2021

Srijith Balakrishnan (srijith@utexas.edu), University of Texas, Austin
Taehoon Lim, Texas Department of Transportation
Zhanmin Zhang, University of Texas, Austin

Among natural disasters, hurricanes pose the most significant threat to port infrastructure in the United States, especially to those along the coasts of the Atlantic Ocean and the Gulf of Mexico. The operational continuity of ports is critical because of the growing reliance of domestic and international businesses on ports and their significance in the growth of national and regional economies. While direct physical damages to ports are not rare, a shutdown of port operations is more likely declared by authorities as a precautionary and preparatory measure during hurricanes. Considering the enormous economic importance of ports and their vulnerability to hurricanes, the current study proposes a framework and methodology to analyze the economic risks of such hurricane-related shutdowns using hurricane- and port-related determinants. The risks of shutdowns are modeled using regression analysis based on historical port shutdown data and are combined with extensions of the well-known input-output model to predict the operational and economic risks of ports to hurricanes. The application of the methodology is demonstrated by conducting a case study based on the Texas Port System to evaluate the economic risks of hurricane-related port disruptions on the U.S. economy.

Jia-Dzwan Shen, Federal Highway Administration (FHWA)
Sheila Rimal Duwadi, Federal Highway Administration (FHWA)
Jeffrey Ger, Federal Highway Administration (FHWA)
Derek Soden, Federal Highway Administration (FHWA)
Waider Wong, Federal Highway Administration (FHWA)
Vincent Chiarito, Federal Highway Administration (FHWA)

This paper focuses on highway bridges and structures including tunnels, their importance within the transportation network, and efforts underway to address resiliency. Keeping the system functional throughout an extreme event cycle and beyond is also discussed. The Federal Highway Administration (FHWA) established several programs to increase the knowledge base for improving transportation system resilience, and to assist states. Disasters and disruptions continue to evolve in complexity, functional impact, economic cost, and in diversity of locations and types. Just focusing on weather, the National Oceanic and Atmospheric Administration’s (NOAA) ‘Billion-Dollar Weather and Climate Disasters’ study reports the U.S. sustained 250 weather and climate disasters from 1980 to 2019. These weather impacts caused overall damages/costs reaching or exceeding $1 billion (Consume Price Index, CPI, adjusted to 2019) and with a total cost exceeding $1.7 trillion. (https://www.ncdc.noaa.gov/billions/overview). Government and private entities at all levels are responding in various ways to effectively and efficiently respond to disasters and their effects. Within transportation, the Federal role is to coordinate with State Transportation Departments, Metropolitan Planning Organizations (MPOs) and with organizations representing the transportation field, such as the Transportation Research Board (TRB), and others. For conciseness, most of the discussions in this paper reference bridges. However, it is generally implied that the same concept applies to other highway structures unless specifically discussed separately.

Katherine Idziorek (kidzi@uw.edu), University of Washington
Cynthia Chen, University of Washington
Daniel Abramson, University of Washington

Current work in the area of resource sharing for disaster response and recovery assumes a top-down, centralized perspective. This study addresses a gap in knowledge about how resources might be shared among community members when a centralized supply of resources is not available, as might occur in a large-scale event such as a Cascadia Subduction Zone earthquake. In the case of such a disaster, community members’ willingness to share resources with one another could contribute to the relative success or failure of communities to be locally self-sufficient if required. This research draws upon data gathered from a community-scale sample survey set in the Pacific Northwest, a region in which earthquakes are a certain, though largely unpredictable, hazard. In order to better understand the potential for resource sharing among community members in the event of an earthquake, we analyze three attitudinal variables related to both actual disaster preparedness and anticipated willingness to share: level of concern about disasters, place attachment, and trust. Our findings reveal a negative association between level of concern and actual disaster preparedness, while willingness to share is most strongly influenced by trust. Additional observed relationships between trust, place attachment, and community social network size suggest a need for further research in this area. Better understanding willingness to share and available resources at the community level can help to inform both grassroots efforts and more formal disaster preparedness organizations regarding targeted interventions for improving disaster preparedness.

Wen-Long Shang (wenlong.shang12@alumni.imperial.ac.uk), Beijing University of Technology
Yanyan Chen, Beijing University of Technology
Washington Ochieng, Imperial College London

To date immunity to disruptions of multi-scale urban road networks (URNs) has received increasing attention. This study utilizes robustness as a representation of immunity of URNs. We propose a benchmark index, Relative Area Index (RAI), based on traffic assignment theory to quantitatively measure the robustness of URNs under global capacity degradation due to three different types of disruptions. We also compare the novel RAI with weighted betweenness centrality, a traditional topological metric of robustness. We employ six realistic URNs as case studies for this comparison. Our analysis shows that RAI is a more effective measure of the robustness of URNs when multi-scale URNs suffer from global disruptions. This improved effectiveness is achieved because of RAI's ability to capture the effects of realistic network characteristics such as network topology, flow patterns, link capacity, and travel demand. Also, the results highlight the importance of central management when URNs suffer from disruptions. Our novel method may provide a benchmark tool for comparing robustness of multi-scale URNs, which facilitates the understanding and improvement of network robustness for the planning and management of URNs.

Theodoros Pantelidis, New York University
Joseph Chow, New York University
Oded Cats, Delft University of Technology

Public transportation delays due to systematic failures have a major impact on network users. We propose designing capacity pooling contracts to facilitate horizontal cooperation among operators to mitigate those costs and improve service resilience. When two or more public transport providers agree upon sharing resources, the total transportation costs can be reduced due to added flexibility in the system. These operators may contribute capacity to be used in cost-effective routes owned by other operators. We formulate a two-stage stochastic model to determine the cost savings under different collaboration scenarios. We provide several solution methods: a deterministic equivalent problem, the L-shaped method, and sample average approximation. Coalitional stability under Shapley value, nucleolus, and -value are tested. The proposed model is applied to a regional multimodal network in the Randstad area of the Netherlands, for four operators, 80 origin-destination pairs, and over 1400 links where disruption data is available. Using the proposed method, we identify stable cost allocations among four operating agencies that could yield a 44% improvement in overall network performance over not having any risk pooling contract in place.  Keywords: Two stage-stochastic programming, network disruptions, cost allocation mechanisms, horizontal collaboration, multicommodity flow problem

Simone Burns (simone1.burns@famu.edu), Florida A&M University
Linoj Vijayan Nair Rugminiamma, Florida A&M University-Florida State University College of Engineering
Mahyar Ghorbanzadeh, Florida A&M University-Florida State University College of Engineering
Eren Ozguven, Florida A&M University-Florida State University College of Engineering
Wenrui Huang, Florida A&M University-Florida State University College of Engineering

Hurricane Michael was the first Category 5 hurricane to ever hit the Florida Panhandle, leaving thousands of residents without housing for weeks if not months. Evacuations were especially challenging when vulnerable communities were considered since they needed more time and assistance before the hurricane hit. This takes on additional complexity when older populations are considered because any extra time they incur in reaching safety can be especially confounding in light of their potential health problems. As such, this paper proposes a Geographical Information Systems-based methodology to assess the vulnerability of two Florida counties, namely Bay and Gulf, substantially impacted by the Hurricane Michael. The utmost importance is given to assess the efficiency of evacuations with a focus on both social (i.e., age, vehicle ownership) and physical (i.e., elevation, storm surge zones) factors that impact vulnerability of communities such as the elderly and areas facing inundation risk. This is also supported by real-world evacuation data that provides the changing evacuation status in the area. The assessment conducted in this study has the potential to provide guidance and recommendations to officials towards better planning for hurricane evacuations throughout the Florida Panhandle.

Jingran Sun (jingransun@utexas.edu), University of Texas, Austin
Zhe Han, University of Texas, Austin
Zhanmin Zhang, University of Texas, Austin

Traditional resource allocation strategy mainly focuses on the physical condition and natural deterioration trend of the individual infrastructure. With more frequent and sever extreme events, it is essential to incorporate the resilience of the infrastructure network into the decision-making process. The interdependent nature of infrastructure systems could amplify both the impact of extreme events and the effect of the resource allocation strategy on infrastructure resilience. As a result, infrastructure interdependencies should be considered when analyzing the impact of extreme events and resource allocation strategies, complicating the quantification of resilience and decision-making process. For life-cycle resource allocation, other challenges further increase the complexity of the problem, such as uncertainties associate with the occurrence of extreme events, the effect of maintenance treatment effects, and the balance between improving infrastructure resilience and its physical condition. This study proposes a methodological framework to optimally allocate resources throughout the life cycle of  infrastructure facilities under management with the goal of maximizing the resilience of infrastructure network during that time horizon.

yao xiao (xiaoyao9@mail.sysu.edu.cn), Sun Yat-Sen University
Mofeng Yang
Zheng Zhu, Hong Kong University of Science and Technology
Hai Yang, Hong Kong University of Science and Technology
Lei Zhang, University of Maryland, College Park
Sepehr Ghader, University of Maryland, College Park

Mathematical modeling of epidemic spreading has been widely adopted to estimate the threats of epidemic diseases (i.e., the COVID-19 pandemic) as well as to evaluate epidemic control interventions. The indoor place is considered to be a significant epidemic spreading risk origin, but existing widely-used epidemic spreading models are usually limited for indoor places since the dynamic physical distance changes between people are ignored, and the empirical features of the essential and non-essential travel are not differentiated. In this paper, we introduce a pedestrian-based epidemic spreading model that is capable of modeling indoor transmission risks of diseases during people’s social activities. Taking advantage of the before-and-after mobility data from the University of Maryland COVID-19 Impact Analysis Platform, it’s found that people tend to spend more time in grocery stores once their travel frequencies are restricted to a low level. In other words, an increase in dwell time could balance the decrease in travel frequencies and satisfy people’s demand. Based on the pedestrian-based model and the empirical evidence, combined non-pharmaceutical interventions from different operational levels are evaluated. Numerical simulations show that restrictions on people’s travel frequency and open-hours of indoor places may not be universally effective in reducing average infection risks for each pedestrian who visit the place. Entry limitations can be a widely effective alternative, whereas the decision-maker needs to balance the decrease in risky contacts and the increase in queue length outside the place that may impede people from fulfilling their travel needs.

Yuqing Ji, Tongji University
Dongxiu Ou, Tongji University
Rei ZhG, Tongji University
Visarut Phichitthanaset, Tongji University
Chenkai Tang, Tongji University

When a railway emergency occurs, especially for EMUs (Electric Multiple Units) of higher speed, it often leads to unexpected consequences. Therefore, the railway emergency plan, a pre-established response plan to deal with emergencies, plays an important role in reducing injuries and losses. However, most of the existing railway emergency plans remain as plain-text documents, requiring lots of manual work to capture the important regulations. Many researcher have proposed many methods for emergency modeling and analysis. However, for the research on railway emergency plan modeling, we are still in great lack. Therefore, to form a visualized, formal and digital railway emergency plan, this paper proposed a HTPN (Hierarchical Timed Petri Net)-based modeling method. The framework model was proposed first for all kinds of railway emergency plans. Then the instantiated HTPN model of railway emergency plan of EMUs rescue was built by ExSpect, a Petri net simulation tool. The instantiated HTPN model is of better readability, execution, and visualization. The experiments show that the HTPN model can conform to the practice well. Furthermore, according to the simulation results, this paper put forward some suggestions for the optimization of the EMUs rescue emergency plan.

Ho Yin Chan (13064178d@connect.polyu.hk), Hong Kong Polytechnic University
Wei Ma, Hong Kong Polytechnic University
Nang-Ngai Sze, Hong Kong Polytechnic University
Xintao Liu, Hong Kong Polytechnic University
Anthony Chen, Hong Kong Polytechnic University

Hong Kong was relatively successful in mitigating transmission early in the outbreak of coronavirus disease (COVID-19), with neither complete lockdown nor intensive mobility restrictions. A case study was conducted in Hong Kong to assess the effectiveness in containing the COVID-19 epidemic in the early stage. This paper aims to model the relationship between the stringency of government responses, mobility patterns, and spread of COVID-19 cases. Multiple data sources including epidemiological record, novel performance indictor, mobility, and traffic data were used to examine two key questions: (1) How various government policies and public preparedness influence urban mobility patterns, considering the spread of COVID-19 at different scales? (2) different data sources provide information about the spatio-temporal mobility patterns and real-time conditions of COVID-19, and how they correlate with each other? Results indicate the importance of both government policy interventions and community engagement in promoting public preparedness and response against epidemic crisis.

Pengzi Chu (cpz_myhk@163.com), Tongji University
Yi Yu, Tongji University
Haizhu Hong, Shanghai Shentong Metro Group Co., Ltd.
Jianjun Yuan, Tongji University
Huahua Zhao, Tongji University

As a significant way for citizens to travel at metropolises, metro has the advantages of convenience, efficiency, and environmental friendliness. Considering the interaction of lines in a metro network, this paper discussed a method for constructing a metro interaction network (MIN), proposed an analytical framework on its robustness. The Shanghai metro interaction networks (SMINs) including the network currently in operation (SMINop) and its extension (SMINext) were used for analysis and comparison. The results suggest that compared with SMINop, the efficiency of SMINext reduces by 6.297%, but removing a vertex has less impact on its efficiency ( p < 0.01). SMINext performs better than SMINop in attack-based tests, with larger efficiency retention rates in degree-based attack and betweenness-based attack. For failure propagation, when the recovery probability is small, SMINext performs worse than SMINop, with larger propagation scopes and smaller unaffected efficiency rates. But when the recovery probability is large, the performance of SMINext is better. The workable efficiency rates of SMINext are always greater than SMINop under different recovery probabilities. A larger recovery probability is one of the prerequisites for the robustness of Shanghai metro networks. Maintaining a high emergency handling ability is vital to the network operation of metros.

Karl Kim, University of Hawaii
Eric Yamashita, University of Hawaii
Jiwnath Ghimire, University of Hawaii

Using data from a national survey conducted in the United States during the Spring of 2020, the differences between emergency managers, transportation planners, and others involved in pandemic disaster response in terms of risk perception and protective actions are investigated.  The study found that 92 percent of respondents reported implementing voluntary actions with 35 percent reporting quarantine and 37 percent reporting isolation actions. The attributes of respondents and the agencies and communities they work in are categorized in terms of personal, disciplinary, or professional backgrounds, as attributes such as urban versus rural, coastal versus non-coastal, and other factors.  Three dependent variables are modeled including  1) risk tolerance; 2) level of preparedness (including support for training), and 3) implementation of protective measures for social distancing, quarantine, and isolation to ascertain the influences of personal, professional, and regional, locational characteristics. A risk tolerance score is implemented by asking respondents “what percentage of the population would need to be sick to implement voluntary and non-voluntary actions. Using Poisson regression analysis and correspondence analysis, the patterns, associations, and clustering of backgrounds and other attributes are modeled to show the relationships between risk perceptions, preparedness, and professional backgrounds. In addition to identifying which places and people are more inclined to support protective actions for the pandemic, this analysis also helps to demonstrate the intersections and mutual interests across public health, transportation, and emergency management.  Overall, this study found a low level of preparedness for the pandemic with 70 percent of the respondents supporting additional training.

Alexander Kolpakov, University of South Florida
Austin Sipiora, University of South Florida
Caley Johnson, National Renewable Energy Laboratory (NREL)
Erin Nobler, National Renewable Energy Laboratory (NREL)

While the Tampa Bay region has not been directly hit by a major hurricane since 1921, it is considered one of the most vulnerable areas in the United States to hurricanes and severe tropical weather. A particular vulnerability stems from the fact that all petroleum fuel comes to the area through Port Tampa Bay, which can be (and has been in the past) impacted by hurricanes and tropical storms. The case study discussed in this paper covers previous fuel challenges, vulnerabilities, and lessons learned by key Tampa Bay public agency fleets during the past 10 years (mainly as a result of most recent 2017 Hurricane Irma) in order to explore ways to improve the area’s resilience to natural disasters. Some of the strategies can include maintaining emergency fuel supply, prioritizing fuel use, strategically placing the assets around the region to help with recovery, investing in backup generators (including generators powered by alternative fuels), planning for redundancies in fuel supply networks, developing more efficient communication procedures between public fleets, improving hurricane planning, and upgrading street drainage systems to reduce the threat of local flooding.

Hana Takhtfiroozeh (hana.t@memphis.edu), University of Memphis
Mihalis Golias, University of Memphis
Sabya Mishra, University of Memphis

An important part of transportation network vulnerability analysis is identifying critical links where failure may lead to severe consequences, and the potential of such incidents cannot be considered negligible. The primary objective of this paper is to propose and evaluate the accuracy of nine measures that can be used to rank links based on their criticality in affecting the vulnerability of a transportation network. The proposed measures combine several characteristics of traffic equilibrium inputs and outputs with a topological-based measure (Betweenness Centrality) to balance accuracy and computational complexity. Three benchmark networks from the literature are used to evaluate the performance of the proposed measures by comparing their accuracy to existing traffic-based measures known to be the most accurate and most computationally expensive at the same time.

Seyed Hooman Ghasemi (hooman.ghasemi@auburn.edu), Washington State University
Ji Yun Lee, Washington State University

Resilience analysis has been concerning the interest of many researchers since it can estimate the remaining functionality performance of a system after a disruption. In general, the resilience measure is determined using four main properties including robustness, redundancy, rapidity, and resourcefulness. Nevertheless, there is no unified metric to measure the resilience of a system. Furthermore, resilience analysis is highly sensitive to the intensity of the disruption. The main contribution of this study is to present a new resilience metric for highway bridges system immediately after happening an earthquake event associated with the spectrum intensity of the earthquake. In doing so, a reliability-intensity spectrum theory corresponding to the various intensity levels is presented as the alternative state-of-the-art instead of the fragility curve theory to overcome the insufficiencies’ aspects of the fragility curves. In addition, a new definition of the structural robustness associated with the intensity measure of the earthquake is demonstrated. Also, the conventional definition of the redundancy is modified using the binomial probability distribution corresponding to the structural robustness. Eventually, a new resilience metric is formulated based on the intersection of the structural robustness and redundancy immediately after occurring the disruption, which is called “instantaneous-resilience”. An example is provided to illustrate the general procedure to determine the new instantaneous-resilience metric for the highway bridges network.

Johan Rose Santos, Republic of the Philippines - Department of Public Works and Highways
Varun Varghese, "Hiroshima Daigaku"
Makoto Chikaraishi (chikaraishim@hiroshima-u.ac.jp), Hiroshima University
Tatsuhiko Uchida, "Hiroshima Daigaku"

Road networks are highly vulnerable to risks stemming from both internal factors such as the topological structure of the network and external factors such as natural disasters. The disruptions caused by these potential risk factors could result in severe physical and socio-economical losses. Therefore, understanding the impact and risk associated with road networks will be beneficial in reducing losses and help in preparing better risk mitigation and management strategies. This study proposes an integrated approach to assess risk of sediment hazard on the road network by borrowing concepts from (a) transport vulnerability studies; (b) disaster risk assessment; and (c) spatial risk analysis, and applied it to an identified vulnerable road network in Kure city, Japan. The proposed risk framework holistically incorporates the processes of topological network vulnerability analysis, exposure spatial analysis, hazard occurrence probability estimation through binary logit regression, impact calculation using Monte Carlo simulation, and risk estimation. 12,000 possible multi-link disruption scenarios were simulated using the recorded sediment disaster information and rainfall event on July 2018 in Hiroshima prefecture. Spatial distribution of the risk calculations helped in identifying links with high probability of disruption and high impact i.e. high-risk links. The findings of this study may support policy decisions regarding road risk mitigation and recovery prioritization during disaster and road infrastructure investment through risk-benefit analysis.

Ouassim Manout (ouassim.manout@polymtl.ca), Ecole Polytechnique de Montreal
Idriss El-Megzari, Ecole Polytechnique de Montreal
Francesco Ciari, Ecole Polytechnique de Montreal

The COVID-19 pandemic has rampaged through the world causing hundreds of thousands of fatalities, millions of infections, and acute economic and social consequences. In this context, epidemiological models have been put forward as key tools to help public health authorities and policy makers cope with the pandemic. In some cases, these models have proved to be useful aid-decision tools. Nonetheless, in several instances, these models still face several shortcomings and challenges. First, most popular epidemic models still rely on an aggregate representation of the population. Second, the reliability and accuracy of these models in the presence of biased input data is yet to be tested. In this research, we address these two challenges by conducting a sensitivity analysis on the impact of bias in input data, using a transportation agent-based model (MATSIM), and a MATSIM-dependent epidemic model, EPISIM. Our findings stress that epidemic predictions are sensitive towards bias in input data. The underestimation of confirmed infections and the overestimation of the rate of fatalities in official public health reports, bias modeling outcomes. We found that when such biased data are used to train EPISIM, predictions of severe cases and infections are often overestimated. This overestimation bias conveys a false picture of the dynamics of the pandemic. The policy implications of these findings are of interest to both researchers and policy makers.

Hoyoung Lee, Seoul National University
Dong-Kyu Kim, Seoul National University
Seung-Young Kho (sykho@snu.ac.kr), Seoul National University

Previous literature on incident impact analysis usually utilizes only a single indicator, and thus they have a limitation to reveal the inherent features affecting incident-induced congestion. This study aims to develop a framework to quantitatively assess freeway link performance reduction due to traffic crashes by using multiple indices based on the concept of transport system resilience. Resilience indices are designed based on traffic flow characteristics and the adaptive capability concept of transport systems. To identify the spatiotemporal crash impact region, a traffic state classification technique is applied by utilizing a multivariate Gaussian mixture model (GMM). Experimental analyses are conducted based on the data from traffic crash reports and vehicle detection systems (VDS) between 2010 and 2014 collected from four major expressways in South Korea. The results of a multivariate multiple regression analysis show the influential factors on freeway traffic resilience indices and the uncharted characteristics of crash-induced congestion properties. Potential applications of this study lie in the assessment of existing infrastructure during disruptive events. The suggested assessment framework in this study is also applicable to other roadway types, other transport systems (e.g., subway, airline, and maritime), and even different systems (e.g., electric grid, water supply, etc.).