The principal greenhouse gases whose longwave radiative effects are included in CAM 3.0 are HO, CO, O, CH, NO, CFC11, and CFC12. The prediction of water vapor is described elsewhere in this chapter, and CO is assumed to be well mixed. Monthly O fields are specified as input, as described in chapter 7. The radiative effects of the other four greenhouse gases (CH, NO, CFC11, and CFC12) may be included in CAM 3.0 through specified concentration distributions  or prognostic concentrations .
The specified distributions are globally uniform in the troposphere. Above a latitudinally and seasonally specified tropopause height, the distributions are zonally symmetric and decrease upward, with a separate latitude-dependent scale height for each gas.
Prognostic distributions are computed following Boville et al. . Transport equations for the four gases are included, and losses have been parameterized by specified zonally symmetric loss frequencies: . Monthly averaged loss frequencies, , are obtained from the two-dimensional model of Garcia and Solomon .
We have chosen to specify globally uniform surface concentrations of the four gases, rather than their surface fluxes. The surface sources are imperfectly known, particularly for CH and NO in preindustrial times. Even given constant sources and reasonable initial conditions, obtaining equilibrium values for the loading of these gases in the atmosphere can take many years. CAM 3.0 was designed for tropospheric simulation with relatively coarse vertical resolution in the upper troposphere and lower stratosphere. It is likely that the rate of transport into the stratosphere will be misrepresented, leading to erroneous loading and radiative forcing if surface fluxes are specified. Specifying surface concentrations has the advantage that we do not need to worry much about the atmospheric lifetime. However, we cannot examine observed features such as the interhemispheric gradient of the trace gases. For climate change experiments, the specified surface concentrations are varied but the stratospheric loss frequencies are not.
Oxidation of CH is an important source of water vapor in the stratosphere, contributing about half of the ambient mixing ratio over much of the stratosphere. Although CH is not generally oxidized directly into water vapor, this is not a bad approximation, as shown by Le Texier . In CAM 3.0, it is assumed that the water vapor (volume mixing ratio) source is twice the CH sink. This approach was also taken by Mote et al.  for middle atmosphere studies with an earlier version of the CCM. This part of the water budget is of some importance in climate change studies, because the atmospheric CH concentrations have increased rapidly with time and this increase is projected to continue into the next century (e.g., Alcamo et al. ) The representation of stratospheric water vapor in CAM 3.0 is necessarily crude, since there are few levels above the tropopause. However, the model is capable of capturing the main features of the CH and water distributions.