The Gulf Stream is one of the most enduring features of the North Atlantic circulation, and was well known to ships' masters sailing between Europe and America. On their return voyage to Europe, they sailed with this warm current through the Straits of Florida. They followed it to about 35° N and then headed for Europe along with the westerly winds (Fig. 1). On page 644 of this issue1, Lynch-Stieglitz et al. show that during the Last Glacial Maximum, some 21,000 years ago, the speed of the Gulf Stream was much lower than it is today. Cold climates, therefore, seem to coincide with the reduced transport of warm water to the high latitudes of the North Atlantic.
The Gulf Stream consists of warm water arriving from the equatorial and tropical regions through the Gulf of Mexico. It is driven by the Trade Winds, which cause sea levels to be higher in the western part of the Atlantic basin and generate an area of high water pressure centred on the Sargasso Sea, comparable to the anticyclonic cells of the atmosphere. Just like winds, water tends to flow from areas of high pressure to areas of low pressure. However, the Coriolis force, caused by the Earth's rotation, deflects the flow to the right in the Northern Hemisphere. Ocean currents, therefore, describe a clockwise gyre in the tropical Atlantic, and as the higher sea level is found in the west, the current that flows along the American coast is narrow and fast.
The current is particularly intense and deep in the Florida Straits, through which water leaving the Gulf of Mexico is forced to flow. The narrowest part of the straits is between Florida in the west and the Bimini Island in the east, where it forms a 80-km-wide trench with maximum depth of 800 m. The speed of the current can reach up to 1.5 m s−1 in the strait, and the Coriolis force tends to push the water towards Bimini. This force creates a local pressure and density gradient, which results in a strong tilt in the surfaces of constant temperature and density (see Fig. 2 on page 645). The faster the current, the stronger the tilt.
Lynch-Stieglitz et al.1 use the oxygen-isotope composition of bottom-dwelling foraminifera in a set of cores collected on both sides of the Florida Straits to show that the tilt of the surfaces of constant temperature or density was significantly less during glacial conditions than it is today. Both the 18O content of foraminiferan shells and seawater density increase as a result of increasing salinity and decreasing temperature, and they are linked by an empirical relationship. Lynch-Stieglitz et al. established a similar relationship for the glacial ocean, taking into account the salinity and 18O increase due to build-up of continental ice-sheets. Their isotope measurements indicate only a small temperature and density contrast in the Florida Straits during the Last Glacial Maximum, and they therefore infer that the Gulf Stream was 35% weaker then.
The observation suggests that at that time ocean circulation was sluggish. The Gulf Stream is part of the global ocean circulation (the conveyor belt), which in the Atlantic today carries warm surface water northwards and cold, dense deep water southwards. The net result of the conveyor belt's operation is the transfer of heat, not only from equatorial regions to high northern latitudes, but also from the Southern to the Northern Hemisphere. A huge amount of heat is released to the atmosphere as the conveyor belt's surface water cools and sinks to the abyss, and it keeps Europe much warmer than Canada at the same latitude.
During the Last Glacial Maximum, the Earth's climate was much colder than today and the strongest cooling (of 10 °C or more) was in the North Atlantic and over Europe2,3. Geochemical and sedimentological studies of deep-sea sediments show that the production of North Atlantic Deep Water, the deep southward-flowing limb of the conveyor, was strongly reduced during glacial conditions4,5,6. Computer models have reproduced the main features of the glacial Atlantic circulation and have shown that the annual mean meridional heat transport was reduced by about 30% at low latitudes and close to zero north of 45° N, allowing the cold polar winds to blow unabated over Europe7. Lynch-Stieglitz et al. provide the first evidence that the low-latitude surface branch of the conveyor belt was less active during glacial conditions, and that the weaker Gulf Stream resulted in a large reduction in both northwards and interhemispheric transport of heat.
Obviously, however, determination of past oceanic currents by the tilt of surface of constant density is limited by certain assumptions. First, such reconstructions assume that wind patterns were not significantly different in the past. Atmospheric models driven with glacial boundary conditions support this hypothesis8,9, but they do not apply to the remote past, when the geography was different from today. Second, reliable estimates of water transport require large variations in density, which are found only in warm upper waters. The density contrast is much smaller in the cold, deep ocean and cannot be determined accurately given the experimental errors in the measurements of oxygen isotope ratios in foraminifera.
Nevertheless, the method pioneered by Lynch-Stieglitz et al. shows great promise for the reconstruction of past patterns of oceanic water transport, in particular during the sudden fluctuations in temperature that occurred every 1,000 years or so during the last glacial period. There seems to be little doubt that these abrupt changes are due to massive iceberg discharges and changes in the Atlantic's conveyor10,11, although the associated changes in ocean circulation and northward heat transport are still poorly understood. Nevertheless, there is no doubt that understanding the behaviour of the Gulf Stream during these events will shed new light on the mechanisms of climate change, which may well act in the near future.
Lynch-Stieglitz, J., Curry, W. B. & Slowey, N. Nature 402, 644–648 (1999).
CLIMAP Geol. Soc. Am. Map and Chart Ser. MC-36 (Geol. Soc. Am., Washington DC, 1981).
Guiot, J., Pons, A., de Beaulieu, J. L. & Reille, M. Nature 338, 309–313 (1989).
Duplessy, J.-C. et al. Paleoceanography 3, 343–360 (1988).
McCave, I. N., Manighetti, B. & Beveridge, N. A. S. Nature 374, 149–152 (1995).
Sarnthein, M. et al. Paleoceanography 9, 209–267 (1994).
Fichefet, T., Hovine, S. & Duplessy, J.-C. Nature 372, 252–255 (1994).
Lautenschlager, M. & Herterich, K. J. Geophys. Res. 95, 22547–22557 (1990).
Kutzbach, J. et al. Quat. Sci. Rev. 17, 473–506 (1998).
Oppo, D. W. & Lehman, S. J. Science 259, 1148–1151 (1993).
Vidal, L. et al. Earth Planet. Sci. Lett. 146, 13–26 (1997).
About this article
Paleoclimatic inference of the mid-Holocene record of monk seal (Monachus monachus) in the Cantabrian Coast
Proceedings of the Geologists' Association (2011)
Estimates of the mortality and the duration of the trans-Atlantic migration of European eelAnguilla anguillaleptocephali using a particle tracking model
Journal of Fish Biology (2009)
Calcium carbonate corrosiveness in the South Atlantic during the Last Glacial Maximum as inferred from changes in the preservation of Globigerina bulloides: A proxy to determine deep-water circulation patterns?
Marine Geology (2004)
Proceedings of the Royal Society of London. Series B: Biological Sciences (2003)
Impact of the Azores Front on the distribution of planktic foraminifers, shelled gastropods, and coccolithophorids
Deep Sea Research Part II: Topical Studies in Oceanography (2002)