Page 290 - Earth's Climate Past and Future

266 PART IV • Deglacial Climate Changes

the icebergs would melt and raise sea level. A rise in sea IN SUMMARY, natural internal oscillations within ice
level could then destabilize other coastal ice margins by sheets could plausibly have played a role in at least
floating those ice shelves off bedrock pinning points the larger millennial oscillations but probably not in
and causing them to surge into the ocean as well. the smaller ones.
A key question about this idea is whether or not the
millennial-scale rises in sea level were large enough to
link all the ice sheets. The amount of sea level change 14-8 Greenhouse-Gas Forcing
during these oscillations has been estimated by two Because greenhouse gases play a major role in climatic
methods. Variations in sea level based on coral reefs changes at tectonic and orbital time scales, their behav-
along the slowly uplifting coast of New Guinea indicate ior during millennial oscillations is worth considering.
that sea level changes could have been as large as How large were the millennial-scale oscillations in CO
10–15 m during these major oscillations. In addition, and methane, and do they indicate a forcing or feedback 2
18
marine δ O signals recorded in the shells of benthic role for the gases?
foraminifera from the deep tropical Pacific Ocean show Methane concentrations in ice cores show clear
millennial-scale variations of 0.1‰ or slightly larger millennial oscillations (see Figure 14–8). Because these
during the same intervals. The coral reef and δ O changes lag a few decades behind δ O (temperature)
18
18
evidence both point to sea level changes of 10 m or fluctuations, they appear to have been the result rather
slightly more during the largest millennial oscillations than the cause of the temperature oscillations. The
(the Heinrich events). Smaller shorter-term variations changing methane concentrations would have acted as a
in Pacific δ O values permit sea level changes of only a positive feedback to millennial-scale temperature
18
few meters during the smaller Dansgaard-Oeschger changes that had begun for other reasons. Methane
oscillations. production and release in the wetlands of northern Asia
The massive North American ice sheet is the best
candidate for causing sea level rises large enough to (Siberia) were particularly susceptible to changes in air
temperature over millennial intervals.
trigger reactions in the other ice sheets during major Ice core records are less clear about millennial-scale
ice-rafting events. Yet the evidence in North Atlantic CO changes. Measurements in high-resolution ice
2
sediments shows that the first sand grains deposited cores from Greenland are suspect because the CO in
during the major ice-rafting pulses came from the the air bubbles interacts chemically with CaCO dust
2
3
smaller ice sheets on Iceland and in coastal regions in the ice. In slowly deposited ice from Antarctica,
farther north. Only later did floods of debris arrive in millennial-scale events are too brief to be recorded in
icebergs originating in North America. This sequence full amplitude because of the long time delay in sealing
appears to rule out the North American ice sheet as the air bubbles in the ice. The best records currently
the initial sea level trigger during large ice-rafting available suggest that some CO oscillations may have
events. been as large as 10 ppm or more, but the records are
2
Changes in the sizes of the northern ice sheets
during the shorter-term millennial oscillations were noisy and the patterns are not clearly developed.
smaller and probably produced sea level fluctuations of
just a few meters. It is more difficult to argue that these IN SUMMARY, the role of CO changes in millennial-
2
small changes could have provided the link between the scale oscillations remains unclear. Future drilling on
northern ice sheet margins. the lower flanks of Antarctica should help to clarify
Another possibility is that the ice streams that deliv- the amplitude and timing of the CO oscillations.
2
ered icebergs to the Atlantic Ocean pulled enough ice
out of the interior of the North American ice sheet to 14-9 Other Natural Interactions in the
alter atmospheric circulation. Climate changes over
orbital time scales have been interpreted as a response Climate System
to splitting of the jet stream because of changes in the Another proposed explanation is that the millennial
elevation of the North American ice sheet (see Figure oscillations were produced by natural interactions
12–13B). Some climate scientists have suggested that within key components of the climate system in the
the same explanation might work for the shorter Greenland/North America area, such as the ice sheets,
millennial-scale changes in response to ice volume the surface ocean, and the deep ocean. At orbital time
changes of 1% to 10%. For this explanation to be scales, changes in this region influence temperature and
viable, atmospheric circulation would have to have been precipitation across a broad area of the northern hemi-
extremely sensitive to small changes in the elevation of sphere (see Figures 11–2, 11–3, and 11–4).
the North American ice sheet, perhaps because of the An initial proposal was that a natural oscillation
existence of a critical threshold. existed between the sizes of the (large) northern ice