gfdl's home page > about us > FY02 milestones > 1st quarter FY02 NOAA/OAR milestone
1st Quarter FY02 NOAA/OAR Milestone
Evaluate the model performance of ice age simulations in partnership with the UKMO Hadley Center
R.J. Stouffer and A. J. Broccoli
How cold was the ice age? An international research team, consisting of Anthony Broccoli and Ronald Stouffer of NOAA/GFDL and Chris Hewitt and John Mitchell of the Hadley Centre for Climate Prediction and Research in Bracknell, United Kingdom, are answering this question using a sophisticated model of the coupled atmosphere-ocean system. Their simulation of the climate of the last glacial maximum, which occurred approximately 21,000 years ago, is the first in which coupled, three-dimensional models of the atmosphere and ocean have been run long enough (more than 1000 model years) to produce a relatively stable climate.
The ice age climate response shows the familiar polar amplification of the temperature changes, in which the high latitudes cool more than the lower latitudes (Fig. A). The largest cooling occurs in those portions of North America and Europe covered by continental ice sheets during the ice age, where simulated temperatures are more than 20°C colder than the present day. Tropical regions cool by an average of about 2°C, with the cooling of tropical land enhanced relative to that of the tropical oceans. Over most of the surface of the globe, ice age temperatures are colder than present day. An interesting exception to this pattern occurs in the North Atlantic Ocean, where little or no ice age cooling is simulated in some regions. Why?
The relative warmth of these North Atlantic regions results from a change in the thermohaline circulation (THC), which is the global overturning mechanism by which water sinks in the North Atlantic Ocean and is replaced by the northward flow of warmer, salty water at the surface. This overturning circulation intensifies in the ice age simulation, because changes in atmospheric flow extract large amounts of heat from the region south of Iceland and west of the British Isles, leading to the enhanced formation and sinking of cold, dense waters. A strengthening of the northward flow at the ocean surface accompanies the increase in sinking, and the more vigorous infusion of warm water compensates for the effects of atmospheric cooling. As a result, limited areas in the North Atlantic do not share in the widespread cooling that prevails over the remainder of the earth's surface.
Further evidence for the importance of the change in the THC comes from another ice age simulation, in which the dynamical ocean model is replaced with a simple ocean that assumes no changes in oceanic heat transport from the present day. Over most regions of the globe, the temperature changes simulated in this simpler model (Fig. B) are similar to those obtained from the fully coupled model. However, the temperature response in the North Atlantic Ocean region in the two models is quite different, revealing the effects of the changes in oceanic heat transport associated with the stronger THC.
Because human-induced changes in greenhouse gases are also projected to affect the THC, the knowledge gained from this study will allow modelers to better understand the changes in climate that may result. These results are also of keen interest to paleoceanographers, as there is no firm consensus about the strength of the THC during the last ice age.
Figure Captions
Figure A. Difference in annual mean surface temperature (°C) between ice age and modern simulations using the coupled atmosphere-ocean model, HadCM3.
Figure B. Difference in annual mean surface temperature (°C) between ice age and modern simulations using the atmosphere-mixed layer ocean version of the model, HadSM3. This model is identical to HadCM3, except that a simple ocean model is substituted for the dynamical ocean model.
Nonlinear color scale at the bottom of the figure indicates temperature difference in degrees C.