Adaptive Management: A Paradigm for Remediation of Public Facilities Following a Terrorist Attack
Abstract
Terrorist actions are aimed at maximizing harm (health, psychological, economical, and political) through the combined physical impacts of the act and fear. Immediate and effective response to a terrorist act is critical to limit human and environmental harm, effectively restore facility function, and maintain public confidence. Though there have been terrorist attacks in public facilities that we have learned from, overall our experiences in restoration of public facilities following a terrorist attack are limited. Restoration of public facilities following a release of a hazardous material is inherently far more complex than in industrial settings and has many unique technical, economic, social, and political challenges. For example, there may be a great need to quickly restore the facility to full operation and allow public access even though it was not designed for easy or rapid restoration, and critical information is needed for quantitative risk assessment and effective restoration must be anticipated to be incomplete and uncertain. Whereas present planning documents have substantial linearity in their organization, the “adaptive management” paradigm provides a constructive parallel paradigm for restoration of public facilities that anticipates and plans for uncertainty, inefficiencies, and stakeholder participation. Adaptive management grew out of the need to manage and restore natural resources in highly complex and changing environments with limited knowledge about causal relationships and responses to restoration actions. Similarities between natural resource management and restoration of a public facility after a terrorist attack suggest that integration of adaptive management principles explicitly into restoration processes will result in substantially enhanced and flexible responses necessary to meet the uncertainties of potential terrorist attacks.
1. INTRODUCTION
The probability of terrorist actions in the United States and the magnitude of the potential consequences have increased dramatically in the past several years.(1) Potential threats include attacks on the American public through the use of chemical, biological, or radiological agents. For maximum effect, a likely scenario could involve terrorists releasing hazardous agent(s) into a public facility to maximize fatalities and causing political, social, and economic instability.
The Department of Homeland Security (DHS) has the responsibility to counter these new and evolving threats, to protect the public, and, when necessary, to help manage an effective response and recovery.(2, 3) While a great amount of research and technology development has gone into preventive actions that reduce the probability of such attacks, it is prudent to be prepared to respond to an event should one occur. Therefore, it is essential that an effective management framework is in place to provide for rapid and complete restoration of the affected domestic facilities, despite the unique challenges.(3-6) The management framework must be able to effectively address the environmental and technical complexity of the challenge and the uncertainty associated with the restoration tools at our disposal.
This article describes an alternative management framework called “adaptive management,” and shows how applying this management paradigm might enhance efficient restoration of affected facilities. Historically, adaptive management principles have been primarily applied for management of natural resources and environmental restoration,(7,8) but the many similarities between restoration of disturbed environments and restoration of attacked public facilities suggest that inclusion of adaptive management principles could prove valuable. Specifically, we describe the challenges associated with restoration of public facilities after a terrorist attack, discuss and contrast adaptive management in the context of the more traditional “linear” approach to restoration, and provide examples from restoration experiences following attacks. In the well-documented restoration of facilities affected in the anthrax attacks, we find many aspects of an adaptive management paradigm that grew “organically.” We suggest that integration of these principles explicitly into plans, processes, and exercises will result in substantially enhanced and flexible responses necessary to meet the uncertainties of potential terrorist attacks.
2. RESTORATION CHALLENGES AFTER A TERRORIST ATTACK
A well-managed restoration can dramatically reduce the consequences of a terrorist act, both in terms of the immediate risk posed by the released agent, but also on a broader scale of maintaining and building public confidence.(3,6,9–15) Therefore, beyond saving lives, it is critically important to gain public confidence following such attacks, since the press, public, and government will be carefully watching and judging the actions of the responding organizations. A well-managed restoration will safeguard public health, quickly restore the affected facilities, reassure the public through effective response, and help minimize restoration costs.
It is important to perform a restoration well, yet there is a lack of experience for restoration of public facilities following a terrorist attack relative to restoration experience following accidental releases of hazardous materials inside industrial facilities. Experience from the few chemical and biological incidents in public settings illustrate that many facilities and responding agencies lacked clear response plans that delineated and integrated roles and responsibilities among facility owners, regulators, and responders.(5) While coordination plans have recently been developed under the National Response Plan,(3) these plans are largely untested and immature for response to large-scale terrorist attacks. Similarly, specifically defined cleanup levels for some potential chemical and biological weapon agents have not been well developed or vetted for public facilities,(4–6) though important work has been done recently to develop cleanup standards for a number of chemical weapon agents.(16–18) In addition, response and restoration plans must address or be adapted to a large diversity of agents and threats, making testing and plan maturation a difficult and time-intensive exercise.
Stakeholders involved in response to terrorist attack on a public area will likely be diverse and include facility owners, multiple government agencies, and numerous public groups. These groups will likely have varying views, knowledge, accessible information, and needs. Therefore, the interaction among facility owners, the public, regulators, public officials, and responders can be expected to be relatively complex and uncertain for restoration of public facilities after a terrorist attack, especially when compared to response to accidental releases at industrial facilities.
At a basic level, the challenges of restoration following a terrorist attack in a public facility result from highly complex technological, facility, social, and political environments where there could be poor information and little experience upon which to base a facility-specific restoration plan. Further, terrorist actions are expected to be designed and implemented to increase uncertainty and complexity simultaneously to successful delivery and to extend impact. Therefore, the ability of the restoration team to embrace, communicate, and use uncertainty becomes a very critical component of a management scheme for restoring facilities after a terrorist attack.
3. TRADITIONAL LINEAR MANAGEMENT PLANNING PARADIGM
Cleanup and restoration of industrial facilities that have become contaminated are generally characterized as a linear process.(3,4,11,13)Fig. 1 is a generalized schematic of a linear process categorized into six sequential phases. Phase 1 describes immediate responses following recognition of the accidental release, which generally are performed to limit the exposure to personnel, to provide medical treatment for those severely exposed, and to limit spread of the release. The extent and nature of the contamination are characterized in Phases 2 and 3, with sampling in Phase 3 providing extensive quantitative information for detailed decontamination planning. Following characterization of the contamination, identified areas are decontaminated in Phase 4 and the cleanup verified in Phase 5. Reoccupation of the facility occurs in Phase 6, and verification of cleanup through long-term monitoring is an option to ensure the effectiveness of the cleanup over time and changing conditions.
Typical linear response to restoration of a facility following an intentional or accidental release of a contaminant into a facility.
The linear approach for cleanup of spills and restoration of a facility, as shown in Fig. 1, has proven to be effective and is recommended for use inside industrial facilities (19) due to several factors, and each factor decreases the likelihood of “surprises” during restoration. First, industrial facilities are generally designed and operated to prevent, but if needed, quickly respond in the event of, an accidental release. Ventilation, equipment designs (e.g., such as secondary and tertiary containment), easy access to spill response kits, and nonporous surfaces are all important for effective cleanup of accidental spills and are designed into these facilities. Second, the hazardous materials used in industrial facilities are well-characterized, workers are highly trained to respond to spills, have easy access to personal protective equipment (PPE) and cleaning materials, and have well-developed (and regulated) procedures for disposing of the hazardous waste generated during the cleanup process. Third, industrial facilities generally have a single user, a limited and defined hierarchy of stakeholders, a well-defined set of regulations and regulators, and restoration generally does not have to pass through a real-time public filter (regulatory development provides a priori direct public involvement). Expectations are driven by impacts to the business and worker safety and less by public and political concerns. This provides for clearly defined responsibilities, clear management structure, clear direction, proven restoration tools and processes, and unambiguous criteria for defining the success of the cleanup. Fourth, generally, restoration activities in industrial settings profit from flexible budgets to help get the facility online with minimal disruption of product flow. These factors combined make spill response in industrial settings relatively uncomplicated, effective, and efficient. Therefore, the linear management approaches can be effectively planned and implemented under this set of limited conditions.
4. LIMITATIONS TO LINEAR APPROACHES
While linear approaches have clear demonstrated success for restoration in industrial settings, their application in information/knowledge-poor settings, such as that experienced in the 2001 anthrax attacks and other chemical attacks,(5,6,20) may not provide sufficient flexibility to ensure efficient restoration of public facilities. High priority terrorist targets include public-use facilities that are not designed for handling hazardous materials, resulting in absence of capabilities for easy cleanup of contaminants (e.g., carpets, chairs, and other absorbing materials) and areas that are not easy to isolate (e.g., shared ventilation among rooms, ventilation exhausts that are not filtered, etc.). Further, initial responders may have minimal, if any, experience or knowledge of the hazardous agent(s) or the affected facility. Thus, combined with terrorist action designs, there will be additional potential for ill-defined or constrained choices and “surprises” during the response and restoration. Because confusion is an advantage to a terrorist's ends and means, we must also plan on subterfuge and additional problems that would be termed mistakes in an industrial or environmental restoration situation.
Stakeholder involvement is not always explicitly required in linear approaches to restoration. Depending on the severity of the attack, there will be numerous and varied stakeholders besides the facility owners, such as the affected public, first responders, forensic investigators (from local to federal), and political leaders.(3) Each of these stakeholders will have their own perspective on the restoration activities and goals, and extensive communication will have to be integrated for consensus on the best ways to proceed locally and globally toward the goal(s) of cleanup. In addition, though safe cleanup levels could be scientifically established for restoration activities,(16–18) there is always the reality that the public and experts will have a range of perceptions about risk and acceptable safe levels.(5,6,17) This must be expected to lead to intense scrutiny of the decontamination effectiveness, reliability of sampling and monitoring, restoration requirements, progress, and judgments on restoration goals and results. Ultimately, if not agreed upon by stakeholders, judgment of restoration adequacy may be decided in the legal courts.
Finally, there are many potential complications that could arise during the restoration because of incomplete and uncertain knowledge, changing goals and tactics as more information becomes available, and ambiguity of initial and interim restoration goals. The knowledge space of information following a terrorist attack is illustrated in Fig. 2. Information will begin to flow to the responding officials immediately following the incident and will consist of information that is known with great certainty, information that is uncertain or changes over time, and information that is totally unknown but essential to complete a sound restoration plan. In addition, the information will come from a wide array of sources including witnesses, first responders, medical staff, forensic analysts, facility owners, and supporting agencies. Such diversity of sources will make data analysis and restoration plan development and implementation even more challenging. Clearly, restoration planning will have to proceed under uncertainty, especially at the beginning of the response and recovery. Therefore, decision makers need a flexible and broadly communicating management framework for facility restoration that anticipates “surprises” and uses them as tools for learning and for adjustment of strategies and technologies to help make restoration more efficient.
Schematic showing the knowledge space used for developing and implementing a restoration plan. The goal is to contain and start restoration while seeking to expand the left circle (what we know) and decrease the overlap and right circle (the uncertain and unknown).
4.1. Feedback Loops in the Linear Approach
Though Fig. 1 shows a linear path, in theory and practice there are feedback loops in the linear process. For example, contaminated surfaces discovered during Phase 5 can be resurveyed (Phase 3), decontaminated (Phase 4), then surveyed for verification (Phase 5). A similar recirculating loop was evident during the survey and cleanup of facilities impacted by the anthrax attacks on numerous government- and media-operated buildings in the fall of 2001.(20,21) Information gathered on contamination levels during surveys and decontamination activities in the Hart Building, for example, provided feedback on strategies for further identification of contaminated areas and helped determine the most effective decontamination techniques.(6)
Feedback loops work best when the information on which decisions are based is generally complete and accurate. That is not always the case and was specifically noted during response and restoration of facilities affected by the 2001 anthrax attacks.(6,20,21) Uncertainties in the sampling data and lack of experience hindered the ability to detect and respond to anthrax contamination in several important ways.(6,20,21) First, anthrax spores spread more easily and went beyond the boundaries originally anticipated by first responders resulting in the potential for avoidable exposure and additional prophylactic medical intervention. Second, the lack of information on the sampling efficiency for anthrax spores for each of the various surface sampling techniques (dry swab, wet swab, and HEPA vacuum) resulted in serious questions about the accuracy of the results and added uncertainty to strategic decisions. In some cases, these unknowns prevented or delayed accurate identification of contamination and lowered the confidence that decontaminated sites were free of biological contamination. This was especially important because the cleanup criteria for residual anthrax spores were set at very low levels because of uncertainties in dose-response relationships. Finally, the original sampling protocol for anthrax was not designed to provide a sufficient statistical confidence that the areas were free of spores.(21) Combined, contaminated sites were not always quickly identified, secured, and properly characterized, and thus feedback loops, existing or not, were not by themselves sufficient to ensure highly efficient and effective facility restoration. All of these were missed opportunities driven by unanticipated unknowns and uncertainties associated with restoration in complex, public facilities where we lack experience. These factors likely slowed facility restoration and dramatically increased the cost.(6)
The linear management paradigm does not explicitly address uncertainty nor is it inherently adaptable to new information. Management strategies and actions during real events can change through time as new information becomes available and uncertainties may increase in scale and number rather than consistently be reduced. It is important to resolve growing or emerging uncertainties quickly and consistently communicate the resulting changes in management action for effective response. This requires an alternative management paradigm that anticipates and plans for uncertainty, incorporates public and political feedback with developing technical feedback, and is able to handle multiple phases of the restoration process simultaneously and continuously. We suggest that “adaptive management” is a management paradigm having unique and important elements that should be considered to increase the effectiveness of facility restoration in response to terrorist actions in public settings, in particular.
5. ADAPTIVE MANAGEMENT
5.1. Background
Adaptive management was developed to respond to the challenges of management of natural resources.(7,8) Some of the biggest challenges in natural resource management are the complexity of ecosystems, the diversity of responses of ecosystems to natural and anthropogenic disturbances, and our fundamental lack of complete data and knowledge about the systems and their response to various events. For example, management of trees in a forest requires knowledge of the intimate relationship between the biotic and abiotic environment and the intricate web of interactions among species (structural and processes), as well as recognition of important, but uncontrollable, environmental conditions. These relationships are multi-dimensional, dynamic, often hard to predict, and impossible to completely control. In addition, there are diverse societal perspectives on what constitutes proper management of natural resources (e.g., protection, harvest levels, multi-use) and acceptable goals. Loggers, ranchers, environmentalists, the U.S. Forest Service, recreation enthusiasts, politicians, and so on all have their own unique and valuable perspective on forest management that requires some level of balance in the short and longer term. These challenges required a new management paradigm that recognized (1) there will be incomplete information at the start, (2) the forest system and human society/expectations will respond uniquely to the managed actions, (3) the complexity of the system and potential for making mistakes, (4) limited resources and time to complete the actions, (5) potential for disagreements on management objectives and efficacy, and (6) a need to build continuous consensus. The adaptive management paradigm grew out of these challenges
Murray and Marmorek(22) describe adaptive management as more than just “learning as you go.” Rather, it is a deliberate and systematic approach for gathering information by applying management actions as “experiments” (interim actions in CERCLA environmental operations) but the focus is on experiments whose results are needed to make critical evaluations and increasingly robust consensus decisions to complete the restoration. Fig. 3 illustrates how adaptive management integrates objectives to gather information that is critical for effective action. The y-axis shows the general management paradigm, which is focused on management objectives and processes (similar to that outlined in Fig. 1), with minimum allowance for uncertainty or unexpected results. The x-axis illustrates a focus on learning objectives where basic research is conducted without the necessity of immediate progress toward the goal and no clear, direct application of the experiment's results moving the restoration process forward. Adaptive management is a hybrid integration of these two management approaches, as shown in Fig. 3.
Schematic illustrating the balance between learning objectives (research activities) and management objectives (traditional approach). Adapted from Murray and Marmorek (2003, pp. 417–442).
Murray and Marmorek(22) also describe adaptive management as an inherently circular process (Fig. 4), which recognizes the uniqueness of each situation and does not promote a “standard procedure” or recipe to be used for all situations. There are two critical considerations to Fig. 4, however. It is important to note that the components shown in Fig. 3 do not necessarily have a temporal sequence; rather, both management and learning objective actions can occur out of sequence or simultaneously. It is also important to recognize that the process does not “go-around-in-circles” forever, but it is under continuous pressure to converge on set management objectives (spiral development and execution). The criticism of overstudying a problem without taking substantive steps toward project goals is commonly made, but the intent of experiments in the adaptive management paradigm is to study the problem only to a level where the experimental results can be used in a meaningful way and at the earliest possible time.
General schematic showing the adaptive management process (adapted from Nyberg (1999)). The converging spiral in the center illustrates the goal of adaptive management, which is to effectively move toward a consensus and a balanced, agreed upon set of objectives rather than to remain a never-ending circular process.
The clear similarities between natural resource management and restoration of contaminated sites has promoted the use of adaptive management for cleanup of contaminated lands. The National Academy of Science (NAS) recognized this when it was tasked with studying the best approach to remediate and restore contaminated Department of Defense lands at naval facilities. The NAS committee recommended the use of adaptive management to restore these naval lands because it recognized the need for a more iterative approach and for better information and technology for efficient restoration and for greater public participation and communication.(9) These are the same conditions one will likely find during restoration of public areas following a terrorist act.
5.2. Adaptive Management for Restoration of a Public Site After a Terrorist Attack
The challenges of resource management and restoration of a site following a terrorist attack are very similar. Key to the adaptive management paradigm is to determine the knowledge space. Fig. 2 is a simplified schematic of the knowledge space dividing it into three basic categories; what is known, unknown, and uncertain. Initial information will largely be derived from the description of the attack from witnesses, first responders, forensic specialists, and emergency workers. This information will be in the form of medical information, recorded eyewitness accounts, measurement information (e.g., from contaminated clothes, personnel, or initial contamination characterization), and initial boundaries for contamination control and access points. In addition, facility information (layout, ventilation, construction materials, facility use and tenants, and current facility conditions) may be available or derivable from the owners and/or operators. The credibility of much of this information will be sufficient to be categorized as “known.” Simultaneously, the known information allows definition of unknowns. Some of the known information can be judged to be implausible, uncertain, or inconsistent enough to cast some doubt on the credibility of the “facts.” Good examples of this would be from conflicting eyewitness accounts or other inconsistent or potentially implausible survey information. This information should be weighed to determine if it is important to the restoration to improve the certainty of the information. These unresolved, but important, knowledge gaps would then need to be filled, and the management plan should include this as one of the important actions.
5.2.1. Potential Areas of Uncertainty During Restoration and Recovery Operations
Areas of potential uncertainty and specific real examples that may affect restoration operations follow.
- 1
Sampling and decontamination efficiency for a variety of surfaces: Sampling and decontamination efficiency (relative amounts of agent transferred onto sample matrix and decontaminated, respectively) varies by technique and surface. These efficiencies can be expected to be unknown or uncertain on many of the surfaces found in a public facility, and effective restoration and reoccupation requires that these efficiencies are well understood. It would, for example, be difficult to justify declaration that an area is at safe levels without some reasonable knowledge of the sampling efficiencies of the used techniques on different surfaces. Potentially, survey results from a carpeted area, for example, may show contamination levels below the safe threshold, but if the sampling efficiency is low, then the residual levels in the carpet may still be high. Later vacuuming and/or traffic across the carpet could then bring the residual contamination to the surface and potentially expose people. This uncertainty may affect cleanup levels, postrestoration monitoring requirements, or consideration of alternative restoration actions.
The evaluation of the effectiveness of the decontamination methods will also likely vary in unpredictable and predictable ways depending on the surface material. In some instances, the uncertainty of sampling and effectiveness of decontamination of agents on some porous materials may lead to conservative judgments, resulting in many items with porous surfaces being simply thrown away, perhaps needlessly, with cost, disposal, and handling implications. Similarly, simply removing carpet and other porous surfaces without proper characterizations may have other unintended consequences and uncertainties.
Examples of consequences from uncertain knowledge of sampling are illustrated in the cleanup following the anthrax releases. Decisions regarding the spread of anthrax and the effectiveness of the decontamination relied on surface sampling whose techniques were not fully tested and validated for the surfaces encountered.(20,21) This led to confusion about the best survey approach (e.g., dry swab, wet swab, vacuum sampling) and the sampling efficiencies of each on the various surface materials and texture. These uncertainties significantly complicated and arguably delayed the reoccupation of the affected facilities.(20,21)
- 2
Mobility of the agent(s): The mobility of the agent(s) can vary in predictable and unpredictable ways during and after a release into a facility, and the mobility can be related to environmental conditions and facility use. Examples include the dispersibility of the agent in rooms and ventilation, volatility of the agent under changing temperature and humidity, and off-gassing rates of the agent from surfaces that can vary with ventilation rate and other environmental conditions. Beyond natural processes, some or all of these aspects may be expected to be exploited by the terrorist action design. Combined, these uncertainties can lead to errors in judging and quantifying the spread of the agent throughout the building and into surface pores and decreases the efficiency of the restoration.
Also, changes in facility operations may have significant impacts on mobility of agent(s). For example, going from shutdown or containment operational status to staged restoration operations and then staged public operations may change accessibility of contamination, and therefore mobility, which needs to be accounted for. A study of resuspension of anthrax spore showed 5 of 17 positive bacterial plates during a 1-hour period of little or no activity, but found 16 of 17 positive bacterial plates during a 1-hour period of simulated office activity.(5) This result has important implications for interpretation of surface sample results and decisions regarding safe reoccupation of the facility.
- 3
The spatial distribution of the contaminant: Knowledge of the spatial distribution of the agent is important for setting of boundaries for the public and access points for restoration personnel. Also, this information could be used to determine which survey methods would provide the best statistical sampling technique. Sampling procedures for highly heterogeneously distributed agent (e.g., hot spots) would be very different than for a situation where the agent is more evenly spread through a facility. The importance of selecting the best sampling strategy given the spatial distribution of an agent was illustrated during the restoration of anthrax-affected buildings. In many cases, selection of sampling locations was largely based on professional judgments of where the spores were expected to be found. This led to uncertainties as to the statistical reliability of the sampling results and the sampling design was especially critical given that low levels of anthrax could be fatal.(21)
- 4
Modification in decontamination techniques and technologies: Decontamination techniques and technologies can evolve during a response providing cleaner areas and more efficient procedures. Information on what has been working or not working should be made available to all restoration teams. This information can affect confidence in the outcome of restoration actions. Decontamination of the anthrax in the Hart Building could have proceeded using sodium hypochlorite and a formaldehyde gas, as was recommended in EPA guidelines, but after testing, chlorine dioxide gas was chosen as a more effective technique given the conditions.(5)
- 5
Acceptable endpoints for the residual contamination: The public, safety professionals, facility owners, and politicians may disagree (or reasonably change positions) on acceptable levels of residual contamination before the area is released back to unrestricted use as information becomes available during characterization, decontamination, or clearance phases of restoration. Decisions for safe reoccupation of facilities require defining acceptable thresholds, which is not simple and seldom unanimous. For example, the EPA has a log-6 kill criteria, meaning that the air concentrations had to be reduced by six orders of magnitude. However, the number of anthrax spores released into some of the buildings was so large (e.g., >1012 spores) that reducing the spore levels by six orders of magnitude was insufficient for reoccupation. This led some to state that the only acceptable clean up criteria was no live spores detected.(6)
5.2.2. Response to Uncertainty
Any preevent facility knowledge, e.g., from past studies, can make restoration more effective and is recommended.(6) Good examples are studies of sampling efficiency for anthrax spores that have recently been published.(23,24) However, given the wide range of potential terrorist targets and the uniqueness of releases in complex public spaces, one can confidently expect incomplete information, which will require collection and analysis of additional data (experimentation) and even requirements for development of combinations of new and old understandings and knowledge.(25) Identification and development of additional experimental activities (studies) is an intrinsic component of adaptive management processes that requires integration into restoration planning, often across teams focused on components of linear process plans. These studies should be directly focused on the knowledge gap and directly applicable to the specific restoration activities, as illustrated in Fig. 3. In some cases, the experimental design may have to be balanced with cost and timeliness of the restoration and may, by design, only provide a qualitative assessment. However, key studies must be designed, implemented, and evolved to reduce the uncertainty enough to significantly increase restoration effectiveness. Therefore, to address the listed uncertainties above, examples of studies such as the following could be planned from the beginning of the remediation, with the experimental design dependent on the specific situation and level of prior research.
- 1
Sampling and decontamination efficiency for site-specific surfaces: Set up plots or sections for determining the efficiencies of surveys and decontamination on different materials and surfaces. Tests such as comparing the amount of the agent on the survey material with the total amount in the surface (may require chipping or removal of the top layer of the surface) can be performed to test for sampling efficiency. Side-by-side plots may be used determine decontamination effectiveness tests. This testing can take some time, so it is best to start early or, better yet, to have some good data prior to the event.
- 2
Environmental mobility of the agent(s) and baseline(s): Make continuous measurements of the amount of the agent(s) in air and on same surfaces to ensure that the natural ambient and restoration activities are not spreading the agent into uncontaminated areas. Small-scale measurements under varied conditions (e.g., under heat lamps or changed air velocities) can be done. For example, resuspension measurements made at the Hart Building during high and low activity provided valuable information useful for establishing cleanup levels and decontamination procedures that are robust under a variety of environmental conditions.(6) Environmental mobility measurements also help ensure that the access points for the responders and the boundaries delineating the restoration area are continuously well defined and appropriate during restoration and across changing environmental conditions.
- 3
The spatial distribution of the contaminant: Gather information about the nature of the event from the first responders (ongoing release, multiple release points, physical form of agent). Perform large area survey smears of surfaces to assess spatial patterns of residual contamination, utilizing potentially novel, natural, and scaleable sampling approaches. These results will be qualitative but sufficient for the purpose of the survey. Measurements of airflow and deposition patterns using a tracer gas in a facility and/or modeling prior to the event to establish which zones would be affected by a release could help establish access boundaries and predict the spatial distribution of the contaminant. Also, use of air dispersion models specifically developed for predicting facility air flow may improve information certainty and provide confidence in restoration effectiveness.
- 4
Modification in decontamination techniques and technologies: There are multiple techniques that can be used to neutralize many hazardous materials. Selection of the optimal technique for all surfaces is not always clear at the beginning of the cleanup. Selection of plots for decontamination and initiating testing at the beginning of the cleanup can provide valuable information on the effectiveness of each technique tested.
- 5
Acceptable endpoints for the residual contamination: Decisionmakers and other stakeholders should begin discussions, communications, and initial negotiations on acceptable restoration levels soon after the event.(12) In many cases, such as for radionuclides, regulatory limits for contamination concentrations in air and on surfaces in public areas are tabulated,(26) but it may be less clear for many chemical attack agents, though levels have been published.(16-18) In addition, understanding of the uncertainties in decisions should be identified so that additional potential review/decision points can be identified to improve integration of restoration actions.
Identifying, communicating, and even resolving these uncertainties early in the building restoration is critical for effective response, but there may be others. The general point is that there needs to be consideration of critical knowledge gaps, and the cross-teaming necessary to address gaps, from the beginning of the restoration, and plans made to quickly answer some key questions, even if only qualitatively. Some of these actions will inevitably occur, even with linearly structure plans; however, the adaptive management paradigm explicitly integrates them. The information gathered and documented in these studies will be used to guide and adjust restoration activities and can be used to justify the decisions made during the response, which could later prove to be operationally and legally important.(27)
5.3. Sequence of Restoration Events
The linear restoration model shown in Fig. 1, while useful, implies that the restoration proceeds in a linear sequence of events following a terrorist attack. In reality, various restoration activities will need to proceed simultaneously. Performing activities in parallel can lead to a converged solution and restoration more quickly. Fig. 5 compares the linear restoration with an adaptive management model where events occur simultaneously and feedback of measurements from targeted experiments is continuously evaluated and restoration methods adjusted, as needed (adapted from a NAS report(9)). The adaptive management model illustrates that activities such as decontamination, characterization sampling, informative studies, and recognition of restoration goal(s) can occur simultaneously. Combined, the adaptive management strategy presents the possibility of significantly improving planning, operational actions, and, finally, the performance of the activities toward reaching or exceeding the goals of the restoration.
Generalized comparison of the linear management model and adaptive management model for restoration of a public facility after a terrorist attack. The left axis represents the time required to complete restoration.
5.4. Limitations of Adaptive Management
While the adaptive management paradigm holds much promise in theory and for some operations, the promise has not always been realized in practice.(28,29) The promises of adaptive management include increased rates of acquiring information, enhancing information flow, and creating shared understanding among varied stakeholders, all of which have been shown to result in high-quality decisions.(30) However, a review of several examples of adaptive management of salmon and forests in North America revealed several important shortcomings,(28) some of which are applicable to facility restoration following a terrorist attack. First, autonomous stakeholders can have views that differ so drastically that they do not always agree on the methods, results, or conclusions of the studies that are required to effectively manage the resources. This can lead to a lack of consensus on management plans and stalled implementation. Second, some groups thought that the planned studies were not always worth the additional cost and time and, in some cases, they were concerned that the study results would be disadvantageous to their group's self-interests. Third, uneven influence between various stakeholders can lead to mistrust and cries of unfair influence of elite and powerful groups on decisions. Of course, each of these vulnerabilities can and do exist in any restoration operation.
Despite difficulties, there were several positive outcomes of these adaptive management trials.(28,29,31) First, iterative studies have shed important insights into ecological process that were important to consider for resource management. Second, the process of involving varied stakeholders improved understanding across groups through discussion, even though consensus was not always reached. Finally, through gained experiences with adaptive management, the limitations of this approach have been tested and refined, leading some to suggest that adaptive management would work well for complex problems whose impacts span across medical, social, economical, and political boundaries, and where a holistic approach is needed.(31) These are the likely conditions following a terrorist attack, again suggesting advantages in incorporating adaptive management principles into restoration plans.
6. CONCLUSIONS
In this article, we introduced adaptive management principles in the context of response and recovery of a public facility following a terrorist attack. The goal of this discussion was not to promote adaptive management as a stand-alone standard approach. Rather, the article is intended to raise adaptive management principles into the discussion to help various operational and regulatory agencies develop flexible standards and site-specific plans for remediation. While the techniques of flexible management can be achieved in many cases (e.g., “flying by the seat of the pants” response and/or multi-threaded team response networks), a more formal preplanning, implementation, exercise, and assessment of the value of adaptive management principles creates a process that recognizes and communicates the “certainty of uncertainties,” that processes can often be more efficiently planned and managed and communicated in parallel, that it is important to benefit from missteps and/or uncertainties, and the need to evaluate multiple options to make optimal decisions. Properly developed and implemented, consideration of adaptive management principles has the potential to improve efficiency, communication, cost, and broader/quicker consensus for of the response to potential public attacks by terrorists.
ACKNOWLEDGMENTS
We thank the anonymous reviewers whose suggestions greatly improved the article. This work was supported by the U.S. Department of Homeland Security, Directorate of Science & Technology, Chemical & Biological R&D section, as part of the Response and Recovery Program for the Facility Restoration Operational Technology Demonstration Project.
