Wait a sec. How big is the Universe again?

By Phil Plait | August 8, 2006 12:32 pm

A couple of recent news articles have me scratching my head. And if I don’t get them, then I’m guessing that folks who aren’t astronomers will be even more confused. By themselves the press releases aren’t too bad, but taken together they contradict each other.

Sit tight. This’ll take a minute.

The Universe is expanding. We’re pretty sure about that at this point; we have about a century’s worth of observations backing it up. The idea is actually pretty simple: when we look at distant galaxies, they appear to be moving away from us. The farther away the galaxy is, the faster it’s receding.

There are a zillion lines of evidence for this, but the most basic one is redshift. You’ve already experienced this! When a race car goes past you, the sound it makes goes "rrrrrreeeeeeeeeeeeOOOOOWWWWWWRRRRRR". The pitch is higher when it approaches, then drops as it passes you. This is redshift! A similar thing happens with light: when an object approaches you, the "pitch" of the light — the color — gets higher, which means the light moves toward the blue end of the spectrum. When an object recedes from you, the color goes toward the red end.

And so it goes. We observe galaxies, and the farther away they are, the more redshift we see. That means the farther away they are, the faster they recede from us. So the Universe is expanding.

The key question is, how fast is it expanding? That is, how does the expansion speed relate to the distance? This is the question Edwin Hubble tried to answer back in the 1920s. He came up with the idea that the distance and speed of a galaxy are related by a simple constant, which we call the Hubble constant. The current value is about 70 kilometers per second per megaparsec (+/- 8 km/s/Mpc). What that means is that a galaxy one megaparsec (about 3 million light years) away will recede from us at 70 km/sec. A galaxy 2 Mpc away will recede at 140 km/sec, and so on.

Still with me? Now, the problem is that this is a BIG deal– for one thing, using some math, you can get the age of the Universe from the Hubble constant. Scientists are picky, and they like to have lots of independent ways of measuring something when it’s this important. So they’re always looking for new ways to see how fast the Universe is expanding. If these methods are totally independent of redshift, but get the same results (i.e., a Hubble constant of 70 or so) then you have more confidence we know what we’re doing.

There are lots of such methods. Some are more complicated than others. The folks at the Chandra X-ray Observatory just announced that they have independently confirmed the Hubble constant. Their method is a bit obscure for the non-scientist: they look at light coming from the background glow of the Big Bang, and see how the light is distorted by superhot gas floating around in clusters of galaxies (I told you it was obscure). This is called the Sunyaev-Zeldovich effect after the astronomers who figured it out (fun side note: Sunyaev’s daughter, who was very young at the time, threw up in a van I was driving once). The image below shows some of those clusters of galaxies. The purple stuff is the hot gas.

The thing to know is that this method really is independent of redshifts, and they get about the same value for the Hubble constant: 77 km/s/Mpc +/- 15%. That’s a pretty good value, and agrees with the currently held value.

But there’s a problem…

Just last week, another team of astronomers announced that using a different method, they get a Hubble constant of 61 km/s/Mpc (irritatingly, they didn’t give an error measurement for this value). That’s a lot lower than the currently accepted value, and the way the math works out that means they get a bigger, older Universe.

The method they used (looking at eclipsing binaries, stars that pass in front of each other, in a nearby galaxy called M33, pictured below) is pretty well-established, and should yield solid data. But so should the hot gas method! So we have two methods, both independent, and both generally good, getting different answers. And that’s what has me scratching my head. What do we do?

Well, there is more to know. For one, the eclipsing binary method used only one example of such a system in a nearby galaxy, and so it may not be terribly reliable. You want to have lots of binary systems to measure, to make sure you haven’t made some mistake. And more distant galaxies are better to use, since that gives you a longer baseline for your measurements. The problem is, stars get pretty hard to detect in distant galaxies, so we’re stuck with using ones that are close by. Worse, nearby galaxies are affected by our own Milky Way Galaxy’s gravity– for example, the Andromeda galaxy is actually approaching us, not moving away! The galaxy the binary team used is only a wee bit farther away than Andromeda, so I’m not sure how reliable it is to use– that alone might account for the 15% problem. But then, the astronomers on that team are smarter than I am, and have probably thought of that.

The hot gas method is weirder, but it also conforms to a lot of other independent methods that yield that 70 km/s/Mpc value for Hubble’s constant. And that value, when you crank through the math, gives an age for the Universe of about 14 billion years, which we know is about right from a zillion other lines of evidence!

So what do we do?

Well, as usual, we have to be a bit careful. First, remember we’re arguing over details, and not the overall idea that the Universe is expanding. Any creationist who uses this as "evidence" that even scientists can’t agree on the Big Bang should be poked in the eye. Second, as usual in science, time will tell. More observations are needed, and that will help nail down some of the specifics of all this. We’ve had lots of controversies like this in astronomy before, and they have always — always — resolved themselves in a logical and coherent way.

The Universe is pretty cool. It obeys a set of rules, and we can watch how it does so. Mysteries like this force us to pay better attention, and to rethink our ideas, our observations, and what we know. And from all that, the rules governing the entire Universe can be inferred. How amazing is that?

Comments (32)

  1. philw

    First, it’s the ‘visible’ universe which may have just been measured as larger. The univers most likely extends beyond the speed of light cone. Secondly, I recall that their error bar was around 6% and that additional measurements could reduce it to below 5%. These measurements would take 2 years. Somebody needs to verify or falsify here.

    One or another method has yet undiscovered systematic errors or there is some astrophysics we don’t understand as well as we thought.

  2. One should be able to calculate the error bar on the 61 km/s/Mpc value from the M33 “Distance Modulus” they give in Table 7 (last page of this PDF), but I’d have to learn what all those special astronomy words mean before I could do that.

  3. Rumour Mongerer

    Teach the controversy! If science isn’t right then the Bible must be! The Universe is 6,000 years old!

    (…etc, etc,…)

  4. And what range would that put the universe at, age-wise? 13-14 bill?

  5. Grand Lunar

    Ah, the joys of cosmology! Wait, does this fall under cosmology?

    Somehow, they’ll figure out this issue. They always do.

  6. Diederick

    I don’t get it (about the redshift thing): if farther away objects are more shifted to the red than nearby things (and therefore going faster), and those farther away things are further back in time, this makes me think that the universe’s expansion is actually slowing down, not accellerating, since old objects go faster than nearby objects. What am I missing here?

  7. Personal Crackpot Theory to explain the different results: Hubble’s Constant isn’t really constant!

    Space expands like a balloon. When I blow up a balloon, I pause to take a breath, and then start the next expansion forcefully. Soon I start running out of air and the balloon barely grows while my lungs collapse until I squeak. Measurements taken in different parts of space are therefore taken in different timeframes as well. Perhaps some measurements were taken at the beginning of a breath and others when alveoli were screaming for mercy.

    I did say “Crackpot”.

    This is one of my favourite problems in all of astronomy. Thanks for pointing out the latest contradictions.

  8. Max Fagin


    Your not missing anything (At least I don’t think so. . .) The universal expansion is slowing down.

    The question is, is it slowing down fast enough to colapse again. Scientists would like to know because it would let them predict when the universal “Big Crunch” would occur.

    Now if someone wants to complicate things and play the Dark Energy/Accelrated Expansion card, feel free to. I never understood dark energy anyway.

  9. Albert Einstein said (something like this): “God does not play dice with the universe”

    “rrrrrreeeeeeeeeeeeOOOOOWWWWWWRRRRRR” Could this be the sound of dice rolling on the cosmic table?

  10. Max Fagin

    Ok, something is wrong here.

    From the study of the microwave background, We’ve found that the farthest observable point in our universe is about 26 billion parsecs away.

    Observable Universal Radius = 26,000,000,000 parsecs

    Hubbles law states that the relative velocity of objects in our universe increaces roughly 70 km/sec for every megaparsec of distance.

    V (km/sec) = 1,000,000 parsecs x 70


    V (km/sec) = parsecs x 70,000,000

    so the V of the farthest observable object is:

    V = 26,000,000,000 x 70,000,000 = 1,820,000,000,000,000,000 km/sec = 6 TRILLION TIMES THE SPEED OF LIGHT.

    Things also go wrong if you work backwards:

    300000 km/sec (the speed of light) = D x 70

    D = 5000 parsecs.

    This would seem to indicate that due to Hubbles law, the stars still well inside our own galaxy would be accelerating outward at exactlly C. Does this mean that the 70 km/sec number only works for a certain range, or have I made a cosmilogical mistake? (Pun intended)

    Please someone check my work, I must have dropped several zeros somewhere for it to come out this bad.

  11. CS

    Dr. Bad Astronomer:

    Are these “zillion” lines of evidence really independent? I was under the impression that many of these Hubble results implicitly rely on other estimates of distances to local galaxies to calibrate measurements for longer scales. If there is an unknown/underestimated systematic in the analysis of, e.g. the Key Project, as suggested in the Bonanos paper (eclipsing binaries), then wouldn’t this affect a lot of these Hubble estimates?

  12. CS


    You have the parsecs on the wrong side of the fraction:

    H = 70 (km/s)/Mpc = 7e-5 (km/s)/pc

  13. Markk

    I’ve always felt that adjusting the Hubble Constant is something that must be done every generation just to shake things up. I must admit that this last result seems a little iffy to me because of the fact that the group got an incredible amount of press from a sample size of ONE. They may be right but seems a lot of press before solid results always leaves a taint…

  14. I have lots of comments.

    First: And what range would that put the universe at, age-wise? 13-14 bill?

    If you leave the Omeag_M and Omega_Lambda (current dark matter and dark energy densities, respectively) the same as the WMAP values (which may be foolish, since those measurements are convolved with H0), you get something a little over 16 billion years, whereas 13.7 is what you get with H0.

    Second: Hubble’s Constant isn’t really constant!

    The Hubble Constant is by definition constant– it’s the present-day value of the Hubble Parameter. The Hubble Parameter is not constant, though, and has been changing with time, so your theory isn’t as crackpot as all that! When we look farther away, we see farther back in time, because the light took longer to reach us. If we look back for 7 or 8 billion years, we see that the expansion rate has actually be speeding up, so Hubble’s parameter has been changing.

    Third: The universal expansion is slowing down.

    Actually, it’s speeding up. Dark Energy is the name for “the thing that we don’t know what it is”… it is what is making the Universe speed up. That, right there, is almost all we know about the nature of Dark Energy! The observation that led us to believe that there is Dark Energy is the acceleration of the Universe’s expansion.

    Fourth: Gotta pimp my own blog. Here’s my blog posting on this M33 result: http://brahms.phy.vanderbilt.edu/~rknop/blog/?p=85

    Fifth: cosmological redshift is not the same thing as Doppler shift! You can approximate cosmological redshift as Doppler shift for relatively nearby galaxies, but it really is a different effect. As such, talking about farther galaxies as “moving faster” is subtly wrong. The rate at which the distance to that galaxy is increasing is higher, but discussing the “speed” of a distant galaxy relative to us isn’t really meaningful. I discuss this at some length, and have another astronomer tell me I’m wrong, only to be confirmed that I was right by a General Relativist, here: http://brahms.phy.vanderbilt.edu/~rknop/blog/?p=68

    Sixth: Let me pimp my podcasts about the Big Bang and the expanding Universe:http://brahms.phy.vanderbilt.edu/~rknop/blog/?p=71

    Finally: I don’t get it (about the redshift thing): if farther away objects are more shifted to the red than nearby things (and therefore going faster), and those farther away things are further back in time, this makes me think that the universe’s expansion is actually slowing down, not accellerating, since old objects go faster than nearby objects. What am I missing here?

    You’re missing a few things. First of all, forget about the whole acceleration/deceleration thing. Even with a constant expansion rate, you will observe greater redshift from a more distant object. Consider the following: put seven paperclips on an elastic band, roughly evenly spaced. Now, stretch the elastic band, smoothly and at a constant rate. Each paperclip represents a galaxy.

    Suppose you are in the leftmost paperclip/galaxy. The next paperclip over is moving slower relative to that galaxy than the farthest paperclip over. In the approximation that cosmological redshift is Doppler shift, you can see that even with a constant expansion rate, you’ll get higher redshift for the more distant galaxy.

    Here’s a different way to think about it, that more accurately describes what cosmological redshift is: the wavelength of a photon expands at the same rate as the Universe. If you see a redshift, it is because the Universe expanded between when the light was emitted and when it was detected. Indeed, the ratio of wavelengths is the same as the ratio of the size of the Universe now to the size of the Universe at emission.

    Light from a more distant galaxy takes longer to reach us, because the speed of light is finite. Thus, when we see light from a distant galaxy, the Univesre has had more time to between the emission of the light and now than it has had for a nearby galaxy. Thus, in an expanding Universe, we see higher redshift for more distant galaxies.

    -Rob Knop

  15. PsyberDave

    Wow. Excellent post BA and all you other apparent astronomers! As comic book guy would say “Best post ever”. This stuff fascinates me. The bonus is, I now know how to spell the Doppler effect:


  16. Ben

    The pitch is higher when it approaches, then drops as it passes you. This is redshift

    Bad Bad Astronomer (Thwack! Thwack!) That isn’t a redshift. Its a doppler shift. It only turns into a redshift when light is involved.

    Thwack (with rolled up newspaper)!

  17. Bad Bad Astronomer (Thwack! Thwack!) That isn’t a redshift. Its a doppler shift. It only turns into a redshift when light is involved.

    C’mon, haven’t you ever heard of “White Noise”? What about “Pink Noise”? We use color metaphors when talking about sound all the time.

    A shift from blue to green is a redshift, even though there’s no red involved. Similarly, a shift from B-flat just down to A is also a redshift, even though there’s no red involved.


  18. Dark Jaguar

    Well you said the first method (red shifting) yielded 70 km/s, and that the second yielded 77 km/s, but you didn’t make a big fuss over the difference there. Was the 70 km/s more of a rough range, with that the lower limit of it? Not sure I understand.

  19. Chet

    Galactic Citizens,
    Well we’re all on this topic, I have another marvelously written book to recommend:
    “The View from the Center of the Universe–Discovering Our Extraordinary Place in the Cosmos” by Joel R. Primack and Nancy Ellen Abrams, 2006, Riverhead Books, ISBN 1-59448-914-9.

    Check out page 278 on how to make “The Cosmic Dessert” that will look like the Cosmic Density Pyramid described.

  20. Well you said the first method (red shifting) yielded 70 km/s, and that the second yielded 77 km/s, but you didn’t make a big fuss over the difference there.

    Phil said 77 km/s/Mpc +- 15% — that means it is a measurement of 77+-11 km/s/Mpc. That’s completely consistent with values of 73 or 70 km/s/Mpc, given the uncertainty on the measurement.


  21. “Sunyaev’s daughter, who was very young at the time, threw up in a van I was driving once”

    Does this say something about your driving? ;)

  22. jess tauber

    Sit right back and you’ll hear a tale
    Here at Bad Astron-o-my
    It starts X billion years ago
    With a singu-la-rity

    The past’s red shifted mightily
    The future’s blue and sure
    Blog passengers signed on to see
    Phil’s “rrrrrreeeeeeeeeeeeOOOOOWWWWWWRRRRRR” tour
    His “rrrrrreeeeeeeeeeeeOOOOOWWWWWWRRRRRR” tour…..

    Theme song from ‘Gilligan’s Island Universe’

  23. John Ploss

    The red shift justifies the belief that the universe is expanding.
    That assumes that the observed red shift is analogous to a doppler
    effect. Such easy assumptions are not necessarily true. There
    is no way to be sure what the red shift means. If the red shift
    does not indicate expansion then the red shift merely limits our
    power to observe and not the size of the universe.
    John Ploss

  24. That assumes that the observed red shift is analogous to a doppler

    Actually, it’s not. It’s a gravitational redshift.

    Re: the rest of what you wrote: yeah, yeah, yeah, typical creationist dissembling garbage. You can pick one thing and say, “dunno, therefore creationism might be right.” The fact is that redshift comes out of an extremely well tested theory and paradigm. Redshift is the direct measurement of the expansion of the Universe, but the whole Big Bang picture is supported by a vast quantity of observations. We’re really sure the Universe is expanding, as sure as we are that humans as a species evolved from earlier species.


  25. Crazy Bob-Astronomy To Go

    Jess Tauber,

    You are seriously and dangerously deranged. That is simply one of the best, most clever smart-*ss comments I have ever seen posted on here…and that is saying a lot. I’m scared to think what else might be going on inside your mind. Hope it comes out here more often. Thanks for the hardest laugh I’ve had this summer. Even my wife liked it and, being married to me, lost much of her sense of humor years ago. In her defense, she says the only reason she has let me live this long is her sense of humor. Oh well. Thanks again.

  26. John Crary

    It seems that the estimates for expansion of the universe are largely based on direct redshift of light coming from remote galaxies. The expansion calculations are then based on the amount of these redshifts without further modifications of the observed values.

    This would assume that once the light leaves its source galaxy that no more redshift occurs while the light traverses the space in between the source and the earth. Is this the current presumption? If a light source is 1 billion light years from earth, it then takes 1 billion years for the light to get to us and during this time the universe will experience an appreciable amount of expansion. It seems resonable to assume that the traveling light will also be affected by this additional expansion. If so, then the light we detect will not provide the same redshift that resulted when the light left its source. It will be the sum of the rate at which the source is receding, AND the amount of expansion that has occurred during its travels to earth. The farther away a light source is, then the greater the transitional redshift would be due to this transitional redshift factor, and would give the appearance of faster expansion for farther light sources.

    Any feedback regarding this curiosity of mine would be appreciated.

  27. Glenn Rottingen

    If we are now seeing quasars that are 15blys awary, we are also seeing quasars that are 15blys old. But they are the same observation. Now, bigbang, it would take 15blys at the speed of light for this quasar to get to this position and 15blys before we could observe it. Hence the universe must be at least 30blys old since the big bang.

    Now some believe that the early universe expanded faster than the speed of light. Why do they believe that. when gravity, electromagnetize, light etc all travel at the speed of light, a maximum velocity acording to relativity. The only reason that they believe that universe expanded faster than the speed of light, is that it is the only why their model of the big bang works. Now I would rather say the model is wrong.

    if someone can expain where this logic is wrong, please explain. glennro@gmail.com Astronomers have so accepted this big bang model, they cannot think outside the big bang box.

  28. Glenn Rottingen

    Additional comment, if the universes expansion is accelerating, then it would be decelearaing as you looked back in time. Hence since it was expanding slower in the past, the universe must be older than 30bys.

    (I have used blys billion light years, both a measurement of distance and/or a mesurement of time (should have bys for time).)

  29. E. Brull

    I have a hard time believing that the universe is finite or infinite. All we hear is red-shift, red-shift-, red-shift….[rare exceptions for blue-shift.] If almost everything we see is red-shifted from us, this would imply that we are close to the center of the universe which I find hard to believe. Every time we come up with a better telescope we add [usually] another one billion years distant with a larger red-shift. Taking heed to Einstein who says that gravity bends light rays [and in the same context, it could slow light's speed.]

    Presume that we look in any direction in the sky, could gravity bend the light from our star & galaxy such that we are looking at ourselves [hypothetically] one billion years ago? We know that we are not stationary in our galaxy & universe such that we are moving is some arbitrary non-linear direction and is dependent on perspective. If gravity can slow light, to an outsider we ourselves generate some degree of red-shift. So, if a complete circle/circuit takes a billion light-years we could be looking at ourselves 1 billion years ago with a certain self-inherant red-shift plus that generated by the gravity slowing of light. Then another 1 billion years further back we would see ourselves again PLUS the red-shift from the first circuit adds to that of the second circuit for the next sighting. Every circuit adds to the red-shift which we presume that farther objects are travelling faster & faster as apparent “distance” increases.

    If this theory, which astrophysicsts can probably inject lots of holes into it, the whole concept of “WHAT IS THE UNIVERSE?” would boggle the mind such that we have to say the universe is infinitely finite!!!!!!!!!!!!!!!

  30. Wildman987

    The title of this article is “Wait a sec. How big is the Universe again?” So, how big is the universe? Good question…

    A start might be to ask how far it is to the most distant object we can see. If the universe is about 13.7 billion years old, we could see something that is quite a bit farther than 13.7 billion light-years away now because of the expansion of the universe that has taken place during those 13.7 billion years.

    Imagine a photon emitted from a distant light source 13 billion years ago. How much has space expanded behind that photon as it has travelled to us. How much farther than 13 billion light-years is that light source now?

    This article is dated Feb. 21, 2005 and the exact age of the universe, plus the Hubble constant is a bit more accurately known today, but the idea is still valid.


    The distance they give to the furthest object we could see is about 46 billion light-years. (their guesstimate, not mine).

    As for how much more universe is out there beyond our visible horizon…good question…


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


See More

Collapse bottom bar

Login to your Account

E-mail address:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »