Karin Labitzke and her colleague Harry van Loon pioneered the study of QBO-Solar Cycle relationships, and Labitzke (2006) updates some of her earlier results on Solar Cycle-QBO relationships, to stress the interaction between the Northern and Southern Hemispheres, and to summarize the influence of the QBO on the solar variability signal as well as the influence of solar variability on the QBO itself throughout the year. It should be noted that an additional 19 years data since Labitzke (1987) are used in her analyses.
Her work indicates that the observed solar/QBO influences on the atmosphere involve a modulation of the Brewer-Dobson circulation. Further, it is found that a great deal of her observed signal results from the presence or absence of major midwinter warmings, and that the winter hemisphere influence gives results that extend over both the winter and summer hemispheres. Also, in agreement with Salby and Callaghan (2006), she finds evidence for a solar modulation of the QBO. Labitzke's (2006) analyses of stratospheric data give many observed characteristics against which theory and models can be tested.
Marie-Lise Chanin has been studying the thermal structure of the middle and upper atmosphere for more than four decades. Different sign effects have been seen in different atmospheric regions, and this was rather puzzling for a long time. In Chanin (2006), she gathers this information together and compares her observational results to the results of a mechanistic model to demonstrate that this observational evidence is consistent with the solar UV/planetary wave mechanism. She also notes that attempts to see the solar variability at the mesopause level are greatly complicated by the fact that the response changes sign near this level.
As mentioned previously, an important factor in obtaining the correct solar UV influences on climate is to include the correct solar UV-induced ozone changes. Both Haigh (1996) and Shindell et al. (1999) emphasized the importance of properly including solar UV changes in ozone because that magnifies the UV effect on temperature over what would exist when no ozone changes are considered, but Haigh (1996) and Shindell et al. (1999) used quite different ozone variations. In particular, the latitudinal variations of their ozone changes were different, and this leads to changes in the latitudinal temperature changes, which in turn results in different directly forced changes in the mean zonal winds. Calisesi and Matthes (2006) consider the problem of determining the solar UV induced ozone variations from observations. They conclude that there remains significant uncertainty in the altitude-latitude ozone response to solar variability that should be included in models for solar influences on climate.
In the introduction, it was indicated that the simple linear planetary wave model of Geller and Alpert (1980) implied that solar-induced changes in the zonal mean atmospheric state must extend down to the lower stratosphere to significantly affect the troposphere. The paper by Haigh and Blackburn (2006) addresses the tropo-spheric response to heating perturbations in the lower stratosphere. The direct solar UV effects on temperature and ozone are maximum in the upper stratosphere, so Gray et al. (2006) explore how the effects penetrate downward into the lower stratosphere. The work by Labitzke and her colleagues have shown the very important role of the QBO in solar effects on the troposphere/stratosphere system, and Salby and Callaghan (2006) summarize several of their earlier studies to see how this might occur.
Haigh and Blackburn (2006) report results from two types of numerical experiments. They used an idealized atmospheric general circulation model to explore what effects result from an imposed temperature perturbation on the stratosphere. It should be noted that the model they use has its top level at 18.5 hPa, so their perturbations are in the lower stratosphere. They find a banded response that very much resembles the solar 11-year response derived from observations. Furthermore, they found that the resulting zonal wind changes were primarily maintained by changes in the poleward eddy momentum flux. As an aid to interpreting these results, they conducted some transient spin-up experiments where these stratospheric heatings were suddenly imposed and the evolving responses were examined. Of course, given the limited vertical extent of the models used by Haigh and Blackburn (2006), the question remains by what means does the solar input, whose UV effect is maximum near the tropical stratopause, propagate down to the lower stratosphere?
Gray (2003) has shown, rather surprisingly, that even though the QBO has maximum amplitude at around 25-30 km, it is likely that the solar cycle effect on the QBO at higher levels at 40-50 km plays the largest role in modulating stratospheric sudden warmings. Gray et al. (2006) pursue these ideas further by reviewing their published work that has used observational analysis, mechanistic models, and general circulation models. A particularly interesting result is that they find that the timing of their modeled stratospheric warmings is sensitive to relatively small anomalies in the equatorial, subtropical, upper stratosphere in early winter.
Salby and Callaghan (2006) show statistical evidence that the solar effect first discovered by Labitzke (1987) is related to the solar modulation of the QBO period. They distinguish between the "linear" effect of changing UV radiation on the stratosphere where small changes are seen that maximize in the tropical upper stratosphere and "nonlinear" effects. The "linear" effects are relatively small and are physically straightforward to understand through solar UV heating, both directly and indirectly, by photochemically enhancing stratospheric ozone at these altitudes and latitudes, which increases the absorbed UV radiation. The larger effects that are seen in lower altitudes and higher latitudes are nonlinear and include two distinct physical mechanisms. These are planetary wave interactions with tropical winds that are affected by the direct solar effect and by solar modulations in the QBO period. Both of these latter two effects are imperfectly understood at the present time.
For instance, Gray et al. (2006) find that the QBO winds in the upper stratosphere are key in modulating the lower stratospheric planetary wave activity and the strength of the winter polar vortex. Statistical analysis of observations, mechanistic modeling, and general circulation modeling all indicate this to be the case. Yet, it is somewhat surprising to see that the upper stratosphere tropical winds seem to be more important than the lower and middle stratospheric winds in modulating stratospheric sudden warmings, with the accompanying large changes in the temperatures at high latitudes and the strength of the winter vortex. Perhaps, this just follows from the fact that the planetary wave Eliassen-Palm vectors are normally directed upwards and equatorward (see Andrews et al., 1987). It is also not clear how solar activity modulates the QBO period although some mechanistic model results by Cordero and Nathan (2005) and Mayr et al. (2006) give this result, albeit for different reasons.
Kodera (2006) discusses two different mechanisms by which the solar UV effects on the stratosphere might influence climate in the troposphere. Both start with the solar UV-planetary wave mechanism that has been referred to in previous Sections, which acts to reduce the Brewer-Dobson circulation at solar maximum.
This scenario can produce solar-climate effects in the troposphere in two ways. One is by the downward extension of the planetary waves themselves. Some evidence for this is that the strongest correlation between surface temperature and the F10.7cm solar flux is found in two regions - east of Japan and north of the Eurasian mountains. It is interesting that these two regions are also the two main regions from which planetary waves propagate upwards from the troposphere (see Plumb, 1985, Figure 5).
The other mechanism involves the solar modulation of the tropical upwelling of the Brewer-Dobson circulation. It is suggested that this might modulate equatorial convective activity. Kodera (2006) goes on to show statistical evidence for a solar modulation of the Walker circulation. Kodera (2006) closes with the remark that given that the total solar irradiance (TSI) shows so much less variation with solar activity than does the solar UV radiation, one might expect much less solar-induced variation in globally averaged temperature than in the regional changes that result from solar-induced changes in the circulation.
A new generation of climate models is appearing that extend up from the troposphere to the thermosphere. Schmidt and Brasseur (2006) show results from one of these models, the HAMMONIA general circulation model whose top is at 250 km. The HAMMONIA model is based on the ECHAM-5 climate model and includes fully interactive chemistry and radiation using the MOZART3 chemistry scheme. They point out that this type of a model extends in altitude where the atmosphere absorbs those portions of the solar spectrum that show large variations with the solar cycle, whereas most climate models do not. Also, the HAMMOMIA model includes physics appropriate to the upper atmosphere that traditional climate models lack. Their modeling results for the solar UV resemble observations, but there are also significant differences between their modeled ozone and temperature responses from those derived from observations.
It should be noted that Giorgetta et al. (2002) were successful in modeling the QBO with the ECHAM-5 model, but at the present time, the HAMMONIA model lacks the vertical resolution needed to do this. Given the apparent importance of the QBO in solar influences on the atmosphere, it will be interesting to see how results might change once there is a QBO in HAMMONIA. Also, if the HAMMONIA model were formulated in a similar manner to what was done to successfully model the QBO in Giorgetta et al. (2002), this would better resolve the planetary wave interactions that are essential for the solar UV/planetary wave mechanism.
An important point made by Schmidt and Brasseur (2006) is that their modeled solar response has a great deal of longitudinal structure, which might not be treated properly in analysis of satellite observations. It will be very interesting to analyze the HAMMONIA model, and other similar models, for transient simulations of varying solar flux, including volcanoes, the QBO, and changing atmospheric composition, in the same way as satellite data analyses have been carried out to see if the statistical analyses that have been used properly separate these modes of variation for the length of the data record that has been analyzed.
Was this article helpful?