Rapid Temperature Fluctuation and Recovery

The most recent time the Earth was covered in glaciers (12,000-115,000 years ago, roughly), there are 25 periods where the temperature in the northern hemisphere shot up by up to 5 degrees in a matter of decades.

This much temperature change is a really huge deal; it would cause widespread famines and water shortages if it happened today. Warmer temperatures would improve life for some areas (Siberians and Alaskans might not complain too much!), but humans have built in such a way that optimizes for local resources. It would take time to adjust agricultural production to Alaska (Operation Canadian Freedom also feels likely); during that time, prices would skyrocket, and people would starve. (That also assumes the ground soil can be made to work; after thousands of years of glaciers, there probably isn’t much topsoil.)

What’s the Evidence

Ice cores from Greenland mark the Dansgaard-Oeschger events. Heinrich events are supported by North American marine sediment layers, and correspond to 6 of the coldest intervals between D-O events. I think the hypothesized interaction is that 6 of the D-O events were shortened or interrupted by some climactic event that swung the temperature seesaw towards cold.

Criticisms of Ice Core Samples

Critical evaluation of the sources of evidence is crucial. The existence of D-O events is entirely based on ice core samples from Greenland.

• Temperatures are not well mixed; an event that makes Greenland 5 degrees warmer might not affect the rest of the world. ** Counter-evidence: There have been some efforts to connect Greenland ice sheets to the Antarctic, although I’m not sure of their applicability here.
• “Ice Core Sampling” measures variation in CO2 and CH4. This variation is correlated with freeze/thaw cycles as well as temperature. One can imagine an event where the earth itself heats up to melt ice in Greenland, triggering a freeze/thaw cycle to spike CO2/CH4 levels in ice sheets. This would not necessarily affect air temperature, nor would it imply global temperature change.

Causes

We really don’t know. There is some evidence they occur on a multiple of 1,470 years, but this seems like overfitting. One credible hypothesis is from NOAA’s site:

The cause of these glacial events is still under debate. During the last glacial time, large ice sheets rimmed the North Atlantic. At certain times, these ice sheets released large amounts of freshwater into the North Atlantic. Heinrich events are an extreme example of this, when the Laurentide ice sheet disgorged excessively large amounts of freshwater into the Labrador Sea in the form of icebergs. Scientists have hypothesized that these freshwater dumps reduced ocean salinity enough to slow deepwater formation and the thermohaline circulation. Since the thermohaline circulation plays an important role in transporting heat northward, a slowdown would cause the North Atlantic to cool. Later, as the addition of freshwater decreased, ocean salinity and deepwater formation increased and climate conditions recovered.

For Heinrich events (extreme cold in between 6 of the D-O periods), there is corresponding evidence found in deep sea sediments. Evidence for the Dansgaard-Oeschger Events is more ambiguous.

It does appear likely these rapid temperature fluctuations occurred, at least in Greenland. While we don’t know what caused them, it’s interesting to learn how climate scientists infer climate history from indirect evidence, and to understand the limitations and possible incorrect beliefs about the conclusions drawn.

More information is available on the Wikipedia page or the NOAA website

I also recommend you check out the XKCD comic that started this exploration.