CHAPTER 4


The Way Forward

The phrase “Here be dragons” appears on one of the first known globes produced in Europe following Christopher Columbus’s voyage across the Atlantic Ocean (Figure 4.1). That globe shows some of the vast uncertainties that Columbus’s exploration highlighted for Europe—indeed, North America is portrayed as just a few islands (e.g., De Costa, 1879). The phrase may have been referring to Komodo dragons, but it could easily have evoked the dragons that we now associate with fairy tales in the imagination of Columbus’s contemporaries. The initial discovery of the “New World” revealed unsuspected possibilities (“unknown unknowns”), effectively increasing uncertainty. Subsequent exploration has long since filled in those uncertainties, and if real dragons had lived between those islands, they would have been found.

By analogy, many investigations into the climate and Earth system have revealed possible dangers. Some of those have been confirmed or even amplified, such as the impacts of chlorofluorocarbons on the ozone layer that were understood in the 1970s before the Antarctic Ozone Hole was discovered the following decade (see Box 1.1). Other possible dangers, such as the sudden release of methane from ocean sediments, have been greatly reduced by subsequent research that showed a clear lack of dragons.

In looking back at the previous studies of abrupt climate change in the Introduction to this report, the committee notes that even when dragons, i.e., possible threats, are identified and clearly pointed out, they may then be ignored and their presence not acted upon. This is not an unusual situation, and ignoring early warnings is a welldocumented phenomenon in environmental research (e.g., EEA, 2001). In this chapter we briefly examine some of the major lessons learned in Chapters 2 and 3 and then propose one possible way forward, namely an Abrupt Change Early Warning System (ACEWS).

WHAT HAS BEEN LEARNED?

Paleoclimatologists have long known that the slower changes of the ice age were punctuated by relatively large events with approximately millennial spacing and sharp onsets and terminations. Some of these were given names, such as the Younger Dryas. A range of studies showed that these events represented changes in wintertime sea



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 147
CHAPTER 4 The Way Forward T he phrase “Here be dragons” appears on one of the first known globes produced in Europe following Christopher Columbus’s voyage across the Atlantic Ocean (Figure 4.1). That globe shows some of the vast uncertainties that Columbus’s exploration highlighted for Europe—indeed, North America is portrayed as just a few islands (e.g., De Costa, 1879). The phrase may have been referring to Komodo dragons, but it could easily have evoked the dragons that we now associate with fairy tales in the imagination of Columbus’s contemporaries. The initial discovery of the “New World” revealed unsuspected possibilities (“unknown unknowns”), effectively increas- ing uncertainty. Subsequent exploration has long since filled in those uncertainties, and if real dragons had lived between those islands, they would have been found. By analogy, many investigations into the climate and Earth system have revealed possible dangers. Some of those have been confirmed or even amplified, such as the impacts of chlorofluorocarbons on the ozone layer that were understood in the 1970s before the Antarctic Ozone Hole was discovered the following decade (see Box 1.1). Other possible dangers, such as the sudden release of methane from ocean sedi- ments, have been greatly reduced by subsequent research that showed a clear lack of dragons. In looking back at the previous studies of abrupt climate change in the Introduction to this report, the committee notes that even when dragons, i.e., possible threats, are identified and clearly pointed out, they may then be ignored and their presence not acted upon. This is not an unusual situation, and ignoring early warnings is a well- documented phenomenon in environmental research (e.g., EEA, 2001). In this chapter we briefly examine some of the major lessons learned in Chapters 2 and 3 and then propose one possible way forward, namely an Abrupt Change Early Warning System (ACEWS). WHAT HAS BEEN LEARNED? Paleoclimatologists have long known that the slower changes of the ice age were punctuated by relatively large events with approximately millennial spacing and sharp onsets and terminations. Some of these were given names, such as the Younger Dryas. A range of studies showed that these events represented changes in wintertime sea 147

OCR for page 147
abrupt impacts of climate chang E FIGURE 4.1 The Hunt-Lenox Globe. Wording near China says “hic sunt dracones,” which translates to “here be dragons.” This is a metaphor for unknown threats. Often, as humans have explored more of the world, threats have become less—in this case, the “dragons” may have referred to Komodo dragons, rather than dragons of fairy tales. ice coverage in the North Atlantic, that they had a near-global footprint but regionally distinct impacts, and that even slow changes in freshwater flux to the North Atlantic could cross a threshold and trigger a sudden event (e.g., Alley, 2007). The very large magnitude of the changes in some regions, the wholesale and rapid reorganization of ecosystems, and the very rapid rates of change that affected certain places (10°C over 10 years in Greenland), together with the realization that greenhouse warming would cause significant changes in freshwater fluxes to the North Atlantic, caused concerns that extended beyond the scientific community into popular culture (e.g., the 2004 science-fiction/disaster movie The Day After Tomorrow). As discussed in Chapter 2, subsequent research has shown that an abrupt disruption of the Atlantic Meridional Overturning Circulation (AMOC) is much less likely under modern boundary conditions than during the ice age, and that the regional cool- ing impact of the shutdown in heat transport to the high northern latitudes would 148

OCR for page 147
The Way Forward probably be smaller than the warming from rising greenhouse gas concentrations. In sum, while the North Atlantic remains a probable site of strong climate variability and even slow changes in the circulation of this basin could result in significant local and regional impacts, recent research has shown that the earlier worries of a total AMOC collapse are unwarranted (IPCC, 2007c; USCCSP, 2008; and Chapter 2). Thus in this case, improved knowledge has allayed some fears of the worst types of outcomes occurring this century from this possible abrupt climate change (Box 4.1). In a similar manner, rapid or catastrophic methane release from sea-floor or perma- frost reservoirs has also been shown to be much less worrisome than first considered possible (see section in Chapter 2 on High-Latitude Methane and Carbon Cycles). The discovery of vast methane deposits, clearly vulnerable to warming, motivated discussion of a potential primary role for methane in Earth’s climate. Fast changes in atmospheric methane concentration in ice cores from glacial time correlated with abrupt climate changes (e.g., Chappellaz et al., 1993). However, subsequent research has revealed that the variations in methane through the glacial cycles (1) originated in large part from low-latitude wetlands, and were not dominated by high-latitude sources that could be potentially much larger, and (2) produced a relatively small radiative forcing relative to the temperature changes, serving as a small feedback to climate changes rather than a primary driver.1 Looking to the future, the available source reservoirs for atmospheric methane release—from both methane hydrates and permafrost—are expected to respond to climate on a time scale slow enough that the climate impact from the methane will probably be smaller than that from rising CO2 concentrations. Nonetheless, there is still much to explore. For example a cause for concern is that wildfires have been spreading into some permafrost regions as local climatic conditions promote increasingly dry conditions (Yoshikawa et al., 2002). Little is known about the potential of such burning to thaw and release stored carbon faster than expected. This possible mechanism of rapid, unexpected carbon release merits research to evaluate its efficacy and climatic impacts. But despite the comfort one might take from knowing that these examples are not the dragons they were once thought to be, “dragons” in the climate system still may exist. 1  Methane was also proposed as the origin of the Paleocene–Eocene thermal maximum event, 55 million years ago, in which carbon isotopic compositions of CaCO3 shells in deep sea sediments reflect the release of some isotopically light carbon source (like methane or organic carbon), and various temperature proxies indicate warming of the deep ocean and hence the Earth’s surface. But the longevity of the warm period has shown that CO2 was the dominant active greenhouse gas, even if methane was one of the im- portant sources of this CO2 , and the carbon isotope spike shows that if the primary release reservoir were methane, the amount of CO2 that would be produced by this spike would be insufficient to explain the extent of warming, unless the climate sensitivity of Earth was much higher than it is today (Pagani et al., 2006). 149

OCR for page 147
abrupt impacts of climate chang E BOX 4.1  ASYMMETRICAL UNCERTAINTIES The current understanding of many aspects of the climate system and its influence on ecosystems and economies projects a most-likely response to changing CO2 but a skewed dis- tribution of uncertainties, such that the outcome may be a little “better” (smaller changes, lower costs), a little “worse” (larger changes, higher costs), or a lot worse than the most-likely case, but with little chance of being a lot better (see Figure). Skewed distributions, and those with “fat tails” that allow finite chances of very large changes, are known or have been suggested in many contexts (e.g., Mandelbrot, 1963; Weitzman, 2011). Vigorous research continues on the effects of such issues on optimal policy paths (e.g., Keller et al., 2004; Keller et al., 2008; McInerney and Keller, 2008; Weitzman, 2009; Tol, 2003). As described above, research can in some cases lead to an improved understanding of risks where the highest impact outcomes are discovered to be less likely than originally feared, i.e., in some cases research can remove these “fat tails” (see Figure). The value of reducing the deep uncertainty associated with fat tails or abrupt climate chang- es is clear, motivating research to stimulate learning. A large possibility of an AMOC “shutdown” , for example, would have notably increased the optimal response to reducing climate change, so the reduction of the estimated chance of such an event provided by recent research has large economic consequences (Keller et al., 2004). Many of the potential abrupt climate changes listed in Table 4.1 are listed as having a “mod- erate” or “low” probability/likelihood of occurring within this century. Recent research on expert elicitations (Kriegler et al., 2009) related to the probability of abrupt climate changes is in general agreement with the assessments provided in Table 4.1. For the five tipping point examples that they examined, they found “significant lower probability bounds for triggering major changes in the climate system.” The collected understanding of these threats is summarized in Table 4.1. For example, the West Antarctic Ice Sheet (WAIS) is a known unknown, with at least some potential to shed ice at a rate that would in turn raise sea level at a pace that is several times faster than is happening today. If WAIS were to rapidly disintegrate, it would challenge adaptation plans, impact investments into coastal infrastructure, and make rising sea level a much larger problem than it already is now. Other unknowns include the rapid loss of Arctic sea ice and the potential impacts on Northern Hemisphere weather and climate that could potentially come from that shift in the global balance of energy, the widespread extinction of species in marine and terrestrial systems, and the increase in the frequency and intensity of extreme precipitation events and heat waves. The com- mittee reviews the various abrupt changes described in Chapter 2 in Table 4.1. Looking beyond the physical climate system, many of the research frontiers high- lighted in this report focus on the potential for gradual physical changes to trigger 150

OCR for page 147
The Way Forward Mean Mean severity severity “A little “A little Likelihood of occurrence Likelihood of occurrence better” worse” “A lot worse” “fat tail” Severity of impacts Severity of impacts FIGURE The graph on the left represents the skewed distribution of uncertainties with a “fat tail.” The mean likelihood of occurrence at the level of severity anticipated is represented by a dotted line. The area to the left of the mean represents the likelihood for impacts less severe (the “a little better” case), while the area to the right shows the greater likelihood for extreme impacts (spanning “a little worse” to “a lot worse” cases). The graph on the right compares the normal distribution (black line) to the “fat tail” distribution (blue line). For some changes, more research has shown that the distribution of possible outcomes includes less likeli- hood of the most severe outcomes. abrupt ecological, economic, or social changes (see Chapter 3). In these cases, reach- ing a certain threshold of change in various climatic parameters can trigger abrupt, irreversible changes in the affected ecological, economic, or social system, even if the trajectory of climate change itself is gradual. There is still much to learn about the po- tential for and possible prediction of these kinds of abrupt changes, but a sound body of theory and empirical data (Barnosky et al., 2012; Carpenter et al., 2011; Hastings and Wysham, 2010; Mumby et al., 2007; Scheffer, 2010; Scheffer et al., 2001, 2009, 2012b) confirm that there are real “dragons” out there to be discovered. In the ecological arena, some of these metaphorical dragons are already becoming evident—for instance, gradual ocean acidification leading to the shutdown of devel- opmental pathways in ecologically important marine species, or gradual warming of ocean waters to exceed temperature thresholds that result in death of coral reefs (see sections on Changes in Ocean Chemistry and Extinctions in Chapter 2). Both kinds 151

OCR for page 147
abrupt impacts of climate chang E of “threshold effects” have been confirmed experimentally (ocean acidification) or from observational data accumulated over the past few decades (coral bleaching and death). A major area of uncertainty, however, is the potential for threshold-induced impacts on one species to cascade to impact others, and the pathways by which such cascades could lead to wholesale ecosystem collapse (see section on Ecosystem Collapse and Rapid State Changes in Chapter 2). For example, while it is clear that species are chang- ing their geographic distributions and seasonal cycles individualistically in response to climate change, resulting in pulling apart of species that have co-existed in the same place at the same time of the year, the magnitude and importance of longer-term ecological changes that will result are still unclear. Answering such species-interaction and ecological-network questions holds the key to assessing the likelihood of figuring out the ‘fat-tail’ probabilities of ecosystem collapse induced by climate change, which in turn would impact economic and social systems (for instance, through loss of fisher- ies, forests, or agricultural productivity). Anticipating the potential for climatically-induced abrupt change in social systems is even more difficult, given that social systems are actually extremely complex sys- tems, the dynamics of which are governed by a network of interactions between people, technology, the environment, and climate. The sheer complexity of such systems makes it difficult to predict how changes in any single part of the network will affect the overall system, but theory indicates that changes in highly-connected nodes of the system have the most potential to propagate and cause abrupt down- stream changes. Climate connects to social stability through a wide variety of nodes, including availability of food and water, transportation (for instance, opening Arctic seaways), economics (insurance costs related to extreme weather events or rising sea level, agricultural markets, energy production), ecosystem services (pollination, fisher- ies), and human health (spread of disease vectors, increasing frequency of abnormally hot days that cause physiological stress). Reaching a climatic threshold that causes rapid change in any one of these arenas therefore has high potential to trigger rapid changes throughout the system. For example, at the time Arctic shipping routes become routinely passable, world trade routes and the related economic and political realities could change dramatically within a single decade. Much remains unknown about how climate maps onto the complex networks that de- fine social systems at local, regional, and global scales. Using network modeling tech- niques to identify the nodes and connections that construct social systems at a variety of spatial scales, and how projected climate changes would be expected to propagate through the system, may well lead to better predictive ability. A more empirical ap- 152

OCR for page 147
The Way Forward proach, already feasible with existing modeling techniques, is to identify communities that are geographically situated where small shifts in global climatic patterns—such as in position of the jet stream—would have large impacts at the local scale, such as an “on-off” switch for local drought. Identifying those geographic areas and assess- ing how the local impacts might propagate spatially through the regional and global social network may provide a viable means of anticipating regional vulnerabilities, and which of those vulnerable regions have high potential of influencing global dynamics. “Stress tests,” or scenario based modeling exercises have been recommended as a way to reveal vulnerabilities and likely effects of disruptive climate events on particular countries, populations, or systems; stress tests provide “a framework for integrating climate and social variables more systematically and consistently within national secu- rity analysis” (NRC, 2012b). ANTICIPATING SURPRISES The recognition of the importance of tipping-point behavior in physical, biological, and social systems has prompted a growing body of research to provide as much early warning as possible of incipient or ongoing abrupt changes (Box 4.2). Theory and experiment agree that some systems approaching tipping points exhibit signs of the impending change. This behavior may include a flickering behavior, in which a system jumps back and forth between two states, or a shift to slower recovery from small perturbations (e.g., Taylor et al., 1993; Drake and Griffen, 2010; Carpenter et al., 2011; Veraart et al., 2012; Wang et al., 2012a), presaging future failure to recover. However, as emphasized by Boettinger and Hastings (2013), there probably is not a generic signal of an impending shift, with different signals in different systems, and there is very real possibility that no warning signal will be evident. Considering this, plus the various challenges to interpreting signals that do occur, the goal of successfully predicting tip- ping points, and providing policy-relevant choices on how to avoid them or deal with the consequences, is likely to be realized in some cases but is unlikely to be universally possible. Nonetheless, identifying potential vulnerabilities is valuable. In some cases science may be able to provide accurate information that a tipping point is imminent, creat- ing time for adaptation or even possibly mitigation. Science can also identify when tipping is occurring, or has recently occurred. Such knowledge could greatly reduce societal damages. And, in light of the potentially very large costs of some possible abrupt changes, additional research to improve this knowledge can have large eco- nomic benefits (e.g., Keller et al., 2007; McInerney et al., 2012). 153

OCR for page 147
abrupt impacts of climate chang E BOX 4.2  EARLY WARNING SIGNALS OF ABRUPT TRANSITIONS Examining the spectral properties of time series data can, in theory, predict the onset of some abrupt changes. There is a burgeoning literature examining early warning signals in systems as they approach tipping points (e.g., Boettiger and Hastings, 2012a, b; Dakos et al., 2008; Ditlevsen and Johnsen, 2010; Lenton et al., 2008; Scheffer et al., 2009; Scheffer et al., 2012a). Several ex- amples of early warning signs are described below: 1.  some systems, the return time from a small perturbation increases as the system ap- In proaches a critical threshold. This “critical slowing down” leads to an increase in autocor- relation in the pattern of variability, which can serve as an indicator of impending abrupt change (Dakos et al., 2008; Scheffer et al., 2009). 2.  There may also be “flickering” as stochastic forcings move the system back and forth across a threshold to sample two alternative regimes; some data suggest that past climatic shifts may have been preceded by such flickering behavior (Scheffer et al., 2009). 3.  is also possible for the macrostructure of systems to indicate proximity to a transition It point (e.g., spatial patterns of vegetation in a landscape as it transitions from patchy to barren; Scheffer et al., 2009). Research in this field is extending to examine highly con- nected networks, where connectivity and heterogeneity patterns may be used to antici- pate state changes (Scheffer et al., 2012a). Significant challenges exist in implementing early warnings to anticipate tipping points. Ac- curate forewarnings that avoid significant false alarms involve tradeoffs between specificity and sensitivity (Boettiger and Hastings, 2012a, b). Additionally, successful detection of an impending change does not imply that an effective intervention is possible. Finally, there is some evidence that previous abrupt changes in Earth’s history have been noise-induced transitions (see Box 1.4); such events will have very limited predictability (Ditlevsen and Johnsen, 2010). In general, be- cause many systems have the property of being sensitive to initial conditions (i.e., the “butterfly effect”), forecasting future behavior is only feasible on short timescales for those systems, and timely early warning signs for abrupt changes may not be possible. An Abrupt-Change Early-Warning System In light of the potential great importance and value of accurate anticipation of the occurrence and impacts of abrupt changes, the committee recommends development of an abrupt change early warning system (ACEWS). An ACEWS would be part of an overall risk management strategy, providing required information for hazard identifi- cation and risk assessment (Figure 4.2). This information would then inform an overall risk strategy, which uses the risk assessment to prioritize hazard mitigation options and their implementation (Basher, 2006). A central part of an effective risk manage- ment framework is continued evaluation of the efficacy of the risk management strat- 154

OCR for page 147
The Way Forward FIGURE 4.2 Continuous and reinforcing process of disaster risk management as a foundation for building resilience. Central to the risk management process is the collective evaluation by the partners regarding goals, values, and objectives for the risk management strategy and for community resilience. The entire process, divided for convenience of discussion into six steps, encompasses the ability to identify and as- sess the local hazards and risks (steps 1 and 2), to make decisions as to which strategies or plans are most effective to address those hazards and risks and implement them (steps 3 and 4), and to review and evalu- ate the risk management plan and relevant risk policies (steps 5 and 6). An Abrupt Change Early Warning System (ACEWS) would be part of such an overall risk management strategy, providing required informa- tion for hazard identification and risk assessment (adapted from NRC, 2012c). 155

OCR for page 147
abrupt impacts of climate chang E egy and adjustment based on new information about the hazard and vulnerabilities (NRC, 2012c). Given the number of critical Earth system components that might cross tipping points and the potentially data intensive monitoring and modeling needs, some strategic decisions will have to be made with regard to which hazards to moni- tor and for which to attempt to develop warning systems. For example, if the lead time necessary to prepare the appropriate social response is very long, the detection of the tipping point might not occur early enough to avoid major impact on the socio- economic system. Some of these decisions are based on what risks are socially accept- able compared to the cost of the hazard mitigation effort and involve value judgment by the affected people or their political representatives (Plattner, 2005). In general, an ACEWS system would (1) identify and quantify social and natural vul- nerabilities and ensure long-term, stable observations of key environmental and economic parameters through enhanced and targeted monitoring, (2) integrate new knowledge into numerical models for enhanced understanding and predictive capa- bility, and (3) synthesize new learning and advance the understanding of the Earth system, taking advantage of collaborations and new analysis tools. These aspects are discussed below, followed by a discussion of the some special considerations for de- signing and implementing an ACEWS. The development of an ACEWS will need to be an ongoing process, one that goes beyond the scope of this report to include multiple stakeholders. As such, there are numerous nuances and issues not addressed here. Vulnerabilities will need to be prioritized, and how the needs and desires of various stakeholder groups are con- sidered can change relative priorities. Some economic costs are clear (threats to an airport from sea level rise, for example) and some are less clear (potential loss of eco- system services, for example), making triage difficult. It is noted that communication is a crucial component of any early warning system to ensure the timely delivery of information on impending events, and prepare potential risk scenarios and prepared- ness strategies. Special considerations need to be given to the importance of accuracy, lead time, warning message content, warning transmission, and the appropriate social response to minimize negative consequences from the hazard (Kasperson et al., 1988; Mileti, 1999). For example, the fact that Superstorm Sandy was not labeled a hurricane by the National Hurricane Center as it came on shore (the storm lost its tropical char- acteristics before landfall) may have confused those in harm’s way and led unneces- sarily to a lessened sense of urgency and danger.2 In addition, an overall risk manage- ment system requires a preparedness and adaptations sub-system that feeds back on 2 For example, http://www.wsfa.com/story/21807734/whats-in-a-name-sandy-hurricane-or-superstorm. 156

OCR for page 147
The Way Forward BOX 4.3  LESSONS LEARNED FROM EARLY WARNINGS IN PAST ENVIRONMENTAL ISSUES Lessons from the European Environment Agency 2001 report on “Late lessons from early warnings: the precautionary principle 1896-2000.”   1.  Acknowledge and respond to ignorance, as well as uncertainty and risk, in technology appraisal and public policymaking.   2. Provide adequate long-term environmental and health monitoring and research into early warnings.   3. Identify and work to reduce “blind spots” and gaps in scientific knowledge.   4. Identify and reduce interdisciplinary obstacles to learning.   5. Ensure that real world conditions are adequately accounted for in regulatory appraisal.   6. Systematically scrutinize the claimed justifications and benefits alongside the poten- tial risks.   7. Evaluate a range of alternative options for meeting needs alongside the option under appraisal, and promote more robust, diverse and adaptable technologies so as to mini- mize the costs of surprises and maximize the benefits of innovation.   8. Ensure use of “lay” and local knowledge, as well as relevant specialist expertise in the appraisal.   9. Take full account of the assumptions and values of different social groups. 10. Maintain the regulatory independence of interested parties while retaining an inclu- sive approach to information and opinion gathering. 11. Identify and reduce institutional obstacles to learning and action. 12. Avoid “paralysis by analysis” by acting to reduce potential harm when there are rea- sonable grounds for concern. SOURCE: EEA, 2013. loss and damage by informing actions needed to reduce impacts from an impending event. An excellent summary of lessons learned from early warnings in past environmental issues can be found in a 2001 report by the European Environment Agency, and are shown in Box 4.3. In this section, the Committee provides further thoughts on selected aspects of an ACEWS: the monitoring, modeling, and synthesis aspects, as well as some special considerations for designing and implementing an ACEWS. Monitoring An ACEWS will require sustaining and integrating existing observing capabilities, as well as adding new capabilities targeted at improving understanding or early warn- 157

OCR for page 147
abrupt impacts of climate chang E BOX 4.4  A DECISION TREE FOR EXAMINING MONITORING NEEDS As a first step, identify the known and suspected thresholds. Some triage needs to be employed initially to avoid the temptation to monitor everything, and to ensure that the most obvious threats are prioritized. Prior reports and this one provide guidance, although additional effort may be needed in social and economic areas. As a second step, ask whether there is there an existing monitoring system already in place that could be used as is, or modified to meet the needs of an ACEWS (see Figure). In many cases, science already monitors key functions of the climate and other Earth cycles, and these systems should be exploited in an ACEWS. They may need more frequent data updates, or data handling may need to be modified to better use the data. If the answer to the above question is yes, then an important action is to protect that net- work to ensure that it continues to operate. A security camera that does not work, or does not watch all entrances and exits, is not very useful. Current resources to provide high-quality and continuous monitoring are at risk, and there are notable examples including critical time series that have been compromised due to reductions of in-situ and remotely sensed observational networks (e.g., the NOAA Cooperative Air Sampling Network). To improve our understanding of the evolving Earth system, monitoring resources must be protected and in some cases expanded. Another example of this category of monitoring system is the GRACE satellite mission. Originally launched in 2002, GRACE (actually two satellites) measures gravity, which in continental areas varies largely as a function of overall water content, and thus highlights areas with changing Identify known and suspected thresholds Does a Yes monitoring No system exist? Support/project Determine requirements monitoring and feasibility for developing network new monitoring network Yes, but inadequate Augment existing monitoring network FIGURE Simplified ACEWS decision tree for monitoring. Box 4-4 Figure_R02461.eps 160

OCR for page 147
The Way Forward BOX 4.4  Continued groundwater storage or withdrawal as well as changes in glaciers and ice sheets. In its 10-year lifetime, GRACE has revealed unsustainable rates of groundwater withdrawal in the southeastern United States and in parts of India, for example, and obtaining such data from other sources would be quite difficult in the United States and much more so in India. A successful ACEWS that monitors for food security should include such monitoring, as groundwater withdrawal is the common approach to combatting the impact of drought on crop yields. A failure of groundwa- ter to backstop rain would be a tipping point in the production of food, and a central part of a system to forecast famine.a If the answer to the above question is no (there is no existing monitoring system in place), then the next steps are to determine what is needed to implement a monitoring system, and to do so if it is feasible. A national inventory of the types and values of coastal resources vulnerable to some set level of sea level rise and storm surge is an example of such a monitoring system, one that does not currently exist but could be created and updated using information that is currently collected, but not collated into a central database.b Another example is a system to monitor the interaction between the ocean and the outlet glaciers of land ice. It is clear that ocean currents and temperatures play a key role in melting ice (Joughin et al., 2012a), and will be important in monitoring for catastrophic instability in the marine-based West Antarctic Ice Sheet as well as marine portions of the Greenland and East Antarctic ice sheets (Chapter 2, sec- tion on Ice Sheets and Sea Level), which in turn present the greatest risk of an unexpected and rapid contribution to sea level rise. A third possible answer to the above question is yes, a current monitoring system exists but it is inadequate to meet the needs of an ACEWS and it therefore needs to be augmented. An example of this is the Circumpolar Active Layer Monitoring (CALM) program, which is designed to observe the response of near-surface permafrost to climate change over long (multi-decadal) time scales. CALM, which is part of the National Science Foundation’s Arctic Observing Network, would likely need to be expanded and more automated to function as an ACEWS, especially be- cause, as described in the section on Potential Climate Surprises Due to High-Latitude Methane and Carbon Cycles in Chapter 2, there are large geographic coverage gaps in important regions. The science of abrupt climate change is not settled; monitoring needs will evolve over time and an iterative mechanism would allow ongoing assessment and evaluation. There needs to be some mechanism to allow for evolution of the ACEWS. One way to do this would be the creation of an ACEWS steering committee that would regularly visit the state of ACEWS monitoring efforts, critically examine proposals for new monitoring systems, and ensure that the current systems are meeting their stated goals. This is described more below as part of the implementation section. a Note: A GRACE follow-on mission is currently planned (see the FY 2014 President’s budget: http:// www.nasa.gov/pdf/740512main_FY2014%20CJ%20for%20Online.pdf ). b Any inventory effort could build upon existing efforts like those that follow on from work of the En- vironmental Protection Agency (http://plan.risingsea.net/index.html) and Climate Central (http://sealevel. climatecentral.org/). 161

OCR for page 147
abrupt impacts of climate chang E • Improved collaborative networks to entrain new communities of scientists and researchers to take fresh looks at the problem; for example, integrating climate scientists with applied mathematicians, statisticians, dynamical systems ex- perts, policy analysts, engineers, city planners, ethicists and others who ensure that the ACEWS meets the needs of a broad spectrum of stakeholders will result in a better ACEWS that produces products that are more useful to those affected; • Enhanced educational activities to provide a platform for innovation in pro- ducing a workforce that is comfortable working across the boundaries; and • Innovative tools—including new data analysis and modeling techniques, to allow for a novel perspective on abrupt change, as well as more robust statisti- cal tools that are needed for analyzing and understanding non-linear dynamic systems and inter-connection among various climate fields; in some cases, this will include applications of tools created in a different context to the abrupt change problems. These research elements are a key part of any monitoring or early warning system efforts. Special Considerations for Designing and Implementing an ACEWS Although the committee discusses some general concepts about implementing an ACEWS below, the implementation of an ACEWS should be planned and executed by those agencies that are tasked to do so and/or contribute funding to the effort. How that effort is organized is beyond the purview of the committee, and thus the ideas below should be viewed as suggestions intended to be helpful, and not prescriptive. One important aspect of an ACEWS should be the integration of the effort, from the monitoring, through modeling and interpreting of the data, to producing scientific products including peer-reviewed publications and consumer-friendly data products. Examples of monitoring efforts in which the data have not been analyzed regularly are unfortunately common,5 and given the inherently time-sensitive nature of an ACEWS, such an outcome would be anathema to the intent of the system. A few examples of current monitoring programs that integrate monitoring with active interpretation of the data include the Long Term Ecological Programs of the National Science Founda- tion, NOAA’s atmospheric gas monitoring program, NASA’s Stratospheric Observatory 5  The initial failure of satellite monitoring to identify the ozone hole over Antarctica because an au- tomatic routine flagged anomalous data as potential errors shows what can happen without sufficiently integrated interpretation (see Grundmann, 2002 for a more complete historical account). 162

OCR for page 147
The Way Forward (SOFIA), and the National Ecological Observatory Network (NEON). Those programs that are currently operational fully integrate an active science program with observa- tions, and have all been extremely successful in their respective missions. An ACEWS should build upon the success of these programs and others like them. Being mindful of stakeholder priorities and meeting stakeholder needs is another as- pect of a successful ACEWS that should be stressed. The National Integrated Drought Information System (NIDIS) is an example of a successful early warning program that integrates the needs of the user community with the monitoring and modeling sys- tems. NIDIS incorporates a spectrum of drought-related products, including regional products tailored to the needs of specific regions, as a regular part of the system. In general, research focus and early warning signal detection would be most beneficial if they were prioritized based on societal impacts and likelihood of occurrence of the extreme events and resultant abrupt changes. Another important aspect of a successful ACEWS is for the system to be flexible and adaptive. It is not enough to integrate data and interpretation, but the science in- volved in the interpretation should inform the whole system and help it to evolve to better meet the needs of society. This step is important for an ACEWS as in some cases we know what to watch for, such as changes in AMOC, and in some cases we are less sure, such as changes in Northern Hemisphere weather patterns that may accom- pany the large energy changes in the Arctic as sea ice melts, trading white sea ice that reflects solar energy for blue ocean that absorbs solar energy. Also, the system should be nimble enough to change focus if necessary as knowledge about abrupt change improves. It is clear today that there is much to learn about the threats of abrupt change, and an ACEWS would be best served if it were designed to evolve as knowl- edge, monitoring tools, and societal needs, evolve. NIDIS is an example of a system that strives to meet this goal. For example, the NIDIS Regional Drought Early Warning Systems explore a variety of early warning and drought risk reduction strategies, seek- ing to match user needs with observations and analyses while allowing for system adaptation and evolution. Organization of an ACEWS would benefit from capitalizing on existing programs, but there will be a need to capture the interconnectedness of the various parts of the cli- mate and human systems. ACEWS could eventually be run as a large, overarching pro- gram, but might better be started through coordination, integration, and expansion of existing and planned smaller programs like the Famine Early Warning System Network (FEWS NET), which is “a USAID-funded activity that collaborates with international, regional, and national partners to provide timely and rigorous early warning and 163

OCR for page 147
abrupt impacts of climate chang E vulnerability information on emerging and evolving food security issues.”6 One pos- sible mechanism to eventually achieve this overall coordination would be to start with a steering group who could provide efficient guidance. Such a steering committee could be made up of representatives of funding agencies, scientists, representatives of various user communities (including national security and interested businesses), and international partners, to name a subset of the possibilities. Subgroups or work- ing groups may be able to bring focus to specific issues that require more attention as needed, e.g., water, food, or ecosystem services. Beyond a steering group, a number of interagency coordinating mechanisms exist, and the committee is not specifically rec- ommending one over another. Whatever the mechanism, the committee does stress that coordination—to reduce duplication of efforts, maximize resources, and facilitate data and information sharing—is key to a successful ACEWS. ACEWS: NEED FOR ACTION As noted earlier, the proper design and implementation of an ACEWS will need to be an ongoing process and will require expertise from many different disciplines beyond just the physical sciences, as well as input from many different stakeholder groups. Providing a complete roadmap to a successful ACEWS was beyond the scope of this report, but the committee has outlined its initial thoughts on what would make such a system successful above. Much is known about the design, implementation, and sustainability of early warning systems that can be leveraged in addition what is de- scribed in this report. In summary, this report should be viewed as a call for an ACEWS to be designed, for to not make such a call would be to willfully ignore “dragons,” and that was an approach the committee strongly opposed. The committee views this call as being particularly salient in light of its analysis of the previous reports on abrupt climate change, where a common theme emerged. Begin- ning with the 2002 NAS study (NRC, 2002), recommendations have been made to ad- dress the problem, but little follow-up action has been taken. To gain the benefits that science can offer on this topic, action is needed now. An ACEWS need not be overly expensive and need not be created from scratch, as many resources now exist that can contribute, but the time is here to be serious about the threat of tipping points so as to better anticipate and better prepare ourselves for the inevitable surprises. 6  http://www.fews.net/Pages/default.aspx. 164

OCR for page 147
The Way Forward “No matter how clear our foresight, no matter how accurate our computer models, a belief about the future should never be mistaken for the truth. The future, as such, never occurs. It becomes the present. And no matter how well we prepare ourselves, when the imagined future becomes the very real present, it never fails to surprise.” —Alan AtKisson, Believing Cassandra 165

OCR for page 147
abrupt impacts of climate chang E TABLE 4.1 State of knowledge on potential candidate processes that might undergo abrupt change. These include both abrupt climate changes in the physical climate system and abrupt climate impacts of ongoing changes that, when certain thresholds are crossed, can cause abrupt impacts for society and ecosystems. The near term outlook for this century is highlighted as being of particular relevance for decision makers generally. Near Term Outlook Long Term (for an Outlook Potential Abrupt Climate Abrupt (for a Change or Impact Change Significant Level of Critical Needs and Key Examples of Current within This Change1 Scientific (Research, Consequences Trend Century) after 2100) Understanding Monitoring, etc.) Disruption to Atlantic Trend not Low High Moderate • Enhanced understanding Meridional Overturning clearly of changes at high latitudes Circulation (AMOC) detected in the North Atlantic (e.g., • Up to 80 cm sea level rise in warming and/or freshening North Atlantic of surface waters) • Southward shift of tropical • Monitoring of overturning at rain belts other latitudes • Large disruptions to local • Enhanced understanding of marine ecosystems drivers of AMOC variability • Ocean and atmospheric temperature and circulation changes • Changes in ocean’s ability to store heat and carbon Sea level rise (SLR) from Moderate Low2 High High • Maintenance and expansion Abrupt Changes in the Ocean ocean thermal expansion increase in of monitoring of sea level • Coastal inundation sea level rise (tide gauges and satellite • Storm surges more likely to data), ocean temperature cause severe impacts at depth, local coastal motions, and dynamic effects on sea level Sea level rise from Losing ice Unknown Unknown Low • Extensive needs, including destabilization of WAIS ice to raise sea but broad remote-sensing, sheets level Probably and modeling research • 3-4 m of potential sea level Low rise • Coastal inundation • Storm surges more likely to cause severe impacts Sea level rise from other ice Losing ice Low High High for some • Maintenance and expansion sheets (including Greenland to raise sea aspects, Low for of satellite, airborne, and and all others, but not level others surface monitoring capacity, including WAIS loss) process studies, and • As much as 60m of potential modeling research sea level rise from all ice sheets • Coastal inundation • Storm surges more likely to cause severe impacts 166 TableS-1- page1-sm2

OCR for page 147
The Way Forward TABLE 4.1 Continued Near Term Outlook Long Term (for an Outlook Potential Abrupt Climate Abrupt (for a Change or Impact Change Significant Level of Critical Needs and Key Examples of Current within This Change1 Scientific (Research, Consequences Trend Century) after 2100) Understanding Monitoring, etc.) Decrease in ocean oxygen Trend not Moderate High Low to • Expanded and standardized ...in the Ocean (cont.) (expansion in oxygen clearly Moderate monitoring of ocean minimum zones [OMZs]) detected oxygen content, pH, and • Threats to aerobic marine life temperature • Release of nitrous oxide • Improved understanding gas—a potent greenhouse and modeling of ocean gas—to the atmosphere mixing • Improved understanding of microbial processes in OMZs Changes to patterns Trends not Low Moderate Low to • Maintaining continuous of climate variability detectable for Moderate records of atmospheric most patterns pressure and temperatures • Substantial surface weather of climate from both in-situ and changes throughout variability remotely sensed sources much of extratropics if the Exception • Assessing robustness extratropical jetstreams were is southern of circulation shifts in to shift abruptly annular individual ensemble mode— members in climate change detectable simulations Abrupt Changes in the Atmosphere poleward shift • Developing theory on of middle circulation response to latitude anthropogenic forcing jetstream Increase in intensity, Detectable Moderate High High • Continued progress on frequency, and duration of increasing (Regionally understanding climate heat waves trends variable, dynamics • Increased mortality dependent • Increased focus on risk • Decreased labor capacity on soil assessment and resilience • Threats to food and water moisture) security Increase in frequency Increasing Moderate Moderate to Low to Moderate • Continued progress on and intensity of extreme trends for High understanding climate precipitation events dynamics (droughts/floods/ Trends for • Increased focus on risk hurricanes/major storms) drought and assessment and resilience • Mortality risks hurricanes • Infrastructure damage not clear • Threats to food and water security • Potential for increased TableS-1- page2-sm2 167

OCR for page 147
abrupt impacts of climate chang E TABLE 4.1 Continued Near Term Outlook Long Term (for an Outlook Abrupt (for a Potential Abrupt Climate Change Significant Level of Critical Needs Change or Impact and Key Current within This Change1 Scientific (Research, Examples of Consequences Trend Century) after 2100) Understanding Monitoring, etc.) Increasing release of Neutral trend Low High Moderate4 • Improved models of carbon stored in soils and to small trend hydrology/cryosphere permafrost in increasing interaction and ecosystem • of human- soil carbon response induced climate change3 release • Greater study of role of in rapid carbon release • Expanded borehole temperature monitoring networks • Enhanced satellite and ground-based monitoring of atmospheric methane concentrations at high latitudes Increasing release of Trend not Low5 Moderate Moderate6 • Field and model based methane from ocean clearly characterization of the Abrupt Changes at High Latitudes methane hydrates detected sediment column • of human- • Enhanced satellite and induced climate change ground-based monitoring of atmospheric methane concentrations at high latitudes Late-summer Arctic sea ice Strong trend High Very high High • Enhanced Arctic disappearance in decreasing observations, including • Large and irreversible effects sea ice cover atmosphere, sea ice, and on various components of ocean characteristics the Arctic ecosystem • Better monitoring and • Impacts on human society census studies of marine and economic development ecosystems in coastal polar regions • Improved large-scale • Implications for Arctic models that incorporate the shipping and resource evolving state of knowledge extraction • Potential to alter large-scale atmospheric circulation and its variability Winter Arctic sea ice Small trend Low Moderate High • Same as late summer disappearance (Decreasing Arctic sea ice • Same as late summer Arctic but not disappearance above sea ice disappearance disappearing) above, but more pronounced due to year-round lack of sea ice TableS-1- page3-sm2 168

OCR for page 147
The Way Forward TABLE 4.1 Continued Near Term Outlook Long Term (for an Outlook Abrupt (for a Potential Abrupt Climate Change Significant Level of Critical Needs Change or Impact and Key Current within This Change1 Scientific (Research, Examples of Consequences Trend Century) after 2100) Understanding Monitoring, etc.) Rapid state changes in Species Moderate High Moderate • Long term remote sensing ecosystems, species range shifts and in-situ studies of key range shifts, and species systems boundary changes others not • Improved hydrological and • Extensive habitat loss clearly ecological models • Loss of ecosystem services detected Abrupt Changes in Ecosystems • Threats to food and water supplies Increases in extinctions Species and High Very high Moderate • Better understanding of of marine and terrestrial population how species interactions species losses and ecological cascades • Loss of high percentage accelerating might magnify extinctions of coral reef ecosystems (Portion intensity (already underway) attributable • Better understanding of • percentage of to climate is how interactions between land mammal, bird, and uncertain) climate-caused extinctions amphibian species extinct or and other extinction drivers endangered7 (habitat fragmentation, overexploitation, etc.) multiply extinction intensity • Improved monitoring of key species 1 Change could be either abrupt or non-abrupt. 2 To clarify, the Committee assesses the near-term outlook that sea level will rise abruptly before the end of this century as Low; this is not in contradiction to the assessment that sea level will continue to rise steadily with estimates of between 0.26 and 0.82 m by the end of this century (IPCC, 2013). 3 Methane is a powerful but short-lived greenhouse gas. 4 Limited by ability to predict methane production from thawing organic carbon 5 No mechanism proposed would lead to abrupt release of substantial amounts of methane from ocean methane hydrates this century. 6 Limited by uncertainty in hydrate abundance in near-surface sediments, and fate of CH4 once released 7 Species distribution models (Thuiller et al., 2006) indicate between 10–40% of mammals now found in African protected areas will be extinct or critically endangered by 2080 as a result of modeled climate change. Analyses by Foden et al.(2013) and Ricke et al. (2013) suggest 41% of bird species, 66% of amphibian species, and between 61% and 100% of corals that are not now considered threatened with extinction will become threatened due to climate change sometime between now and 2100. 1 TableS-1- page4-sm3 169

OCR for page 147