According to this
http://www.ucar.edu/learn/1_3_1.htm it will be re-emmitted as infrared. I don't know a lot about the physics of this, but I think it can, in theory, be reemitted at a higher frequency, but the odds of that happening shorten very quickly as the frequency gets higher.
The radiation re-emmitted can't be at a higher frequency that it was absorbed at, can it?
Looks like it's time to do a proper science post on how CO2 interacts with photons.
The short answer is, a CO2 molecule that absorbs a photon of a particular frequency, assuming it hasn't absorbed any others, and assuming it was quiescent, can't radiate a photon of a higher frequency, because higher frequency is higher energy, and that would violate conservation of energy; the energy has to come from somewhere. Conservation of energy is a really important concept, and one that's followed in all of physics. We first found out about it in thermodynamics, but it turns out to be important everywhere. What it means is that energy never just disappears; if it appears to, it went somewhere, and finding that somewhere can lead to all sorts of interesting things. No experiment has ever been performed that violated conservation of energy, and physicists expect none ever will be. As far as we can tell, conservation of energy is an absolute principle of how our universe works. Physicists call it a Law of Nature. That's very important, so keep it in mind.
Now, there are different ways CO2 can absorb a photon. The simplest is by the carbon atom, or one of the oxygen atoms, absorbing the photon in their electron shells. This increases the energy of an electron (the one that absorbed the photon), and it jumps to a higher energy level (farther out from the nucleus). In technical terms, it changes to a different orbital. This state is unstable; the electron can jump back down to the lower state it came from, by re-emitting the photon, and it generally does so. One point to note very carefully is that an electron can only absorb a photon of precisely the right frequency to jump the electron to the next orbital; a little too much, or too little, and the photon cannot be absorbed. And, of course, the photon emitted is the same frequency as the one absorbed. Physicists refer to this phenomenon as deep elastic electron-photon scattering.
But there is a little problem; the frequencies this happens at are too high to be infrared; they're in the visible light spectrum, or in the ultraviolet. Luckily, there is another mechanism at work in molecules that permits the absorption of photons. And this mechanism works in the infrared.
Molecules aren't zero sized. What that means is that the molecule has a shape that can be bent; and what's more important, it can resonate between different shapes. In other words, it can vibrate, kind of like a tuning fork does; but there's a difference between the way a molecule vibrates and the way a tuning fork does. The tuning fork is constantly losing little bits of energy to the air, and so it eventually runs out of energy and stops vibrating; but there's no "air" around a molecule, so it doesn't lose energy that way. (That's actually an oversimplification, and there's a thing that happens that does that to molecules, but I'll come back to it.) A molecule that has absorbed a photon of the correct frequency to set it vibrating this way is said to be "excited." And just like the electron, the situation isn't stable. The molecule can return to its original unexcited state by simply emitting a photon of the same frequency, and in general, that's exactly what it does.
Now, the ways a molecule can vibrate depend on the shape of the molecule, and there's a formula for that. A molecule like CO2 is called a linear molecule, because the atoms are arranged in a straight line, like this: O-C-O. For linear molecules, the formula is 3N - 5, where N is the number of atoms; for CO2, that's three. Therefore, the CO2 molecule can vibrate in (3 x 3) - 5, or four, different ways. For a picture of this, see
here, search on the string "CO2" to see a discussion of it. Each of those ways has a specific energy. There are also specific considerations having to do with deeper rules of quantum mechanics that dictate that some modes cannot be excited by photons; this has to do with how the electric charges move relative to one another, and is outside the scope of this discussion. One of the four modes cannot for CO2. So that leaves three. Of those three, two are identical, just rotated 90 degrees; so they have the same energy. Therefore, CO2 has two frequencies it can absorb in the infrared, one for the first two, and one for the single remaining one.
Now, it's important to understand that quantum phenomena like the emission and absorption of photons don't proceed like clockwork. They're random; and the chance of them happening is governed by a probability equation. The probability equation that governs the emission of photons by excited CO2 has the highest probability right after the absorption, and it lowers over time quite quickly. So most CO2 molecules that absorb a photon and start vibrating emit the photon again pretty quickly, although a few may hang onto it longer. We're talking nanoseconds to milliseconds here.
But there are another couple of variables to consider as well. The molecule can only absorb the photon if it is coming in a certain direction, and it can only emit it in that same direction (or perhaps in a plane parallel to the direction of vibration, or in a line parallel to it- the details depend on the particular vibration mode, and CO2 turns out to have more than one). In addition, the CO2 molecule isn't trapped in a solid, it's floating as a constituent of a gas, so it will be tumbling and spinning in various ways. The direction the photon emerges in is therefore random; it depends on the way the molecule is tumbling, and the vibration mode the photon engendered, and ultimately on that random timing of the quantum probability equation. It's therefore safe to say that if you divide all directions into a pair of hemispheres, about half the time the photon will come out toward one of them, and the other half, the other, and that's just what we see when we experiment with them.
Well, a photon floating in the Earth's atmosphere is going to see about half of the directions be into space, and the other half back to Earth. The Earth's atmosphere is very thin compared to its diameter. It's probably 50.00000001% space and 49.999999999% Earth; close enough.
Slimething pointed out, quite correctly, that there's something else that can happen to an excited CO2 molecule than emitting a photon. Since the molecule is floating in a gas, it might collide with another molecule, and if it does, then it might transfer some or all of its energy to that other molecule. This can result in that other molecule vibrating, but only if the energy is just the right amount; most likely, that can only happen with another CO2 molecule. Another thing that can happen is that that other molecule can be "bumped" and accelerate to a higher speed. This is called "quenching." It's important to remember, however, that this can only happen for the various vibrational modes, not for the excited electron in an electron shell; of course, we're not talking about that situation, but it is something to keep in mind.
It's important to keep that quantum probability equation in mind; because what it says is that the photon is most likely to be re-emitted almost immediately. As a result, in a fairly low-pressure gas like our atmosphere, it's fairly unlikely that the CO2 molecule will encounter another molecule before it re-emits the photon. So for every 1,000 absorptions, probably only one or so will result in a quenching event; the other 999 will be re-emissions.
But there's another thing to keep in mind. What we call "temperature," at least in a gas, is really the average kinetic energy of the molecules in the gas. (It's a bit different in a solid or a liquid, but that's off-topic.) And remember, the most likely molecule for a CO2 molecule to encounter is a nitrogen molecule (nitrogen molecules contain two nitrogen atoms, thus N2). Since it can't absorb the vibration (doesn't have a vibration mode of the proper energy), if anything happens, and it might not other than the molecules just bouncing off one another, then what happens is the nitrogen atom is accelerated; that means it speeds up. And that of course means its kinetic energy goes up; and
that means that the gas (the atmosphere) gets warmer. That nitrogen molecule will be bouncing off other molecules in the atmosphere, and at each bounce it will give up some of its kinetic energy, increasing the kinetic energy of the other molecules it hits. Eventually, that tiny bit of energy will be distributed among all the molecules.
So either the CO2 molecule re-emits the photon, and half the time it returns to the Earth where it came from; or else the atmosphere gets a tiny bit warmer. Either way, some of the energy radiated from the Sun-warmed surface of the Earth doesn't escape into space; it's trapped, and makes the surface and the atmosphere warmer. And of course, since the Earth's surface can lose heat by conduction to the atmosphere, either way, the atmosphere gets warmer, too. This is the greenhouse effect, and the fact that I can just about completely describe it without much mathematics should tell you that this is a fairly simple piece of physics.
Now, not all photons leaving the Earth hit CO2 molecules. In fact, most of them don't hit any molecule at all. Otherwise, obviously, we wouldn't be able to take infrared pictures of the Earth with satellites. But the more CO2 molecules there are, the more likely photons are to hit them; and we've already discussed what happens then. So the more CO2 there is in the atmosphere, the warmer the Earth's surface and atmosphere will get. It's not a simple linear relationship; but qualitatively speaking, the more CO2, the more trapped heat, and the higher the temperature. And that's global warming.
Also, according to the absorption bands, CO2 is transparent at higher frequencies, that is, sunlight passes pretty well right through it, but not at certain infrared bands.
Yes, that's pretty much true; CO2 has a few narrow visible light absorption bands, but because sunlight is white, it has many frequencies, and we don't notice the CO2 absorbing much of the light. If the atmosphere were pure CO2, and we had a spectroscope, we could find those bands; but we would only probably notice a slight brightening if it were all taken away.
Now, doubtless Slimething will find various minor errors, mostly of omission, in that explanation; but from the viewpoint of the layman, that's about as detailed as you're going to get without engaging all sorts of math. And you can always refer to that page I linked. So I'd say anyone with an ounce of gumption can just go ahead and look over there and see whether what Slime's got to say holds water for themselves without me needing to point anything out.