Transportation Resilience: Adaptation to Climate Change (2016)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

40 APPENDIX A: WHITE PAPER Transportation Resilience Adaptation to Climate Change and Extreme Weather Events H. G. Schwartz, Jr., Consultant, USA Lori Tavasszy, Delft University of Technology and TNO, Delft, Netherlands 1 introduCtion This white paper is intended to set the stage for the June 16–17, 2016, European Union–United States (EU–U.S.) symposium Transportation Resilience: Adaptation to Climate Change and Extreme Weather Events. Climate change is a matter of increasing concern worldwide, and nowhere will its impacts be felt more strongly than with the built infrastructure—the transportation, energy, water and wastewater, health care, and com- munications systems that underpin our economy and society. The focus of this symposium is on the research needs to design, build, operate, and maintain transpor- tation systems that are better adapted to the predicted changes in Earth’s climate. In other words, it is on how to develop more resilient transportation systems. Resilience has been defined as “the ability to pre- pare and plan for, absorb, recover from, and more suc- cessfully adapt to adverse events” (1). It is a complex problem, especially for the vast transportation net- works of the world, encompassing not only the physi- cal infrastructure, but people, environment, operation, maintenance, and emergency response. Moreover, the interactions between transportation systems and other sectors, such as power generation and distribution, agriculture, manufacturing, retailing, waste manage- ment, health care, and communications, must be under- stood and addressed. There are so many aspects to the term transportation resilience that it will be a challenge to the symposium participants to identify the critical research needs of the transportation community to create a more resilient future. A few of the overarching questions that might be addressed include the following: • What do we know or need to know about climate change and extreme weather events? • How can broad geographic climate projections be downscaled to local and regional levels? • Do we understand how to make decisions when faced with inherently uncertain conditions? • What makes a transportation system, or its parts, resilient enough? • What technologies might be developed or refined to reduce or even prevent damage from extreme weather events? To set the stage, the next section (Section 2) addresses the science of climate change: what scientists are pre- dicting and why. Section 3 discusses decision making under conditions of uncertainty. As climate change is itself complex and our understanding of it is continually improving, how do transportation professionals make decisions today for systems that have lifetimes of 50 or more years? The fourth section of the paper deals with risk assessment and suggests the need to incorporate risk management techniques into decision making. Finally, in Section 5 we attempt to draw on the preceding sections to address the fundamental subject of the symposium, transportation resiliency.

41A P P E N D I X A : W H I T E P A P E R 2 Climate Change imPaCts on transPortation The preponderance of scientific research indicates that Earth is warming at an accelerating rate and that this change is due in large measure to the use of fos- sil fuels. Earth is surrounded by an envelope, the tro- posphere, filled in part with what we call greenhouse gases (GHGs) such as carbon dioxide (CO2), methane, and water vapor. These GHGs are essential in main- taining a temperate climate, but increases in their con- centrations are causing changes to the climate, namely global warming. Fifty million years ago CO2 levels in the atmosphere are believed to have averaged 1,400 parts per million (ppm), and temperatures were 10°C (18°F) higher than in the preindustrial period—there was no ice on Antarctica (2). Data from ice cores have been used to reconstruct Antarctic temperatures and atmospheric CO2 concentrations over the past 800,000 years. Over the past 800,000 years, CO2 concentrations in the atmo- sphere have increased from a range of 170 to 300 ppm to just over 400 ppm, mostly in the past 40 years. As the record shows, the recent increase in atmospheric CO2 concentration is unprecedented in the past 800,000 years (Figure 1) (3, 4). During that period Earth’s temperature has also risen significantly. The data clearly demonstrate a steady increase in global temperatures (Figure 2) (5). Indeed, although the first decade of the 21st century was the warmest on record, 2013 through 2015 were even warmer, with 2015 being the warmest on record (5). Temperatures on land, in the oceans, and in the troposphere have all increased over this time period, with the greatest increase in the Arctic. These climate changes coincide with the great increase in the use of fossil fuels, beginning with the Industrial Revolution, but accelerating greatly since the 1970s. Natural variability due to sun spots, Earth’s wobble on its axis, and events like El Niño and La Niña can- not explain the climate changes over the past 30 years. Only when one factors in the use of fossil fuels can the recent changes be explained. The clear conclusion is that Earth’s climate is warming and that the changes are due in large measure to anthropogenic activities. Much of the discussion about climate change has revolved around actions to decrease GHG emissions to reduce or mitigate the extent of climate change. The COP [Conference of Parties] 21 meetings in Paris this past fall [2015] established new targets for reducing CO2 emissions for both the industrialized world and develop- ing countries. The new agreement accommodates inter- national differences in political attitudes and approaches to climate change, especially the perspectives of develop- ing countries. Regardless of the policies and indeed actions imple- mented today and in the near term, the impact of increas- ing worldwide GHG emissions will continue for decades. 4 0 –4 –8 –12 400 350 300 250 200 150 800,000 600,000 200,000 0400,000 Years before present CO2 concentration, ppmv Current Antarctic temperature, °C FIGURE 1 Temperature and CO2 from Antarctic ice cores over the past 800,000 years. The cyclical pattern of temperature variations constitutes the ice age and interglacial cycles, during which changes in CO2 concen- trations (blue line) track closely with changes in temperature (red line) (ppmv = parts per million by volume) (4, Figure 3, p. 10, and https://www2.bc.edu/jeremy-shakun/FAQ.html). Accessed May 10, 2016.

42 t r a n s p o r t a t i o n r e s i l i e n c e Mitigation is a long-term solution, perhaps, but the world must take proactive steps now to adapt to climate changes. 2.1 Major Impacts on Transportation In transportation, as with much of the built infrastruc- ture, five specific impacts will need to be addressed (6): • Sea level rise, • Higher temperatures and longer heat waves, • Changes in precipitation patterns, • Rising Arctic temperatures, and • Increased intensity of storms and hurricanes. These impacts may occur simultaneously, which will aggravate the final effects. Sea Level Rise Globally, sea levels are projected to rise by as much as 1.5 to 3 feet (0.5 to 1.0 meters) or more (Figure 3c) by the end of this century as a result of melting glaciers, most notably the Greenland ice sheet, and expanding oceans as the sea warms (3). But the increases are not uniform around the world; instead, they reflect local or regional factors such as subsidence, land rebound as glaciers melt, and prevalent wind conditions and ocean currents. For example, in many regions such as the Gulf Coast of the United States, relative sea level rise will be greatly exac- erbated by land subsidence. Relative sea level rise could be as much as 4 feet in this region by 2100, inundating as much as 2,400 miles of roadway as well as rail lines and ports (6). Couple storm surge with higher sea levels, and the impacts and damage will extend much farther inland. Hurricane Katrina in New Orleans in 2005 brought storm surges of 25 feet, literally lifting major bridge structures off their piers and flooding the New Orleans airport. Similarly, storm surges from Hurricane Sandy on the East Coast of the United States flooded miles of the New York subway system as well as two of the three New York airports. Impacts on other elements of the transportation system as well as most other seg- ments of the built infrastructure, the economy, and social systems are too numerous to discuss in detail in this paper. 1.2 1.0 0.8 0.6 0.4 0.2 0 –0.2 –0.4 –0.6 –0.8 C h an g e fr om A ve ra g e (° F) Decade 2001–2012 even warmer. Every year warmer than 1990s average. 1990s even warmer. Every year warmer than 1980s average. 1980s warmest decade on record at the time. 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s FIGURE 2 The past five decades have seen a progressive rise in Earth’s average surface tempera- ture. Bars show difference between each decade’s average temperature and the overall average for 1901–2000 (5).

43A P P E N D I X A : W H I T E P A P E R Higher Temperatures and Longer Heat Waves Scientists are confident that Earth will continue to warm, perhaps by as much as 2.6°C to 4.8°C (4.7°F to 8.6°F) by 2100 in the absence of significant mitigation measures. Figure 3a shows projected temperature changes across the globe under two modeling scenarios. Recognizing the political realities of implementing carbon mitigation mea- sures rapidly, it is very likely that the Earth will continue to warm for decades to come. High temperatures and heat waves will become more intense, longer lasting, and more frequent. The impacts FIGURE 3 Worldwide (a) temperature, (b) precipitation, and (c) sea level increases predicted by the end of this cen- tury under two models: (left) RCP 2.6, with aggressive emission reductions, and (right) RCP 8.5, with emissions con- tinuing on their present trajectory (7).

44 t r a n s p o r t a t i o n r e s i l i e n c e on the built infrastructure will be severe. In transporta- tion, the heat will affect expansion joints, rail stresses, and pavement deterioration. Construction work in many regions will have to shift to night hours to protect work- ers from extreme temperatures. More severe drought in many areas will affect not only agriculture, but transpor- tation systems, most notably by creating the kindling for potentially huge wildfires that can disrupt many modes of transportation. Changes in Precipitation Patterns Warmer atmospheric temperatures will lead to important changes in precipitation patterns: more drought in some regions and heavier storms in others (Figure 3b). To sim- plify, warm air causes more evaporation, but it also can carry more moisture so that rain- and snowstorms may well become more intense. For example, in the United States, the Southwest is currently suffering from extreme droughts, a condition that is likely to continue, while the Midwest and Northeast receive more intense rainfall events and flooding. Not all changes in precipitation patterns can be attrib- uted to anthropogenic-driven climate changes. In fact, natural variability, especially the El Niño and La Niña Southern Oscillation in the Pacific Ocean, has an enor- mous impact on precipitation patterns worldwide. The extreme weather events arising from these phenomena are equally disruptive and should be a part of any adap- tation strategy. It would certainly appear that the tradi- tional return frequency calculations based on historical data no longer will accurately guide designers and opera- tors of transportation systems. What was once a 1:100 year storm may become a 1:20 year event. These shifts in precipitation have profound effects on the transportation industry. Strong storms and flood- ing will affect and are affecting many communities in the EU and the United States. Heavy rains and changing waterway levels will affect operations of the transpor- tation services sector. Engineering design standards and operating practices must change to accommodate more frequent heavy storm events. By contrast, as mentioned previously, severe drought can create a different set of challenges for transportation professionals. Rising Arctic Temperatures Climate change, particularly warming, is manifest most strongly over land and in the far northern latitudes. Arc- tic Alaska, for instance, has already witnessed tempera- ture rises of 3°F to 5°F, double those of the continental United States. Thawing of permafrost will create major disruptions to highways, railroads, pipelines, and even buildings. Sea ice is retreating rapidly, particularly during the summer months, but the maximum extent of Arctic sea ice in March 2016 was a record 431,000 square miles below the long-term average (8). The fabled Northwest Passage through Canada and Russia is now open for shipping in the summer months. Increased Intensity of Storms and Hurricanes There is some uncertainty about how or if climate change affects hurricane frequency, but there are basic scientific reasons to believe that the intensity of hurricanes may be increasing. As the oceans warm and the atmosphere becomes more moist, especially in the North Atlantic and the Caribbean, more energy will build up in these storms. With stronger hurricanes and higher sea levels, result- ing storm surges will reach much farther inland. Chang- ing wind directions may disrupt airport operations, and strong or gusty winds may cause delays in traffic and accidents due to truck rollover crashes. Although hurricanes are not a problem in Europe, very strong storms with sustained high winds are a major con- cern. Storm surge accompanying high winds and high tides can cause extensive flooding of both urban and rural com- munities. London’s Thames Barrier and the Dutch Delta Works system of dams and surge barriers are examples of the extensive engineering structures built to protect coastal cities. Moreover, high winds can reduce vehicular speeds, especially the speed of large trucks, affecting overall traffic flow. Superposition of Changes The above impacts will occur simultaneously and together determine the final effects on the transportation user. Figure 4 illustrates this for road transport. The superposition of effects can have several conse- quences. It may cause a culmination of problems in critical places. Impacts may be softened or exacerbated. Knowl- edge about the propagation of effects and their interplay is vital to be able to design cost-effective policy packages. 2.2 Attribution of Extreme Weather Events Not surprisingly, many of these climate changes are reflected in extreme weather events, but at present few singular events can be directly attributable to climate change. A new area of climate and weather research is emerging, however, called event attribution. Using either observational records or modeling or, more likely, a combination of both, the science of event attribution is advancing rapidly according to a recently released report by NAS (10). Confidence in making a connection between

45A P P E N D I X A : W H I T E P A P E R a specific weather event and climate change is greatest when the type of event has a long-term historical record, such as events related to temperature. Such information would be of great value to transportation professionals, but there is much more research needed before meaning- ful information will be available to the practitioner. 2.3 Climate Models Forecasting future climate changes is a very complex matter, but one of immense importance if we are to adapt successfully to these changes. Climate scientists have expended great effort developing a panoply of mod- els. Consider some of the variables: atmospheric–ocean circulation, population and economic growth, energy sources and utilization, new clean technology develop- ment, continued deforestation, and the CO2 absorption capacity of the ocean. In 2000, IPCC9 (Intergovernmental Panel on Cli- mate Change 9) developed a series of socioeconomic scenarios that were used to calculate global GHG emis- sions. These scenarios, known as Special Report on Emission Scenarios, have been widely used to estimate future conditions. The National Climate Assessment in the United States published in 2014 used two scenarios, B1 and A2, to create an envelope of possible futures as shown in Figure 5a (11, 12). They also included more recent models, adopted by the IPCC in 2013 (13), with the representative concentration pathways shown in Figure 5b. Both sets of models show fair agreement in temperature and CO2 projections through about 2050 but, not surprisingly, the models clearly diverge after the mid-century mark. The lower values reflect faster and greater control of fossil fuel emissions, and the upper curves anticipate greater population growth and higher levels of emissions. A related issue concerns scale. Most of these mod- els are at a global or at least continental scale. What transportation and other infrastructure designers and operators need are projections at the local or regional scale. So-called downscaling techniques can be used to Long-term changes in travel patterns Cost of cancelled or postponed trips Vehicle loss hours Congestion potential, gridlock Closed or blocked roads and bridges Reduced severity of incidents Reduced road speed and capacity Increased probability of incidents Closed or blocked roads and bridges Reduced road speed and capacity Instability of road substructure Opening or closing problems of steel bridges Roadside fires Melting asphalt and rutting Maintenance costs Water, snow, or ice on the roads Reduced visibility Infrastructure costs Flooded roads or tunnels Evacuation Increased noise Overturned trucks, caravans, etc. Trees & branches on the road Drought Extreme temperatures Extreme rain, snow, or glazed frost Increase in average sea level Flooding Exreme wind Infrastructure costs Indirect economic effects FIGURE 4 Aggregation of impacts on road infrastructure (9).

46 t r a n s p o r t a t i o n r e s i l i e n c e provide more localized information, but the accuracy may be limited. Certainly one important need is for accurate climate projections at the regional and local levels. 2.4 Intermodal and Cross-Sector Issues Transportation in its entirety can be thought of as a sys- tem of systems, many interconnected and dependent on each other both in normal operations and in emergen- cies. Just this past winter, floods in South Carolina closed Interstate 95, one of the busiest highways in the United States, for weeks, necessitating a 150-mile detour par- tially on surface roads for passenger and freight traffic. In the United Kingdom, a series of severe storms in Decem- ber 2015 destroyed a railway viaduct on the main line linking Scotland and England, interrupting train services for 2 months—a major disruption with serious economic consequences. Critical nodes and routes must be identi- fied and plans made to operate in times of stress. Although it may be obvious, transportation cannot be considered in isolation, but as an essential, integral part of the world’s social, economic, and environmental fabric. Like the warp and weft of a fabric, transporta- tion interacts with many, if not most, sectors of today’s society. Extreme weather events and climate change will affect the whole cloth and will also affect energy, water and wastewater, information technology (IT) and com- munications, health care, agriculture, the economy, the environment, and more. Consider a few examples of the cross-sector impacts: • Power system failures can disrupt many transpor- tation modes and transportation-related infrastructure, such as airports, pipelines, traffic signalization, and com- munications. • Conversely, transportation disruption may prevent fuel from reaching power plants and chemicals from being delivered to water and wastewater treatment facilities. • Hurricanes and floods can totally disrupt com- merce and freight movement, with major impacts on a nation’s economy. Logistics and supply chains can be affected. • IT and communications breakdowns will bring parts of the transportation networks to a halt, notably 12 10 8 6 4 2 0 –2 1900 1950 2000 Year Te m p er at u re C h an g e (° F) Year 2050 Observations Modeled Historical SRES A2 SRES B1 Observations Modeled Historical RCP 8.5 RCP 2.5 SR ES B 1 SR ES A 2 RC P 8. 5 RC P 6. 0 RC P 4. 5 RC P 2. 6 SR ES E A 1B (a) (b) 2100 1900 1950 2000 2050 2100 12 10 8 6 4 2 0 –2 FIGURE 5 Temperature forecasts from (a) Special Report on Emission Scenarios (SRES) and (b) representative concentration pathway (RCP) scenarios (11, 12).

47A P P E N D I X A : W H I T E P A P E R passenger and freight rail service, air service, transit sys- tems, and pipeline fuel distribution. • In emergency conditions, transportation is criti- cal to provide safe evacuation and maintain health care operations. • Major flooding can make movement of agricul- tural products difficult or impossible. • With rising sea levels comes increased risk to coastal communities, their homes, businesses, and virtu- ally all aspects of their infrastructure. Detailed maps can be built of cascading impacts across infrastructures. Little research has been done on these effects from a systems perspective (14). 2.5 Key Challenges for Adaptation Practice There is general acceptance that the climate is indeed changing and that our planning, design, operation, and maintenance practices must proactively address these changes. No longer can we rely on historical data to plan, develop, and operate the transportation systems of the 21st century. Rather, we must recognize the impor- tance of climate science, acknowledge that our scientific understanding of the climate is improving, and apply new techniques of decision making using more sophisti- cated risk management and uncertainty methodologies. In essence, the development of future transportation sys- tems has become more complex, but ever more important to the overall welfare of society and the environment. The past decade has witnessed a great change in the approach to climate change within the transporta- tion community. The management of most public and private transportation agencies and companies now insists that climate change be addressed not only in new facilities, but in the operation of existing systems and the development of response mechanisms for extreme weather events. From individual passenger drivers, to freight haulers, to shipping lines, to the businesses they serve, all want more resilient transportation systems. Policies directed at increasing system resilience have to operate simultaneously at different geographical scales (local, national, global), at several temporal scales (short, medium, and long term), and in different stake- holder dimensions (public and private) and are therefore inherently complex. Depending on this scoping, different strategies will prevail, from short-term approaches based on early warning and coping to long-term-oriented adap- tive or transformative policies (Figure 6). Adaptation policy efforts have accelerated in the past decades. For a detailed overview of achievements we refer the reader to the IPCC’s climate adaptation reports (11–13). In the next subsection, we provide a brief impression of the state of research and practice within the EU and the United States and the key areas for research and development (R&D) that have been identi- fied by the IPCC. 2.6 Achievements in the European Union and the United States The IPCC’s Fifth Assessment Report on adaptation rec- ognizes the globally increasing attention to adaptation policies, both in terms of policies and of actual adapta- tion measures (7). The IPCC also points out differences in how adaptation has developed in the EU and the United States. Adaptation efforts appear to have been more equally spread across different levels of govern- ment in Europe as compared to the United States. The EU has seen large international R&D programs directed at climate adaptation of infrastructures. The priorities and implementation approaches of the two regions dif- Maintain or Recover Status Quo Adapt Incremental Innovation Transform Disruptive Innovation RESILIENCE APPROACHES City Functions Tolerance Range City Functions Tolerance Range TIME FIGURE 6 Illustration of impact of temporal scale on preferred strategy (15).

48 t r a n s p o r t a t i o n r e s i l i e n c e fer due to differences in adaptation priorities, availability of funding, and degree of top-down regulation of infra- structure policies. Nevertheless, the EU and U.S. regional reports are broadly in agreement about three key gaps in climate adaptation research. First, there is agreement on the need for integra- tive research that allows understanding of phenomena across different sectors and types of impact, toward accumulation at the local level. Overall, R&D efforts have primarily focused on incremental adaptation measures to be mainstreamed into existing asset poli- cies, rather than on comprehensive adaptation policies. Particular attention would need to go to cobenefits or counterproductive effects of combined measures and into approaches that explain place-based resilience from a complex of factors. Second, these regional reports mark similar substan- tive gaps in the current body of knowledge. Both call for more research on critical infrastructures, includ- ing transport, water and energy supplies, and health services, including related urban and rural planning and governance challenges. Other caveats include the following: • Costing methods and statistics are lacking for spe- cific cases, including biodiversity, business and industry, and population health costs. • The impacts of the new high-end scenarios of cli- mate change (>4°C global average change, with higher temperature change in Europe) need to be developed. • Rural development, including resilience of cultural landscapes (e.g., old cities, heritage sites) and communi- ties and managing adaptation in low-technology (pro- ductively marginal) landscapes, needs consideration. • Information is needed to manage agricultural and forestry systems. Third, they identify a need for additional method- ological work, in particular through increased alignment of regional monitoring and evaluation approaches, for adaptation policies and climate change knowledge. As we discuss further in Section 5, R&D efforts have so far focused on creating framework conditions for adaptation policy (in terms of data, instruments, and assessment methods) and incremental adaptation measures that have been mainstreamed into existing asset policies rather than on comprehensive adapta- tion policies. One could derive from these efforts that climate adaptation is still at an early stage of the policy cycle. To our understanding, an important stumbling block for transformative policy actions is the phenomenon of deep uncertainty that is associated with climate change. We discuss this issue in the next section. 3 deCision maKing under Conditions of unCertainty 3.1 Uncertainty in Climate Change: What Is It About? The notions of probability and uncertainty are deeply embedded in the recent global agreements about climate change. The United Nations Framework Convention on Climate Change of 1992 uses danger as a central concept in its aims, in which safe levels of climate change allow “ecosystems to adapt naturally, food production not to be threatened and economic growth to proceed in a sustain- able manner.” As danger can be associated with risk and risk with probability, the notion of uncertainty becomes more than just a numerical fact that we need to deal with in our calculations. It becomes instrumental in measur- ing the magnitude of the real problem and our progress in managing the outcome. The IPCC’s Fifth Assessment Report on adaptation mentions the word “uncertain” and its variants around 1,100 times, at an average frequency of slightly less than once per page (11, 12). TABLE 1 Breakdown of Uncertain Factors Behind Impacts of Climate Change on Society Natural Variability of the Climate System Severity of climate Human influence on GHG stock Autonomous development Demography, economy, social, change technology, politics Effect of mitigation measures International agreements and their implementation Climatic response Response mechanisms Impact on society Socioeconomic impacts Vulnerability Future welfare levels Exposure Detailing climate predictions Responsive capacity Resilience, robustness Impact on well-being Economy Interest rates and sustainability Environmental Absorptive capacity, valuation Social equity Intergenerational and social redistribution NOTE: Shaded cells indicate the current focus of climate adaptation measures.

49A P P E N D I X A : W H I T E P A P E R There is substantial uncertainty in the expected impact of climate change on the functioning of society (see Table 1). Measuring the effects of climate change in economic terms, the 2006 Stern Review predicted a per- manent impact of a reduction of between 5% and 20% of the gross domestic product (16). There has been a lot of debate about the assumptions behind these figures, and the resulting uncertainty, despite the already broad bandwidth. To predict a decrease in quality of the trans- portation system, as experienced by its users, several fac- tors need to be taken into account. It is possible to reduce uncertainty in projected impacts but impossible to eradicate it completely. Epistemic uncertainty (due to limited knowledge) can be reduced to a certain extent by research and improved measure- ment, aleatory uncertainty (due to inherent variability of the system) by using scenario-based or stochastic projections, and ambiguity (due to multiple definitions of phenomena) by improved communication (17). It is therefore important to understand the different factors that cause uncertainty in impact projections (Figure 7). The natural variability of the climate system and, primarily, our lack of understanding of these variations create significant uncertainty in predictions of atmo- spheric concentrations of GHGs, global warming, and phenomena such as sea level rise and weather changes. Emissions due to human activity are strongly dependent on economic and population growth as well as tech- nology and geopolitics. The physical impact of climate change on society is the primary concern of adaptation policies. The severity of these impacts depends on vul- nerability (i.e., the natural sensitivity of areas to changes such as drought), the expected exposure to changes in climate, and the capacity to respond to this exposure, either through coping (resilience) or by adaptive mecha- nisms. As adaptation measures only affect part of the causes of climate change impacts, they will be limited in their ability to reduce the uncertainty of these impacts. The propagation of the physical impacts into wider and final socioeconomic and environmental damage is a separate question. Especially surrounding cost–benefit analysis and sustainability impact analysis, important topics are second- and third-order effects, discount- ing for capital loss, and avoiding double counting of impacts. As an illustration, the numbers from the Stern Review mentioned above would be an order of magni- tude lower (i.e., gross domestic product impact below 2%) if the assumed discount rate were a couple of per- centage points higher. Uncertainty occurs in each of these factors and cas- cades through the different steps by which we predict climate change. In order to reduce uncertainties, besides understanding the main phenomena, their formal description and modeling need to be improved. Trans- portation system models generally have a much higher spatial granularity than climate models, which makes it necessary to detail forecasts to a level at which one can distinguish individual roads, for example. But even if accurate and detailed predictions of extreme weather sit- uations were available, often the transportation system models are incapable of calculating the consequences, due to a lack of empirical knowledge of these relations. Future GHG emissions Transportation system impacts Transportation models Transportation system conditions under future climate Adaptation measures Socioeconomic impacts Socioeconomic scenarios Downscaling and statistical correction Climate projections Climate models FIGURE 7 Climate impact projection pathway (17).

50 t r a n s p o r t a t i o n r e s i l i e n c e 3.2 The Science of Decision Making Under Great Uncertainty Uncertainty in projections will remain and, therefore, it is important for decision makers to develop approaches that use this information and embrace uncertainty rather than deny it. Climate change has been termed by some as “post-normal science” (18). Normal science uses expert knowledge–based decision making and conventional tools for policy analysis such as utility theory, contin- gent valuation, cost–benefit analysis, and statistical deci- sion theory. Post-normal science functions in a world of intractability of facts, deep uncertainty, disputed values, high stakes and, sometimes, urgent decisions. Under these circumstances, the decision-making process becomes as important as the facts that support decisions. Decision makers have fundamental problems dealing with uncertain factors. First, there are different types of uncertainty (e.g., on the one hand caused by knowledge that is incomplete but potentially attainable, such as the effects of extreme weather, and on the other hand caused by unknowable factors, such as the future). Not only are uncertainties difficult to understand, they also require responses that sometimes lie far away from decision mak- ers’ capabilities or mandates. Second, the different ways in which calculation and presentation of uncertainties can be accomplished make their processing by decision makers a complicated task and, more importantly, intro- duce the subjectivity of the researcher into the process. Policy makers, often unknowingly, act with information that appears much more certain than it actually is (19). Once uncertainty is of a different nature or a higher level than what the decision maker is accustomed to, decision- making processes become more difficult (20). The concept of deep uncertainty is central to under- standing climate change impacts. Figure 8 shows differ- ent levels of uncertainty, from a certain world to a fully unknown world. Level 4 and Level 5 uncertainties depict the so-called deep uncertainties. These are the ones that cannot be treated probabilistically and include uncer- tainties of model structure or those that experts cannot agree upon (21). Climate change uncertainties in general can be classified as such, although there will be varia- tions in predictability between factors (e.g., temperature anomalies are easier to predict than rainfall, and global phenomena generally easier than local ones). The approach toward policy analysis will be heavily dependent on the level of uncertainty recognized. Prediction- based, linear policy measures are usually based on a percep- tion of the world at Level 1. This approach is equivalent to ignoring uncertainty. Already more sophisticated policies are based on projections that take into account historical A clear-enough future A single (deterministic) system model Alternate futures (with probabilities) Alternate futures with ranking A multiplicity of plausible futures An unknown future Unknown weights; know we don’t know Unknown outcomes; know we don’t know Unknown system model; know we don’t know A known range of weights A known range of outcomes Several system models with different structures Several system models, one of which is most likely Several sets of weights, ranked according to their perceived likelihood Several sets of point estimates, ranked according to their perceived likelihood Several sets of weights, with a probability attached to each set A confidence interval for each outcome A single (stochastic) system model A single set of weights A point estimate for each outcome Context System model System outcomes Weights on outcomes Level 1 Level 2 Level 3 Level 4 Level 5 Total Ig n oran ce LO C AT IO N C om p le te C er ta in ty LEVEL FIGURE 8 Increasing levels of uncertainty and underlying assumptions (21).

51A P P E N D I X A : W H I T E P A P E R uncertainty (Level 2). Level 3 policies are usually based on a choice of a most likely future scenario. The typical answers to Level 4 and Level 5 uncertainties include planning for the worst conceivable situation (resistance), planning for quick system recovery (resilience), or for adjustment for different scenarios (robustness) (22). We elaborate on these strategies in Section 4. 3.3 Implications for Adaptation Policy Deep uncertainty requires a different approach to policy design than shallow uncertainty. The most fundamen- tal change from current practices will be that we can no longer base our policies on simple predictions. We have to learn to become adaptive to hedge our investments against severe uncertainty. A relatively new approach, adaptive policy making, which is gaining interest among decision makers, replaces the forecasting-based planning paradigm with a dynami- cally responsive approach (23, 24). As prediction becomes harder, proactive planning loses its value, and preparation for response in multiple scenarios is preferable. By setting predefined thresholds (e.g., sea level rise at a certain point in time), including an appropriate response policy, the necessary policy will only be activated once these thresh- olds are passed. Depending on the circumstances, different policies will be needed. Several policies are kept in stock to cater to all realistically conceivable cases. To select the best alternative policies, the expected value of different policies is determined under different circumstances. To this end, option-valuation approaches are used instead of conven- tional benefit–cost analysis. Currently, however, most infrastructure planning and management systems still build on notions of mild uncer- tainty, using means and probabilities. Essential for asset management practice is the recent inclusion of risk as a new concept on which to base decisions. Including risk is a precondition to address events that have a low-frequency, high-impact nature. In a world of deep uncertainty, risk can be used as a dynamic concept in which thresholds for risks are used to trigger the implementation of adaptation actions. The next section introduces risk management, in particular the U.S. risk-based system for building resil- ience into transportation assets, along with recent R&D efforts in the same direction within the EU. 4 risK management 4.1 Risk Risk can be defined as “the positive or negative effect of uncertainty or variability upon agency (or personal) objectives” (6). Statistically, risk is the probability of an event occurring times some measure, often cost, of the consequences of that event. In terms of climate change and its impact on transpor- tation, risk analysis is the identification of the hazards of concern (e.g., sea level rise, extreme weather events, storm surge); the vulnerable infrastructure assets (e.g., bridges, highways, airports); the potential direct and indirect consequences, including cost to the economy and social and environmental costs; and the probability that the hazardous event will occur. The challenge is to balance the risks with the benefits and costs in a rational manner. Situations will vary from high-probability, low- consequence events such as flooding of low-traffic roads located in the floodplain to low-probability, high- consequence events such as Hurricane Sandy in New York and New Jersey or the extensive flooding in Cen- tral Europe in 2013. Some climate changes will develop gradually over years or decades (e.g., sea level rise), but other changes relate to extreme weather events, whether a direct result of climate change or other more transient climate impacts such as El Niño. So-called “black swan” thinking emphasizes the importance of unpredictable, extreme events in con- trast to the traditional normative analysis. This school of thought has much to recommend it with regard to climate change and extreme weather events. Transpor- tation systems are developed with a long-time horizon, 50 or more years, and traditionally have used historical weather data for planning and design purposes. Yet so often damage to these systems arises from unexpected or, at least, unplanned-for events. It might be argued that the planning process underestimates the impact of unpredictable events, extreme weather, or otherwise dis- rupting events. There has been much discussion about climate change tipping points or thresholds beyond which changes may be irreversible. Determining specific tipping points is, at best, an inexact science, but we must recognize that cli- mate change is likely to be a nonlinear phenomenon. 4.2 Risk Analysis Risk analysis as it relates to climate change is anything but simple. Many complexities have to be considered, including the following: • Uncertainty. As discussed in the preceding sec- tion, various uncertainties must be considered, including uncertainty in the level of future GHG emissions, socio- economic impacts, natural perturbations, the Earth’s response to GHG increases, and the capacity to respond. • Gradual versus sudden change. Gradual change must be contrasted with extreme weather events.

52 t r a n s p o r t a t i o n r e s i l i e n c e • The issue of scale. Climate scientists are much more confident about projecting climate changes, espe- cially temperature increase and sea level rise, at the global scale. Accurately downscaling this global infor- mation to the regional and local levels is difficult, but transportation professionals need data at the local and regional scale. • Multiple stresses. Climate change cannot be con- sidered in isolation, but rather in the context of many other stresses (environmental, economic, and social) that affect the human experience. Climate change then becomes an additional stress on the system, one that may become the tipping point that causes the system to be permanently altered. Continued sea level rise coupled with higher storm surge, for example, will place coastal communities—homes, businesses, and the infrastructure that supports them—at greater risk. • System and modal interconnectivity. The intercon- nectivity of different modes of transportation and differ- ent infrastructures (e.g., power, water and wastewater, health care, communications) must be considered in risk analysis. • Probability of an event occurring. Predicting prob- ability is not as simple as using the return frequency alone as a surrogate for probability. For example, over the 50-year life of a specific transportation project, a 1:500 year event has a probability of happening during the project’s design life of 9.5%. 4.3 Risk-Based Asset Management In recent years, the Federal Highway Administration of the U.S. Department of Transportation has pro- moted risk-based transportation asset management (RBTAM), which calls for building resilience into transportation assets. Their 2013 report focuses on the “three R’s”: redundancy, robustness, and resil- ience (25), defined as • Redundancy: “duplicative or excess capacity that can be used in times of emergency.” • Robustness: “the capacity to cope with stress or uncertainty. . . . Well-maintained assets generally are bet- ter able to withstand stresses of storm events and other disasters.” • Resiliency: “the ability to prepare and plan for, absorb, recover from, and more successfully adapt to adverse events, . . . better anticipation of disasters, better planning to reduce disaster losses, and faster recovery after an event.” Although much of this U.S. Department of Transpor- tation report deals with climate change, it importantly notes that good RBTAM practices will better prepare an agency, a community, or a business to absorb other unexpected disasters (an “all hazards” approach) (25). The actions needed to develop a risk-based approach to asset management include the following: • Maintenance of accurate inventories of transporta- tion assets, their vulnerabilities, and criticalities; • A prioritization of vulnerable assets that will under- gird capital improvement plans, maintenance programs, and recovery actions; • Better maintenance practices to strengthen assets (e.g., bridges with well-maintained retaining walls and scour protection will be more robust during floods); • Asset inventories coupled with good repair cost data to speed recovery efforts; and • Thorough geographic information service mapping and preplanning, which are essential for rapid evacua- tion of affected areas. One interesting comparison and possible guideline is the earthquake preparedness program for bridges in California. Following two major earthquakes in the late 1980s, the California Department of Transportation (Caltrans) developed a risk-based approach to prioritiz- ing the seismic upgrading of every bridge in the state, some 24,000 bridges. The primary objective was to prevent loss of life at every location, but not to prevent all damage. Caltrans based the prioritization on three factors: site hazard, structure vulnerability, and system impact. For the seven major toll bridges, the criteria were a little different. Several, including the Golden Gate and Oakland Bay bridges, had to remain serviceable imme- diately after the design earthquake (8.5 on the Richter scale); others might suffer damage but had to be restor- able to service within 6 months. All have been retrofitted to meet the appropriate criteria. A somewhat different or less regimented approach to risk analysis, the observational method, is suggested by the American Society of Civil Engineers and has been applied in the European engineering community. The key elements of the observational method are as follows (26): • Project design is based on the most probable cli- mate condition(s) rather than the most unfavorable. • The most unfavorable conceivable deviations are identified. • A course of action is devised (in advance) for every foreseeable unfavorable climate deviation from the most probable condition(s). • The performance of the project is observed over time and the response of the project to observed changes is assessed. • Design and construction modifications can be implemented in response to observed changes.

53A P P E N D I X A : W H I T E P A P E R The technique has been used for decades in geotech- nical engineering and in its simplest form is an iterative cycle of analyze, plan, design–construct–operate, monitor, and revise. It seems most appropriate for gradual climate changes such as temperature and sea level rise and least applicable when extreme weather events are considered. Recent overview studies on climate adaptation rec- ommend climate risk management as a tool but also indicate that this practice is not at all commonplace (27– 30). In Europe, less than a handful of countries (Aus- tria, France, Spain, and Switzerland) have adopted such explicit risk management plans (28). A notable R&D effort in Europe, funded by 11 countries, was directed at developing a risk management approach that can be adopted by the diverse set of road agencies in Europe. The RIMAROCC framework operationalizes the ISO 31000 risk-management process standard and provides a systematic yet qualitative approach to identify and respond to climate-related risks for road infrastructure (31). It primarily stresses the process dimension of risk management and positions the qualitative method next to analytic approaches. It distinguishes structure, (road) section, and network levels of analysis, which may be a useful feature in aligning the levels of analysis. A detail- ing of quantitative approaches was provided recently for tunnels (32). 5 aChieving Climate resilienCe The IPCC defines adaptation as “an adjustment in natu- ral or human systems in response to actual or expected climatic stimuli or their effects, which moderates harm or exploits beneficial opportunities” (33). Interestingly, there appear to be two distinct streams of thinking about climate adaptation policy. When research and policy are discussed, care must be taken about the existence of this difference in perspectives. The first approach is identified as bottom-up as it deals with the social, or citizen’s per- spective, in the form of a preexisting condition of vulner- ability. The second approach can be seen as top-down, taking the perspective of larger social-technical or bio- physical systems, focusing on physical exposure (Figure 9). Resilience is the positive result of low vulnerability from both perspectives. Ultimately, the two directions need to meet in sensible policies in which biophysical stressors and social stressors are reduced so that resil- ience is increased (33). The results of our adaptation efforts so far have been a mixed bag in terms of types of effort and outputs. A 2011 report of the worldwide state of adaptation practice (34) concluded that a majority of actions involve inten- tions to adapt, without implementation. Adaptation appears to be motivated by short-term climate variabil- Global Past Present Future Local Top-down approach Bottom-up approach Climate adaptation policyVulnerability (social) Vulnerability (physical) Impacts Regionalization Global climate models Global greenhouse gases World development Institutions Equity Infrastructure Information & skills Economic resources Technology Indicators based on: Adaptive capacity FIGURE 9 Dominant perspectives on climate vulnerability (18).

54 t r a n s p o r t a t i o n r e s i l i e n c e ity, especially extreme events; is typically mainstreamed into existing policies; is mainly national government driven; and is directed at final economic impacts. A more recent and larger study (35) analyzed over 4,104 adapta- tion projects from 117 countries worldwide from 2008 through 2012. Although it ranked the United States and five EU countries among the top 10 leading countries in adaptation practice, actions mostly consisted of so-called groundwork activities (assessments and development of assessment tools); only around 15% of efforts related to infrastructural, technological, or innovation projects. This characterization of current adaptation practice indi- cates that we are only at the beginning of a longer and deeper adaptation process that aims to improve overall system resilience. To be able to determine appropriate levels of effort for adaptation, it is important to ask what level of resilience should be achieved, for which activities or assets, and by when. Especially when considering how to address research opportunities in the area of climate impacts on transportation, it helps to establish what a reasonable balance of investment for resiliency of transportation assets and services is, and what set of goals should pro- vide the context for achieving a resilient system. A base- line can be established in part by first asking the question “When is something resilient enough?” Investing too much in making transportation assets resilient can lead to shortfalls for other societal needs, but underinvest- ing in transportation assets can lead to premature repair costs and increased system disruptions due to service interruptions. Additionally, resiliency must be compared with our current thresholds of acceptance for levels of transportation infrastructure, operational state of repair, and available mobility services. In the face of climate change impacts, do we want to maintain levels similar to today, which may not be desirable to begin with? Or do we want instead to maintain infrastructure in a state of good repair, and is this feasible in the face of climate change? How do we ensure compatibility of adaptation measures with longer-term mitigation plans? With these questions in mind, it is also helpful to identify goals for research that can help set the larger context for resiliency at both the modal and intermodal levels. In this section, we look at the baseline and goals for transportation resiliency and also explore a set of topical areas in which research collaboration is needed to effec- tively address the increasingly complex scenarios being created by climate change impacts on transportation. 5.1 Define Acceptable Levels of Transportation Resilience Having the capacity to understand whether a transpor- tation asset is resilient enough to withstand projected impacts of climate change and extreme events without overinvesting scarce resources is a critical need. If a trans- portation project is overbuilt, it may preclude other, more useful investments in the transportation sector and else- where in society. If it is underbuilt, it may be subject to risks of premature damage or destruction that require premature repair or replacement and impose an addi- tional cost of being out of service to the public and to the industry. What do we need to know to have this capacity? Is it a particular level of skill in climate projections and downscaled models necessary to design corresponding transportation asset attributes? Is it a better understand- ing of how those assets’ attributes can be designed “just well enough” to withstand increasing environmental stresses through time? If we can determine what makes a transportation system “resilient enough,” how do we then develop a process for achieving overall transporta- tion resiliency? Our existing transportation infrastruc- ture, the transportation infrastructure we are building and will build, and the overall performance and dynam- ics of the system of transportation need to have goals for resiliency in order for us to know what research is needed to help identify pathways for achieving desired outcomes. Critical Issues and Research Needs • Cost-effective materials that are robust enough to withstand climate impacts are needed, as well as alterna- tive structural solutions or concepts in addition to mate- rials (e.g., floating elements, extra capacity activated in an emergency situation, deliberate weak points, alterna- tive ways of foundation). • Improved climate and weather models are needed to provide more precise projections of regional impacts for better understanding the level of resiliency needed before building. • Better weather forecast systems will allow the application of early warning systems. • Better communication is needed between climate scientists and transportation researchers on other Earth observation tools that are more closely aligned or better “fit for purpose” with transportation needs. • An investigation of opportunities for cost saving and increased speed (without dilution of quality) would improve construction and rebuilding mechanisms. • Remote sensing applications are needed that provide for more frequent, cost-effective, and detailed monitoring of transportation assets to address problems sooner for longer life of assets. • An understanding of current capacity to map vul- nerability (e.g., identification of critical infrastructure, geographical locations, crucial nodes, specific construc- tions) is required, as well as an understanding of the gap between this capacity and what is needed. In addition,

55A P P E N D I X A : W H I T E P A P E R acceptable risk levels to define intervention needs must be identified. Also needed are models and assessment approaches that are accurate enough and practicable to allow thorough vulnerability assessment, which ideally will include the full effects of climate change, including direct and all indirect economic effects, such as supply chain impacts and economic growth. • Available technologies and operations that can prevent and/or mitigate disruptions caused by weather and other extreme natural events should be assessed, as well as the gap between these technologies and opera- tions and what is needed. 5.2 Implement Risk-Based Asset Management Once organizations responsible for transportation assets achieve a better understanding of what an acceptable level of resiliency means for a transportation asset and the assets and services connected to it in a system, those organizations must have the capacity to integrate this knowledge into their asset management and perfor- mance management programs. Risk-based approaches to managing transportation assets can help make this transition. Broader in scope than traditional transpor- tation asset management and performance management systems, RBTAM is the application of risk management to these systems. If we define risk as “the positive or negative effects of uncertainty or variability upon agency objectives,” then risk management is the “cultures, pro- cesses, and structures that are directed towards the effec- tive management of potential opportunities and threats.” Within the context of this paper, climate impacts are risks that can constrain, or in some cases enhance, a transportation organization’s ability to meet its objec- tives. Risk management is the effective organizational response to those climate risks that results in resiliency. Transportation organizations that implement some form of risk-based approach to asset management will be able to better communicate climate risks to their stakehold- ers and provide a clear understanding of the suite of responses needed to ensure resiliency against those risks. Critical Issues and Research Needs • What transportation organizations have used RBTAM or other risk-based approaches to develop and institutionalize resiliency, and what are some of the com- mon best practices among them? • How can risk-based asset management be insti- tuted in organizations throughout the transportation sector to help achieve consistency and effectiveness in overall transportation resiliency? • In some countries, the implementation of good risk analysis processes will require new skills at many trans- portation agencies. What resources will these agencies need to acquire these skills? • The variability of robustness in different trans- portation agencies’ climate-based risk assessments and vulnerability assessments may well determine the effec- tiveness of risk-based decisions in developing resiliency for transportation assets and services. Can a necessary baseline for the effectiveness of an agency’s risk assess- ment and vulnerability assessment for climate impacts be established in order to have greater confidence in the effectiveness of a subsequent implementation of a risk- based approach to asset management at that agency? • Involving all stakeholders, from users to suppliers of the system as well as cross-modal and cross-sector partners, is essential to the ultimate success of the plan- ning and implementation of resiliency measures. It is to be expected that different stakeholder groups will have different views of what constitutes adequate resilience in the transportation system, and these differences need to be resolved. In addition, cooperative efforts to adapt are expected to be more effective. 5.3 Improve Sense-and-Respond Capabilities Existing transportation infrastructure is owned and operated by various public agencies and private firms and covers an enormous range of ages, service life, and levels of sophistication. Existing infrastructure has been built to many different design standards, and its current and future environmental risks are similarly varied. As environmental risks change, the probability of unex- pected failures may increase. Further, as existing infra- structure approaches the end of its service life, decisions about replacement or abandonment should, but may not currently, take into account changing future risks. Research is needed to better understand how disparate levels of resiliency in existing transportation assets that were not necessarily built with the foresight of climate change impacts can be managed to adapt as well as pos- sible in the decades to come. Critical Issues and Research Needs • What is the state of technologies such as laser imag- ing detection and ranging, or lidar, and remote sensing in terms of their application for monitoring and determin- ing asset integrity? • What is the gap between the state of the art of these technologies and what is needed to more accurately and economically gauge the level of robustness of existing

56 t r a n s p o r t a t i o n r e s i l i e n c e transportation assets relative to their ability to withstand increasing climate change impacts? • What innovative technologies can enhance adapta- tion capacity (e.g., the protection of tunnels, the mainte- nance of bridges, soil stabilization, and drainage)? • How can we improve our measurements and pre- dictions of weather phenomena, especially at the local level? 5.4 Adopt Planning and Engineering for Climate Resilience Newly constructed infrastructure should be designed and built in recognition of the best current understanding of future environmental risks. For this to happen, under- standing of projected climate changes would need to be incorporated into infrastructure planning and design processes across the many public and private builders and operators of transportation infrastructure. A com- prehensive long-term vision should encompass this inte- gration of climate projections with a strategy laid out for revision of existing construction standards and guidelines and definitions of new targeted ones in order to ensure adequate redundancy, accessibility measures, and spatial planning. Additionally, the confluence of new technolo- gies taking place in the sector brings new considerations of how climate change impacts may negatively affect or be mitigated by vehicle connectivity, automated vehicles, electrification of transportation fleets, advanced materi- als, and renewable energy, and energy storage deploy- ment in support of transportation assets. Consideration should also be given to the effects that extreme weather conditions can have on the functionalities and reliability of these new technologies. Finally, challenges for long- term planning and governance include group design processes under uncertainty, cross-jurisdictional collabo- ration, mainstreaming of climate policies, and budgeting of climate change measures. Critical Issues and Research Needs • What fundamental research in nanomateri- als inspired by transportation-specific concerns could lead to dramatic improvements in the “toughness” of materials—or even self-healing attributes—used in new construction? How do new technologies improve or make more challenging conventional plans to adapt to climate change in the transportation sector? • How can spatial planning and governance approaches be adopted to address climate adaptation in a responsive rather than prediction-based manner? Should we factor in competition between mitigation and adaptation funds in climate change policy? • With the increasing ties to the power sector through electrification of the fleet and to the communications sec- tor with the advent of autonomous vehicles, what are the crossover impacts for these three sectors? What research is needed to better understand how more interconnected economic sectors either endure larger impacts or have the ability to combine for greater resistance to impacts? • What innovative multimodal governance methods can support new options in land use and planning, acces- sibility plans, and other variables that can improve the capacity for functional redundancy in the face of climate change? For example, if switching to alternative transpor- tation modes is part of the contingency plan, the vulnerabil- ity of the intermodal hubs becomes even more important. 5.5 Address System Resilience Transportation systems are more than just the sum of their individual parts. Some elements are of particular importance because of their vital economic role, absence of alternatives, heavy use, or critical function. Transpor- tation systems are potentially vulnerable to the loss of key elements. Therefore, selectively adding redundant infrastructure may be a more efficient strategy than hard- ening many individual facilities on the existing system. In addition, smart solutions on the user’s side, such as buf- fer inventories or excess vehicle capacity, may obviate the need for expensive infrastructural measures. System resilience is best viewed across transportation modes and multiple system owners, building on component resil- ience to prevent system failure (Figure 10) (9). Lately, new holistic engineering approaches have been devel- oped for climate adaptation adoption of sociotechnical systems (14) or systems-of-systems perspectives (36). Cooperation among stakeholders is indispensable to allow integrated, complementary, and mutually support- ive actions in the sector and outside it. Although some key elements are obvious, other dependencies may be less well recognized. For example, some airports rely on petroleum pipelines, which may depend, in turn, on electric power for pumping. Transportation systems are also interdependent when passengers or freight carriers rely on multiple transportation modes to reach their des- tinations. For these reasons, research is needed to better understand the collective potential set of ripple effects that may be induced by climate change impacts on sys- tem dynamics within the transportation sector. Critical Issues and Research Needs • Disruptions in waterborne shipping due to sea level rise, major flooding, or extended drought may force goods delivery to less efficient rail services, and even less

57A P P E N D I X A : W H I T E P A P E R efficient than that, truck service. Both rail and truck ser- vice might struggle to absorb additional volumes given up by water-borne shipping. A similar scenario could occur with rail goods shifting to truck–freight service due to climate change–related problems like rail buck- ling and cracking. • What areas of research can address lock- and port- related technologies to better prepare waterborne ship- ping for climate change impacts? • Although research is ongoing for predictive capa- bilities in rail buckling and cracking, what materials research might find solutions to prevent the frequency of these occurrences in the first place? • What algorithm and modeling applications can better understand ripple effects that spread through the transportation system when a critical transportation node (or nodes) is increasingly impaired by a changing environment? • How can we assess the overall systems impact of increasing environmental stressors (direct and indirect) on critical transportation nodes at the aggregate level? • What research can help identify critical nodes in transportation that are not obvious to us today? • What indicators assess the efficiency of adaptation measures (e.g., prevention of disruption in vulnerable zones, reduction of downtime, efficiency of rerouting)? 5.6 Assess Societal Impacts Several areas in society may be affected by climate change, adaptation to climate change, and resilience efforts. Older populations and those with disabilities are particularly sensitive to the availability of mobility services. Communication and information, emergency response, and evacuation management are important to ensure efficient use of transport in case of major disrup- tions of service. The broader impacts of disturbances in road transport can escalate quickly. The disruption of a country’s road freight system can paralyze its economy and social welfare system in as little as 4 to 5 days (37). Critical Issues and Research Needs • Is there a potential for impacts on transit, demand- response services, pedestrians and sidewalk quality, protection from the elements, or simply accessibility to transportation assets? • Can automation and improved logistics serve in mitigating potential negative consequences in this area? More responsive supply chain organization, early warn- ing systems, and adaptive planning systems could, indi- vidually and cooperatively, create built-in automated resilience. • What will be the general impact on logistics in terms of production sourcing, inventory levels, storage location, and delivery routing and scheduling? • How will the relocation of tourist destinations affect the transport system? What will be the socioeco- nomic impact of relocation of settlements for adaptation purposes? • Lastly, is there a social equity issue in climate change, adaptation, and resilience efforts? How do we define the FIGURE 10 Conceptual framework of impact propagation (9). Measures Measures Measures Governance Loads, eventsClimate change e.g., extreme rainfall event e.g., more- frequent extreme rainfall events In fr as tr uc tu re 2 In fr as tr uc tu re n ENVIRONMENT SOCIAL SYSTEM NetworksLinks, nodesComponents e.g., network gridlock e.g., congested transport corridor e.g., tunnel flooding In fr as tr uc tu re 1 TECHNICAL SYSTEM

58 t r a n s p o r t a t i o n r e s i l i e n c e issues and the collective set of objectives to address social equity in the face of climate change impacts? 6 synthesis and ConCluding remarKs This paper attempts to lay out the broad issues confront- ing the transportation community as it addresses the need to build greater resilience into the transportation infra- structure and its users to meet the increasing demands and threats of a changing climate and more extreme weather. Developing sound resilience measures begins with an understanding of the basic science of climate change. Sim- ply put, there are five major changes of import to trans- portation and much of the built infrastructure: • Sea level rise, sometimes exacerbated by land subsidence; • Higher temperatures and longer heat waves; • Changes in precipitation patterns leading to both droughts and stronger rain- and snowstorms; • Rising Arctic temperatures resulting in melting per- mafrost and reduced ice sheets; and • Increased intensity of hurricanes: not more fre- quent but stronger storms. But the devil is in the details. When? Where? How often? These are the questions that designers, builders, and operators of transportation systems need to have answered. Today, much of the information on climate change is at the global or continental scale, but transpor- tation professionals need information at the regional or, ideally, the local scale. The science of climate change, as with other areas of science, is one lacking stationarity. Historical records are of limited value in predicting the future. One can model changes in the climate, but even the input to the models is uncertain. How aggressively will the nations of the world address fossil fuel consumption? What new technologies will evolve to reduce the need for fossil fuels or to cap- ture and sequester CO2? What is the impact of naturally occurring events such as El Niño and volcanic eruptions? How will the ecosystem itself react to increased levels of GHGs? And how resilient are the socioeconomic systems that depend on transportation? There are techniques for dealing with problems of deep uncertainty, but by and large they have not been used in the engineering profession. These techniques, such as robust decision making or adaptive planning, must become ingrained in the decision process for cli- mate change impacts. Equally important and essential is the development of sound risk management method- ologies. RBTAM is but one of several approaches that incorporate probability analysis with decision matrices to reach optimum decisions on the development and operation of transportation and other infrastructure systems. References to several of these approaches are provided in the text. This paper also examines the essential question raised by the conference organizers: How do we achieve bet- ter resilience to climate change in the transportation industry? We attempt to identify the critical issues that must be addressed through needed research. What is an acceptable level of transportation resilience? Can we develop more precise climate and extreme weather mod- els and forecasts? Are there new structures or materials, including nanoproducts, that can better withstand heat, drought, or flooding? The issue of developing sound risk management approaches to managing the entirety of transportation assets is discussed. Monitoring systems, including lidar, must be improved to measure the robustness of assets and the responsiveness of those assets to extreme events. The transportation system itself is undergoing signifi- cant change with new technologies such as automated vehicles, electrification of transportation fleets, and new materials. How will climate change enhance or hinder the implementation of these new developments? One cannot consider climate change just in relationship to today’s systems, but how it will affect the systems of the future. Transportation is not simply a set of individual parts, but rather an amalgamation of many components linked together by nodes and forming a giant network. It oper- ates with a multiplicity of interdependencies with other critical infrastructure elements (e.g., energy, water and wastewater, IT and communications, buildings, and health care). Furthermore, transportation is essential to agriculture, forestry, and, most importantly, to our social and economic systems. The task for this symposium is to begin to prioritize the major research needs of the transportation industry in relation to creating stronger resilience in transporta- tion systems. Perhaps no more important research need exists than to better understand the interdependencies that exist between transportation modes and between transportation and all the sectors with which it inter- faces. This paper suggests other critical issues that must be addressed, most of which require more research. The main areas of interest for researchers and practitioners appear to be the following: 1. Defining acceptable levels of transportation resil- ience and acquiring practice in the design of measures at an appropriate scale; 2. Implementation of risk-based transportation asset management, including improved information provision as well as design methodological and institutional aspects; 3. Improvement of sense-and-respond capabilities to allow adaptive policies to be formulated and imple- mented;

59A P P E N D I X A : W H I T E P A P E R 4. Adoption of planning approaches for climate resil- ience, recognizing that spatial planning and governance practices need to be reviewed in the light of climate change; 5. Addressing system resilience by developing cross- modal, cross-sectoral, and cross-infrastructure asset- management approaches; and 6. Assessment of societal impacts, with the aim of ensuring that top-down adaptation initiatives and bottom-up absorptive capacities and resources are in balance. aCKnowledgments We gratefully acknowledge contributions to this paper from the planning committee for the symposium. These contributions included the symposium concept paper, which formed the basis for Section 5, and comments to the draft version of this white paper. The production of the paper was supported by the Transportation Research Board and the European Commission. referenCes Abbreviations IPCC Intergovernmental Panel on Climate Change 1. National Research Council. Disaster Resilience: A National Imperative. National Academies Press, Wash- ington, D.C., 2012. doi:10.17226/13457. 2. Anagnostou, E., E. H. John, K. M. Edgar, G. L. Foster, A. Ridgwell, G. N. Inglis, R. D. Pancost, D. J. Lunt, and P. N. Pearson. Changing Atmospheric CO2 Concentration Was the Primary Driver of Early Cenozoic Climate. Nature, Letter, 2016. doi: 10.1038/nature17423. 3. U.S. National Oceanic and Atmospheric Administration. Earth Systems Research Laboratory, Global Monitoring Division, 2014. 4. Climate Change Evidence and Causes: An Overview from the Royal Society and the US National Academy of Sciences. National Academy of Sciences and The Royal Society, 2014. http://nas-sites.org/americasclimatechoices/ events/a-discussion-on-climate-change-evidence-and- causes/. 5. U.S. Global Change Research Program. Climate Change Impacts in the United States. U.S. National Climate Assessment, 2014. http://nca2014.globalchange.gov. 6. U.S. National Oceanic and Atmospheric Administration. Global Analysis, Annual 2015. National Centers for Envi- ronmental Information, 2015. 7. IPCC. Climate Change 2014: Synthesis Report. Contribu- tion of Working Groups I, II, and III to the Fifth Assess- ment Report of the Intergovernmental Panel on Climate Change (Core Writing Team, R. K. Pachauri, and L. A. Meyer, eds.), IPCC, Geneva, Switzerland, 2014. 8. Special Report 290: Potential Impacts of Climate Change on U.S. Transportation. Transportation Research Board of the National Academies, Washington, D.C., 2008. 9. Bollinger, L. A., C. W. J. Bogmans, E. J. L. Chappin, G. P. J. Dijkema, J. N. Huibregtse, N. Maas, T. Schenk, M. Snelder, P. van Thienen, S. de Wit, B. Wols, and L. A. Tavasszy. Climate Adaptation of Interconnected Infra- structures: A Framework for Supporting Governance. Regional Environmental Change, Vol. 14, No. 3, 2014. 10. National Academies of Sciences, Engineering, and Medi- cine. Attribution of Extreme Weather Events in the Context of Climate Change. National Academies Press, Washington, D.C., 2016. doi:10.17226/21852. 11. IPCC. Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assess- ment Report of the Intergovernmental Panel on Climate Change (C. B. Field, V. R. Barros, D. J. Dokken, K. J. Mach, M. D. Mastrandrea, T. E. Bilir, M. Chatterjee, K. L. Ebi, Y. O. Estrada, R. C. Genova, B. Girma, E. S. Kissel, A. N. Levy, S. MacCracken, P. R. Mastrandrea, and L. L. White, eds.), Cambridge University Press, Cam- bridge, United Kingdom and New York, USA, 2014. 12. IPCC. Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part B: Regional Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (V. R. Barros, C. B. Field, D. J. Dokken, M. D. Mastrandrea, K. J. Mach, T. E. Bilir, M. Chatterjee, K. L. Ebi, Y. O. Estrada, R. C. Genova, B. Girma, E. S. Kissel, A. N. Levy, S. MacCracken, P. R. Mastrandrea, and L. L. White, eds.), Cambridge University Press, Cambridge, United Kingdom and New York, USA, 2014. 13. IPCC. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assess- ment Report of the Intergovernmental Panel on Climate Change (T. F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P. M. Midgley, eds.), Cambridge University Press, Cam- bridge, United Kingdom and New York, USA, 2013. 14. Chappin, E. J., and T. van der Lei. Adaptation of Inter- connected Infrastructures to Climate Change: A Socio- Technical Systems Perspective. Utilities Policy, Vol. 31, 2014, pp. 10–17. 15. Chelleri, L., J. Waters, M. Olazabal, and G. Minucci. Resilience Trade-Offs: Addressing Multiple Scales and Temporal Aspects of Urban Resilience. Environment and Urbanization, Vol. 27, No. 1, 2015, pp. 181–198. 16. Stern, N. H. Stern Review: The Economics of Climate Change, Vol. 30. HM Treasury, London, 2006. 17. Refsgaard, J. C., K. Arnbjerg-Nielsen, M. Drews, K. Hal- snæs, E. Jeppesen, H. Madsen, and J. H. Christensen.

60 t r a n s p o r t a t i o n r e s i l i e n c e The Role of Uncertainty in Climate Change Adaptation Strategies—A Danish Water Management Example. Miti- gation and Adaptation Strategies for Global Change, Vol. 18. No. 3, 2013, pp. 337–359. 18. Dessai, S., and M. Hulme. Does Climate Adaptation Policy Need Probabilities? Climate Policy, Vol. 4, No. 2, 2004, pp. 107–128. 19. Clark, M. P., and R. S. Pulwarty. Devising Resilient Responses to Potential Climate Change Impacts. Ogmius: Newsletter of the Center for Science and Technology Pol- icy Research, Vol. 5, 2003, pp. 2–3. 20. Schenk, T. E. W. Institutionalizing Uncertainty: Exploring How Infrastructure Stakeholders Can Prepare for Uncer- tain Climate Futures. Doctoral dissertation. Massachu- setts Institute of Technology, Cambridge, 2015. 21. Walker, W. E., R. J. Lempert, and J. H. Kwakkel. Deep Uncertainty. In Encyclopedia of Operations Research and Management Science, Springer, 2016, pp. 395–402. 22. Van Drunen, M., A. Leusink, and R. Lasage. Towards a Climate-Proof Netherlands. In Water Management in 2020 and Beyond, Springer, Berlin and Heidelberg, Ger- many, 2009, pp. 165–194. 23. Walker, W. E., S. A. Rahman, and J. Cave. Adaptive Poli- cies, Policy Analysis, and Policy-Making. European Jour- nal of Operational Research, Vol. 128, No. 2, 2001, pp. 282–289. 24. Lempert, R., N. Nakicenovic, D. Sarewitz, and M. Schlesinger. Characterizing Climate-Change Uncertain- ties for Decision-Makers: An Editorial Essay. Climatic Change, Vol. 65, No. 1, 2004, pp. 1–9. 25. U.S. Department of Transportation. Risk-Based Trans- portation Asset Management: Building Resilience into Transportation Assets. Report 5: Managing External Threats Through Risk-Based Asset Management. FHWA, Washington, D.C., 2013. 26. Olsen, J. R., ed. Adapting Infrastructure and Civil Engi- neering Practice to a Changing Climate. American Society of Civil Engineers, Reston, Va., 2015. 27. Royal Academy of Engineering. Infrastructure, Engineer- ing and Climate Change Adaptation. Royal Academy of Engineering, London, 2011. 28. European Environment Agency. Adaptation of Transport to Climate Change in Europe. European Environment Agency, Copenhagen, Denmark, 2014. 29. United Nations Economic Commission for Europe. Cli- mate Change Impacts and Adaptation for International Transport Networks. United Nations Economic Commis- sion for Europe, Geneva and New York, 2013. 30. Nemry, F., and H. Demirel. Impacts of Climate Change on Transport: A Focus on Road and Rail Transport Infra- structures. Joint Research Center, European Commission, Brussels, 2012. 31. Bles, T., Y. Ennesser, J. Fadeuilhe, S. Falemo, B. Lind, M. Mens, and F. Sandersen. Risk Management for Roads in a Changing Climate: A Guidebook to the RIMAROCC Method. ERA-NET ROAD Program, 2010. http://www .cedr.fr/, 2010. Accessed May 21, 2016. 32. Huibregtse, E., N. Morales, L. Hellebrandt, D. Paprotny, and S. De Wit. Climate Change in Asset Management of Infrastructure: A Risk-Based Methodology Applied to Disruption of Traffic on Road Networks Due to the Flooding of Tunnels. European Journal of Transport and Infrastructure Research, Vol. 16, No. 1, 2016. 33. IPCC. Summary for Policymakers. In Climate Change 2007: Impacts, Adaptation and Vulnerability (M. L. Parry, O. F. Canziani, J. P. Palutikof, P. J. van der Linden, and C. E. Hanson, eds.), Contribution of Working Group II to the Fourth Assessment Report of the IPCC, Cam- bridge University Press, 2007, pp. 7–22. 34. Berrang-Ford, L., J. Ford, and J. Paterson. Are We Adapt- ing to Climate Change? Global Environmental Change, Vol. 21, No. 1, 2011, pp. 25–33. 35. Lesnikowski, A. C., J. D. Ford, L. Berrang-Ford, M. Bar- rera, and J. Heymann. How Are We Adapting to Climate Change? A Global Assessment. Mitigation and Adapta- tion Strategies for Global Change, Vol. 20, No. 2, 2015, pp. 277–293. 36. Hall, J. W., M. Tran, A. J. Hickford, and R. J. Nicholls (eds.). The Future of National Infrastructure: A System- of-Systems Approach. Cambridge University Press, 2016. 37. McKinnon, A. Life Without Trucks: The Impact of a Temporary Disruption of Road Freight Transport on a National Economy. Journal of Business Logistics, Vol. 27, No. 2, 2006, pp. 227–250. additional resourCes National Academies. America’s Climate Choices: Advancing the Science of Climate Change. National Academies Press, Washington, D.C., 2010. U.S. National Oceanic and Atmospheric Administration. Global Summary Report. National Centers for Environ- mental Information, 2016.