A variety of climate forcings are now thought to be able to influence planetary wave dynamics in the troposphere by affecting the propagation of planetary waves out of the troposphere. However, this propagation pattern is sensitive to the details of the corresponding zonal wind changes. Here we discuss two forcing mechanisms that alter zonal winds and subsequent tropospheric responses: changes in atmospheric CO₂ concentrations, and solar forcing in conjunction with the QBO. Increased atmospheric CO₂ concentrations can be shown to influence planetary wave refraction so as to produce an intensified residual circulation in the subtropical lower stratosphere (which increases transport of tropospheric species into the stratosphere). In our GCM experiments, the low latitude response appears qualitatively robust over a wide range of tropical warming magnitudes, although the quantitative circulation change depends upon the degree of tropical warming as influenced by convection and cloud cover changes; it varies by a factor of three with a factor of three change in tropical warming. At higher latitudes, this equatorward planetary wave refraction has been associated with an increase in the high phase of the Arctic Oscillation. In the model experiments, the extratropical response depends upon the magnitude of both low and high latitude warming in the troposphere; with SST and sea ice changes that result in a weaker Hadley Cell and greater high latitude warming, the Arctic Oscillation phase change may be negative.
The QBO alters the latitudinal gradient of the zonal wind in the stratosphere, and solar heating, in association with ozone response, alters the vertical gradient of the zonal wind. Both gradients affect the refractive properties of planetary waves uniquely for each individual combination of tropical east/west winds and solar maximum/minimum activity. In the model, when we consider solar maximum compared to solar minimum conditions, the east (west) phase of the QBO results in a relative high (low) phase of the Arctic Oscillation with corresponding temperature changes. Observed and modeled surface air temperature variations calculated between the solar cycle extremes in the different QBO phases are similar in magnitude to those derived from regression of monthly data on the AO, both being on the order of observed interannual variations.