Michael E. Summers*
It is now well established that anthropogenic pollutants such as chlorofluorocarbons (CFCs) released in the troposphere cause depletion of stratospheric ozone on a global scale (3). Effects attributable to other anthropogenic pollutants have also been predicted to occur but are less well established. Middle atmosphere model results presented by Guy Brasseur (National Center for Atmospheric Research) predict that increasing levels of atmospheric carbon dioxide--associated with global warming in the troposphere as a result of an enhanced greenhouse effect--should, in contrast, lead to global cooling of the middle atmosphere (4, 5). This is a robust prediction and is based on the fact that CO2 is not only an efficient absorber of infrared radiation but also an efficient emitter. At stratospheric altitudes and above where infrared emission from CO2 can "escape" to space because of the low atmospheric density at these heights, dramatic atmospheric cooling is expected. Tropospheric climate models generally predict a 1º to 4ºC increase in the tropospheric temperature in a doubled-CO2 scenario. Corresponding middle atmospheric models for the same scenario predict a 10º to 20ºC decrease in middle atmospheric temperatures (5).
Such a large middle atmospheric signal of global temperature change suggests that the atmospheric effects of increasing levels of CO2 may become apparent at higher altitudes before an unambiguous trend is observed in the troposphere. A large change in the middle atmospheric temperature structure would also drive significant changes in the middle atmospheric circulation. This would affect tropospheric climate because stratospheric dynamics provide the upper boundary mechanical forcing of tropospheric weather patterns. Recent results suggest that high-altitude cooling is already being observed, sooner than expected (6, 7). However, our ability to explain this cooling is complicated by other perturbations, including solar activity and volcanic eruptions.
To understand the effects of anthropogenic pollutants in the middle atmosphere, we must determine the dynamic processes by which tropospheric trace gases are transported into the middle atmosphere. Both observations and models have shown that most tropospheric trace gases enter the stratosphere through the tropical tropopause (see the figure) and are then transported through the middle atmosphere by winds and various mixing processes. For greenhouse gases like CO2 and CH4, which show increasing abundances and have a long atmospheric lifetime, the time difference between the appearance of a given abundance in the troposphere and the appearance of that same abundance in the stratosphere is known as the "age-of-air." In addition to providing a measure of the delay between the release of the gases at Earth's surface and their effects on the middle atmosphere, the age-of-air can be used to diagnose model descriptions of dynamics (8).
Darryn Waugh (Johns Hopkins University) and colleagues have compared more than 20 two- and three-dimensional middle atmosphere models and showed that all of the models significantly underpredict the age-of-air, some by as much as a factor of 2, compared with observations of the tracer sulfur hexafluoride (SF6). These discrepancies indicate major deficiencies in chemical-dynamical models, possibly due to inadequate treatment of mesospheric chemical loss processes or of the time history of tracers in the upper troposphere. These results emphasize the effect of interlayer coupling on middle atmosphere trace gas distributions and their variation with time, and their importance for predicting, for example, long-term changes in stratospheric ozone.
In addition to its role as a greenhouse gas, methane is oxidized in the stratosphere to yield water vapor. The photochemical breakup of the water molecule in turn produces hydroxyl (OH), one of the most important oxidizing agents in the atmosphere. In the stratosphere and mesosphere, OH is a key catalytic agent for the destruction of ozone. The ozone budget in the lower and middle stratosphere is well understood, but models of upper stratospheric and lower mesospheric ozone continue to underpredict the observed ozone abundance (9). Robert Conway (Naval Research Laboratory) presented the first satellite observations of OH in the upper stratosphere and showed that the observed abundances are consistent with standard models of ozone chemistry. Thus, the underprediction of ozone near the stratopause continues to be a mystery.
Recent analysis of observations from the NASA Upper Atmosphere Research Satellite (UARS) (10), presented by Lin Callis (NASA Langley Research Center) and David Siskind (Naval Research Laboratory), has established a potentially significant chemical coupling between the stratosphere and thermosphere. Nitrogen oxides produced in the upper mesosphere and lower thermosphere by energetic electrons impacting the upper atmosphere (which also produce the visible emissions of the aurora) are carried downward by the atmospheric circulation into the polar stratosphere (see the figure) (11). These nitrogen oxides react with ozone, increasing the chemical loss of ozone in the polar stratosphere. However, the global impact of this effect on stratospheric ozone is uncertain and controversial.
It is well known that diurnal and semidiurnal atmospheric tides have a strong influence on mesospheric and thermospheric winds. It is also clear that dissipation of gravity waves and planetary waves, both of which are generated in the troposphere by the interaction between winds and orography (see the figure), provide momentum drag on the circulation of the stratosphere and mesosphere. Yet there are currently no dynamical models that encompass the entire atmosphere from the ground to the thermosphere. Most upper atmospheric models have a lower boundary above 90 km altitude, whereas most lower atmospheric models have an upper boundary below 50 km altitude. Ray Roble (National Center for Atmospheric Research) has coupled the NCAR Thermosphere-Ionosphere-Mesosphere-Electrodynamics General Circulation Model (TIME-GCM), which has a lower boundary near 30 km altitude, with the NCAR Community Climate Model (CCM3) model. This "coupled" model was run for one model year with the CCM3 model forcing the TIME-GCM but without feedback from the TIME-GCM on the CCM3 model. Initial results show that lower atmospheric forcing affects the dynamical variability in the mesosphere and the layers above significantly. The results illustrate the need for a fully coupled ground-to-space global atmospheric model with which to study global atmospheric weather and climate, and to understand solar-terrestrial relations.