universe expanding faster than light

camilus

I read it is possible for the universe to be expanding faster than the speed of light.

What would be the implications of this?

DaveC426913

The speed limit of the universe applies locally. The expansion of the universe is the expansion of the space - no object is moving faster than c locally, so it does not violate SR.

However, the expansion will prevent us from seeing anything that's moving away faster than c.

Search this forum for "superluminal expansion".

sylas

Most of the galaxies we see in deep space are "moving" (or receding) faster than the speed of light. They have always been receding faster than the speed of light.

You need to be careful. There are sometimes people who confidently make statements that are not true. Sorry Dave, you just did this here yourself.

That's part of the fun... many people here are amateurs trying their hand at learning and explaining what they've learned; but we sometimes make mistakes. There are also a number of professionals here who know and understand the details very well. It can take time to sort out who is who. I'm not one of the experts; but I know a bit about it... and I've been picked up by the experts from time to time as well!

I'll see if I can find a good thread on this. There are some confusing threads as well.

Cheers -- sylas

camilus

I perfectly understand this, I know that expanding faster than c doesnt violate the 'cosmic speed limit' because it is space itself thats expanding, not a particle in space moving faster than c.

Thats my initial thought as well. But if the rate of this expansion is in fact superluminal, what can we say about the size of the universe? Cant it be a lot bigger than we think. since we cant see that far?

Im gonna do the search now, thanks.

sylas

Perhaps more useful would be a tutorial reference. See, for example, the well regarded cosmology tutorial by Ned Wright. Also his FAQ on this very question:

Can objects move away from us faster than the speed of light?

Again, this is a question that depends on which of the many distance definitions one uses. However, if we assume that the distance of an object at time t is the distance from our position at time t to the object's position at time t measured by a set of observers moving with the expansion of the Universe, and all making their observations when they see the Universe as having age t, then the velocity (change in D per change in t) can definitely be larger than the speed of light. This is not a contradiction of special relativity because this distance is not the same as the spatial distance used in SR, and the age of the Universe is not the same as the time used in SR. [...]

marcus

That is not quite true. For example there are thousands of galaxies with redshift around z = 1.5.

A galaxy which we see with redshift 1.5 is today moving away from us faster than c.

However according to standard cosmo, someone in such a galaxy could send us a message today, and it would get here.

The reason this works is because the Hubble rate is decreasing and will be decreasing for the foreseeable future. This means that the Hubble distance (c/H, reciprocal to the rate) is increasing.

This is explained in Lineweaver's SciAm article "Misconceptions about the Big Bang". I have a link to it in my sig.

DaveC426913

That one always throws me. I defer to greater wisdom.

marcus

Sylas, I think the word is "kudos" or maybe it is "thanks". Not for your contribution to this thread in particular, but on several astro/cosmo threads I've been seeing clear reliable helpful posts of yours. It seems to help.
Thanks and kudos.

I also would warn against searching PF for "superluminal expansion". The fact that someone uses that term even suggests they may not understand normal cosmo very well. It is typical for largescale distances to expand at rates > c.

I think your idea of Ned Wright's tutorial is a good suggestion.

Most of the galaxies we can see are at distances which were expanding > c when they emitted the light and which continue now to expand > c. That is true for anything with redshift 1.7 or greater. And of course we are getting the light from them--it reaches us nevertheless.

And of course these galaxies are not moving significantly, they are not going anywhere in particular. Individual motions are generally only a few 100 km/s which is negligible compared with c.

marcus

Well it's intelligent to ask if there are consequences of expansion. But basically expansion is just Hubble law, which says that on large scaler the actual (the present day or "now") distance is increasing 1/140 of a percent per million years. That doesnt apply to local in-galaxy distances, among gravitationally bound objects.

Superluminality is a non-issue because of course if you take a large enough distance, like an actual distance of 14 billion LY, that distance will be increasing c or more, just work out the percentage! At 1/140 percent, it will grow by a million LY in the course of a million years. I'm talking actual distance. That rate of increase---a million LY in a million years---is c.

And astronomers have been living with Hubble law (which is what I'm talking about) since 1929.

So they have had time to get used to expansion, and their ideas of the size of the universe and the distances to various horizons are comfortably compatible with the expansion model and the Hubble law.

==============

You asked about current size estimates. This is a separate issue. If space is infinite then it is infinite, right? So what if it is finite. When astronomers discuss that case they typically are talking about a 3D hypersphere and the question comes down to estimating a lower bound on the circumference.

Currently, according to the most recent authoritative report, Komatsu et al, March 2008 from NASA's WMAP mission, there is a 95 percent confidence estimate for the circumference of 600 billion LY.
I'll get the link for you later, unless someone else does. I have to go to supper now.
Back now. To get that report you just google "komatsu wmap" and it will be the fourth hit or so or you google
"komatsu wmap cosmology" and it will be the first hit. The reports were published in 2009 although the preprints came out in 2008.
The key thing is you want "Cosmological Interpretations" report. Table 2 on like page 3. You have to know the units they are using, but if you work it out it comes to a circumference of some 600-630 billion LY, lower bound with 95% confidence.

That is actual distance ("now distance") meaning you freeze expansion and send off a flash of light and it will take 600 billion years before it comes around and you see it coming in from opposite side of the sky.

There are other types of distance and coordinates, which are interesting to discuss once you learn the basics, and of course distance depends on coordinates. But this now distance or actual presentday distance is what the professionals most commonly use, and it's what works in Hubble Law. So it is the good one to get started with.

Phrak

If the answer is yes, what does it mean?

It means you take a local map of spacetime that is nice and flat (said to have a Minkowski metric) and apply it to the spacetime of distant regions.

Are the values of velocities for distant objects and radiation intrinsic properties, or are they an artifact of the map?

marcus

Once you choose a map, then everything you derive from using that map must derive ipso facto from that map.

That's not the point. The question is what kind of coordinates and what kind of basic distance measure do cosmologists actually use. We have an educational responsibility to use terms and concepts compatible with normal cosmology. They use the Friedman metric and the Friedman equations are the basic model.

They don't use the Minkowski metric---trying to force the universe to fit on a Minkowski rack would be horribly impractical. Minkowski spacetime has no expansion. You can think of the universe as spatially flat or nearly flat, so you might think Minkowski would be a good fit. But don't confuse spatial flatness with overall spacetime being Minkowski.

By contrast the 1922 Friedman metric, which is what cosmologists are still using, and mostly very happy with, has built in expansion and is considerably more comfortable to use. Friedman metric is the language the Hubble Law talks and the Friedman equations which are about that metric, are the basic equations of cosmology.

Friedman lets you say easily what it means to be at rest with respect to the ancient matter of the universe, or the afterglow from it, the cosmic microwave background. And it turns out that galaxies have very small individual motion with respect to background, most matter is approximately rest.

If you pick some peculiar coordinates then to get a fit you will have to strain and force things---like it may force galaxies to be whizzing away from some arbitrarily chosen centerpoint at stupendous speeds. Actual motion (in terms of your different choice of coordinates.)

Chronos

Other interpretations are possible, marcus. I perceive no need to portray any particular model as the only one worth considering.

Phrak

[QUOTE=marcus;2231055]Once you choose a map, then everything you derive from using that map must derive ipso facto from that map.

That's not the point. The question is what kind of coordinates and what kind of basic distance measure do cosmologists actually use. We have an educational responsibility to use terms and concepts compatible with normal cosmology. They use the Friedman metric and the Friedman equations are the basic model.

I didn't read between the lines. What did I not see.

Chalnoth

I don't think this is actually true. First, I'd like to make a pedantic point:

In General Relativity, the operation of subtracting one velocity from another to get relative motion is only well-defined at a single point. So if I want to talk about recession velocity, it turns out that the question is very ambiguous, and whether or not such velocities exceed the speed of light just depend upon what ambiguous definition we choose. For example, if a far-off object is at rest with respect to the universe, and we are at rest with respect to the universe (we aren't, quite, but this is just a thought experiment), then we could just as easily state that we are neither moving towards or away from this far-off object as we could state that this object is receding at faster than the speed of light. This ambiguity is unavoidable due to the nature of General Relativity, and is precisely why "faster than light recession" is not a problem: if the way in which you compare velocities of objects far away from one another is ambiguous, then there certainly can be no law limiting such velocities.

The actual speed of light limitation in GR is that no object can outrun a light beam. That is, no matter where the object is, and no matter what your perspective of looking at that object is, that object will always be moving more slowly than the photons in that object's vicinity.

Now, with that out of the way, we can define a very simple recession velocity:

v = Hd.

This is sort of what we naturally think of when we think recession, and I'll use this definition from here on out. With this definition, the distance at which things start to exceed the speed of light is about 14 billion light years away. As marcus points out, in terms of redshift this is about z = 1.5 or so. We currently can see galaxies out beyond redshift 7, and the CMB is at a whopping redshift of 1090.

The solution to this apparent problem is just that we have to take into account how the universe has expanded from the time the light was emitted to now. To take a look at the distance today of these objects, and notice that the Hubble law predicts a recession velocity greater than c, and then scratch our heads wondering why we can see these objects, is missing the point.

First, we can't say, without looking in more detail, whether or not these objects were receding faster than light. They might have been: as long as the universe reduced its expansion enough in the interim, we might see them today.

But what if the object today is expanding faster than light, and always will be? Well, in that case, we can never observe it. The distance at which this occurs depends upon the future expansion of the universe, which we don't yet know for certain, but if the correct explanation for dark energy is a cosmological constant, or something that has the future behavior of the cosmological constant, then stuff that is currently receding at faster than the speed of light (beyond about 14 billion light years, or redshift 1.5) is currently emitting photons that we will never ever see, no matter how long we wait.

Yes, this means that many objects visible today have already passed out of causal contact. This does not mean that they will disappear. But what it does mean is that we will never see them age past 13.7 billion years old (the current age of the universe): all of the photons they emit after that time will never reach us.

If, on the other hand, the expansion were to slow again, then we may start to see some of the photons from later in these objects' lives. For instance, for much of the early history of our universe, the expansion was slowing down. So the norm would have been to see far-off objects that were, when the light was emitted, receding faster than light, but currently are not. Today the converse is true due to the recent accelerated expansion.

All that said, it is just patently incorrect to state that we can see objects today there have always been receding at faster than the speed of light.

marcus

Chalnoth, I don't understand your reasoning, or why you say that. Let me explain.
There are objects that we see today which have always been receding faster than c, and indeed this is typical.
Have a look at Lineweaver, the link is in my sig, to understand how that happens intuitively. He has a graphic illustration. What you seem to be saying is one of the common misconceptions about expansion cosmo, i.e. that it is impossible for us to be be seeing light from objects which have always been receding faster than c.

We can do the numbers if you would google "cosmos calculator" and put in .25, .75, 74 (the current best parameters) and try some redshift like 1.7.
If you put in z = 1.7 you will get that the recession rate then (when light emitted) was 1.01c and the recession rate now (when light received) is 1.14c.
So that thing was receeding all that time > c. But the light got here.

Now you can put in z = 1.8 , or 2.0, or 2.5, and so on and you will find the same story except all the recession rates are higher. In other words what Sylas said is even more true the farther past z = 1.7 you go.

But there are way more objects with z > 1.7 than with z < 1.7. So one can easily conclude that it is typical, that when we see some galaxy probably that galaxy was receding > c when it emitted the light, and also all the time since then.

Physically it is easy to understand how the light gets here despite losing ground at first. The hubble rate H(t) decreases and that makes the hubble distance increase and the hubble distance so to speak reaches out to the struggling photon. Once the photon is within the hubble distance then it can make forwards progress and get closer to us and gradually make it here. That is kind of figurative, not a rigorous description, but you can make it rigorous.

Then also notice that even though we have acceleration and a''(t) > 0, the Hubble rate has always been decreasing and is destined to continue to decrease according to the standard cosmo model.

I guess the asymptotic value for H(t) will be about sqrt(.75) times 74. Let us see what that is. It comes out to 64. So H(t) will decline from present 74 and approach 64 from above. This is just what the standard model predicts. Of course we do not know the future.
But going by the standard model, we see that the Hubble distance still has another 15 percent to expand. From the present 13.2 out to 15.3 billion LY.
That means that any photon which can manage to stay within 15.3 billion LY of us, and is heading our way, will eventually make it to us.

I don;t think that is right. A galaxy which is today just outside the Hubble radius could send us a message today which would reach us. The Hubble radius is the current distance of objects which are receding at c. Using the most recent parameters the Hubble distance is about 13.2 billion LY or a redshift of 1.4 (more precisely maybe 1.38 but let's say 1.4). The Hubble distance is supposed to expand in the future out to 15.3 billion LY (that is the "cosmological event horizon").

Something out there at 15.3 billion LY can not send us a message. And we could not get to them even if we could leave today and travel at speed c.
So it's like you said about out of causal contact. But 13.2 is not the same as 15.3. There is some leeway, so to speak. Some slack. I estimate a galaxy which for example is now out at 14 billion LY could send us a message today that would get here even though it is receding at a rate > c. I will try to double check that. Havent done the numbers rigorously on it.
14 billion LY corresponds to a redshift z = 1.5 and a presentday recession rate of 1.06 c. So at first the photons would be "swept back" so to speak at the rate 0.06c. But they have 1.3 billion LY of leeway before they hit the limit of 15.3 billion LY. Even if they lost ground at an average rate of 0.1c for 10 billion years that is still only a loss of 1 billion LY and they would then be 15 billion LY from us. Before those 10 billion years have passed I think the Hubble distance will have extended out to them and they will be making progress towards us. I'm sure you see how it works. I'll have to figure out how to do the example more rigorously though.

yogi

I would agree - and while having learned much from reading marcus's informative and well written tutorials, the standard model of mainstream cosmology should be qualified. History has shown that the majority is seldom right.

marcus

Thanks for your courtesy, yogi! I am a fan of some nonstandard cosmology and am excited by the progress it is making, but I think with newcomers asking basic questions we have a responsibility to give them the mainstream standard model first. They should get clear on that, as basics.

I personally perceive no need to portray any particular model as the only one worth considering.

And I do not myself portray any particular model as the only one worth considering.

My experience is that most newcomers arrive with some misconceptions about the mainstream model, gotten from some badly written popularization, or some oversimplification they heard, and they are puzzled by it. What they want first is to understand the mainstream picture and see why it makes sense. So that's our first job.

Once everybody is up to speed on mainstream basics it's great to explore how you can go beyond the standard cosmology model. For example with quantum cosmology---cosmology based on quantum general relativity---models that remove the singularity and go back earlier in time, for example.