Most of the Antarctic ice sheet is a high polar desert receiving no more than an inch or two of snowfall each year, the polar equivalent of the arid core of the Sahara. Ice cores from such locations cannot capture the detail needed to reveal the short-term variations in CO2 and other gas and solid constituents that occur over centuries or decades. But along the lower margins of the ice sheet, snowfall is heavier, and in favorably sheltered sites it is blown into thicker piles by strong winds. In these places, more detailed records can be recovered. Two such locations in Antarctica yielded the high-resolution records of the last 2,500 years shown in figure 12.1.
The shading shows a projection of the long-term increase in CO2 over the last 8,000 years that I attribute to human activities. Not long after 2,000 years ago, that rising trend slowed and then was interrupted by several dips in CO2. The two ice-core records in figure 12-1 disagree in places but indicate two or perhaps three CO2 minima: a broad but not very deep one from about AD 200 to 600, a short one near 1300-1400, and a final minimum that is both deep and broad from 1500 until about 1750.
One reason the CO2 measurements from the two ice cores disagree in the area of overlap is dating error. Of the two records shown, the one from Law Dome is superior: it contains layers of very fine ash particles deposited by volcanic explosions of known age. It was also deposited at a faster rate than the core from Taylor Dome. In any case, these records agree on one important point: negative CO2 oscillations as large as 4 to 10 parts per million occurred within the last 2,000 years. Processes within the ice can smooth the true CO2 record and reduce its amplitude, but it is difficult to create negative oscillations where none actually existed.
What could explain these drops? The most obvious explanation is that they are tied to natural changes in climate. In the last decade, one of the most exciting areas of climate research has been the discovery of abrupt oscillations that occur over decades, centuries, and millennia, intervals that are all much shorter than changes in Earth's orbit. For most of Earth's ice-age history, these short-term variations must have had a natural origin because they occurred well before any possible human effect on climate. When ice sheets were present, these oscillations tended to be very large: short-term temperature changes across Greenland and the North Atlantic amounted to as much as one-third of the difference between o
O Taylor Dome
Was this article helpful?