Carbon dioxide/circulation: Models indicate the possibility that CO₂ has an impact on both past and future ozone. There are two effects; cooling of the stratosphere and changes in the residual circulation. These were described in detail in Li et al. [2009] and elsewhere. The upper stratosphere should be dominated by cooling, which slows the ozone loss rates leading to an increase in ozone. The lower stratosphere is dominated by a change in the residual circulation, which results in a decrease of ozone in the tropics and an increase of ozone in the mid and high latitudes (at least according to models).

The results from Li et al. [2009] indicate that ozone should increase in the upper stratosphere (red curve) as CO₂ cools that region leading to slower ozone recombination. The effect is nearly independent of latitude. On the other hand, the lower stratosphere is expected to have a strong latitudinal signature to increased CO₂ with increases in ozone at mid and high latitudes and decreases in the tropics. This is due to an increase in the speed of the Brewer-Dobson circulation in the model. Other models get a similar increase in the circulation, but the magnitude and latitude signature vary. In the Li et al. [2009] results, the total column ozone amount in the tropics shows a slight decrease due to the net effects of CO₂. It should be noted that the circulation speed-up in the model is not directly due to absorption of radiation in-situ by CO₂. Rather, the speed-up is an indirect effect of the warming of sea-surface temperatures by the CO₂. Mark Olsen got this result when he ran constant CO₂ in a CTM, but used time-dependent measured SSTS (I don't know if this fact appeared in any publication).

1) In the upper stratosphere, the cooling should have decreased the trend that one would expect due to chlorine pre 1996. The cooling should be increasing the trend post 2000. Thus, when fitting to EESC, the ratios of downward and upward trend are modified. When fitting independent linear trends to the two periods, the expectation of the ratio of these trends is modified. To me, the important question is whether ozone concentrations behave as we would expect from the provisions of the Montreal Protocol. Therefore, I like to either use the EESC or interpret the linear trends in terms of this expectation of ratios.

2) The lower stratosphere is quite different. The models all indicate that the residual circulation should be speeding up. The search for a clear demonstration of this prediction has been fruitless thus far. The predicted change of circulation impacts the expectation for ozone trends. The tropical lower stratosphere may see no recovery as the circulation speed-up continues to bring low ozone concentrations into the tropical lower stratosphere even as chlorine is recovering. For the mid-latitudes, the implication is that the ozone “recovery” may be faster than predicted for a pure chlorine decrease.