Spatiotemporal Analysis of Highway Traffic Patterns in Hurricane Evacuation: A Case Study of Hurricane Irma in Florida
Mahyar Ghorbanzadeh (firstname.lastname@example.org), Florida A&M University-Florida State UniversityShow Abstract
Simone Burns, Florida A&M University
Linoj Vijayan Nair Rugminiamma, Florida A&M University-Florida State University
Eren Ozguven, Florida A&M University-Florida State University
Wenrui Huang, Florida A&M University-Florida State University
Hurricanes devastate cities and entire regions in the U.S. every year, causing widespread major infrastructure damages and claiming lives. Florida is one of the U.S. states which is significantly vulnerable to these hurricanes where planning for evacuations becomes extremely critical. In 2017, Hurricane Irma has proven a unique challenge for Floridians trying to escape. Such storms usually hit the state from the east or west, allowing residents to flee north or south to avoid the damage. However, with Irma to make an unusual landfall from the south, it enveloped the entire Florida peninsula forcing evacuees to drive farther north, creating an evacuation route longer than any recent storm. As such, this study aims to assess the spatiotemporal traffic impact of Irma on Florida’s major highways based on the real-time traffic data before, during, and after the hurricane made landfall. First, a time series-based analysis was conducted to evaluate the temporal evacuation patterns of this large-scale evacuation. Second, a metric, namely the congestion index (CI), was developed to assess the spatiotemporal traffic patterns with an application on the Hurricane Irma evacuation using data on I-95, I-75, I-10, I-4, and turnpike (SR-91) highways in an interval of six hours starting on Tuesday, September 5, 2017 through Friday, September 15, 2017. Results show the impact of the uncertainty for Irma’s predicted paths on the evacuation patterns due to the imperfect forecasts, indicating the changing levels of congestion on major highways as a result of the provided information on the hurricane’s path.
Understanding the Willingness to Share Resources in the Hurricane Irma Evacuation: A Multi-Modeling Approach
Stephen Wong (email@example.com), University of CaliforniaShow Abstract
Mengqiao Yu, University of California, Berkeley
Anu Kuncheria, University of California, Berkeley
Susan Shaheen, University of California, Berkeley
Joan Walker, University of California, Berkeley
Recent technological improvements have greatly expanded the sharing economy (e.g., Airbnb, Lyft, and Uber), coinciding with growing need for transportation and sheltering resources in evacuations. To understand influencers on sharing willingness in evacuations, we employed a multi-modeling approach across four sharing scenarios using three model types: 1) four binary logit models that capture each scenario separately; 2) a multi-choice latent class choice model (LCCM) that jointly estimates multiple scenarios via latent classes; and 3) a portfolio choice model (PCM) that estimates dimensional dependency. We tested our approach by employing online survey data from 2017 Hurricane Irma evacuees (n=368). The multi-model approach uncovered behavioral nuances undetectable with a single model. First, the multi-choice LCCM and PCM models uncovered scenario correlation, specifically willingness to share for both transportation scenarios and both sheltering scenarios. Second, the multi-choice LCCM found three classes – transportation sharers, adverse sharers, and interested sharers. Transportation sharers were more likely to be female, lower-income, and residents of Southwest Florida compared to adverse sharers. Interested sharers were more likely to be male, long-time residents, and higher-income compared to adverse sharers. Third, families with children were unwilling to share regardless of the model, while spare capacity (i.e., seatbelts, spare beds) had a positive but somewhat insignificant influence on sharing. Fourth, experienced home sharers were more willing to share shelter in the binary logit and PCM models. We suggest that local agencies consider holistic sharing mechanisms across resource types and time (i.e., before, during, and after a hurricane evacuation).
A dynamic traffic assignment model in emergency evacuation considering background traffic
Tao Zhang (firstname.lastname@example.org), Taiyuan University of Science and TechnologyShow Abstract
Gang Ren, Southeast University
Gang Cheng, Tibet University
Yang Yang, Taiyuan University of Science and Technology
Minjie Jin, Taiyuan University of Science and Technology
This paper builds a dynamic traffic assignment model in the emergency evacuation considering the background traffic. First, the equilibrium functions of entry time and path choice are both described based on the Logit model, whose constrained conditions includes the state equation, propagation function, and conservation constraint; the proposed equilibrium functions are formulated as a variational inequality problem, which is solved by the dynamic traffic network demand loading iterative method. Then, based on the case study of Wenling city, the proposed model and solution method are verified according to the iterative evolution and path assignment sections, and their characteristics are researched by sensitivity analyses of four relevant parameter, i.e., length of time interval, total number of evacuation vehicles, adjusting parameters related to entry time, and path choice. The results can provide the basic traffic assignment means for evacuation planners to propose the reasonable plan and organization.
Development of an Evacuation Decision Support Tool: A Combined Optimization and Traffic Microsimulation Modeling Approach
MD Jahedul Alam, Dalhousie UniversityShow Abstract
Muhammad Habib, Dalhousie University
This study develops a novel framework to formalize how to optimally use all available modes of transportation, particularly, auto, transit and school bus for mass evacuation. The study develops an “All Mode Evacuation Decision Support Tool (AMEDST)” to estimate an optimum composition of auto-bus mix that demonstrates an improvement in evacuation times and network congestion. The tool integrates two components: an “All Mode Allocation Module (AMAM)” and a traffic evacuation microsimulation model. The first module prioritizes evacuees for bus allocation accounting for vulnerabilities of populations. The study follows a Knapsack problem formulation to develop AMAM. A Dynamic Programming algorithm is used to solve the Knapsack optimization problem within a Python platform. Individuals not prioritized are assumed to be evacuated by personal vehicles. Traffic microsimulation model follows a dynamic traffic assignment process to simulate evacuation scenarios using all available modes. It informs AMAM the maximum number of individuals that can be evacuated by buses. The results suggest that individuals with higher level of vulnerabilities are prioritized for bus allocation within AMEDST. Moreover, time to allocate bus within an evacuation period is sporadic for a high bus demand. Results from the traffic simulation suggest that traffic congestion and evacuation times significantly decrease as reflected in vehicular traffic reduction of 3.9-7.7% and evacuation time reduction of 9-22.7% if 5-20% auto-based demand can be served by buses. The tool will help emergency personnel evaluate alternative scenarios regarding resource allocation and evacuation planning during extreme events that necessitate large-scale evacuations.
The Isolated Community Evacuation Problem with Two-Stage Stochastic Mixed Integer Programming
Klaas Fiete Krutein (email@example.com), University of WashingtonShow Abstract
Anne Goodchild, University of Washington
With the risk of natural disasters affecting isolated communities, the demand for efficient evacuation plans is increasing. However, isolated areas such as islands, often have characteristics that make conventional methods, such as evacuation by private vehicle impractical to infeasible. Further, most previous research on emergency evacuations focuses on densely populated areas with a rich infrastructure, which makes these models not applicable to the characteristics isolated communities. Instead, when road connections are disrupted, isolated communities are dependent on recovery resources, such as vessels, to evacuate the population. In this paper, we introduce the Isolated Community Evacuation Problem (ICEP) and a corresponding mixed integer programming formulation that aims to minimize the evacuation time of an isolated community through optimally routing a coordinated fleet of recovery resources. To enable planning, we expand the formulation to a two-stage stochastic problem that allows scenario-based resource planning, while ensuring minimal evacuation time. We further provide objective functions with a varying degree of risk and present the sensitivity of the model to different objective functions and problem sizes through numerical experiments. The result gives emergency planners of remote areas a tool to evaluate their community's level of preparedness for evacuation.
Willingness to Share During Multi-Hazard Events: An Exploratory Study of Flood Evacuations During the COVID-19 Pandemic in the United States
Elisa Borowski (firstname.lastname@example.org), Northwestern UniversityShow Abstract
Victor Limontitla Cedillo, Northwestern University
Amanda Stathopoulos, Northwestern University
Volunteered sharing of resources is commonly observed in response to disaster events, but during a multi-hazard scenario that requires physical distancing, such as the COVID-19 pandemic, the factors influencing willingness to share resources face-to-face remain unknown. Given the ability of on-demand ridesourcing to provide emergency transportation to individuals without access to alternatives, this study examines the personal and situational factors that contribute to willingness to share a flood evacuation ride with a stranger during the COVID-19 pandemic. We hypothesize that willingness to share will be significantly correlated with traditional evacuation decision factors, traditional sharing factors, and current COVID-19 risk factors. To test these hypotheses, we distribute a survey to 586 individuals in three Midwestern and three Southern U.S. states with high risk of flooding. To determine willingness to share a ride as a driver and passenger, we estimate two binary logistic regression models. Our findings show that the evacuation variables correlated with willingness to share for passengers are destination type and belongings. Significant sharing variables include age, employment, type of residential area, and ridesourcing experience for drivers and having children and living with others without paying rent for passengers. Significant COVID-19 related variables include race for both drivers and passengers, as well as perceived vulnerability and political affiliation for passengers. These findings suggest the decision to share a flood evacuation ride with an unknown neighbor during the COVID-19 pandemic is impacted by different factors for drivers and passengers, although personal and situational factors influence both. Ridesourcing provider implications are discussed.
Assessing the Crash Risks of Evacuation: A Matched Case-Control Approach Applied over Data Collected during Hurricane Irma
Rezaur Rahman, University of Central FloridaShow Abstract
Tanmoy Bhowmik, University of Central Florida
Naveen Eluru, University of Central Florida
Samiul Hasan (email@example.com), University of Central Florida
Recent hurricane experiences have created major concerns for transportation agencies and policymakers to find better evacuation strategies, especially after hurricane Irma—which forced about 6.5 million Floridians to evacuate and caused a significant amount of delay due to heavy congestion. However, a major concern for issuing an evacuation order is that it involves a high number of crashes in highways. In this study, we present a matched case-control based approach to understand the factors contributing to the increase in the number of crashes during evacuation. We use traffic data for a period of 5 to 10 minute just before the crash occurred. For each crash observation, traffic data are collected from two upstream and two downstream detectors of the crash location. We estimate models for three different conditions: regular period, evacuation period, and combining both evacuation and regular period data. Model results show that, if there is high volume traffic at an upstream station and high variation of speed at a downstream station, the probability of crash occurrence increases. Using a panel mixed binary logit model, we estimate the effect of evacuation itself on crash risk and find that, after controlling for traffic characteristics, during evacuation the chance of an accident is higher than in a regular period. Our findings have implications for evacuation declarations and highlight the need for better traffic management strategies during evacuation. Future studies should develop advanced real-time crash prediction models which will work for evacuation traffic conditions and design proactive countermeasures to reduce crash occurrences during evacuation.
Constructing Evacuation Evolution Patterns and Decisions Using Mobile Device Location Data: A Case Study of Hurricane Irma
Aref Darzi (firstname.lastname@example.org), University of Maryland, College ParkShow Abstract
Vanessa Frias-Martinez, University of Maryland, College Park
Sepehr Ghader, University of Maryland, College Park
Hannah Younes, University of Maryland, College Park
Lei Zhang, University of Maryland, College Park
Understanding individuals’ behavior during hurricane evacuation is of paramount importance for local, state, and government agencies hoping to be prepared for natural disasters. Complexities involved with human decision-making procedures and lack of data for such disasters are the main reasons that make hurricane evacuation studies challenging. In this paper, we utilized a large mobile phone Location-Based Services (LBS) data to construct the evacuation pattern during the landfall of Hurricane Irma. By employing our proposed framework on more than 11 billion mobile phone location sightings, we were able to capture the evacuation decision of 807,623 smartphone users who were living within the state of Florida. We studied users’ evacuation decisions, departure and reentry date distribution, and destination choice. In addition to these decisions, we empirically examined the influence of evacuation order and low-lying residential areas on individuals’ evacuation decisions. Our analysis revealed that 57.92% of people living in mandatory evacuation zones evacuated their residences while this ratio was 32.98% and 33.68% for people living in areas with no evacuation order and voluntary evacuation order, respectively. Moreover, our analysis revealed the importance of the individuals’ mobility metrics in modeling the evacuation decision choice. Historical mobility behavior information such as number of trips taken by each individual and the spatial area covered by individuals’ location trajectory estimated significant in our choice model and improve the overall accuracy of the model significantly.
Strategic Evacuation for Regional Events: With and Without Autonomous Vehicles
Jooyong Lee, University of Texas, AustinShow Abstract
Kara M. Kockelman (email@example.com), University of Texas, Austin
An evacuation scheduling algorithm is developed for optimal planning of large-scale, complex settings to minimize total delay plus time in transit across residents. The algorithm is applied to the 8-county Houston-Galveston region and land use setting under the 2017 Hurricane Harvey scenario, with multiple shelters as destinations, far from the Gulf of Mexico. Autonomous vehicle (AV) use under central guidance is also tested, to demonstrate the evacuation time benefits of AVs. Having a higher share of AVs delivers more efficient evacuation performance, due to greater reliability on routes selected, lower headways, and higher road capacity. 100% AV use delivers lower overall evacuation costs and network clearance times (from 89 hr. to 68 hr. network clearance time, assuming 1.88 vehicles per household) and lower uncertainty in travel times (from reduced standard deviation of 12 hr. to 9 hr.). Other scenarios were also tested. For example, a 3% to 5% compressed network clearance time added 10% to 25% longer travel times and network congestion. A 6% longer network clearance time reduced residents’ total travel time and network congestion by 10%, but increased the evacuation cost, demonstrating the benefits of scheduling (and enforcing) evacuations across residents and neighborhoods more thoughtfully. Optimal combination of departure times by neighborhood and household helps balance these conflicting objectives.
Developing Transportation Response Strategies for Wildfire Evacuations via an Empirically Supported Traffic Simulation of Berkeley, California
Bingyu Zhao, University of California, BerkeleyShow Abstract
Stephen Wong (firstname.lastname@example.org), University of California
Government agencies must make rapid and informed decisions in wildfires to safely evacuate people. However, current evacuation simulation tools that could help resource-strapped agencies largely fail to compare possible transportation responses or incorporate empirical evidence from past wildfires. Consequently, we employ online survey data from evacuees of the 2017 Northern California Wildfires (n=37), the 2017 Southern California Wildfires (n=175), and the 2018 Carr Wildfire (n=254) to inform a policy-oriented traffic evacuation simulation model. We test our simulation for a hypothetical wildfire evacuation in the wildland urban interface (WUI) of Berkeley, California. We focus on variables derived from empirical surveys and past studies, including fire speed, departure time distribution, towing of items, transportation mode, GPS-enabled rerouting, phased evacuations (i.e., allowing higher-risk residents to leave earlier), and contraflow (i.e., switching all lanes away from danger). We found that reducing household vehicles (i.e., to 1 vehicle per household) and increasing GPS-enabled rerouting (e.g., 50% participation) lowered exposed vehicles (i.e., total vehicles in the fire frontier) by over 50% and evacuation time estimates (ETEs) by about 30% from baseline. Phased evacuations with a suitable time interval reduced exposed vehicles most significantly (over 90%), but produced a slightly longer ETE. Both contraflow and slowing fire speed were effective in lowering exposed vehicles (around 50%), but not ETEs. We recommend agencies develop a communication and parking plan to reduce evacuating vehicles, create and communicate a phased evacuation plan, and build partnerships with GPS-routing services.
One-way Coupling of Fire and Egress Modelling for Realistic Evaluation of Evacuation Process
He-in Cheong (email@example.com), Imperial College LondonShow Abstract
Zhiyu Wu, Imperial College London
Arnab Majumdar, Imperial College London
In the discipline of fire engineering, computational simulation tools are used to evaluate the Available Safe Egress Time (ASET) and Required Safe Egress Time (RSET) of a building. The evaluations are often carried out separately on separate tools – ASET using a computational fluid dynamics (CFD) fire simulation package and RSET using a crowd dynamics modelling tool. However, there are advantages to coupling the ASET and RSET analysis to quantify tenability conditions and reevaluate evacuation time within a building. The disadvantages of such coupling are that the results are specific to the modelled scenario and that the coupling process is computationally complex and time-consuming. This paper presents the successful one-way coupling of CFD and crowd dynamics modelling via the tenability condition analysis of fractional effective doses (FED) and its effect on the evacuees. The simulation tool used for fire analysis was Fire Dynamics Simulator (FDS), and the software package used for the crowd dynamics modelling was Oasys MassMotion. The coupling is carried out with the help of the software development kit (SDK) of the Oasys MassMotion in two different example geometries: 1) an open-plan room and 2) a floor with six rooms and a corridor. The paper shows that the application of the new method of coupling ASET and RSET analysis has been successful and that the evacuation time is calculated to be longer when considering the effects of the smoke on the evacuees and the crowd densities show different profiles.
Methodology to Quantify Statewide Evacuations
Scott Parr (firstname.lastname@example.org), Embry Riddle Aeronautical UniversityShow Abstract
Lorraine Acevedo, Embry Riddle Aeronautical University
Pamela Murray-Tuite, Clemson University
Brian Wolshon, Louisiana State University
Statewide mass evacuations are among the largest single-event traffic movements. They can last several days, cover thousands of roadway miles, and include hundreds of thousands of people and vehicles. Often, they are also marked by enormous delay and congestion and are nearly always criticized for their inefficiency and lack of management. Despite the critical importance and the potential to impact lives and safety, there are no recognized methods to systematically quantify traffic characteristics at statewide scales. This paper describes research to develop and apply an analytical method to measure and describe statewide mass-evacuations in a practical, cost-effective manner. The research methods are based on simple, yet widely available, and easily understood traffic count datasets that support both qualitative and quantitative analyses. By spatially and temporally arranging sensor-based statewide traffic volume data from Hurricane Irma (2017) and Michael (2018) evacuations, these methods are applied describe and answer several key questions about statewide mass evacuations. The methods developed in this research are able to estimate the start and end of the auto-based evacuation, the loading and peaking characteristics of traffic, the total number of vehicles involved in the evacuation, and the effective start and end time of the auto-based reentry. Among the key findings of this work were that the Hurricane Irma and Michael evacuations began several days before landfall, peaking two to three days prior to the storm. It is expected that state departments of transportation and emergency management officials can apply similar methods to assess and better plan future evacuations.
Modelling the Roadway Temperature during Wildfire for Evacuation and Assessment of Pavement Damage: A Case Study of Camp Fire
Mohammadreza Barzegar, Washington State UniversityShow Abstract
Haifang Wen (email@example.com), Washington State University
Wildfires have caused enormous loss of assets, resources, and human lives. During wildlife, timely evacuation is a critical element to save lives. Therefore, it is important to understand the conditions of evacuation routes, such as traffic, temperature, etc., either of which, if adverse, can cause loss of lives. The temperature of the evacuation roadway has not been studied in the past. In addition, the damage to the roadway as a result of high wildfire temperature warrants a quantification of temperature of the roadway to estimate the asset loss as a result of wildfire. This study aimed to model the temperature conditions of roadway within a wildfire zone, using the 2018 Camp Fire as a case study. Numerical modelling program, FlamMap, was used to analyze the temperature conditions of an evacuation route. Topography, canopy, fuel, ignition, and weather conditions specific to the Camp Fire site were retrieved for simulation. It was found that the temperature of roadway can increase quickly in a matter of a few hours. The spatial and temporal distribution of temperature along the roadway was also determined. The roadway temperatures ranges from 74℃ to 363℃ just 5 hours after the ignition, depending on the specific location and other conditions. In addition, temperature decreases quickly away from the roadway. The results from this study can be used to understand the temperature conditions for evacuation and estimate of roadway damage. Similar site-specific analysis can be used to prepare for evacuation and develop an evacuation plan for a specific site.
DISCLAIMER: All information shared in the TRB Annual Meeting Online Program is subject to change without notice. Changes, if necessary, will be updated in the Online Program and this page is the final authority on schedule information.