IPCC Report.pdf - Adam Curry

Climate Change: New Dimensions in Disaster Risk, Exposure, Vulnerability, and ResilienceChapter 11.2.2. Extreme Events Defined in Physical Terms1.2.2.1. Definitions of ExtremesSome literature reserve the term ‘extreme event’ for initial meteorologicalphenomena (Easterling et al., 2000; Jentsch et al., 2007), some includethe consequential physical impacts, like flooding (Young, 2002), and somethe entire spectrum of outcomes for humans, society, and ecosystems(Rich et al., 2008). In this report, we use ‘extreme (weather or climate)event’ to refer solely to the initial and consequent physical phenomenaincluding some (e.g., flooding) that may have human components tocausation other than that related to the climate (e.g., land use or landcover change or changes in water management; see Section 3.1.2 andGlossary). The spectrum of outcomes for humans, society, and physicalsystems, including ecosystems, are considered ‘impacts’ rather than partof the definition of ‘events’ (see Sections 1.1.2.1 and 3.1.2 and theGlossary).In addition to providing a long-term mean of weather, ‘climate’characterizes the full spectrum of means and exceptionality associatedwith ‘unusual’ and unusually persistent weather. The World MeteorologicalOrganization (WMO, 2010) differentiates the terms in the following way(see also FAQ 6.1): “At the simplest level the weather is what is happeningto the atmosphere at any given time. Climate in a narrow sense isusually defined as the ‘average weather,’ or more rigorously, as thestatistical description in terms of the mean and variability of relevantquantities over a period of time.”Weather and climate phenomena reflect the interaction of dynamic andthermodynamic processes over a very wide range of space and temporalscales. This complexity results in highly variable atmospheric conditions,including temperatures, motions, and precipitation, a component ofwhich is referred to as ‘extreme events.’ Extreme events include thepassage of an intense tornado lasting minutes and the persistence ofdrought conditions over decades – a span of at least seven orders ofmagnitude of timescales. An imprecise distinction between extreme‘weather’ and ‘climate’ events, based on their characteristic timescales,is drawn in Section 3.1.2. Similarly, the spatial scale of extreme climateor weather varies from local to continental.Where there is sufficient long-term recorded data to develop a statisticaldistribution of a key weather or climate variable, it is possible to find theprobability of experiencing a value above or below different thresholdsof that distribution as is required in engineering design (trends may besought in such data to see if there is evidence that the climate has notbeen stationary over the sample period; Milly et al., 2008). The extremityof a weather or climate event of a given magnitude depends ongeographic context (see Section 3.1.2 and Box 3-1): a month of dailytemperatures corresponding to the expected spring climatological dailymaximum in Chennai, India, would be termed a heat wave in France; asnow storm expected every year in New York, USA, might initiate adisaster when it occurs in southern China. Furthermore, according to thelocation and social context, a 1-in-10 or 1-in-20 annual probabilityevent may not be sufficient to result in unusual consequences.Nonetheless, universal thresholds can exist – for example, a reductionin the incidence or intensity of freezing days may allow certain diseasevectors to thrive (e.g., Epstein et al., 1998). These various aspects areconsidered in the definition of ‘extreme (weather and climate) events.’The availability of observational data is of central relevance for definingclimate characteristics and for disaster risk management; and, while datafor temperature and precipitation are widely available, some associatedvariables, such as soil moisture, are poorly monitored, or, like extremewind speeds and other low frequency occurrences, not monitored withsufficient spatial resolution or temporal continuity (Section 3.2.1).1.2.2.2. Extremes in a Changing ClimateAn extreme event in the present climate may become more common, ormore rare, under future climate conditions. When the overall distributionof the climate variable changes, what happens to mean climate maybe different from what happens to the extremes at either end of thedistribution (see Figure 1-2).For example, a warmer mean climate could result from fewer cold days,leading to a reduction in the variance of temperatures, or more hot days,leading to an expansion in the variance of the temperature distribution,or both. The issue of the scaling of changes in extreme events with respectto changes in mean temperatures is addressed further in Section 3.1.6.In general, single extreme events cannot be simply and directly attributedto anthropogenic climate change, as there is always a possibility theevent in question might have occurred without this contribution (Hegerlet al., 2007; Section 3.2.2; FAQ 3.2). However, for certain classes ofregional, long-duration extremes (of heat and rainfall) it has provedpossible to argue from climate model outputs that the probability ofsuch an extreme has changed due to anthropogenic climate forcing(Stott et al., 2004; Pall et al., 2011).Extremes sometimes result from the interactions between two unrelatedgeophysical phenomena such as a moderate storm surge coincidingwith an extreme spring tide, as in the most catastrophic UK storm surgeflood of the past 500 years in 1607 (Horsburgh and Horritt, 2006).Climate change may alter both the frequency of extreme surges andcause gradual sea level rise, compounding such future extreme floods(see Sections 3.5.3 and 3.5.5).1.2.2.3. The Diversity and Range of ExtremesThe specification of weather and climate extremes relevant to theconcerns of individuals, communities, and governments depends on theaffected stakeholder, whether in agriculture, disease control, urbandesign, infrastructure maintenance, etc. Accordingly, the range of suchextremes is very diverse and varies widely. For example, whether it falls40