22

There is something with CMB radiation that does not sit well with me... It seems very counterintuitive that we are able to see it. If CMB radiation formed at the early phases of the universe, would it not make sense that it expands and propagates "outwards" with the Big bang so that we would never see it? Most space is out of our reach since it does not belong to the observable universe. But CMB radiation formed earlier than most space, should it not also be far outside the observational universe on its way away from us?

How should I visualize the CMB radiation? I have a master in theoretical physics but I never went back to understand this. Any help in understanding this is very welcomed.

Qmechanic
  • 201,751
Johan Hansen
  • 563
  • 5
    Are you aware that all points (x,y,z) in our present space were also present at the beginning of the universe? see this https://physics.stackexchange.com/questions/136860/did-the-big-bang-happen-at-a-point/136861#136861 – anna v Aug 15 '21 at 20:02
  • 13
    It doesn't propagate "outward" or "inward" towards us. It started everywhere and propagates in every direction. We happen to see the bits that started at the right distance and had the right direction to hit us now. – knzhou Aug 15 '21 at 20:27
  • @knzhou: Yes, you can only detect radiation that propagates directly towards you and strikes your detector. – jamesqf Aug 16 '21 at 03:27
  • 1
    This might be of interest: Why is CMB not considered as the edge of the universe? – Dhruv Saxena Aug 16 '21 at 08:19
  • @knzhou how convenient... – Michael Aug 16 '21 at 17:43
  • 11
    @Michael Perhaps it's convenient, but it's not any more convenient than being able to see a tree in the daylight. For that to be possible, a photon from the Sun needs to be aimed perfectly to hit the tree, and then reflect at the perfect angle to hit your eye... a conspiracy? – knzhou Aug 16 '21 at 18:03

4 Answers4

46

The biggest misconception I see in the question is the idea of the Big Bang as something that propagates "outwards", like an explosion. There is no outwards direction, the universe didn't expand into something.

Even though recent observations seem to suggest that the universe is closed, for the sake of simplicity let me assume that the universe is flat and infinite.

The first thing to notice is that if the universe is infinite now, it was infinite also at shortly after the Big Bang, it was just more dense. Imagine a flat plane on which it is drawn a lattice of dots, equally spaced.

before expansion

The plane is the universe at a certain time. The dots represent objects in the universe, electrons, atoms, stars, whatever.

Now, imagine scaling up the plane, making it bigger. Of course the plane is infinite, so scaling it doesn't change its size, but the dots get further away from each other while keeping their size fixed (otherwise, you wouldn't be able to tell that an expansion took place).

after expansion

You see that the dots aren't expanding into anything, the universe (the plane) was already infinite, it didn't grow in size.

Critically, there is not a center of the expansion. The distance from any two dots has doubled (in this example), regardless of their position.

Now, let's talk about the CMB. Imagine that at time $t_0$, every dot emitted a pulse, an expanding circular wave. This wave symbolizes the photons of the cmb that are emitted at the same time from every point towards every direction.

cmb emitted

The last pictures refers to our situation. Earth is the black dot that is touched by the wave fronts. We see the CMB coming from every direction (in the picture only from four directions) because it was emitted from everywhere.

cmb gets to us

anna v
  • 233,453
Prallax
  • 2,859
  • 3
    Actually, why bother assuming anything about the size of the universe when it does not matter? All that matters is that at some point after the big bang there were lots of photons being radiated from many different locations, so some of them might reach our position now. – user21820 Aug 16 '21 at 06:33
  • 3
    @user21820 Well, it has to be big. Infinite is one kind of big. I believe "30 billion light year radius" would also be big enough. But 1 billion light years would not be big enough, the CMB wouldn't look like it does. So the size of the universe does matter. – Yakk Aug 16 '21 at 15:18
  • 1
    @Yakk: Yes of course. We both know that my point was just that it does not have to be infinite. – user21820 Aug 16 '21 at 15:32
  • "Critically, there is not a center of the expansion". This doesn't work in a flat 2D plane, right? There must be a center somewhere on the plane. Your sentence can apply to the surface of a sphere, though. Right? – Eric Duminil Aug 17 '21 at 07:17
  • 2
    @EricDuminil As long as the plane is infinite there is no center. – Taemyr Aug 17 '21 at 10:18
  • 1
    @Taemyr sure, an infinite plane has no center (or an infinity of it). There must be a center for the scaling, though. Each point has to move in a specific direction, with a specific velocity. Points cannot move just anywhere, at infinite speed. – Eric Duminil Aug 17 '21 at 10:47
  • 5
    @EricDuminil No, there does not need to be a center for the scaling. The pertinent question is; move relative to what? – Taemyr Aug 17 '21 at 10:56
  • 2
    @Taemyr Thanks. It's counterintuitive, but I think I now get it. There's no external reference, since the plane is the whole universe. Every observer has the impression that every other point is flying away from them. – Eric Duminil Aug 17 '21 at 14:38
  • @user21820 I think it does matter if the universe is infinite. If it is finite, then either its size must change, or some of the dots in the first example plane have to vanish to make room for the greater distance between them. Having an infinite plane solves that, although I have to admit that it boggles my mind. Also, if the universe is finite, there would be a reference you could use to find the "center". – Kevin Keane Aug 19 '21 at 00:00
  • @KevinKeane you can still have a finite universe and avoid the problems you mention if it is closed: e.g. the surface of a sphere (just the surface!) – Prallax Aug 19 '21 at 05:48
  • @KevinKeane: I do not get what you are saying at all. We only can observe data from the observable universe, which is finite. What has happened or will happen outside of that boundary is not observable by us, nor is that relevant to why the CMB can take a long time to reach us. – user21820 Aug 19 '21 at 14:46
  • @user21820 The issue here is where the boundary of the observable universe is. Naively, one would expect it to be 13.8b light years away. But that is ignoring the expansion of space itself. A point that today is 13.8b light years away would have been much closer to our current position when the light left. And conversely, a point where light needed 13.8b years to reach us will in the last 13.8b years have moved to 40+b light years away thanks to the expansion of the universe. So the effective observable universe is dictated by time+expansion, not just by time. – Kevin Keane Aug 20 '21 at 02:53
  • @KevinKeane: What did I say that made you think I was using a "naive" definition of the observable universe? My point stands that there is no need or reason to assume an infinite universe. – user21820 Aug 20 '21 at 05:32
  • 1
    @user21820 I agree with you, to explain CMB there is no need to assume an infinite universe. But I also wanted to address the misconception that the universe is expanding "outwards" into something. I just thought it would be simpler to explain it in an infinite flat universe, than to postulate a closed universe, like an expanding sphere, and then explain that a topological manifold needs not to be embedded in 3D space – Prallax Aug 20 '21 at 05:46
  • @Prallax: Ah okay that makes sense. Thanks for addressing my point! – user21820 Aug 20 '21 at 07:32
15

The CMBR did not occur in a single location, but rather throughout the entire early universe (from over 13 billion years ago). Therefore, any theoretical observer throughout the full history of the universe is able to see the CMBR passing through their area of the universe in their present moment (however, with a different thermal signature). What we see today is light that has traveled through a distance of over 13 billion light years. This total distance is comprised of both the initial distance (what your distance would have been from a certain region of the universe at the time the CMBR was emitted) plus the additional distance created by the stretching of spacetime (and hence the observed cooling of the CMBR).

Since the CMBR is not an event (a point in spacetime defined specifically by three spatial dimensions and one temporal) but rather occurred everywhere, you simply need to align the distance traveled with the time that has transpired and you will be able to observe the CMBR that was emitted from the corresponding region of the universe at the time the CMBR was emitted.

JRL
  • 778
  • The photons would have traveled for 13 billion years, but probably from much further away than points that today are 13 billion lightyears away. This is because the expansion of space itself is not limited by the speed of light. I vaguely recall that the actual number is something like 46 billion light years, although don't quote me on that. – Kevin Keane Aug 19 '21 at 00:06
10

I struggled with visualising this for some time. What we are seeing as the CMB is the "surface of last scattering"; in other words the (receding) surface where photons were emitted after recombination.

I found Lineweaver's "surface of last screaming" a helpful analogy:

Surface of last screaming

Consider an infinite field full of people screaming. The circles are their heads. You are screaming too. (Your head is the black dot.) Now suppose everyone stops screaming at the same time. What will you hear? Sound travels at 330 m/s. One second after everyone stops screaming you will be able to hear the screams from a `surface of last screaming' 330 meters away from you in all directions. After 3 seconds the faint screaming will be coming from 1 km away...etc.

https://ned.ipac.caltech.edu/level5/March03/Lineweaver/Lineweaver7_2.html

PM 2Ring
  • 11,873
James Random
  • 201
5

Imagine 13 billion years ago, the Universe is a hot soup with many photons bouncing around in random directions. Every direction you look, hot soup photons.

Now expand everything and come to today. The same hot photon soup is still here, but redshifted. Every direction you look, wherever you are, soup photons. We are in the same universal cloud of photons.

The CMB is essentially our view from inside a photon gas or perfect blackbody.

The issue with understanding the "observable" Universe is that there are many horizons to be confused between: https://en.wikipedia.org/wiki/Cosmological_horizon

Since the Universe is expanding, distance travelled by a photon 13 billion years ago is worth more distance today (similar to how money a long time ago was worth more). This means that we can see photons from a very long distance that were emitted in the past (particle horizon), even though we can never visit the place where it came from if we started today (event horizon).

Alwin
  • 4,955
  • 6
    So where are the angry cosmologists protesting in front of the universe that we must stop printing new spacetime and pay back the intergalactic debt? – user253751 Aug 16 '21 at 08:28
  • 1
    @user253751 They are very closed cosmologically minded. – Yakk Aug 16 '21 at 15:20
  • 1
    @user253751 they were shouted down by the cosmic economists who just came up with a new way they could leverage inflation to make a quick buck for their employers off the back of working class bosons, who are clearly being forced apart by social distancing. We totally need a Justice for Bosons movement! – Corey Aug 18 '21 at 23:35