Figure 14.13. V ariations of the number of total detritic grains, percent volcanic glass (Icelandic tracer), and dctrital carbonate (Laurcntidc tracer) in core ODP 6()9 (Bond and I.otti, 1995). HE indicates I leinrich events. Letters a-g indicate smaller detrital events in the sediment, coinciding «ith the eold phase of D-O events recognized in the GRIP ice core.

Figure 14.13. V ariations of the number of total detritic grains, percent volcanic glass (Icelandic tracer), and dctrital carbonate (Laurcntidc tracer) in core ODP 6()9 (Bond and I.otti, 1995). HE indicates I leinrich events. Letters a-g indicate smaller detrital events in the sediment, coinciding «ith the eold phase of D-O events recognized in the GRIP ice core.

iceberg discharge, but deep convection occurred in areas that were not affected by the salinity decrease. Soon after the Heinrich event, the oceanic pattern became similar to that before the event (Vidal et al., 1997). In addition, benthic rccords from dccp-sea sediment cores raised from a depth shallower than 2 km exhibit light 5,sO peaks coinciding with Heinrich events. These peaks indicate a rapid transfer of ¿1H0-depleted melt water to depth. However, mcltwatcr has a low salinity and is unable to sink to great depth unless its salinity increases significantly. Paleo-oceanographers were therefore searching for a physical process allowing salinity increase for surface water without changing its ¿l80 value. Evaporation and advection of highly evaporated tropical water are therefore ruled out, because evaporation results in large <5,s() variations. The most likely process responsible for a transfer of ¿lsO-deplctcd meltwatcr to depth is sea-ice formation during winter, because freezing occurs without significant isotope fractionation and, therefore, without variation of the seawater <$lsO value. Katabatic winds would maintain the process activity and would inducc the formation of high-salinity surface water (brine), because sea ice is strongly depleted in salt. The more sea icc is formed, the saltier becomes surface water without <5,sO variation. Brine formation might have been particularly active along the Norwegian margin, where the <5lsO signals were also larger (Rasmusscn et al., 1996; Vidal et al., 1998).

Rapid changes in ocean circulation arc not limited to glacial conditions. During the last interglaciation, SST and SSS records from the North Atlantic Ocean and Nordic Seas show that summer SST and SSS decreased, sometimes rapidly, during the interval of minimum ice volume at high-latitude sites (>52 N), whereas they were stable or increased during the same period at low-latitude sites (3 I N to 41JN). This increase in meridional gradients of SST and SSS may have been caused by changcs in the latitudinal distribution of summer and annual-average solar radiation, and associated oceanic and atmospheric feedbacks (Cortijo et al., 1994; (Cortijo et al., 1999; Fronval and Janscn, 1996,1997). In addition, the end of the last interglaciation was marked by an abrupt event that begins the transition to a harsher climate within about 4(X) years at most (Adkins et al., 1997).

14.5 Concluding Remarks

The study of high-resolution dccp-sea sediment cores, together with that of Greenland and Antarctic ice cores, has demonstrated that the Earth has switched repeatedly and abruptly between cold and warm climates over the coursc of ice-agc cyclcs. These changes developed within a few decades. The palco-occanographic record of the North Atlantic Ocean revealed phenomena that may be related to changes in surface salinity much larger than the Great Salinity Anomaly in the 1970s (Dickson et al., 1988) during both glacial and interglacial conditions. These changes were closely correlated with sea surface temperature variations. Low-salinity events were associated with brine formation in the Norwegian Sea but with rcduccd convection in the North Atlantic, where surface waters remained stratified even during winter. These events were associated with low SST and low air temperatures over Greenland and the northern European continent. The atmosphere adjusted quickly to the perturbation, and most

North Atlantic climatic changes were felt throughout the Northern Hemisphere cither as a temperature or a hydrological cycle change. By contrast, warm episodes were associated with higher SST and convection in the Norwegian Greenland Sea.

It should be stressed that we still do not know whether the thermo-haline circulation has experienced significant changes during intcrglacial conditions, although some records suggest that this may he the case (Duplessy etal., 1992; Keigwin, 1996). It should also he stressed that intcrglacial climate variability is characterized primarily by major changes in the hydrological cycle (Gasse and Van Campo, 1994). As a consequence, the global warming induced by human greenhouse gas emissions might well result in precipitation changes over both continental and oceanic areas, possibly triggering lower North Atlantic surface water salinity. Such changes would reduce the mean flux of deep water formed every year during winter, the thermo-haline circulation, and the oceanic heat flux brought to the Norwegian Sea. This might deeply affect the European climate (Broecker, 1997; Manabe and Stouffer, 1995).

One of the most fruitful interactions between the modern and paleo scientific communities has been the study of multiple states of the thermo-haline circulation under various forcing factors. This characteristic feature of coupled ocean-atmosphere models would have been left aside without the numerous evidences brought by paleo-oceanographers and palcoclimatologists. Documenting the variability of the thermo-haline circulation on decadal-to-centennial time scales through the construction of high-resolution time series will lead to progress in our understanding of the thermo-haline circulation and will be useful for forecasting its future behavior.

However, we have discovered only a small part of the climatic puzzle. We are not yet able to explain the glacial-interglacial pCX)2 variations. We still don't know the complicated feedbacks that amplify North Atlantic circulation and SST changes to influence the Firth's climate on a large scale. How is it that millennial-scale variability appears nearly synchronously in California (Behl and Kennett, 1996) and in Indonesia (Linsley, 1996)? Or that the concentration of methane in the atmosphere (presumably a reflection of microbial activity on tropical land surfaces) varied along with other climatic variables (Stauffcr et al., 1998)? Or that Antarctic air temperatures sometimes varied nearly inversely with those in the Arctic (Blunicr et al., 1998)? These observations point to the involvement of the atmosphere and other oceanic processes, something that in turn raises the possibility that the North Atlantic oscillations were themselves the product of other, stronger sources of variability. In particular, the relationship between tropical and high-latitude oceans should be investigated. «--

Figure 14.14. Variations of the <5|H0 in the GRIP ice core, a proxy for air temperature variations (Dansgaard et al., 1993), compared with magnetic susceptibility, a proxy for bottom water current activity, and <5lsO variations of the planktonic foraminiferal species ;V. pachy derma in core ENAM 9321 (Rasmussen etal., 1996) and core SU 90-08 (Corti jo etal., 1997). Sl3C in benthic foraminifera in core NA 87-22 show major circulation changes coinciding with Heinrich events and low salinity periods in the North Atlantic (Vidal et al., 1997). I IE indicates I Icinrich events recognized in deep-sea sediments. Their position in the GRIP record has been inferred from a detailed comparison with North Atlantic sediment records (Bond et al., 1993).

15.2 Rapid Climatic Changes

It was only during the 1980s that the possibility of rapid climatic changes, occurring at the time scale of a human life or less, was fully recognized, thanks largely to the oxygen 18 record (a proxy for local temperature change) measured along the Greenland ice corc drilled at Dvc 3 in southern Greenland (Dansgaard et al., 1982, 1984, 1989). A possible link between such events and the mode of operation of the ocean was then suggested (Oeschger et al., 1984; Broecker et al., 1985; see also Broecker, 1997, for a recent review). The occurrence of rapid changes was fully confirmed by the central Greenland ice record (GRIP and GISP2), which allows a resolution approaching annual over the entire period of the last climatic transition (Figure 15.1). The return to the cold conditions of the Younger Dryas from the incipient interglacial warming

0 0

Post a comment