Fire Management today Volume 64 • No. 2 • Spring 2004
TTHEHE TTENEN AND AND EEIGHTEENIGHTEEN—— AA MMEMORYEMORY AAIDID
United States Department of Agriculture Forest Service Fire Management Today is published by the Forest Service of the U.S. Department of Agriculture, Washington, DC. The Secretary of Agriculture has determined that the publication of this periodical is necessary in the transaction of the public business required by law of this Department.
Fire Management Today is for sale by the Superintendent of Documents, U.S. Government Printing Office, at: Internet: bookstore.gpo.gov Phone: 202-512-1800 Fax: 202-512-2250 Mail: Stop SSOP, Washington, DC 20402-0001
Fire Management Today is available on the World Wide Web at
Ann M. Veneman, Secretary April J. Baily U.S. Department of Agriculture General Manager
Dale Bosworth, Chief Robert H. “Hutch” Brown, Ph.D. Forest Service Managing Editor
Jerry Williams, Director Madelyn Dillon Fire and Aviation Management Editor
The U.S. Department of Agriculture (USDA) prohibits discrimination in all its programs and activities on the basis of race, color, national origin, sex, religion, age, disability, political beliefs, sexual orientation, or marital or family status. (Not all prohibited bases apply to all programs.) Persons with disabilities who require alternative means for communication of program information (Braille, large print, audiotape, etc.) should contact USDA’s TARGET Center at (202) 720-2600 (voice and TDD).
To file a complaint of discrimination, write USDA, Director, Office of Civil Rights, Room 326-W, Whitten Building, 1400 Independence Avenue, SW, Washington, DC 20250-9410 or call (202) 720-5964 (voice and TDD). USDA is an equal opportunity provider and employer.
Disclaimer: The use of trade, firm, or corporation names in this publication is for the information and convenience of the reader. Such use does not constitute an official endorsement of any product or service by the U.S. Department of Agriculture. Individual authors are responsible for the technical accuracy of the material presented in Fire Management Today. Fire Management today Volume 64 • No. 2 • Spring 2004
On the Cover: CONTENTS Visualizing the Ten and Eighteen—With Humor ...... 4 Kathy Murphy Managing Fire-Dependent Ecosystems: We Need a Public Lands Policy Debate ...... 6 Jerry Williams New Interagency Fire Program Analysis System ...... 12 S.L. Robertson and H. Roose A New Tool for Fire Managers—An Electronic Duff Moisture Meter...... 15 Peter R. Robichaud and Jim Bilskie One of a series of images by artist and firefighter Kari Fuel Mapping for the Future...... 19 Cashen to help fellow fire C.W. Woodall, G.R. Holden, and J.S. Vissage fighters remember the 10 Diversifying Fuels Management To Offset Uncertainty . . . . . 22 Standard Fire Orders and 18 John Hof Watch Out Situations. See the story by Kathy Murphy begin Probability of Spot Fires During Prescribed Burns ...... 24 ning on page 4. John R. Weir Fires in the Wildland/Urban Interface: Best Command Practices ...... 27 Michael S. Rohde Fire Regimes in the Northern Rockies ...... 32 Stephen W. Barrett Keetch–Byram Drought Index: Can It Help Predict Wildland Fires? ...... 39 Daniel W. Chan, James T. Paul, and Alan Dozier The FIRE 21 symbol (shown below and on the cover) stands for the safe and effective use of Fire Prevention Team Shows Its Worth in Georgia ...... 43 wildland fire, now and throughout the 21st cen tury. Its shape represents the fire triangle (oxy James T. Paul, Daniel Chan, and Alan Dozier gen, heat, and fuel). The three outer red triangles represent the basic functions of wildland fire Building Group Cohesion in Type 2 Fire Crews ...... 48 organizations (planning, operations, and aviation management), and the three critical aspects of Bill Lee wildland fire management (prevention, suppres sion, and prescription). The black interior repre Wildland Fire Education: Going the Distance in Alaska . . . . . 51 sents land affected by fire; the emerging green points symbolize the growth, restoration, and Sandi Sturm and Matt Weaver sustainability associated with fire-adapted ecosystems. The flame represents fire itself as an ever-present force in nature. For more informa tion on FIRE 21 and the science, research, and SHORT FEATURES innovative thinking behind it, contact Mike Apicello, National Interagency Fire Center, 208-387-5460. Websites on Fire...... 38 Canadian Fire Weather Index System: Training Now Available on CD-ROM ...... 54 Paul St. John and Martin E. Alexander “Brewer Fire Mystery” Discussion ...... 56 Guidelines for Contributors...... 58 Firefighter and public safety is our first priority. Annual Photo Contest ...... 59
Volume 64 • No. 2 • Spring 2004 3 VISUALIZING THE TEN AND EIGHTEEN— WITH HUMOR
Kathy Murphy
or 2 years, Kari Cashen, a fire fighter with the Truckee hand F crew on the Tahoe National Forest, spent the summer digging line, setting backfires, sleeping in the dirt, and enjoying the cama raderie of a fire crew. Recently, Cashen combined her passion for art and firefighting to create color ful, humorous posters to help her and others memorize the 10 Standard Fire Orders and the 18 Watch Out Situations—required for all firefighters.
For Cashen, every 10-mile run or uphill training hike was an oppor tunity to draw mental pictures. Throughout the gauntlet of exten Kari Cashen posing with her artistic rendering of the 10 Standard Fire Orders. sive fire and physical-fitness train Photo: USDA Forest Service. ing, Cashen focused on a Fire Order or Watch Out and crafted a related scenario based on her crew’s fire fighting effort. “I would draw the scene in my mind. Before I knew it, I had climbed the hill or run the 10 miles, and I would be giggling to myself about the cartoon I had envisioned.”
Although the mental art was help ful and fun, seeing the images on posterboard really helped Cashen focus on the essential firefighting safety messages. “The colors are bright and eye-catching and the cartoons appeal to the sense of fun that many firefighters have,” says Cashen. She shared the artwork with crew members, who especially liked the cartoon characters that
Kathy Murphy is the fuels management specialist for the USDA Forest Service, Tahoe National Forest, Truckee Ranger District, Truckee, CA.
Fire Management Today 4 These sketches just might help you remember the Fire Orders and Watch Outs!
showed them and the fires that they had battled! They encouraged Cashen to perfect her work and develop the sketches into a poster format for the benefit of other fire fighters.
Cashen’s firefighting experience and engaging artwork visually com municate vital fire safety messages. The posters are fun and colorful, and they clearly depict the conse quences of violating the 10 Standard Fire Orders and the 18 Watch Out Situations. Cashen is working to widely circulate her art work to help wildland firefighters remember the essential safety rules.
Four examples of Cashen’s artwork are shown here, including two Fire Orders and two Watch Outs. The full series is available as a screen saver, in a PowerPoint presentation, and for display at
Volume 64 • No. 2 • Spring 2004 5 MANAGING FIRE-DEPENDENT ECOSYSTEMS: WE NEED A PUBLIC LANDS POLICY DEBATE*
Jerry Williams
n wildland fire management today, we know that sustaining It is not so much that our suppression policy was I healthy, resilient fire-dependent flawed as it is that our fire use policy is too ecosystems is the key to protecting constricted. people and property. We have departed from the policy of fire exclusion that characterized our The risk is due to altered fire that uses fire as a bedrock. fire management for most of the regimes. Fire regimes are an Historically, the ponderosa pine 20th century. There will always be a expression of fire’s role in terms of canopies were very open, with trees need to fight fire, but the wholesale historical or natural fire frequency that were very big and widely exclusion of fire was a major factor and burning intensity. Fire man spaced. Low-severity fires burned in putting our fire-dependent agers expect large, stand-replace through on the ground every few ecosystems at risk, particularly our ment fires in our long-interval fire years without doing much damage long-needle pine forests, such as regimes. Ecologically, that is how to the big trees. But fire exclusion ponderosa pine. It is not so much these forests established. Alarm and other factors allowed small that our suppression policy was ingly, however, we are beginning to trees and brush to build up in the flawed as it is that our fire use poli see landscape-scale, stand-replace understory. Today, where we once cy is too constricted. ment wildfires in our short-interval had a hundred large trees per acre, fire regimes, such as ponderosa we might have thousands of small Things Coming Due pine. trees that “choke” the overstory. In a way, things are “coming due” for wildland fire operations in the Sustaining these forests will In a drought, we now have continu United States. Things are coming require a management approach ous fuels from the ground into the due for our workforce—we rely on retirees during difficult fire sea sons. Things are also coming due for some of our equipment, such as our air tankers—our average air tanker is 46 years old. And things are coming due for our forests— nationwide, we have about 397 mil lion acres (161 million ha) at risk from wildland fires that compro mise human safety and ecosystem health.**
Jerry Williams is the Director of Fire and Aviation Management, USDA Forest Service, Washington Office, Washington, DC. Winema Interagency Hotshot Crew member walking towards a drip torch in a May under- * Based on a paper presented at the 3rd International burn in open ponderosa pine, Crescent Ranger District, Deschutes National Forest, OR. Wildland Fire Conference in October 2003 in Sidney, Long-needle forest types such as ponderosa pine need frequent fire use to stay healthy. Australia. ** The area in fire regimes I and II, condition classes 2 Photo: Brendan O’Reilly, USDA Forest Service, Klamath National Forest, Klamath Falls, and 3 (Schmidt and others 2002). OR, 1999.
Fire Management Today 6 canopy. When we get a fire, it climbs into the canopy and becomes severe and stand-replac ing. In 2002, four States in the West had their biggest fires ever, and a fifth State came close, partly because the fire regime has changed in our long-needle pine forests.
The USDA Forest Service recently mapped fire regime condition class es in relation to wildfire activity in the United States (Schmidt and others 2002). In many of our ecosystems, fire regimes have sig nificantly changed from their his View of the Hayman Fire on June 18, 2002. The largest fire in Colorado history, Hayman torical range. The 397 million acres burned through many areas of overgrown ponderosa pine forest in or near the (161 million ha) most at risk wildland/urban interface. Photo: Cindy Nowak, USDA Forest Service, Klamath National Forest, Klamath Falls, OR, 2002. nationwide constitute an area almost three times the size of France. In the West, nearly all of Fire protection in the WUI is not just about the area most at risk is ponderosa pine in the prolonged absence of protecting houses—it’s about protecting quality periodic underburning. of life.
From a social perspective, pon derosa pine forests are most com fire management is evolving. Why? Because people are moving to mon at lower elevations, where Although the physical sciences will places they value for a better quali most people live, work, and play. remain essential for understanding ty of life. People value forested set That makes them of particular con ecosystems and fire behavior, we tings. They value places with water, cern because of the huge fire dan will need a deeper understanding of mountains, and amenities, such as ger they represent. It is no coinci the social sciences to help us widen hunting or hiking on public land. dence that many of our most costly, the decision space we will need for People are moving to the West or damaging, and destructive wildfires ensuring the health, resilience, pro South to find these places. These occur in these changed ponderosa ductivity, and safety of fire-depend are also the regions dominated by pine forests, often in close proximi ent ecosystems. long-needle pine ecosystems with ty to the wildland/urban interface altered fire regimes. (WUI). Stand structure is much The reason is that altered fire more dense, with small trees and regimes in our long-needle pine The result is often a dangerous mix. undergrowth choking the forest. forests are increasing the fire dan People are moving in record num Species composition has often ger to communities. In the 2000 bers into forests that are increas shifted to Douglas-fir and other census, the five fastest growing ingly susceptible to crown fire. The fire-intolerant species. And people States were all in the Western very qualities that people value— have moved into the forest. United States. By 2020, our 20 dense forests that provide a sense of fastest growing counties are all seclusion and screening from Need for Social expected to be in the South and neighbors—these same qualities Science West (Cordell and Overdevest put people at risk. The risks are 2001). Our population is gradually enormous, and they go way beyond That brings me to a second thing shifting from the Northeast and individual homes. If their houses that we are learning to recognize: Upper Midwest to the South and are saved but the surrounding land The kind of science we will need in West. scape is blackened, then as far as
Volume 64 • No. 2 • Spring 2004 7 they’re concerned, people in the We might argue that the extended-attack fire and WUI have lost the very values that the megafire are our two most important kinds of brought them there. fire—one in terms of safety, the other in terms of Fire protection in the WUI is there cost. fore not just about protecting hous es—it’s about protecting quality of life. The wildland fire community is Although accumulated fuels and when fire behavior has become too expected to protect the entire land drought predispose many of our extreme for initial-attack tactics to scape—not only communities, but forests to wildfires, we are coming be safe and effective. also watersheds, viewsheds, recre to realize that there are four dis ational opportunities and other tinctly different kinds of fire. We Large fires and megafires are less amenities, and forest health— have good suppression strategies than 1 percent of all of our fires, everything people value in the WUI, for two of them. But there are two but they account for a dispropor everything they move there to find. other kinds of fire for which we do tionately high percentage of our not have good strategies, and it total suppression costs—about 80 We will therefore need a better shows in our statistics. percent—and of our total area understanding of the social sci burned—about 90 percent. The ences. If we are going to protect These four kinds of fire occur along megafire accounts for the majority quality of life in the WUI, then we a spectrum of size and complexity. of these costs and acres burned, have got to do more to understand They range from the small initial- even though these fires probably people’s motivations so we can bet attack fire to the enormous and only comprise one-tenth of 1 per ter influence social attitudes and complex “megafire.” We have sound cent of all fires. behaviors. We have to do a better approaches for dealing with the job of addressing public biases and small initial-attack fire and with We’ve learned that we can’t go toe fears in connection with fuels man the large fire. We train, organize, to-toe with these big fires under agement and fire use in our fire- and staff to address the unique extreme burning conditions. We’ve dependent ecosystems. We also characteristics of these two types of got to back off and take a defensive have to do a better job of address fire. But for the transition or posture. Megafires are qualitatively ing public preferences and lifestyles extended-attack fire and the so- different from large fires and need a in the WUI. For that, we will need called megafire, we do not do this qualitatively different type of man to take such fields as sociology, well. We tend to treat the extended- agement, just as extended-attack communications, community rela attack fire like we do the initial- fires need a qualitatively different tions, and public administration attack fire, only we fight it harder. type of management from initial more into account when we formu And we tend to treat the megafire attack. For both kinds of fire, we late policy for public lands. like the large fire, only—believing need to develop discrete strategies more is better—we fight it with in terms of policy, procedures, and Four Kinds of Fire more people, more equipment, and practices. more money. A third thing we are learning has to Many of us believe that the sup do with our suppression program We might argue that the extended- pression fight against large fires in the context of the fuels and fire attack fire and the megafire are our and megafires will ultimately be environment. Despite significant two most important kinds of fire— won or lost on the fuels front, advances in our firefighting tech one in terms of safety, the other in where we’re using fire and mechan nology, budgets, and personal pro terms of cost. Some 70 percent of ical fuels reduction tools to take a tective equipment, we are seeing an our fireline fatalities occur on tran little heat out of the woods. upward trend in the number of sition fires, such as South Canyon Basically, we need to fight fire acres burned per acre protected. in 1994 or Thirtymile in 2001. We where we must but use fire where Also, again in spite of all the get into trouble when we keep we can. We are getting megafires in advances we’ve made, the number using initial-attack tactics on a fire long-needle pine forests because of entrapments and fatalities we’re that requires a shift in thinking fire regimes there have been seeing remains a major concern. about potential fire behavior— altered. The long-term solution is
Fire Management Today 8 to restore these forests to some fire policymakers in the United policy provides for fire use, sup thing more resembling their histor States. pression, and prevention. But I am ical condition and then get the afraid it is not balanced enough. I’ll right kind of fire back into the Our objectives in wildland fire explain by giving a little history. ecosystem. management are clear. Our aim is to protect values—to protect quali In 1995, the five Federal agencies Perhaps one of our lessons in accel ty of life by restoring fire-depend with fire management responsibili erating fuels reduction work ent ecosystems such as long-needle ty in the United States wrote a col involves learning to mobilize for lective policy for fire management. fire use operations like we mobilize In 2001, we updated the Federal for fire suppression operations. Many of us believe that fire policy. As part of the implemen Although we’ve made progress the suppression fight tation process, we gave the revised toward a more balanced wildland fire policy to two outside panels for fire policy, we still have to work on against large fires and their review. overcoming the bias toward fire megafires will ultimately suppression that stems from a lega be won or lost on the One panel was made up of fire cy of fire exclusion. fuels front. experts. They were satisfied that our revised Federal fire policy Next Big Step reflected good science and sound The three things we are learning— pine. For that, we need to establish fire management. The other team the need for more fire use, for a a total, balanced program of fire was made up of policy experts. better understanding of the social management where there is no They, too, were generally satisfied sciences, and for discrete strategies longer any bias toward fire suppres that we had provided a coherent on the four kinds of fire—are all sion or fire use. fire policy. interconnected. In fact, our ability to make progress in one area Given these objectives, we have But one of these reviewers, from depends on understanding all three. probably pushed our fire manage the JFK School of Government at That brings me to what lies ahead: ment policy about as far as we Harvard University, said our fire the next big challenge for wildland might effectively go. Today, our policy was missing something: a much larger public land policy debate. We were setting ourselves up for failure, he said, without a broad public debate—a debate that addresses all the long-term social, legal, and economic factors that drive how we manage our fire- dependent ecosystems. These fac tors go way beyond our fire policy per se.
In other words, a sound fire policy must be predicated on a public lands policy that is not only socially acceptable, but also ecologically appropriate and economically effi cient over time. Our fire policy is somewhat “stuck” until we can do three things:
View of the Biscuit Fire on August 1, 2002, the largest fire in Oregon history. Large fires and megafires are less than 1 percent of all fires but account for 80 percent of total sup pression costs and 90 percent of total area burned. Photo: Gary Percy, USDA Forest Service, Siskiyou National Forest, Gold Beach, OR, 2002.
Volume 64 • No. 2 • Spring 2004 9 • More effectively influence devel opment or growth behaviors in the WUI; • Better align regulatory controls for clean air, clean water, and endangered species with the dis turbance processes that define our fire-dependent ecosystems; and • More specifically tailor resource objectives to be consistent with the ecological dynamics of fire- prone forests and grasslands.
Let me give a few examples to illus trate what I mean about the impor tance of a public lands policy debate for the viability of a bal anced wildland fire policy.
Technical solutions are not enough. We also need social, legal, and regulatory solutions that focus on the dynamics of fire-prone forests.
First, let’s consider the social influ ence on wildland fire policy. We know that we need to thin over crowded long-needle pine forests to A home in Clark County, WA, before (top) and after (bottom) removal of hazardous fuels reduce fire danger in the WUI. The through a State project funded by the National Fire Plan. Many people prefer secluded result would be a forest that is very homes in densely forested settings and resist vegetation removal—a serious constraint to open, with maybe only a hundred wildland fire management. Photo: Thomas Iraci, USDA Forest Service, Pacific Northwest Region, Portland, OR, 2002. trees per acre. But people move to the WUI partly because they value ordinances against tree cutting. protect habitat for threatened, the sense of seclusion and “natural Other people might see it as affect endangered, and sensitive species. ness” they get from lots of trees. ing their quality of life if we remove In the case of the northern spotted They are used to seeing thick most of the trees near where they owl and Mexican spotted owl, we do forests, with thousands of trees per live. In fact, our projects are often that partly by managing for late acre. It’s what they think of as nat appealed and even litigated for just seral stand conditions to maximize ural and healthy, even if it isn’t this reason. canopy cover. really natural, healthy, or resilient. Now let’s look at the regulatory side But managing for closed canopies So people often object to a thinning of wildland fire policy. Under the might keep us in some places from project. Some people might object Endangered Species Act, Federal restoring the more open forests in principle to cutting any trees at land managers are legally bound to that existed historically. The regula all—there are even counties with
Fire Management Today 10 tory context can actually put us at Although today we use fire more, we still have to cross-purposes. In fact, two of the work on overcoming the bias toward fire megafires we had last year—the Biscuit Fire in Oregon and the suppression that stems from a legacy of fire Rodeo–Chediski Fire in Arizona— exclusion. burned partly in areas we were managing for late-seral stand con ditions. Ironically, such fires not ger a prohibition under the Clean As wildland fire professionals, we only consume the old-growth forest Water Act. This is another exam need to prompt a larger public we are trying to protect, but also ple of a tradeoff between short- lands policy debate that deals with imperil the very species we are try term environmental impacts and values and tradeoffs if we hope to ing to sustain. long-term environmental bene redeem our fire protection man fits. date. And we need to do it in the We are in some serious quandaries. • When we try to get people to be context of the dynamics of fire- Social and regulatory factors can smarter about building houses dependent ecosystems. That is the freeze our ability to reduce fuels and maintaining their property in next big step in the evolution of and restore long-term forest health. the woods, they might see it as a wildland fire policy in the United Here are some more examples: States’ rights issue or as Federal States—and maybe in other coun meddling in private affairs. Local tries as well. • When we use fire, people some building codes often favor eco times object to the smoke. Under nomic expansion and develop References provisions of the Clean Air Act, ment, even though development Cordell, H.K.; Overdevest, C. 2001. Footprints on the land: An assessment of prescribed fire emissions count as in some cases puts people, busi nesses, and local communities at demographic trends and the future of air pollution, whereas wildfire natural lands in the United States. emissions do not—even though, risk in fire-prone forests. Champaign, IL: Sagamore Publishing. over time, wildfire emissions have Schmidt, K.M.; Menakis, J.P.; Hardy, C.C.; We think we have the ecological Hann, W.J.; Bunnell, D.L. 2002. actually increased due to our Development of coarse-scale spatial data attempts to exclude fire. People science to restore fire-dependent for wildland fire and fuel management. tend to focus on immediate ecosystems and better protect the Gen. Tech. Rep. RMRS–87; Missoula, MT: people we serve, and technically USDA Forest Service, Rocky Mountain impacts, not future benefits. Research Station, Fire Sciences • When we mechanically thin trees, maybe we do. But technical solu Laboratory. ■ the reduction in vegetative cover tions are not enough. We also need can temporarily impair local social, legal, and regulatory solu water quality, which might trig tions that focus on the dynamics of fire-prone forests.
Volume 64 • No. 2 • Spring 2004 11 NEW INTERAGENCY FIRE PROGRAM ANALYSIS SYSTEM*
S.L. Robertson and H. Roose
he 2000 and 2001 fire seasons shook the wildland fire com The new Fire Program Analysis System will T munity in the United States. In replace the budget and analysis systems currently May 2000, an escaped prescribed in use. fire roared through parts of Los Alamos, NM. That was followed by one of the largest fire seasons in 50 The new Fire Program Analysis FIREPRO, and Fire Base. The new years. The next summer brought (FPA) System will replace the budg system is scheduled for field testing the Thirtymile Fire, a tragedy fire et and analysis systems currently in in summer 2004, with release of in the State of Washington that use, such as the National Fire the first module (preparedness) entrapped 14 firefighters and took Management Analysis System, planned for October 1, 2004. 4 lives.**
These events raised calls for more accountability in the Nation’s wild- The System in a Nutshell land fire program. The five Federal Features the National Fire Plan and the agencies responsible for the pro The Fire Program Analysis (FPA) Government Performance and gram—the USDI Bureau of Indian System is a strategic tool for Results Act of 1993; and Affairs, Bureau of Land Manage helping managers move from • Facilitates partnerships ment, National Park Service, and measuring outputs to measuring between Federal interagency U.S. Fish and Wildlife Service and results and outcomes. The sys fire planning units and tribal, the USDA Forest Service—are tem: State, and local governments. working together to develop a mutually compatible, performance- • Uses objectives in land and Scope based system for analyzing the resource management plans When completed, the FPA System interagency fire program and budg and fire management plans as will address the full scope of fire et (see the sidebar). the cornerstone for planning; management activities, including: • Provides a common approach to fire management program • Initial response for both wild Sarah Robertson is the national intera planning and budgeting for all fire suppression and wildland gency fire planner for the USDA Forest Service and USDI National Park Service; five Federal wildland fire man fire use; and Howard Roose is the Fire Program agement agencies; • Extended attack, large fire sup Analysis Business Team leader for the USDI • At any given budget level, dis port, and national resources; Bureau of Land Management, National Interagency Fire Center, Boise, ID. plays tradeoffs involved in • Prescribed fire and fuels man meeting various fire manage agement; * Based on a paper presented at the 3rd International ment objectives, such as pro • Wildland fire prevention and Wildland Fire Conference in October 2003 in Sidney, Australia. tecting natural and cultural education; and ** For more on these events, see Jim Paxon, • Postfire emergency stabilization “‘Remember Los Alamos’: The Cerro Grande Fire,” resources as well as infrastruc Fire Management Today 60(4) [Fall 2000]: 9–14; Mike ture and property; and rehabilitation. Dombeck, “How Can We Reduce the Fire Danger in the Interior West?”, Fire Management Today 61(1) [Winter • Helps quantify meaningful per 2001]: 5–13; and Hutch Brown, “Thirtymile Fire: Fire formance requirements under Behavior and Management Response,” Fire Management Today 62(3) [Summer 2002]: 23–30.
Fire Management Today 12 Land Management The FPA System focuses on the goals, strategies, Planning Framework and objectives identified in developing fire The FPA System focuses on the management plans. goals, strategies, and objectives identified in developing fire man agement plans (FMPs). FMPs are • Defines measures of accomplish use. Weighting takes into account derived from broad land and ment (fire management objec such variables as time of year and resource management plans, which tives) over time; fire intensity levels (fig. 1). assess the landscape condition and • Addresses public and firefighter define the desired future condition. safety; and Weights relate to objectives in Wildland fire specialists then deter • Describes sensitive social, eco other fire management units. For mine the appropriate role of wild- nomic, and resource issues relat example, fire suppression is more land fire management in achieving ed to fire management strategies urgent in the wildland/urban inter the desired condition, and they for and objectives. face (WUI) than in the backcountry, mulate FMPs accordingly. so the suppression objective has a Fire management objectives are greater weight in a WUI fire man The FMPs define fire management important triggers for the FPA sys agement unit than in a backcoun units—specific areas that are dis tem. For each fire management try unit. tinct from adjacent units in terms unit, the FPA system documents of fire regime, management objec fire management objectives from By assigning weights, the FPA tives, topographic features, values the FMP and translates them into System helps managers set priori to be protected, or other key fac meaningful measures for the analy ties and budgets. For example, a tors. For each unit, the FMP: sis model (fig. 1). planning unit might have several different fire management units, • Formulates suppression goals and Weighted Objectives with strategies ranging from pro objectives that support the Through an interdisciplinary and/or tecting communities in the WUI, to desired condition; interagency process, the FPA restoring the natural role of fire in • Specifies programmatic System weights each fire manage wilderness, to protecting threat approaches or strategies for man ment objective based on such fac ened species and restoring habitat. aging fire; tors as values to be protected from By weighting the associated objec wildfire or enhanced through fire tives, the FPA System lets man-
Example: Suppression objectives Fire management plan for a particular fire man Resource management agement unit Fire Program Analysis Land management plan plan (Fire management System input (General direction) (More specific direction) objective) (Damage threshold)
Suppress all wildfires.* Protect critical habitat for Over the next 5 years, 100 Wildfires at or greater threatened and endan percent of all wildfires in than fire intensity level 3 gered (T&E) species from habitat for T&E species should be kept to less damage by wildfires, espe are controlled during ini than 100 acres from April cially high-intensity fires. tial attack. to June. Weight = 8 (high relative importance)
* A wildfire is an unwanted wildland fire. Figure 1—Relationship of land and resource management planning through fire management plans to the Fire Program Analysis System for a sample fire management objective (wildfire suppression). The inputs are designed to be meaningful performance measures.
Volume 64 • No. 2 • Spring 2004 13 agers prioritize the needed work For example, many outcomes from objectives is to ensure that indica and set corresponding funding lev prescribed fire are very long term. tors are meaningful to field-level els. Interim measurable objectives are fire managers and high-level deci therefore needed for short-term sionmakers alike. Formulating program accountability. Moreover, Measurable Objectives the National Fire Plan and the 10 A Big Step Forward In the past, accountability systems Year Comprehensive Strategy Implementing the FPA System will for Federal land management agen Implementation Plan prescribe pro take time. Coming up with appro cies have focused on outputs rather grams in the WUI with social and priate short-term indicators as than outcomes. The reason was cultural outcomes that are interim measures for long-term simple: Outputs were deemed more extremely difficult to measure. wildland fire management out measurable. comes will not be easy. Moreover, The key to successful performance the new system cannot be expected However, under the Government measurement is knowing which to make the hard political choices Performance and Results Act of program indicators have the most for decisionmakers. Many factors in 1993, program activities must be significant relationship to desired addition to performance must still directly linked to desired outcomes, outcomes. Defining performance go into final decisions on resource and they must be measured for indicators is where science and art allocation. effectiveness in achieving those results (Robertson 1998). Tradi In fact, the FPA System will only be tional performance measures such By weighting each fire as good as the process of land man as costs and number of acres treat management objective, agement planning it is based on. ed reflected a program’s efficiency the FPA System helps Stakeholders and the public must but not its effectiveness. Such managers set priorities be meaningfully engaged to ensure measures say nothing about that land and resource manage whether results were achieved. and budgets. ment plans reflect a full range of ecological, social, and economic The FPA System addresses the values. Plans must also contain come together as managers link the problem by drawing on the fire goals, desired conditions, and logic and rationale of policy and management objectives specified in resource management objectives programs with the primary activi FMPs. FMPs are based on land and that support an appropriate role for ties designed to produce results. resource management plans, which wildland fire in the landscape. focus on outcomes—on achieving For every set of activities, there desired conditions on the land. By However, the FPA System is an should be a set of indicators that deriving performance indicators important step toward increased test whether program logic and from objectives designed to achieve accountability in wildland fire man assumptions truly deliver intended the desired future condition, the agement. It will move the Federal results. Many indicators will be FPA System helps managers evalu wildland fire program from report interdisciplinary ecological meas ate not only the efficiency of a pro ing one-dimensional output data ures of resilient and sustainable gram, but also its effectiveness. toward providing more meaningful ecosystems; others will be social information based on outcomes. and economic measures of fire pro But that raises another problem. Managers need such information to tection for communities and the The greatest challenge ahead will make decisions that will stand the general public. be identifying fire management test of time in the public service. objectives and performance indica An additional challenge will be to tors that are both meaningful and ensure that the sum of the indica Reference measurable—that are both out Robertson, S.L. 1998. Implementing the tors at the national level translates come based and useful for showing Government Performance and Results into a defensible and coherent accountability. Act: Performance measurement in the budget structure for Congress. The U.S. Forest Service. Masters thesis. ■ challenge in developing measurable Eugene, OR: University of Oregon.
Fire Management Today 14 A NEW TOOL FOR FIRE MANAGERS—AN ELECTRONIC DUFF MOISTURE METER
Peter R. Robichaud and Jim Bilskie
rescribed fires are increasingly being used to reduce haz The DMM600, a duff moisture meter, provides P ardous fuels, a major objective reliable, real-time measurements of duff moisture of the National Fire Plan. Despite content. advancing technology and ever- improving models, fire managers still find it challenging to deter closure, slope, aspect, and microto through a cooperative research mine the right time for a pre pography, duff moisture levels can effort by three of the USDA Forest scribed burn. vary significantly across the land Service’s research and development scape, even at the hillslope scale. units—the Forestry Sciences Measuring Duff Laboratory in Moscow, ID; and the Moisture Content These variations make it important Fire Sciences Laboratory and The effect of fire on the forest floor to use real-time duff moisture Missoula Technology and can vary from merely removing the measurements to estimate duff Development Center, both in litter to totally consuming the duff, consumption and, more important Missoula, MT. The Forest Service which exposes the mineral soil and ly, to achieve desired postfire duff collaborated with Campbell alters the surface soil structure. depths. Using a measured duff Scientific, Inc., in Logan, UT, for Fire managers often design pre moisture content in models like production and marketing. scribed fires to leave some of the FOFEM (Keane and others 2003), duff to protect the mineral soil. fire managers can better estimate How the Meter Works Duff thickness and moisture con the duff remaining after a burn. The fire manager collects a sample tent are the most important factors from the portion of the duff layer in determining duff consumption In the past, the fire manager might just above the soil mineral horizon during fires. pick up a bit of duff, squeeze it and pushes it through a #4 mesh firmly, and check the moisture sieve that fits in the opening of the In comparison to other woody before giving approval to begin the sample chamber (fig. 2). Passing fuels, duff has greater spatial and burn. The fire manager used past the duff through the sieve breaks temporal variation in moisture con experience, weather information, tent. Small precipitation events and and a “feel” for the current condi heavy dew accumulations that have tions as a final check. negligible effect on the moisture content of large fuels can signifi An additional tool is now available. cantly increase the moisture con The DMM600,* a duff moisture tent of fine fuels and litter. And meter (fig. 1), provides reliable, subsurface duff can lose moisture real-time measurements of duff through evaporation much more moisture content. The DMM600 quickly than the large woody fuels. was patented and developed Due to subtle differences in canopy
* The use of trade, firm, or corporation names in this Pete Robichaud is a research engineer for publication is for the information and convenience of the USDA Forest Service, Rocky Mountain the reader. Such use does not constitute an official Research Station, Moscow, ID; and Jim endorsement of any product or service by the U.S. Department of Agriculture. Individual authors are Bilskie is the soil physicist at Campbell responsible for the technical accuracy of the material Figure 1—The duff moisture meter Scientific Inc., Logan, UT. presented in Fire Management Today. (DMM600).
Volume 64 • No. 2 • Spring 2004 15 up large organic fragments and A B removes sticks and rocks, allowing for more uniform packing.
After the sieved material fills the chamber, the cap is put on and the compression knob turned until an audible indicator signals that the sample is properly compressed (at 15 pound-force [66 N]). The meter then automatically takes the meas urement and displays it at the base of the instrument. Readings are dis played in real time only; measure C D ments are not stored. Total time needed to sieve and measure each sample is about 30 seconds.
The tough, lightweight DMM600 is a portable, battery-powered sensor that was developed from frequency domain reflectometry technology. When proper pressure triggers a measurement, a high-frequency signal of 42 MHz is applied to the sensor electrodes at the base of the Figure 2—Using the duff moisture meter in the field is simple and fast. Toggle the power sample chamber, and the sensor switch on and select the duff to be sampled—usually from the lower half of the duff layer. electronics detect the change in (a) Push the duff sample through the sieve into the chamber. (b) Fill the chamber with frequency of the reflected signal the duff. (c) Compress the sample using the hand-turn knob, stopping when the audible signal is sounded. (d) Read the moisture content from the display window at the base of (fig. 3) (Robichaud and others the instrument. 1999, 2000). The frequency change depends on the dielectric constant each sample is pressed evenly of the medium adjacent to the sen against the sensor electrodes, sor electrodes. Because the dielec which reduces measurement vari tric constant of the medium varies ability. with moisture content, the frequen cy change can be easily related, Calibration through a simple calibration func tion, to provide moisture content. The factory-supplied calibration for The unit’s microprocessor uses a the DMM600 is derived from labo factory-supplied calibration to con ratory measurements of the volu vert the frequency to a volumetric metric moisture content of duff moisture content and displays the from eight different forested sites. value in the LCD readout. Depending on elevation, the cover species included Engelmann spruce (Picea engelmannii), subalpine fir Figure 3—The bottom of the sample cham Air voids in the duff can reduce the ber is a printed circuitboard containing apparent dielectric constant and/or (Abies lasiocarpa), lodgepole pine two interlocking finger sensor electrodes create a poor contact between the (Pinus contorta), western larch attached to a movable piston. duff and the sensor electrodes. (Larix occidentalis), ponderosa pine Using the meter’s compression fea (Pinus ponderosa), and Douglas-fir ture on sieved duff ensures that (Pseudotsuga menziesii). Because
Fire Management Today 16 Specifications The standard calibration curve will likely meet the Physical needs of most fire Dimensions: 3.5 inch (9 cm) diameter 10 inch (25 cm) length managers across a Weight: 3.7 lbs. (1.7 kg) range of vegetation Sieve: #4 mesh = 0.203 inch (5.16 mm) types. 3 inch (7.6 cm) diameter
Power cent at 60 percent volumetric mois Battery: 9 V alkaline ture content and approximately Measurements per battery: > 2000 ±1.5 percent at 30 percent volu metric moisture content. It is rec Performance ommended that the average of sam ples from several nearby locations Accuracy: ± 5 percent for full scale range be used to reduce the effects of nat Resolution: 1 percent volumetric moisture content ural variability.
Included with the DMM600: The meter’s response to changing • CD containing (a) video instruction, (b) PCDMM software, moisture content is best described (c) DMM600 Instruction Manual, and (d) Excel spreadsheet for with a quadratic calibration equa calibrations. tion: • 6-foot (1.8-m), 9-pin serial cable. • Soft-sided, padded carrying case with screwdriver and spare 9-volt Volumetric Water Content = battery. 5.288 + 5.905 × freq – 0.142 × freq2,
where freq is the DMM600 readout the individual calibration curves The error bars indicate that meas- frequency in MHz. User-derived cal were similar, the data were com- urement accuracy decreases as ibrations can be determined using bined to develop a single, standard moisture content increases. The the laboratory procedures described calibration curve (fig.4). accuracy is approximately ±4.0 per- in the DMM600 Instruction Manual and the DMM600 calibration spreadsheet.xls provided in the PCDMM software package (Campbell Scientific 2000). User- defined calibrations are entered into the PCDMM interface and loaded to the DMM600 through a serial port connection.
Studies done on eastern hardwood duff from Massachusetts show little deviation from the standard calibra tion curve. It is likely that the stan dard calibration curve will meet the needs of most fire managers across a range of vegetation types.
Figure 4—DMM600 standard calibration curve. The variability is smallest in the lower moisture/higher frequency area of the curve, where accuracy is most critical.
Volume 64 • No. 2 • Spring 2004 17 Gravimetric Moisture Duff moisture content is critical information for Content fire managers making operational and planning The basic operation of the DMM600 decisions for prescribed burns. gives the volumetric moisture con tent of the sampled duff. Using a simple calibration process, the and uploaded to the DMM600 gravimetric (“dry-weight-based”) through a serial port connection on Ordering moisture content—the measure the base. Information ment most commonly used by fire managers—can be added to the Because duff bulk density is rela Ordering information can be instrument readout. tively constant, this calibration obtained from the Campbell process can be completed before Scientific, Inc., Website at: A value for duff bulk density is going to the field to make duff
Fire Management Today 18 FUEL MAPPING FOR THE FUTURE
C.W. Woodall, G.R. Holden, and J.S. Vissage
he large wildland fires that raged during the 2000 and The lack of a nationally consistent and T 2002 fire seasons highlighted comprehensive inventory of forest fuels has the need for a nationwide strategic hindered large-scale assessments. assessment of forest fuels. The lack of a nationally consistent and com prehensive inventory of forest fuels Phase 3. Subsets of the permanent for a portion of all permanent plots has hindered large-scale assess FIA field plots are sampled for indi to maintain current fuels data. ments—essential for effective fuel cators of forest health. Down woody hazard management and monitor material (DWM)—duff, litter, and Seamless Dataset. When all phases ing reduction treatments. Data fine and coarse woody debris—is of the FIA inventory program are from the USDA Forest Service’s one indicator. Both planar intercept integrated, the result is a seamless enhanced Forest Inventory and methodologies and fixed-radius dataset connecting DWM invento Analysis (FIA) Program is key to plots are used to measure DWM. In ries, traditional forest inventory creating fuel maps to improve addition, inventory data from indi data, and remotely sensed imagery. large-scale assessments. cators, such as soils and vegetative The resulting multidimensional diversity, when combined with fuel maps, developed using the The Program DWM data, help refine estimations most current data available, provide The enhanced FIA Program has of nonwoody fuels. When using the national fuel estimates from the three phases that contribute to fuel enhanced FIA Program, all three duff layer to the top of the forest map synthesis. phases should be sampled annually canopy (fig. 2).
Phase 1. Remotely sensed images help determine whether field crews can access potentially permanent FIA field plots. Forest and nonforest stratification data layers from phase 1 might both constrain fuel loading predictions in forested areas and facilitate certain analyses, such as locating fuels adjacent to wild land/urban interfaces.
Phase 2. Permanent FIA field plots are established (fig. 1), and stand ing live and dead trees are invento ried. The data from this phase pro vide information about tree species, crown ratio, and height. They are useful for estimating fuels in each plot’s live and dead standing tree stands.
Chris Woodall is a research forester, Geoff Holden is a GIS specialist, and John Vissage is a research forester for the USDA Figure 1—Annual forest inventory analysis phase-2 plots (forested condition) for the Forest Service, North Central Research North-Central United States from 1999 to 2001. The plots allow for timely fuel hazard Station, St. Paul, MN. assessments at strategic scales.
Volume 64 • No. 2 • Spring 2004 19 Mapmaking spatial interpolations to estimate maps. A third way to create effec Modeling, interpolation, and ancil fuel loadings for forests not directly tive fuel maps is to use mean fuel lary stratification are methods for sampled by DWM inventory. When loadings from an ancillary data creating fuel maps using all phases combined with phase-1 forest and layer, such as ecological units or of the enhanced FIA Program to nonforest data layers, the fuel load forest type, combined with phase-1 satisfy a variety of needs. Selecting ing estimates can produce fuel forest and nonforest data layers. the appropriate fuel-mapping tech nique depends on the objectives, acceptable levels of error, regional DWM sampling intensity, and the proximity of the map region to international borders and large bodies of water.
A modeling approach is appropriate when fuel loads are predicted based on a DWM inventory from phase-2 Figure 2—Components of forest fuels. Individual components were sampled during a plots. Another approach is to use comprehensive, vertical fuel hazard assessment.
Figure 3—Strategic assessment of fuels in the North-Central United States, in tons of fuel per acre, for duff (A); litter (B); fine woody debris (C); coarse woody debris (D); and standing live trees—small (E) (0 to 5 inches [0–12.7 cm] diameter at breast height [dbh]), medium (F) (5.1–10 inches [13–25 cm] dbh), and large (G) (greater than 10 inches [> 26 cm] dbh).
Fire Management Today 20 When all phases of the inventory program are integrated, a seamless dataset connecting downed-wood material inventories, traditional forest inventory data, and remotely sensed imagery results.
For example, figure 3 shows seven regional maps for each of the seven fuel layers (duff, litter, two layers of woody debris, and three layers of standing live trees). We produced the maps by interpolating fuel load ings from DWM sample plot loca tions, combined with phase-1 forest and nonforest data layers. The maps Figure 4—Relative fuel loadings (duff layer to canopy top) for the North-Central United were based on 8,569 phase-2 plots States, based on combining scaled inventory maps for duff, litter, two layers of woody inventoried from 1999 to 2001 and debris, and three layers of standing live trees. 249 DWM plots inventoried in 2001. Value of Fuel Mapping toring fuel hazards across forest Creating comprehensive and con ecosystems, might mitigate the Figure 4 shows a single map inte sistent national fuels maps is possi potential for another devastating grating all the data in figure 3. Fuel ble using the extensive forest fire season. loadings for each individual map inventory data available through were scaled to a uniform data range the enhanced FIA program, coupled For more information, contact (0 to 1), then combined and with new mapping and remote- Chris Woodall, USDA Forest rescaled, yielding a comprehensive sensing technologies. Based on the Service, North Central Research assessment of fuels on a strategic information provided in such maps, Station, 1992 Folwell Ave., scale. completing large-scale assessments, St. Paul, MN 55108, 651-649-5141 ■ essential for managing and moni- (tel.), [email protected] (e-mail).
Volume 64 • No. 2 • Spring 2004 21 DIVERSIFYING FUELS MANAGEMENT TO OFFSET UNCERTAINTY
John Hof
s the USDA Forest Service embarks on a campaign to Diversifying a land management portfolio helps A reduce and better manage for reduce the chance of catastrophic loss. est fuels, the outcomes are uncer tain. How do we manage for uncer tainty? One approach is to diversify Reducing uncertainty and igno ject to the same random events management actions to reduce the rance through research or assess so that the responses to the chance of severe loss and to ments is always good but takes events are different; and enhance adaptive management. time. Besides, the systems we man • Performing the same activities in age are often inherently stochastic. different areas only when they are Risk, Uncertainty, and In such systems, converting uncer far enough apart so that they are Ignorance tainty and ignorance into risk is not subject to the same random Ecosystems are stochastic (that is, usually all we can hope to achieve. events. subject to random variability), In the short term, we face consider which renders our knowledge of able uncertainty or ignorance. Yet In both cases, the object is to them imperfect. There are three doing nothing while trying to learn reduce the random variability of conditions of imperfection, each is a gamble. system response through diversifi more difficult to deal with than the cation. We can do this without last: knowing the actual variances Eliminating randomness because we know that landscape • Risk, where we know both the is impossible—all we independence increases with dis probabilities of the possible out can do is reduce the tance—the farther apart two land scapes are, the greater their inde comes and the effects that our chance of making a alternative courses of action will pendence from each other. have on those probabilities; really big mistake. • Uncertainty, where we don’t Long-Term Adaptive know the probabilities or the Management effects, but we do know possible Short-Term Diversification is also a good basis outcomes; and Diversification for adaptive management, whereby • Ignorance, where we don’t even There is a solution: diversification. we experiment with management know what outcomes are possi Diversifying land management actions to learn more about the ble. activities is similar to a stock port systems we manage. By diversifying folio manager diversifying invest our approach to management, we Although analytical approaches are ments to reduce the chance of cata learn more—and sooner. Of course, effective for managing risk, they strophic loss. Like diversifying a we must monitor the results to typically fall short when uncertain stock portfolio, diversifying a land learn from the “management exper ty and ignorance are involved. For management “portfolio” reduces its iments” and adapt management example, for stochastic program overall variability. strategy accordingly. ming and expert systems, we need information on outcome probabili For the short term, two principles Let’s take fuels management as an ties to manage randomness rigor of diversifying a land management example. To prevent an unwanted ously. portfolio are: stand-replacing fire, fuels managers often tend to look for Best Manage John Hof is a chief economist for the USDA • Performing different land man Forest Service, Rocky Mountain Research ment Practices, assuming that in Station, Fort Collins, CO. agement activities in areas sub- any given situation, there is always
Fire Management Today 22 Alternative methods of reducing hazardous fuels and restoring long-needle pine forest. Left, the sun shines through smoke and flames during a prescribed fire on the Colville National Forest in Washington. Right, a crew member limbs up a tree as part of a thinning proj ect in the Jemez Mountains of northern New Mexico. Diversifying treatments can reduce the chance of unwanted outcomes. Photos: Thomas Iraci, USDA Forest Service, Pacific Northwest Region, Portland, OR, 2001; and Kristen Honig, National Park Service, Los Alamos, NM, 2002.
a “best” thing to do. In reality, we are usually uncertain about the When land management is diversified, effectiveness of different fuels we learn more, sooner. reduction methods, and we are also typically uncertain or ignorant about the effects of fuels reduction we learn about the actual fire and to see what happens. But because methods on the ecosystem. ecosystem responses and can adapt we can never eliminate random our management strategies accord ness, we do not always know for Diversification means using a vari ingly. sure what management activity is ety of fuels reduction methods in a best in a given situation. given forest area and using the Facilitating Active same method only in well-separated Management We can make the best of a bad situ areas. That reduces risk: If a fuels Land managers today face a diffi ation by diversifying land manage reduction method harms the cult conundrum. High fuel loads ment activities. By accepting the ecosystem, it affects only a small and their associated hazards mean chance of making small mistakes, area; and if it fails to prevent a that we do not have the time to we can reduce the chance of mak crown fire, the affected area is also ■ experiment in a few areas and wait ing big ones. limited. As we monitor the results,
Volume 64 • No. 2 • Spring 2004 23 PROBABILITY OF SPOT FIRES DURING PRESCRIBED BURNS
John R. Weir
pot fires have always been a problem on prescribed burns. If we can narrow spot fire causes down to a S Just the possibility of a spot single main weather factor, burn bosses might fire can cause mental and physical focus on that variable, possibly reducing the stress on burn bosses and crews. Actual spot fires can cause personal chance of spot fires. injury or even loss of life, as well as costly damages and loss of public ambient air temperature is below that variable, possibly reducing the support for prescribed fire pro 60 ºF (15 ºC) when burning. chance of spot fires. grams. • Windspeed. Windspeeds of at least 8 miles (13 km) per hour Many private and public land man Key Variable are needed to ignite and burn agers in Oklahoma have told me At the Oklahoma State University standing fuels (Britton and that they avoid prescribed burning Research Range (OSURR), we con Wright 1971). However, wind- for fear of spot fires and escaped duct prescribed burns during differ speeds of more than 20 miles (32 fires. Many have the resources ent seasons all over Oklahoma. km) per hour can create prob needed to conduct prescribed fires, Fuels include tallgrass prairie lems with firebrands and other but lack the experience or knowl (NFES fuel models 1 and 3—see blowing debris (Wright and Bailey edge to deal with spot fires. A sim Anderson 1982), post oak–blackjack 1982). ple guideline or rule-of-thumb oak (fuel models 3, 8, and 9), erod might help. ed mixed prairie (fuel models 1 and If we can narrow spot fire causes 3), sandsage grassland (fuel model down to a single main weather fac 4), and oak–pine (fuel models 3, 8, Variables Affecting tor, burn bosses might focus on Fire Behavior 9, and 11). Since 1996, we have Weather factors are the main vari ables that burn bosses can use to predict and monitor prescribed fire behavior. In general, there are three main weather factors:
• Relative humidity. Burning when relative humidity exceeds 40 percent significantly slows rates of spread (Lindenmuth and Davis 1973) and reduces danger from firebrands (Green 1977). • Temperature. Bunting and Wright (1974) found that danger from firebrands was lower if the
John Weir is the superintendent of the Oklahoma State University Research Range and prescribed burning instructor in the Firewhirl on a prescribed fire in Oklahoma. Fuels were tallgrasses, sand sagebrush, and Rangeland Ecology and Management pro- scattered eastern redcedar. About 20 feet (6 m) tall, the firewhirl left the burn unit and gram, Plant and Soil Science Department, started two small spot fires that were quickly contained. Photo: John Weir, Oklahoma Oklahoma State University, Stillwater, OK. State University Research Range, Stillwater, OK, 2001
Fire Management Today 24 been keeping track of spot fires on our prescribed burns. We consider a spot fire to be any fire outside the burn unit, no matter what the size or cause.
The size of a spot fire depends on fuel loads outside of the burn unit, crew size, crew experience, equip ment present, equipment depend ability, firebreak type and size, and weather conditions. Our spot fires have ranged in size from less than one square foot (0.09 m2) to 120 acres (264 ha). Most of our spot fires were caused by firebrands from crowning eastern redcedar Figure 1—Spot fires on prescribed burns in relation to relative humidity. Twenty-one of (Juniperus virginiana). The rest 99 prescribed fires conducted across Oklahoma from 1996 to 2002 were associated with were usually caused by smoke, fire spot fires, but only two spot fires occurred when relative humidity was greater than 40 whirls, or flaming oak leaves or percent. tallgrasses floating or blowing across the fireline. When we reviewed the burn data, one weather When the OSURR crew conducts variable stood out in association with spot fires: prescribed fires, we record weather low relative humidity. data onsite before, during, and after the burn. We also note whether or not a spot fire occurred. When we That does not mean that managers Spot Fire Probability reviewed the burn data, one weath should never prescribe-burn when er variable stood out in association What is the probability that a spot relative humidity falls below 40 fire will occur when relative with spot fires: low relative humidi percent. Some parts of the United ty (fig. 1). From 1996 to 2002, we humidity is below 40 percent—or, States and some particular burn for that matter, at any level? conducted 99 burns, 21 of which units might require low relative produced spot fires. All but two Knowing spot fire probability can humidity to achieve the goals and be vital in preparing and safely con occurred when relative humidity objectives of a prescribed burn. was at or below 40 percent. ducting prescribed burns. But it should be etched into every 40-Percent Threshold We used the information from our burn boss’s mind that a spot fire 99 prescribed burns to develop a set Research has shown that fine fuels might well occur on a prescribed of spot fire probabilities at various ignite and burn easily when relative burn when relative humidity is levels of relative humidity. We used humidity is below 40 percent, below 40 percent. Of course, burn the following formula (based on whereas ignition slows when rela bosses should always be ready for Steele and Torrie 1980): tive humidity is above that thresh spot fires on every prescribed burn, old (Britton and Wright 1971; no matter what the relative humid P = SF ÷ PF, Green 1977; Lindenmuth and Davis ity. We recorded one spot fire when 1973). Our experience validates the the relative humidity was as high as where P is the probability of a spot research. A threshold value of 40 73 percent. It was caused by fire fire occurring, SF is the number of percent relative humidity might brands thrown by a crowning east spot fires, and PF is the number of suggest an excellent rule-of-thumb ern redcedar into heavy fuels across prescribed fires. for conducting prescribed burns. the firebreak.
Volume 64 • No. 2 • Spring 2004 25 Our data showed a 21.2-percent probability of a spot fire occurring on a prescribed burn when relative humidity was between 20 and 80 percent, or about one out of five burns. For the 40-percent relative humidity threshold, the probability of a spot fire was 41.3 percent when relative humidity was below the threshold and only 3.8 percent when it was above the threshold—a substantial difference.
The data also showed that, below Figure 2—The probability of spot fires as a function of relative humidity, based on 99 the 40-percent threshold, spot fire prescribed fires conducted across Oklahoma from 1996 to 2002. probability rose with each 5-per cent drop in relative humidity (fig. Within the range of 20- to 40-per of a crew’s anxiety about spot fires 2). At 25-percent relative humidity, cent relative humidity, there is a on prescribed burns when relative there appears to be another thresh large difference in the probability of humidity exceeds 40 percent. Best old: Below this point, there was a a spot fire occurring. When relative of all, it can help managers reduce 100-percent probability of a spot humidity is below 25 percent, burn risk and increase safety for their fire occurring. But in the 25- to 29 bosses should be prepared for a crews. percent relative humidity range, spot fire probability dropped from References 100 percent to just 46.2 percent; It should be etched into Anderson, H.E. 1982. Aids to determining and in the 30- to 35-percent range, every burn boss’s mind fuel models for estimating fire behavior. only one out of three burns was Gen. Tech. Rep. INT–122. Ogden, UT: that a spot fire might USDA Forest Service, Intermountain likely to produce a spot fire. Forest and Range Experiment Station. well occur on a Britton, C.M.; Wright, H.A. 1971. prescribed burn when Correlation of weather and fuel variables Lessons for Burn to mesquite damage by fire. Journal of Bosses relative humidity is Range Management. 23: 136–141. Bunting, S.C.; Wright, H.A. 1974. Ignition What does all this mean for burn below 40 percent. capabilities of nonflaming firebrands. bosses? It does not mean that man Journal of Forestry. 72: 646–649. agers should never prescribe-burn Green, L.R. 1977. Fuel breaks and other when relative humidity falls below 100-percent probability of a spot fuel modification for wildland fire con trol. Ag. Hb. 499. Washington, DC: USDA 40 percent. But managers should fire. But they can cut the risk by Forest Service. still take the 40-percent threshold about half with just a slight Lindenmuth, A.W.; Davis, J.R. 1973. into account, particularly when increase in relative humidity. Predicting fire spread in Arizona oak chaparral. Res. Pap. RM–101. Fort inexperienced personnel are con Collins, CO: Rocky Mountain Forest and ducting prescribed burns, when This information can help burn Range Experiment Station. heavy fuel loads are adjacent to the bosses determine spot fire potential Steele, R.G.D.; Torrie, J.H. 1980. Principles when considering burn units or and procedures of statistics. New York: burn unit, or when a fire escape McGraw-Hill. could result in terrible publicity or burn days. It can also help them Wright, H.A.; Bailey, A.W. 1982. Fire ecolo even litigation. determine crew size and equipment gy: United States and Southern Canada. needed. It might help relieve some New York: Wiley and Sons. ■
Fire Management Today 26 FIRES IN THE WILDLAND/URBAN INTERFACE: BEST COMMAND PRACTICES
Michael S. Rohde
recently completed a study (Rohde 2002) providing insight Within the ICS, experts prefer to organize initial I into critical decisions by com- response to a major WUI incident by branches mand officers on some of rather than divisions. California’s most notorious wild fires in the wildland/urban interface (WUI) (see the sidebar). My study • Calculating resource needs, and Incident Command focused on the first several hours of • Projecting fire behavior and System response to the fires, a period of spread. time when organizational develop Within the Incident Command System (ICS), most experts consult ment and control can be as com All experts consulted in the study ed in the study prefer to organize plex as the fire itself, and State or had developed prefire plans for initial response to a major WUI Federal incident management their respective areas of responsi incident by branches rather than teams have not yet been mobilized. bility, and many had involved coop divisions. They establish branches I consulted with experts with erating agencies, including law for each flank of the fire and possi exceptional command experience enforcement. Some had followed up bly for structural protection, law on WUI fires. on prefire planning with intera enforcement and evacuation, or gency tabletop exercises. The study shows the fire environ ment common to catastrophic WUI fires (see the sidebar). It also identi Wildland/Urban Interface fies best command practices that might be used by incident com Fires Studied manders and others responsible for The study focused on command ately adjacent to heavily urban leadership on such fires in the complexities and key decisions on ized areas. On each fire, the vast future. Some of the practices are six notorious wildland/urban majority of loss occurred during summarized below. The practices interface fires in California: the first 12 hours. are best utilized in a “systems fash ion”—by integrating multiple • The 1990 Paint Fire near Santa Subject matter experts who com concepts into a command method Barbara, mented on these and other fires ology. • The 1991 Tunnel/Berkeley Hills included: Fire in Oakland and Berkeley, Prefire Planning • The 1993 Old Topanga Fire in • Bill Teie, Bill Clayton, Tim The study found that planning for Malibu, Sappok, John Hawkins, and wildfire risks in the WUI and in his • The 1993 Kinneloa Fire near Chuck Manor from the torical fire corridors is critically Altadena, California Department of important. Planning might include: • The 1993 Laguna Fire in Forestry and Fire Protection Orange County, and (CDF); • Conceiving strategies and tactics, • The 1996 Harmony Fire near • Gary Nelson from the Los • Identifying values at risk, Carlsbad. Angeles County Fire • Planning deployments and evacu Department; and ations, Collectively, these six fires caused • Mike Warren from the Corona 30 fatalities, burned 4,907 struc Fire Department (formerly with Michael Rohde is a battalion chief for the tures and 52,422 acres (21,215 the USDA Forest Service and Orange County Fire Authority, Rancho ha), and occurred in or immedi CDF). Santa Margarita, CA.
Volume 64 • No. 2 • Spring 2004 27 other specific needs. Operations branch directors may then establish divisions or groups as resource availability or situational needs dic tate.
This organizational approach has several advantages. By establishing branches first, the incident com mander (IC) can immediately organize the entire projected fire area, leaving no part of the fire unsupervised. Quick establishment
Experts suggested that the best way to The May 13, 2002, Antonio Fire threatened 500 Orange County homes and burned 1,500 overcome conflicting acres. Photo: Orange County Fire Authority, Orange County, CA. demands on the OSC’s time early in a WUI Experts consulted in the study sug perimeter control in favor of struc incident is to delegate gested that the best way to over tural protection risks unabated fire more authority to come the conflicting demands on expansion, increased structural the OSC’s time is to delegate more risk, and difficulty of control. In branch directors. authority to branch directors for some situations, perimeter control operations section management. might have to be abandoned for a With branch directors responsible period of time, but it must be of a basic command framework can for managing operations, the OSC reestablished as soon as possible. alleviate concerns about independ can provide less direct oversight ent actions. ICs are also immediate and devote more time to partner Crews, dozers, air tankers, and ly able to place initially responding ship with the IC. The OSC is also some engines are best assigned to command officers into high- freer to interact in other necessary perimeter control, whereas type 1 responsibility positions, thereby relationships and attend planning or 2 engines are ideally assigned to best using their local experience and strategy meetings as needed. structure protection, supported by and knowledge while capitalizing helicopters capable of working on the preexisting basis of trust On the Harmony Fire, for example, close to the ground in heavy they are likely to have. the five branch directors assumed smoke. A common strategy is to responsibility for a great deal of the “pinch the flanks” through perime Role of the Operations operational leadership. The OSC ter control to limit the width of the Section Chief was able to focus more exclusively fire’s head as it enters areas with It is critical for the operations sec on coordination and mobility of structures. tion chief (OSC) to communicate resources and on ensuring that nearly constantly with the IC dur resources were responding to what Experts consulted in the study ing early incident development. the fire might do rather than what acknowledged the difficulty of sort However, the OSC must also over it was already doing. ing out the key issues from all the see suppression and related activi minutiae on a WUI fire. They rec ties, demands that can interfere Strategy and Tactics ommended that ICs be careful to with communication between the The ideal strategy is to provide both focus on key issues such as: IC and OSC. Physical collocation is offensive perimeter control and not necessarily the solution. defensive structural protection • Protecting lives and property, simultaneously. Abandoning • Supporting effective operations with additional resources,
Fire Management Today 28 • Taking advantage of containment A common strategy is to “pinch the flanks” opportunities, and through perimeter control to limit the width of the • Holding line. fire’s head as it enters areas with structures. Experts recommended allocating resources to structural areas par be managed. Of the 900 engines Many experts suggested that med tially damaged by fire to prevent assigned to the Old Topanga Fire, ical personnel be prepositioned for additional loss from residual fires only 20 percent were actually com firefighter support, and that addi after the main fire has passed. mitted. Fire apparatus on the tional resources for technical res Pacific Coast Highway was backed cue and extraction be placed at Command Post Staff up for miles. their immediate disposal. In one Without timely logistical support, case, a burnover involving two fire key firefighting needs go unmet, Experts agreed that overordering fighter fatalities and two serious such as drinking water, food, or has become a serious problem. injuries required hours for person fuel. Unless command staff develop- Most could not visualize an inci nel extraction due to difficult ter ment—including logistics—match dent requiring more than 300 rain. Accident scenes should be es resource commitment, it risks engines. Prefire planning was sug quickly designated as “incidents falling hopelessly behind. To assist gested as key to effective resource within incidents,” with separate in command staff development, ordering and deployment. One command, communications, and firefighters are generally assigned expert suggested preestablishing resources. to initial logistics functions, such resource orders that can be placed as situation and resource status when fires reach certain bench Unified Command tracking, until they can be relieved marks, perhaps shown on fire pro On the six fires studied, the most by fully qualified ICS staff. jection maps with time ellipses at successful incident commands hourly intervals. Past fire history immediately organized a unified One expert suggested having can also guide resource needs command and ordering point. This responding resources report to one assessment and planning. helped reduce independent actions, of at least two staging areas on increase command cohesiveness, opposite sides of the fire for assign Risk Acceptance and and concentrate available firefight ment. The IC would communicate Mitigation ing resources on the most signifi potential assignments directly to Experts recommended allowing cant needs. the staging area manager, who operations involving elevated risk would fill the requests through only under very specific circum Incidents that included law face-to-face contact with available stances—generally, only when civil enforcement in the unified com resources and report the action to ian lives are directly at risk and mand were highly successful in the IC. In this manner, the IC could then only with strong planning and mounting evacuations. On the ensure check-in of resources and support. Under other circum Laguna Fire, for example, 26,000 reduce radio traffic. stances, operational risk must be people were evacuated from Laguna addressed on a continuing basis for Beach and the surrounding area in Ordering Resources all line assignments and mitigated 2 hours. The evacuation included California has a well-developed as much as possible through air planning for fire service access and mutual aid system, allowing more craft support, construction or iden separate civilian egress. In contrast, than 900 engines to be assigned to tification of safety zones, communi lack of effective traffic management two of the fires studied. However, cations and lookouts, varied tactical during the Berkeley Hills/Tunnel that was more than could possibly approaches, and other means. Fire contributed directly to loss of life.
Experts agreed that overordering has become a It is essential to assign an informa serious problem and that prefire planning is the tion officer to deal with the media on wildland fires in the WUI. Com- solution. mands that utilized the media to
Volume 64 • No. 2 • Spring 2004 29 Catastrophic WUI Fires: Common Factors Forty-seven factors were common clearance, and some had concentra Situation-driven tactics compro to all six WUI fires studied in tions of combustible landscaping mised and elevated firefighter risk, California. Some are summarized adjacent to structures. In each case, as did the need to effect rescues below. the presence of threatened or and civilian evacuations. Coordin endangered species was an obstacle ation with police agencies for traffic General Factors to presuppression activities. control and evacuation was diffi All fires occurred during critical cult, as was acquiring accurate fire weather patterns, involving Water systems were often unable to information on the situation and Santa Ana or other foehn winds. provide adequate fire flow or failed status of resources. Inability to pro They occurred during seasonal during the fires, and arterial road vide adequate and timely logistical drought, with critical live and access and egress were generally support, including water and fuel, dead fuel moistures, following a limited. Many burned structures compromised firefighting. Loss of wet winter preceded by multiple were isolated and served by sub momentum occurred in perimeter years of drought. They generally standard or hazardous roads or control activities as a result of con occurred in steep mountainous bridges. In some areas, dense struc current structural protection terrain with a history of wind- tural spacing aided conflagration demands. Effective command orga driven wildfire and structural loss. development. nizational development was hin dered by lack of qualified staff and Native chaparral fuels were abun Emergency Response the rapidly changing fire condi dant in the fire areas. The fires Each fire caused significant injury, tions. exhibited conflagration behavior and half cost lives. Available region immediately following mass struc al fire resources were overwhelmed Initial command post locations tural involvement, including by the initial fire problem, and were generally inadequate, with extreme burning intensity, fire massive structural loss occurred command posts burned over on whirls, long-range heavy spotting, during the first 12 hours on each three of the six studied fires. After mass ignition, high energy incident. Firefighters practiced resources were initially deployed, release, and rapid rates of spread. structural triage to select defensible incident commanders had difficulty homes, and a period of independent mobilizing them to address new Community-Related action by firefighting resources and evolving threats. Despite Factors occurred on each of the fires. California’s well-developed Fire and Regional commitments to multiple Rescue Mutual Aid System, mobi Affected communities were largely lization of mutual-aid resources constructed of nonfire-resistant fires compromised availability of aircraft, hand crews, and dozers. was slow; problems were exacerbat materials, including wood shake ed by overordering resources on roofs. Many properties lacked ade Communications centers and fire radio systems were overwhelmed. some incidents. Initial use of multi quate fuel modification or brush ple resource ordering points com-
get out their messages succeeded in ern California, burning more than a communicating information about On the six fires quarter of a million acres and evacuations and other fire-related studied, the most destroying thousands of homes. matters. On incidents where the successful incident Such fires are common in media were not engaged, misinfor California and spreading across the mation ensued, including misdirec commands immediately Nation. All six of the fires I studied tion of civilians in the fire area. organized a unified confounded the best cooperative command and efforts of local, State, and Federal Improving Command ordering point. firefighters, who shared responsibil In fall 2003, more than 10 large ity for initial response. fires ringed urban centers in south-
Fire Management Today 30 plicated organizational and com Independent Action • On all six incidents, independ mand activities. Limited communi Particularly problematic during the ent action contributed to situa cations and interaction between early periods of firefighting was tions where safety and efficiency fires and governmental emergency independent action by firefighting were compromised or organiza operations centers contributed to resources. In fact, independent tional control and resource confusion and support inadequa action was so prevalent that it accountability lost. cies. might be seen as typical during the • During several of the studied initial phases of a major WUI fire. fires, high-risk firing operations Human Factors with low probability of success On all six fires, firefighters and fire Independent action occurred either resulted from independent management officers acknowledged through intentional delegation by action. Some succeeded, but stress associated with high-risk, command or through unselected others directly contributed to high-consequence decisionmaking. organizational evolution. What structural loss. Many seemed frustrated by the drove it were overwhelming and • Firefighter entrapments result large structural losses. Public fear dangerous fire conditions during ed from independent action, and panic also affected firefighters, the first hours on an incident. including 20 separate entrap as did stress associated with fatali Often, an incipient command ments on the Old Topanga Fire. ties, entrapments, missing-person organization saw no other choice. • Independent action made it dif reports and searches, and concern ficult for command organiza for civilians defending their proper Independent action did achieve tions to deploy resources to new ties. some tactical benefits where indi threats and to coordinate vidual crews were well trained and searches and evacuations. On all fires, public volunteerism highly motivated. One expert con proved unmanageable and an sulted during the study conceded In general, the command impediment to firefighting opera that independent action yielded “a approach to independent action is tions. Firefighters used recogni lot of productivity,” but noted that to end it as quickly as possible by tion-primed decisionmaking, for the incident commander might not consolidating command and better or worse, in exercising tac know or be able to direct “what developing field supervisory posi tics and strategy, emphasizing the that productivity is.” tions, but this takes time and need for high-quality training and resources. Although experts have experience. Demands for fire serv All of the experts consulted during recommended ways of overcoming ice involvement in postfire recovery the study viewed independent some of the negative aspects of and political events exceeded all action as a “strategy of last resort.” independent action, its prevalence expectations. Independent action raises serious during the early phases of major issues and concerns: WUI fires—desired or not—sug gests the need for specific training in such operations.
Command decisions and actions I hope that my findings will help CA 92688, 949-858-8659 (tel.), 949 can and should be preplanned for fire management officers achieve 858-9168 (fax), such incidents, partly for firefighter superior leadership and command [email protected] (e-mail). safety and efficiency. Decisions for future incidents in the WUI. For must be rapidly made and imple a more detailed summary of my Reference mented on such rapidly evolving study, please contact Michael S. Rohde, M.S. 2002. Command decisions dur incidents, often in the absence of Rohde; Battalion Chief; Orange ing catastrophic urban-interface wildfire: organized management teams and County, CA, Fire Authority; 21 A case study of the 1993 Orange County Laguna Fire. M.A. thesis. California State with the best local capability avail Aloysia, Rancho Santa Margarita, University, Long Beach, CA. ■ able.
Volume 64 • No. 2 • Spring 2004 31 FIRE REGIMES IN THE NORTHERN ROCKIES*
Stephen W. Barrett
and managers today recognize the importance of understand Knowing the current status of the historical fire L ing natural disturbance regimes is critical for land management planning. regimes for maintaining sustain able ecosystems (USDA 1999, 2000a, 2000b). Ecosystem-based including one nonlethal (NL) type, stand (fig. 2). In dry ponderosa pine plans require information on long three mixed-severity (MS) types, stands, fire-scarred veterans often term disturbance history (Cissel and two stand replacement (SR) contain from 10 to 20 scars per tree and others 1999; Morgan and oth types. All six are described below. (Arno 1976; Arno and others 1995; ers 1994; Quigley and others 1996), Barrett 1988; Heyerdahl 1997). I because most terrestrial and aquat NL (< 25-year MFI). During the classified about 30 percent of the ic systems in the Northern Rockies presettlement era, stands in the NL database as NL, where MFIs ranged are disturbance adapted. fire regime (also known as “fre from about 10 to 26 years long and quent surface fire regime” [Brown averaged 17 years. Fire regimes classifications describe 2000]) experienced frequent under- in a general way the periodicity, story fires that promoted lightly The NL fire regime occurs largely severity, sizes, and patterns of fire stocked, uneven-aged structures in relatively dry forest types, for disturbance (Brown 2000), largely (Arno 2000). Brown (2000) defines example, in the ponderosa pine, dry at the stand scale (Arno and NL fires as those killing less than Douglas-fir, and dry grand fir Peterson 1983). To date, most clas 20 percent of the mature trees in a potential climax types. Most of the sifications have been largely theo retical. Such systems are useful at a broad conceptual level but lack pre cision and usefulness for fine- to mid-scale management planning (see the sidebar). In contrast, I developed an empirically based classification for forested land scapes in the northern Rocky Mountains. Six Fire Regimes I reviewed all published and unpub lished fire history studies in the Northern Rockies to establish a database containing 1,440 plot sam ples from 95 studies (fig. 1). Mean fire intervals (MFIs) from about 1600 to the present suggest that there are six different fire regimes in the Northern Rockies (table 1),
Steve Barrett is a consulting fire ecologist in Kalispell, MT. Figure 1—Study area in the northern Rocky Mountains. The fire regimes database * The article is distilled from a contract final report (Barrett 2002) prepared for a USDA Forest Service land contains 1,440 plot samples from 95 fire history studies, each represented by a dot. scape modeling project (Jones and others 2002). Shading depicts national forest land.
Fire Management Today 32 The National Classification: How Does It Compare? Schmidt and others (2002) devel • Fire regime IV: 35- to 100-year • Fire regimes I and III do not oped a national fire regimes clas MFI or greater, stand-replace show the relatively fine grada sification used by the USDA ment fires. tions found in the Northern Forest Service and other agen • Fire regime V: 200-year MFI or Rockies from nonlethal to cies. Their system is useful for greater, stand-replacement fires. mixed-severity fire regimes. I assessing landscape condition in used the more refined NL, MS1, terms of vegetation composition Schmidt and others (2002) devel and MS2 fire regimes to reflect and structure, potential fire oped a broad classification in order those gradations. severity, and other variables. They to make assessments on a national • Similarly, fire regimes III, IV, identified five different fire scale. By contrast, my classification and V are too broad to adequate regimes: is tailored to a regional scale— ly reflect natural variation at specifically, to the Northern higher elevations in the • Fire regime I: 0- to 35-year Rockies. Differences therefore exist Northern Rockies, especially mean fire interval (MFI), low- between our classifications: near upper treeline. I defined severity fires. the MS3, SR1, and SR2 fire • Fire regime II: 0- to 35-year • Fire regime II is found primarily regimes to reflect those condi MFI, stand-replacement fires. in grass- and shrublands tions. • Fire regime III: 35- to 100-year (Schmidt and others 2002) and is MFI or greater, mixed-severity thus not covered by my Northern fires. Rockies classification for forests. data are from ponderosa-pine-dom Table 1—Northern Rockies database: number of plot samples and inated stands, for example, in east descriptive statistics for historical fire frequency, by fire regime type. ern Oregon (Heyerdahl 1997), cen Median tral Idaho (Barrett 1988, 2000), and fire Mean fire Standard 10th 90th west-central Montana (Arno 1976; Fire interval interval error percentile percentile Arno and others 1995; Gruell and regime a Number b years others 1982). East of the Continental Divide, the NL regime NL 407 16.00 17.11 0.31 10 26 also occurs in open-grown Douglas- MS1 216 32.50 32.13 0.49 23 41 fir stands bordering intermountain MS2 334 65.50 73.22 1.46 43 117 valleys (Arno and Gruell 1986; MS3 36 122.00 135.22 13.82 47 275 Bakeman 1983; Barrett 1997). SR1 288 129.00 133.41 2.00 96 180 SR2 159 220.00 244.19 4.11 200 325 MS1 (25- to 40-year MFI). The a. NL = nonlethal; MS1 = short-interval mixed severity; MS2 = moderate- to long-interval MS1 fire regime essentially is a mixed severity; MS3 = variable-interval mixed severity; SR1 = moderate- to long-interval variant of the NL regime, but with stand replacement; SR2 = long-interval stand replacement. longer fire intervals that occasion b. Out of a total of 1,440 plot samples from 95 fire histories (fig. 1). ally promote locally severe burning (Arno and others 1997; Barrett and even-aged patches (unlike the pri- types as the NL regime, and stand others 1991). For example, such marily uneven-aged NL stands). I structures and species composition fires might kill up to 30 percent of classified about 15 percent of the are much more variable. West of the overstory trees, in a highly database as MS1, where MFIs the Continental Divide, MS1 occurs patchy pattern (fig. 2) (Arno and ranged from about 25 to 40 years on relatively productive montane others 1997; Barrett and others long and averaged 32 years. sites, usually in warm–dry forests 1991). The oldest trees generally such as dry Douglas-fir and dry have from 3 to 10 fire scars each, The MS1 regime contains nearly grand fir. Such stands often are and the stands often contain small twice as many potential vegetation codominated by various combina-
Volume 64 • No. 2 • Spring 2004 33 Barrett andothers1991)but can (fig. 2)(Arnoandothers1993; from 50to100percentof thestand MS1 regime.Suchfiresoften kill erally moreseverefiresthanthe tially longerfireintervalsandgen MS2 fireregimecontainssubstan MS2 (40-to120-yearMFI). page 35). Gruell 1985)(seethesidebaron Indians (BarrettandArno1982; have beenfrequentlyburnedby often occurinareasthatmight As withtheNLregime,MS1stands 1983; Barrett1996;Houston1973). mountain valleys(ArnoandGruell northern GreatPlainsandinter- and lodgepolepineborderingthe stands dominatedbyDouglas-fir example, theMS1regimeoccursin widely varyingfireresistance.For subalpine zone,wherestandshave adjacent cold–drysitesinthelower cool–dry montanesitesandon ern Idaho.Suchburningoccurson the ContinentalDivideandineast MS firesalsoarecommoneastof and others1991). size (Arnoandothers1997;Barrett patches lessthan2.5acres(1ha)in cohorts, generatingeven-aged ally producedmultipleseral Lethal fireseveritylocallyhasusu 1997; Barrettandothers1991). pine (Arno2000;Arnoandothers larch, Douglas-fir, andlodgepole tions ofponderosapine,western SR2 =long-intervalstandreplacement. variable-interval mixedseverity;SR1=moderate-tolong-intervalstandreplacement; = short-intervalmixedseverity;MS2moderate-tolong-intervalMS3 Figure 2—Hypothesizedpostfirestandmortalitybyfireregimetype.NL=nonlethal;MS1
34 Fire Regimes MS1 MS2 MS3 SR1 SR2 NL 0 20 The Percent Mortality 40 1994; ZackandMorgan1994). (Barrett 1982;Brownand others redcedar andgrandfirhabitat types northern Idaho’s uplandwestern mixed-conifer stands,suchasin MS2 regimealsoincludesmany Douglas-fir, andsubalpinefir).The ern hemlock,moistgrandfir tivity (e.g.,westernredcedar–west in foresttypesofvaryingfiresensi relatively productiveorsteepsites others 1991).Moststandsoccuron (Arno andothers1993;Barrett 5 to100acres(2–40ha)insize well-defined patchesrangingfrom one tothreeseralcohorts,oftenin 1991). Standsusuallycontainfrom and others1993;Barrett than theMS1orSRtypes(Arno more diverseageclassmosaics The MS2fireregimecanpromote 73 years. 40 to120yearslongandaveraged where MFIsgenerallyrangedfrom percent ofthesamplesasMS2, fire scarseach.Iclassifiedabout25 trees rarelyhavemorethanthree also burnatlowseverities.Old forested areawithlow- Over thepastcentury, has declinedbymore severity firepotential than 80percent. 60 0 100 80 , moist whitebark pine,subalpinefir stands typicallyaredominatedby upper subalpinezone),where near uppertreeline(i.e.,inthe MFI). MS3 FireRegime(50-to275-year zones. montane andlowersubalpine 1997, 1999;Pierce1995)inthe pine (ArnoandGruell1986;Barrett nated byDouglas-firorlodgepole ly productiveorsteepterraindomi regime oftenisfoundonmoderate in easternIdaho,theMS2fire East oftheContinentalDivideand 135 years).However about 50to275yearslong(mean: estimated fireintervalsrangedfrom base containsonly36plots,andthe studies todatearesparse.Thedata cult, andthedatafromfew upper treelineisinherentlydiffi Interpreting firefrequencynear weather. meadows, andhighlyvariablefire ers suchasrocklandsandwet arrangements, extensivefirebarri widely varyingfuelloadsandspatial dently resultsfromsuchfactorsas variability intheMS3regimeevi Morgan andBunting1990).High 1996; Keaneandothers1990; terns varywidely(fig.2)(Barrett cy, sizes,severities,andspreadpat alpine larch.Historicalfirefrequen formly thanintheMSregimes to belargerandspreadmore uni (Brown 2000).Suchfires also tend ly killmosttreesinastand (fig.2) MFI). SR1 FireRegime(100-to180-year vague term“highlyvariable.” patterns iswiththeadmittedly best waytocharacterizeMS3fire and Bunting1990).Soperhapsthe (Fischer andClayton1983;Morgan involve justoneortwotrees ited meaningbecausefiresoften “stand firefrequency”hasonlylim The MS3fireregimeoccurs Stand-replacing firestypical Fire Management Today , theconceptof , and (Barrett 1996; Barrett and others Case Study: Indian-Influenced 1991). As a result, stands usually are dominated by a single seral Fire Regimes cohort, fire-scarred trees are Fire regimes can be difficult to Grove near Seeley Lake, MT. My uncommon, and patch sizes often differentiate. For example, evi- Northern Rockies database sug exceed 1,000 acres (400 ha) (Arno dence of a “stand-replacing” fire gested that such habitat types 2000). might simply reflect locally heavy generally are dominated by rela mortality during the last mixed- tively severe fire regimes (e.g., The data suggested two SR fire severity fire. Other factors that > 100-year fire return intervals). regimes for the Northern Rockies. can hinder classification include Conversely, Arno and others For example, I classified about 20 sparse or skewed data; sampling (1997) found frequent low-severi percent of the samples as occurring errors; natural or topographical ty fires at the site, with a mean in the moderately long-interval SR variation; and anthropogenic fire interval of just 24 years. That regime (SR1), where fire intervals influences, such as fire use by primeval Indian campsite and ranged from about 100 to 180 years American Indians (Arno and oth- adjacent travel corridor evidently long and averaged 133 years. ers 1997; Barrett and Arno 1982; had been heavily used and repeat- Gruell 1985). edly burned for centuries (Ayres SR fire regimes often occur on pro 1901; Barrett and Arno 1982)— ductive or steep terrain. Eighty per An example of the latter occurred indirect evidence of human- cent of the samples classified as in a moist subalpine fir habitat altered landscapes and fire SR1 were in the lower and upper (potential climax) type at Girard regimes in an earlier era. subalpine forest zones, where stands usually are dominated by trees with moderate to high fire sensitivity (such as lodgepole pine, During the presettlement era, stands in the subalpine fir, and Engelmann nonlethal fire regime experienced frequent spruce). Biophysical differences understory fires that promoted lightly stocked, between SR1 stands and the long- interval SR2 stands (see below) are uneven-aged structures. not always readily apparent. However, many SR1 stands are jux ern hemlock (Arno and Davis 1980; In those areas, site MFIs average taposed with stands in the MS Barrett 1993) and moist subalpine about 350 years and 200 years, regimes, whereas SR2 stands gener fir forest (Barrett 1994; Barrett and respectively. ally are not (Barrett and others others 1991; Tande 1979). Con 1991; Brown and others 1994; versely, evidence of the SR2 regime Hawkes 1979). Management has also been found on highly Implications unproductive terrain, such as in cli SR2 Fire Regime (200- to 325-year Fire regimes classifications by max lodgepole pine stands on the MFI). The SR2 regime exhibits Brown (2000) and Hardy and others Yellowstone Plateau (Romme 1982), substantially longer fire intervals (1998) are useful for national-scale where long fire-free intervals might than SR1, but other fire character work, whereas this article presents be necessary to develop sufficient istics are similar, including fire a more refined, empirically based fuel for stand-replacing fires. sizes, spread patterns, and postfire system for the Northern Rockies. mortality levels (fig. 2). I classified This system helped support terrain Stands in the SR2 regime often are about 10 percent of the samples as modeling of historical and current juxtaposed with SR1 stands rather the SR2 fire regime, where stand fire regimes for land management than with more frequently burned MFIs ranged from about 200 to 325 planning (Jones and others 2002). terrain. For instance, the lodgepole years long and averaged 244 years. Specifically, the modeling results pine stands on the Yellowstone suggested that low-severity fire Plateau (Romme 1982) adjoin seral The SR2 regime often occurs on regimes (NL and MS1) occupied lodgepole pine stands in the steep highly productive terrain, for exam about 35 percent, moderate-severity Absaroka Mountains (Barrett 1994). ple in wet western redcedar–west- fire regimes (MS2 and MS3) occu-
Volume 64 • No. 2 • Spring 2004 35 pied about 40 percent, and high- The mixed-severity fire regimes contain more severity fire regimes (SR1 and SR2) potential vegetation types than the nonlethal occupied about 25 percent of the forested terrain (fig. 3). regime, and stand structures as well as species composition are much more variable. Over the past century, forested area with low-severity fire potential has western fire regimes. Northwest Science. declined by more than 80 percent ties, and other factors (Morgan and others 2001). Given the complexity 72(4): 24–34. due to long-term fire exclusion, Arno, S.F. 1976. The historical role of fire inappropriate logging, and other of forested ecosystems, manage on the Bitterroot National Forest. Gen. activities. These results generally ment decisions at all scales, from Tech. Rep. INT–187. Ogden, UT: USDA national-level planning to site-spe Forest Service, Intermountain Forest and agree with the Interior Columbia Range Experiment Station. Basin Ecosystem Management cific treatments, require such Arno, S.F. 2000. Fire in western forest Project interpretations for detailed information. ecosystems. In: Brown, J.K.; Smith, J.K., eds. Wildland fire in ecosystems: Effects Northern Rockies forests (Morgan of fire on flora. Gen. Tech. Rep. and others 1998; Quigley and oth For more information, please con RMRS–GTR–42, vol. 2. Ogden, UT: USDA ers 1996). tact Stephen W. Barrett, 995 Ranch Forest Service, Rocky Mountain Research Lane, Kalispell, MT 59901, 406-756 Station: 97–120. 9547 (tel.), [email protected] Arno, S.F.; Davis, D.H. 1980. Fire history of Knowing the current status of fire western redcedar/hemlock forests in regimes is critical for planning. For (e-mail). northern Idaho. In: Stokes, M.A.; Dieterich, J.H., tech. coords. Proceedings example, fire regimes models can of the Fire History Workshop. Gen. Tech. be used in conjunction with other Acknowledgments Rep. RM–81. Fort Collins, CO: USDA data in a geographic information The author thanks Flathead Forest Service, Rocky Mountain Forest system to help design and prioritize National Forest Landscape and Range Experiment Station: 21–26. Arno, S.F.; Gruell, G.E. 1983. Fire history at strategies for forest restoration, Ecologist J.L. Jones for invaluable the forest–grassland ecotone in south wildland fire planning, and habitat assistance and support during this western Montana. Journal of Range protection for threatened and project. S.F. Arno, J.K. Brown, and Management.. 36(3): 332–336. Arno, S.F.; Gruell, G.E. 1986. Douglas-fir endangered species. Legislators, E.K. Heyerdahl also provided encroachment into mountain grasslands policymakers, and forest managers detailed reviews that improved the in southwestern Montana. Journal of might also use such information in report on which this article is Range Management. 39(3): 272–276. based. Arno, S.F.; Peterson, T.D. 1983. Variation in developing funding priorities for estimates of fire intervals: A closer look at Federal lands and management fire history on the Bitterroot National units. References Agee, J.K. 1998. The landscape ecology of Land managers are increasingly SR2 13% NL 13% focusing on ecosystems rather than on individual stands. Therefore, fire regimes sampling at multiple scales would foster a better understanding SR1 14% of the varying roles of disturbance MS1 21% (Lertzman and others 1998; Morgan and others 2001). Land- MS3 2% scape-scale research, particularly in areas dominated by the inherently complex MS fire regimes (Agee 1998; Arno and others 2000; Rollins and others 2001), would further MS2 37% refine our understanding of fire Figure 3—Percent of forested terrain by fire regime type in the general study area (Jones regimes by revealing the influence and others 2002). NL = nonlethal; MS1 = short-interval mixed severity; MS2 = moderate- of macroclimate, microclimate, to long-interval mixed severity; MS3 = variable-interval mixed severity; SR1 = moderate- topography, anthropogenic activi to long-interval stand replacement; SR2 = long-interval stand replacement.
Fire Management Today 36 Forest. Res. Pap. INT–301. Ogden, UT: In the stand replacement regime, stands usually USDA Forest Service, Intermountain Forest and Range Experiment Station. are dominated by a single seral cohort, fire- Arno, S.F.; Reinhardt, E.D.; Scott, J.H. scarred trees are uncommon, and patch sizes 1993. Forest structure and landscape pat terns in the subalpine lodgepole pine often exceed 1,000 acres. type: A procedure for quantifying past and present conditions. Gen. Tech. Rep. INT–294. Ogden, UT: USDA Forest Service, Intermountain Forest and Range Management Division, Environment forest in western Montana: Implications Experiment Station. Canada, Canadian Parks Service Waterton for resource management. Gen. Tech. Arno, S.F.; Scott, J.H.; Hartwell, M.G. 1995. Lakes National Park, Waterton Townsite, Rep. INT–130. Ogden, UT: USDA Forest Age-class structure of old growth pon Alberta. Service, Intermountain Forest and Range derosa pine/Douglas-fir stands and its Barrett, S.W. 1997. Fire regimes on the Experiment Station. relationship to fire history. Res. Pap. Beaverhead–Deerlodge National Forest Hardy, C.C.; Menakis, J.P.; Long, D.G.; INT–481. Ogden, UT: USDA Forest (Beaverhead Portion). Unpublished report Brown, J.K.; Bunnell, D.L. 1998. Mapping Service, Intermountain Research Station. on file, Fire Management Division, USDA historic fire regimes for the Western Arno, S.F.; Smith, H.Y.; Krebs, M.A. 1997. Forest Service Beaverhead–Deerlodge United States: Integrating remote sensing Old growth ponderosa pine and western National Forest, Wise River Ranger and biophysical data. In: Proceedings of larch stand structures: Influences of pre District, Wise River, MT. the 7th Biennial Forest Service Remote 1900 fire regimes and fire exclusion. Res. Barrett, S.W. 1999. Fire regimes on the Sensing Applications Conference; 6–9 Pap. INT–RP–495. Ogden, UT: USDA Pine Butte Swamp Preserve, Montana. April 1998; Nassau Bay, TX. Bethesda, Forest Service, Intermountain Research Unpublished report on file at The Nature MD: American Society for Station. Conservancy, Pine Butte Swamp Preserve, Photogrammetry and Remote Sensing: Arno, S.F.; Parsons, D.J.; Keane, R.E. 2000. Choteau, MT. 288–300. Mixed-severity fire regimes in the Barrett, S.W. 2002. A fire regimes classifica Hawkes, B.C. 1979. Fire history and fuel Northern Rocky Mountains: tion for northern Rocky Mountain appraisal study of Kananaskis Provincial Consequences of fire exclusion and forests: Results from three decades of fire Park. Thesis. University of Alberta, options for the future. In: Cole, D.N.; history research. Contract final report on Edmonton, Alberta. McCool, S.F.; Borrie, W.T.; O’Laughlin, J., file, Planning Division, USDA Forest Heyerdahl, E.K. 1997. Spatial and temporal comps. Wilderness ecosystems, threats, Service, Flathead National Forest, variation in historical fire regimes of the and management. Proceedings, Kalispell, MT. Blue Mountains, Oregon and Washington: Wilderness Science in a Time of Change Barrett, S.W.; Arno, S.F. 1982. Indian fires The influence of climate. Ph.D. Diss. Conference; 1999 May 23–27; Missoula, as an ecological influence in the University of Washington, Seattle, WA. MT. RMRS–P–15–VOL–5. Ogden, UT: Northern Rockies. Journal of Forestry. Houston, D.B. 1973. Wildfires in northern USDA Forest Service, Rocky Mountain 80(10): 647–651. Yellowstone National Park. Ecology. 54: Research Station: vol. 5, 225–232. Barrett, S.W.; Arno, S.F.; Key, C.H. 1991. 1111–1117. Ayres, H.B. 1901. The Lewis and Clark Fire regimes of western larch–lodgepole Jones, J.L.; Barrett, S.W.; Berglund, D. Forest Reserve, Montana. In: 21st Ann. pine forests in Glacier National Park, 2002. Modeling historical and current fire Rept., part V, 1899–1900. Washington, Montana. Canadian Journal of Forest regimes for major vegetation types in DC: USDI Geological Survey: 35–80. Research. 21: 1711–1720. northern Idaho and western Montana: A Bakeman, M.E. 1983. The genesis of mol Brown, J.K. 2000. Introduction and fire first approximation. Unpublished report lisols under Douglas-fir. Thesis. regimes. In: Brown, J.K.; Smith, J.K., eds. on file, Planning Division, USDA Forest University of Montana, Missoula, MT. Wildland fire in ecosystems: Effects of fire Service, Flathead National Forest, Barrett, S.W. 1982. Fire’s influence on on flora. Gen. Tech. Rep. RMRS–GTR–42, Kalispell, MT. ecosystems of the Clearwater National vol. 2. Ogden, UT: USDA Forest Service, Keane, R.E.; Arno, S.F.; Brown, J.K.; Forest: Cook Mountain fire history inven Rocky Mountain Research Station: 1–7. Tomback, D.F. 1990. Modeling stand tory. Unpublished report on file, Fire Brown, J.K.; Arno, S.F.; Barrett, S.W.; dynamics in whitebark pine (Pinus albi Management, USDA Forest Service, Menakis, J.P. 1994. Comparing the pre caulis) forests. Ecological Modeling. 51: Clearwater National Forest, Orofino, ID. scribed natural fire program with preset 73–95. Barrett, S.W. 1988. Fire suppression’s tlement fires in the Selway–Bitterroot Lertzman, K.; Fall, J.; Dorner, B. 1998. effects on forest succession within a cen Wilderness. International Journal of Three kinds of heterogeneity in fire tral Idaho wilderness. Western Journal of Wildland Fire. 4(3): 157–168. regimes: At the crossroads of fire history Applied Forestry. 3(3): 76–80. Cissel, J.H.; Swanson, F.J.; Weisberg, P.J. and landscape ecology. Northwest Barrett, S.W. 1993. Fire regimes on the 1999. Landscape management using his Science. 72: 4–23. Clearwater and Nez Perce National torical fire regimes: Blue River, Oregon. Morgan, P.; Bunting, S.C. 1990. Fire effects Forests, north-central Idaho. Unpublished Ecological Applications. 9(4): 1217–1231. in whitebark pine forests. In: Schmidt, report on file, Fire Management Division, Fischer, W.C.; Clayton, B.D. 1983. Fire ecol W.C.; McDonald, K.J., comps. USDA Forest Service, Nez Perce National ogy of Montana forest habitat types east Proceedings, Symposium on Whitebark Forest, Grangeville, ID. of the Continental Divide. Gen. Tech. Pine Ecosystems: Ecology and Barrett, S.W. 1994. Fire regimes on Rep. INT –141. Ogden, UT: USDA Forest Management of a High-Mountain andesitic mountain terrain in northeast Service, Intermountain Research Station. Resource. Gen. Tech. Rep. INT–270. ern Yellowstone National Park. Gruell, G.E. 1985. Fire on the early western Ogden, UT: USDA Forest Service, International Journal of Wildland Fire. 4: landscape: An annotated record of wild- Intermountain Research Station: 65–76. land fires, 1776–1900. Northwest Science. 166–170. Barrett, S.W. 1996. The historical role of 59(2): 97–107. Morgan, P.; Aplet, G.H.; Haufler, J.B.; fire in Waterton Lakes National Park, Gruell, G.E.; Schmidt, W.C.; Arno, S.F.; Humphries, H.C.; Moore, M.M.; Wilson, Alberta. Unpublished report on file, Fire Reich, W.F. 1982. Seventy years of vegeta D.W. 1994. Historical range of variability: tion change in a managed ponderosa pine
Volume 64 • No. 2 • Spring 2004 37 A useful tool for evaluating ecosystem Rollins, R.G.; Swetnam, T.W.; Morgan, P. Jasper National Park, Alberta. Canadian change. Journal of Sustainable Forestry. 2001. Evaluating a century of fire pat Journal of Botany. 57: 1912–1931. 2: 87–111. terns in two Rocky Mountain wilderness USDA (U.S. Department of Agriculture). Morgan, P.; Bunting, S.C.; Black, A.E.; areas using digital fire atlases. Canadian 1999. Sustaining the people’s lands: Merrill, T.; Barrett, S.W. 1998. Past and Journal of Forest Research. 31: Recommendations for stewardship of the present fire regimes in the Interior 2107–2123. national forests and grasslands into the Columbia River Basin. In: Close, K.; Romme, W.H. 1980. Fire history terminolo next century. Report by the Committee of Bartlette, R.A., eds. Proceedings, Fire gy: Report of the ad hoc committee. In: Scientists. Washington, DC: USDA Forest Management Under Fire: Adapting to Stokes, M.A.; Dieterich, J.H., tech. coords. Service. Change. Fairfield, WA: International Proceedings of the Fire History USDA (U.S. Department of Agriculture). Association of Wildland Fire: 77–82. Workshop. Gen. Tech. Rep. RM–81. Fort 2000a. National Forest System land Morgan, P.; Hardy, C.C.; Swetnam, T.W.; Collins, CO: USDA Forest Service, Rocky resource management planning: Final Rollins, M.G.; Long, D.G. 2001. Mapping Mountain Forest and Range Experiment rule. Fed. Reg. 36 CFR, parts 217 and fire regimes across time and space: Station: 135–137. 219. Washington, DC: USDA Forest Understanding coarse- and fine-scale pat Romme, W.H. 1982. Fire and landscape Service. terns. International Journal of Wildland diversity in subalpine forests of USDA (U.S. Department of Agriculture). Fire. 10: 329–342. Yellowstone National Park. Ecological 2000b. Protecting people and sustaining Pierce, J.R. 1995. Fire history data from the Monographs. 52(2): 199–221. resources in fire-adapted ecosystems—A Cutreef–Willow Creeks area, Montana. Schmidt, K.M.; Menakis, J.P.; Hardy, C.C.; cohesive strategy. Washington, DC: USDA Unpublished report on file, Planning Hann, W.J.; Bunnell, D.L. 2002. Forest Service. Division, USDA Forest Service, Northern Development of coarse-scale spatial data Zack, A.C.; Morgan, P. 1994. Fire history on Region, Missoula, MT. for wildland fire and fuel management. the Idaho Panhandle National Forest. Quigley, T.M.; Haynes, R.W.; Graham, R.T., Gen. Tech. Rep. RMRS–GTR–87. Fort Unpublished report on file in the tech. eds. 1996. Integrated scientific Collins, CO: USDA Forest Service, Rocky Planning Division, USDA Forest Service, assessment for ecosystem management in Mountain Research Station. Coeur d’Alene National Forest, Coeur the Interior Columbia Basin and portions Stokes, M.A.; Smiley, T.L. 1968. An intro d’Alene, ID. ■ of the Klamath and Great Basins. Gen. duction to tree ring dating. Chicago, IL: Tech. Rep. PNW–GTR–382. Portland, OR: University of Chicago Press. USDA Forest Service, Pacific Northwest Tande, G.F. 1979. Fire history and vegeta Research Station. tion pattern of coniferous forests in
WEBSITES ON FIRE* Fire Consortia for Environmental Protection Agency, northeastern regions of the Advanced Modeling and universities. Located in East United States. Participants—the of Meteorology and Lansing, MI; Athens, GA; Fort USDA Forest Service’s North Smoke Collins, CO; Riverside, CA; and Central and Northeastern Seattle, WA, each consortium is Research Stations, the Eastern In support of the National Fire conducting research on fire/atmos Region of the National Forest Plan, Federal and State land phere interactions and developing System, the Forest Service’s management agencies have cre improved predictive models and Northeastern Area State and ated a national framework of decision support tools for the fire Private Forestry, and the regional modeling consortia. management community. The Interagency Eastern Area Fire Consortia for Advanced FCAMMS Website displays a nation Coordination Center—work Modeling of Meteorology and al map that links to the consortium together to address the need for Smoke (FCAMMS) members sites where you can learn about better predictions and decision include the National Oceanic and their research and development support tools for fire and air Atmospheric Administration, the objectives, relevant meetings and quality management. Site visi National Weather Service, the presentations, and current projects tors can quickly view real-time and products. fire weather model maps for the * Occasionally, Fire Management Today briefly contiguous United States, the describes Websites brought to our attention by the wildland fire community. Readers should not con- The Eastern Area Modeling North-Central and Northeastern strue the description of these sites as in any way exhaustive or as an official endorsement by the Consortium (EAMC), in East States, New England, and the USDA Forest Service. To have a Website described, Lansing, MI, one of the five consor- Lake States. contact the managing editor, Hutch Brown, at USDA Forest Service, 202-205-0878 (tel.), tium sites, addresses fire weather, [email protected] (e-mail). fire behavior, and smoke transport Found at issues for the north-central and
Fire Management Today 38 KEETCH–BYRAM DROUGHT INDEX: CAN IT HELP PREDICT WILDLAND FIRES?
Daniel W. Chan, James T. Paul, and Alan Dozier
he Georgia Forestry Commission uses the Georgia’s typical fire season from 1957 to 2000 T Keetch–Byram Drought Index ran from February through April—when the (KBDI) (Keetch and Byram 1968) Keetch–Byram Drought Index was lowest. to determine potential wildland fire hazards. (For an overview of KBDI, see the sidebar.) The objectives of Airport, Macon Regional Airport, Daily records for the 24-hour peri our study were to better under and Savannah International Airport od ending at 1 p.m. and daily stand the relationship between from 1957 to1995. From these data, records ending at midnight can KBDI and fire activities in Georgia we constructed a fire-weather-type yield different maximum tempera and to evaluate KBDI computed observation for both 1 p.m. and ture and rainfall data for the previ from National Weather Service midnight. Then we used the data to ous 24-hour period. If a heavy rain (NWS) observational data compared calculate a KBDI for the two incident occurred after 1 p.m., the with KBDI computed from fire defined observation times. KBDI numbers from 1 p.m. and weather observations. What We Did What is the Keech–Byram Traditionally, fire weather observa tions for determining wildland fire Drought Index? hazards are recorded at 1 p.m. According to Melton (1989), the If a location has been dry during daily. This means that the maxi Keetch–Byram Drought Index the previous 24 hours, the KBDI mum temperature recorded at this (KBDI) is an index based “on a will increase, depending on the time usually occurs during the pre measurement of 8 inches (0.2 m) maximum temperature in the vious day’s afternoon hours. Like of available moisture in the upper previous 24 hours, the previous wise, the 24-hour precipitation soil layers that can be used by day’s index, and the annual rain recorded is from 1 p.m. on the pre vegetation for evapotranspiration. fall amount at that location. vious day until 1 p.m. on the pres Generally, high temperature and ent day. By contrast, the NWS The index measure is in hun a low KBDI mean big increments. reports maximum temperature and dredths of an inch of water and 24-hour precipitation for the 24 has a range of 0 to 800, with 0 When an area has received rain hour period ending at midnight. being saturated and 800 repre during the previous 24 hours, the senting the worst drought condi index changes, depending on the To compare NWS data to traditional tion. The index indicates deficit rain-adjusted KBDI—for each fire weather data, we used NWS inches of available water in the 0.01 inches (0.03 cm) of net rain hourly data from Athens Municipal soil. A K/B reading of 250 means fall, one point is subtracted from there is a deficit of 2.5 inches the previous day’s index—the Daniel Chan is a meteorologist for the (6.4 cm) of ground water avail maximum temperature, and Georgia Forestry Commission, Macon, GA; able to the vegetation. As drought annual rainfall amount at that James Paul is the President and Chief progresses, there is more avail location. ® Scientist for SCITRAN , Inc., Gray, GA; and able fuel that can contribute to Alan Dozier is the Chief of Forest Protection for the Georgia Forestry fire intensity.” Commission, Macon, GA.
Volume 64 • No. 2 • Spring 2004 39 In Georgia, the Keetch–Byram Drought Index alone is not a good indicator for fire activity. midnight could yield a difference of a hundred points or more. Al though KBDI computed from 1 p.m. averaged higher than KBDI computed from the midnight data, we decided that the midnight KBDI is a good approximation of the 1 p.m. KBDI (fig. 1).
Figure 1—Average Keetch–Byram Drought Index (KBDI) climatology at Athens, GA, for After determining that the mid midnight (KBDI00) and 1 p.m. (KPDI13), 1957–95. There is no significant difference. night and 1 p.m. indexes were com parable, we obtained daily weather data from selected NWS cooperative weather stations in Georgia between 1950 and 2001 through Georgia’s State Climatologist Office. These cooperative daily sta tions record data once each day for maximum and minimum tempera tures and 24-hour precipitation. Although observation times vary at these stations, our analysis of the NWS hourly data indicated minimal differences between a 1 p.m. and midnight observation time. Based on this analysis, we assumed that the differences due to varying observations at the cooperative sta tions would also be minimal, and we chose 14 stations across Georgia (fig. 2). What We Found Average quarterly temperatures from 1961 to 1990 at the stations we sampled are shown in figure 3. The hottest months were June, July, and August at all stations. The Figure 2—The 14 National Weather Service cooperative weather stations chosen for the stations in northern Georgia had KBDI study, located across Georgia. lower temperatures than those in Average quarterly rainfall from March are the rainy season. But at the south. The temperature differ 1961 to 1990 for the selected sta- stations near the coast, such as ence between the hottest and the tions is shown in figure 4. At sta- Savannah, Waycross, Brunswick, coolest stations was less than 10 °F tions in northern and central and Alma, the rainy months are (5.5 °C). Georgia, January, February, and June, July, and August. We found
Fire Management Today 40 Wildland fire incidents in Georgia followed the burn seasons of spring and fall rather than the high Keetch–Byram Drought Index months of summer. that KBDI was low at coastal sta tions during June, July, and August due to high rainfall levels in this region. Conversely, when rainfall was low at coastal stations during January, February, and March, KBDI was higher. Therefore, the seasonal range of KBDI at coastal Figure 3—Average quarterly temperature at 14 National Weather Service cooperative stations was smaller than at sta weather stations, 1961–90. tions located elsewhere (fig. 5).
In Georgia, KBDI is typically lowest in February or March and highest in August. However, the number of fires and acres burned from 1957 to 2000 suggested just the opposite (fig. 6). Georgia’s typical fire season for those years ran from February through April, when KBDI was low est. Fire activities then dropped off gradually from May through September and picked up again in October as KBDI began to drop. Obviously, other factors besides KDBI were influencing the fire episodes.
According to fire reports collected Figure 4—Average quarterly rainfall at 14 National Weather Service cooperative weather by the Georgia Forestry Commis stations, 1961–90. sion, we found that human activi ties caused 95 percent of wildland fires in Georgia from 1957 to 2000 There were exceptions to a low activities in southeastern Georgia (fig. 7). Almost half of the fires KBDI during the spring. In 2001, burned almost 16,000 acres (6,500 were caused by outdoor burning, KBDI at Waycross in southeastern ha). More than 360 people were which is mostly done in spring and Georgia rose steadily from 100 in involved in the firefighting effort, autumn. Therefore, wildland fire April to more than 600 by the end and about $1 million was spent to incidents corresponded to the of May (fig. 8), which is more than control the dangerous fires. burning seasons—especially 150 points above normal. This find spring—rather than the high KBDI ing suggests that fuels were very What We Think months of summer. dry and fires could be difficult to All this information helped us to control—which was exactly what conclude that KBDI alone is not a happened. In late May, severe fire good indicator for fire activities. It
Volume 64 • No. 2 • Spring 2004 41 Figure 5—Average KBDI climatology at Georgia stations on the Figure 7—Causes of wildland fires in Georgia, 1992–2001. coast (Brunswick) and inland (Gainesville), 1957–95. The seasonal range of KBDI is smaller on the coast.
Figure 6—Average number of wildland fires and acres burned in Figure 8—KBDI values for Waycross, GA, from April to June 2001 Georgia by month, from January (1) to December (12), differed greatly from normal, suggesting that fuels were much 1957–2000. drier than usual and fires could be more difficult to control. should be used only in conjunction with other reliable References sources of information to predict wildland fires. Keetch, J.J.; Byram, G.M. 1968. A drought index for forest fire con However, a higher-than-normal index or a sustained trol. Res. Pap. SE–38. Asheville, NC: USDA Forest Service, rise in the index could mean that the potential for Southeastern Forest Experiment Station. Melton, M. 1990. The Keetch/Byram Drought Index: A guide to fire wildland fire is high. conditions and suppression problems. Fire Management Notes. 50(4): 30–34. ■ Knowing how KBDI varied across Georgia could be helpful to fire managers when planning resource allocation. Managers should be alert to a potential fire hazard when KBDI is higher than normal during summer.
Fire Management Today 42 FIRE PREVENTION TEAM SHOWS ITS WORTH IN GEORGIA
James T. Paul, Daniel Chan, and Alan Dozier
mokey Bear’s familiar message, “Only you can prevent wild- Short-term intensive fire prevention efforts during S land fires,” comes from one of times of high fire danger can make people more the oldest fire prevention programs cautious if they become aware that the situation in the nation. Most forestry organi zations have active fire prevention is critical. programs, largely centered around long-term education. Less common clearing, burning brush, weeds, The GFC also decided, for the first are short-term efforts to get out grass, trash, garbage, etc.” (GFC time in its history, to put a fire pre fire prevention messages through Fire Staff 1996). vention team to the test. The Rome the media when fire potential is District (district 1) in Georgia’s high. Such efforts can make people Finally, on November 11, 2001, the northwestern corner (fig. 1) was more cautious by making them GFC announced a ban on all out chosen as the site due to its histori aware that the situation is critical. door burning. It was an unusual cally high number of escaped move; the GFC seldom bans burn debris-burning fires (table 1). The In 1999 and 2000, the group of Fire ing, in part due to the beneficial team was dispatched to the Rome Chiefs in the Southern States pro uses of prescribed fire (Moorman District on November 9, 2001. moted the formation of specially 2001; Wade and Lunsford 1989). trained wildland fire prevention teams.* Joining the effort, Georgia trained 10 people in wildland fire prevention. Put to the Test In fall 2001, fire potential was increasing in northern Georgia and adjacent states due to a continuing drought and severe burning condi tions. Districts administered by the Georgia Forestry Commission (GFC) were cautiously issuing per mits for debris burning on a case by-case basis. A debris fire is “any fire intentionally set for any pur pose other than campfire or incendiary [burn] such as land
James Paul is President and Chief Scientist of SCITRAN TM, Inc., Gray, GA; Daniel Chan is a meteorologist for the Georgia Forestry Commission, Macon, GA; and Alan Dozier is the Chief of Forest Protection, Georgia Forestry Commission, Macon, GA.
* For more on Cooperative Wildland Fire Prevention/Education Teams, see Judith W. Kissinger, “Interagency Teams Prevent Fires From Alaska to Florida,” Fire Management Notes 59(4): 13–17. Figure 1—Districts administered by the Georgia Forestry Commission.
Volume 64 • No. 2 • Spring 2004 43 The team had a twofold purpose Table 1—Average annual number and acres of escaped debris-burning (Lane and others 2001): fires in districts administered by the Georgia Forestry Commission, 1990–2000. • To raise public awareness about District Escaped fires the increasing fire danger throughout northwestern Number Name Number Acres burned Georgia while providing basic fire 1 Rome 522 1,387 prevention information for people 2 Gainesville 240 460 to use in their daily activities, and 3 Athens 157 436 • To communicate the urgency for residents to respond preventative 4 Newnan 296 883 ly to the severe fire conditions. 5 Milledgeville 318 1,052 6 Washington 162 768 7 Americus 252 1,070 In fall 2001, the 8 Tifton 432 1,901 Georgia Forestry 9 Camilla 369 1,270 Commission mobilized a 10 Statesboro 446 1,658 fire prevention team— 11 McRae 408 1,134 for the first time in its 12 Waycross 390 3,145 history. Note: Districts 1–4, covering northern Georgia, were used in this study. The other districts are shown here only for comparison purposes.
Flurry of Activity The team sought to accomplish its purpose through media contacts distributed throughout the Rome District (fig. 2). But that wasn’t all. The team’s flurry of activity from November 9 to November 21, when the next rain fell, was summarized by Lane and others (2001):
• 314 personal contacts with hand outs for players and spectators at the State soccer championship; • 47 door-to-door contacts with handouts in the Cherokee County wildland/urban interface; • 39 phone calls with prevention messages; • Daily faxes and e-mails with fire prevention messages, included with the daily fire situation update from the Rome Severity Information Center, to— – 5 television stations, – 19 radio stations, and – 25 newspapers. • Prevention messages to the Rome District office and county unit Figure 2—Media contacts in the Rome District for the wildland fire prevention team.
Fire Management Today 44 dispatchers to help in handling callers; • Visits to 38 radio stations, includ ing— – 30 contacts, – 9 interviews, and – 1 on-air 20-minute program in Spanish; • Visits to 25 newspapers, includ ing— – 22 contacts, – 13 interviews, and – 1 interview with a Spanish- language newspaper; • Visits to three television stations, including four contacts and three interviews; and • Visits to three visitor centers in Figure 3—Keetch–Byram Drought Index at Dallas, GA, from October 26 to Bartow, Floyd, and Whitfield November 22, 2001. Counties. Effectiveness Analysis If one assumes the same level of fire activity How well did the team do? One way before and after the team was in place, the drop to tell is to compare the number of in fire activity can be attributed to the burn ban escaped debris-burning fires and plus the team’s work. acres burned before and after the team arrived. We looked at the 14 days before and the 14 days after (fig. 3). According to Melton (1997), lower to the upper ends of this the team arrived, focusing primari fires burning when the KBDI is in range.” ly on district 1 (the Rome District) the range of 400 to 600 can be very and using districts 2–4 (the adja intense. “Under these levels, most Both the acres burned and the cent districts) for comparison. of the duff and associated organic number of fires declined sharply layers will be sufficiently dry to after the burn ban was announced The entire evaluation period had ignite and contribute to the fire (table 2, fig. 4). If one assumes the severe burning conditions, with the intensity and will actively burn,” same level of fire activity before and Keetch–Byram Drought Index noted Melton (1997). “The intensity after the team was in place, the (KBDI) at 454 on October 26 and can be expected to increase at an drop in fire activity can be attrib increasing to 545 by November 23 almost exponential rate from the uted to the burn ban in districts
Table 2—Occurrence of escaped debris-burning fires on the 14 days before and the 14 days after the arrival of a wildland fire prevention team in Rome District, GA.
District 1 a Districts 2–4 a Data Before b After b Total Before b After b Total Total Number of fires 128 74 202 52 39 91 293 Percent of total 63 37 100 57 43 100 100 Acres burned 412 126 538 109 43 152 690 Percent of total 77 23 100 72 28 100 100 a. District 1 = Rome District; districts 2–4 = Athens, Gainesville, and Newnan Districts (fig. 1). b. Before = 10/26/01 to 11/8/01; after = 11/9/01 to 11/22/01.
Volume 64 • No. 2 • Spring 2004 45 The fire prevention team prevented 18 fires and kept 34 acres from burning, for a net savings of $10,400.
tricts, then the substantially greater savings for the Rome District are at least partly attributable to the effec tiveness of the fire prevention team.
Just how many fires did the team actually prevent? Table 2 shows that:
• 293 debris-burning fires escaped during the entire 28-day period evaluated, burning a total of 690 acres; and • 6 percent fewer fires escaped in district 1 than in districts 2–4 after the fire prevention team arrived, burning 5 percent fewer acres.
If we therefore assume that the number of fires prevented by the team is 6 percent of the total num ber of fires and that 5 percent fewer Figure 4—The number of escaped debris-burning fires (top) and acres burned (bottom) in acres burned, then the team pre the Rome District before and after a wildland fire prevention team was deployed. vented 18 fires and saved 34 acres 2–4. In the Rome District, the drop Based on table 2 and an estimated from burning. can be attributed to the burn ban suppression cost of about $500 per plus the work of the fire prevention acre, savings due to the burn ban in Now we can calculate cost savings. team. districts 2–4 can be calculated as: If the team kept 34 acres from burning at a suppression cost of Cost/Benefit Analysis [(109 acres burned before – $500 per acre, then the team saved × × $17,000 in suppression costs. The Can a dollar value be assigned to 43 acres burned after) 3] $500 ≅ $99,000 team itself cost $6,600, so net sav the team’s effectiveness? ings are $10,400.
In Georgia, we estimate the total In district 1 (the Rome District), total savings due to both the burn In summary, the fire prevention cost of suppressing fire on 1 acre team’s activities: (0.4 ha) to be about $500, based on ban and the fire prevention team can similarly be calculated as: the average number of acres • Reduced the number of fires by burned annually and the total 18, annual GFC fire budget. This is (412 acres burned before – × ≅ • Reduced the number of acres higher than the actual on-the 126 acres burned after) $500 $143,000 burned by 34, and ground cost for an individual fire, • Produced a net savings of because the $500 includes all orga $10,400. nizational costs. If burn ban effectiveness is assumed to be constant across all four dis-
Fire Management Today 46 Of course, there are important All things considered, it seems quite reasonable to caveats. Our data set was small, and say that the fire prevention team contributed more extensive analysis might show different results, although the gen- measurably to reducing fire activity in the Rome eral pattern is likely to be similar. District. Moreover, parts of the Newnan and Gainesville Districts were probably influenced by the fire prevention sentative of the savings actually References effort in the Rome District through realized by the team. GFC Fire Staff. 1996. Georgia Forestry the reach of radio, television, and Commission fire report manual. Macon, All things considered, it seems GA: Georgia Forestry Commission. newspapers across districts. Con Lane, R.W.; Slauson, G.H.; Haire, B.J. 2001. sequently, the fire prevention team quite reasonable to say that the fire Georgia Forestry Commission Wildland might have been even more effec prevention team contributed meas Fire Prevention Team—final report. urably to reducing fire activity in Unpublished report on file with the tive than calculated. Georgia Forestry Commission, Macon, the Rome District. The work of the GA. Positive Balance team was considered a success—an Melton, M. 1997. Keetch–Byram Drought incentive to deploy similar teams in Index revisited: Prescribed fire applica The benefits of fire prevention are the future. tions. Fire Management Notes. 56(4): often difficult to measure and 7–11. Moorman, C. 2002. Burning for wildlife. therefore intuitive at best. But the Acknowledgments Forest Landowner. 61(3): 5–9 fire prevention team in the Rome Wade, D.D.; Lunsford, J.D. 1989. A guide District did produce measurable The authors wish to thank Dr. for prescribed fire in southern forests. results. The estimate of $10,400 James A. Pharo, a forest economist, Tech. Pub. R8–TP–11. Asheville, NC: for his helpful suggestions in the USDA Forest Service, Southern Research produced in savings is likely repre Station. ■ cost/benefit analysis for this article.
Volume 64 • No. 2 • Spring 2004 47 BUILDING GROUP COHESION IN TYPE 2 FIRE CREWS
Bill Lee
uch of the training for wild- land firefighters, from Group cohesion can mean the difference M smokejumpers to ground between life and death. pounders, centers on fire behavior, weather, fuels, fireline construc tion, safety, physical conditioning, Before Fire Season it—the district ranger needs to say, and equipment use. The training is Type 2 crews usually include per “Besides learning policy, proce extensive and thorough, yet one sonnel from different units (such as dures, and programs, we want you element might not not always ranger districts or national forests) to get to know each other better so receive enough attention at the or agencies (such as the Forest that when we move into our fire local or national level—and it can Service, USDI Bureau of Land season, we can pull together more mean the difference between life Management, or State fire divi effectively.” Such prefire experi and death. sions). Different affiliations among ences will improve communication crew members make it difficult for on a fire. That element is group cohesion of a a crew boss to build a cohesive unit fire crew. Group cohesion affects (Putnam 1995b). Right when a type During Staging group communication, decision- 2 crew is first mobilized, the crew This approach works well with making, and survivability. In a 12 boss should openly acknowledge adjoining units for local fire sup year study of USDA Forest Service group cohesion as a crew weakness. pression. But crews are often fire crews, Driessen (1990) discov formed from people from around a ered an inverse correlation between Crew members not knowing each State or across a region. How do we crew cohesion and accident rates. other can hamper communication, build cohesion in such crews? In his report on the collapse of trust, and leadership, all key to a decisionmaking and organizational cohesive crew (fig. 1). If crew mem Again, we look for prefire opportu structure on the 1994 South bers are to better communicate, nities to help crew members get Canyon Fire, Putnam (1995a) make better decisions based on better acquainted and understand pointed out that too little training group input, and watch out for each other’s strengths. In the stag time for firefighters is spent on each other, then crew bosses must ing phase on the way to a fire, crew improving thinking, leadership, and speed up the process of building members typically meet at a prede crew interactions. group cohesion. termined location and caravan to Driessen (2002) offered a basis for One way is through gatherings of increasing the level of cohesion on crew members in neighboring a type 2 crew as quickly as possible. ranger districts before fire season This article takes Jon Driessen’s begins. For example, the Washakie Leadership Trust report on crew cohesion and pro Ranger District on the Shoshone vides specific steps to increase National Forest has an orientation cohesion in type 2 crews. Group campout with the neighboring Cohesion ranger district at the beginning of summer to get people better Bill Lee is a school social worker in Lander, acquainted with each other for WY, during the schoolyear; and a seasonal coming work projects and fire duty. Communication recreational specialist/firefighter since 1979 for the USDA Forest Service, Shoshone National Forest, Washakie Such socializing builds group cohe Figure 1—Three key elements make up the Ranger District, Lander, WY. sion. The purpose should be explic- group cohesion triangle.
Fire Management Today 48 the fire. During this phase, the Right when a type 2 crew is first mobilized, the crew boss should assemble the crew crew boss should openly acknowledge group and acknowledge the need for crew members to get to know each other cohesion as a weakness. better. Crew members can connect with each other, building cohesion work as crew members wait to be safety of the crew, and make deci within the crew, through a formal dispatched. sions that can have life-or-death discussion on crew cohesion. For implications. Type 2 crews require a example, crew members might: Throughout a crew’s assignment, crew boss who is skilled at: from standby to actual fire duty, • Introduce themselves, crew bosses should work to build • Fostering good communication, • Explain the special skills they group cohesion (see the sidebar • Communicating clear expecta bring to the fireline, below). At the beginning or end of tions, • Say something personal about each shift, crew bosses can discuss • Briefing crews well, themselves, observed examples of good commu • Understanding the strengths and • Discuss what they can contribute nication and teamwork. They can limitations of each crew member, to a more cohesive crew, and also ask the crew to share their and • Add anything else they want. own observations of good commu • Creating an atmosphere in which nication and teamwork. Finally, the crew members look out for each Afterwards, the crew members crew needs to talk about what other. should be assigned to travel in things are helping the crew become groups other than the groups they more cohesive. In short, we need type 2 crew boss arrived with. Travel and conversa es who can build group cohesion. tion with unfamiliar people can The toughest and most important help break down communication position on a fire is the crew boss. Through Training barriers and build alliances. Such He or she must carry out plans The wildland fire community mixing can break up cliques and from overhead, look out for the foster group cohesion. knows the importance of group
When a crew arrives on a fire, it is best to assemble crew members Sample Training Card into squads by the groups they The words below are printed on a business card I hand out at crew originally arrived with for mobiliza boss training. It can help trainees remember steps to follow in devel tion. The time spent split apart for oping group cohesion with the type 2 crews. travel will still enhance communi cation between squads, encourage Type 2 Wildland Fire Crew At Staging Area squads to look out for each other • Good decisions occur by pro • Crew members introduce them on the fireline, and make it easier moting crew cohesion. selves. for crew members from different • Acknowledge the need for crew • Explain skills/experiences they squads to further get to know each members to get to know each bring to the fire crew. other during break times and in other better. • Share something personal fire camp. • Encourage input from all about themselves. crewmembers. Express contra • Explain what they can con When on Standby dictory observations in a tribute to better a cohesive Under the severe fire conditions of respectful manner. crew. recent years, some fire manage • Brief/debrief often. ment officers have ordered type 2 • Reassign to traveling groups. crews to stand by for potential fires. • Know strengths/limitations of The standby strategy can allow each crew member. more time for a crew to become • Model behavior you want from more cohesive through project the crew.
Volume 64 • No. 2 • Spring 2004 49 Training Opportunities for We should use prefire Building Group Cohesion opportunities to help crew members get The National Wildfire making—in field simulations. Coordinating Group offers several It allows fire crew leaders to better acquainted and courses on group cohesion. practice crew cohesion-building understand each other’s Though not mandatory, they are techniques, such as providing strengths. widely used within the wildland vision, estabishing standards, fire community, including the conducting after-action reviews, USDA Forest Service. mitigating stress, and resolving 6 weeks that it typically takes to conflict. develop a cohesive type 2 crew. • Human Factors on the Fireline Type 2 crews can benefit from very (L–180): This entry-level Two more courses are currently purposeful training prior to the fire course, intended for all fire- under development, one for those season to enhance group cohesion. fighters, addresses basic situa- stepping up to the Division Additional training can improve tional awareness and communi- Supervisor or IC Type 3 level communication, decisionmaking, cations skills, along with the (L–381), the other a seminar for and survivability. challenges of integrating into incident management team can- an effective team operating in a didates (L–480/580). References high-risk environment. Driessen, J. 1990. The supervisor and the • Followership to Leadership In addition, a Web-based self work crew. Missoula, MT: USDA Forest (L–280): Intended for aspiring study resource (
Fire Management Today 50 WILDLAND FIRE EDUCATION: GOING THE DISTANCE IN ALASKA
Sandi Sturm and Matt Weaver
eaching educators in rural Alaska is not easy. A midsized Communicating with educators in rural Alaska is R rural district, Iditarod, is about hard, but it is critical that Alaskan educators have the same size as the State of Ohio. access to exciting, interdisciplinary fire education Anchorage and Barrow, AK, are about 800 air miles (1,300 km) materials. apart. Many rural communities are in roadless areas and rely on air or as the USDI U.S. Fish and Wildlife water travel most of the year. Face-to-Face Workshops Service’s Role of Fire in Alaska, the USDA Forest Service’s FireWorks, Very few opportunities for addition Fire crosses all boundaries—and so and the national FireWise pro al training exist in the home com does Project Learning Tree (PLT). gram—and contributions from munities of rural educators. The In partnership with the U.S. Alaskan agency fire experts to build alternative is to travel to “the city” Department of the Interior (USDI) a program that addresses the for training. However, a weekend Bureau of Land Management and in unique needs of Alaskan educators. workshop can cost $2,000 or more, cooperation with the National not to mention the inconvenience Interagency Fire Center, PLT deliv More than 100 educators have com of leaving the classroom for up to 4 ers annual fire education work pleted the 2- to 3-day Fire! In days. Extreme weather can extend shops to hundreds of teachers in Alaska face-to-face workshop. The such trips indefinitely. about 20 States. first workshop was in Homer, a town nestled between the boundary In spite of these obstacles, it is With support from agencies, organ of Alaska’s coastal rain forest and more critical than ever for Alaskan izations, and colleagues, Alaska PLT the interior boreal forest on educators to get exciting interdisci developed a fire education program Alaska’s south-central coast—often plinary fire education materials. In called Fire! In Alaska. The program called the “end of the road.” Homer, 2002, Alaska had its fifth largest is designed to help educators learn the epicenter of a decade-long wildland fire season on record, with about wildland fire ecology, behav spruce bark beetle epidemic that 539 fires burning more than 2.2 ior, and prevention (see the sidebar has affected more than a million million acres (880,000 ha). on page 52). Alaska PLT used a vari acres (400,000 ha), was the perfect ety of established curricula—such Most of these fires were human caused, and several threatened vil Alaska Division of Forestry Fire lages and towns, recreational areas, Manager Joe Stam and cultural and historic sites. The and Alaska teachers answer is more fire education evaluate a property within beetle-killed statewide. Because teachers can’t spruce in Homer, AK, always come to us, we must find a using FireWise evalu way to get to them. ation sheets. Photo: Matt Weaver, Alaska Department of Natural Resources, Division of Forestry, Sandi Sturm is the online learning design Anchorage, AK, 2002. er with Creative Conservation, Wasilla, AK; and Matt Weaver is the Alaska Project Learning Tree coordinator, Alaska Department of Natural Resources, Anchorage, AK.
Volume 64 • No. 2 • Spring 2004 51 Fire! In Alaska Course Objectives Alaska Project Learning Tree • Define, understand, model, and • Use the fire triangle and fire developed a program for educa- predict primary and secondary experiments to explain how fire tors in rural Alaska called Fire! In succession in the boreal forest. is used safely in prescribed Alaska. The 2- to 3-day workshop • Provide examples of how biota burning and fire suppression. is designed to help teachers do are adapted to fire-dependent • Define, describe, and model how the following: ecosystems. controlled burning influences • Understand the Alaska Wildland ecosystems and people in the Fire Ecology Fire Coordinating Group suppres- wildland/urban interface. • Locate and describe three sion plan for Alaska. • Define, describe, and model the major Alaska biomes. three main types of wildland fire – Explain how variables such Fire Behavior and the variables that influence as temperature, rainfall, • Explain the chemical nature of them. soils, and topography influ- combustion. ence biota. – Compare combustion to photo- Fire Prevention/Fuels Mitigation • Apply anatomical similarities synthesis and cellular respira- • Understand and apply precepts and differences to taxonomy. tion. of defensible space to hypotheti – Use a dichotomous key. • Design and conduct new experi- cal and real situations. – Name major Alaskan trees by ments that explain how the parts – Use the FireWise defensible sight. of the fire triangle interact. space checklists to evaluate • List how boreal forest fire his- • Identify, explain, and model how their own homes. tory is studied and explain fuels, weather, and topography • Design a FireWise house. how/why fire affects the biotic influence the behavior of wild- • Plan and conduct a simulated and abiotic features of the for- land fire. prescribed burn to safely miti est. gate fuel loads near subdivisions. backdrop for a field trip after par ticipants learned about fire ecology Distance education is an excellent tool to help and conducted fire behavior experi- reach wildland fire ecology, behavior, and ments. prevention goals. On the field trip, teachers evaluated private homes in the wildland/ ing connectivity issues and over fire experts from several agencies— urban interface threatened by the coming numerous obstacles, the the Alaska Department of Natural dry, heavy fuels left behind by first distance PLT course was avail Resources, Division of Forestry; and marauding beetles. Although the able in May 2002. the USDI Bureau of Land workshops have been a great suc Management and National Park cess—we trained about 100 teach By April 2004, five online courses Service—joined educators in pilot ers in 2003, who will reach an esti had reached 75 teachers in rural ing the program and fine-tuning its mated 2,000 students and house and remote villages, who in turn content. holds—they are not easily accessi reach 1,500 students per year or ble to teachers working in rural more. Even before the online The term “online course” might be Alaskan communities. courses ended, the teachers were deceiving. The course uses a variety already using the materials. of methods and technologies. Fire! In Alaska Online Rather than a “sit-and-click” kind One answer was online training. In Alaska PLT and the professional of experience, the course is fully 2000, Alaska PLT enlisted support facilitator followed up by develop facilitated and very interactive, from a PLT facilitator to develop a ing an 8-week Fire! In Alaska bringing in local fire experts for a distance learning course on wild- course, to be delivered through dis local twist. land fire. After months of research tance education. In October 2003,
Fire Management Today 52 Alaska Project Learning Tree teaches teachers about fire ecology, behavior, and prevention through a large national network of State coordinators and volunteer facilitators.
tance education initiative offers the most effective long-term solution for reducing the number of new fire starts, informing the public about wildland management, and preventing catastrophic loss in remote communities due to wild- land fire.
Figure 1—Since May 2002, 47 educators from across Alaska have taken an online Project Research has already begun to take Learning Tree workshop. the program to other regions of the country. Sponsors are coming for Fire! In Alaska Online focuses on The 8-week course is scheduled to ward to develop a regional or the same three areas as the face-to be delivered twice per year, in the national Fire! Online course and a face workshop (see the sidebar). fall and spring of the schoolyear. train-the-trainer program to build a Teleconferencing, compact discs, Teachers are currently on a waiting force of online facilitators. PLT is and the Internet are incorporated list for the fall 2004 offering. committed to offering the opportu throughout the course. Participants Several times per year, face-to-face nity to its more than 3,000 coordi share local cultures and interact workshops are still offered to edu nators and facilitators nationwide. with their individual communities cators living in more urban areas. while working on fire ecology, For more information about dis behavior, and prevention projects Future Training tance-delivered wildland fire educa and experiments. One advantage is Distance education can play a vital tion, contact Sandi Sturm at the interaction among the many role in Alaska’s wildland fire man
Volume 64 • No. 2 • Spring 2004 53 CANADIAN FOREST FIRE WEATHER INDEX SYSTEM: TRAINING NOW AVAILABLE ON CD-ROM
Paul St. John and Martin E. Alexander
nderstanding the Fire Weather Index (FWI) The course offers a comprehensive introduction to “U System” is the latest CD the Canadian Forest Fire Weather Index System, a ROM-based wildland fire training major subsystem or module of the Canadian course produced by Alberta’s Hinton Training Centre in concert Forest Fire Danger Rating System. with Christie Communications* to utilize interactive multimedia tech ing/persistence, spread rate, total • 14 video clips, 219 audio clips, nology (Alexander and others fuel consumption, and intensity— and 656 graphics/photos; 2002). The course, completed in based on four weather observations. • Online help and a glossary; August 2002, also involved the • An SI-to-imperial-unit conversion Canadian Interagency Forest Fire “Understanding the Fire Weather calculator; Centre’s National Training Working Index (FWI) System” contains: Group and was produced in associa tion with the Canadian Forest Service. Course Content The course offers a comprehensive introduction to the Canadian Forest Fire Weather Index (FWI) System, one of the major subsys tems or modules of the Canadian Forest Fire Danger Rating System. The FWI System consists of three fuel moisture codes and three fire behavior indexes that provide rela tive numerical ratings for six aspects of wildland fire potential— ignition, duff consumed, smolder-
Paul St. John is a fire management instructor with Alberta Sustainable Resource Development, Hinton Training Centre, Hinton, Alberta, Canada; and Marty Alexander is a senior fire behavior research officer with the Canadian Forest Service, Northern Forestry Centre, Edmonton, Alberta, Canada.
* The use of trade, firm, or corporation names in this publication is for the information and convenience of the reader. Such use does not constitute an official The CD-ROM training course “Understanding the Fire Weather Index (FWI) System” endorsement of any product or service by the U.S. Department of Agriculture. Individual authors are encapsulates more than three decades of knowledge and experience with the FWI System responsible for the technical accuracy of the material in Canada. Photo: M.E. Alexander, Canadian Forest Service, Northern Forestry Centre, presented in Fire Management Today. Edmonton, AB, 2002.
Fire Management Today 54 The course was developed and reviewed by a national team of fire danger rating specialists representing research, operations, Missing and training. [person at computer screen = FWI.jpg]
• The “FWI Calculator,” which allows for the calculation of the six standard components of the FWI System for two broad regions in both the northern and the southern hemispheres; and • A calculator allowing for the overwinter adjustment to the spring starting value of the Drought Code component of the A training session in progress using the “Understanding the Fire Weather Index (FWI) FWI System. System” CD-ROM. Photo: N. Merrifield, Alberta Sustainable Resource Development, Hinton Training Centre, Hinton, AB, 2003. The course was developed and The course can be run on a stand the Hinton Training Centre Website reviewed by a national team of fire alone computer or over a network. at
Volume 64 • No. 2 • Spring 2004 55 “BREWER FIRE MYSTERY” DISCUSSION
Editor’s note: Occasionally, Fire Management Today publishes comments from readers on topics of concern, offering authors a chance to respond. To have your comments considered for publication, write to Managing Editor Hutch Brown at USDA Forest Service, Office of the Chief, Yates Bldg., 4th Floor NW, 201 14th St., SW, Washington, DC 20024, tel. 202-205-0878, fax 202-205-1765, e-mail: [email protected].
Reader Comment: Montana was even more severe Author Response: “Brewer Fire than during the Dustbowl. The What Triggered the recorded moisture was 3.35 inches Mystery” Not So (8.49 cm), compared to 5.11 inches Brewer Fire Blowup Mysterious (12.98 cm) per year in the 1930s. Remains the Mystery Normally, surface fires in open dry Stephen A. Eckert ponderosa pine forest stay on the Martin E. Alexander ground; but in summer 1988, had n article in the Fall 2002 issue you touched a lighted match to the irst of all, I wish to state for the A of Fire Management Today pine duff, the flame would have eas record that my mention of the mentions a blowup on the 1988 ily crawled all the way up even the F Brewer Fire in my article Brewer Fire in Montana that forced biggest yellow pine. Throughout (Alexander 2002) was in no way shelter deployment by the that summer, winds that ranged meant to criticize the people Wyoming Interagency Hotshot from southwest to northwest were involved in fire suppression or the Crew.* The article states that “that consistently and unusually strong, subsequent investigation into the there has never been an explana both during the day and in the shelter deployment incident. tion for what triggered the Brewer evening. However, I do believe that Mr. Fire blowup.” Eckert has missed the point as to So conditions were ripe for extreme what “mystery” I was referring to in But what happened on the Brewer fire behavior. They included record my article. I hope this response will Fire is no mystery. drought, record low fuel moistures, clarify matters, and I appreciate the erratic and strong winds, extreme opportunity to elaborate on my ini From 1982 to 1990, I was the fire temperatures, and very low relative tial thoughts concerning the control officer for the Bureau of humidities. Under these conditions, Brewer Fire blowup. Land Management, Miles City Field the fire quickly went from a surface Office, Miles City, MT. From the fire to a running crown fire. The My article was not intended to outset of the Brewer Fire, I was the hotshot crew was flanking the fire, serve as a case study of the Brewer air attack supervisor. I also ordered building fireline. Had a lookout Fire. My sole purpose was to show the overhead team on the evening been posted in the meadow where how fire investigations are often the fire started, and I was a mem the deployment took place, the rushed and the root causes of an ber of the fire investigation team crew would have had more time for incident on a fire (e.g., a fatality or that later explored and reported on escape or shelter deployment. near-miss) are often inadequately the shelter deployment incident. explored due to more pressing Under the severe burning condi issues, in this case a rapidly escalat In 1988, the drought in eastern tions at the time, what happened ing fire season in the Western on the Brewer Fire is no mystery. United States. I wanted to support Steve Eckert is the assistant fire manage Instead, it was entirely predictable. my call for creating wildland fire ment officer for the Bureau of Land Obviously, weather factors created behavior research units. Management, Wyoming State Office, an explosive environment. No other Cheyenne, WY. explanation is needed. Marty Alexander is a senior fire behavior * See Martin E. Alexander, “The Staff Ride Approach to Wildland Fire Behavior and Firefighter Safety Awareness research officer with the Canadian Forest Training: A Commentary,” Fire Management Today Service at the Northern Forestry Centre, 62(4) [Fall 2002]: 25–30. Edmonton, Alberta.
Fire Management Today 56 It is absolutely true that the critical nomena could be made (Schroeder examined by a fire weather meteor level of dryness in live and dead and Countryman 1960). A lot more ologist using all the available data fuels contributed to extreme fire work is needed. The meteorological from both synoptic and mesoscale behavior on the Brewer Fire. But conditions associated with the 1953 standpoints (e.g., upper air data, the real question is this: What actu Rattlesnake Fire in California, satellite and radar imagery, and ally triggered the temporary escala which involved 15 firefighter fatali hourly airport observations). tion in extreme fire behavior on ties, are a specific case in point Simply examining the data collect that particular evening in late June (Maclean 2003). ed at a single remote automatic 1988? weather station might not neces The whole point of my bringing up sarily suffice to detect an HB, We simply don’t know. That is the the Brewer Fire was the need for because existing research suggests mystery, not the fire as a whole, thorough followup, because investi that HBs are so localized that they which is readily explainable. gations are often rushed and we are not picked up at a single obser Certainly, other fires have burned don’t necessarily learn as much vation point on the landscape. under similar critically dry fuel sit about what influenced a fire’s uations over the years, and yet we behavior as we should or could have. In closing, if extreme fire behavior haven’t always seen events like As a result, we set ourselves up for was really so predictable on the those reported on the Brewer Fire. the possibility of repeating the same Brewer Fire, then I must ask, with scenario sometime in the future— all due respect: Why was the I am a strong believer in not perhaps with a fatal outcome. Wyoming IHC allowed to be in attributing unusual fire behavior to such a dangerous position? To my an “act of God.” So I speculated I believe that mesoscale phenome knowledge, neither the fire weather that perhaps a heat burst (HB) was na such as an HB should be looked forecast nor the fire behavior fore responsible for causing (i.e., trig into as a possible factor in the cast mentioned the possibility of gering) the blowup or flareup that blowup of the Brewer Fire. An HB what transpired. forced the Wyoming Interagency would seem to explain what hap Hotshot Crew (IHC) to move away pened. Associated with nocturnal References from the fire to a clearing and thunderstorms, HBs are character Alexander, M.E. 2002. The staff ride deploy fire shelters. ized by a sudden and dramatic approach to wildland fire behavior and localized increase in air tempera firefighter safety awareness training: A commentary. Fire Management Today. An HB is a recognized meteorologi ture and a drop in relative humidity, 62(4): 25–30. cal phenomenon (Bernstein and coupled with strong, gusty winds. If Bernstein, B.C.; Johnson, R.H. 1994. A Johnson 1994; Johnson 1983). we were to find that the Brewer dual-doppler radar study of an OK PRE-STORM head burst event. Monthly Perhaps HBs happen a lot more Fire blowup was in fact triggered by Weather Review. 122: 259–273. often than we think. We think we an HB, we might in the future be Johnson, B.C. 1983. The heat burst of 29 have studied our fire environments able to use a sudden increase in air May 1976. Monthly Weather Review. 111: 1776–1792. really well, but the truth of the temperature—like the one reported Maclean, J.N. 2003. Fire and ashes: On the matter is that we haven’t—we just by the Wyoming IHC foreman just front lines of American wildfire. New think we have. In the late 1950s, before the blowup—as an “early York: Henry Holt and Company. Mark Schroeder and Clive warning system.” Schroeder, M.J.; Countryman, C.M. 1960. Exploratory fireclimate surveys on Countryman conducted a series of prescribed burns. Monthly Weather “fireclimate surveys” to begin col The possibility that an HB ulti Review. 88(4): 123–124. ■ lecting case histories or studies mately triggered (not set up) the from which generalizations about Brewer Fire shelter deployment the dynamics of mesoscale phe incident should, in my opinion, be
Volume 64 • No. 2 • Spring 2004 57 GUIDELINES FOR CONTRIBUTORS Editorial Policy Paper Copy. Type or word-process the manu etc.; photograph A, B, C, etc.). At the end of the Fire Management Today (FMT) is an internation script on white paper (double-spaced) on one manuscript, include clear, thorough figure and al quarterly magazine for the wildland fire com side. Include the complete name(s), title(s), affili photo captions labeled in the same way as the munity. FMT welcomes unsolicited manuscripts ation(s), and address(es) of the author(s), as well corresponding material (figure 1, 2, 3; photo from readers on any subject related to fire man as telephone and fax numbers and e-mail infor graph A, B, C; etc.). Captions should make pho agement. Because space is a consideration, long mation. If the same or a similar manuscript is tos and illustrations understandable without manuscripts might be abridged by the editor, being submitted elsewhere, include that informa reading the text. For photos, indicate the name subject to approval by the author; FMT does tion also. Authors who are affiliated should sub and affiliation of the photographer and the year print short pieces of interest to readers. mit a camera-ready logo for their agency, institu the photo was taken. tion, or organization. Submission Guidelines Electronic Files. See special mailing instruc Submit manuscripts to either the general man Style. Authors are responsible for using wildland tions above. Please label all disks carefully with ager or the managing editor at: fire terminology that conforms to the latest stan name(s) of file(s) and system(s) used. If the man dards set by the National Wildfire Coordinating uscript is word-processed, please submit a 3-1/2 USDA Forest Service Group under the National Interagency Incident inch, IBM-compatible disk together with the Attn: April J. Baily, F&AM Staff Management System. FMT uses the spelling, cap paper copy (see above) as an electronic file in one Mail Stop 1107 italization, hyphenation, and other styles recom of these formats: WordPerfect 5.1 for DOS; 1400 Independence Avenue, SW mended in the United States Government WordPerfect 7.0 or earlier for Windows 95; Washington, DC 20250-1107 Printing Office Style Manual, as required by the Microsoft Word 6.0 or earlier for Windows 95; tel. 202-205-0891, fax 202-205-1272 U.S. Department of Agriculture. Authors should Rich Text format; or ASCII. Digital photos may e-mail: [email protected] use the U.S. system of weight and measure, with be submitted but must be at least 300 dpi and equivalent values in the metric system. Try to accompanied by a high-resolution (preferably USDA Forest Service keep titles concise and descriptive; subheadings laser) printout for editorial review and quality Attn: Hutch Brown, Office of the Chief and bulleted material are useful and help read control during the printing process. Do not Yates Bldg., 4th Floor NW ability. As a general rule of clear writing, use the embed illustrations (such as maps, charts, and 1400 Independence Avenue, SW active voice (e.g., write, “Fire managers know…” graphs) in the electronic file for the manuscript. Washington, DC 20024 and not, “It is known…”). Provide spellouts for Instead, submit each illustration at 1,200 dpi in a tel. 202-205-0878, fax 202-205-1765 all abbreviations. Consult recent issues (on the separate file using a standard interchange format e-mail: [email protected] World Wide Web at such as EPS, TIFF, or JPEG, accompanied by a
Contributors Wanted We need your fire-related articles and photographs for Fire Management Today! Feature articles should be up to about 2,000 words in length. We also need short items of up to 200 words. Subjects of articles published in Fire Management Today include: Aviation Firefighting experiences Communication Incident management Cooperation Information management (including systems) Ecosystem management Personnel Equipment/Technology Planning (including budgeting) Fire behavior Preparedness Fire ecology Prevention/Education Fire effects Safety Fire history Suppression Fire science Training Fire use (including prescribed fire) Weather Fuels management Wildland–urban interface To help prepare your submission, see “Guidelines for Contributors” in this issue.
Fire Management Today 58 PHOTO CONTEST ANNOUNCEMENT Fire Management Today invites Rules job descriptions; and descriptions you to submit your best fire-related • The contest is open to everyone. of any vegetation and/or wildlife). photos to be judged in our annual You may submit an unlimited • You must complete and sign a competition. Judging begins after number of entries from any place statement granting rights to use the first Friday in March of each or time; but for each photo, you your photo(s) to the USDA Forest year. must indicate only one competi Service (see sample statement tion category. To ensure fair eval below). Include your full name, Awards uation, we reserve the right to agency or institutional affiliation All contestants will receive a change the competition category (if any), address, and telephone CD–ROM with all photos not elimi for your photo. number. nated from competition. Winning • An original color slide is pre • Photos are eliminated from com photos will appear in a future issue ferred; however, we will accept petition if they have date stamps; of Fire Management Today.In high-quality color prints, as long show unsafe firefighting practices addition, winners in each category as they are accompanied by nega (unless that is their express pur will receive: tives. Digitally shot slides (pre pose); or are of low technical ferred) or prints will be accepted quality (for example, have soft • 1st place—Camera equipment if they are scanned at 300 lines focus or show camera move worth $300 and a 16- by 20-inch per inch or equivalent. Digital ment). (Duplicates—including framed copy of your photo. images will be accepted if you most overlays and other compos • 2nd place—An 11- by 14-inch used a camera with at least 2.5 ites—have soft focus and will be framed copy of your photo. megapixels and the image is shot eliminated.) • 3rd place—An 8- by 10-inch at the highest resolution or in a • Photos are judged by a photogra framed copy of your photo. TIFF format. phy professional whose decision • You must have the right to grant is final. the Forest Service unlimited use Categories Postmark Deadline • Wildland fire of the image, and you must agree • Prescribed fire that the image will become public First Friday in March • Wildland-urban interface fire domain. Moreover, the image • Aerial resources must not have been previously Send submissions to: • Ground resources published. Madelyn Dillon • Miscellaneous (fire effects; fire • For every photo you submit, you CAT Publishing Arts weather; fire-dependent commu must give a detailed caption 2150 Centre Avenue nities or species; etc.) (including, for example, name, Building A, Suite 361 location, and date of the fire; Fort Collins, CO 80526 names of any people and/or their
Sample Photo Release Statement
Enclosed is/are (number) slide(s) for publication by the USDA Forest Service. For each slide submitted, the contest category is indicated and a detailed caption is enclosed. I have the right to grant the Forest Service unlimited use of the image, and I agree that the image will become public domain. Moreover, the image has not been previously published.
Signature Date
Volume 64 • No. 2 • Spring 2004 59