Changing the Resilience Paradigm Igor Linkov, Todd Bridges, Felix Creutzig, Jennifer Decker, Cate Fox-Lent, Wolfgang Kröger, James H

opinion & comment

COMMENTARY: Changing the resilience paradigm Igor Linkov, Todd Bridges, Felix Creutzig, Jennifer Decker, Cate Fox-Lent, Wolfgang Kröger, James H. Lambert, Anders Levermann, Benoit Montreuil, Jatin Nathwani, Raymond Nyer, Ortwin Renn, Benjamin Scharte, Alexander Scheffler, Miranda Schreurs and Thomas Thiel-Clemen Resilience management goes beyond risk management to address the complexities of large integrated systems and the uncertainty of future threats, especially those associated with climate change.

he human body is resilient in its ability associated with the vulnerabilities of these

to persevere through infections or V systems, combined with the unpredictability ulner trauma. Even through severe disease, of climatic extremes, challenges our ability T abilit critical life functions are sustained and the Threat to understand and manage them. To address body recovers, often adapting by developing Risk y these challenges, risk analysis should be immunity to further attacks of the same used where possible to help prepare for and type. Our society’s critical infrastructure — Consequence prevent consequences of foreseeable events, cyber, energy, water, transportation and but resilience must be built into systems to communication — lacks the same degree help them quickly recover and adapt when of resilience, typically losing essential adverse events do occur.

functionality following adverse events. y A roadmap for enabling the development Plan Although the number of climatic extremes Adapt of such capability should include: (1) may intensify or become more frequent1, specific methods to define and measure there is currently no scientific method Absorb resilience; (2) new modelling and simulation available to precisely predict the long-term over techniques for highly complex systems; evolution and spatial distribution of tropical Rec (3) development of resilience engineering;

Critical functionalit System cyclones, atmospheric blockages and extra- resilience (4) approaches for communication with tropical storm surges; nor are the impacts stakeholders. Strategies for communicating on society’s infrastructure in any way with policy makers are needed to support quantified2. In the face of these unknowns, the shift to resilience management by building resilience becomes the optimal legislative, regulatory and other means. Time course of action for large complex systems. The National Academy of Sciences (NAS) Resilience, as a property of a system, defines resilience as “the ability to prepare must transition from just a buzzword Figure 1 | A resilience management framework and plan for, absorb, recover from, and to an operational paradigm for system includes risk analysis as a central component. more successfully adapt to adverse events”4. management, especially under future Risk analysis depends on characterization of Conceptually, risk analysis quantifies the climate change. Current risk analysis the threats, vulnerabilities and consequences of probability that the system will reach the methods identify the vulnerabilities of adverse events to determine the expected loss lowest point of the critical functionality specific system components to an expected of critical functionality. The National Academy profile. Risk management helps the system adverse event and quantify the loss in of Sciences definition of resilience places risk in prepare and plan for adverse events, whereas functionality of the system as a consequence the broader context of a system’s ability to plan resilience management goes further by of the event occurring3. Subsequent risk for, recover from and adapt to adverse events integrating the temporal capacity of a system management has focused on hardening over time. In the system functionality profile, risk to absorb and recover from adverse events, these specific system components to in a system is interpreted as the total reduction and then adapt (Fig. 1). Resilience is not a withstand the identified threats to an in critical functionality and the resilience of the substitute for principled system design or acceptable level and to prevent overall system is related to the slope of the absorption risk management5. Rather, resilience is a system failure. curve and the shape of the recovery curve — complementary attribute that uses strategies Two factors make this form of protection indicating the temporal effect of the adverse of adaptation and mitigation to improve unrealistic for many systems. First, event on the system. The dashed line suggests traditional risk management. Strategies to increasingly interconnected social, technical that highly resilient systems can adapt in such build resilience can take the form of flexible and economic networks create large complex a way that the functionality of the system may response, distributed decision making, systems and the risk analysis of many improve with respect to the initial performance, modularity, redundancy, ensuring the individual components becomes cost and enhancing the system’s resilience to future independence of component interactions time prohibitive. Second, the uncertainties adverse events. or a combination of adaptive strategies to

NATURE CLIMATE CHANGE | VOL 4 | JUNE 2014 | www.nature.com/natureclimatechange 407

used risk management to address flooding

y in this region. Building resilient systems Low risk High risk to reduce the impact of future events and High resilience High resilience speed recovery will involve the integration of a combination of structural measures (such as seawalls and levees), non-structural

Critical functionalit measures (for example, the management of floodplains) and natural and nature-based features (such as beach-dunce complexes and wetlands)10. Developing a better road Low risk infrastructure and training emergency Low resilience staff could improve the overall resilience High risk of coastal communities where flooding Low resilience hampers transportation, communication and the deployment of crucial services. Critical functionality Management strategies for one network (for example, telecommunications, water, Time Time gas, electricity or transportation) often rely Figure 2 | Schematic representations of changes in critical functionality over time show the interplay of on the functionality of another network risk and resilience in a system’s performance during an adverse event. The size of the initial perturbation so that all could be considered part of an reflects the total risk to the system while the shape of the recovery curve is controlled by the system overall system of systems. Resilience can resilience. The area under the curve is indicative of the overall system functionality. Systems that face be enhanced by studying and improving high risks with high resilience perform better than those facing similar risks but with low resilience. the interconnectivity of networks. For Systems with low risk but also low resilience may perform the same as, or possibly worse than, systems example, during Hurricane Sandy, power with high risk and high resilience. remained off in many coastal areas because street flooding prevented repair trucks from accessing damaged facilities. Systems- minimize the loss of functionality and to Once the system is defined and models level resilience analysis can help identify increase the slope of the recovery (Fig. 2). to support resilience quantification are the need for alternative capacities and the Identifying the need for system resilience developed, the next step is to design functional autonomy of a given network requires defining the system. Current efforts interdependent infrastructures to be more in the case where multiple networks are often focus on defining systems in just resilient. Unlike risk-based design, which concurrently impacted. one domain (for example, the physical or focuses on one component at a time, Enhancing social networks is also a information domain), but the complexity resilience engineering identifies critical component of resilience. The meltdown of of threats affecting multiple domains, and system functionalities that are valuable to nuclear reactors at the Fukushima Dai-ichi increasing interdependencies between stakeholders and society. It also involves power plant in 2011 following an earthquake them, requires expanding this definition. the development of customized socio- and tsunami led to a crisis situation from In one effort, the concept of military technological methods and solutions to which Japan is still recovering. In some ways, network-centric operation across physical, ensure these functionalities are sustained Japanese society has proven resilient in the information, social and cognitive domains under broad categories of threats. Thus, face of this calamity as communities have was combined with the NAS’s definition resilience engineering builds resistance, come together to rebuild and move forward. of resilience4 to build a resilience matrix adaptability and the ability to recover quickly At the same time, radiation has turned some where four life-cycle stages of a resilient in the face of adverse events (for example, communities into ghost towns and daunting system (plan, absorb, recover and adapt) see ref. 9). Examples of technologies for efforts are required to decontaminate are assessed in each domain6,7. Of course, resilience engineering are self-healing, large stretches of land. Resilience must be assessing resilience either through this adaptive materials or energy-self-sufficient considered over different time horizons — matrix approach or otherwise requires and automated sensor networks. Such immediate (for example, evacuation and a detailed understanding of a system’s technologies alone do not improve the medical services), intermediate (establishing behaviour and functions, especially during resilience of a system, but they can be used temporary communities to maintain social catastrophic events8. Modelling and as components to help reduce unanticipated connections and clean-up operations) simulation of complex, interconnected socio- risks and guide efficient recovery where it is and long-term (permanent relocation and technical systems allows system managers most necessary. funding to rebuild). Japan has persevered to identify weak spots, plan counter- Recent natural disasters and their in the face of a large calamity, but it has also measures in advance, fix errors and prepare, effects on local socio-techno-economic experienced policy shortcomings that have in a comprehensive way, for diverse and systems provide examples where resilience further strained affected communities and heterogeneous threats and vulnerabilities. management could supplement risk prevented rapid recovery11. Growing complexity and the emergence of ex management to improve outcomes following Climate change, among other stressors, ante unprecedented and unpredicted threats the events. In 2012, Hurricane Sandy struck might produce events that cannot be necessitate sound principles, innovative the Atlantic Coast of the United States, precisely predicted, analysed or prepared thinking, databases, models, methods and flooding major metropolitan areas in New for with traditional risk analysis techniques. simulations of socio-economic technical York and New Jersey, and leaving large Moving forward, resilience will be a key systems (especially concerning cascading populations without fresh water or sanitation component of sustainable development — effects) and the application of systems theory facilities. The US Army Corps of Engineers, striving to meet the needs of the present and network science in resilience analysis. among other agencies, has traditionally without compromising the ability of

408 NATURE CLIMATE CHANGE | VOL 4 | JUNE 2014 | www.nature.com/natureclimatechange

future generations to meet their own for transparent dialogue on resilience Dynamics, Eckerstraβe 4, 79104 Freiburg, Germany, needs12. Resilient systems share qualities of management with stakeholders, such as 13Hamburg University of Technology, Kasernenstraβe sustainable systems because they are able to industry associations and policymakers, is 12, 21073 Hamburg, Germany, 14Free University minimize the negative impacts of adverse essential for the timely and broad acceptance of Berlin, Ihnestraβe 22, 14195 Berlin, Germany, events on societies and sustain or even of resilience concepts. ❒ 15Hamburg University of Applied Sciences, Lohbrügger improve their functionality by adapting to Kirchstrasse 65, 21033 Hamburg, Germany. and learning from fundamental changes Igor Linkov 1*, Todd Bridges2, Felix Creutzig 3, *e-mail: [email protected] caused by those events. Jennifer Decker 4, Cate Fox-Lent 1, Wolfgang Kröger 5, 6 7 References In summary, risk analysis and risk James H. Lambert , Anders Levermann , 1. IPCC Managing the Risks of Extreme Events and Disasters to 8 9 management based on probabilistic Benoit Montreuil , Jatin Nathwani , Advance Climate Change Adaptation. A Special Report of Working quantitative methods have been widely Raymond Nyer 10, Ortwin Renn11, Benjamin Scharte12, Groups I and II of the Intergovernmental Panel on Climate Change Alexander Scheffler 13, Miranda Schreurs14 and (eds Field, C. B. et al.) 582 (Cambridge Univ. Press, 2012). adopted and have been useful for dealing 2. Levermann, A. Nature 506, 27–29 (2014). 15 1 with foreseeable and calculable stress Thomas Thiel-Clemen are at United States Army 3. Linkov, I. et al. Env. Sci. Tech. 47, 10108–10110 (2014). situations. Benchmarks and thresholds for Corps of Engineers — Engineer Research and 4. National Research Council Disaster Resilience: A National risk analysis are built into the regulations Development Center, Environmental Laboratory, Imperative (The National Academies Press, 2012). 5. Park, J., Seager, T. P., Rao, P. C. S., Convertino, M. & Linkov, I. and policies of organizations and nations; 696 Virginia Road, Concord, Massachusetts 01742, Risk Analysis 33, 356–367 (2013). however, this approach is no longer USA, 2United States Army Corps of Engineers — 6. Linkov, I. et al. Environ. Syst. Decisions 33, 471–476 (2013). sufficient to address the evolving nature of Engineer Research and Development Center, 7. Roege, P. et al. Energy Policy (in the press). 8. Dauphiné, A. & Provitolo. D. Annales de Géographie risks in the modern world. Moreover, the Environmental Laboratory, 3909 Halls Ferry Road, 654, 115–125 (2007). increased complexity and interdependency Vicksburg, Massachusetts 39180, USA, 3Mercator 9. Lambert, J. H., Tsang, J. & Thekdi, S. Am. Soc. Civ. Eng. J. Infrastr. Syst. of many of society’s critical networks Research Institute on Global Commons and Climate 19, 384–394 (2013). 10. Bridges, T. et al. Coastal risk reduction and resilience presents a fundamental challenge to even Change, Torgauer Straβe 12–15, 10829 Berlin, (US Army Corps of Engineers, 2013). the most comprehensive and sophisticated Germany, 4Embassy of Canada, Leipziger Platz 17, 11. Hollnagel, E. & Fujita, Y. Nuc. Eng. Tech. 45, 1–8 (2013). risk analysis. Therefore, early integration 10117 Berlin, Germany, 5Swiss Federal Institute 12. World Commission on Environment and Development Our common future (Oxford Univ. Press, 1987). of resilience into the design of systems of Technology Zürich (ETH), Scheuchzerstrasse 7, and the regulatory structures of systems 8092 Zürich, Switzerland, 6University of Virginia, Acknowledgements management is needed to address the 151 Engineer’s Way, Charlottesville, Virginia This paper resulted from discussions at a workshop on the use emerging issues associated with complexity 22903, USA, 7Potsdam Institute for Climate of risk and resilience assessment methodologies in guiding and uncertainty. An urgent need exists to Impact Research, Telegrafenberg A 31, 14191 policy development, held at the Embassy of Canada in Berlin, 8 Germany on 4 February 2014. Partial financial support for the complement the existing knowledge-base Potsdam, Germany, Université Laval, 2325 Rue de workshop was provided by grant W911NF-13-1-0168 to RNC 9 of risk analysis and management by further l’Université, Québec G1V 0A6, Canada, University Conseil, Neuilly-sur-Seine, France. Additional funding was developing frameworks and models enabling of Waterloo, 200 University Ave W, Waterloo, provided by the German National Academy of Technology system-wide and network-wide resilience Ontario N2L 3G1, Canada, 10RNC Conseil and and Engineering (Acatech), the Helmholtz Association, analysis, engineering and management. Ecole Centrale de Paris, 56 Rue Charles Laffitte, the Embassy of Canada and the Society for Risk Analysis. 11 Permission was granted by the USACE Chief of Engineers to Although research and development 92200 Neuilly-sur-Seine, France, University publish this material. The views and opinions expressed in on methods and tools is progressing, of Stuttgart, Seidenstraβe 36, 70174 Stuttgart, this paper are those of the individual authors and not those of establishing channels of communication Germany, 12Fraunhofer Institute for High-Speed the US Army or other sponsor organizations.

COMMENTARY: Capturing provenance of global change information Xiaogang Ma, Peter Fox, Curt Tilmes, Katharine Jacobs and Anne Waple Global change information demands access to data sources and well-documented provenance to provide the evidence needed to build confidence in scientific conclusions and decision making. A new generation of web technology, the Semantic Web, provides tools for that purpose.

he topic of global change covers resources, atmospheric composition and/or makes the provenance of global change changes in the global environment chemistry and ecological systems. Data information especially important. In Tthat may alter the capacity of the and findings associated with global change addition, because different types of decisions Earth to sustain life and support human research are of great public, government benefit from different types of information, systems1. This includes changes to climate, and academic concern and are used understanding how to capture and present land productivity, oceans or other water in policy and decision making, which the provenance of global change information

NATURE CLIMATE CHANGE | VOL 4 | JUNE 2014 | www.nature.com/natureclimatechange 409