Previous: Observations and Model Comparisons Next: References Up: Ext. Abst.
Discussion and Possible Mechanisms
The major areas of disagreement between current model estimates of the stratospheric response to solar variability on the 11-year time scale are: (1) the large differences between the observationally derived ozone response and model predictions in the middle and upper stratosphere; and (2) the evidence for a substantial lower stratospheric response component in both temperature and ozone that is not predicted by available models.
With respect to the upper stratospheric ozone response difference, it should first be noted that ozone at these lev- els is nearly in a state of photochemical equilibrium so that transport effects are less important. Assuming that current model treatments of ozone photochemistry are approximately valid, significant long-term changes in ozone concentration at these levels beyond that expected from solar ultraviolet forc- ing should therefore reflect changes in the concentrations of key trace constituents that determine the ozone catalytic loss rate. Evidence for such changes in the concentrations of up- per stratospheric CH4, NO2, and H2O (all of which indirectly influence the abundance of O3) has been obtained using data from the Halogen Occultation Experiment (HALOE) on the Upper Atmosphere Research Satellite (UARS) [17]. Using a simple photochemical model, Siskind et al. [18] showed that the observed upper stratospheric O3 decrease from 1992 to 1995 (approximately solar maximum to minimum) could be attributed to a combination of (i) decreases in odd oxygen production due to decreased solar flux; (ii) increases in total chlorine; (iii) decreases in CH4 (which affects the abundance of reactive chlorine); and (iv) increases in H2O (which affects the rate of O3 catalytic loss). The solar flux contribution was less than half of that resulting from catalytic loss rate trends.
In principle, long-term variations in tropical upwelling, and
hence in the supply of trace constituents affecting the O3 balance
in the upper stratosphere, may be forced either by vol- canic
aerosol variability or by solar variability, or both. Since increases
in lower stratospheric heating and tropical upwelling occurred
following the injection of Mt. Pinatubo aerosols in 1991, a gradual
decrease in upwelling rate could have occurred from this source
during the 1992 to 1995 period [17]. This could be a potential
cause of the observed decline in upper stratospheric CH4 over
the same period. On the other hand, if significant changes in
lower stratospheric heating occur be- tween solar minimum and
maximum such that the tropical upwelling rate is modified, then
this would be an alternate po- tential cause of the observed CH4
decline. Randel et al. [19] have recently analyzed additional
HALOE data through 1998 to show that the decreasing CH4 concentration
in the upper stratosphere reached an apparent minimum in 1996-1997
and may have begun to increase slightly thereafter. If this rep-
The Solar Component of Stratospheric Variability: Hood and Soukharev
resents, in part, a solar cycle variation of upper stratospheric
CH4, then such a variation would need to be incorporated in existing
models in order to accurately simulate the solar cycle variation
of ozone.
With respect to evidence for a substantial lower strato- spheric component of the 11-year response (e.g., Figure 3), one possibility that is not currently considered in stratospheric models is that the QBO may be modulated slightly by the 11-year solar cycle [20]. Statistical studies of the NCEP equa- torial wind data set demonstrate significant decadal variability of the QBO. However, it is not yet clear that this variability is solar-driven.
As shown in Figure 4, based on NCEP data for a 40-year period, near the 50 hPa level, some evidence exists for a longer duration of QBO westerlies and a higher amplitude of QBO easterlies under solar minimum conditions. Such a modula- tion would tend to increase the vertical wind shear during the transition to westerlies at levels above 50 hPa, thereby mod- ulating the QBO-induced meridional circulation in the sub- tropics. Specifically, during the transition to westerlies, the induced meridional circulation is characterized by upwelling in the subtropics that reduces the ozone column at these lati- tudes. An increased vertical wind shear under solar minimum conditions would result in deeper ozone minima at these times relative to solar maximum conditions.
As shown in Figure 5, there is some evidence for such a modulation of the column ozone minima in the northern subtropics. The solid line is the column ozone deviation from the long-term monthly mean while the dashed line is the NCEP 50 hPa zonal wind at the equator. It is seen that deeper ozone minima tend to occur in years when the easterly wind amplitude is larger and the deepest minima tend to occur near solar minima.
However, this possible solar modulation of the QBO- induced meridional circulation does not easily explain the larger column ozone amounts in the tropics as a whole at solar maximum as well as the associated higher tropical mean tem- peratures in the 40-120 hPa layer (middle panel of Figure 3). A resolution of these issues is required if general circulation models are to more accurately simulate the solar component of long-term climate change.