IPCC_Managing Risks of Extreme Events.pdf - Climate Access

Climate Change: New Dimensions in Disaster Risk, Exposure, Vulnerability, and ResilienceChapter 1society and nature have accomplished on many occasions spontaneouslyin the past, if over a different range of conditions than expected in thefuture.Within the sphere of adaptation of natural systems to climate, amongtrees, for example, natural selection has the potential to evolveappropriate resilience to extremes (at some cost). Resistance towindthrow is strongly species-dependent, having evolved according tothe climatology where that tree was indigenous (Canham et al., 2001).In their original habitat, trees typically withstand wind extremes expectedevery 10 to 50 years, but not extremes that lie beyond their averagelifespan of 100 to 500 years (Ostertag et al., 2005).In human systems, communities traditionally accustomed to periodicdroughts employ wells, boreholes, pumps, dams, and water harvestingand irrigation systems. Those with houses exposed to high seasonaltemperatures employ thick walls and narrow streets, have developedpassive cooling systems, adapted lifestyles, or acquired air conditioning.In regions unaccustomed to heat waves, the absence of such systems,in particular in the houses of the most vulnerable elderly or sick,contributes to excess mortality, as in Paris, France, in August 2003(Vandentorren et al., 2004) or California in July 2006 (Gershunov et al.,2009).The examples given above of ‘spontaneous’ human system adjustmentcan be contrasted with explicit measures that are taken to reduce riskfrom an expected range of extremes. On the island of Guam, withinthe most active and intense zone of tropical cyclone activity on Earth,buildings are constructed to the most stringent wind design code in theworld. Buildings are required to withstand peak gust wind speeds of76 ms -1 , expected every few decades (International Building Codes,2003). More generally, annual wind extremes for coastal locations willtypically be highest at mid-latitudes while those expected once everycentury will be highest in the 10° to 25° latitude tropics (Walshaw,2000). Consequently, indigenous building practices are less likely to beresilient close to the equator than in the windier (and storm surgeaffected) mid-latitudes (Minor, 1983).While local experience provides a reservoir of knowledge from whichdisaster risk management and adaptation to climate change are drawing(Fouillet et al., 2008), it may not be available to other regions yet to beaffected by such extremes. Thus, these experiences may not be drawnupon to provide guidance if future extremes go outside the traditionalor recently observed range, as is expected for some extremes as theclimate changes (see Chapter 3).1.3. Disaster Management, Disaster RiskReduction, and Risk TransferOne important component of both disaster risk management andadaptation to climate change is the appropriate allocation of effortsamong disaster management, disaster risk reduction, and risk transfer,as defined in Section 1.1.2.2. The current section provides a brief surveyof the risk governance framework for making judgments about such anallocation, suggests why climate change may complicate effectivemanagement of disaster risks, and identifies potential synergiesbetween disaster risk management and adaptation to climate change.Disaster risks appear in the context of human choices that aim to satisfyhuman wants and needs (e.g., where to live and in what types ofdwelling, what vehicles to use for transport, what crops to grow, whatinfrastructure to support economic activities, Hohenemser et al., 1984;Renn, 2008). Ideally, the choice of any portfolio of actions to addressdisaster risk would take into consideration human judgments aboutwhat constitutes risk, how to weigh such risk alongside other valuesand needs, and the social and economic contexts that determine whosejudgments influence individuals’ and societal responses to those risks.The risk governance framework offers a systematic way to help situatesuch judgments about disaster management, risk reduction, and risktransfer within this broader context. Risk governance, under Renn’s(2008) formulation, consists of four phases – pre-assessment, appraisal,characterization/evaluation, and management – in an open, cyclical,iterative, and interlinked process. Risk communication accompanies allfour phases. This process is consistent with those in the UNISDR HyogoFramework for Action (UNISDR, 2005), the best known and adhered toframework for considering disaster risk management concerns (seeChapter 7).As one component of its broader approach, risk governance usesconcepts from probabilistic risk analysis to help judge appropriateallocations in level of effort and over time and among risk reduction,risk transfer, and disaster management actions. The basic probabilisticrisk analytic framework for considering such allocations regards riskas the product of the probability of an event(s) multiplied by itsconsequence (see Box 1-2; Bedford and Cooke, 2001). In this formulation,risk reduction aims to reduce exposure and vulnerability as well as theprobability of occurrence of some events (e.g., those associated withlandslides and forest fires induced by human intervention). Risk transferefforts aim to compensate losses suffered by those who directly experiencean event. Disaster management aims to respond to the immediateconsequences and facilitate reduction of longer-term consequences (seeSection 1.1).Probabilistic risk analysis can help compare the efficacy of alternativeactions to manage risk and inform judgments about the appropriateallocation of resources to reduce risk. For instance, the frameworksuggests that equivalent levels of risk reduction result from reducing anevent’s probability or by reducing its consequences by equal percentages.Probabilistic risk analysis also suggests that a series of relatively smaller,more frequent events could pose the same risk as a single, relatively lessfrequent, larger event. Probabilistic risk analysis can help inform decisionsabout alternative allocations of risk management efforts by facilitatingthe comparison of the increase or decrease in risk resulting from thealternative allocations (high confidence). Since the costs of available44