190
mohaline overturning in glacial times. The Manabe and Bryan low CO2
experiment was actually accompanied by reduced thermohaline overturning which they attributed to nonlinearity in the equation of state and reduced midlatitude heat loss through the expanded sea ice cover. Another
potentially important effect is the influence of ice sheets upon the zonal
distribution of precipitation. To our knowledge, the ice age atmospheric
simulations have not been analyzed to determine the net change in precipitation minus evaporation over the North Atlantic ocean and its drainage
basin.
We may look to a different mechanism to explain these glacial thermohaline oscillations. A more promising correlation seemingly exists between
cooling and thermohaline stability than between freshwater forcing and
thermohaline instability. The warm climates of the interglacial Holocene
and Eemian periods were apparently accompanied by steady thermohaline
circulation, while the intervening cold period exhibits instability nearly
throughout its duration. Atmospheric models have shown that the ice
sheets exert a powerful cooling influence upon the subpolar North Atlantic
(Kutzbach and Wright, 1985; Manabe and Broccoli, 1985). Anti-cyclonic
low level winds over the ice sheets funnel Arctic air into this region. The
strength of this circulation has been reported to be sensitive to the size
of the ice sheet (Kutzbach and Ruddiman, 1993). Bond et al (1993) have
noted that the D-O events are grouped into long term cooling cycles bracketed by Heinrich events. The reduction in size of the Laurentide ice sheet
after a Heinrich event is consistent with the subsequent appearance of the
warmest part of these "Bond cycles." It is noteworthy that the interstadials decrease in duration with the cooling in the Bond cycles - suggesting
a correlation between surface cooling and thermohaline instability within
the glacial period as well as between the glacial and interglacial periods. In
Section 7 we will address these problems by using some simple models to
investigate the way changes in thermal forcing affect the level of freshening
required to break down a steady thermally direct deep overturning.
Let us close this section by noting a distinct type of oscillation associated with a normal thermohaline overturning, which is often called a loop
oscillation. A salinity anomaly can be traced as it passes through the deep
overturning cell. The overturning is enhanced when it is in the polar region,
and retarded when it is in the low latitudes. When the salinity anomaly
dominates, it can grow or maintain itself against dissipation through surface fluxes by its effect on the overturning (advective feedback). When the
mohaline overturning in glacial times. The Manabe and Bryan low CO2
experiment was actually accompanied by reduced thermohaline overturning which they attributed to nonlinearity in the equation of state and reduced midlatitude heat loss through the expanded sea ice cover. Another
potentially important effect is the influence of ice sheets upon the zonal
distribution of precipitation. To our knowledge, the ice age atmospheric
simulations have not been analyzed to determine the net change in precipitation minus evaporation over the North Atlantic ocean and its drainage
basin.
We may look to a different mechanism to explain these glacial thermohaline oscillations. A more promising correlation seemingly exists between
cooling and thermohaline stability than between freshwater forcing and
thermohaline instability. The warm climates of the interglacial Holocene
and Eemian periods were apparently accompanied by steady thermohaline
circulation, while the intervening cold period exhibits instability nearly
throughout its duration. Atmospheric models have shown that the ice
sheets exert a powerful cooling influence upon the subpolar North Atlantic
(Kutzbach and Wright, 1985; Manabe and Broccoli, 1985). Anti-cyclonic
low level winds over the ice sheets funnel Arctic air into this region. The
strength of this circulation has been reported to be sensitive to the size
of the ice sheet (Kutzbach and Ruddiman, 1993). Bond et al (1993) have
noted that the D-O events are grouped into long term cooling cycles bracketed by Heinrich events. The reduction in size of the Laurentide ice sheet
after a Heinrich event is consistent with the subsequent appearance of the
warmest part of these "Bond cycles." It is noteworthy that the interstadials decrease in duration with the cooling in the Bond cycles - suggesting
a correlation between surface cooling and thermohaline instability within
the glacial period as well as between the glacial and interglacial periods. In
Section 7 we will address these problems by using some simple models to
investigate the way changes in thermal forcing affect the level of freshening
required to break down a steady thermally direct deep overturning.
Let us close this section by noting a distinct type of oscillation associated with a normal thermohaline overturning, which is often called a loop
oscillation. A salinity anomaly can be traced as it passes through the deep
overturning cell. The overturning is enhanced when it is in the polar region,
and retarded when it is in the low latitudes. When the salinity anomaly
dominates, it can grow or maintain itself against dissipation through surface fluxes by its effect on the overturning (advective feedback). When the