The purpose of the study is to enhance the hurricane preparedness of populations in the coastal zones of Rhode Island. The aim is to determine the extent of flows to expect for varying modes of transportation under hurricane threat through survey conduct, and the development of a multi-modal evacuation plan. Our objectives are to 1) assess the motivational factors that provoke the evacuation behavior, 2) assess the level of correlation of the pre-determined factors with the actual risk posed by the hurricane, 3) implement a phased multi-modal evacuation that minimizes overall risk to society, assuming a high level of correlation. (Assuming a low correlation, phasing complies with the motivational factors rather than actual risk.) A strategic risk analysis based on engineering methods combined with a social analysis of a community and its resources determines the actual risk posed by worse-case scenario hurricanes. Actual risk then determines the population segments that demand priority consideration assuming a phased evacuation. Such phased evacuation may prove self-enforcing given its synchronization with the emergent evacuation behavior.
Intermodal
Recent hurricanes, Katrina and Rita, have demonstrated a need for multi-modal evacuation under hurricane threat. A commonly accepted framework of mass disaster evacuation postulates the dependency of the actual risk posed to actual resilience and actual exposure. Actual resilience relates closely to the material and conceptual resources available within reach or through the larger community including the availability of transportation means. Actual exposure varies with such factors as actual location relative to disaster agent’s pathway, magnitude and scope. Actual resources can be employed to control actual exposure through population evacuation, for instance, thereby controlling risk.
This proposal derives and mitigates the actual risk posed by hurricanes to the coastal population segments of Rhode Island. To this end, it assesses the actual resilience and the actual exposure to hurricanes at varied coastal transportation analysis zones. It further scrutinizes the feasibility of enhancing resilience, to minimize overall societal risk, through the enactment of a multi-modal phased evacuation. The empirical data necessary to study conduct is gleaned from disaster-related literature, proposed surveys of representative groups of potential or past evacuees, emergency management agencies and readily available GIS databases.
Task 1 – Review of the literature on Evacuation Behavior, Evacuation Management and Operational Plans
All efforts will be made to review results from the social science disciplines on the predictors of the evacuation behavior. Further, the engineering literature will be scrutinized for evacuation models of multi-modal participation rates, phased evacuation, evacuation destinations and plans.
Task 2 – Assessment of Socio-Economic Factors Motivating the Evacuation Behavior (data collection and analysis)
This research will utilize two approaches to gather insights into the hurricane evacuation behavior from local residents and various agencies/officials. They will be: 1) focus groups and 2) in-depth interviews. There will be four focus groups designed as an outlet for individuals to discuss critical information regarding evacuation behavior. The focus groups will be: a) health care professionals, b) emergency management relief officials, c) political leaders, and d) community leaders. Groups A-C will be videotaped and presented to Group D for their reaction and commentary. All focus groups, including Group D, will be videotaped. Researchers will study the videotapes to acquire information regarding repeated themes in the focus groups' discussions.
A purposive sample of local residents (approximately 25-30) who have experienced hurricanes and residents without hurricane experience will be interviewed to obtain information on their evacuation behavior as outlined by the various focus group spokespersons and their attitudes about the necessity and timing of evacuation. In addition, residents will be queried about their resources in an evacuation situation, such as, financial resources, familial support structure, etc. The interviews will use a semi-structured interview schedule (questions), which will consist of closed and open-ended questions. The interviews will be transcribed; and Nu*Dist or Ethnograph programs will be employed to tease out important themes presented in the interview data.
Task 3 – Derivation of Actual Risk Posed by Hurricane to Transportation Analysis Zones
Risk is defined as the product of the probability of harm and the harmful consequences. The harm agents resulting from a hurricane event include: storm surge, inland flooding and winds. Various software packages provided by either the Federal Emergency Management Agency (FEMA) or the National Hurricane Center (NHC) enable assessment of the probability and the consequences of these harm agents. The below sections introduce the software packages that will be employed to the derivation effort of both, the risk components.
Assessment of Harm Probability
FEMA maps provide information on “A” floodplain zones, subject to a 100-year flood, “V” zones subject to a 100-year flood characterized by breaking wave action, and “X” zones subject to a 500-year flood. SLOSH enables determination of the extent of winds and flooding resulting from hurricanes. “SLOSH is a computerized model run by the National Hurricane Center (NHC) that estimates storm surge heights and winds resulting from historical, hypothetical, such as worse-case scenario, or predicted hurricanes by taking into account hurricane pressure, size, forward speed, track and winds. Graphical output from the model displays color coded storm surge heights for a particular area in feet above the model's reference level, the National Geodetic Vertical Datum (NGVD), which is the elevation reference for most maps”, (National Hurricane Center Website, 2005).
“The SLOSH calculations are applied to a specific locale's shoreline, incorporating the unique bay and river configurations, water depths, bridges, roads and other physical features. The model accounts for astronomical tides (which can add significantly to the water height) by specifying an initial tide level, but does not include rainfall amounts, river flow, or wind-driven waves. However, this information is combined with the model results in the final analysis of at-risk-areas”, (National Hurricane Center Website, 2005).
The SLOSH and FEMA maps will help determine the actual populations at-risk to flooding, notwithstanding evacuation, for hypothetical (worse-case scenarios) hurricanes within the Saffir/Simpson scale. Needless mention that overall probability of harm does not depend solely on the potential risk from flooding but also on the availability of resources, such as means of transportation and strong social links, that can mitigate those risks. The risk tree developed will most certainly account for a variety of resources.
Assessment of Hurricane Consequences
HAZUS-MH MR1, a FEMA’s software program, will estimate potential losses from hypothetical (worst-case scenarios) hurricanes in all categories of the Saffir/Simpson scale. “In HAZUS-MH MR1, current scientific and engineering knowledge is coupled with the latest Geographic Information Systems (GIS) technology to produce estimates of hazard related damage before, or after, a disaster occurs. HAZUS-MH takes into account various impacts of the consequences of a hazard event such as: physical damage (damage to residential and commercial buildings, schools, critical facilities, and infrastructure), economic loss (lost jobs, business interruptions, repair and reconstruction costs), and social impacts (impacts to people, including requirements for shelters and medical aid).
Task 4 – Correlation Analysis of Motivational Factors to Actual Risk
The motivational factors derived will be correlated with the actual risks, derived as well. A strong correlation indicates the appropriateness of a risk-based phased evacuation. Phasing will attempt to minimize overall societal risk over time in complete synchronization with the emergence of the evacuation behavior. (Hence, phasing will in essence be self-enforcing. The reader may note that no actual mechanisms exist at present to enforce a phased evacuation.) Needless mention that Task 2, 3 and 4 will be conducted iteratively. The discovery, in Task 2, of risk-related motivational factors not accounted for in the risk models of Task 3 will prompt revisions of the latter models towards factor inclusion. This revision in turn will impact the correlation achieved in Task 4.
Task 5 – Implementation of Multi-Modal Phased Evacuation
A phased evacuation plan that provides priority evacuation consideration to varied population segments in decreasing order of the actual risk confronted will be implemented. This dynamic trip generation, modal-split and traffic assignment solution will attempt to satisfy all evacuation needs as they become manifest. It will employ a rolling horizon to periodically update input demand levels. It is expected that a combined logit model will be implemented in this model task using TRANSCAD or other specialized evacuation software. TRANSCAD is readily available for use at URITC. All attempts will be made to satisfy, through public means of transportations, the demand not met by private automobiles and/or shelter. Further, the realisms of highway delay costs may be enhanced using simulated travel time actually or the Highway Capacity Manual’s model instead of the Bureau of Public Roads delay functions. A hypothetical scheme that evacuates populations toward the nearest destination located outside of the evacuation zone will be considered. Compliance of threatened populations with such a scheme is doubtful. However, it may provide a benchmark for gauging the efficiency of the normal scheme of evacuation toward preferred destination, also analyzed. (The use of FEMA and SLOSH (Sea, Lake and Overland Surges from Hurricanes) maps is anticipated to estimate the extent of the evacuation zone given the probability of flooding due to heavy rains and worse-case scenarios storm surge, respectively. Also, the motivational factors of the evacuation behavior may shed light on compliance levels with a nearest safe destination scheme.)
Task 5 (Alternate) – Prediction of Evacuation Behavior, Multi-Modal Origin-Destination Flow Rates over Time (Use of Census and other Data)
The motivational factors, pre-determined in Task 2, will be employed to the determination of the evacuation behavior using census and other data for varied transportation analysis zones or census blocks of the Rhode Island coastal regions. Analysis will then proceed similarly to Task 5, above.
Task 6 - Transportation Supply Analysis
An assessment will be made of the multi-modal means of satisfying the derived demand not met by private means. Conceivably a percentage of the population may opt to shift to public rather than private transportation. Survey results shall illustrate as well the prevalence or lack of such “doubtful” shift. Combined models developed shall illustrate this shift if confirmed. This task will thus be synchronized with Task 5.
Task 7 – Report Writing
A summary of study methods and findings will be submitted for review to URITC and end-use advisors. Comments will be incorporated into a final report.
The below chart provides the scheduling of the proposed research tasks. It is anticipated that the proposed research effort will last for 1 year. Task 1 will last 4 months, from January to February. Task 2 and 3 will last 8 months. Task 4 and 7 will last each 4 months, whereas tasks 5 and 6 will last 6. A single month overlap is typically anticipated toward the completion of a task and the initiation of the next. However, Tasks 2 and 3 overlap over 6 months given their relative independence.
Task 1 Jan. Feb.Mar. Apr.
Task 2 Mar. Apr.May Jun.Jul. Aug.Sep. Oct.
Task 3 May Jun.Jul. Aug.Sep. Oct.Nov. Dec.
Task 4 Jan. Feb.Mar. Apr.
Task 5 Mar. Apr.May Jun.Jul. Aug.
Task 6 May Jun.Jul. Aug.Sep. Oct.
Task 7 Sep. Oct.Nov. Dec.
$457,262.00
Two graduate research assistants, utilized full-time during the academic year and half-time in the summer will contribute to the research efforts described in this paper. Alolade Campbell, a graduate student in the department of Civil and Environmental Engineering, will conduct in part the research analysis. A graduate student from the Sociology and anthropology department will supplement her work. Further, the investigators, including Dr. Thomas, will present internal, on-campus, seminars of study methods and findings. Students from diverse backgrounds, including social science and engineering, will benefit from seminar attendance. It is anticipated that much of the material utilized during research conduct will be employed to the derivation of new course material in traffic engineering safety.
This proposed research builds on efforts expanded by the principal investigator and her students using seed money provided last summer by URITC. Prior efforts reviewed the literature on engineering risk assessment, the very them addressed by Task 3. Further, by all indications, valuable federal and state research funds may be freed, in the near future, for conducting hurricane evacuation studies at the local, state or regional level. The proposed research would provide the investigators with the necessary background in hurricane evacuation to undertake writing efforts of proposals seeking to attract these funds. It will further help the investigators cultivate a working rapport with the Rhode Island Department of Transportation and the Federal Emergency Management.
To address the outreach goals of this grant, the faculty at URI will promote the participation of underrepresented groups in project conduct and will further derive training and educational material from project results, presentations also constitute anticipated means of information dissemination:
1) Results will be made part of graduate and undergraduate courses in collective behavior, risk analysis, and emergency planning at URI and hopefully at other institutions of higher learning in the US and abroad, as well as become part of programs and training at FEMA.
2) Oral presentations will be made to FEMA and to leadership in the state and local emergency management community. They will also be made available to other members of the disaster research and management community and to officials in other emergency management organizations.
3) Findings will be published in professional journals and presented at professional meetings and conferences.
4) A brochure describing the findings of the study and its policy implications will be distributed to individuals and organizations from the International Research Committee on Disasters, subscribers of journals in the disaster specialty such as the International Journal of Mass Emergencies and Disasters, the Natural Hazards Review, and the Natural Hazards Observer, and agency personnel, academicians, and other attendees at the Annual Hazards Research and Application Workshop in Boulder, Colorado.
5) The project will also support the training of two master students; either one or both will be Black and/or Latino.
The project will derive a method for assessing the risk posed to the coastal regions of Rhode Island by hurricane threat. Risk will be dependent upon population resilience as relayed by the availability of resources including public and personal evacuation plans. The direct application of the evacuation plans derived will result in mitigation based on the minimization of overall societal risk under hurricane threat for the coastal regions of Rhode Island. The plans are also expected to be self-enforcing given their alignment with the emergence of the evacuation behavior. As population density in the coastal regions increase while the roadway network remains unchanged efficient methods of hurricane evacuation minimize significant losses of lives. Direct beneficiaries will be RIDOT and FEMA whose representative, Ed Parker and Robert Warren, directly stated their interest in study results.
Hurricane evacuation behavior, plan, operation and participation rates; phased evacuation; risk from hurricane threat.