Carbon Dioxide and Global Temperature: To lag or not to lag

 One of the persistent puzzles about the ice ages is the appearance of an 600 to 800-year lag between Antarctic temperatures and atmospheric CO₂ levels.

Image from http://www.skepticalscience.com/co2-lags-temperature-intermediate.htm

The argument goes that if CO₂ lags global temperature, how can CO₂ be a primary driver of temperature?  The general model looks like this:

1. A warming phase of the Milankovitch cycles caused more direct sunlight to fall on the northern high latitudes ~19,000 years ago (Shakun et al. 2012), starting the rise in global temperatures.
2. That triggered melting in the Northern Hemisphere, with large amounts of freshwater flooding into the Atlantic Ocean and disrupting the Atlantic Meridional Overturn Circulation.
3. That disruption causes a seesawing of heat between the hemispheres, ending when the Southern ocean starts warming ~18,000 years ago.
4. The Southern ocean also warmed the atmosphere slightly, starting ~18,000 years ago.
5. Warm water holds less CO₂ than cold, so CO₂ was released as ocean temperatures rose (Martin et al. 2005), with the release starting ~17,500 years ago (Shakun et al. 2012).
6. The increase in CO₂ amplified the warming from the Milankovitch cycle, warming the atmosphere even more and spreading the warmth from the Southern Hemisphere to the Northern Hemisphere, triggering full deglaciation and a further temperature rise of 3.2ºC from ~17,500 years ago to ~12,000 years ago.
7. The process stopped when the ocean/atmosphere system reached equilibrium with respect to CO₂ and a particular insolation level

Previous research showed that atmospheric warming in the Southern Hemisphere preceded CO₂ by 620 years whereas CO₂ rose before Northern Hemisphere atmospheric temperatures by 720 years and the overall global average by 460 years (Shakun et al. 2012).  However, newer research (Parrenin et al. 2013) shows that the apparent 620-year lag between the Southern Hemisphere air temperatures and CO₂ is due to a mistake in how the age of air bubbles within Antarctic ice was calculated—and that once the age was corrected, the lag largely disappears, with the rise in CO₂ now within the margin of error of the rise in Antarctic air temperature.  In fact, Perrenin et al. noted that CO₂ may actually lead the rise in Southern Hemisphere air temperatures.  In that light, the revised model would look like this:

1. A warming phase of the Milankovitch cycles caused more direct sunlight to fall on the northern high latitudes ~19,000 years ago (Shakun et al. 2012), with global air temperatures rising by 0.3ºC from 19,000 years ago to 17,500 years ago.
2. That triggered melting in the Northern Hemisphere, with large amounts of freshwater flooding into the Atlantic Ocean and disrupting the Atlantic Meridional Overturn Circulation.
3. That disruption causes a seesawing of heat between the hemispheres, ending when the Southern ocean starts warming ~18,000 years ago.
4. Warm water holds less CO₂ than cold, so CO₂ was released as ocean temperatures rose (Martin et al. 2005), with the release starting ~18,000 years ago, simultaneous with the rise in Southern Hemisphere air temperatures.
5. The increase in CO₂ amplified the warming from the Milankovitch cycle, spreading the warmth from the Southern Hemisphere back to the Northern Hemisphere, triggering full deglaciation and a further global air temperature rise of 3.2ºC from 17,500 years ago to ~12,000 years ago.
6. The process stopped when the ocean/atmosphere system reached equilibrium with respect to CO₂ and a particular insolation level

Even before the Antarctic lag was corrected, Shakun et al. (2012) found that 91% of the total warming (3.2ºC out of a total of 3.5ºC) at the end of the last ice age came after CO₂  levels rose, with both global and Northern Hemisphere air temperatures lagging behind the rise in CO₂ and only Southern Hemisphere air temperatures (mostly Antarctica) leading the increase in CO₂.  Perrenin et al. (2013) now shows that CO₂ and Antarctic air temperatures rose together, so close in time that it's difficult to discern which came first, resolving the puzzle of why CO₂ appeared to lag Antarctic air temperatures.

The only puzzle left?  When "skeptics" will update their talking points.  I'm guessing "never."