, Adaptive Management Presenter Bio: Becky Miller is lead technician of the Malheur National Forest CFLRP Forest Vegetation and Fuels Monitoring Program. She is a first generation college graduate, earning her B.S in Bioresource Research from Oregon State University and currently works for the Wildland Fire Lab in OSU’s College of Forestry. She has conducted several independent research projects, most notably investigating relationships between intra-canopy microclimate and phenological response of Douglas-fir trees. Her current work with the Malheur National Forest exemplifies the dynamic and highly successful collaborative effort between community stakeholders, the USFS and private forest managers in the Southern Blue Mountains.

125. Fire Adapted Communities - Networking on a Local & National Scale Jerry McAdams, Wildfire Mitigation Coordinator, Boise Fire Department Forest Shafer, Tahoe Fire and Fuels Team Abstract: The Fire Adapted Communities Learning Network (FACLN) encourages the development and sharing of best practices to accelerate the adoption of fire adapted community concepts nationwide. The FACLN supports eighteen hub organizations that have committed to implementing and sharing the FAC-centric work that they are conducting, in order to increase their communities’ resilience to a real and historic threat of wildfire. Funding for the FACLN is provided by the USDA Forest Service’s Fire Adapted Communities Program and participants’ matching funds. The FACLN is cooperatively managed by the Watershed Research and Training Center and The Nature Conservancy. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Meeting one of the goals of the National Cohesive Wildland Fire Management Strategy of creating Fire Adapted Communities, the Boise Fire Department, in partnership with the FACLN, acts as the hub organization for the Ada County FAC. One of the underpinning values of the FACLN is that “Collaboration and partnerships are keys to successful adaptation.” In this spirit, many of the eighteen hub organizations have taken great strides to identify stakeholders and build collaborative partnerships, both internally and externally. By successfully partnering with federal agencies, not-for-profits, local university research teams, local environmental study groups, volunteer organizations, civic groups, small businesses, developers, as well as homeowners’ and neighborhood associations, these groups are reducing wildfire risk and increase community awareness. These hub organizations are also networking with each other on a national scale to build efficiencies and share best practices, an example of which was the Austin/Boise Fire Department learning exchange in 2015. The IAWF plays a part in this exchange of information by providing a venue to share FAC information. This presentation will highlight many of these collaborative partnerships, as well as some of the great FAC-centric work taking place around the nation. Keywords: Fire Adapted Communities Learning Network FAC FACLN Partnerships Best Practices Collaboration National Cohesive Strategy Presenter Bio: Jerry McAdams is the Wildfire Mitigation Coordinator for the Boise Fire Department, where he has worked for 15 years. He serves on City of Boise and Multiagency Wildfire Mitigation Committees. He is a Board Member for the International Association of Wildland Fire, a Council Member of the Boise District BLM Resource Advisory Council, and holds an NWCG certification as a Wildland Fire Investigator. Jerry has previously presented at an IAWF Conference, Backyards & Beyond, a Ready, Set, Go! webinar, and at the Southwest Idaho Wildfire Mitigation Forum. Jerry has a Bachelor of Science in Psychology from Boise State University. He enjoys working with communities and seeking new, cooperative partnerships.

126. Think bigger: statewide wildfire risk perceptions in Idaho. Thomas Wuerzer, Assistant Professor, Boise State University, Idaho Abstract: Risk perception matters! This project targets five of Idaho’s 13 Priority Landscape Areas (PLAs) with a state-wide risk perception survey across 20,000 households, and presents interagency and intergovernmental collaboration. These PLAs are designated as the highest areas of need for increasing the planning capacities to address development pressures and wildfire risks found in watersheds, forests, and communities. The PLAs are exemplar for Idaho’s landscapes and forests, as many of Idaho’s counties are at high risk for wildland fire and face significant development pressure in the WUI; but are mostly of rural character and lack planning capacity. Smaller local governments often lack knowledge, planning and legal resources to develop land-use policies that reduce fire risk in the Wildland Urban Interface (WUI). The work presented seeks to understand and address landowners’ risk perceptions and provides planning resources to rural communities where demand for new developments in the WUI is high. In doing so, the presentation shows the findings of a statewide resident’s survey with 20,000 participants investigating Idahoan’s risk perception. This provides not just insights on the effectiveness of preparedness programs but also pin-points the actually perceived high-hazards' areas via interactive mapping. Hence, the survey results are georeferenced, the findings are likely transferable to other regions in the North Pacific West with similar urban/rural development patterns and topology; therefore offers replication to other Western States Keywords: Risk perception, large residents' survey, novel mapping Presenter Bio: Dr. Thomas Wuerzer is Assistant Professor in Boise State University’s Department of Community and Regional Planning. His work and research are bridged by an in-depth background in 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Geographic Information Systems (GIS). He is actively researching regional planning issues of wildfires within the Wildland Urban Interface with focus on wildfire and related hazards' impacts on natural and built environment, wildlife and human habitat.

127. How Wildland Fire Leaders are Co-Managing Risk Michael Zupko, Executive Manager, Wildland Fire Leadership Council Abstract: Following one of the more extreme fire seasons in the last decade, US fire managers and decision makers are faced with a future of “the new normal” of larger, more intense, more destructive fires. To change this trajectory, the National Cohesive Wildland Fire Management Strategy provides strong guiding principles, rooted in cross boundary collaboration, to make meaningful progress towards resilient landscapes, fire adapted communities and a safe and effective wildland fire response. Change will not happen overnight however. And it comes at a cost. RISK. How much risk are we going to accept when lives, natural resources and property are in jeopardy? Wildfire and land managers were faced with this question routinely throughout the 2015 fire season. Hear how these risks are being evaluated, strategies that are being developed to address them and leaders' intent across multiple national partnerships. Keywords: Risk co-management leaders leadership Presenter Bio: Executive Manager, Wildland Fire Leadership Council

128. New Approaches for Mapping the Wicked Problem of Wildfire Cody Evers, PhD Student, Portland State University Additional Authors: Alan Ager, Operations Research Analyst, USFS Max Nielsen-Pincus, Assistant Professor, Portland State University Zahid Chaudhry, GIS Program Manager, USFS Abstract: The emerging national strategy for coordinating wildfire risk mitigation and suppression efforts calls for a broader, more integrated and interdisciplinary approach to managing wildfire. Developing an all-lands approach to risk management requires assembling, analyzing, and disseminating a large amount of GIS information. It also requires functionally integrating different kinds of data in new ways. While a number of states in the Western US are developing wildfire risk portals as planning tools, they are built on standard GIS functionality: layering of geo-referenced data and the ability to summarize that information within a given area. We argue that such information platforms need to be matched to the multiple scales and dimensions of wildfire. We present pioneering work that describes wildfire risk in a way that integrates landscape-level biophysical dynamics and community-level social information. Our work builds from on the concept of firesheds. Firesheds are spatial containers for determining which lands surrounding a community generate wildfire risk and which do not. Not surprisingly, risk transmission is complex and this complexity can be represented in multiple forms. One form is a network. Such transmission networks give further insight into how wildfire is a risk shared between different landowners. In order to display these data in novel ways, we use an online map journal to merge a number of emerging technologies and host them in a single place. This modular approach to application development allowed us the flexibility to integrate different datasets and modes of visualization (maps, D3.js, text). Modular application development has its unique challenges. It required to use multiple scripting languages and find ways to pass information from one scripting environment to another. It also required developing content using multiple online tools and hosting these tools and data in a many different places. The complexity of wildfires and the coupled social- ecological nature of risk require new forms of visualizing and exploring data. Our work illustrates a number of advances that can support a more targeted mitigation of wildfire risk. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Keywords: Wildfire risk; Risk transmission; Web portal; Decision support; Wildfire simulation; Risk mitigation Presenter Bio: Cody Evers is a PhD student in the Department of Environmental Studies, School of the Environment, Portland State University. He is a fellow of the NSF IGERT program at PSU examining how ecosystem services support urbanizing regions. His research focuses is on multi-scalar assessment and management of wildfire risks to communities as well as the benefits they provide. He has been working with wildfire simulation modeling for more than 5 years under projects funded by NSF, USDA and the JFSP.

129. Quantifying avoided wildfire emissions from significant wildfires in California Thomas Buchholz Additional Authors: David Saah, Managing principal, Spatial Informatics Group LLC and Associate Professor and Director of Geospatial Analysis Lab, Environmental Studies Department, University of San Francisco Jason Moghaddas, Research Scientist, Spatial Informatics Group LLC John Gunn, Director, Spatial Informatics Group – Natural Assets Laboratory David Schmidt, Research Scientist, Spatial Informatics Group LLC Abstract: The western U.S. has millions of acres of overstocked forestlands at risk of large, uncharacteristically severe or catastrophic wildfire owing to a variety of factors, including anthropogenic changes from nearly a century of timber harvest, grazing, and particularly exclusion of fire as an ecosystem process. Methods for fuel treatment intended to modify or reduce fire severity include mastication, chipping, or removal of sub-merchantable timber and understory biomass, pre-commercial and commercial timber harvest, and prescribed fire. Mechanisms for cost recovery for low value woody material removed or re-arranged for fuel treatments are not well established, and return on investment may only be realized as reduced potential costs of avoided future wildfires reduced by treatment implementation. While the benefits of fuel treatments in reducing wildfire severity are clear and well- documented in the scientific literature, the absolute probability of wildfire impacting fuel treatments or nearby areas within their effective lifespan are difficult to account for with certainty and are variable across the landscape. As market-based approaches to global climate change are being considered and implemented, one important emerging strategy for changing the economics of fuels treatments is to generate carbon emission offset credits. Carbon credits can theoretically be generated by projects that reduce potential emissions from wildfire, as by reducing effects of wildfire for a given portion of land, but avoided emissions must be weighed along with carbon sequestration form forest growth, emissions from treatment activities, risk of fire occurrence, wood product life cycles and many other factors. Development of carbon emission offsets as an effective tool for forest and fire mangers thus requires an integrated approach that considers wildfire probabilities and expected emissions, as well as net expected carbon sequestration or loss over time. We present a “real world” framework that is currently being developed on a case study site in El Dorado County, California. The study area is within mixed conifer forest and representative of 2nd growth forests across the Sierra Nevada. This framework integrates scientifically based models for predicting changes in fire behavior and related emissions, both with and without hazardous fuel treatments. Keywords: Wildfire emission accounting, fire and fuel modeling, life cycle assessment, carbon offsets Presenter Bio: Dr. Thomas Buchholz leads the Forest and Agriculture Team at the Spatial Informatics Group. Thomas has more than 12 years of experience in working with governments, academia, non- profits and the private sector in the management and economics of natural forests, timber plantations, and short rotation energy crops (e.g. willow shrub plantations) for biomass production. His recent work includes micro- and macroeconomic analysis and carbon accounting of forest wildfires and forest based 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

bioenergy use in the US, Europe and Sub-Saharan Africa. Thomas is an affiliate of the Gund Institute for Ecological Economics at the University of Vermont.

130. Estimates of biomass consumption based on MODIS Fire Radative Power overestimate global biomass consumption and carbon release. Bryce Kellogg Additional Authors: Nicole Vaillant , Ecologist, USFS Brian Wing, Research Forester, USFS Crystal Kolden, Assistant Professor, University of Idaho Alistair Smith, Assistant Professor, University of Idaho Abstract: Carbon emissions from wildfire are a significant contributor to global carbon flux. Studies quantifying global carbon emissions, such as the Global Fire Assimilation System (GFAS) use measurements of fire radiative energy (FRE) from MODIS as a proxy measure of biomass consumption. We evaluate this method using a unique pre and post fire LiDAR dataset to quantify biomass consumption from a large forest fire (10,844 ha) west of Sisters, Oregon. Our estimate of the combustion factor is an order of magnitude lower than previously reported estimates. This study constitutes a first test of the relationship between FRE and biomass consumption at a scale appropriate to MODIS data. These results suggest that either current systems, such as GFAS, drastically overestimate carbon release from global fires, or the temporal and spatial resolution of the MODIS sensor is simply too course to accurately characterize FRE. Both possibilities call into question the accuracy of estimates of global biomass consumption and carbon emissions based on MODIS derived FRE. Presenter Bio: Bryce Kellogg recently received his masters of environmental science from the Yale school of Forestry and Environmental Studies. His work focuses on remote sensing and wildfire ecology in the Western Untied States.

131. A new top-down method for estimating aerosol emissions applied to large wildfires in North America. Tadas Nikonovas, Swansea University, Wales, UK Additional Authors: Stefan Doerr, prof., Swansea University Peter North, prof., Swansea University Abstract: Particulate matter emissions from wildfires affect climate, weather and air quality. Global and regional aerosol emission estimates differ by a factor of up to 4 between different methods. Bottom-up inventories based on burned area and emission factors generally indicate smaller aerosol emissions than top-down approaches employing satellite fire radiative power (FRP) and aerosol optical thickness (AOT) observations. Using a novel approach, we estimate daily total particular matter (TPM) emissions from large wildfires in North American boreal and temperate regions. Moderate-resolution Imaging Spectro- radiometer (MODIS) fire location and aerosol optical thickness (AOT) datasets are coupled with HYSPLIT dispersion simulations. MODIS smoke plume observations are linked to sources estimating TPM emissions for up to two diurnal cycles prior to the observation. Unlike other top-down approaches, our method does not rely on emission factors or FRP to dry matter burned conversion factors. Daily TPM estimates include full diurnal cycle of emissions and do not depend on instantaneous FRP and biomass emission rate relationships observed during a satellite overpass. Day-to-day variability in TPM emissions for individual fires agree well with bottom-up Global Fire Emission Database (GFEDv4) and top-down Global Fire Assimilation System (GFASv1.0), Fire Energetics and Emissions Research (FEERv1) and Quick Fire Emission Dataset (QFEDv2) inventories. Regional TPM estimates show very close agreement with QFED and are higher by a factor of 2 to 4 compared to other 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

approaches. Significant variability occur amongst the existing methods, and also our estimates, for fires in different land cover types. TPM emission estimates for individual fire events based on our new method, (I) which do not rely on emission factors and (II) take into account the whole diurnal cycle of combustion, suggest higher aerosol emissions than most existing inventories. The results provide a benchmark for inventory performance comparison and contribute towards development of top-down emission factors. Keywords: Aerosol emissions, large wildfires, top-down estimates. Presenter Bio: Tadas Nikonovas is undertaking a PhD at Swansea University, Wales, UK. His research is focused on coupling satellite smoke aerosol retrievals with plume dispersion modelling to (I), estimate particulate carbon emissions from wildfires and (II), investigate current large discrepancies between different emission inventories.

132. Snag Dynamics and Fuel Succession Following Wildfires in the Eastern Cascade Mountains David Peterson, Research Forester, USDA Forest Service, PNW Research Station Additional Authors: Erich K. Dodson, Forester, USDA Forest Service Abstract: Large wildfires often produce pulses of fire-killed trees that become habitat for a wide range of vertebrate and invertebrate species; downed woody debris that contributes to soil stability and nutrient retention; and forest fuels that influence subsequent fire behavior and effects. The relative contributions of fire-killed trees to each of these functions depends, in part, on temporal patterns of decay and transitions from standing dead trees to downed woody debris. I will present results from a regional study designed to describe temporal patterns of snag decay and fall, surface fuel accumulations, and snag usage by cavity-nesting birds following stand-replacing wildfires in dry coniferous forests of the interior Pacific Northwest. We sampled fire-killed trees and surface fuels within a chronosequence of 159 forest stands that burned in stand-replacing wildfires between 1970 and 2007. Tree species and diameter strongly influenced snag longevity, with small-diameter snags falling faster than large-diameter snags and ponderosa pine snags falling faster than Douglas-fir and true fir snags. Most standing snags developed broken tops within 10-15 years following fire, after which snag fall rates declined notably. Wildlife cavities were found in about 2.5% of standing snags. Cavities were most commonly found in medium-diameter (30-60 cm) snags with broken tops in stands that had burned 10- 20 years prior to our survey. Surface woody fuels increased with time since fire, reaching maximum levels within 5-20 years after fire for small-diameter fuels (≤ 7.5 cm) and within 10-30 years after fire for large-diameter fuels (> 7.5 cm). Small diameter snags generally fell without being used as nesting habitat by cavity-nesting species, serving as short-term foraging habitat and as a source of downed woody debris and fuels. Larger diameter snags stand longer, on average, but still deposit surface woody debris during the first decade after fire as branches and tops break off. Although large-diameter snags are often highly valued as potential nesting habitat, our study suggests that cavity-nesting wildlife select and use snags of varying sizes and species. Keywords: CWD, fuel dynamics, snag dynamics, habitat Presenter Bio: David W. Peterson is a Research Forester with the USDA Forest Service, Pacific Northwest Research Station, in Wenatchee, Washington. Dave’s research focuses primarily on restoration and management of dry coniferous forests of the interior Pacific Northwest, with emphases on forest ecosystem responses to wildfires and post-fire forest management practices, including emergency slope stabilization treatments and post-fire logging. He also maintains ongoing research interests in dry forest restoration treatment effects, forest vegetation responses to climatic variability and change, and oak savanna ecology.

133. Estimating canopy bulk density distribution using calibrated t-LiDAR indices 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

François Pimont, Engineer scientist, INRA Additional Authors: Jean-Luc Dupuy, Research Scientist, INRA URFM Maxime Soma, PhD Student, INRA, URFM Eric Rigolot, Engineer Scientist, INRA URFM Frédéric Jean, Engineer assistant, INRA URFM Abstract: Canopy leaf biomass distribution is a factor of fire behaviour, which affects rate of spread, intensity and crown fire potential. At plot scale, the inventory-based approach combines a stem inventory, allometric equation for leaf mass and its vertical cumulative distribution to estimate leaf load and bulk density profile. This approach is still very time-consuming and allometric equation performance can be highly variable among sites. We propose a method to estimate the spatial distribution of leaf bulk density from TLS scans. Our method is based on relative density indices within the point cloud that have already been computed from TLS data to estimate leaf area density. Here, the indices are computed in spherical volumes of 70 cm and a field calibration is used to provide a rapid but robust estimation of the bulk density distribution. To proceed to the calibration, some calibration spherical volumes of 70 cm are selected within plots and identified with a polystyrene ball at their center. These polystyrene balls entail to identify the location of the calibration volumes within the point clouds once scanned, but also permit field crew to collect leaf biomass inside these volumes. Relative density indices can thus be computed at these locations and be calibrated with the biomass collected in the field at the same location. This method was applied to four 12 m diameter contrasted plots selected in a Quercus pubescens forest. The calibration was carried out over 40 calibration volumes. The calibration was satisfactory, with a determination coefficient of 0.73 and unbiased calibrated indices. Calibrated indices were then computed to all nodes of a grid to estimate the 3D leaf bulk density of each plot. Vertical profiles were evaluated against the inventory-based approach, using an allometric equation and a vertical cumulative distribution fitted on 10 trees harvested and sampled on the study site. Our method provides very encouraging results at a labor cost less than ten times cheaper than the inventory-based approach. In the future, separating leaf and wood returns should improve the results and potentially entails the application of the method to thin roundwoods. Keywords: Canopy fuel; Quercus pubescens; LIDAR; TLS. Presenter Bio: François Pimont holds an Engineer-Scientist position in the Mediterranean Forest Research Unit (URFM, Avignon) of the National Institute for Agricultural Research (INRA, France). He is an engineer from Ecole Polytechnique and AgroParisTech and obtained his PhD thesis of Environmental Sciences in 2008. His domains of experience and expertise are wildfire physics and wildfire fire behaviour modelling, as well as fuel modelling and fuel remote sensing. Since 2005, he has developed a close collaboration with LANL on fire modelling and more recently with USFS (RMRS) on fuel modelling. He obtained the Silver Medal of French Academy of Agriculture in 2011.

134. Effects of Stand Thinning in Modifying Crown Fire Behavior in a Black Spruce Stand in Interior Alaska Eric Miller, Fire Ecologist, BLM Alaska Fire Service Additional Authors: T. Scott Rupp, Director, Scenarios Network for Alaska & Arctic Planning, University of Alaska Fairbanks Bret W. Butler, Research Mechanical Engineer, USFS Rocky Mountain Research Station Roger D. Ottmar, Research Forester, USFS Pacific Northwest Research Station Robert E. Vihnanek, Supervisory Forester (Retired), USFS Pacific Northwest Research Station Randi R. Jandt, Fire Ecologist, Alaska Fire Science Consortium, University of Alaska Fairbanks Robert Schmoll, Operations Forester, State of Alaska, Division of Forestry 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

T. Kato Howard, Fuels Management Specialist, BLM Alaska Fire Service Abstract: The effects of stand thinning in modifying fire behavior was tested by exposure to active and passive crown fire in a black spruce forest in interior Alaska. The forest was thinned to a nominal 8 foot bole spacing and pruned to a height of 4 feet in 150 X 150 m treatment blocks. Canopy fuel load was reduced from 2.5 to 1.1 kg/m^2. Stand characteristics and fuel attributes were measured in canopy, surface, and subsurface fuel layers inside and outside the treatment blocks. Weather was logged inside and outside the thinning over two seasons and indicates negligible differences in fine dead fuel moisture during the afternoon burn period. The thinning was windier at all hours of the day. One thinning block was exposed to an active crown fire in a fully instrumented and monitored prescribed fire in 2009. At the point of impingement the fire was spreading at 40 m/min with a fireline intensity of 39 MW/m. Mass flow rate was comparable to other measured crown fires in mostly boreal forests. The second thinning block was exposed to mostly passive crown fire during a wildfire in 2015. Crown fire immediately dropped to the surface in both thinning blocks. Fire continued to penetrate into the thinnings some distance through spotting, surface flaming, and smoldering in the duff but eventually self-extinguished, leaving 38 and 63% of the thinning areas unburned. Analysis of weather, fuel bed structure and moisture content, and fire behavior suggests co-dependency between the surface and crown fire which was broken by removal of canopy fuels. Inbibition of surface spread may have been due to high moisture content in feathermosses or to post-thinning changes in plant abundance. Keywords: Hazard fuel reduction, stand thinning, crown fire, mass flow rate, canopy bulk density Presenter Bio: Eric Miller began his fire career in Yellowstone National park with the Fire Effects Monitoring Program. He has been Fire Ecologist with the BLM Alaska Fire Service in Fairbanks, Alaska since 2008.

135. A Legacy of Fire Use: Fire Management and Fire Use in Eastern Province of Zambia LaWen Hollingsworth, Fire Behavior Specialist, Rocky Mountain Research Station Additional Authors: Darren Johnson, Climate Change Advisor, USDA Forest Service International Programs Gift Sikaundi, Principal Information Systems Officer, Zambia Environmental Management Agency Sylvester Siame, Forestry Extension Officer, Zambia Forestry Department Abstract: A team of four natural resource management experts convened in October, 2014 to conduct a fire regime assessment in Eastern Province of Zambia. The team included a multinational assemblage from the United States and Zambia tasked with the following objectives: 1) gain preliminary insights into the current status of fire management and fire use in Eastern Province, Zambia, 2) assess the role of fire in the regeneration and maintenance of the dominant vegetation present in the terrestrial ecosystems, and 3) provide recommendations for fire management training and equipment needs for Zambia Forestry Department and partner agencies in Eastern Province. The team traveled 2,500 km from Lusaka, through Eastern Province, and then returned to Lusaka. In Eastern Province the team met with community leaders and members in remote villages to understand how the villagers use fire and the unique threats to their villages from unmanaged wildfires. The team also met with personnel from the Zambia Forestry Department, Zambia Wildlife Authority, and other critical agencies active in Eastern Province, such as Community Markets for Conservation and the Center for International Forestry Research. In Lusaka, the team met with officials from pertinent ministries as well as additional stakeholders to understand the nature of fire management from the governmental and political perspective and to learn about relevant fire management policies and projects in other provinces within Zambia. Eight recommendations have been identified as a result of this assessment. Instead of proposing that fire be excluded from Eastern Province, recommendations focus on potential changes or reviews within all levels of government, as well as essential formal training for personnel working in agencies that have 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

been legislatively assigned fire management duties. In addition, we recommend enhanced outreach within communities to decrease potential negative effects of fire to communities, wildlife, soils, and air quality while allowing for fire to be used in a sensible and judicious manner. Keywords: Africa, fire management, fire ecology, community outreach Presenter Bio: LaWen Hollingsworth is a Fire Behavior Specialist for the Fire Modeling Institute at the Rocky Mountain Research Station in Missoula, Montana. Professional interests include observing and analyzing fire behavior and fire effects, data preparation and calibration, and designing projects in multiple vegetation types. Most of her federal career has been spent in the field working as a Fire Ecologist, firefighter, and conducting long-term vegetation monitoring in upland and riparian systems. She has a M.S. from the University of Montana in Forestry/Fire Ecology and a B.S. from the University of Idaho in Forest Resources.

136. Living with Fire – Lessons Learned from Central Africa Grass Savannas and how it relates to Fire Management in the United States Jim Menakis, Fire Ecologist, USDA/Forest Service/Fire and Aviation Management Abstract: The Bateke Plateau is about 6 million hectares that spans across the countries Gabon, Republic of Congo and Democratic Republic of Congo with most of the Plateau in the Republic of Congo. Fire has long been part of the Bateke Plateau savanna ecosystems. The savannas on the plateau are classic fire adapted ecosystems and are rich with fire tolerant species both in numbers and diversity. Today almost all of the fires are caused by humans for several objectives including: hunting, gathering, safety from fire and wildlife, travel, and many others. Forests represent approximately 25% of the area, and can mostly be found along rivers, pockets in the upper plateau, and fingers stemming from the Massif forest along the western edge. Recent concerns over decreasing forest area on the plateau due to “uncontrolled” savanna fires and the conversion to agricultural by local population has resulted in the need to better understand the fire ecology of Bateke Plateau in order to develop land management plans for maintaining the forest system. Also this area has been identified as Reducing Emissions from Deforestation and Forest Degradation (REDD+) project with the goal of increasing potential carbon storage by increasing the gallery forest in the area. This presentation will briefly describe the fire ecology and local use of fire on the Bateke Plateau and then make direct comparisons to fire management in the United States and the goals outlined in the Cohesive Strategy. This presentation will explore the following questions: What are the lessons learned from the Bateke Plateau? How to do they apply to the Cohesive Strategy goals? And, what does it mean to live-with- fire? Keywords: fire management, fire ecology Presenter Bio: Jim Menakis has been the National Fire Ecologist for Forest Service Washington Office Fire and Aviation Management since 2010. Prior to 2010, Jim worked at RMRS Missoula Fire Science Laboratory for 20 years on various: research projects relating to fuels and fire ecology, national mapping projects, and applying the best available science to support management needs. Today Jim is working on evaluating the effectiveness of fuel treatment when tested by wildfire, supporting multi-scale wildfire risk assessments, and supporting the FS fire ecology program. Jim Menakis received his B.S. in Forestry and M.S. in Environmental Studies from the University of Montana, Missoula.

137. Introduction to STARFire: wildland fire spatial planning and budgeting Douglas Rideout, Professor, Colorado State University Additional Authors: Andy Kirsch, Wildland fire Analyst, National Park Service Niki Kernohan, Research Associate, Colorado State University Abstract: STARFire is a scalable spatial fire planning and budgeting system developed in cooperation with Colorado State University, the Bureau of Land Management, the National Park System, and the U.S. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

Fish and Wildlife Service. It operates at the local, state or national levels. It combines the full spectrum of values (life, property, protection and nature’s services) with fire behavior information to evaluate the risks of wildland fire with the benefits of restoring and maintaining resilient landscapes. STARFire assesses the current condition of the landscape to generate a ‘landscape value’unit. Planning alternatives are evaluated through STARFire’s four assessment modules to that quantify their impacts on the marginal value of the landscape. These modules include i) a risk assessment that estimates the risk and benefit of wildfire at every location on the landscape; ii) the identification of fuel treatment locations that maximize value added and return on investment; iii) how dispatch locations and preparedness strategies (loss mitigation and wildfire benefit) respond to different budgets and fuel treatment programs and iv) an estimate of the impact of smoke emissions from any location on the landscape. A comparison of pre-treated and post-treatment value quantifies performance, demonstrate return on investment and validate the impact of budget changes. STARFire’s outputs including spatial data, graphs and charts that can be used to support spatial fire management plans. The resulting ‘landscape value’ unit generated for a planning area is directly comparable to another planning area’s ‘landscape value’ unit. Consequently, STARFire provides the ability to generate strategic surfaces to inform program tradeoffs and national budget allocations at the state and national levels. These concepts will be demonstrated using the results of the STARFire analysis conducted at several parks within the U.S. National Park System. Keywords: Planning budgeting landscape economics ROI Presenter Bio: Doug Rideout is Professor of Forest and Wildland Fire in the Department of Forest and Rangeland Stewardship at Colorado State University. He received is Ph.D. at the University of Washington in Seattle where he developed an interest in wild land fire economics. As the co-director of the WESTFIRE Research Center at Colorado State University he had worked closely with the U.S. Department of Interior in wild land fire planning and budgeting.

138. A Survey of Fire Managers: Characterization of Resource Importance, Scarcity, and Substitutability by Resource Type Crystal S. Stonesifer, Ecologist, USDA Forest Service/ Rocky Mountain Research Station Additional Authors: David E. Calkin, Research Forester, USDA Forest Service, Rocky Mountain Research Station Michael S. Hand, USDA Forest Service, Rocky Mountain Research Station Abstract: Each year, United States fire managers draw from a pool of federal and contract suppression resources to meet fire suppression objectives. This relatively static resource pool, in terms of quantity of specific resource types, must respond to national suppression demand presented by highly variable fire seasons with respect to both the frequency of fire occurrence and the spatial and temporal coincidence of new and ongoing fire activity. We surveyed federal incident managers and operations personnel to characterize ordering patterns and perceptions related to the relative importance of suppression resources by type, resource scarcity, and substitutability of different resources. Currently, requests for nationally-available resources (e.g., crews, engines, and aircraft) are placed by incident-level fire managers through an interagency dispatch system, which allows resources to be sent where they are requested, if available. Generally, a request is categorized as “unable-to-fill” (UTF) if it cannot be met as specified, and UTF numbers are frequently used in federal Agency reports as the metric for unmet resource demand. This survey provides a novel exploration of the relationship between the likelihood of having a request returned UTF and the consequences of this to fire suppression strategies. Survey questions were designed to elicit responses related to UTF requests to characterize the ordering practices in the current system and to depict supply and demand levels that may not be fully quantified by a list of available resources and a corresponding number of UTFs. Additionally, as a proxy for resource scarcity, we explore how survey responses change at varying national and regional Preparedness Levels. 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

The interplay between resource importance, scarcity, and substitutability is poorly understood, and the implications of this relationship to national suppression capacity are significant. We present survey results related to these three themes and discuss how these results may further affect tradeoff analyses, operational efficiency, and risk management practices. Keywords: resource scarcity, supply and demand, unable-to-fill Presenter Bio: Crystal Stonesifer is an Ecologist with the US Forest Service, Rocky Mountain Research Station, Human Dimensions Program in Missoula, MT. Her research in recent years has focused primarily on resource use, risk, and efficiency in fire management. Specific areas of interest and expertise include federal fire suppression-related data systems, aviation use in fire suppression, particularly large airtankers, resource ordering practices and data, and incident-specific production calculations and risk indices.

139. Water Quality Above All Else: Fire Management in the Greater Victoria (British Columbia) Water Supply Area Robert Walker, Manager, Wildfiree, Security and Emergency Response, Capital Regional District Abstract: The Greater Victoria Water Supply Area (GVWSA) supplies drinking water to urban and rural areas within the Capital Regional District on Vancouver Island, BC including 13 municipalities and approximately 350,000 residents. The GVWSA has a history of large wildfires in the past and wildfire is considered the most serious risk to drinking water quality in the watershed. A new fire management plan recommends a landscape level approach to fire management in the GVWSA with elements of risk assessment, preparedness, research, fuel management and fire restoration. This talk will describe the approach being taken, the opportunities and the challenges inherent in managing fire to protect drinking water quality. Keywords: Fire Management, water quality, drinking water, planning Presenter Bio: Rob worked in wildfire management in national parks for 21 years before moving to the Capital Regional District in 2013. He has extensive experience in fire suppression and in the use of fire to achieve ecological objectives. Rob was a member of a National Incident Management Team for 20 years and is an ICS instructor. Rob has a BSc in Ecology from the University of Calgary and has been actively involved in research including reconstructing paleoecological fire and disturbance regimes, quantifying prescribed fire effects, whitebark pine conservation and wildfire risk.

140. NASA Fire Science and Applications: Technology, Satellites, Airborne Data and Models Amber Soja, Associate Program Manager, NIA / NASA Additional Authors: Vince Ambrosia, NASA AMES, SCUMB Dr. Lawrence Friedl, NASA HQ, Washington DC Abstract: NASA supports fire research and the application of fire data, models and technology in many cross-cutting Earth Science programs to include Terrestrial Ecology, Carbon Cycle and Ecosystems, Climate Variability and Change, Atmospheric Composition, Interdisciplinary Science and Applied Science. In this presentation we will discuss NASA Missions that have data that could support fire research, land management, fire recovery and active firefighting. We will also provide several examples of the successful use of NASA satellite and model data in fire science research and the application of those data. NASA has an Applied Science program element, Wildland Fires, that specifically targets the use of NASA data in ‘customer organizations’ or communities that manage related fire science, which includes regional, national or global active firefighting, rules and regulatory communities, air quality, ecosystem protection, disaster and recovery organizations, fire weather, fuels and other modeling. Currently there are nine active projects that support pre-, active-, and post-fire applications within customer 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

organizations. For instance, the US Forest Service actively uses MODIS and VIIRS active fire detection data to locate and track fire during the fire season. Another application of NASA data is preparing post- fire recovery tools to aid ecosystem rehabilitation and manage flood/erosion control for the Burned Area Emergency Response (BAER) teams. Additionally, an international team in making active fire detection and fire risk data available to local communities in remote regions in Peru, Brazil, Madagascar, and Indonesia in an effort to protect critical environments. Keywords: NASA satellite data, Applied Science, Wildland Fire Management Presenter Bio: Amber’s research interests focus on connections between fire regimes, the atmosphere and biosphere, and feedbacks to and from the climate system. She has two decades of research experience, where she has taken part in and led numerous national and international teams of research scientists. Specifically, she uses Geographic Information Systems (GIS) and satellite-derived data as tools to explore these dynamic relationships. Dr. Soja is currently an Associate Research Fellow at the National Institute of Aerospace, resident in Climate Sciences at NASA LaRC. She has recently taken a part-time Associate Program Manager position in the NASA Applied Sciences Program, Wildland Fire.

141. Moisture Exchange Models for Standing Dead Grass in Alaska Eric Miller, Fire Ecologist, BLM Alaska Fire Service Abstract: Several models to predict the moisture content of standing dead grass were developed from a set of 295 measurements of weather and environmental variables collected during prescribed burning at 17 sites over 80 days and six years in Alaska. Purely empirical-based linear regression models based on air temperature and relative humidity or dew point depression fit well. However several non-linear models with a basis in thermodynamics better represent the physics of sorption and are more conservative at very low humidity. Sigmoid-shaped function forms for equilibrium moisture content fit the data well because the response time of moisture content in dead grass to changes in weather is very short. Existing equilibrium moisture content function forms prevalent in the science of wildland fire behavior are used but others from the industries of agriculture and food science are introduced and found to be superior. The best models predict moisture content from temperature and relative humidity and yield root mean squared errors of approximately 1.7\% moisture content. Keywords: Moisture exchange, fine dead fuel moisture, equilibrium moisture content, moisture sorption, log drying rate, standing dead grass Presenter Bio: Eric Miller began his fire career in Yellowstone National park with the Fire Effects Monitoring Program. He has been Fire Ecologist with the BLM Alaska Fire Service in Fairbanks, Alaska since 2008.

142. Examination of pyrophytic plant combustion and the relationship between fuel moisture, energy released, and emissions Evan Ellicott, Assistant Research Professor, University of Maryland Additional Authors: Dr. Michael J. Gollner, Assistant Professor University of Maryland Colin Miller, Graduate Student, University of Maryland Robert Kremens, Research Professor, Rochester Institute of Technology Matthew Dickinson, Research Ecologist, U.S. Forest Service - Northern Research Station Abstract: In this paper we present preliminary findings from our experimental fires to examine the influence of fuel moisture and volatile compounds in biomass burning. Our research seeks to better understand the effects of these two factors at it relates to instantaneous release of fire radiative energy (FRE), or fire radiative power (FRP), and emission of carbon oxides (CO and CO2). Fire experiments were conducted under the University of Maryland, Department of Fire Protection Engineering’s laboratory hood, providing a climate-controlled space where exhaust gasses could also be 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

collected and analyzed. Fuel moisture content (FMC) was controlled by oven drying fuels to a stable, consistent proportion and then allowing them to equilibrate outside the oven to various water contents. A brief list of fuels examined includes Pinus monticola (western white pine), Kalmia species (sheep and mountain laurel), Eucalyptus (Eucalyptus spp.), and Adenostoma fasciculatum (chamise) and Myrica cerifera (wax myrtle), as well maple and oak species. Fire intensity was measured using a FLIR (T640) focal-plane array uncooled microbolometer with a resolution of 640x480 and a nominal spectral bandpass of 7.5–13 µm. We also used a new generation of thermopile radiometers developed at the Rochester Institute of Technology to estimate FRP per unit area (Wm-2) at 1s intervals using dual-band thermometry. The new radiometers measure fire-leaving radiative flux density using 5 infrared channels: 3.95 µm, 10.68 µm, 0.8 – 50 µm, 0.1 – 5 µm, and 8-14 µm. These were digitized at 24 bit resolution at a 1Hz rate. The system was radiometrically calibrated using standard blackbody radiation sources between 373K and 1250K. Exhaust gasses were analyzed in real-time to give heat-release rates (HRR) of the fire using oxygen-consumption calorimetry, an indication of completeness of combustion, and an assessment of soot and other emitted particulates. Combined with fuel mass loss measurements, an assessment of total convective HRR and radiative losses were compared with measurements from overhead radiometers. Relationships were developed between the HRR, total energy, rates of fuel consumption, and total emissions. Keywords: FRE, fire energy, fuel moisture, volatiles Presenter Bio: My research focus is on the examination of geospatial data to describe and characterize biophysical phenomena and the interactions with society. One of my topics of study is the examination of biomass burning, in particular wildland fires, in a changing physical and social environment. I have developed methods to estimate biomass burning emissions and fuel consumption using satellite and field-based measures of fire energetics to provide synoptic level detail of how fire emissions are changing in time and space. My work also includes characterizing fire intensity and post-fire severity for applications as varied as habitat selection and reproduction of the Mexican Spotted owl (Strix occidentalis lucida) to fuel treatment mitigation efficacy strategies and policy. I am a Joint Polar Satellite System (NOAA and NASA) Suomi NPP science team member tasked with VIIRS Active Fire product calibration and evaluation. I am also the PI for the Proving Ground and Risk Reduction (PGRR) project at UMD which includes research and outreach components to develop and improve operational deployment of VIIRS fire products for hazard management end-users (e.g. National Weather Service, US Forest Service).

143. Climatic and eco-hydrological drivers of fuel moisture dynamics in complex terrain Petter Nyman, Dr, School of Ecosystem and Forest Sciences, The University of Melbourne Additional Authors: Craig Baillie, Technician, School of Ecosystem and Forest Sciences, The University of Melbourne Thomas Duff, Research Fellow, School of Ecosystem and Forest Sciences, The University of Melbourne Kevin Tolhurst, A/Professor, School of Ecosystem and Forest Sciences, The University of Melbourne Gary Sheridan, Senior Research Fellow, School of Ecosystem and Forest Sciences, The University of Melbourne Abstract: Fuel moisture is a critical parameter for predicting fire behaviour and for planning prescribed burning operations. Moisture content in fuels located on or near the forest floor is particularly important because this fuel source can 1) comprise a large component of the overall fuel load, 2) have strong impact on fire spread, and 3) in many cases be effectively managed with prescribed burning. Being able to predict surface fuel moisture content is therefore an important research topic. Moisture dynamics in surface fuel is a function of microclimate above the litter layer, rainfall, interception, soil moisture and the hydraulic properties of the fuel itself. Process-based fuel moisture models include 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

these factors in their predictions. However, the data needed to parametrise and test such models at a landscapes scale are often lacking. The relative importance of various components of the water balance in the litter layer is therefore unknown. In this research we seek to quantify how climate, vegetation and eco-hydrological feedback contribute to variation in net radiation and potential evaporation at the forest floor. Research sites were established at 16 locations in eucalypt forests in south-east Australia with variable elevation, solar exposure, and drainage areas. Forests ranged from open woodland to tall temperate forests. At these sites we measured solar radiation, air temperature, relative humidity, throughfall, litter moisture, soil moisture, and litter temperature. Forest structure was characterised with hemispherical photos. With these data on microclimate and vegetation structure we develop and parametrise a Penman-Monteith model of potential evaporation on the forest floor at daily timescales. Using this model of potential evaporation alongside landscape-scale information on the aridity and net radiation we quantify the effects of topography, long-term climate and eco-hydrological feedbacks on the energy and water balance at the air-litter interface. The implications of these effects for fuel moisture dynamics are evaluated with fuel moisture time series that were collected simultaneously at the 16 sites. Results from the study are used in the development of a general framework for incorporating vegetation and topography into landscape-scale fuel moisture models. Keywords: Fuel moisture, radiation, litter, net radiation, complex terrain Presenter Bio: Petter is a research fellow at the School of Ecosystem and Forest Sciences, University of Melbourne, working on applied and theoretical research questions regarding water and sediment fluxes in forests. His research spans a range of disciplines from soil physics and geomorphology to forest meteorology and wildland fire. In his current project Petter aims to develop models that can be used to represent climatic and topographic drivers of fuel moisture dynamics in complex terrain.

144. Flammability of Live Vegetation: Combustibility and Ignitability Assessment Jan Christian Thomas, PhD Student, University of Edinburgh Additional Authors: Michael Gallagher, Research Technician and PhD student, USDA Forest Service, Northern Research Station, Morgantown, WV, USA Kenneth Clark, Research Forester, USDA Forest Service, Northern Research Station, Morgantown, WV, USA Rory Hadden, Lecturer, University of Edinburgh, BRE Center for Fire Safety Engineering, Edinburgh, Scotland, UK Nicholas Skowronski, Research Forester, USDA Forest Service, Northern Research Station, Morgantown, WV, USA Albert Simeoni, Senior Manager, Exponent Engineering and Scientific Consulting, Natick, MA, USA Abstract: The occurrence of wildfire is largely known to be part of the natural lifecycle of selected ecosystems. These have adapted, yet even require fire to prosperously flourish. Fire spread and intensity of wildfires is influenced by a number of factors, related to fuel properties, geographical and climate conditions. Many of the parameters influencing the fire spread and intensity are dynamic. For example, in an environmental context, ambient wind and temperature conditions vary on short timescales. It is also well known that the fuel properties, such as fuel moisture content (FMC) and density, exhibit this dynamic behavior due to (1) the metabolism and growth cycle of the plant and (2) the climatic conditions in which the plant exists. The density and FMC are two specific parameters that are known to control the heat and mass transfer in the fuel matrix and consequently affect the ignitability and combustibility. Small-scale laboratory experiments have been designed to investigate the effect of the growth cycle and FMC on the ignition and burning dynamics of Pitch pine (Pinus rigida) needles. The experiments allow good control of the other variables which influence the heat and mass transfer (e.g. wind, fuel loading, 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

and heating conditions). The work presented in this paper illustrates a framework in which not only ignitability, but also combustibility is assessed in a controlled manner. Experiments are conducted using the Fire Propagation Apparatus, which has been proven to be a useful tool in the flammability characterization of vegetative fuel. The FMC and density of Pitch pine needles are evaluated over the course of one year, and related to flammability parameters obtained from combustion tests. Results are presented to quantify the ignitability and combustibility. Findings indicate that the season where the ignition potential is highest (early summer) does not coincide with the season where combustibility is at its peak level (late summer early fall). These insights can be used to better understand the occurrence of fire seasons from a fuels perspective and also for planning and executing preventative wildfire measures, such as prescribed burning. Keywords: time to ignition, heat release rate, oxygen consumption calorimetry, live foliage, FMC, plant chemistry Presenter Bio: Jan Christian Thomas is currently pursuing his Ph.D. in Fire Safety Engineering, under the supervision of Prof. Albert Simeoni at the University of Edinburgh, UK. His research is in forest fires, with a thesis topic on flammability and burning dynamics of forest fuel. Besides his thesis work, Christian is involved in large scale forest fire experiments, investigating the effectiveness of prescribed burning, fire behavior, and the generation of short range firebrands. The latter topic is of special interest to Christian, where information from large scale testing is used in small scale laboratory application to investigate the ignition potential of combustible structural elements in the Wildland-Urban Interface by firebrands.

145. Critical Examination of the Haines Index and its Use Brian Potter, Research Meteorologist, Pacific Wildland Fire Sciences Lab, USDA Forest Service Abstract: The Haines Index is used to assess large fire growth potential that could be caused by near- ground atmospheric instability and dryness. It was developed in 1988, and is now a standard piece of information contained in fire weather forecasts. Is it based on sound science? If so, is it used in a manner consistent with that science? If it is not based on sound science, can or should it be used? This presentation will examine these questions and more related to the Haines Index and its use. Keywords: Haines Index; instability; fire weather Presenter Bio: Brian Potter studies fire-atmosphere interactions from the perspective of atmospheric dynamics. His research interests span spatial scales from the fire front (a few meters) to synoptic weather patterns affecting fire behavior and fire danger (hundreds of kilometers.)

146. Developing new references for fine dead fuel moisture in the Southeastern United States W. Matt Jolly, Research Ecologist, USFS, RMRS, Fire Sciences Laboratory Additional Authors: Jim Brenner, Fire Management Administrator, Florida Forest Service Fred Turck, Associate Director of Resource Protection, Virginia Department of Forestry Daniel Chan, Meteorologist, Georgia Forestry Commission Dennis Dauterive, Fire Chief, Mississippi Forestry Commission Mark Goeller, Assistant Director, Oklahoma Forestry Services Balsie Butler, Fire Operations Chief, Alabama Forestry Commission Luke Saunier, Chief, Kentucky Division of Forestry Abstract: Today’s commonly-used fuel moisture field guides were developed decades ago without both the aide of recent technology and without extensive field verification. As such, these guides often poorly predict fine dead fuel moisture in many areas, particularly those of the humid Southeastern states. Here we present the development of a set of fuel moisture field guides that are tailored for use in these areas. We combine a field sample collection with laboratory analyses to generate fuel moisture over a range of environmental conditions and we test and modify a physically-based fuel moisture model to 5th International Fire Behavior and Fuels Conference  Portland, Oregon  April 11-14, 2016 Oral Presentation Abstracts

best reflect the influence of environmental conditions on fuel moisture dynamics. We will then use this calibrated fuel moisture model to generate field references similar to those found in the Incident Response Pocket Guide (IRPG). These tables will allow the proper calculation of fine dead fuel moisture as a function of temperature, humidity, solar radiation and rainfall. Additionally, we discuss the potential development of a prototype mobile application that will allow the rapid distribution of these tools to firefighters and fire managers throughout the Southeastern US. Keywords: Fuel Moisture, Fire Behavior

Presenter Bio: Matt Jolly is a Research Ecologist in the Fire, Fuel and Smoke Science Program of the US Forest Service, Fire Sciences Laboratory in Missoula, MT. He received a BA in Environmental Science from the University of Virginia and a PhD in Forestry from the University of Montana. His main research interest is to improve our understanding of the roles that live and dead fuels play in wildland fires and to use this improved understanding to develop or improve predictive tools that can help support fire management decisions.