How? Models that show an increase in water vapor coming from the increase in CO2. The slight (very slight relative to other gases) warming from CO2 allows the atmosphere to hold more water vapor. This much more potent GH gas then causes further warming. The problem with those models is the assumption that clouds can be described with a single parameter ignoring or trivializing the effect that the increased water vapor has on the weather.
So to say that CO2 changes the radiative balance contains several assumptions and a huge simplification that doesn't bear out in reality. Again today there are lenticular clouds leftover from overnight convection. Those affect the temperature and climate. They do not have to be modeled accurately for my particular area, but they must be modeled somehow since water vapor will create more convection and more of those effects (clouds near the top of the troposphere). But they are not modeled except in aggregate and ultimately arbitrary parameters.
Sorry for the long Easter weekend delay.
Clouds and aerosol effects are the single largest source of uncertainty in global warming predictions/projections. And there's no doubt about that (insert Mona Lisa smiley here)