Ask Ethan #39: Why does light stretch as the Universe expands? (Synopsis)

“They say the universe is expanding. That should help with the traffic.” -Steven Wright

When you think of the expanding Universe, you very likely think of galaxies moving away from one another, of space getting stretched, and quite possibly, of light redshifting as it travels through this expanding, stretching space.

Image credit: James Imamura of University of Oregon, via Image credit: James Imamura of University of Oregon, via

But does it have to be this way? In other words, these galaxies are moving away from us; are we sure that the light is getting redshifted because space is expanding and not, for instance, because of the rapid recession of these individual objects? After all, Doppler shifts can redshift light, too.

Image credit: Larry McNish of RASC Calgary Center, via Image credit: Larry McNish of RASC Calgary Center, via

We do know the answer. Go read the whole thing and find out for yourself.


More like this


In your "Why does light stretch" it seems like you have avoided a conceptual issue that many folks would raise. That is: If I have a meter stick and watch if for a very long time (while space expands), I would be tempted to assume that it is still a meter long. And if I find any other meter stick of the same age, they will all appear to be the same length, yes? But if I get a new meter stick, defined by so many waves of sodium light or whatever, only then will I realize that the old meter sticks are all too long. Yes?

I am always amazed that these kinds of discussions don't include this kind of question.

The balloons represent different quantum states of spacetime. The size of the balloon is governed by Heisenberg's Uncertainty Principle.The larger the balloon the lower its energy. The energy levels are separated by an energy gap of E=hH where h=Planck const. H =Hubble const. in SI units. There are 10^60 energy levels. The cosmological constant is lambda=3(E/hc)^2= 1.76 x 10^-52/m^2

Proving gravitational blueshift is not the same as proving special relativistic redshift ! Plz correct.

And more. You write "It’s no secret that our observable Universe began with the Big Bang: a set of initial conditions that tells us our Universe was hotter, denser, more compact, and expanding rapidly in the past. "

Wich i agree with. But do you understand the conseqences ? If we look deep in space, we also look back in time. We see closer to the big bang. We should see a hotter universe. We should see a more scattered freq. band. More colors, less red. Only relativistic redshift can explain why we dont.

By scienceblogs fan (not verified) on 30 May 2014 #permalink

So, she asked me: why can only relativistic redshift (RR) explain the cosmic microwave background (CMB) as we see it?

Well. Look at the picture of the CMB. Choose any 2 pixels. Lets name them A and B.

Did they travel towards us at the same speed ? I would say yes. Its save to assume that.

Did they travel the same distance ? I would say no. Its save to assume that they never touched each other.

Should there be a RR in A ? Yes. It traveld a long time. So it should show a redshift linked to a expanding universe.

Should there be a RR in B ? Yes. It traveld a long time. So it should show a redshift linked to a expanding universe.

Did they travel te same distance? No. So they did not travel the same time. Lets call those tA and tB.

Because tA is not the same as tB. The RR in A is not the same as the RR in B. Lets call them RRA and RRB.

Are there other forces that cause a RR ? Yes anything on the path on any given time on any given distance. Lets call this RRr. The r from random. Those effects can only be random.

Is A almost the same temp. as B ? Yes. And that feels as a problem.

They start out at the same temp ? Probaly not. We know also that if A traveld longer then B, than A started in a hotter universe. We can assume at the same time that the most energitic radiance has more chances to reach us after such a long time and distance traveling.
So we have tAs and tBs. We can savely state that they stand for the hottest of there time.

Have they cooled off since they travelled ? Yes. They traveled trough the same cooling universe. Lets call that Td. From temperatuur drop.

Can we legaly simplify and say A and B have the same temperature ? Yes. We the differences are very small and we simplify a lot to understand a lot of things. So there temperature is a constant. Lets call it T0.

So for A we can say:T0 = tAs - (RRr + RRa + Td)
So for B we can say:T0 = tBs - (RRr + RRb + Td)

Now lets focus on how it is possible for T0 to be a constante. Lets accept that RRr stands for the noise in the picture. Lets accept that Td would not produce a differend temp for A than for B.

Then we can assume.
t0 = TAs - RRa = TBs - RRb

Or. RR leveld out the differences. And. If we take a CMB fotograph, it will be the same at any given moment. Now and over a biljoen years.

By scienceblogs fan (not verified) on 31 May 2014 #permalink

maybe it's nothing to do with expansion

maybe light from distant places is more red-shifted because the light just got real tired from traveling such a long way

or maybe we live in a really (gravitationally) dense part of the universe (i offer sum human IQ as supporting evidence) and light that comes here slows down a lot just as it does when entering any denser medium

i remember reading about Ptolemy's model getting ever more complex as more and more evidence was squeezed into it

it predicted everything

but it was wrong wrong wrong and the truth was more simple and more beautiful


By The Peak Oil Poet (not verified) on 31 May 2014 #permalink

Peak oil Poet,

Your hypothesis falls down on one single, simple observation. According to your "we live in a gravitationally dense part of the universe" idea, ALL light from ALL celestial bodies should be red shifted. There should be no observations of blue-shifted light from any celestial object. That flies in the face of actual observations, however. There are many galaxies in our local group that do indeed show a blue shift. The local gravitational force simply overcomes the expansion of space for these (relatively) close objects.

Great article Ethan.

I especially liked the intuitive example of the kinetic energy of the particle pair shifting the wavelength of the photons produced from annihilation, which is needed for conservation of energy.

My question is, if photons traveling through expanding space already, and their wavelength shifts down, does the energy from the photon "get absorbed" by the expanding spacetime or something to that effect. Energy obviously can't just disappear, so I've always wondered where the photon's energy goes (and I assume the above because the photons created by the dropping particles have basically "stolen" energy from the gravitational field of the planet, though I guess if they radiated outward that energy would then be returned to the gravitational field).

Any help with this concept would be greatly appreciated.

Sean T

i did not posit that all light arriving would appear red shifted. Blue shifted light would be slowed but would still be blue shifted if it was going fast enough to begin with

you don't have to leave our galaxy to see blue shift

my idea is simply that our explanation for what we observe could be quite wrong

after all, what seems more logical - that the universe started from nothing and expanded to be as big as it is

or that it's always been as big as it is but we can not yet understand why it appears to be expanding

even if we, like Ptolemy, think we do


By The Peak Oil Poet (not verified) on 08 Jun 2014 #permalink

Peak oil poet,

You seem to be confused about doppler shifting of light. The doppler shift reflects the relative velocity of the source and receiver, not the velocity of the light itself. The light does not red shift because it slows down. It red shifts because the source is moving away from the detector. The speed of light is always c. BTW, that's true even in a dense medium, at least in the strictest sense. When we measuere a reduced speed of light in a dense medium, we are looking at the large-scale behavior of light. When travelling in a dense medium, photons travel for some distance at the speed c before interacting with an atom. That photon then is absorbed and a photon of the same wavelength is reemitted. In that manner, the light is slowed as it travels through the medium, but between atomic interactions, the photons still travel at speed c.

Now your idea implies that the photons are redshifted by a large gravitational field. As Ethan has explained, this is a real effect in relativity. However, unless you have some other explanation as to why relatively distant galaxies ALL are moving away from us whereas relatively close ones show a mix of moving away or toward us, then universal expansion is the best explanation. Without this expansion, there's no explanation for the fact that only relatively close galaxies have velocities directed toward us. Your idea would make it much more likely that a given galaxy would be redshifted rather than blueshifted, but it gives no explanation of the fact that the only galaxies we see blueshifted are relatively close ones. If your idea were true, there should be no correlation between blueshifted galaxies and distance. Without universal expansion, why should distant galaxies be expected to be moving away from us?

You always talk about inflation happening before the Big Bang, but just about every other science article I read says that inflation followed the Big Bang. So which is it really?

By Neil Nelson (not verified) on 26 Sep 2014 #permalink

@Neil Nelson #10: The confusion arises because the term "Big Bang" has evolved over the 70+ years since the model was first introduced to cosmology.

At that time, the purely classical, general-relativity based model started with a singularly (infinite density, zero scale factor) in a Lemaitre-Friedman-Robertson-Walker metric, which then expanded according to Hubble's law in the form we see it now. In that original model (Gamow, Alpher and Herman), the term "Big Bang" referred to the initial singularity, and necessarily everything that happened in the universe happened _after_ that "Big Bang."

When Guth developed his inflationary model in the 1970s, he was still working in that original paradigm, and posited inflation as happening during a very short (~10^-30ish seconds) period after an initial singularity. In such a model, the inflation itself would drive the universe into a cold state (you could think of it like a Hubble expansion lasting ~forever), and the end of inflation had to be followed by a "reheating" in order to get "hot big bang" conditions for nucleosynthesis etc. But still, this ~40-year-old model has everything happening after an initial singularity.

Modern cosmology is more fully informed by ideas from quantum field theory, and less tied to Lemaitre and Gamow's initial, classical vision. One key point, and the point which Ethan uses to justify his formulation, is that inflation itself erases all evidence of an initial singularity! Because of inflation, our entire observable Universe (the 13.8 billion year old, 92 billion light-year diameter ball we can see) is only a very tiny, tiny piece of whatever was doing the inflation.

More interestingly, because we only have access to a tiny patch of the pre-inflation universe, we can't make any statement at all about how long inflation might have been going! Once even a few e-foldings happened, our tiny patch would be so close to flat that any more expansion wouldn't make a measurable difference. The very interesting, and counterintuitive, consequence of this is that inflation _could_ have been going on "forever" (indefinitely)! All that we can experimentally touch (e.g., through CMB anisotropies, polarization, etc.) is the _final_ 10^-30ish seconds of inflation.

That last little bit is followed by the _reheating_ necessary to get a hot initial state for particle production, nucleosynthesis, etc. What Ethan is doing, is basically taking the idea of "eternal inflation" above at face value, and defining the t=0 time of the universe, the "Big Bang", to be the reheating event itself.

This is, quite clearly, in contradiction to the way most other popular descriptions of cosmology are written. But it is much closer to the way actual, modern cosmology is formulated and analyzed by the active researchers in the field.

By Michael Kelsey (not verified) on 26 Sep 2014 #permalink

Could somebody answer Peter's question from June 5, 2014.

How is energy conserved as e.g. a CMB photon's wavelength increases due to space expanding?

I have been wondering the exact same thing for 5 years, since I first heard about cosmological redshift.