IPCC Report.pdf - Adam Curry

Changes in Climate Extremes and their Impacts on the Natural Physical EnvironmentChapter 3mid-latitude storms averaged over each hemisphere (Meehl et al.,2007b) and a poleward shift of storm tracks in both hemispheres(particularly evident in the Southern Hemisphere), with greater stormactivity at higher latitudes (Meehl et al., 2007b).A poleward shift in the upper level tropospheric storm track due toincreased greenhouse gas forcing is supported by post-AR4 studies(Lorenz and DeWeaver, 2007; O’Gorman, 2010; Wu et al., 2011). Itshould be noted that other studies indicate that the poleward shift isless clear when models including a full stratosphere or ozone recoveryare used (Huebener et al., 2007; Son et al., 2008; Morgenstern et al.,2010; Scaife et al., 2011) and the strength of the poleward shift is oftenseen more clearly in upper-level quantities than in low-level transientparameters (Ulbrich et al., 2008). Post-AR4 single model studies supportthe projection of a reduction in extratropical cyclones averaged over theNorthern Hemisphere during future warming (Finnis et al., 2007;Bengtsson et al., 2009; Orsolini and Sorteberg, 2009). However, neitherthe global changes in storm frequency or intensity were found to bestatistically significant by Bengtsson et al. (2009), although they wereaccompanied by significant increases in total and extreme precipitation.Models tend to project a reduction of winter cyclone activity throughoutthe mid-latitude North Pacific and for some models a north-easternmovement of the North Pacific storm track (Loeptien et al., 2008; Ulbrichet al., 2008; Favre and Gershunov, 2009; McDonald, 2011). However, theexact geographical pattern of cyclone frequency anomalies exhibitslarge variations across models (Teng et al., 2008; Favre and Gershunov,2009; Laine et al., 2009).Using band-passed sea level pressure data from 16 CMIP3 coupledGCMs, Ulbrich et al. (2008) showed regional increases in the storm trackactivity over the Eastern North Atlantic/Western European area. Thiseastward or southeastward extension of the storm track is also found inother studies (Ulbrich et al., 2008; Laine et al., 2009; McDonald, 2011) andmay be attributed to a local minimum in ocean warming in the centralNorth Atlantic and subsequent local changes in baroclinicity (McDonald,2011). In line with the eastward shift, Donat et al. (2010a) projected anincrease in wind storm days for central Europe by the end of the 21stcentury. The increase varies according to the definition of storminessand one model projects a decrease. A common deficiency among manyAR4 models is a coarsely resolved stratosphere and there are stillconcerns that this may lead to systematic biases in the Atlantic stormtrack response to increased anthropogenic forcing (Scaife et al., 2011).A reduction in cyclone frequency along the Canadian east coast hasbeen reported (Bengtsson et al., 2006; Watterson, 2006; Pinto et al.,2007a; Teng et al., 2008; Long et al., 2009). New results for SouthernHemisphere cyclones confirm the previously projected poleward shift instorm tracks under increased greenhouse gases (Lim and Simmonds,2009). That study projected a reduction of Southern Hemisphereextratropical cyclone frequency and intensity in mid-latitudes but aslight increase at high latitudes. The poleward shift due to increasedgreenhouse gases may be partly opposed by ozone recovery (Son et al.,2010).Detailed analyses of changes in physical mechanisms related to cyclonechanges in coupled climate models are still few. O’Gorman (2010) showedthat changes in mean available potential energy of the atmosphere canaccount for much of the varied response in storm-track intensity toglobal warming, implying that changes in storm-track intensity aresensitive to competing effects of changes in temperature gradients andstatic stability in different atmospheric levels. Using two coupled climatemodels, Laine et al. (2009) indicate that the primary cause for synopticactivity changes at the western end of the Northern Hemisphere stormtracks is related to the baroclinic conversion processes linked to meantemperature gradient changes in localized regions of the westernoceanic basins. They also found downstream changes in latent heatrelease during the developing and mature stages of the cyclone to be ofimportance and indicated that changes in diabatic process may beamplified by the upstream baroclinic changes [stronger (weaker)baroclinic activity in the west gives stronger (weaker) latent heatrelease downstream]. Pinto et al. (2009) found that regional increasesin track density and intensity of extreme cyclones close to the BritishIsles using a single model was associated with an eastward shift of thejet stream into Europe, more frequent extreme values of baroclinicity,and stronger upper level divergence.The modeled reduction in Southern Hemisphere extratropical cyclonefrequency and intensity in the mid-latitudes has been attributed to thetropical upper tropospheric warming enhancing static stability anddecreasing baroclinicity while an increased meridional temperaturegradient in the high latitudes is suggested to be responsible for theincrease in cyclone activity in this region (Lim and Simmonds, 2009). Inaddition to details in the modeled changes in local baroclinicity anddiabatic changes, the geographical pattern of modeled response incyclone activity has been reported to be influenced by the individualmodel’s structure of intrinsic modes of variability (Branstator and Selten,2009) and biases in the climatology (Kidston and Gerber, 2010).In summary it is likely that there has been a poleward shift in themain Northern and Southern Hemisphere extratropical stormtracks during the last 50 years. There is medium confidence in ananthropogenic influence on this observed poleward shift. It hasnot formally been attributed. There is low confidence in pastchanges in regional intensity. There is medium confidence thatan increased anthropogenic forcing will lead to a reduction in thenumber of mid-latitude cyclones averaged over each hemisphere,and there is also medium confidence in a poleward shift of thetropospheric storm tracks due to future anthropogenic forcings.Regional changes may be substantial and CMIP3 simulations showsome regions with medium agreement. However, there are stilluncertainties related to how the poorly resolved stratosphere inmany CMIP3 models may influence the regional results. In addition,studies using different analysis techniques, different physicalquantities, different thresholds, and different atmospheric verticallevels to represent cyclone activity and storm tracks result indifferent projections of regional changes. This leads to lowconfidence in region-specific projections.166