IPCC Report.pdf - Adam Curry

Chapter 3Changes in Climate Extremes and their Impacts on the Natural Physical Environmentthe 11 coupled climate model simulations examined by Yeh et al. (2009),three projected a relative decrease in the frequency of these centralPacific episodes, and only four of the models produced a statisticallysignificant change to more frequent central Pacific events.A caveat regarding all projections of future behavior of ENSO arisesfrom systematic biases in the depiction of ENSO behavior through the20th century by models (Randall et al., 2007; Guilyardi et al., 2009).Leloup et al. (2008) for instance, demonstrate that coupled climatemodels show wide differences in the ability to reproduce the spatialcharacteristics of SST variations associated with ENSO during the 20thcentury, and all models have failings. They concluded that it is difficultto even classify models by the quality of their reproductions of thebehavior of ENSO, because models scored unevenly in their reproductionof the different phases of the phenomenon. This makes it difficult todetermine which models to use to project future changes in ENSO.Moreover, most of the models are not able to reproduce the typicalcirculation anomalies associated with ENSO in the SouthernHemisphere (Vera and Silvestri, 2009) and the Northern Hemisphere(Joseph and Nigam, 2006).There was no consistency in projections of changes in ENSO variabilityor frequency at the time of the AR4 (Meehl et al., 2007b) and thissituation has not changed as a result of post-AR4 studies. The evidenceis that the nature of ENSO has varied in the past apparently sometimesin response to changes in radiative forcing but also possibly due tointernal climatic variability. Since radiative forcing will continue tochange in the future, we can confidently expect changes in ENSO andits impacts as well, although both El Niño and La Niña episodes willcontinue to occur (e.g., Vecchi and Wittenberg, 2010). Our current limitedunderstanding, however, means that it is not possible at this time toconfidently predict whether ENSO activity will be enhanced or dampeddue to anthropogenic climate change, or even if the frequency of El Niñoor La Niña episodes will change (Collins et al., 2010).In summary, there is medium confidence in a recent trend towardmore frequent central equatorial Pacific El Niño episodes, butinsufficient evidence for more specific statements aboutobserved trends in ENSO. Model projections of changes in ENSOvariability and the frequency of El Niño episodes as a consequenceof increased greenhouse gas concentrations are not consistent,and so there is low confidence in projections of changes in thephenomenon. However, there is medium confidence regarding aprojected increase (projected by most GCMs) in the relativefrequency of central equatorial Pacific events, which typicallyexhibit different patterns of climate variations than do theclassical East Pacific events.3.4.3. Other Modes of VariabilityOther natural modes of variability beside ENSO (Section 3.4.2) that arerelevant to extremes and disasters include the North Atlantic Oscillation(NAO), the Southern Annular Mode (SAM), and the Indian Ocean Dipole(IOD) (Trenberth et al., 2007). The NAO is a large-scale seesaw inatmospheric pressure between the subtropical high and the polar lowin the North Atlantic region. The positive NAO phase has a strongsubtropical high-pressure center and a deeper than normal Icelandiclow. This results in a shift of winter storms crossing the Atlantic Oceanto a more northerly track, and is associated with warm and wet wintersin northwestern Europe and cold and dry winters in northern Canadaand Greenland. Scaife et al. (2008) discuss the relationship betweenthe NAO and European extremes. Paleoclimatic data indicate that theNAO was persistently in its positive phase during medieval times andpersistently in its negative phase during the cooler Little Ice Age (Trouetet al., 2009). The NAO is closely related to the Northern Annular Mode(NAM); for brevity we focus here on the NAO but much of what is saidabout the NAO also applies to the NAM. The SAM is the largest mode ofSouthern Hemisphere extratropical variability and refers to north-southshifts in atmospheric mass between the middle and high latitudes. Itplays an important role in climate variability in these latitudes. The SAMpositive phase is linked to negative sea level pressure anomalies overthe polar regions and intensified westerlies. It has been associated withcooler than normal temperatures over most of Antarctica and Australia,with warm anomalies over the Antarctic Peninsula, southern SouthAmerica, and southern New Zealand, and with anomalously dry conditionsover southern South America, New Zealand, and Tasmania and wetanomalies over much of Australia and South Africa (e.g., Hendon et al.,2007). The IOD is a coupled ocean-atmosphere phenomenon in the IndianOcean. A positive IOD event is associated with anomalous cooling in thesoutheastern equatorial Indian Ocean and anomalous warming in thewestern equatorial Indian Ocean. Recent work (Ummenhofer et al., 2008,2009a,b) has implicated the IOD as a cause of droughts in Australia, andheavy rainfall in east Africa (Ummenhofer et al., 2009c). There is alsoevidence of modes of variability operating on multi-decadal time scales,notably the Pacific Decadal Oscillation (PDO) and the Atlantic MultidecadalOscillation (AMO). Variations in the PDO have been related toprecipitation extremes over North America (Zhang et al., 2010).Both the NAO and the SAM exhibited trends toward their positive phase(strengthened mid-latitude westerlies) over the last three to four decades,although the NAO has been in its negative phase in the last few years.Goodkin et al. (2008) concluded that the variability in the NAO is linkedwith changes in the mean temperature of the Northern Hemisphere.Dong et al. (2011) demonstrated that some of the observed late 20thcenturydecadal-scale changes in NAO behavior could be reproduced byincreasing the CO 2 concentrations in a coupled model, and concludedthat greenhouse gas concentrations may have played a role in forcingthese changes. The largest observed trends in the SAM occur inDecember to February, and model simulations indicate that these aredue mainly to stratospheric ozone changes. However it has been arguedthat anthropogenic circulation changes are poorly characterized by trendsin the annular modes (Woollings et al., 2008). Further complicatingthese trends, Silvestri and Vera (2009) reported changes in the typicalhemispheric circulation pattern related to the SAM and its associatedimpact on both temperature and precipitation anomalies, particularly157