Rethinking the Wildland Fire Management System

Home , Journal of Forestry

J. For. 116(4):382–390 PRACTICE OF FORESTRY doi: 10.1093/jofore/fvy020 Published by Oxford University Press on behalf of the Society of American Foresters 2018. This work is written by (a) US Government employee(s) and is in the public domain in the US. fire & fuels management Rethinking the Wildland Fire Management System

Matthew P. Thompson, Donald G. MacGregor, Christopher J. Dunn, David E. Calkin, and John Phipps

In the western United States and elsewhere, the need to change society’s relationship with wildfire is well-recog- the toolkit; however, application of fire is nized. Suppressing fewer fires in fire-prone systems is promoted to escape existing feedback loops that lead to ever often necessary for hazardous fuel reduc- worsening conditions and increasing risks to responders and communities. Our primary focus is how to catalyze tion and forest restoration treatments to be changes in fire manager behavior such that responses are safer, more effective, and capitalize on opportunities for effective (Schwilk et al. 2009; McIver et al. expanded use of fire. We daylight deep-seated, systemic drivers of behavior, and in so doing, challenge ingrained 2013; Fischer et al. 2016; Kalies and Kent ways of thinking and acting that may be inconsistent with current intentions around wildland fire management. We 2016). Further, feasible scales for application pose the questions of whether all fires are emergencies that require rapid deployment and concentration of sup- of mechanical treatment and prescribed fire pression resources, whether rhetoric and actions align with policy and guidance, and whether we can unambiguously are often dwarfed by the area that burns from define and measure what a safe and effective response looks like. Using the Forest Service of the US Department unplanned ignitions annually (North et al. of Agriculture (USDA) as a relevant test case for systemic investigation, we argue that fundamental changes in how 2012; Collins et al. 2013; North et al. 2015b; the fire management community thinks about, learns from, plans for, and responds to wildland fires may be neces- Barnett et al. 2016a). Opportunistically sary. Our intention is to initiate a broader dialog around the current and future state of wildland fire management. managing unplanned ignitions under less extreme weather conditions could in the Keywords: risk management, systems thinking, restoration, resiliency, suppression future limit fire spread and occurrence, alter the extent and spatial pattern of burn severity, and enhance suppression effectiveness t is increasingly apparent that in many risks, (i.e., the “fire paradox”) (Calkin et al. (Regos et al. 2014; Parks et al. 2015; Stevens- places around the globe, society must 2015; Olson et al. 2015; Ingalsbee 2017). Rumann et al. 2016; Parks et al. 2016; Iforge a new relationship with fire If void of significant social and managerial Thompson et al. 2016b; Stevens et al. 2017). (Moritz et al. 2014; Curt and Frejaville changes, the future of wildland fire man- Therefore, changing responses to unplanned 2017; Otero and Nielsen 2017). In the agement in the western United States likely ignitions provides a largely untapped but western United States, a changing cli- entails increasing costs, damages, and loss important—if not essential—opportunity mate, expanding human development, and of life (Thompson et al. 2016a). to restore landscape conditions and reduce accumulating fuels (among other factors) While any progress toward sustainable future hazard and risk. require new approaches that manage for solutions will necessarily be multifaceted A principle focus here is how best to cat- resilience by suppressing fewer fires and (cross-boundary collaboration and planning, alyze desired changes in fire manager behavior accelerating forest restoration ( Stephens community adaptation, etc.), a key biophysi- in terms of individual fire events, but more et al. 2016; Schoennagel et al. 2017). cal requisite is reintroducing fire to fire-prone importantly in terms of consistent patterns Perpetuating business-as-usual attempts systems (North et al. 2015a; Hessburg et al. of behavior over time. We do so because, to exclude fire from systems that evolved 2016; Boisramé et al. 2017; Prichard et al. ultimately, changes in fire manager decisions with frequent fire will in some cases simply 2017). Forest management approaches such regarding response strategies and tactics will amplify feedbacks that increase long-term as forest thinning remain a valuable tool in be necessary to change fire outcomes. Helping

Received October 26, 2017; accepted April 23, 2018; published online June 9, 2018. Affiliations: Matthew P. Thompson ([email protected]), David E. Calkin ([email protected]), John Phipps ([email protected]), Rocky Mountain Research Station, US Department of Agriculture Forest Service. Donald G. MacGregor ([email protected]), MacGregor-Bates, Inc. Christopher J. Dunn (Chris.Dunn@ oregonstate.edu), Department of Forest Engineering, Resources, and Management, Oregon State University. Acknowledgments: We would like to thank David Chapman, Christopher O’Connor, Jennifer Hayes, Jan Engert, Chuck McHugh, Scott Stephens, Malcolm North, Matt Rollins, David Cleaves, Courtney Schultz, and Sarah McCaffrey for their reviews and insights.

382 Journal of Forestry • July 2018 fire managers better arrive at response strate- accounts for the role of fire managers and as possible under nearly all circumstances. gies that reduce unnecessary exposure and their response decisions. We emphasize these The Forest Service, by contrast, has the pol- increase the probability of success is a primary decisions in light of their potentially signifi- icy flexibility to pursue a range of responses, motivation of this article. So too is helping fire cant impacts in terms of cost and responder including managing fires for ecosystem ben- managers better determine where, when, and exposure and because their accumulated efits, which not only complicates the man- how to appropriately expand the footprint of effects over time can alter landscape condi- agement problem but also creates potential fire—in the right places, at the right times, tions, influence perceptions and expectations, for conflict when managing fires across under the right conditions, for the right rea- restrict fire managers’ decision space, and ulti- jurisdictional boundaries. sons. A variety of approaches have been pro- mately perpetuate a fire exclusion paradigm. In recent years, the Forest Service has posed, including expanding public support In other words, insofar as fire response committed to strengthening its risk manage- for fire use, developing methods to quan- decisions can influence landscapes and ment acumen and capacity and is in the ini- tify the effectiveness of suppression actions, communities, ST compels us to consider tial stages of developing an ERM framework strengthening incentives that discourage the fire management system, especially in (Thompson et al. 2016a, 2016c; US Forest ineffective suppression actions, streamlining contexts where there is flexibility to change Service 2017a). The Forest Service is focused the use of analytics to enhance prefire plan- response decisions as a lever for broader on improving fire management because of ning and decision support, and increasing socioecological change. A key concept here increasing concerns over responder safety, the relevance and flexibility of fire manage- is that not only do we need to examine the critiques from oversight agencies, and a ment plans (Steelman and MCaffrey 2011; various management systems that have been severely compromised ability to achieve core Plucinski and Pastor 2013; Thompson et al. devised and are in place, but also how these missions, including addressing a backlog of 2013a; Thompson 2014; Meyer 2015; North various systems interact with each other forest restoration needs, due to the grow- et al. 2015a; Fischer et al. 2016; O’Connor and with properties of the larger business ing budgetary impacts of fire (North et al. et al. 2016; Dunn et al. 2017; Thompson environment. Systems thinking can help 2012; US Forest Service 2015; Thompson et al. 2017a). achieve a core tenet of ERM, that of being et al. 2016a). Forest Service managers face Here, we expand upon these themes proactive rather than reactive, by more fully regulatory barriers that inhibit proactive by drawing insights from enterprise risk characterizing the environment in which mitigation in advance of fires (e.g., air qual- management (ERM) and systems think- fire management decisions are made and ity concerns limiting controlled burning) ing (ST) in order to examine the manage- anticipating factors that may lead to com- and during a fire event are often pushed and ment organizations responsible for making promised decision making. pulled by conflicting forces (e.g., sociopolit- and implementing wildland fire response In this article, we discuss how cog- ical pressures to “put the fire out,” opposing decisions. Essential aims and processes of nitive, institutional, and cultural factors natural resource objectives). When facing ERM include (1) ensuring that strategy and influence fire response decisions and ask time pressures, managers may rely on quick objectives are aligned with organizational whether the existing system sets up man- intuitive judgments that inadequately ana- mission, vision, and core values; (2) ensur- agers for success in an increasingly com- lyze a situation or overlook relevant infor- ing that decision makers are well-equipped plex fire environment. Contextually, we mation (Wilson et al. 2011; Hand et al. to effectively manage risks and capital- focus on the Forest Service, US Department 2015). These concerns are compounded ize on opportunities; and (3) embedding of Agriculture, which is one of the largest given the dynamic and unpredictable assessment, planning, monitoring, control, wildland fire management organizations in nature of wildfire incidents, which renders learning, and continual improvement into the world. Although most fire management intuitive judgements unlikely to be consist- the fabric of the organization (US General organizations at state and county levels in ently reliable (Kahneman and Klein 2009). Accountability Office 2016; Committee of the United States have similar challenges in Cognitive biases combined with misaligned Sponsoring Organizations of the Treadway identifying and implementing effective sup- incentives and other factors often lead Commission 2017). ERM requires that pression response strategies, their directives to excessive use of suppression resources, organizations systematically evaluate risks, are to suppress fires as small and as quickly with accompanying increases in responder recognize the systemic nature of some risks, and develop commensurate capacity for risk-informed decision making, especially Management and Policy Implications for high-impact decisions (Thompson et al. 2016a). Changing responses to unplanned ignitions provides a largely untapped but important, if not essential, We use ST as a lens to understand sys- opportunity to restore landscape conditions and reduce future risk. Effectuating this change in fire man- temic forces driving fire manager behaviors ager behavior is challenging because ambiguity and incomplete information surround issues of responder and the systemic risks that those behaviors safety, suppression effectiveness, and performance measurement. We propose that by more rigorously present (Sterman 2001; Meadows and Wright researching suppression actions and refocusing on evidence rather than intuition as the basis for manage- 2008; Collins et al. 2013). In Figure 1, we ment decisions, the US Forest Service could better understand and improve the quality of its management introduce some key themes from ST that operations. By capitalizing on recent advances in risk management acumen and capacity, the US Forest motivate and guide our discussion. Our Service and the broader fire management community can achieve a vision for fire management in the stylized depiction of a coupled human and 21st century where decisions and actions are risk-informed, evidence-based, enriched with analytics, and natural fire-prone system, although simpler aligned with long-term objectives. than others (e.g., Spies et al. 2014), explicitly

Journal of Forestry • July 2018 383 exposure and suppression expenditures managers operating within this system may on individual events are seen, but generally (Thompson 2014). seem rational given their perceived risks and unseen are the patterns of behavior, system Our discussion begins from the insight rewards, even if these decisions from the structure, and mental models that underlie that these observed behaviors are a natural outside-in appear inconsistent with broader those events (Monnat and Gannon 2015). result of the Forest Service’s fire manage- agency objectives (Donovan and Brown That is, the major drivers of behavior com- ment system, its culture, values, organi- 2005). To emphasize these points, we apply prise the largest part of the iceberg that is zational structure, core assumptions and the “iceberg model” to the fire management below water. Daylighting these drivers can beliefs, and operating norms and protocols system (Figure 2). Like an iceberg whose be difficult and even uncomfortable but (Thompson et al. 2015; Thompson et al. bulk is typically unseen below the surface of often provides the greatest leverage for 2017b). In other words, the decisions of fire the water, here the analogy is that behavior learning and improvement, especially when

Figure 1. Key themes of systems thinking, along with examples and conceptual representations relevant to wildland fire management.

Figure 2. “Iceberg” model applied to fire management system behavior (modified from Thompson et al. 2017b).

384 Journal of Forestry • July 2018 ingrained ways of thinking and acting are model perpetuates a default response based appropriate; we will return to this point challenged (Meadows and Wright 2008). on automatic intervention (i.e., a bias for later in the article. Simply put, fundamental changes in action), which often drives a “do whatever how the US fire management community it takes” mentality to limit potential neg- What Does Safe Response thinks about, learns from, plans for, and ative impacts with few constraints placed Look Like? responds to wildland fires may be neces- on the magnitude of suppression response The Forest Service initiated a series of sary to close the gap between intentions (Calkin et al. 2015). One of the potentially Life First engagements (US Forest Service and actions. We propose that reevaluating undesirable effects of emergency response 2017b) intended to expand dialogues the fire management system is a necessary is a tendency to drift into a “sunk cost” and highlight areas of concern regarding first step to identifying leverage points that situation, thereby continuing to expend responder safety. It is a subject infused with could achieve desirable and sustainable resources to achieve an unachievable out- controversy and emotion, which is embod- changes in behavior. We do not propose come (McLennan et al. 2006). ied in the following quote from Mark that addressing internal factors alone will A possible alternative to emergency Smith’s essay, “The Big Lie” (Smith 2016): be sufficient to foster more resilient land- response is incremental intervention on the “The lie that wildland fire firefighting is scapes and communities; rather, we stress basis of changing conditions and accompa- safe . . . The lie is so insidious that it per- that any comprehensive strategy for change nying risks. Incremental analysis can insert meates the thinking of many fire manag- ought to critically examine the fire man- pauses in the evaluation of a situation and ers and agency administrators to the point agement system from within. We present help attenuate a sunk cost bias. In areas of denial, despite a steady flow of coffins a series of interrelated questions with the where imminent threats to public safety do standing as evidence to the contrary . . . If intention of initiating a broader dialogue not exist (i.e., much of the forested land- the [interagency policy] definition of safety around the current and future state of scapes managed by the Forest Service) or in is meaningless, and in contravention of its wildland fire management in the western circumstances where probability of success true nature, so too will be all the policies, United States. is very low, the default could become non- rules, and checklists that flow from it.” Is Every Fire an Emergency? intervention. Managers would be required There is a clear line of sight between to successively analyze the conditions under this perspective and the well-known Can We Change Default which incrementally increasing levels of Standard Firefighting Orders, which adorn Responses? intervention are justified, given the values at the back cover of Incident Response Pocket How we understand situations in which we risk and the exposure of incident respond- Guides carried by fire responders and which find ourselves and how we typically respond ers to the inherent risks of wildland fire. It are unequivocally stated to apply to all fire to those situations can exert significant influ- changes the assumption from one of aggres- situations (US Forest Service 2017c). Of ences on decision making, especially when sive action (that leads to analyses to justify particular importance is the 10th order, facing complex choices (Johnson-Laird why not to intervene) to one of intervention which states: “fight fire aggressively, hav- 2010). This can result in the “status quo only if needed. We recognize that there will ing provided for safety first.” Parsing this bias,” wherein individuals revert to default be locations and conditions under which order raises questions of inconsistency sur- behaviors even though those behaviors may emergency response is absolutely essential, rounding guidance and messaging. Is always be inappropriate given the circumstances, especially where communities and critical fighting fire aggressively consistent with the a phenomenon that has been documented infrastructure are threatened, but at the agency’s rhetoric and stated fire manage- among fire managers (Wilson et al. 2011). same time, there is ample reason to believe ment objectives? Is the notion that safety The default response to an unplanned igni- the Forest Service has not taken advantage is something that can be “provided for” in tion has tended to be aggressive suppression of historical opportunities for more flexible the operational fire environment consistent to contain the fire as quickly and as small response (Thompson et al. 2013b; North with risk management principles? Ensuring as possible, and when that fails during peri- et al. 2015a). near absolute responder safety would mean ods of extreme fire weather, to mobilize and Resetting default responses could be putting individuals nowhere near the flam- deploy additional suppression resources a powerful tool for changing behavior and ing front and rarely, if ever, engaging fires. even when control opportunities may be would ideally expand the decision space of This would undoubtedly lead to increased severely restricted. fire managers while dampening tendencies risks to the public, natural resources, and This pattern of behavior can be toward fire exclusion and the minimization property in some cases. A more actionable explained in part from a perspective that of short-term risks over longer-term risks. perspective is to conceive of safety in terms views wildfire suppression as emergency Similarly, reframing how information is pre- of acceptable or tolerable risks, that is, a response. Applying the model of emer- sented (i.e., describing expected fatality rates balancing of the gains from actions with gency response carries an accompanying set instead of duration of suppression activity) the risks that taking those actions incur. of assumptions, such as a need to be deci- could induce behaviors more consistent In some situations, engagement may be sive, a need to take immediate action, that with safety-related objectives (Hand et al. prohibitively dangerous regardless of con- time is of the essence, that action is better 2015). Importantly, the premise of shifting sequences; in others, higher levels of risk than analysis, and that life is imminently at baselines to nonintervention requires that might be tolerated, for instance, to protect risk. In some cases, this may be an accurate managers proactively differentiate landscape irreplaceable resources or to buy time for interpretation of the situation, but in some locations and fire weather conditions under evacuations. This complex balancing act cases it may not. The emergency response which incremental intervention would be is articulated well by the direction former

Journal of Forestry • July 2018 385 Forest Service Chief Tom Tidwell provided federal reporting systems (K. Short, per- and improvement fail to keep pace with heading into the 2017 fire season: “Our sonal communication). Tabulating this the increasing risk and complexity of the strategies and tactics commit responders to metric on the basis of all fires that received fire environment. By more rigorously operations where and when they can be suc- a response assumes in effect that all fires researching suppression actions and refo- cessful and under conditions where import- are unwanted, seemingly in conflict with cusing on evidence rather than intuition ant values at risk are protected with the least Federal Wildland Fire Policy stating that as the basis for management decisions, exposure necessary . . .” (US Forest Service “wildland fire will be used to protect, main- the agency could better understand and 2017b). Thus, exposing responders to haz- tain, and enhance resources and, as nearly as improve the quality of its management ards in order to protect low-value resources possible, be allowed to function in its natu- operations. while perpetuating feedbacks that increase ral ecological role” (Fire Executive Council Only through development of more long-term risks might be beyond the pale. 2009). This appears to create a powerful robust systems of accountability and per- However, implementing the former chief’s cognitive dissonance that possibly perpet- formance measurement can effective behav- direction requires clarity and agreement on uates the fire exclusion paradigm and the iors be identified, evaluated, and rewarded. what types of risks are acceptable and what model of emergency response. That is, there An objective and attainable approach to are not. Does such clarity exist? Do existing may be intrinsic problems with how initial performance measurement could begin by operational tools for assessing safety (e.g., attack is conceptualized and executed. The examining environmental conditions that the job hazard analysis) begin with evaluat- Frog Fatality Report (US Forest Service influence probability of success. It has been ing the worthiness of the mission’s objective 2016) aptly summarizes the situation as fol- established that under extreme fire weather with respect to operational risks, or do they lows: “In the current wildland firefighting conditions, certain types of suppression begin with accepting the mission and then culture, aggressive initial attack is highly operations are of little use, exceedingly working through the safety implications? valued and is generally the standard oper- expensive, and can be highly dangerous. Building from the ideas introduced earlier, ating procedure . . . . Firefighters currently Thus, metrics could, for instance, examine could responders instead avoid engagement see themselves as most successful when they the number and type of assignments lead- as the norm, only initiating suppression catch the fire and feel like they’ve failed ing to active suppression operations under actions when a clear set of signals exists when they lose a fire.” extreme weather conditions, with a target indicating that actions are likely to be suc- Questions similarly exist about sup- of minimizing these types of assignments. cessful, have a purpose, and limit exposure pression effectiveness in the more complex Measures could be downscaled to focus on to hazards? large fire environment (i.e., for those rare specific mission types, for example, evalu- fires that escape initial attack efforts or are ating aerial retardant drops by factors like What Does Effective Response intentionally managed from detection as wind, topography, and fuel type (Stonesifer a resource benefit fire—the latter is even et al. 2016). Similarly, evaluation of con- Look Like? How Would rarer). Research findings indicate that fire structed fire control lines could proceed We Know? weather can be the dominant influence on on the basis of whether they engaged the The fire community has struggled to containment probability, that managers fire and were effective in restricting fire define meaningful and actionable perfor- exhibit wide variability in use of suppres- spread, and if not, why not (Thompson mance measures for some time (Booz Allen sion resources when responding to fires et al. 2016b). Such measures could help Hamilton 2015). Perhaps the starkest exam- with similar characteristics, that manag- improve understanding the conditions ple is the oft-cited metric that initial attack ers commit substantial levels of resources under which actions are likely to be suc- to unplanned ignitions on federal lands after fire growth has largely ceased, and cessful, while also shedding light on is successful 95–98 percent of the time. that suppression resources are often used upstream decisions about the timing and Peering behind the curtain, however, reveals outside of conditions where use is likely to location of fire-control activities. Ideally, some illuminating insights. First, natural be effective (Finney et al. 2009; Stonesifer these measures would inform and incen- constraints on fire size (e.g., the episodic et al. 2016; Hand et al. 2017; Katuwal tivize fire managers to restrict engagement nature of weather events conducive to high et al. 2017). A critical take is that default under conditions where success is unlikely, fire-spread potential) means that there is responses are possibly ineffective during thereby reducing unnecessary exposure of some nontrivial baseline level of fires that extreme fire weather and possibly counter- fire responders. would have extinguished absent human productive during moderate fire weather Is there a future where the absence intervention (M. Finney, personal commu- (while recognizing that quiescent peri- rather than the presence of active suppres- nication). Second, choosing to extinguish ods of weather may provide critical win- sion is rewarded? Could fire managers face those fires that might have spread under dows for containment in advance of more positive incentives related to the number nonextreme conditions forecloses opportu- extreme weather). of days where suppression resources were nities to yield ecological benefits, reinforc- The aforementioned patterns of beha- not engaged due to weather, the length ing the fire paradox ( Calkin et al. 2015; vior are sustained in part by insufficient of perimeter where direct attack was not Barnett et al. 2016b). Third, and perhaps monitoring and investigation of fire man- pursued, or other similar constructs? Is most important, this metric is effectively agement system behavior. Even where it possible to, in effect, reverse engineer premised on the differentiation of fires on possibly ineffective behaviors can be iden- performance measures that subtract away the basis of “wanted” versus “unwanted,” tified, the lack of effective feedback control unnecessary exposure and maximize prob- a classification that is nonexistent in any mechanisms means that system learning ability of success?

386 Journal of Forestry • July 2018 Can Recent Risk Management capitalize on opportunities to manage fire played out on the Tonto National Forest Successes Be a Springboard for for resource benefit. It would lead to the in Arizona in the 2017 fire season, as illus- Improvement? recognition that the practice of risk manage- trated in Figure 3. The top panel shows stra- ment applies not just to suppression activi- tegic response zones categorized according What would adopting ERM look like in the ties but across the organization (and further to risk-based management objectives. These Forest Service? To begin with, the agency extends to communication and comanage- categories are practical simplifications of a would look at the systems it has in place ment with communities, stakeholders, and response continuum from resource protec- in a new way, using ST concepts and lan- partners). tion to resource benefit and are intended to guage to look across functional areas and Fortunately, there are emerging risk facilitate response decisions. These catego- identifying instances where systems inter- management success stories that the agency ries are not binding and in no way diminish act to product unintended and undesirable could build upon and that directly address the flexibility or decision space of managers. consequences when viewed from a larger several of the issues raised earlier in this They do, however, reset defaults in such a systemic perspective, despite working fairly article related to improving decisions and way that aggressive initial attack may no well within their own “locally” defined per- outcomes. Perhaps the most relevant exam- longer be the status quo option in zones spectives. More concretely, it would lead ple is more rigorous prefire assessment and where management objectives call for the to the recognition that escalating costs of planning, which helps fire managers strat- use of fire. suppression and degrading forest conditions ify the landscape on the basis of risk and This example also illustrates how pose risks to the core mission of sustaining predetermine acceptable response strate- improved planning and monitoring can the nation’s forests and grasslands. It would gies and operational priorities accordingly enhance responder safety and performance force the agency to critically evaluate how it (O’Connor et al. 2016; Thompson et al. measurement. The boundaries of each stra- could better enhance responder safety and 2016d; Dunn et al. 2017). This scenario tegic response zone are potential fire-control locations (e.g., roads, ridgetops), the identi- fication of which facilitates deployments of suppression resources to where they can be safer and more effective. The bottom panel displays daily fire progression contours in relation to pre-identified potential control locations, along with maps of constructed fire-control lines (“hand line”), for the 2017 Pinal Fire. Local fire managers were able to leverage these pre-identified control locations to check fire spread in accordance with land and resource objectives (K. O’Connor, personal communication). The collection of accurate geospatial data on fire-control line construction allows for post hoc evaluation of control line effectiveness, ideally leading to more targeted and discriminating decisions of where to send suppression resources to minimize unnecessary exposure. Two logical next steps are expanding capacity for widespread use of this type of risk-based planning and disseminating incremental intervention as a model for engagement through training exercises. A last point on ERM relates to being anticipatory of future challenges and road- blocks associated with such a strategic, systemic, organizational redirection. It will be important to discuss with candor the full spectrum of near- and long-term tradeoffs that may arise. It could be the case, for exam- Figure 3. Demonstration of how new analysis tools can help managers more systematically ple, that suppression expenditures increase plan for and evaluate suppression response. Specific information highlighted in this figure in the near-term, as fewer fires are rapidly includes strategic response zones categorized according to risk-based management objec- extinguished and more fires become longer tives, pre-identified fire control locations, and daily fire spread of the 2017 Pinal Fire, on duration events. The overall expectation the Tonto National Forest in Arizona, US image and analysis provided by Dr. Christopher however may be that the trend will even- D O’Connor. tually reverse as both landscape condition

Journal of Forestry • July 2018 387 Reacve response to unplanned ignions → Proacve predetermined range of responses to unplanned ignions under diﬀerent scenarios

All fires require emergency response → Some fires require emergency response, under locaons and condions that can be pre-idenfied

Minimize or exclude disturbance → Capitalize on and leverage disturbance

Jusfy avoiding intervenon → Jusfy intervenon

Reward on the basis of outcomes → Reward on the basis of quality of decisions

Scarcity governs suppression resource use → Risk-risk tradeoﬀ analysis governs suppression resource use

Intuive experse governs decision processes → Proacve assessment and planning governs decision processes

Figure 4. Vision for fire management in the 21st century, expressed as transition arcs from the status quo to desired endpoints. and system behavior set the stage for mod- agency, investments in workforce education engagement and extinguishment of wild- erated suppression response. Similarly, and capacity, and the adequacy of available fire, capacity will be built (and rewarded) increased near-term exposure to smoke information, all with respect to responsi- around deployments of force instead of may be problematic for public health but bilities imposed upon and pressures faced deployments of planning, analysis, and must be evaluated in light of the observa- by managers. Emphasis on acceptability of mitigation; long-term risks will likely tion that suppression cannot eliminate but risk, mindful engagement, and justification continue to grow. Barring major change, rather defers and possibly intensifies smoke of actions based on value and probability the next generation of fire managers may exposure (Schweizer and Cisneros 2017). As of success would necessarily infuse such inherit the same ways of thinking, corre- managers attempt to judiciously expand the analyses. sponding competencies, and capacities that footprint of ecologically beneficial fire, there Aligning performance measures, incen- have led in part to the problems the agency is a nonzero chance that fires could escape tives, workforce capacity, and information faces today. Hence, the call to rethink a containment efforts and cause damage. This systems with the agency’s intentions around currently flawed wildland fire management underscores the importance of pre-iden- fire will require a clearer articulation of a system. tification of potential control locations to vision for fire management in the 21st cen- facilitate more rapid containment when tury (Figure 4). Embracing this vision rep- Literature Cited warranted to preclude fire from spreading resents significant cultural shifts, challenging Barnett, K., S.A. Parks, C. Miller, and to threaten communities and infrastructure. notions that fire must be fought, that fire H.T. Naughton. 2016a. Beyond fuel treat- can be eliminated or at least minimized, and ment effectiveness: Characterizing interac- The Road Ahead: Continual that overwhelming displays of suppression tions between fire and treatments in the US. Improvement response are the safest or the least costly Forests. 7(10):237. doi:10.3390/f7100237. Barnett, K., C. Miller, and T.J. Venn. 2016b. Improving the safety and effectiveness of option in the long-term (e.g., Houtman Using risk analysis to reveal opportunities response to each and every wildfire is essen- et al. 2013). It may require taking a hard for the management of unplanned igni- tial in pursuit of broader socioecological look at the existing organizational structures tions in wilderness. J. For. 114(6):610–618. objectives. To achieve this, it will be import- surrounding planning and decision-making doi:10.5849/jof.15-111. ant to go “upstream” in the Forest Service processes, evaluating whether the precondi- Boisramé, G.F., S.E. Thompson, M. Kelly, J. Cavalli, K.M. Wilkin, and to identify broader opportunities for mean- tions are in place for the incident response S.L. Stephens. (2017). Vegetation change ingful change, and in doing so, revisiting system to effectively function (Jensen and during 40 years of repeated managed wild- how organizational culture, structure, and Thompson 2016), and ensuring that tacti- fires in the Sierra Nevada, California. Forest. other factors may drive behaviors inconsis- cal and operational decisions are enriched Ecol. Manag. 402:241–252. doi:10.1016/j. tent with strategic objectives and the agen- with analytics and aligned with long-term foreco.2017.07.034. Booz Allen Hamilton. 2015. 2014 Quadrennial cy’s mission. As a starting point, the Forest agency objectives. It will also require recog- Fire Review: Final Report. Available at: https:// Service could introduce periodic analyses nizing that, in an uncertain world, bad out- www.forestsandrangelands.gov/QFR/docu- of doctrine and policy (US Forest Service comes can result from well-made decisions, ments/2014QFRFinalReport.pdf; last accessed 2017d) and their relation to current goals and vice versa; supporting the former while September 8, 2017. and intentions (US Forest Service 2005). avoiding reinforcing the latter are critical to Calkin, D.E., M.P. Thompson, and M.A. Finney. 2015. Negative consequences Such a review could evaluate the transpar- effectively manage risks over time. of positive feedbacks in US wildfire manage- ency and accountability of fire managers, Insofar as cultural and institutional ment. For. Ecosyst. 2(1):1–10. doi:10.1186/ how risks are shared across levels of the factors remain aligned around rapid s40663-015-0033-8.

388 Journal of Forestry • July 2018 Collins, R.D., R. de Neufville, J. Claro, Johnson-Laird, P.N. 2010. Mental mod- and opportunities. Geosciences. 6(3):35. T. Oliveira, and A.P. Pacheco. 2013. Forest els and human reasoning. Proc. Natl. Acad. doi:10.3390/geosciences6030035. fire management to avoid unintended conse- Sci. USA. 107:18243–18250. doi:10.1073/ Olson, R.L., D.N. Bengston, L.A. DeVaney, quences: A case study of Portugal using sys- pnas.1012933107. and T.A.C. Thompson. 2015. Wildland Fire tem dynamics. J. Environ. Manage. 130:1–9. Kahneman, D., and G. Klein. 2009. Management Futures: Insights from a Foresight doi:10.1016/j.jenvman.2013.08.033. Conditions for intuitive expertise: A fail- Panel. Gen. Tech. Rep. NRS-152. US Committee of Sponsoring Organizations of ure to disagree. Am. Psychol. 64(6):515. Department of Agriculture, Forest Service, the Treadway Commission. 2017. Enterprise doi:10.1037/a0016755. Northern Research Station, Newtown Square, Risk Management: Integrating with Strategy and Kalies, E.L., and L.L.Y. Kent. 2016. Tamm PA. 44 p. Performance. Available at: https://www.coso.org/ Review: Are fuel treatments effective at achiev- Otero, I., and J.Ø. Nielsen. 2017. Coexisting Documents/2017-COSO-ERM-Integrating- ing ecological and social objectives? A system- with wildfire? Achievements and challenges with-Strategy-and-Performance-Executive- atic review. Forest. Ecol. Manag. 375:84–95. for a radical social-ecological transformation Summary.pdf; last accessed October 25, 2017. doi:10.1016/j.foreco.2016.05.021. in Catalonia (Spain). Geoforum. 85:234–246. Curt, T., and T. Frejaville. 2017. Wildfire Katuwal, H., C.J. Dunn, and D.E. Calkin. doi:10.1016/j.geoforum.2017.07.020. policy in Mediterranean France: How far is 2017. Characterising resource use and poten- Parks, S.A., L.M. Holsinger, C. Miller, it efficient and sustainable? Risk Anal. doi: tial inefficiencies during large-fire suppres- and C.R. Nelson. 2015. Wildland fire as 10.1111/risa.12855. sion in the western US. Int. J. Wildland. Fire. a self‐regulating mechanism: The role of Donovan, G.H., and T.C. Brown. 2005. An 26(7):604–614. doi:10.1071/WF17054. previous burns and weather in limiting fire alternative incentive structure for wildfire Monat, J.P., and T.F. Gannon. 2015. What is progression. Ecol Appl. 25(6):1478–1492. management on national forest land. For. Sci. systems thinking? A review of selected liter- doi:10.1890/14-1430.1. 51(5):387–395. ature plus recommendations. Am. J. Sys. Sci. Parks, S.A., C. Miller, L.M. Holsinger, Dunn, C.J., D.E. Calkin, and M.P. Thompson. 4(1):11–26. doi:10.5923/j.ajss.20150401.02. L.S. Baggett, and B.J. Bird. 2016. 2017. Towards enhanced risk management: McIver, J.D., S.L. Stephens, J.K. Agee, Wildland fire limits subsequent fire occur- Planning, decision making and monitoring et al. 2013. Ecological effects of alterna- rence. Int. J. Wildland. Fire. 25(2):182–190. of US wildfire response. Int. J. Wildland. Fire. tive fuel-reduction treatments: Highlights of doi:10.1071/WF15107. 26(7):551–556. doi:10.1071/WF17089. the National Fire and Fire Surrogate study Plucinski, M.P., and E. Pastor. 2013. Criteria Finney, M., I.C. Grenfell, and C.W. McHugh. (FFS). Int. J. Wildland. Fire. 22(1):63–82. and methodology for evaluating aerial wild- 2009. Modeling containment of large wildfires doi:10.1071/WF11130. fire suppression. Int. J. Wildland. Fire. using generalized linear mixed-model analysis. McLennan, J., A.M. Holgate, M.M. Omodei, 22(8):1144–1154. doi:10.1071/WF13040. For. Sci. 55(3):249–255. and A.J. Wearing. 2006. Decision making Prichard, S.J., C.S. Stevens-Rumann, and Fire Executive Council. 2009. Guidance for effectiveness in wildfire incident management P.F. Hessburg. 2017. Tamm Review: Shifting Implementation of Federal Wildland Fire teams. J. Contingencies Crisis Manag. 14(1):27– global fire regimes: Lessons from reburns and Policy. Available at: https://www.nifc.gov/ 37. doi:10.1111/j.1468-5973.2006.00478.x. research needs. Forest. Ecol. Manag. 396:217– policies/policies_documents/GIFWFMP.pdf; Meadows, D., and D. Wright. 2008. Thinking 233. doi:10.1016/j.foreco.2017.03.035. last accessed October 25, 2017. in systems: a primer. Chelsea Green Publishing, Regos, A., N. Aquilué, J. Retana, M. De Fischer, A.P., T.A. Spies, T.A. Steelman, White River Junction, VT USA. Cáceres, and L. Brotons. 2014. Using et al. 2016. Wildfire risk as a socioecological Meyer, M.D., S.L. Roberts, R. Wills, unplanned fires to help suppressing future pathology. Front. Ecol. Environ. 14(5):276– M. Brooks, and E.M. Winford. 2015. large fires in mediterranean forests. PLOS 284. doi:10.1002/fee.1283. Principles of effective USA federal fire man- ONE. 9(4):e94906. doi:10.1371/journal. Hand, M.S., M.J. Wibbenmeyer, D.E. Calkin, agement plans. Fire Ecol. 11(2). doi:10.4996/ pone.0094906. and M.P. Thompson. 2015. Risk prefer- fireecology.1102059. Schoennagel, T., J.K. Balch, H. Brenkert- ences, probability weighting, and strategy Moritz, M.A., E. Batllori, R.A. Bradstock, Smith, et al. 2017. Adapt to more wild- tradeoffs in wildfire management. Risk Anal. et al. 2014. Learning to coexist with wildfire in western North American forests 35(10):1876–1891. doi:10.1111/risa.12457. fire. Nature. 515(7525):58–66. doi:10.1038/ as climate changes. Proc. Natl. Acad. Sci. Hand, M., H. Katuwal, D.E. Calkin, and nature13946. USA. 114(18):4582–4590. doi:10.1073/ M.P. Thompson. 2017. The influence of inci- National Interagency Fire Center. 2017. Red pnas.1617464114. dent management teams on the deployment of Book 2017. Interagency Standards for Fire and Schweizer, D.W. and R. Cisneros. 2017. wildfire suppression resources. Int. J. Wildland. Fire Aviation Operations. Available at: https:// Forest fire policy: change conventional think- Fire. 26(7):615–629. doi:10.1071/WF16126. www.nifc.gov/policies/pol_ref_redbook.html; ing of smoke management to prioritize long- Houtman, R.M., C.A. Montgomery, last accessed October 17, 2017. term air quality and public health. Air Qual A.R. Gagnon, et al. 2013. Allowing a wildfire North, M., B.M. Collins, and S. Stephens. Atmos Health. 10(1):33–36. doi:10.1007/ to burn: Estimating the effect on future fire sup- 2012. Using fire to increase the scale, bene- s11869-016-0405-4. pression costs. Int. J. Wildland. Fire. 22(7):871– fits, and future maintenance of fuels treat- Schwilk, D.W., J.E. Keeley, E.E. Knapp, et al. 882. doi:10.1071/WF12157. ments. J. For. 110(7):392–401. doi:10.5849/ 2009. The national Fire and Fire Surrogate Hessburg, P.F., T.A. Spies, D.A. Perry, jof.12-021. study: Effects of fuel reduction methods on et al. 2016. Tamm review: Management of North, M., S. Stephens, B. Collins, et al. forest vegetation structure and fuels. Ecol Appl. mixed-severity fire regime forests in Oregon, 2015a. Reform forest fire management. 19(2):285–304. doi:10.1890/07-1747.1. Washington, and Northern California. Forest. Science. 349(6254):1280–1281. doi:10.1126/ Smith, M. 2016. Honor the Fallen Essay – The Ecol. Manag. 366:221–250. doi:10.1016/j. science.aab2356. Big Lie. Available at: http://wildlandfire- foreco.2016.01.034. North, M., A. Brough, J. Long, et al. 2015b. leadership.blogspot.com/2016/06/the-big- Ingalsbee, T. 2017. Whither the paradigm shift? Constraints on mechanized treatment signifi- lie.html; last accessed September 8, 2017. Large wildland fires and the wildfire paradox cantly limit mechanical fuels reduction extent Spies, T.A., E.M. White, J.D. Kline, et al. offer opportunities for a new paradigm of eco- in the Sierra Nevada. J. For. 113(1):40–48. 2014. Examining fire-prone forest land- logical fire management. Int. J. Wildland. Fire. doi:10.5849/jof.14-058. scapes as coupled human and natural sys- 26(7):557–561. doi:10.1071/WF17062. O’Connor, C.D., M.P. Thompson, and tems. Ecol Soc. 19(3):14. doi:10.5751/ Jensen, J., and S. Thompson. 2016. The Incident F. Rodríguez y Silva. 2016. Getting ES-06584-190309. Command System: A literature review. Disasters. ahead of the wildfire problem: Quantifying Steelman, T.A., and S.M. McCaffrey. 40(1):158–182. doi:10.1111/disa.12135. and mapping management challenges 2011. What is limiting more flexible fire

Journal of Forestry • July 2018 389 management—public or agency pressure? Thompson, M.P., D.G. MacGregor, and Wilson, R.S., P.L. Winter, L.A. Maguire, J. For. 109(8):454–461. doi:10.1093/ D.E. Calkin. 2016a. Risk Management: Core and T. ascher. 2011. Managing wild- jof/109.8.454. Principles and Practices, and their Relevance fire events: Risk-based decision mak- Stephens, S.L., B.M. Collins, E. Biber, and to Wildland Fire. Gen. Tech. Rep. RMRS- ing among a group of federal fire P.Z. Fulé. 2016. US federal fire and forest GTR-350. US Department of Agriculture, managers. Risk Anal. 31(5):805–818. policy: Emphasizing resilience in dry forests. Forest Service, Rocky Mountain Research doi:10.1111/j.1539-6924.2010.01534.x. Ecosphere. 7(11):19. Station, Fort Collins, CO. 29 p. US Forest Service. 2005. Foundational Sterman, J.D. 2001. System dynamics mod- Thompson, M.P., P. Freeborn, J.D. Rieck. et al. Doctrine Guiding Fire Suppression in the eling: Tools for learning in a complex 2016b. Quantifying the influence of previ- US Forest Service. Available at: https://www. world. Calif. Manag. Rev. 43(4):8–25. ously burned areas on suppression effectiveness fs.fed.us/fire/doctrine/; last accessed July 31, doi:10.2307/41166098. and avoided exposure: A case study of the Las 2017. Stevens, J.T., B.M. Collins, J.D. Miller, Conchas Fire. Int. J. Wildland. Fire. 25(2):167– US Forest Service. 2015. The Rising Cost of M.P. North, and S.L. Stephens. 2017. 181. doi:10.1071/WF14216. Wildfire Operations: Effects on the Forest Changing spatial patterns of stand-replacing fire Thompson, M.P., T. Zimmerman, D. Mindar, Service’s Non-Fire Work. Available at: https:// in California conifer forests. Forest. Ecol. Manag. and M. Taber. 2016c. Risk Terminology www.fs.fed.us/sites/default/files/2015-Fire- 406:28–36. doi:10.1016/j.foreco.2017.08.051. Primer: Basic Principles and a Glossary for the Budget-Report.pdf; last accessed October Stevens-Rumann, C.S., S.J. Prichard, Wildland Fire Management Community. Gen. 26, 2017. E.K. Strand, and P. Morgan. 2016. Prior Tech. Rep. RMRS-GTR-349. US Department US Forest Service. 2016. Frog Fire Fatality: wildfires influence burn severity of subsequent of Agriculture, Forest Service, Rocky Mountain Learning Review Report. Available at: https:// large fires. Can. J. Forest. Res. 46(11):1375– Research Station, Fort Collins, CO. 13 p. www.wildfirelessons.net/orphans/viewinci- 1385. doi:10.1139/cjfr-2016-0185. Thompson, M.P., P. Bowden, A. Brough, et al. dent?DocumentKey=e44eeaab-e525-4174-9 Stonesifer, C.S., D.E. Calkin, M.P. 2016d. Application of wildfire risk assess- e72-60a54cd89ffd; last accessed September Thompson, and K.D. Stockmann. 2016. ment results to wildfire response planning 8, 2017. Fighting fire in the heat of the day: An anal- in the Southern Sierra Nevada, California, US Forest Service. 2017a. Fiscal Year 2018 ysis of operational and environmental con- USA. Forests. 7(3):64. doi:10.3390/ Budget Justification. https://www.fs.fed.us/ ditions of use for large airtankers in United f7030064. sites/default/files/usfs-fy18-budget-justifica- States fire suppression. Int. J. Wildland. Fire. Thompson, M.P., F. Rodríguez y Silva, tion.pdf; last accessed October 25, 2017. 25(5):520–533. doi:10.1071/WF15149. D.E. Calkin, and M.S. Hand. 2017a. A US Forest Service. 2017b. Chief’s Letter of Thompson, M.P., D.E. Calkin, M.A. Finney, review of challenges to determining and Intent for Wildland Fire – 2017. Available K.M. Gebert, and M.S. Hand. 2013a. A demonstrating efficiency of large fire manage- at https://www.nifc.gov/PIO_bb/Agencies/ risk-based approach to wildland fire budgetary ment. Int. J. Wildland. Fire. 26(7):562–573. USFS/ChiefLetterofIntent.pdf; last accessed planning. For. Sci. 59(1):63–77. doi:10.5849/ doi:10.1071/WF16137. October 26, 2017. forsci.09-124. Thompson, M.P., C.J. Dunn, and D.E. Calkin. US Forest Service. 2017c. Standard Firefighting Thompson, M.P., C.S. Stonesifer, R.C. Seli, 2017b. Systems Thinking and Wildland Fire Orders and 18 Watchout Situations. https:// and M. Hovorka. 2013b. Developing stan- Management. In Proceedings of the 60th www.fs.fed.us/fire/safety/10_18/10_18.html; dardized strategic response categories for fire Annual Meeting of the ISSS-2016. Boulder, last accessed July 31, 2017. management units. Fire Management Today. CO, USA (Vol. 1, No. 1). 17 p. US Forest Service. 2017d. Forest Service Manual 73(1):18–24. Tidwell, T. 2015. Statement of Tom Tidwell, (FSM) Directive Issuances, Series 5000 – Thompson, M.P. 2014. Social, institutional, Chief, Forest Service, United States Protection and Development. Available at: and psychological factors affecting wildfire Department of Agriculture, Before the https://www.fs.fed.us/im/directives/dughtml/ incident decision-making. Soc. Nat. Resour. Committee on Agriculture, Conservation fsm_5000.html; last accessed October 26, 2017. 27(6):636–644. doi:10.1080/08941920.201 and Forestry Subcommittee, United States US General Accountability Office. 2016. 4.901460. House of Representatives. October 8, 2015. Enterprise Risk Management: Selected Thompson, M., C. Dunn, and D. Calkin. Available at: https://www.fs.fed.us/sites/ Agencies’ Experience Illustrate Good Practices 2015. Wildfires: Systemic changes required. default/files/media/types/testimony/Oct-8- in Managing Risk. GAO-17–63. Available at: Science. 350(6263):920. doi:10.1126/science. Testimony-HAgC-Fire-Forest-Management. http://www.gao.gov/assets/690/681342.pdf; 350.6263.920-b. pdf; last accessed September 8, 2017. last accessed October 26, 2017.

390 Journal of Forestry • July 2018