Figure 14,1. SST difference between Last Ulacial Maximum and modem conditions for Northern Hemisphere winter and summer (CI JMAP, 1981),

Recently, several studies suggested that the CLIMAP reconstruction may be erroneous hoth in very cold (Norwegian Sea) and very warm (tropical ocean) areas. Using a new transfer function and a large set of Norwegian-Greenland sea cores, Weinelt (1996) suggested that SS'ls were higher than those reconstructed by CLIMAP and that these seas were ice-free during glacial summer. These revised SS I estimates are probably too high, because the transfer function used in this stud) does not allow us to estimate accurately very low SST (see Weinelt's Figure 4, page 289), hut the high abundance of foraminifera in deep-sea sediments deposited during the last glaciation supports the hypothesis of seasonally ice-free Nordic seas, allowing a high summer foraminifera! productivity.

In the Tropics, fossil populations of foraminifera exhibit only small variations during the last climatic cycles. They were interpreted as an indication that the tropical ocean experienced only a minor cooling during the last glaciation. This result is challenged by selective measurements performed on corals: Measurements on a fossil coral indicate that 10,300 years ago mean annual SSTs near Vanuatu in the southwestern Pacific Ocean were about 5 C cooler than today and that seasonal variations in SST were larger (Heck et aL, 1997). Similarly, ls0/If,0 and Sr/Ca measurements performed in Atlantic corals collected along t he slopes of Barbados indicate that SSTs were 5 C cooler than today 18,(KM) 19,000 years ago, during the LGM (Guilderson et aL, 1994). Such significant coolings during the glaciation are in agreement with several continental evidences, noticeably the glacial depression of snowlines in tropical mountains (1 Iawaii, Colombia, East Africa, and New Guinea) and vegetation changes (Bonncfille et al., 1992; Bonnefille et al., 1990; Rind and Peteet, 1985). However, it should be stressed that most continental vegetation changes in the Tropics may also be explained by the low CO2 content of the atmosphere and a minor air cooling (Jolly and Haxeitine, 1997), In addition, the validity of the Sr/Ca paleothcrmomcter, which is generally used for detecting a large SST drop in the Tropics, is not yet fully established. Other comparisons reiving on micropaleontological transfer functions or alkenone paleothermometry show a good agreement between low-altitude-continental and marine-shelf temperature estimates, and both indicate minor tropical cooling (Bard et al., 1997; Rostek et al., 1993; Van Campo et al., 1990). I Iowever, marine SST estimates, which were generated at proximity of the African continent (Van Campo et al., 1990), are slightly colder than those of CLIMAP, which were obtained in the open ocean (CLIMAP, 1981). This suggests that foraminifera may be faithful indicators of past SSTs but that the sediment core coverage originally used by CLIMAP should be enlarged in tropical oceans in order to obtain a precise reconstruction.

In conclusion, the CLIMAP (1981) reconstruction still provides the best maps of SST available for the LGM, although some inaccuracies probably exist. Model simulations have been performed with tropical SSTs colder than CLIMAP estimates as boundary conditions. These simulations exhibit land climates colder than those computed with CLIMAP SSTs (Webb et aL, 1997). This result has been put forward in support of a revision of the CLIMAP reconstruction of the Last Glacial Maximum, However, this argument rests on a very simple interpretation of continental vegetation changes, an interpretation that does not take into account the impact of the low atmospheric CO2 content. Meaningful results could be obtained only by coupling atmospheric general circulation models with models that include enough physiology to compute the response of the continental vegetation to atmospheric CO> changes. Sensitivity experiments have also been made to simulate the impact of the uncertainties of the CLIMAP reconstruction on our understanding of the glacial ocean circulation. They are discussed later, wf w

0 0

Post a comment