CMB photons, why are they still around?

[Gerinski]

10 billion years ago, photons emitted at the surface of last scattering were being absorbed by whatever stuff was around. Those atoms would absorb those photons and release other photons, but these would not be primordial photons from the CMB anymore but photons emitted at a later time by the stuff existing at that later time. The same 9 billion years ago. The same 6 billion years ago, The same 2 million years ago and the same in the XIX century.

We are more than 13 billion years from the surface of last scattering and we are still detecting photons from the CMB. It seems reasonable to believe that if humanity will stay alive, it will keep on detecting photons from the CMB for many centuries or millennia to come. How come they have not all been absorbed by some atom in all this time? Were there indeed so many photons released that no matter how much stuff there is to absorb them, there are still plentiful of them to keep on traveling forever to every edge of the Universe? Is it conceivable that at some stage an hypothetical civilisation would not be able to detect the CMB anymore because all of its photons had been absorbed by the stuff 'in between'?

 

[Orodruin]

But this is precisely what the name "last scattering surface" means. At that time, the mean free path (the length a photon travels on average before interacting) became larger than the size of the Universe. Hence, most photons will not interact in between then and now. There are simply not enough stuff in the Universe to absorb them.

 

[Ken G]

10 billion years ago, photons emitted at the surface of last scattering were being absorbed by whatever stuff was around. Those atoms would absorb those photons and release other photons, but these would not be primordial photons from the CMB anymore but photons emitted at a later time by the stuff existing at that later time.

But the latter is what we call the CMB, not the former. I think that is what is confusing you-- we never see those primordial photons, we see the ones that have been absorbed and re-emitted, probably many times over. Just not in the last 10 billion years, all the absorbing and re-emitting was over by then. So the CMB is what gets re-emitted, but because energy is conserved, it is in some sense the "same energy" if not the "same photons." The point is, we can still learn about the early universe using the CMB, even if it is not the same photons.

 

[Orodruin]

But the latter is what we call the CMB, not the former. I think that is what is confusing you-- we never see those primordial photons

This is plain wrong. The mean free path of the CMB photons are larger than the size of the observable Universe. If what you said was true the CMB polarisation would have no information on the early Universe as scattering would destroy it.

 

[Orodruin]

Just for some perspective, the mean free path today if we would assume all matter to be free protons and electrons and use the Thomson cross section would be ca 400 times the Hubble length. Now, everything is not free protons and electrons, but even with this overestimate only one in 400 photons would interact within the lifetime of the Universe.

 

[Ken G]

This is plain wrong. The mean free path of the CMB photons are larger than the size of the observable Universe. If what you said was true the CMB polarisation would have no information on the early Universe as scattering would destroy it.

I don't think you understand the original question. Gerinski was thinking that the CMB were the photons from the early moments of the universe, which are constantly being absorbed and replaced by other photons that aren't the CMB. But the CMB is the photons after they get replaced like that.

 

[Orodruin]

I don't think you understand the original question. Gerinski was thinking that the CMB were the photons from the early moments of the universe, which are constantly being absorbed and replaced by other photons that aren't the CMB. But the CMB is the photons after they get replaced like that.

I do understand the question perfectly. The last scattering surface is not just a name given on a whim. Then it depends if you consider 300000 years "early". Do you understand that a mean free path orders of magnitude larger than the size of the observable universe means that scattering are not "constantly happening"? It did happen, but only until the Universe was 300000 years old.

Edit:

Gerinski was thinking that the CMB were the photons from the early moments of the universe

No, this is not what the OP states. Gerinski is asking whether photons have been continuously scattered over the last 10 billion years. There is no mention of times before the LSS and he is asking what is happening to photons emitted at the LSS in the times between the LSS and now.

 

[Ken G]

Well, you are welcome to have a different interpretation of the question, but my answer is perfectly accurate-- given what I still think was being asked. The CMB are the photons after they have been absorbed and re-emitted many times, and very little of that absorption and re-emission has happened in the last 10 billion years, but lots and lots of it happened before that.

 

[Bandersnatch]

I don't think you understand the original question. Gerinski was thinking that the CMB were the photons from the early moments of the universe, which are constantly being absorbed and replaced by other photons that aren't the CMB. But the CMB is the photons after they get replaced like that.

I'm with Orodruin, Ken. The only bit I find hard to understand in this thread is your answers.
What do you mean by 'CMB photons get replaced'? By interacting with what?

 

[Orodruin]

Well, you are welcome to have a different interpretation of the question, but my answer is perfectly accurate-- given what I still think was being asked.

The question is obviously asking about scatterings between 10 billion years ago and now:

The same 9 billion years ago. The same 6 billion years ago, The same 2 million years ago and the same in the XIX century.

I am sorry but I do not see how this can be misinterpreted and I believe your answer has only served to confuse the OP. Nowhere is a time before the LSS mentioned. The entire point is that the premise is flawed as the following statement is not true:

10 billion years ago, photons emitted at the surface of last scattering were being absorbed by whatever stuff was around.

 

[Ken G]

The issue is what Gerinski meant by "primordial" versus emitted "at a later time." It is probably true that Gerinski was confused that expansion of the universe, and recombination of hydrogen, makes the photons able to go farther and farther with time, until they simply don't get absorbed any more at all. That's what Orodruin was answering. I was thinking that perhaps Gerinski had been told that the CMB is the "light from the Big Bang," or some such thing, and he is wondering why any of it is left, given all that absorption and re-emission going on. I'm saying the CMB is what you get after all absorption and re-emission is over, not the original "light from the Big Bang." However, it is certainly true that all this absorption and re-emission had to happen prior to encountering the surface of last scattering, that's what those words mean. So it amounts to guessing what parts of that story is creating the confusion, but we can agree there are two crucial aspects: the CMB is the result of absorption and re-emission, and there was a surface of last interaction long ago and far away. On re-reading the original question, I think perhaps the problem is in the term "scattering", which Gerinski is thinking is different from "absorption and re-emission", whereas no crucial distinction between those terms is relevant here.

What's more, CMB photons have been absorbed in the last 10 billion years, just not most of them. So yes, I agree with you that Gerinski was confused about that, but I still think he is also confused about the fact that the CMB is the result of an absorption and re-emission process, it is not really "primordial" if primordial takes its literal meaning (before history, or "existing at the beginning of time".) The bottom line to my answer was that the CMB is not the primordial light that was present at the beginning of time, it is the result of interactions with matter. That's perfectly correct. I also said that those interactions just happened a long time ago (because the universe has become very rarified), though that part had already been answered. Combining those two elements is crucial to answering the question, because the CMB is not "primordial," though its existence can be traced back to a primordial source.

 

[Gerinski]

Thanks to all. Orodruin got my question right, I was asking about absorptions between the LSS and now, not before the LSS. The thing is, the photons of the LSS are continuously being absorbed by the atoms in our telescopes, everytime we look up (with a sensitive enough device) we see the CMB, we see those photons meaning that they are hitting our detectors and being absorbed by them. If we had had sensitive enough devices 500 years ago we would also have seen them, and the same at every epoch of the Earth. And the same can be said from any location in the Universe, any being with a sensitive enough device will capture photons from the CMB. So how can you say that those photons are most unlikely to ever get absorbed by any atom in the Universe? It seems to me that they are constantly being absorbed by every atom in the Universe aren't they?

 

[Orodruin]

So how can you say that those photons are most unlikely to ever get absorbed by any atom in the Universe?

Because they are very unlikely to ever get absorbed. The thing to realize here is that your telescope is not very representative for how much matter there is to scatter on in the Universe. Naturally, photons are more likely to interact in regions of larger matter density - but those regions are very very rare. So rare that it is unlikely that any given photon will encounter them.

 

[Drakkith]

Gerinski, consider that the Hubble space telescope (and others) can see galaxies that are about 40 billion light years away because there is practically nothing in the way of the light. The same goes for the CMB. Overall, space is just mostly empty, with a smattering of objects here and there.

 

[Fantasist]

Just for some perspective, the mean free path today if we would assume all matter to be free protons and electrons and use the Thomson cross section would be ca 400 times the Hubble length. Now, everything is not free protons and electrons, but even with this overestimate only one in 400 photons would interact within the lifetime of the Universe.

Thomson scattering is not really a very efficient process due to the small cross section. Photoionization could be more efficient: CMB radiation can ionize excited neutral atoms in quantum states around n=100 and above. The cross section for this would be about 10 orders of magnitude larger than Thomson scattering (facilitated by the fact that the cross section strongly increases with increasing n). Only a very small density of these excited atoms would be required to have a mean free path smaller than the observable universe.

It would be interesting to know whether the spectrum of galaxies shows any indication of absorption in this frequency range (as a function of distance).

 

[Gerinski]

Thanks to all, it's clearer now.