One dilemma has been the apparent observation that the increase in global warming may have preceded a rise in CO2 by about 800±600 years. This has now been largely explained as a snow to ice effect, in a new summary article in Science, Vol. 339, p. 1042-3, 1 March 2013, by Edward J. Brook of the College of Earth, Ocean, and Atmospheric Sciences of Oregon State University. The title of the article is Leads and Lags at the End of the Last Ice Age. Brook describes the work of F. Parrenin et al., Synchronous Change of Atmospheric CO2 and Antarctic Temperature During the Last Deglacial Warming in the same issue Science, Vol. 339, p. 1060-3 (2013).
First of all, this is not directly related to our current situation where our man-made rapid CO2 and other pollutant increase is driving increased radiative forcing, which is warming the planet. At the end of the last ice age about 18,000 years ago, it was supposedly the 100,000 year orbital cycle that provided the warming that increased the temperature initially. Then, secondary effects such as a warming ocean would have expelled dissolved CO2 and increased it in the atmosphere. If you don’t understand that effect, just open a warm carbonated soda can. Then, of course, the enhanced CO2 greenhouse gas would have further increased the warming.
The experimental effect is seen by comparing the temperature proxy in isotope ratios in ice, with the capture of air bubbles with CO2 into ice. The rare heavy oxygen isotope of O18 in a water molecule makes it heavier than water with both O16 isotopes. Then it is less likely to be evaporated. (Technical aside: this is a Boltzmann distribution effect.) As the water temperature increases at the surface, it becomes more likely to be evaporated, and the ration of O18 to O16 becomes the proxy. The water is evaporated at the warm tropics and makes its way to the poles, where it rains out and is frozen into the ice. The same is true of the Deuterium isotope of Hydrogen, which has a proton and a neutron, while the main Hydrogen isotope is just a proton. So a deuterated water molecule is one atomic mass heavier than an ordinary one, and its ratio to ordinary water is increased with increasing temperature. It also is a proxy for temperature found in ice cores.
In trapping air containing the CO2, the top snowpack or firn is porous, and can be exchanging air for thousands of years before it freezes air pockets into the ice. So the level of CO2 measured in the air pockets generally lags the collection of the isotope temperature measurement.
Parrennin et al. used the ratio of N15 to N14 in nitrogen molecules N2 in the firn column. While the molecules can still diffuse through the firn column, the heavier molecules with N15 isotopes settle more toward the bottom (as an exponential of their potential energy difference -(Δm g h)/kT, or Boltzman factor). This relates the observed enhancement of N15 at the bottom to the depth at which the CO2 is encapsulated in air. Knowing the depth, the CO2 at that depth can be compared with the ice collected at the top, since they were both fixed at the same time. Correcting for this in the comparison of CO2 content and temperature proxies show that the CO2 increase and temperature increases occurred at the same time to within the error of plus or minus 200 years.