Flood Risk Management: US Army Corps of Engineers and Layperson Perceptions

Risk Analysis DOI: 10.1111/j.1539-6924.2012.01832.x

Flood Risk Management: US Army Corps of Engineers andLayperson Perceptions

Matthew Wood,1 Daniel Kovacs,2 Ann Bostrom,3 Todd Bridges,1 and Igor Linkov1,∗

Recent severe storm experiences in the U.S. Gulf Coast illustrate the importance of an in-tegrated approach to flood preparedness planning that harmonizes stakeholder and agencyefforts. Risk management decisions that are informed by and address decision maker andstakeholder risk perceptions and behavior are essential for effective risk management policy.A literature review and two expert models/mental models studies were undertaken to iden-tify areas of importance in the flood risk management process for layperson, non-USACE-expert, and two USACE-expert groups. In characterizing and mapping stakeholder beliefsabout risks in the literature onto current risk management practice, recommendations foraccommodating and changing stakeholder perceptions of flood risks and their managementare identified. Needs of the U.S. Army Corps of Engineers (USACE) flood preparedness andresponse program are discussed in the context of flood risk mental models.

KEY WORDS: flood; mental models; risk analysis

1. INTRODUCTION

In the fall of 2005, hurricanes Katrina and Ritarevealed inadequacies in severe storm and flood pro-tection plans for the U.S. Gulf Coast. Initial criticismcentered around engineering design and manage-ment issues (e.g., impact and loss projections, floodprotection infrastructure) and degradation over timeof the region’s wetland defenses by various industries(e.g., energy, transportation).(1,2) Recent work hasalso noted the importance of human factors in dis-aster prevention planning. For example, Gheytanchiet al.(3) stresses the importance of considering psy-chological factors before engineering and other con-straints when designing disaster management plans.

1U.S. Army Corps of Engineers, Engineer Research and Develop-ment Center, Environmental Laboratory, Vicksburg, MS

2Decision Partners, LLC, Pittsburgh, PA3University of Washington Evans School of Public Affairs, Seattle,

WA∗Address correspondence to Igor Linkov, US Army Engineer Re-

search and Development Center, US Army Corps of Engineers;tel: 6172339869; [email protected].

The Institution of Civil Engineers panel in theUnited Kingdom made similar recommendations aspart of their assessment of the state of flood manage-ment and their proposed improvements in responseto severe flooding in 1998 and 2000. Two key recom-mendations of their report were to “learn to live withrivers” by accommodating waterway expansion fromrainfall, and to provide greater weight to human andsocial factors when assessing flood risk.(4,5)

The U.S. Army Corps of Engineers (USACE)and other agencies are working on increased in-teragency coordination and stakeholder inclusionin restoration planning for coastal Louisiana andMississippi.(6−8) This approach includes multiobjec-tive management tailored to the needs of specificcommunities and the region in general. As the defacto leader of U.S. flood risk management (FRM)efforts, USACE is looking to evolve from its historicemphasis on engineering for addressing flood risks towell-informed strategic planning in coordination withstate and local management authorities. These stateand local agencies typically manage floodplains at a

1 0272-4332/12/0100-0001$22.00/1 C© 2012 Society for Risk Analysis

2 Wood et al.

micro-level in ways that often do not consider theeffects of mitigation plans on upstream or down-stream communities.(9,10)

To date the USACE is responsible for 383 lakesand reservoirs, 8,500 miles of dikes and levees, and240 miles of shoreline protection split across 90projects in the United States. In addition, the Corpshas constructed hundreds of smaller local flood riskreduction projects that have been turned over tonon-Federal authorities for operation and mainte-nance.(11) With such a large stake in the nation’sFRM activities, effective policies for designing, coor-dinating, and implementing FRM plans are particu-larly essential to USACE’s core mission, as well asthe work of public safety and risk management pro-fessionals generally.

Further complicating matters, many citizens ex-pect government agencies to bear the burden ofFRM, even on matters of personal liability or prop-erty loss.(3,12,13) USACE design of architecture toreduce flooding probability, along with governmentcertification of levees and National Flood InsuranceProgram (NFIP) requirements, helps to perpetuatethe phenomenon of personal lack of responsibilityfor flood planning.(9) Though the National FloodRisk Management Program established by USACEis helping to address some personal responsibility is-sues,(9,10) approaches which incorporate stakeholderperspectives and encourage them to take an activerole in disaster planning are critical to future sus-tainable disaster management in the Gulf Coast. Theneed for effective FRM is especially pressing giventhe passive stance that many citizens take towardFRM. Flood risk is often underestimated by indi-viduals, who frequently recall only the most severe(and therefore most infrequent) events and believethe government is the sole provider and responsibleparty for FRM.(12)

USACE and others are working to increaseinter- and intra-agency coordination and stakeholderinclusion in FRM planning. Interagency coordinationof FRM is a complex issue, as USACE is conduct-ing FRM planning alongside FEMA, DHS, and ahost of other federal, state, and local agencies. Theimportance of accommodating social and human di-mensions in disaster preparedness has been recog-nized, but specific tools for integrating knowledge,interests, and values of stakeholders4 require addi-tional development. Mental modeling has been rec-

4A stakeholder for the purpose of this report is defined as an in-dividual or organization with a direct or indirect investment inflood preparedness and response.(14)

ognized as a useful framework for better understand-ing and addressing deeply held risk and value beliefsthat can enhance stakeholder involvement in strate-gic planning, and applied to represent layperson per-ceptions of floods.(12,15,16) While there are compar-isons of layperson and expert perceptions of floodrisk;(17) addressing the revealed differences betweenthese groups remains a challenge.

This article explores perceptions of FRM fromUSACE expert and layperson perspectives, with afocus on how experts from a variety of organiza-tional units and disciplines within USACE under-stand FRM, and identifies opportunities for FRMimprovement in general. A key interest here is toexamine any differences in FRM perceptions andbeliefs between planners and engineers, using for-mal analysis of qualitative data derived from men-tal models interviews. These findings should aid US-ACE in redeveloping its FRM process, and also otheragencies with similar risk management responsibili-ties. Comparisons between diverse FRM stakehold-ers’ perceptions and beliefs can be derived fromseparate studies of each, or comparative studies. Weapply both approaches in this article, triangulatingbetween original investigation of how two stake-holder groups within USACE perceive FRM, andpublished research on lay perceptions and mentalmodels of flood risks and FRM.

A common approach to identifying discrepan-cies between expert and lay perceptions or beliefsis to compare the results of studies of each. Theseresearch efforts typically focus on a single imple-mentation context. For instance, Bostrom et al.(18)

conducted a literature review to better understandexpert perceptions on the sources of radon in thehome and risks of radon exposure. They then sur-veyed laypersons on the same issues to understandwhere lay perceptions differed from that of experts.This approach has the advantage of drawing on theaccumulated research evidence, but the disadvantageof comparing studies from varying decision and riskcontexts, such that the expert and lay views may notbe truly analogous. A second systematic approach isto formally model potential FRM decisions with in-put from experts, then conduct mental models re-search to elicit the mental models people (with anylevel of expertise) apply when making such decisions.

This article synthesizes results from both ap-proaches to compare and contrast what is knownabout lay perceptions of FRM with the views of twogroups of experts. First, a literature review was con-ducted (Section 2) which identified layperson con-cepts and beliefs related to FRM. These layperson

Flood Risk Management 3

perceptions are aligned with expert knowledge. Next,an expert modeling effort was undertaken, focusingon USACE FRM experts, as well as the resultinginfluence diagram (Section 3). Finally, comparativeanalyses of telephone interviews informed by the ex-pert model (EM) influence diagram are conducted(Section 4).

2. LITERATURE REVIEW OF LAYPERCEPTIONS

2.1. Review Methodology & Structure

A literature review was conducted in order to un-derstand the mental models of laypersons and non-USACE experts on topics related to USACE FRMin order to understand how USACE FRM prac-tice may better accommodate stakeholder knowl-edge and beliefs. The literature search for this re-view was conducted from May through July of 2008.It began with a keyword search of the ISI Web ofKnowledge database for the term “mental model”along with “flood manage∗,” “flood respon∗,” or“flood recov∗.” Subsequent keyword searches wereconducted in both ISI Web of Knowledge and APAPSYCInfo R© with terms identified from the initialquery, including “concept mapping,” “decision sup-port systems,” “strategic environmental assessment,”“flood risk management,” “diagrammatic reason-ing,” “Logical Argument Mapping,” “risk-ranking,”“shared mental model,” and “team mental model.”A limited search was also done on both databasesusing the terms “schema∗” and “folk model∗.” How-ever, this search returned many results unrelated torisk issues discussed in this article. “Schema∗” iden-tified over 8000 entries in Web of Knowledge span-ning a large domain of interests (e.g., computationalneuroscience, language acquisition), many of whichwere not relevant to understanding how individualsrelate to floods or natural disasters in general. Manydiscussions of folk models exclusively describe causalattributions of another individual’s behavior (theoryof mind), and a few authors(19) describe them as aspecial case of mental models. Social Science Cita-tion Index Expanded was used to identify papers thatcited Morgan et al.(20) or Doyle and Ford.(21,22) Theseworks provide a context for defining mental modelsof complex real world phenomena.

Queries identified by Web of Knowledge,PSYCInfo R©, and Social Science Citation Index Ex-panded searches were supplemented with reviews ofquery reference sections to identify sentinel works,

and a Google Scholar search of key authors to lo-cate books or other published works on the topic notavailable via peer-reviewed journal sources. USACEinternal documents were also reviewed. Though byno means exhaustive, this literature search producedover 600 documents, of which the resources citedhere make up a fraction. Documents were retained ifthey elucidated stakeholder perceptions of the FRMprocess and matched one or more components of thesimple EM (Fig. 1 & Section 3). Documents wereexcluded that detailed existing risk managementDecision Support Systems or focused extensively ontechnical aspects of meteorological, hydrological, orgeological processes implicated in flooding. These ex-clusions were made because these articles, while elu-cidating the state-of-the-science related to floodingevents, did little to elucidate why laypersons holdopinions of flood events and their management thatdiffer from those of experts. Links to the USACEFRM expert model (Section 3) have been made whenpossible (Table I), though it should be noted thatthe match between USACE FRM and informationuncovered in the literature review is sometimes im-precise. Terms used in the model can be abstract innature, and any particular piece of reviewed infor-mation might fit several different categories.

2.2. Results

While the review is by no means exhaustive, itdoes capture some of the key issues faced by USACEand other agencies tasked with managing risks fromnatural hazards like flooding. A summary of reviewfindings and their relationship to the EMs (Section 3)can be found in Table I. In particular, literature re-view results provide lay perspective information re-lated to Societal Drivers, Influences on Flood Risks,Quality of FRM Planning, and Quality of FRM Im-plementation.

2.2.1. Societal Drivers

Societal drivers are factors determined by soci-ety at-large that influence the FRM process. Willing-ness to prepare for an event, and knowing how toprepare, are among the most significant of these fac-tors. These depend, in part, on the availability andaccuracy of timely weather forecasts, the decision-making of emergency managers and support staff,and local laypersons who decide when and how toprepare for a storm.(27) Case study reports of dis-connects between what laypeople and experts know

4 Wood et al.

USACE Flood RiskManagement

Desired OutcomesOptimal USACE Flood

Risk ManagementEffective Public EngagementOptimal Public Preparedness

& Response Positive Perceptions of

USACE Trustworthiness &Competence

Quality of FloodRisk ManagementImplementation

Flood Control ConstructionAsset Management (e.g., Ops., Mtce., Inspection)Non-Structural Activities

PreparednessResponse

Quality ofStakeholder

Collab., Coord., & Comm.Interest, Advisory, &

TransactionalGroups

USACE DriversMission & MandatePrinciples & Values

TechnologyPrioritization

Funding

Quality of FloodRisk Management

Planning

Quality of Internal USACE Collaboration,

Coordination, & CommunicationHorizontal & Vertical Coordination

Organizational StructureLeadership & CultureCollaboration Tech.

Quality ofFlood Risk Issue

Identification

Quality ofGovernment Partners’

Collaboration, Coordination, & Communication

Federal, State &Municipal

Societal DriversEconomic DevelopmentPublic Health & Safety

Social & Community ImpactPublic Expectations (e.g., flood risk, flood control structure performance)

Environmental Protection& Climate Change

Political DriversCongressional Directives

Federal Policy National Flood Insurance

Program

Qualityof Flood Risk Assessment

Workforce CapacityCapability & Expertise

Communities of PracticeDistribution & MobilizationRetirement & Replacement

Knowledge RetentionWorkforce Distribution

Workload SharingMorale

Individuals’Mental Models of Flood

Preparedness and ResponsePerceptions of Flood Risks

& Flood Risk ControlPerceptions of USACE

Activities

HistoricFlooding Events

Mississippi FloodsHurricane Katrina

Others

Quality ofPublic Engagement

Risk CommunicationsNEPA Processes

EducationOutreach

Influences on Flood RisksWeather & Climate

Geography, Topography & DemographyDevelopment and Land Use

Flood Control StructuresNon-Structural Measures

Flood Response &Mitigation

•

•

•

•

••

••

••

••

•••

••

•

•

••

••

••

•

••

••

•••

••

••

••

•

••

••

•••

•

••

•

Fig. 1. Simple expert model of influences on USACE flood risk management.

about flooding suggest this discrepancy can inhibitthe process of reducing public flood exposure andsensitivity, thereby reducing the effectiveness of pro-posed adaptations.(28) This is problematic as layper-sons generally know little about how to decreasetheir exposure and sensitivity to floods. This is a caseof “ . . . ignorance of what to do, not rejection [of ex-pert plans, which] caused people to ignore advicefrom flood control experts” (p. 265).(12) Not only doexperts need to inform laypeople on what to do foreffective execution of disaster plans, but they shouldalso attempt to forecast changes in future compli-ance if a false alarm is issued,(27) a nontrivial task asdisagreement exists about the average net effect ofissuing evacuation orders without subsequent stormevents occurring.(23) Rasid and Haider(24) also makean argument for properly informing laypersons aboutthe flooding process and clarifying misunderstand-ings. They state that adequate explanation of flood-

ing risks might help to counter resistance to manda-tory evacuation orders that help laypersons avoid ad-verse consequences. Resistance to evacuation ordersmay also come from ignoring the possibility of an-other major flooding event.(12)

In a study on the Red River Valley flood of1997, Buckland and Rahman(25) found that commu-nities with more physical, human, and social capi-tal were better prepared for flooding. However, so-cial capital somewhat complicated the preparationprocess. While social capital (quantity and qualityof civic organization community involvement) canhelp to gather people in an effort to prepare, itcan also delay important decisions. Other researchsuggests that those who live in flood-prone areasand downstream portions of urban watersheds aremore willing to pay for community-wide measures todecrease exposure and sensitivity.(26) In a study ofMilwaukee metropolitan area residents, those who


Table I. Summary of Key Research Findings

Literature Review

USACE FRM Model Concepts Expert Knowledge Layperson Knowledge & Beliefs

Societal Drivers Experts disagree about effects of issuingevacuation false alarms.(23)

Preference for home flood-proofing versusless risky buy-out option, even after aflood.(24)

Communities with more physical, human, &social capital are better prepared forfloods.(25)

Those in flood-prone areas willing to paymore for community-wide riskreduction.(26)

Experts need to communicate disaster planeffectiveness & anticipate costs of issuingan evacuation “false alarm”.(27)

People know about home protectionmeasures, but not how these measureswork.(16)

Flood risk education may counteractresistance to mandatory evacuationorders.(24)

Resistance to evacuation may come fromignoring the possibility or likelihood of aflooding event.(12)

Lack of flood exposure and sensitivityknowledge inhibits reductions in exposure& sensitivity.(28)

Influences on Flood Risks: Geography,Tomography, & Demography

Catchment area & precipitation create flashflood risk(16)

Focus on local land/engineering features &precipitation.(16)

Believe past damages approximate futurerisk.(28)

Flood-prone area residents lack flood riskexposure knowledge & motivation tolearn.(12)

Influences on Flood Risks: Developmentand Land Use

Dredging & development can adverselyimpact natural flood protection systems.(7)

Laypersons blame poorly coordinatedactivities to increase waterway benefits forcreating flood conditions.(12)

Quality of FRM Planning Fewer structural solutions may reduce publicblame of governments or businesses forflooding.(12)

Public and risk management experts havedifferent concepts of acceptable risk.(30)

Shifting priorities toward land managementversus structural solutions to reduce floodrisk.(32)

Risk reduction policies can increaserisk-taking behavior (lulling effect or riskcompensation).(31)

Control structures can reduce local risk atthe expense of increased system-widerisk.(34)

Public misunderstands how flood protectionstructures work.(33)

Structural solution advocates recommenddams/levees accommodate ProbableMaximum Flood.(35)

Believe (incorrectly) that structures preventdamage.(28)

Quality of FRM Implementation Emergency managers and scientists lackcommon FRM language.(36)

Poor infrastructure knowledge promotesactivities that increase exposure.(37)

Forecast accuracy impacts its economicvalue in FRM.(38)

Postevent compensation policies promotereliance on protective structures.(24)

Dynamic response approaches can reduceunnecessary preparation & evacuationcosts.(27)

Memory of past floods influences willingnessto prepare.(39)

Improving FRM requires changing landregulations and pricing risk into propertyinsurance.(40)

Belief in a just world helps individuals tocope with flood consequences.(41)

Preparedness depends on forecasting,emergency manager decisions, and laybehavior.(27)

Mandatory flood insurance is unnecessarybecause government and NGOs providebetter aid at lower cost.(12)

6 Wood et al.

lived outside the 100-year floodplain were most will-ing to pay for flood reduction measures, fearing in-clusion in the 100-year plain if flood risk were to in-crease. Household income was positively correlatedwith willingness to pay, and belonging to an envi-ronmental group also had a small positive effect onwillingness to pay for flood mitigation technologies.In addition, laypersons from flood-prone areas some-times hold suboptimal beliefs even when it concernstheir own personal property. For example, respon-dents to a study in the Red River Basin in Canadaafter the 1997 flood tended to prefer home flood-proofing techniques for flood mitigation, even whenfiscal incentives of other strategies (e.g., governmentbuyout) were better.(24) In a different study relatedto flash flooding, most laypersons knew about flood-proofing devices to keep their homes safe. However,they rarely knew how these devices worked to pro-tect their homes.(16)

2.2.2. Influences on Flood Risks

Many factors can directly influence flood risk.The most salient factors that arose in materials re-viewed include Weather & Climate, Geography, To-mography, & Demography, and Development andLand Use. Lave and Lave(12) find that those who livein flood-prone areas generally know little about theirrisk of flooding and how floods begin, and are usu-ally unmotivated to learn about either. Those withmore education who enjoy reading self-identify asscientifically savvy, or own their own home, tendto know more about the flooding process. Layper-sons also have significant misunderstandings relatedto flood prediction. For example, they believe thatflood frequencies are well understood by experts, andthat recent flood damages are unprecedented (andtherefore unpredictable) because of global warming.Some anecdotal evidence also suggests that manyindividuals believe flood forecasts are available toeveryone.(28) Even if everyone did have access toaccurate forecasts of precipitating event likelihoods,current weather forecasting techniques predict ex-treme weather events with limited temporal and spa-tial resolution.(29)

Many laypersons believe experts know moreabout precipitating factors that cause flooding thanis actually the case.(28) Concerning flash floods inthe Bavarian Alps, Wagner(16) found that layper-sons believed weather conditions, debris problems,and local watershed authority engineered structuresare important in determining whether a flash flood

will occur. In truth, the main factors that influenceflash flooding are weather conditions, characteristicsof the catchment area, and the processes that oc-cur within mountain torrents in the region of theBavarian Alps where this work was done. Manylaypeople also thought global warming would lead toan increased incidence of flash floods. Though humaninfluence on global warming does not influence flashflooding directly, human impact in terms of land useand torrent control clearly do change flash flood like-lihoods. Personal experience and visibility of processare two main influencing factors that explain the ac-curacy of layperson flash flood knowledge. Layper-sons were better informed about floods than land-slides, in part because flooding processes are easierfor laypersons to observe directly. Those with moreexperience with flash floods had a more completeunderstanding of the process. Layperson knowledgeabout flash flooding changed very little over a two-year period, demonstrating that public perception offlood events can be stable over time.(16)

Many laypeople may be skeptical of governmentability to manage water resources effectively. Asmentioned earlier (Section 1), waterways are oftenmanaged in a piecemeal fashion without consider-ing the interaction of management strategies on ad-joining sections of a waterway. Some residents inLave & Lave’s(12) sample of residents near one south-west Pennsylvania river system tended to blame thislack of coordination for flooding events, with theircommunity being sacrificed to save larger urban cen-ters downstream from the consequences of flooding.Dredging operations and development similarly leadto the destruction of wetlands protecting the city ofNew Orleans from tropical storms. These modifica-tions tended to benefit some groups (e.g., energycompanies) at the expense of increasing risk for oth-ers (e.g., tourism companies).(7)

2.2.3. Quality of FRM Planning

Lack of knowledge about structural defensesmay couple with ignorance of flood or precipitat-ing event likelihoods to increase flood damage costs.People have developed property in floodplains underthe belief that structural techniques for flood man-agement provide greater protection than they actu-ally do and often disregard the frequency of floodevents when developing flood-prone areas. In re-sponse to this trend, experts have been moving fromstructural mitigation strategies to those that rely onland management to make room for some flooding


to occur.(42) These strategies may reduce laypersons’tendency to pass on blame (typically to a governmentor water management authority) for flood events anddamages, since manmade structures have a smallerrole in flood management and protecting individualsin the first place.(12)

Some advocates of structural flooding solutionsrecommend increasing dam heights and modifyingspillways so structures can accommodate the prob-able maximum flood (PMF). However, others havesuggested that fixing dams to accommodate the PMFcould be ineffective and even wasteful or harmful. Asit concerns dam spillways:

. . . Additional spillway capacity, while preventing damfailure if the PMF occurs, can have very serious nega-tive consequences.5 Additional spillway capacity is fre-quently chosen by itself or in combination with a damraise to bring a dam into compliance with existing pol-icy and law. At some ‘unsafe’ dams, accepting the statusquo (i.e., doing nothing) may be the best course of ac-tion.(35)

Though dams and levees are intended to protectindividuals from flooding, their construction may ac-tually serve to increase layperson exposure becauselaypersons tend to believe (incorrectly) that thesestructures “prevent” damages.(28) Levees can reducethe likelihood of inundation by water, but not elimi-nate it. Also, while these structures can reduce floodlikelihoods in the short-term for the area nearest thelevee, they can have ancillary effects on the water ta-ble or nearby wetlands that actually lead to a long-term increase in flood likelihood and severity.(34) AsLave and Lave state, “No engineering structure canguarantee protection for people living in a flood plain(p. 257).”(12)

This mindset highlights the discrepancy in riskplanning and mitigation between laypersons who de-mand certainty in risk (e.g., Is X safe or unsafe?) andrisk management experts, who take a more nuancedinterpretation (e.g., How safe is X?).(30) Advocatesof dams who make claims about provided safety maycause people to believe a dam reduces flooding ex-posure to negligible levels. Laypersons then becomemore likely to build in the flood plain. Dams lead in-directly to increased sensitivity to floods as measuredby property damage rates and loss of life.(12) Thoughconstruction within a flood plain appears to lead toa paradox in net levels of flood risk mitigation, the

5These negative consequences include increased annual down-stream flood losses, attenuated benefit–cost ratios, constructionaccidents from modifying spillways, and high monetary costs formodest safety returns.(35)

effect seen here is generally consistent with a lullingeffect, or risk compensation,(31) whereby individualsengage in counterproductive risky behaviors in re-sponse to regulations that require consumer productsafety features. People behave less responsibly in thepresence of these safety features because individualsbelieve the features provide a higher degree of pro-tection than is actually the case. Classic cases of thiseffect can be seen in the transportation safety indus-try.(43,44)

In general, the protection imparted by infrastruc-ture is poorly understood by the public.(33) Peoplethen build within the flood plain, and as time goesby people forget about their exposure and sensitiv-ity to flooding.(37) Furthermore, compensation poli-cies after a flood promote reliance on these struc-tural mitigation methods.(24) These are unfortunateconsequences of both a lulling effect and plannerssuggesting to laypersons that flood control structurescan do more than they are designed to in order to re-duce concerns arising from misunderstandings of ex-posure to flood risk.(12) To counteract this trend andreduce flood costs, Lenntorp(40) argues for more ef-fective use of land regulation rules and pricing floodrisk into property insurance rates.

2.2.4. Quality of FRM Implementation

One impediment to FRM implementation is thedifficulty that expert groups have communicatingwith each other to take appropriate action in theface of a flood event. These groups can often havedifferent technical terminology for which to referto FRM processes. For example, professional emer-gency managers working with the public have dif-ficulty understanding weather forecast uncertaintiesprovided by meteorologists without assistance.(36)

Although forecast information must be used on acase-by-case basis when making disaster manage-ment decisions, recent work has shown that changesin forecast accuracy impact the economic value ofthat forecast.(38) If decision-makers have the flexibil-ity to wait for updated forecasts (for hurricanes inthis case), they can gain significant value from adopt-ing a dynamic decision model. The gains from adopt-ing a dynamic approach, as opposed to depending onan instantaneous strike probability, can be greaterthan those from reduced preparation time after anevacuation decision is made.(27)

After a flood, laypersons purchase flood insur-ance reactively to protect against future financialloss. Large increases in insurance coverage have been

8 Wood et al.

shown to occur after flooding, but, consistent witha reactive strategy based on event salience, manynew policies are cancelled as time passed and mem-ories of the flood fade.(12) This is an illustration of ageneral trend in individuals’ willingness to plan forthe consequences of a flooding event, in that thenumber of adaptations people are willing to maketo protect themselves from flood risk typically de-creases with the perceived probability of an eventand as past events are forgotten.(24,39) Though fewin Lave & Lave’s study(12) carried flood insurance,even though it was mandated, several individuals ex-pressed that insurance was unnecessary because so-cial service organizations (e.g., American Red Cross)provided benefits and compensation that were lessexpensive and better than insurance coverage. Somefederal program requirements also help to perpetu-ate the phenomenon of personal lack of responsibil-ity for flood planning (Section 1).

After a catastrophic flood event, some may feelthat their quality of life has suffered permanent neg-ative impact and that their life will never be the sameagain.(37) The personality trait of belief in a just world(BJW)6 may help people to protect themselves psy-chologically when faced with a perceived injustice,like experiences in the aftermath of flooding. Whenthose high in BJW are faced with an apparent injus-tice, they try to account for it or compensate in someway; this buffering process appears to help such peo-ple cope with the event. BJW has been shown to pro-vide psychological protection for victims of floods,and may be important when predicting posteventmental health after other natural disasters.(41)

3. EXPERT MODELING

3.1 Method

Developed initially for facilitating design of riskcommunication plans, the expert modeling approachused here for identifying beliefs about a risk manage-ment topic of interest(20) has been applied in a varietyof domains.(45−47) Past successes with mental model-ing in understanding environmental management is-sues include projects on expert and lay knowledge ofclimate change,(48−50) environmental manager viewson climate change and biodiversity,(51) public percep-tions of carbon sequestration techniques,(52) lay per-

6Belief in a just world is the illusory belief that good things happento good people and bad things to bad people.

ceptions of flashfloods and landslides,(16) and numer-ous unpublished projects with private clients.(53)

An important step in the mental modeling pro-cess is the development of an EM, a visual rep-resentation in the form of an influence diagram(54)

that illustrates the key variables in the decision con-text, viewed as a system, as well as the directionand strength of their influence on other variables inthe system. In this case, a draft EM was first devel-oped with a small number of USACE experts. Thisdraft model was elaborated into an EM and seriesof submodels illustrating drivers of USACE FRMin a workshop with 11 USACE experts in Septem-ber 2008, and a series of supplemental interviewswith five engineers in October 2008.(55) The EM pro-vided the basis for developing a semistructured men-tal model interview protocol, with the goal of gain-ing further insights into specific drivers of the qualityof USACE FRM, as well as to understand any dif-ferences in perspectives between planners and engi-neers within USACE stemming from differences ineducation or roles within USACE.

3.2 Results

The simple EM of Influences on USACE FloodRisk Management (Fig. 1) is in the form of an influ-ence diagram,(54) a directed graph in which “influ-ences” (arrows) link related “nodes” (variables) inthe system. The direction of the arrow indicates thedirection of influence. The following is a narrative de-scription of the simple EM.

3.2.1. Drivers

The model starts in the upper left-hand cornerwith Political, Societal and USACE Internal Drivers,factors that establish the fundamental conditions thatdrive USACE activities.

3.2.1.1. Political drivers. Political drivers come fromother government entities that influence or controlUSACE FRM Activities. Political Drivers include:Congressional Directives that fund and mandate spe-cific FRM activities; Federal Policy that mandatesUSACE FRM priorities (e.g., economic develop-ment); and the NFIP which mandates insurance forareas with specified flood risk levels. NFIP drivesmuch flood control structure construction, aimed atreducing flood risk and removing areas from man-date to purchase flood insurance.


3.2.1.2. Societal drivers. Societal drivers come fromsociety at large, including: Economic Developmentpriorities that encourage national/regional economicdevelopment and damage protection; Public Health& Safety priorities to protect public health/welfare;Social & Community Impact priorities to protect so-cial networks and community organizations that re-flect the collective impact to individuals and the di-rect impact on social structures and organizations;Public Expectations regarding flood risk levels andflood control structure effectiveness; and Environ-mental Protection & Climate Change priorities re-garding protection of environmentally sensitive ar-eas (i.e., wetlands) and response to climate change,which is likely to influence flood risks.

3.2.1.3. USACE drivers. USACE Drivers are fac-tors internal to the USACE that influence FloodRisk Management activities such as: Funding andcost sharing requirements, which limit which projectswith positive benefit-to-cost ratios can be enacted;Mission & Mandate, covering areas such as publicprotection, war fighting, protecting water resourcesand the environment, maintaining waterway infras-tructure, and homeland security; Principles & Val-ues, focusing on relevance, readiness, responsibility,and reliability; Technology and technical innovationssuch as tools to model floods for risk assessmentand planning, enable knowledge sharing/retention,and improve flood control structure performance;Prioritization of Flood Risk Management activitiesbased on economic development and public healthand safety criteria; and Planning Assumptions, par-ticularly the appropriate or acceptable level of floodrisk.

3.2.2. Historic Flooding Events

The Drivers are particularly influenced by His-torical Flooding Events, including Hurricane Katrinaand historic flooding of the Mississippi River Val-ley. Experts believe the impact of these events hashad a significant influence on public perceptions andpublic expectations (Section 3.2.7.), which in turnstrongly influence Societal, Political, and InternalUSACE Drivers.

3.2.3. USACE Flood Risk Management

The nodes grouped together in the middle of themodel depict USACE Flood Risk Management activ-

ities, the actual tasks performed by Corps personnel,with the quality of upstream nodes significantly im-pacting downstream nodes. Specific activities whichimpact Quality of Flood Risk Management Imple-mentation include: Flood Control Construction, forexample, building dams and levees; Asset Manage-ment, for example, flood control structure operationand maintenance; Non-Structural Activities, such asinfluencing changes in zoning and development to re-duce risk via removal of susceptible structures andpeople from flood prone areas; Flood Preparedness,preparation activities such as warning system designand evacuation planning; and Flood Response, re-sponse activities such as emergency repair of leveesand other flood control structures. These Flood RiskManagement activities, in turn, influence the nodefound in the lower left portion of the Simple Modeldepicting Influences on Flood Risks.

3.2.4. Influences on Flood Risks

The node in the lower left corner of the SimpleModel represents Influences on Flood Risks. Theseinclude naturally occurring variables (e.g., weather,geography) and human influences such as flood con-trol structures, policies and individual decision mak-ing. It includes the following influences: Weather& Climate, such as hurricanes and other extremeweather events and the impact of climate change onthe frequency, magnitude and distribution of events;Geography, Topography & Demography, the com-bination of land masses, waterways, and coastal ar-eas potentially impacted by floods and populationdensity that gives rise to flood risk exposure; De-velopment and Land Use, the concentration of resi-dential development and other land uses that affectrisk exposure magnitude and impact potential; FloodControl Structures, such as dams and levees that pro-tect against flood event damage; Non-Structural Mea-sures, such as activities that reduce flood risk expo-sure by limiting development in flood prone areasor improving warning and evacuation procedures toprotect public health and safety; and Flood Response& Mitigation, activities such as emergency repair oflevees and other flood control structures, drainage offlooded areas, and evacuation of people.

3.2.5. Collaboration, Coordination, andCommunication

The nodes in the upper and right section of themodel depict influences on Corps activities related

10 Wood et al.

FEMA

OrganizationalStructure

DivisionsDistricts

SharedStandards &Processes

Quality ofCollaboration, Coordination,

& Communication with Other Stakeholder

Groups

WorkforceCapacity

Quality of Internal USACE Collaboration,

Coordination, & Communication

Knowledge & Workforce

Sharing

TechnologyComputer Modeling

e-Collaboration

ParkService

HUD

USGS

EPA

Communications

RoleClarity

RegionalDifferences

USACE CultureSilo Mentality

USACELeadership

Quality

Quality ofPublic Engagement

Risk CommunicationsNEPA Processes

EducationOutreach

Quality ofCollaboration, Coordination,

& Communication withGovernment Partners

E-MailWebsitesList Serves

Communitiesof Practice

HistoricActivities & Interactions

SharedUnderstanding

Appropriate Level of Flood Risk

MOUs

WorkforceCapabilityExperienceExpertise

Time

KnowledgeRetention

DocumentationTraining

WorkforceMorale

RecognitionReward

Retirement& Recruitment

Annuitants

WorkloadSharing &Balancing Workforce

Distribution &Mobilization

RegionalWorkload

Assessment WorkforceResourceLeveling

ResourceIndustries

State & LocalCost-Sharing

Partners

Bureauof Land

Reclamation

DevelopmentGroups

ResidentialCommercial

AquacultureGroups

(Fisheries)

Utilities &Hydropower

Groups

AgricultureGroups

RecreationGroups

EnvironmentalGroups

NavigationGroups

Tech./Sci.AdvisoryGroups

StatutoryAdvisoryGroups

Information QualityRepresentativeness

Completeness

USACEFlood Risk ManagementFlood Risk Issue Identification

Flood Risk AssessmentFlood Risk Management Planning

Flood Risk ManagementImplementation

•

••

•

•

•

•

••

•

••

••

•

••

••

••

••

••

••

Fig. 2. Detailed submodel of collaboration, communication, and coordination.

to Workforce Capacity and to USACE Collabora-tion, Coordination, and Communications, internallyand externally as well as the Quality of USACE Pub-lic Engagement. These nodes are presented in moredetail in the Detailed Sub-Model of Collaboration,Communication, & Coordination (Fig. 2). WorkforceCapacity is the effectiveness of USACE workforcebased on alignment of workforce resources withworkload requirements. Quality of Internal USACECollaboration, Coordination, & Communication rep-resents the effectiveness of internal procedures, poli-cies and activities that affect the quality and effec-tiveness of working relationships within USACE.Quality of Collaboration, Coordination, & Commu-nication with Government Partners represents the ef-fectiveness of internal procedures, policies, and ac-tivities that affect the quality of working relation-ships with other Federal, State and Local agencies

and entities, such as Bureau of Land Reclamation,FEMA, EPA, USGS, and HUD, among others. Qual-ity of Collaboration, Coordination, & Communica-tion with Other Stakeholders represents the effective-ness of procedures, policies, and activities that affectthe quality of working relationships with other stake-holders. Quality of Public Engagement represents theUSACE activities directed at the public with respectto their Flood Risk Management activities, includingrisk and informational communications, educationalprograms and other outreach activities.

3.2.6. Desired Outcomes

The end point of the model is the Desired Out-comes node in the lower right hand corner. This re-flects the degree to which USACE FRM activitiescan achieve desired goals and objectives, such as:


Optimal USACE FRM; Effective Public Engage-ment; Optimal Public Preparedness & Response; andPositive Perceptions of USACE Trustworthiness &Competence. These outcomes are directly influencedby the Quality of USACE Flood Risk Managementactivities and by Individuals’ Mental Models of FloodPreparedness and Response.

3.2.7. Individuals’ Mental Models of FloodPreparedness and Response

This node represents the perceptions held by in-dividuals that determine their assessment of floodrisks of the effectiveness of flood risk controls andof USACE Flood Risk Management activities. Sec-tion 3 discusses the influence of mental models onpublic decision making and behavior, which can inturn influence the level of individual flood risk as wellas Societal Drivers expressed via public expectationsregarding flood risk and FRM (Section 3). This nodeis influenced by the Quality of Public Engagement.

4. MENTAL MODEL INTERVIEWCOMPARATIVE ANALYSIS

In order to clarify concepts defined in the EMs ofUSACE FRM (Section 3), follow-up mental modelinterviews were conducted. These interviews lever-aged the EM to develop a semistructured interviewprotocol with the goal of gaining further insightsinto specific drivers on the quality of USACE FRM,as well as to understand differences in perspectivebetween planners and engineers within the organi-zation on these topics. We were also interested ininvestigating differences between planners and engi-neers in the drivers they emphasize. Past work identi-fied differences in responsibilities and focus betweenthese two groups as they relate to the FRM process.This is in part because of differences in education,and in part due to role differences within USACE.Engineers tend to emphasize specific technical ele-ments of a process they are tasked with creating oroptimizing. In contrast, planners tend to focus moreon the integration of a technical solution with othersystems, as well as the role of stakeholders in systemutilization and management.(55)

These differences are consistent with those inother organizations and civil engineering efforts. Forexample, in the transportation sector, it has beensuggested that the U.S. interstate highway system’spoor aesthetics are due to engineers who primarilyfocused on maximizing traffic capacity, and were less

concerned about the effects designed structures hadon dividing or reorganizing communities.(56) Relat-ing to civil works more broadly, others have referredto the structures that evolve from an engineering fo-cus on utilitarian design as “stupifyingly bleak” interms of appearance and apparent lack of harmonywith the environments and communities where thesestructures are built.(57) The American Planning Asso-ciation, one of the major professional organizationsfor planners, emphasizes the profession’s holistic per-spective, working “ . . . to improve the welfare of peo-ple and their communities by creating more conve-nient, equitable, healthful, efficient, and attractiveplaces for present and future generations.(58)”

4.1. Method

4.1.1. USACE Mental Models InterviewParticipants

Engineers (n = 10), planners (n = 8), and pro-gram managers (n = 4) participated in one-on-one semistructured telephone interviews. Interviewsfrom some participants in the engineer cohort (n =5) were also used in the first part of this effort forEM refinement. Although explicit reference has notbeen made in past work to similarities or differencesbetween program managers and either engineers orplanners, the job functions of program managers inUSACE makes their role in the FRM process moresimilar to that of planners than to that of engineers.Therefore, for purposes of the current analysis, theywill be considered part of the planning cohort, yield-ing engineer and planner cohorts that are roughlyequal in size (n = 10 & 12, respectively). Inter-view participants were a convenience sample of se-nior personnel from a variety of organizational unitswithin USACE.

4.1.2. Interview Process

An interview protocol was adapted from thatused in previous research on USACE personnel’sbeliefs and knowledge related to the Corps currentFRM process(59) to identify important areas that re-quired further clarification or where additional infor-mation was otherwise valuable. As these changes tothe interview protocol were modest, and the infor-mation provided by earlier engineer participants wasconsistent with that from new engineer participants,data from the October 2008 sample of five engineersis included in the following analysis. Of particular in-terests was information related to:

12 Wood et al.

(1) The most important drivers of USACE FRM.(2) The scope and quality of USACE FRM activ-

ities.(3) Drivers of the quality of collaboration, coor-

dination, and communication.

(i) Internally (within USACE).(ii) Externally (with other government, NGO,

and layperson stakeholders).

Questions were asked to assess key features, be-liefs, and perceptions of the above as they exist incurrent FRM practice, as well as how these driversare forecasted to change in the future. In general, in-terview questions were structured to address a topicbroadly (e.g., “Thinking at the most general level,what would you say tends to drive, or significantly in-fluence the Army Corps’ flood risk management ac-tivities?”) and then focus on a series of different as-pects of the topic (e.g., “Has the significance of anyof these drivers changed over time?”; “How wouldyou describe the quality of Army Corps personnelworking together internally?”). This structure is con-sistent with past interviews developed from EM dia-grams.(20,47) Interviews were an hour long on aver-age (mean = 66 minutes; range = 45–82 minutes)and were collected from October 2008 through June2010. All telephone interviews were recorded.

4.1.3 Transcript Analysis.

Recordings of interviews were transcribed, andinvestigated via content analysis. An analyst with aPh.D. in Engineering and Public Policy with over 14years experience using mental model methodologyviewed responses for questions one-at-a-time, andidentified common themes that these responses ex-emplified using a two-pass process. In the first pass,each interviewee’s responses are read and associatedwith particular elements of the EM to allow for sub-sequent comparison of EM topics across interviewsand across questions. In the second pass, once in-terviewee responses are entered into a database; re-sponses are examined by question, across interviews,to identify more specific fundamental themes withintopic areas (and elements of the EM). Frequenciesof the number of participants mentioning a themeare recorded for each question, converted to a per-centage of total participants overall and by cohort(engineers or planners) and rounded to the nearest5%. Differences between cohorts are reported onlyif they are greater than 20%. Results will be reported

for a select subset of questions, based on their util-ity in clarifying concepts and processes in the EM re-lated to FRM, and in providing insights concerningdifferences between cohorts. Unless otherwise indi-cated, percentages reported in-text refer to the entiresample.

4.2. Results

4.2.1. Drivers of USACE FRM Activities

Participants provided a variety of factors con-sistent with the EM when asked about drivers ofFRM activities. They placed particular emphasis ondrivers related to public safety and reducing riskto the public (55% [n = 12]). These items consti-tute a key element of USACE’s mission. Respon-dents also frequently mentioned Congressional man-date, funding, and politics (45% [n = 10]); as wellas historical events, and a propensity for reactivityrather than proactive protection (45% [n = 10]). Re-spondents suggest Congressional mandate and pol-itics are related to funding in that project fund-ing depends on approval from Congress, especiallyfor large projects, and projects that Congress feelsare more closely related to USACE’s Congressionalmandate are more likely to be funded. Sometimes itis difficult to demonstrate the need for a project un-less a catastrophe has occurred, even though Corpspersonnel may feel problems are imminent. Politicalwill (and therefore funding) for FRM projects is of-ten closely tied to public outcry over these events,which is often strong at its onset but lacks persistence.

Responses related to local interest, support,sponsorship, and cost sharing are consistent withthis story (30% [n = 7]). A sizable proportion ofparticipants mentioned public interest in their re-sponses. Economics, National Economic Develop-ment (NED), and economic efficiency (25% [n =5]) rounded out the field of drivers participants feltdrove the USACE FRM process. Although interestand support for a project often occurs on a local orstate basis, the benefit/cost ratio used to determinewhich projects are suitable for funding only takes Na-tional Economic Development into account. Somerespondents perceived this as unfair, since localitiesoften bear a significant portion of the costs, but thebenefits to localities are underrepresented in cost–benefit analyses.

Across the dominant five themes related to FRMdrivers, engineers tended to mention more of thesethan planners in their responses. Engineers who were


interviewed for EM refinement were allowed to dis-cuss earlier questions more freely, thereby mention-ing more topics and perhaps driving this. In partic-ular, engineers focused on Congressional mandatesand political drivers (60% [n = 6] vs. 35% [n = 4] ofplanners), and on National Economic Development(NED)(59) assessment requirements (40% [n = 4] vs.10% [n = 1]). Several engineers explained that Con-gressional mandates tend to drives the Corps currentobjectives, especially when it comes to big projects.Concerning NED, engineers tended to focus on thefact that the assessment emphasizes national (versusregional or local) benefits, and efficient risk reduc-tion rather than optimal risk reduction.

4.2.1.1. Most Significant Drivers.

4.2.1.1.1. Currently most significant. Overall, re-spondents cited politics and congressional mandates(especially related to funding; 40% [n = 9]), eco-nomic development (30% [n = 7]), and response toweather and significant effects (25% [n = 6]), Corps’mission and mandate (15% [n = 3]), and guidingprinciples were also identified (15% [n = 3]) as themost significant drivers behind FRM activity. A smallnumber of participants also mentioned local inter-ests, environmental protection and impact, and tech-nological drivers. Between cohorts, planners weremuch more likely to emphasize politics and congres-sional mandates. In contrast, engineers tended tofocus on economic development and responding toweather and significant events. This contrast makessense in terms of organizational roles, as planners aremore likely to seek project financing. In contrast, en-gineers are more likely to conduct site visits in ar-eas currently or recently impacted by weather andother natural hazard events. Engineers’ tendency tofocus on NED figures is consistent with responses ondrivers of FRM activities, and indicates frustrationwith a system that appears to emphasize national eco-nomic development over local safety.

4.2.1.1.2. Ideally most significant. Respondents var-ied in their judgments regarding which drivers aremost significant. Two on which there was some agree-ment were public safety (20% [n = 4]) and re-sponding to storm events, including historical events(15% [n = 3]). Other topics that respondents thoughtshould be most significant included USACE guid-ing principles, environmental protection and climatechange, politics, and a systematic approach. A fewother participants indicated that something other

than economics (especially NED), or funding con-straints that limit ability to conduct proper initialstudies and complete projects correctly the first time,should be driving the FRM process.

4.2.1.2. Changes in Driver(s) Significance. Concern-ing the question of whether any of the drivers inthe EM have changed significance over time, respon-dents noted changes in five key areas: planning crite-ria, primarily based on past events (55% [n = 12]);political and legislative changes, especially funding(40% [n = 9]); technology (25% [n = 5]); and en-vironmental protection and climate change (25%[n = 5]). A few participants also mentioned resourcesas a driver that has changed in significance in recentyears. Between cohorts, engineers overwhelminglyemphasized changes in the importance of public pri-orities based on past events (80% [n = 8] vs. 35%[n = 4] for planners). Changes in public prioritieshave a resulting influence on planning criteria, andthe frequency of changes recently in planning crite-ria has lead to “ . . . an almost schizophrenic kind ofbehavior . . . ” as described by one planner. Severalrespondents reflected on the impact of HurricaneKatrina. One engineer noted:

Hurricane Katrina has forced us to enforce policies thatwe already had in place that we had became inconsis-tent in applying and still struggle with. From a corpo-rate perspective it has elevated public safety to the highlevel. We’re shifting how we go about planning theseprojects away from economics only to where we have amore balanced approach, where we try to balance eco-nomics, environmental, and societal impact.

4.2.2. Current Strengths in FRM Process &Capability

When asked what USACE does well in the areaof FRM, many participants cited the quality of struc-tural activities and projects (50% [n = 11]) and riskanalysis and flood risk modeling techniques (50%[n = 11]). Engineers (70% [n = 7]) were more likelythan planners (35% [n = 4]) to mention quality ofstructural work, focusing specifically on innovationsand efficiencies in materials and design, as well ashigh maintenance standards for legacy structures.Engineers also emphasized USACE’s exemplary re-sponse to flood emergencies (40% [n = 4]) relativeto planners (10% [n = 1]; 25% [n = 5] overall).Emergency efforts were hailed for short lags in re-sponse time and a high degree of technical expertiseand flexibility. Other strengths referenced were the

14 Wood et al.

quality of the overall planning process (25% [n = 5]),the expertise and commitment of Corps employees(25% [n = 5]), nonstructural activities (<10% [n =1]), and environmental protection (<10% [n = 1]).

4.2.3. Opportunities for Improvement in FRM

When it comes to opportunities for USACE toimprove the current FRM process, engineers wereespecially sensitive to the difference between pub-lic expectations and reality (90% [n = 9] vs. 40%[n = 5] planners; 65% [n = 14] overall), especiallyas it concerns the amount of protection a structure isdesigned to provide and the longevity of that protec-tion. The public is often unaware that the planningand protection horizon on a structural project is of-ten finite, that “[a] structure is going to be good for50 years and then, that it is going to give after that,”or that an area protected by a structure is one of theriskiest places to be if the structure is compromisedin any way. Planners and engineers both expresseda need for public education, citing misunderstand-ings of the meaning of probabilistic terms like “thehundred-year flood,” and lack of awareness of prob-abilistic terms.

Engineers were also quite critical of currentproject evaluation criteria (70% [n = 7]) relative toplanners (35% [n = 4]; 50% [n = 11] overall). Manyparticipants reported a perceived mutual exclusivityin NED and ecosystem restoration metrics for decid-ing which projects to fund. Some also found NED tobe an unfair metric, calculating costs for both federaland regional/local groups, but only concerning itselfwith benefit at the national (and not local) level. Oth-ers focused on a lack of certainty on how best to usethe growing variety of alternative metrics for projectplanning. One planner noted how more metrics maynot necessarily indicate better planning:

. . . you could have as many as three criteria besidesNED and NER [National Environmental Restoration]and other social effects and you’re supposed to some-how add up both the monetary and the nonmonetarybenefits so that means you go from a one-variable de-cision rule to as many as a six-variable decision rule.A one-sided game, I can understand. A six sided game,now we’re pressing the edges of my comprehension.

In contrast, planners tended to focus more onimproving collaboration internally and with partnerorganizations (40% [n = 5] vs. 0% engineers; 25%[n = 5] overall), as well as on shortening often ex-orbitantly long project timelines (35% [n = 4] vs.10% [n = 1] engineers; 25% [n = 5] overall). Some

planners recognized that improvements in collabo-ration, especially with external partners, might helpto hurry things along, but they also recognized that“our policies and processes are not for the faintof heart . . . . . . there are cities and towns out therewho are not engaging with us because of the timeand funding it takes us to get through our process.”The planning cohort (which includes project man-agers; see Methods) was also more sensitive thanthe engineering cohort (35% [n = 4] vs. 0%, re-spectively; 20% [n = 4] overall) to inefficiencies inthe project management process. Specifically, plan-ners commented that the process does not allow fornonlinearity. More critically, project management iscurrently yoked by a very cumbersome peer reviewprocess. One planner identified as many as seven lay-ers of review in the current process. This planneradded “There are times when I think we’ve spentmore on review than on the biggest possible mis-take you could make. In one project I’ve worked on,we’ve spent more on study than we will on imple-mentation. This is a baleful influence on our effec-tiveness.” Notably, several engineers in the samplecited peer-review processes as one of the strengthsof the Corps FRM process. Other areas for improve-ment identified by the sample included proper fund-ing levels and mechanisms to improve design qualityand limit inefficiencies with the current process (30%[n = 7] overall), slowing degradations to the Corpstechnical expertise base (25% overall [n = 6]; 40%[n = 4] engineers, 15% [n = 2] planners), accountingfor watershed impacts (20% [n = 4]), follow-up onCorps structures handed off to other organizationsfor maintenance (15% [n = 3]), incorporating morenonstructural and nonproject-based solutions (15%[n = 3]), and translating technical capabilities intopractical methods (15% [n = 3]).

4.2.4. Organizational Structure and Alignment withFRM Activities

Participant sentiment was mixed on the extentto which USACE’s current organizational structurewas aligned with FRM activities. While most partic-ipants (65% [n = 14]) commented positively, manyalso had negative comments on organizational andfunctional alignment (45% [n = 10]), with a few in-dividuals providing both positive and negative com-ments. The extent of positive and negative commentswas comparable across cohorts. Positives includedimproving alignment between planning and engi-neering functions/groups concerning FRM planning,


improved communication between the Corps keybusiness lines (e.g., navigation, recreation, FRM),and tactically sensible district and regional manage-ment frameworks organized at the basin and wa-tershed level. One key negative cited in the orga-nization’s structure was that the heavy emphasis onproject-based funding makes it very difficult to im-plement changes to FRM (or other processes) at aprogrammatic level. In addition, since flood risk isdetermined by the activities across the Corps otherbusiness lines, expertise in FRM is dilute across thoselines, making it difficult to improve current FRMprocess.

While the organization’s hierarchical structurewas hailed by some participants for the consistencyin procedure it creates, it was criticized by others asdiscouraging interdistrict, and especially interregion,collaboration. Many participants reported being en-couraged to search their districts and regions exhaus-tively for necessary expertise before reaching out toother organizational units. The peer-review processwas also hailed as a mixed blessing as it relates to theCorps’ historically hierarchically dominant structure.While the shift from a supervisor-approval model ofproject critique was appealing to several participantsbecause of its increased transparency and the diver-sity of viewpoints it can bring to a project, it was citedby others as leading to unnecessary study-phase costsand eroding accountability for later design flaws.

4.2.5. Quality of Internal USACE Collaboration

Opinions concerning the quality of collabo-ration, coordination, and communication withinUSACE were quite mixed, with some participants re-sponding positively (35% [n = 8]), while others ex-pressed a negative viewpoint (30% [n = 7]). A thirdgroup (25% [n = 5]) noted that the quality of col-laboration is heavily dependent on the individualsand districts composing the project team or work-ing group. A planner summarized this view as it con-cerned collaboration between districts:

It varies across the organization . . . . . . Some Districtsare very good at working with each other with no ‘turfjealousy’ or anything like that. There are other Districtsacross the Corps where . . . . . . they won’t even talk toeach other, even on issues, which may have relevance toboth of them.

When asked to reflect on strengths of Corpscollaboration, participants cited the effectiveness ofmultidisciplinary project teams (30% [n = 7]), and

the willingness of Corps personnel to interact witheach other (20% [n = 4]). In terms of opportunitygaps, participants frequently mentioned changes inmanagement and organization to promote culturalchange (60% [n = 13]), and a requirement for moreappropriate levels of resources, especially fundingand manpower (55% [n = 12]). A few also mentioneda need for technical expertise and oversight (15%[n = 3]). Specific suggestions for improvement in-cluded creating structures that encouraged a balanceof power between project managers and engineers,an increase in the number and diversity of engineer-ing personnel, and a shift from project-based Con-gressional funding to more “continuing authority-type” funding.

4.2.6. Factors Impacting USACE Collaboration,Coordination, & Communication Quality

Participants were asked about the ways in whichseveral factors influenced how well Corps personnelwork together. These factors were identified from theEM in an earlier investigation,(55) and included work-place quality, resources (especially funding and per-sonnel), organizational culture, and role clarity. Onmatters of workplace quality, participants noted thatthe organization has made good investments in lead-ership and leader development that are paying offin improving the quality of the workforce, collabo-ration, and the Corps capacity to complete projects.While sentiment on leadership training was good,that on training in general was more nuanced. Someparticipants noted “great improvements” in the pastten years, while several others stated that the abilityto conduct training is limited because of the project-based nature of most Corps funding.

On the impact of resources, particularly fundingand personnel capacity and availability, most engi-neers and some planners noted that the Corps work-force is stretched thinly across many projects, andthat coordination is one of the first items to be com-promised when everyone is trying to meet a sched-ule. Some participants believed constrained work-force issues were related to constraints in funding. Inparticular, respondents noted that many functionalunits (e.g., Communities of Practice, Major Subor-dinate Commands) have difficulty regularly meet-ing mission objectives because of limited fundingavailability, let alone communicate progress on thoseobjectives to others. Increased workforce turnoverfrom retirement and shorter employee tenure was

16 Wood et al.

also cited as a negative impact on collaboration, com-munication, and coordination. This secession meansthat established relationships between employees indifferent units need to be recreated as new talententers the organization to replace the veteran work-force.

On matters of organizational culture, many par-ticipants noted that the culture of the Corps encour-ages districts and regions to work autonomously, de-creasing the likelihood and willingness of districts tocooperate. That said, many also indicated that thiscultural feature was slowly starting to change towardincreased collaboration. The addition of younger,more flexible staff members and the introduction ofvirtual teaming and other collaboration tools werecited as some of the reason for this culture shift.

Most engineers and some planners noted thatrole clarity was an area of concern as it relates to in-ternal collaboration, coordination, and communica-tion. Some comments suggested that recent reorgani-zations in workflow and responsibilities have blurredwhat team members are specifically responsible forwhat tasks. Some cross-functional responsibilities,like decision documentation and other tasks that donot clearly fall within one discipline, are sometimesneglected because no team member wants the ad-ditional responsibility. One respondent noted caseslike these as a result of the communication lag thatoften follows a swift organizational change. Someother respondents noted that role clarity was good,but only in some situations, providing credence to thereorganization-in-progress view of recent role clarityissues.

4.2.7. Quality of Collaboration, Coordination, andCommunication with Partner Organizations

Most participants stated that relationships andpartnering efforts with other government organiza-tions are typically good (55% [n = 12]). Particularcited successes included the Advisory Committee onWater Information, recent work in dam safety pro-grams with the Bureau of Reclamation and the Fed-eral Energy Regulatory Commission, and the SilverJackets program done in cooperation with FEMA.Of the individuals that reported mixed results inworking with other agency partners (30% [n = 7]),complaints included some of the barriers that werereported to limit internal collaboration. These in-cluded limits to resources of the other agency andUSACE’s available resources to partner with that

agency, role ambiguity that arises from overlappingjurisdiction, and a history of poor interaction withsome partners (particularly FEMA).

Concerning work with nongovernment organi-zations (NGOs; community, industrial, commercial,environmental, and other interests), nearly all engi-neers (80% [n = 8]) and some planners (35% [n =4]) regarded interactions with these groups as usu-ally positive. These respondents typically regardedNGOs as partners with whom shared goals could beestablished. Nature Conservancy was cited by severalparticipants as a specific partner who shared partic-ularly good relations with the Corps, though otheragencies were also mentioned. Interestingly, of indi-viduals who reported problems working with NGOs,some were planners (40% [n = 5]), but only onewas an engineer (10%). Most individuals who re-ported problems with NGOs noted that those groupsthat share a formal partnering agreement (like Na-ture Conservancy) share a good relationship with theCorps. NGOs that do not have a financial stake in aproject were noted by several planners as being tol-erated, but not considered or included when it ulti-mately comes to project decisions.

4.2.8. Quality of Public Interaction and Partnering

When asked about interaction with the public,some participants (40% [n = 9]) reported that in-teraction does happen with the public writ largethrough regulated public notices and hearings, butthat specific individuals were rarely consulted andthat the public was rarely considered a formal part-ner on projects. Many participants noted that re-lations could be improved with public stakeholders(50% [n = 11]). Specific suggestions included bet-ter communicating risk mitigation plans and currentrisk exposure to the public, being more responsive topublic concerns and needs, and being more consis-tent as an organization in the degree to which projectprogress and risks are communicated. Several whoreported that public interactions could be better alsonoted that work was in progress to improve pub-lic communication, but that these initiatives are stillunder development or otherwise have yet to bearfruit.

4.2.8.1. Accomplishing Synergy with ExternalPartners. When asked whether working relation-ships with external stakeholders were synergistic,


many participants reported that they were (45%[n = 10]), though just as many others (45% [n =10]) reported that they could be more synergisticthan they are currently. There were no meaningfuldifferences between planners and engineers on thisissue. Those who reported synergy with externalstakeholders noted that these partners often shareinformation with USACE that the Corps does nothave itself, and that the purpose of many of the pro-cesses in place is to bring groups together. Of thosereporting opportunities for improvement, severalreported that USACE was trying to improve andexpand cooperative relationships with other groups,and a few others reported that these relationshipswere improving already.

4.2.9. Engineers and Planners: An Essential Tensionand Potential for Agreement

Engineers and planners did show some meaning-ful differences in terms of their opinions regardingcurrent FRM process and collaboration with internaland external partners. Many areas in which engineerswere more vocal than planners are consistent with aneed for certainty on the part of engineers. For ex-ample, engineers tended to believe that more pro-grammatic (versus project-based) funding was nec-essary to make sure technical capacity is maintainedand kept up-to-date with current technologies. Theyalso noted that the layered peer review approach waseffective in ensuring that structural solutions to FRMeffectively achieved plan parameters. Engineers alsotended to emphasize public safety more than plan-ners, with a desire to make safety a more prominentconcern in designing risk management plans. Evenwith regard to communicating with the public, engi-neers expressed a desire to align expectations closerwith reality, which they thought would increase thecertainty of appropriate public utilization of FRMplans.

In contrast, planners were much more inter-ested in making the FRM process more efficient,and tended to take more of a systems view of FRMand other risk management processes. Planners wereslightly more likely than engineers to express frustra-tion over Congressional mandates and the way thatlarge projects are funded. If Congress asks for higherlevees in Region X, the Corps must build them, eventhough they may know that more effective and lessexpensive ways exist to reduce flood risk in that re-gion. While agreeing with engineers that a need existsfor better public communication about flood risks,

it appears that this interest in communication is tar-geted toward getting the public to better understandwhat tools are reasonable to reduce flood risk. Engi-neers in this sample appeared more focused on com-municating what a tool or structure does given that ithas already been put in place.

While planners and engineers tended to agreethat more funding and manpower was required to en-sure the Corps could do its work effectively, plannerswere much more likely to point out inefficiencies inthe existing planning process. Engineers hailed thepeer-review process as key for quality assurance, butplanners noted that the many layers of review meantincreased costs in terms of money and manpower.Engineers tended to focus on maintaining technicalcapacity so it could be used to fully understand arisk management problem in the event such exper-tise may be needed; planners, in contrast, noted thatsome information is not worth generating, in that itwill be unlikely to change a current course of actioneven after that information is known. Although plan-ners agree that multicriteria assessments that includecriteria other than NED can be helpful for moreeffective planning, generating additional data can be-come very costly, and the additional factors to con-sider can make it more difficult to identify the bestcourse of action.

These differences between engineers and plan-ners might signal an essential tension: engineers tendto focus on doing the most good with each projectthey are tasked with, while planners are more fo-cused on maximizing the public good across projects.As is manifest in USACE’s FRM process, engineerstend toward more project study than may be practi-cally useful because they discount costs involved inacquiring that information, and how those increasedcosts impact ability to conduct other projects. So,contrary to the sentiment expressed by many en-gineers, staffing levels may be sufficient but morestaff time is used during study phase for each projectthan may be useful. This leads to a net shortfall inavailable staffing. Planners and program managers,interested in making the most out of available re-sources across the portfolio of current projects, useNED (and perhaps a few other metrics) to formu-late a quick decision rule about the best projects tofund. This rule may discount the importance of pub-lic safety and other noneconomic indicators. In addi-tion, to the engineer, a planning heuristic like the onedescribed here may lend itself to the appearance thatthe planner or program manager is not doing all thathe or she can for any given project.

18 Wood et al.

5. DISCUSSION

The preceding review makes it clear that, de-spite high profile cases of catastrophic flooding inrecent history (e.g. 2004 Indian Ocean Earthquake,Hurricane Katrina in 2005), lay stakeholders remainlargely unaware of the processes that create floodrisk and the measures that may be taken to mitigatethis risk. Several recommendations are available inthe literature to improve this knowledge, and tak-ing advantage of these recommendations may facil-itate improvement in the FRM process not only fromUSACE’s perspective, but potentially also from theperspective of other government agencies and stake-holder groups.

The Expert Model of Influences on USACEFRM (Fig. 1) shows that stakeholderperceptions (Individuals’ Mental Models of FloodPreparedness and Response) have a significant anddirect influence on Desired Outcomes concernedwith improving the USACE FRM process. Themodel also suggests that these individual mentalmodels can be influenced in two ways: directlyfrom improvements to USACE’s FRM process,and indirectly through improving the Quality ofPublic Engagement. The direct pathway mightinclude initiatives to address public perceptions andunderstanding of USACE’s FRM process. USACEFRM improvements could improve the quality ofpublic engagement by explicitly encouraging col-laboration and communication between laypersonsand USACE in the planning process. In addition,the reviewed literature suggests that educatingstakeholders about the risks they are susceptible toin an actionable way has great beneficial potential, ifdone in a way that accounts for what they can do toreduce the risks they face from floods. Stakeholdereducation may also decrease pressure on Congressto fund structural flood management projects, whichmay be counterproductive in the long-term.

Interview results suggest that USACE is makingan effort to partner with external stakeholder groups,but has difficulty growing these partnerships becauseof resource limits. Efforts are also underway to im-prove internal collaboration, coordination, and com-munication, but barriers that are creating difficultiesin working with external partners also create inter-nal issues. Limited resources, in addition to increasedemployee turnover and some organizational cultureartifacts, were cited as causes for why internal col-laboration is not better. However, many expressedthat current internal coordination was already good,

and some blamed a lack of collaboration on idiosyn-crasies of individual working groups or group mem-bers, rather than systemic issues.

Future work should seek effective methods foridentifying stakeholder perceptions and improvingstakeholder engagement in USACE FRM processes.Future work should also expand use of decisionanalysis-based mental modeling methods, for exam-ple in optimizing other processes that are part ofUSACE’s mission. A few possible avenues may becoupling a mental model influence diagram methodwith statistical techniques (e.g., structural equationmodeling) or machine learning algorithms to pro-vide more refined information about the degree ofinfluence between nodes in the model. More specificknowledge about how concepts relate to each otherin mental model influence diagrams will help to in-form future actions of USACE research teams, man-agers, and policymakers alike.

The comparative analysis of engineer and plan-ner perceptions represents a first effort to understandthe current views of USACE engineers and plannerson the agency role and needs for FRM in a post-Katrina world. Future research in this area shouldtake two directions. First, concerning the FRM pro-cess specifically, efforts should be taken to under-stand the key issues that a FRM process should ad-dress, and each step of the current process should beassessed for the extent to which it meets these needs.Steps in the process that add less value to the out-come should be considered for revision or removalfrom the FRM process. Second, concerning the ten-sion this research has uncovered between engineersand planners in their need for project versus sys-tem optimization, future work should focus on iden-tifying methods and processes that reduce this ten-sion. Ultimately, both groups are interested in doingthe best they can to help USACE effectively servecitizens and the various stakeholder groups theymake up.

The engineer’s focus on project performance andthe planner’s goal of portfolio performance are notmutually exclusive goals. Concerning the balancebetween project study phase costs and staff availabil-ity, decision analytic techniques like value of infor-mation(60,61) can be used to estimate the utility of ac-quiring additional information before resources arespent to obtain that information. The peer reviewprocess might also be simplified to the extent thatsome redundancies can be eliminated without ad-versely affecting project quality. In principle, thesechanges may serve to reduce the amount of staff


time and resources used unnecessarily, balancing thework-effort equation in a way that does not compro-mise project quality for system efficiency. A crucialconstraint in identifying the best way to serve thepublic good is in identifying an acceptable tradeoffbetween similar, though seemingly mutually exclu-sive, ends.

ACKNOWLEDGMENTS

We would like to thank Drs. Jack Davis andEdmond Russo of the US Army Corps for advice anddiscussions. Gordon Butte, Sarah Thorne, Alexan-der Tkachuk, and Laure Canis helped in manuscriptpreparation and review. Information described andthe resulting data presented herein, unless otherwisenoted, were obtained from research conducted underthe sponsorship of the Civil Works Basic ResearchProgram by the U.S. Army Engineer Research andDevelopment Center. Permission was granted by theUSACE Chief of Engineers to publish this material.

REFERENCES

1. Cigler BA. The “Big Questions” of Katrina and the 2005 greatflood of New Orleans. Public Administration Review, 2007;67:64–76.

2. Johnson BB. Testing and expanding a model of cognitive pro-cessing of risk information. Risk Analysis, 2005; 25(3):631–650.

3. Gheytanchi A, Joseph L, Gierlach E, Kimpara S, Housley J,Franco ZE, Beutler LE. The Dirty Dozen – Twelve failuresof the Hurricane Katrina response and how psychology canhelp. American Psychologist, 2007; 62(2):118–130.

4. Fleming G. Learning to live with rivers: The ICE’s re-port to government. Proceedings of the Institution of CivilEngineers-Civil Engineering, 2002; 150:15–21.

5. Fleming G. How can we learn to live with rivers? The findingsof the Institution of Civil Engineers Presidential Commissionon flood-risk management. Philosophical Transactions of theRoyal Society of London Series a-Mathematical Physical andEngineering Sciences, 2002; 360(1796):1527–1530.

6. National Research Council. First Report from the NRC Com-mittee on the Review of the Louisiana Coastal Protection andRestoration (LACPR) Program. Washington, DC: The Na-tional Academies Press, 2008.

7. U. S. Army Corps of Engineers, New Orleans District.Louisiana Coastal Protection and Restoration Technical Re-port Draft. 2008.

8. Working Group for Post-Hurricane Planning for theLouisiana Coast. A New Framework for Planning the Futureof Coastal Louisiana after the Hurricanes of 2005. Cambridge,MD: University of Maryland Center for Environmental Sci-ence, 2006.

9. Hecker EJ, Zepp LJ, Olsen JR. Presented at the The Euro-pean Conference on Flood Risk Management. Keble College,Oxford, UK, 2008.

10. Rabbon PD, Zepp LJ, Olsen JR. Presented at the EuropeanConference on Flood Risk Management. Keble College, Ox-ford, UK, 2008.

11. U.S. Army Engineer Institute for Water Resources. Alexan-dria, VA: U.S. Army Corps of Engineers, 2009.

12. Lave TR, Lave LB. Public perception of the risks offloods: Implications for communication. Risk Analysis, 1991;11(2):255–267.

13. Vari A, Linnerooth-Bayer J, Ferencz Z. Stakeholder views onflood risk management in Hungary’s Upper Tisza Basin. RiskAnalysis, 2003; 23(3):585–600.

14. USEPA. Terms of Environment: Glossary, Abbreviations,and Acronyms. Washington, DC: United States Environmen-tal Protection Agency, 1997.

15. Kolkman MJ, van der Veen A, Geurts P. Controversies in wa-ter management: Frames and mental models. EnvironmentalImpact Assessment Review, 2007; 27(7):685–706.

16. Wagner K. Mental models of flash floods and landslides. RiskAnalysis, 2007; 27(3):671–682.

17. Siegrist M, Gutscher H. Flooding risks: A comparison of laypeople’s perceptions and expert’s assessments in Switzerland.Risk Analysis, 2006; 26(4):971–979.

18. Bostrom A, Fischhoff B, Morgan G. Characterizing mentalmodels of hazardous processes: A methodology and an appli-cation to radon. Journal of Social Issues, 1992; 4:85–100.

19. Godfrey-Smith P. On folk psychology and mental represen-tation. Pp. 147–161 in Chlapin H, Staines P , Slezak P, (eds).San Diego, CA: Elsevier Inc., 2004.

20. Morgan MG, Fischhoff B, Bostrom A, Atman C. Risk Com-munication: A Mental Models Approach. New York: Cam-bridge University Press, 2002.

21. Doyle JK, Ford DN. Mental models concepts for system dy-namics research. System Dynamics Review, 1998; 14(1):3–29.

22. Doyle JK, Ford DN. Mental models concepts revisited: Someclarifications and a reply to Lane. System Dynamics Review,1999; 15(4):411–415.

23. Gruntfest E, Ruin I, Hayden M, Schmidt C. Presented at the33rd Annual Hazards Research and Applications Workshop,Broomfield, CO, 2008.

24. Rasid H, Haider W. Floodplain residents’ preferencesfor non-structural flood alleviation measures in the RedRiver basin, Manitoba, Canada. Water International, 2002;27(1):132–151.

25. Buckland J, Rahman M. Community-based disaster manage-ment during the 1997 Red River Flood in Canada. Disasters,1999; 23(2):174–191.

26. Clark DE, Novotny V, Griffin R, Booth D, Bartosova A,Daun MC, Hutchinson M. Willingness to pay for flood andecological risk reduction in an urban watershed. Water Sci-ence and Technology, 2002; 45(9):235–242.

27. Regnier E, Harr PA. A dynamic decision model applied tohurricane landfall. Weather and Forecasting, 2006; 21:764–780.

28. Pielke RA. Nine fallacies of floods. Climatic Change, 1999;42(2):413–438.

29. Lynch AH, Brunner RD. Context and climate change: Anintegrated assessment for Barrow, Alaska. Climate Change,2007; 82:93–111.

30. Johnson BB, Slovic P. Lay views on uncertainty in envi-ronmental health risk assessment. Journal of Risk Research,1998; 1:261–279.

31. Viscusi WK. The lulling effect: The impact of child-resistantpackaging on asprin and analgesic ingestions. The AmericanEconomic Review, 1984; 74(2):324–327.

32. Pahl-Wostl C. The implications of complexity for integratedresources management. Environmental Modelling & Soft-ware, 2007; 22(5):561–569.

33. Fischhoff B, Slovic P, Lichtenstein S. Lay foibles and expertfables in judgments about risk. The American Statistician,1982; 36(3, pt. 2):240–255.

34. Fischetti M. Drowning new orleans. Scientific American,2001; 284:76–85.

20 Wood et al.

35. Graham WJ. Should dams be modified for the probable max-imum flood? Journal of the American Water Resources As-sociation, 2000; 36(5):953–963.

36. Faulkner H, Parker D, Green C, Beven K. Developing atranslational discourse to communicate uncertainty in floodrisk between science and the practitioner. Ambio, 2007;36(8):692–703.

37. Tapsell SM, Turnstall SM. “I wish I’d never heard of Ban-bury”: The relationship between ‘place’ and the health im-pacts from flooding. Health & Place, 2008; 14:133–154.

38. Considine TJ, Jablonowski C, Posner B, Bishop CH. Thevalue of hurricane forecasts to oil and gas producers inthe Gulf of Mexico. Journal of Applied Meteorology, 2004;43(9):1270–1281.

39. McClure J, Williams S. Community preparedness: Counteringhelplessness and optimism. Palmerston North, Netherlands:Dunmore Press, 1996. Pp. 237–254.

40. Lenntorp E. Hell or high water? An economic analysis of theSwedish institutions for flood risk management. The GenevaPapers, 2008; 33:323–336.

41. Otto K, Boos A, Dalbert C, Schops D, Hoyer H. Posttrau-matic symptoms, depression, and anxiety of flood victims: Theimpact of the belief in a just world. Personality and IndividualDifferences, 2006; 40(5):1075–1084.

42. Pahl-Wostl C. The importance of social learning inrestoring the multifunctionality of rivers and flood-plains. Ecology and Society, 2006; 11(1). Available at:http://www.ecologyandsociety.org/vol11/iss1/art10/.

43. McCarthy P, Talley WK. Evidence on risk compensation andsafety behaviour. Economics Letters, 1999; 62:91–96.

44. Pretzman S. The effects of automobile safety regulation. Jour-nal of Political Economy, 1975; 83(4):677–725.

45. Downs JS, Bruine de Bruin W, Fischhoff B. Parents’ vacci-nation comprehension and decisions. Vaccine, 2008; 26 1595–1607.

46. Wood MD, Mukherjee A, Bridges T, Linkov I. A mentalmodeling approach to study decision-making in dynamic taskenvironments. In Chinyio E, Olomolaiye P, (eds). London:Wiley-Blackwell, 2010.

47. Darisi T, Thorne S, Iacobelli C. Influences on decision makingfor undergoing plastic surgery: A Mental Models and quanti-tative assessment. Plastic and Reconstructive Surgery, 2005;116:907–916.

48. Lowe TD, Lorenzoni I. Danger is all around: Eliciting expertperceptions for managing climate change through a mental

models approach. Global Environmental Change-Human andPolicy Dimensions, 2007; 17(1):131–146.

49. Read D, Bostrom A, Morgan MG, Fischhoff B, Smuts T.What do people know about global climate change? 2. Surveystudies of educated laypeople. Risk Analysis, 1994; 14(6):971–982.

50. Bostrom A, Morgan MG, Fischhoff B, Read D. What do peo-ple know about global climate-change .1. mental models. RiskAnalysis, 1994; 14(6):959–970.

51. Hill SD, Thompson D. Understanding managers’ viewsof global environmental risk. Environmental Management,2006; 37(6):773–787.

52. Palmgren CR, Morgan MG, De Bruin WB, Keith DW. Initialpublic perceptions of deep geological and oceanic disposal ofcarbon dioxide. Environmental Science & Technology, 2004;38(24):6441–6450.

53. Decision Partners LLC. Success Stories, 2010. Available at:http://www.decisionpartners.com/success.html, Accessed onSeptember 20, 2010.

54. Howard RA, Matheson JE. Influence diagrams. DecisionAnalysis, 2005; 2(3):127–143.

55. Linkov I, Wood MD, Bridges T, Kovacs D, Thorne S,Butte G, Presented at the 2009 IEEE International Confer-ence on Systems, Man, and Cybernetics. San Antonio, TX,2009.

56. Brown JR, Morris EA, Taylor BD. Planning for cars in cities:Planners, engineers, and freeways in the 20th century. Jour-nal of the American Planning Association, 2009; 75(2):161–177.

57. Russell JS. Uncivil engineering + infrastructure, a role for ar-chitects. Architectural Record, 1993; 181(6):98–99, 145

58. American Planning Association. What is planning? 2010.Available at: http://www.planning.org/aboutplanning/whatisplanning.htm#2, Accessed on July 8, 2010.

59. Yoe C. National Economic Development ProceduresManual: National Economic Development Costs. WestChester, PA: Greely-Polhemus Group, IWR Report 93-R-12,1993.

60. Linkov I, Moberg EA. Multi-Criteria Decision Analysis: En-vironmental Applications and Case Studies. Boca Raton, FL:CRC Press, 2011.

61. Linkov I, Bates ME, Canis LJ, Seager TP, Keisler JM. Adecision-directed approach for prioritizing research into theimpact of nanomaterials on the environment and humanhealth. Nature Nanotechnology, 2011; 6:784–787

Flood Risk Management: US Army Corps of Engineers and Layperson Perceptions

Documents

Transcript of Flood Risk Management: US Army Corps of Engineers and Layperson Perceptions