
There are three levels in the atmosphere where CO2 could act: at the surface, in the troposphere, and high in the atmosphere where radiation to space is greater than thermal absorption. In each location the conditions are different, and the way CO2 acts is different.
At the surface, the CO2 “back radiation” has been a mystery to many, this author included. Some argue that such a thing doesn’t exist, that it is against the second law of thermodynamics. It does exist, but the mechanism does not work very often, or exactly as advertised. This article will throw you some curves, and many numbers, but will stay away from complicated mathematics. Complex math is unnecessary.

First, the basics: CO2 is a three-atom molecule. It can be visualized as one carbon atom with two oxygen atoms attached, one on each side, in a straight rod configuration. As such, it is “non-polarized”, that is the charges on each atom balance, and the charge on the whole molecule appears to be zero as viewed from any angle. Thus, it has only two modes of vibration when excited by infrared radiation (IR). It can vibrate in and out along the long axis, like a rod that has been struck on one end, or it can vibrate parallel to the axis, like a rod that has been struck in the middle.
The vibration mode along the long axis is only excited in an asymmetric mode, where the end atoms move together in the same direction at the same time. The symmetrical stretch, where the oxygen atoms both move away from the carbon atom then back towards it, does not expose any charge, so doesn’t radiate or absorb radiation.
Figure 3 below shows radiation transmitted by the atmosphere. Conversely, that part that is not transmitted is absorbed. Note: only two peaks of CO2 act to absorb radiation that water vapor does not. Those peaks are at 4.25 µm and 15 µm. The 4.25-µm peak occurs at a wavelength that has little long-wave radiation by the earth.
Click source to read FULL report from