@ Center for Urban Transportation Research at the University of South Florida
Call Us: 1-813-974-3120

Transit Agency Adaptation to Extreme Weather Events: Cataloguing and Modeling, Costs, Decisions and Behaviors

(Center Identification Number: 79050-18)

Principal Investigators:

Dr. Eric Welch, Associate Professor
Science, Technology, and Environment Policy Lab
Department of Public Administration (M/C 278)
College of Urban Planning and Public Affairs
138 CUPPA Hall
Chicago, IL 60607
312-413-2416

Dr. P. S. Sriraj
Director, Metropolitan Transportation Support Initiative (METSI)
Research Assistant Professor
Urban Transportation Center
College of Urban Planning and Public Affairs
412 S. Peoria Street, Suite 340
Chicago, IL 60607
312-413-7568

Sara Hendricks
Senior Research Associate
National Center for Transit Research (NCTR) at the
Center for Urban Transportation Research (CUTR)
University of South Florida
4202 E. Fowler Avenue, CUT 100
Tampa, Florida 33620-5375
813-974-9801

I.  Introduction and Purpose

The project aims to develop greater detail about the decision process, costs, perceived risks and tradeoffs that transit planners and managers consider as part of addressing potentially damaging extreme weather events. The project also aims to work toward the development of decision tools and approaches that can more effectively respond to future weather related challenges that are often associated with climate change.

We plan to conduct a two-part, four-component project that collects different types of data and conducts different types of analysis related to transit adaptation to extreme weather events. Taken together the four components will provide substantial insights into the decision process – specification of costs, occurrence and severity of weather events, recognition of risks and risk perceptions, and the role that inter-organizational network structures play in information dissemination, coordination and support. The four components are:

Part I

  1. Create an inventory of adaptation strategies, with a characterization of costs and benefits of each.
  2. Conduct two case studies of transit agency decisions to reduce the impacts of extreme weather events. The case studies would not only collect cost data, but would also consider perceived benefits, perceived risks, and risk tolerances related – factors that fundamentally affect decision outcomes.  In this regard, the research team will look at the FEMA reimbursement guidelines to understand the extreme weather events eligible for reimbursement.

Part II

  1. Develop a dataset of forecasted climate data, history of extreme weather events data, government finance data, and other institutional data for all transit cities. The data can be used descriptively or integrated in other analyses.
  2. Conduct a national survey of transit agencies to go beyond existing descriptive assessments of adaptation activity, to better explain and inform why transit agencies adopt some strategies or behaviors and others do not. This survey will pay attention to the broader interagency and inter-organizational information and resource networks within which transit agencies operate.

Part I – Analysis of Transit Agency Extreme Weather Adaptation Strategies

The aims of this part of the project are to 1) catalogue the range of extreme weather related adaptation strategies undertaken by transit agencies and 2) collect detailed data and provide in-depth analysis of costs, benefits and perceived risks.  This work will complement existing FTA sponsored climate change adaptation assessment pilot projects by:

  • identifying a broad range of current adaptation strategies and placing them within a recognized analytical framework for ongoing reference
  • partnering with other non-pilot transit agencies to collect cost and other data
  • using different methodologies to assess climate related decision making within transit agencies
  • working toward development of decision analysis tools related to extreme weather event


Part I, Component 1 – Catalogue Adaption Strategies & Collect/Analyze Adaptation Cost Data (To be conducted by USF/CUTR)

This component will collect information and data on a wide range of adaptation strategies that have been developed in transit agencies and elsewhere throughout the world to provide a useful catalogue of adaptation strategies currently underway. Researchers at USF will adopt the framework recommended in the 2010 New York City Panel on Climate Change report, Climate Change Adaptation in New York City: Building a Risk Management Response as a basis for characterizing adaptation strategies. The framework identifies the following adaptation strategy dimensions:

i.    Time frame of the strategy, such as short term, mid-term, long term

ii.    Type of hazard(s) the strategy attempts to address

iii.    Sensitivity of the strategy to underlying assumptions

iv.    Intended effect of the strategy on transit infrastructure, operations, management, planning, and capital investment

v.    Interdependencies of the strategy with other infrastructure including energy, water, and communications.

vi.    Multidimensionality with regard to addressing other goals, such as emergency preparedness and environmental sustainability

vii.    Role of the strategy in prevention, adaptation, preparation and mitigation, response, and recovery

viii.    Flexibility of the strategy to incorporate/adjust to changing conditions and technologies

ix.    Cost of strategy implementation, where cost data are available.  Where cost data are unavailable, characterize potential order of magnitude of cost, relative to other strategies.

x.    Ways to monitor and evaluate the effectiveness of a strategy, including identification of performance measures

xi.    Diversity and level of involvement of stakeholders in strategy development and implementation

Researchers at USF will then conduct a comprehensive review of cases documented in the academic and grey literatures attesting to how transit agencies have made strategic changes in transit service, operations and maintenance, communications, procurement, planning, and other dimensions as they are affected by weather events (snow, ice, cold, heat, flooding, wind, etc.) and climate change.  A review of international experiences will be a part of this sub-task (Holland, Singapore, etc.). Data and information gleaned from this review will be organized according to the framework and will provide an initial catalogue of adaptation strategies linked to different agencies.

In addition to the literature review, it will likely be necessary to follow up with the different transit agencies and other organizations to collect additional data. Therefore the USF team will also contact key individuals to fill in for missing data to ensure completeness of the catalogue.  The effort will result in a more complete catalogue of different strategies associated with weather related transit impacts.

Unlike the seven FTA-sponsored climate change adaptation assessment pilots aimed at specifying the unique hazards and risks, and adaptation strategies that are advantageous to selected transit agencies, this task will seek to identify a wider array of adaptation strategies.  Additionally, the Adaptation Strategy Catalogue could be posted for ongoing transit agency updates and new entries, making it a living document that can be sustained by the transit community and continuously referenced by decision makers and others.

Deliverable: An inventory of adaptation strategies, with a characterization of the different dimensions identified in the framework.

Part I, Component 2 –Analysis of Adaptation Decisions including Cost and Risk Perception Data (To be conducted by UIC)

This component applies a decision analytic methodology to quantify and model choices that transit agencies make in response to extreme weather events. We will identify two cases studies in two different transit agencies that represent two specific types of weather events that are broadly relevant for transit adaptation. The methodology will make use of influence diagrams and decision trees to structure specific decisions such as choices among a set of new infrastructure technologies or emergency service options. The build out of the decision model will include the direct costs of each choice (e.g., technology) but also the indirect costs such as ridership losses associated with events caused by extreme weather. It will include probabilities associated with the different events identified in the model and perceived risks and risk tolerances of decision makers.  The quantification of costs, probabilities, risks, and risk tolerances will require both the collection of historical data from transit agencies and survey data from agency decision makers.  Sensitivity analysis will be performed to determine how variation in the parameters may affect preferences.  The methodology used in this research hopes to complement that work currently underway at the different FTA sponsored adaptation pilot projects. The work will provide an initial decision tool that can be taught to decision makers for use as one input in complex decisions such as those related to climate change adaptation. This component will be undertaken by UIC and will:

a. Identify one or two types of extreme events that have resulted in agency level consideration of an adaptation strategy.

b. Interview key decision makers to understand what the choices were, how they responded to extreme weather events, and which variables they considered when weighing their response.  Based on this information it will be possible to structure (an initial model) the decision problem; however, further effort will likely be needed to articulate specific adaptation strategies that are considered or can be considered by the agency as a result of a particular type of extreme weather event.

c. Collect cost data related to the damages incurred during and after these events. Costs not only include costs from equipment damage, but also consider reputational damages, loss of life, injury, law suits, etc.

d. Conduct structured interviews/surveys of decision makers within each of the two transit agencies to collect data on:

i.    Levels of uncertainty related to extreme weather events and event outcomes.

ii.    Preferences of decision makers, their perceptions of risk and levels of tolerance for risk.

e. Analysis: Calculate potential costs associated with the strategy based on varying probabilities of events and costs. Identify the best alternative. Conduct a sensitivity analysis to show how sensitive the decision is to different levels of perceived risk and different potential cost ranges.

Deliverable: Decision analysis examples that will take into account the range of costs, perceived probabilities associated with decision choices, and sensitivity of different choices to perceptions of risk and cost estimates associated with different strategies. A brief example is attached. The value to FTA would be a greater understanding of the role that variability in costs and risk perceptions play in making adaptation decisions. The project would also provide an initial introduction to the use of decision analysis methods as quantitative tools for aiding decision making resulting from extreme weather events.

Decision Analysis Example: One equipment purchase decision and subsequent combination of price and event effects. The tree includes only a few of the possible nodes.  Note the structure of this decision problem depends heavily on a narrowly defined decision problem, data on cost ranges for equipment, actual or (more likely) perceived probabilities associated with extreme weather events, costs associated with damage and ridership, probabilities associated with ridership declines, and preferences.

Part II. National Level Study of Adaptation to Extreme Weather Events (To be conducted by UIC)

This part of the study is designed to enable national level analysis of the decision making context within which transit agencies operate.  It consists of two components. The first component will develop a national database that will integrate extreme weather data, transit expenditure data, and other city, regional and federal data of relevance for each transit agency. The second component will conduct a national survey of transit agencies to better understand national level extreme weather event experiences and strategies of decision makers. This survey will complement the other related studies including the recent study by Oregon Transportation Research and Education Consortium (OTREC).  The survey design is still under consideration; however the current plan is to survey multiple individuals within the agency to learn more about individual perceptions, behaviors and experiences regarding efforts to consider and respond to extreme weather events; it will also collect data on the social networks used by agencies to anticipate and respond to extreme weather events.  The two components will overlap in some ways. For example, perceived risk of extreme weather events collected in the survey can map onto extreme weather event data collected in component one. Taken together, the two components will provide the basis for analysis of the relationship between extreme weather events and transit and identify gaps in understanding or particular areas of further attention.

Part II, Component 3 – Develop Dataset that Integrates Weather/Transit/Institutional Data

As the overall objective is to collect and integrate different types for each transit city. The data would include historical extreme weather data, agency level revenue and expenditure data, and other related state and federal data.

Extreme Weather Data Because we expect that extreme weather affects adaptation strategies of transit agencies, it is important to collect data on extreme weather trends.  Adapting a prior work by Tschakert (2007), we create two indexes of climatic stress: incidence and severity.  Incidence is measured as the change in frequency of extreme weather events over time while severity is measured as the level of damage sustained over the same time period.  For all transit agencies, we aim to collect data on extreme weather events that have occurred in their service area over the past 10 years and classify them according to the change in incidence and the change in severity of extreme weather events. We expect that some areas will have had substantial increased frequency and severity, while others will have realized no change or a reduction.  Figure 1 depicts the incidence and severity for selected land grant university towns. Data on extreme weather events will be collected from two different sources both of which are provided online by the National Oceanographic and Atmospheric Agency/National Climactic Data Center (NOAA/NCDC).  The first source, North America Climate Extremes Monitoring (http://www.ncdc.noaa.gov/nacem/index.jsp) provides location-based data on extreme climate occurrences such as “percentage of days when the maximum temperature is greater than the 90th percentile” or “maximum length of dry spell”. An online search engine enables generation of change in annual measures between two different points in time.  The extreme weather data produced by this source will be combined with spatial data (latitude and longitude) for each of the cities in the population such that it will be possible to classify cities in terms of the change in severity of weather patterns over time.  This data will be used to measure incidence of extreme weather events.

Figure 1 Climate Change Stress Indexes

 The second source of extreme weather event data will come from the NOAA/NCDC Storm Events data base http://www4.ncdc.noaa.gov/cgi-win/wwcgi.dll?wwEvent~Storms, which allows selection of extreme events within a certain time period at the county level.  The resulting data provides the number and type of events, as well as the deaths, injuries, property damage and crop damage associated with the event.  We will collect ten years of annual data, calculate change over time in human and property impacts as well as number of events, and incorporate the data in the database as indicators of change in weather severity.  Measures calculated from this data will operationalize intensity.

Transit and Other Data. This component will also assemble ten years of financial data from the National Transit Database. Other data that will be collected may include the government financial data available from the survey of local governments, FEMA data, or other state level data.

Deliverable: A dataset of extreme event, transit data, government expenditure, and other data for each transit area that can be used descriptively or integrated in other analyses.

Part II, Component 4 – National Survey of Transit Adaptation Behavior (To be conducted by UIC)

Conduct a national survey of transit agencies to go beyond the descriptive assessment of current adaptation activity to better integrate weather and disaster response behavior, perceptions about climate change and weather impacts, planning and management activities related to extreme weather events, experience with weather related disasters, collaboration, coordination and information exchange with key partner agencies and others, organizational characteristics and other factors to explain:

  1. How transit agencies currently respond to extreme weather events.
  2. Why some strategies or behaviors are undertaken and others are not.
  3. How individuals within agencies perceive weather related risks.
  4. How well developed interagency collaboration and coordination networks are and how they are used when weather events occur.
  5. What information and knowledge exchange patterns are and how those patterns explain types of response or non-response.

 

Sample Frame. We will conduct an online survey of planners and managers in transit agencies in the United States.  The sample frame will comprise all transit agencies that have total revenues of over one million dollars in 2010. From this frame, it is likely that some stratified random sample of agencies will be selected based on at least two strata: region and size. It is anticipated that the survey will be administered to multiple (up to three) transit planners and managers in each selected transit agency. Under faculty guidance, graduate students will identify the populations of public officials. Team members will make follow-up phone calls to confirm mailing lists and to identify errors.  The UIC team has substantial prior experience administering national surveys of between 3,000 and 10,000 individuals.

Survey Design, Question Format, and Content. The surveys will employ a mixed design to collect two types of data: traditional survey items such as information needs, perceptions of risk, and demographics, and social network data related to extreme weather event planning and management, transit investment and other activity relevant for the study.

Questionnaire development will benefit from interviews conducted in Part I, Component Two. Survey questions will be primarily closed-ended, although several open-ended responses will be included in order to capture some limited qualitative data.  Question structure and implementation process will be based on the principles of survey research and designed for maximum response rate and the validity and reliability of data ( Dillman, 1991; Dillman & NetLibrary, 2007b; Dillman, Tortora, & Bowker, 1998b).

Standard principles of survey implementation typically include a pretest process to a small sample of respondents, followed by survey revisions. Actual implementation includes alert letters, instrument mailings, post card follow-ups, second mailings, telephone, and email reminders (D. A. Dillman & NetLibrary, 2007a; D. A. Dillman, Tortora, & Bowker, 1998a). In previous studies of government officials, we have obtained average response rates of 40%. We expect the same for this study. A pretest of the survey will be conducted to check for question content, clarity, and response variability. Following the pretest, an alert letter will be mailed on university letterhead alerting potential respondents to the survey, and providing the URL, individual ID and password. One week later, an email notification will be sent to respondents, with additional follow-up reminders every week or two over the course of seven weeks. As an inducement, respondents will be offered access to study results. As required by the university’s Institutional Review Board, respondents will be informed of anonymity protection and Protection of Human Subjects review.

Deliverables: 1) Descriptive assessment that compares extreme weather event history with perceived risks for transit agencies. This would demonstrate the gap between risk perceptions and reality can be studied with this information. 2) Statistical modeling that would investigate transit agency behavior, risk perceptions, network structure and information/resource exchange, among others. The project will provide more detailed analysis and understand about how and why agencies respond to extreme weather events than currently exists.

II. Synthesis

Overall the project addresses adaptation to extreme weather events, it also examines what current efforts and costs are, how to specify costs and risks within a practical decision making context, and what the gaps are between current agency adaptation efforts and extreme weather realities. The four components of the project, while presented as separate strands of research, will be integrated in several ways.

Key findings from USF analysis (Part I, Component 1) will be compared with UIC decision analysis (Part I, Component 2) to determine best approaches to developing useful decision tools that recognize and incorporate extreme weather events into decision making practices. Component 1 provides a descriptive catalogue of approaches while Component 2 provides a more detailed approach to delineating potential choices and their associated costs, probabilities, risk tolerances, and model sensitivities.

Second, descriptive assessments will be conducted and will compare extreme weather event history and perceived risks of and responses to extreme weather as reported and perceived by transit agencies (Components 3 and 4). Finally, statistical modeling will investigate how and why agencies respond to extreme weather events than currently exists. The findings from these statistical analyses would help interpret and generalize findings from Components 1 and 2 for transit agencies in the US.

III. Implementation Schedule

Part I is to be accomplished within the 18 month time frame of the project. Components 1 and 2 will begin immediately. The final report for Part I, incorporating components 1 and 2, will be delivered in December 2013. Part II will begin in fall 2012 with the collection of extreme weather event data. Activities surrounding the national survey of transit agencies will begin in the spring of 2013 with the identification of a pool or potential respondents, contact list development, and survey development. The national survey will be administered in the fall of 2013. Analysis and presentation of survey and other collected data will extend through summer 2014.

V. Budget

The overall project budget is $375,256. The USF/CUTR portion is $75,000 and will be conducted by S. Hendricks. The UIC portion is $300,256 and will be conducted by Eric Welch and P.S. Sriraj.  Part II of the project is dependent on future funding from the Department of Transportation.

 

 

 

 

2 Comments
  1. Your project is of fundamental importance, and of particular importance to us. We are an ad-hoc group looking at alternatives for public transport in Ulaanbaatar, the world’s coldest capital city (annual mean temperature) and the world’s coldest capital city in winter (minus 30C is quite usual).

  2. Robin: Thank you for your interest in this project. I am on the study team and would like to learn more about the experience of Ulaanbaatar. I can be reached at hendricks@cutr.usf.edu

Leave a Reply

Please complete the simple math problem prior to submitting your comment. This reduces spam. Thanks! * Time limit is exhausted. Please reload CAPTCHA.

Sponsors

  • Florida Department of Transportation
  • US Department of Transportation

Archives

Address

National Center for Transit Research c/o Center for Urban Transportation Research 4202 E. Fowler Ave., CUT100 Tampa, FL 33620-5375