Dr Roy Spencer elaborates further on the greenhouse effect in a recent post. I think that I can follow his arguments to a certain extent although I disagree with him on some fundamental issues. Let's begin with a thought experiment. The main issue here is the cooling of the stratosphere, something that Roy Spencer tries to explain with a "blanket analogy", the same analogy which is often used to explain the elevation of the surface temperature.
Suppose that your house is heated by a constant heat source, that is, there is no thermostat regulating the temperature in the house. In this case, it is widely acknowledged that if you increase the insolation in your house the inside temperature will increase. But will it lower the outside temperature?
My answer to that is no. When the new stationary state is reached, the same amount of energy will be emitted from your house to the ouside, there is no way this heat could disappear. In other words, a squirrel living just beneath your roof will not suffer from your insolation, apart maybe for a certain amount of time while the new stationary state is reached.
I agree with Roy Spencer that back-radition exists, and that this downwelling radiation taken alone probably reduces the rate of cooling of the earth. But so does any downward motion of the atmospheric particles.
I disagree with Roy on the following issues:
1. I believe that any gas can cool radiatively, just check Maxwell's equations for accelerating charged particles.
2. I have not found, despite my best efforts, any arguments to show why an increased IR absorption/emission destabilizes the atmosphere, that is, increases the lapse rate (check out my simple radiative transfer models). I challenge anyone to demonstrate it mathematically using a bottom-up model.
3. I believe that the sunlight destabilizes the atmosphere.
4. I believe that the total atmospheric mass is the primary parameter that determines the heat insolation, there are microscopic arguments supporting this, the best I can think of is Newtons third law together with the fact that radiation carries momentum.
5. Increased heat insolation does not cool any part of the system, in the sense that when the new stationary state is reched the temperature will be higher everywhere.