Over the past 350,000 years (and in fact the last 650,000 years), the carbon dioxide did not exceed 300 ppm, until the last 100 years or so. Some of the shorter bumps and wiggles between the warm and cool periods are accompanied by similar changes in carbon dioxide. This figure establishes that carbon dioxide and temperature are correlated on these long time scales. It does not establish cause and effect.
Here is the second figure, which has been routinely mentioned by people commenting on this blog. This is the figure of carbon dioxide and temperature change as revealed from the Vostok ice core. Vostok is in East Antarctica at about 78 degrees south. Figures like this have appeared in many papers, and an excellent reference for the figure is Petit et al. (Nature, 1999). In that paper you will see curves for methane, dust, sodium, etc. Recently the record has been extended back to more than 650,000 years.
First how is this measurement made? The basic idea is that air is trapped in snow and as more and more layers of snow accumulate, the air is help in bubbles under many meters of ice. Hence, the air in the bubbles is representative of the atmosphere at the time that the snow fell. Temperature is determined by the concentration of a particular isotope of oxygen, whose concentration is correlated with temperature. (How would you validate this measurement?)
Here is a version of the figure from the Koshland Science Museum. There is excellent material in their discussion of global warming, including a lot more about ice cores.
Figure 1: Carbon dioxide (parts per million, ppm) and temperature (Fahrenheit) from the Vostok Ice Core. Recently the record has been extended back more than 650,000 years ( Siegenthalter et al., 2005, Science )
What can we see from this figure? There is an oscillation between warmer and cooler states of the climate. The rise to the warm periods is more rapid than the descent into the cool periods. The carbon dioxide is higher in the warm periods and lower in the cool periods. Over the past 350,000 years (and in fact the last 650,000 years), the carbon dioxide did not exceed 300 ppm, until the last 100 years or so. Some of the shorter bumps and wiggles between the warm and cool periods are accompanied by similar changes in carbon dioxide. This figure establishes that carbon dioxide and temperature are correlated on these long time scales. It does not establish cause and effect.
There are other things that we can note about this figure based on some external information. First the great civilizations of humans only occur in the last 10,000 years or so. This is in a warm time, and humans have thrived in this warm period. The cool periods are categorized as ice ages, ages when glaciers expanded and covered large parts of the northern hemisphere. (As an aside, a year or so ago I attended a fascinating talk about the city of Sagalassos. Over the course of the seminar, there was a discussion about the people who lived in area before the city was formed. It struck me that they were living at the edge of a glacial lake that was first forming, then disappearing, as the glaciers receded. These people saw the end of the ice age.)
The periodicity is closely correlated with many of the orbital parameters of the Earth and the Sun. As many of you will note, this curve does not establish cause and effect. Further, higher scrutiny shows that the temperature increase begins before the carbon dioxide increase. This combination of the orbital parameters, the change in temperature, and the change in carbon dioxide provides a challenge for understanding. A plausible physical argument can be made that the greenhouse gases modulate, perhaps amplify, the radiative changes associated with the orbital parameters. Much of the change in the carbon dioxide would be associated with changes in ocean biology and chemistry. While it is straightforward to make physical arguments and to find correlative observations, the direct determination of cause and effect in these long-term climate fluctuations is difficult. (Here is adiscussion from Realclimate about these difficulties.)
These graphs from ice cores contain more information than discussed here; they are very rich. The curves are used by scientists and non-scientists to make statements about both past and future climates. In both cases it is important to remember that this curve does not stand in isolation. We have many other sources of information from both observations and theory. Plus, there is something very different today; the carbon dioxide is more than 30% higher.
