Some say the world will end in fire,

Some say in ice.

From what I’ve tasted of desire

I hold with those who favor fire.

But if it had to perish twice,

I think I know enough of hate

To say that for destruction ice

Is also great

And would suffice. -Robert Frost

Ever since the discovery of the radiation glow left over from the initial hot, dense state of the Universe — the cosmic microwave background — the Big Bang has proven to be the best description of the early Universe.

Image credit: Stephen van Vuuren.

This hot and dense initial state has, for the past 14 billion years, been expanding, cooling, and slowing down. This last bit is often overlooked, but it’s incredibly important. Because what the Big Bang model doesn’t tell us, on its own, is what the final state of the Universe is.

Because your intuition tells you that, sure, the Universe is expanding now, but gravity is an attractive force. Starting from a hot, dense, expanding Universe, you can easily imagine three different cases for its fate.

Illustration credit: NASA.

  1. Perhaps the Universe begins expanding quickly, but there’s a tremendous amount of matter in it! If there’s enough matter, perhaps your Universe will expand initially, with all the galaxies moving farther apart for some time, but gravity is dominant enough to halt the expansions, and even reverse it! In this case, the Universe will recollapse on itself, ending in a fiery demise known as the Big Crunch.
  2. Perhaps the opposite is true; perhaps the Universe begins expanding quickly but there isn’t nearly enough matter to halt and reverse the expansion. In this case, the bound structures in our Universe — galaxies, clusters of galaxies, and everything contained within them — will all continue to expand away from one another into an infinite abyss of space. Although the expansion rate continues to drop and slow, it never reaches zero, and can never reverse itself. This coasting Universe case is known as either the Big Freeze or the heat death of the Universe; an isolated, icy fate.
  3. Or, I suppose, you could imagine the Goldilocks case, where putting just one more atom in the Universe would give it enough gravitational mass to stop its expansion and recollapse, but instead the expansion rate asymptotes towards zero, never quite getting there.

Each of these cases assumes that the Universe contains matter and radiation, and the geometry of the Universe is simply determined by their presence, and of course by the laws of general relativity.

What’s interesting, astrophysically, is that each of these cases corresponds to a specific spatial curvature of the Universe! What do we mean by spatial curvature, and how would we measure it? Let’s give you a conceptual example.

If you had a flat sheet of paper, and drew a triangle on it, any triangle, you would find that the sum of its three angles is always 180 degrees. This is true for the Universe as well; if you summed the angles between any three points in the Universe, if its geometry is flat, those three angles would indeed sum to 180 degrees as well. This is what we expect to happen for a critical Universe.

But if your Universe were positively curved like a globe, your triangle would always have its angles sum up to more than 180 degrees. Try it if you don’t believe me! If you put one point at the North pole and the other two somewhere on the equator, it’s very easy to see, as each of the base angles are 90 degrees. For the Universe, this corresponds to the case of a recollapsing fate.

And your Universe could also be negatively curved, like the surface of a saddle. In this case, the angles always sum up to less than 180 degrees. And this corresponds to a coasting Universe.

Image credit: NASA.

Each of these cases for the Universe would have a different expansion history, so that if we looked at faraway objects (and hence also looked back in time), we could measure just exactly how the Universe has expanded over its lifetime, and hence what its fate was. And the tool for doing this was none other than the Hubble Space Telescope, capable of making incredible, precise measurements farther away than any other instrument.

In the late 1990s, there were two teams — the High-z Supernova Search and the Supernova Cosmology Project — that went out and made the crucial measurements.

Image credit: NASA / CXC / M. Weiss.

Whether they’re formed by a white dwarf accreting matter until it passes above a critical threshold (above), or by two white dwarfs merging with one another (below), one of the most useful objects in the whole Universe for determining great distances are Type Ia supernovae.

Image credit: NASA / CXC / M. Weiss.

Type Ia supernovae are so useful because their light-curves — how their brightnesses evolve over time — are so well-understood. If you watch a type Ia supernova over a long enough time period, you can determine what the intrinsic brightness of this event was.

Image credit: S. Blondin and Max Stritzinger.

And because you also observed the apparent brightness of the supernova, you can determine how far away it is! Combine that information with the observed redshift (i.e., how fast it’s expanding away from us), and that’s what it takes to determine how the Universe has expanded throughout its history. And as far back as you can accurately measure these supernovae, that’s as far back as you can know the Universe’s expansion history.

So these two teams, using the Hubble Space Telescope, set out to measure these distant supernovae as accurately as possible. In the graph below — from the Supernova Cosmology Project — there are three black lines: the top one corresponds to a “coasting” Universe, the middle one to a “critical” Universe, and the lowest one to a “recollapsing” Universe. So what’s our fate?

Image credit: S. Perlmutter et al.

The disturbing answer is none of them! What both teams found in 1998 was that the expansion rate will not approach zero, even an infinite time into the future, but will always remain some significant positive number!

Image credit: High-z team, left, and SCP, right.

In other words, as objects expand ever farther away from us, the speed that they move away from us increases, rather than slowing down!

This result, when it first came out, was met with a great deal of skepticism, because it would mean that the Universe is full of not just matter (both baryonic and dark) and radiation, but a new type of energy intrinsic to spacetime itself! (This is now known by many names, including dark energy and a cosmological constant.)

But the results have not only held up, but have only grown stronger over time, while alternative explanations — however clever and interesting — have fallen flat.

And when you put together the results of these supernova teams with the other great cosmological observations — that of large-scale-strcture (BAO, above) and the cosmic microwave background (CMB, above) — you find that, in fact, the Universe is dominated by this dark energy. Around 70-75% of the total energy density in the Universe today is given by this dark energy!

What does this mean for the expansion history — and fate — of the Universe?

Illustration credit: NASA.

It means we live in an accelerating Universe, one in which the objects which are not gravitationally bound to us right now (i.e., not in the local group) will eventually speed away from us and accelerate out of the Universe we can observe.

The most distant galaxies and clusters are already doing this! And it was the supernova data collected by these two teams that allowed us to discover the fate of our Universe.

So when I woke up this morning to read that the leaders of these teams — Saul Perlmutter from the Supernova Cosmology Project and (jointly) Adam Riess and Brian Schmidt from the High-z Supernova Search — were awarded the 2011 Nobel Prize in Physics, I couldn’t have been happier. It’s hard to argue that there’s any discovery in physics over the last 15 years that’s been more profound and deserving of this award.

Congratulations to all involved, including all past and current members of these two teams (not just the leaders), for discovering the fate of the Universe, scientifically!

So while our Sun eventually boiling our oceans, becoming a red giant and frying the Earth may be the fate of our planet, for the Universe I suppose Frost had it entirely right after all.

But if it had to perish twice

I think I know enough of hate

That for destruction ice

Is also great

And would suffice.

Congratulations once again, and for a little more info, don’t hesitate to read Sean’s, Steinn’s or Peter’s interesting takes, also, on the 2011 Nobel Prize in Physics!


  1. #1 Jack Dawe
    October 4, 2011

    The Big Rip is the Big Bang.

    At that point at which the individual quarks, or strings, in every atom begin to move away from each other at light speed, (thereby gaining infinite mass for the appropriate ones), a phase shift occurs, resulting in an apparent Big Bang everywhere, and another evolution of the Universe.

    Big Fire, not Big Ice. And twice.

    Or not.

  2. #2 Shahidur Rahman Sikder
    October 4, 2011

    See my comments at

    “An individual respective very location is the present and the rest all the locations are of the depth of the past” In or under the circumstances: The Universe as I see it, Big bang, evolution, gravitational world’s, time dimension or history of the universe or case or series of events or the structure of sketch of universe is the result one site and your lifetime. Early or Copy- such like- Digital CD i.e. Digital Universe found A-DEMO & see (Big Bang Video) Brief History of Universe.

  3. #3 Ben
    October 4, 2011

    So Lawrence Krauss is wrong in his assessment that the Universe is flat? He gave a really interesting talk at AAI 2009.

  4. #4 Ethan Siegel
    October 4, 2011

    Oh no, Ben, it’s flat. It’s just that there’s more to the Universe than matter, radiation and spatial curvature; there’s also this dark energy, which is in fact most of the Universe’s energy content (70-75%)!

  5. #5 The Bobs
    October 4, 2011

    Is Ω related to gaussian curvature? It sure looks similar to me.

  6. #6 Lloyd
    October 4, 2011

    If a universe falls in the space-time continuum and no one can “hear” it, does it make a sound?

  7. #7 Alan L.
    October 5, 2011

    On Sean’s link above one of the comments on the prize by Stacy McGaugh is interesting:

    It is great to see that the physics Nobel prize is recognizing the contribution of astronomers. It has always struck me as odd that to date, only observations of the microwave sky “counted” as physics, especially considering the missed opportunity to honor Hubble himself. Everyone agrees that the detection of dark matter particles – should that ever occur – would warrant a Nobel prize. So why not a Nobel prize for Vera Rubin and Albert Bosma, whose observations did so much to motivate the current search for dark matter?

    Perhaps we can expect to see Nobel prizes awarded to cosmologists and astro physicists more frequently in future as the newly enhanced – if ‘newly’ can still be used – LHC appears to be a damp squib as far as making new discoveries is concerned.

    (Note: I hope a full comment quote from another blog didn’t break any rule. If so, feel free to delete)

  8. #8 Phillip Helbig
    October 5, 2011

    The relationship between geometry and fate above holds only for no cosmological constant (obviously no longer an appropriate assumption). We can now say that the universe will expand forever, but we don’t know whether it is spatially closed (finite) or open (infinite) (for the experts: assuming a simple topology). The stuff above is misleading at best and, taken at face value, wrong. We are now sure the cosmological constant exists and is positive; relationships between geometry and future of the universe which hold only for the case of no cosmological constant are misleading, especially for someone new who might be trying to understand what is going on.

  9. #9 critter42
    October 5, 2011

    Especially in light of the Nobel award, what is your opinion of Christos Tsagas’ theory that the region we’re in is moving more quickly in relation to the area around it, giving rise to the illusion of an expanding universe that’s been popping up in the mainstream news the last week or so?

  10. #10 Phillip Helbig
    October 5, 2011

    “Especially in light of the Nobel award, what is your opinion of Christos Tsagas’ theory that the region we’re in is moving more quickly in relation to the area around it, giving rise to the illusion of an expanding universe that’s been popping up in the mainstream news the last week or so?”

    Several people have suggested similar things. Typical ad-hoc explanations. None of them have panned out.

  11. #11 Torbjörn Larsson, OM
    October 5, 2011

    we don’t know whether it is spatially closed (finite) or open (infinite)

    It is observably flat in the standard cosmology, so unless other predictions trump that, we know it is flat.

    This is also predicted by it (rather, a FRW universe) being a zero energy system, AFAIK. I.e. flat and eternal.

  12. #12 dcpetterson
    October 5, 2011

    What if we live in an oscillating universe, where a Big Bang is followed by a Big Crunch? Please bear with me.

    In the expansion phase, the farther back we look in time, the faster the galaxies are moving away from each other. As time goes on,the force of gravity slows them down, so that nearer (more recent) galaxies are moving slower. This is what physicists expected to see.

    What would it look like if we enter the contraction phase? We see the really distant galaxies still in the Expansion phase, as they were billions of years ago. But the nearer, more recent galaxies are already, like us, in the Crunch phase.

    In this phase, galaxies are accelerating — they are leaving behind the galaxies in the past, moving away from them at an ever-increasing rate as they fall toward the Crunch Point.

    Any galaxy we see has to be in the past, because it took time for its light to reach us. And unless it’s so very far in the past that it’s still in the Expansion Phase, any galaxy in our “immediate” past has to be moving slower than we are — because we’re father along in time, and therefore, farther along in the Crunch.

    Therefore, if we are already in the Crunch Phase of an oscillating universe, wouldn’t we see exactly what we are seeing? The really distant (far past) galaxies are still in the Expansion phase, and are moving faster as we go farther back in time and farther away in space. But the nearer ones are, like us, in the Crunch, and are moving faster as they are closer to us (father along in the Crunch).

    The graph of their velocities would look exactly like the 4-universe graphic in this article. In an Oscillating Universe, we would never see galaxies moving toward us as we approach the Crunch, because everything we can see is in our past and therefore moving slower than we are.

    Perhaps we’re not seeing the effects of a previously-unknown Dark Energy. Perhaps the accelerating we’re seeing seeing is the effect of normal gravity, accelerating us toward the Big Crunch.

  13. #13 Jack Dawe
    October 5, 2011

    No. In that case, light from those objects would be blue-shifted, as they fell back toward us.

    My suggestion (see first comment) is a variant of the oscillating universe. Everything accelerates until almost at light speed (given dark energy) and then undergoes a phase transition into a kind of widespread Big Bang. Wait fifty billion years or so, and dark energy has its day again, and the process repeats. But on an infinite wave front, instead of from a singularity.

    If nothing else, this would account for the virtually flat Universe we see out there, which implies one of those pesky infinities or something close to it.

  14. #14 dcpetterson
    October 5, 2011

    Jack, I understand the idea that the light would be blueshifted. I suggest that it wouldn’t be, because we’d still be moving away from the galaxies “behind” us. Light can only be blueshifted if objects are moving toward each other, or if the space between them shrinks.

    It seems to me there’ be a universal “tidal effect”, as near the event horizon of a black hole, where objects closer to the event horizon are pulled in faster than the objects behind them, and everything stretches out. That stretching would cause the light from objects behind you to be redshifted.

    Your concept is certainly viable as well, and reminds me of some of Hoyle’s “steady state” ideas — but more a sort of cosmological equivalent of the punctuated equilibrium of biological evolution.

  15. #15 Jack Dawe
    October 5, 2011

    If the visible universe were alot smaller than the universe we see, or if all visible objects were alot closer than actuality, there might be room for dcpetterson’s theory to operate.

    Unfortunately, even the visible universe is huge beyond belief (around 27 billion ly in diameter, in every direction), and the unseen universe may well be infinite (there’s at least a remote possibility that it is).

    Galaxies are receding from us almost exponentially according to the distance at which they’re removed — which is enormous already. I just see no way – particularly given the evidence of dark energy — that by some sleight of perspective they’re now (in “actuality,” as if light weren’t its own actuality) moving together again.

    I get that there are models of the universe whereby what’s rushing away from you is actually getting closer in the opposite direction. But, like I said, that would be alot smaller universe, and would require some sort of extra-dimensional fancy footwork to make even theoretically possible.

    As for the Steady State theory, the odd thing is that it started out on the micro-scale (atoms popping into existence, generating material for stars between the stars), but now almost every theory leads back to it on the macro-scale. I’m not sure Fred Hoyle would be pleased by that irony, but it is what it is.

  16. #16 dcpetterson
    October 5, 2011

    Thanks for the feedback, Jack. I’m looking for people to poke holes in this idea (that’s how science works — you try to falsify theories). So I greatly appreciate the criticisms.

    The idea that the universe could have begun in a singularity certainly lends credence to the idea that it could end in one. If it was not too big to start that way, it is not too big to end that way.

    Ethan’s article advances a “recollapsing universe” as one of the plausible models. The reason this model is not currently one of the actively-considered ones is not that the “actual” universe (as opposed to the “visible” universe) is “too big.” It’s that cosmologists are trying to explain continued expansion.

    My theory merely offers an alternate explanation for the observed accelerating expansion between an observer and galaxies in the (relatively recent) past. My model is virtually identical to the “recollapsing universe” model–which, I hasten to mention, doesn’t actually consider what a collapsing universe looks like from the point of view of an observer in the era of collapse. I am thinking about what that would really look like. And from everything I understand, it would look like exactly what we see.

    In a collapsing universe, the observer would always be ahead of the portions of the universe that are visible. All visible galaxies would be redshifted, and the velocity of the observer would be increasing.

    The observer would always be accelerating away from the portions of the universe that are visible, because those portions are in the past. The observer would never see blueshifted galaxies–because anything blueshifted would have to be on the other side of the ending singularity, and that would be an impossibility.

    Again, I thank you for your reactions and your criticisms. They are helping me to refine my model :)

  17. #17 Jack Dawe
    October 5, 2011

    Appreciate it, dcpetterson. You got me rethinking this matter, which is a subject I’m interested in (though, lacking the maths, I’m kind of like the musician who doesn’t read music).

    The most interesting theory out there right now is Dark Flow, which — if it involved some sort of entangled extra-dimensional mass (an extension of Dark Matter) — would dispense with the need for Dark Energy, thereby invalidating the entire scenario I described above (Big Rips creating oscillating Universes, etc).

    Everything, unfortunately, is contingent on observational fact. Only where observation doesn’t reach, are we allowed theories.

  18. #18 OKThen
    October 6, 2011

    Good summary Ethan.
    And all the best regards to the winners. Their award is well deserved.

    Your converging chart (i.e. SNe, CMB, BAO, flat)is new to me; is there a link you can point me to that discusses the convergence of these ideas a bit more?

    Someday, maybe we’ll have an explanation of what this dark energy is. Questioning the observations is pointless; but a theory with a deeper understanding would be nice.

  19. #19 galen
    October 6, 2011

    I’m sure you’ve heard Tsagas’ proposal that the expansion is not accelerating. He claims the apparent acceleration is an artifact of the motion of the motion of our neighborhood (a rather large neighborhood) relative to a family of fundamental observers. I’m not at all sure how it’s supposed to work. Given the recent prize, I’m guessing Tsaga’s hypothesis is pretty unlikely. Can anyone offer some enlightening comments?

  20. #20 Ethan Siegel
    October 6, 2011


    I will write a post on Tsagas’ proposal for you (and everyone) either later today or tomorrow.

  21. #21 galen
    October 6, 2011

    Looking forward to it. Thanks.

  22. #22 Collin
    October 6, 2011

    Ethan, when you say the universe is flat, I’m assuming this means that some components of the Riemann tensor average to zero on a very large scale — since the Riemann tensor is the only observable pertaining to the shape of the universe.

    IIRC, the flat FRW model is infinite. I’ve never understood how an infinite universe could have arisen from a point-like Big Bang. (FWIW, the alternative I’m alluding to is a very small positive curvature.) If the FRW model is wrong (I’m not sure it can accomodate dark energy), does the universe perhaps have an “unnatural” finite topology despite being metrically flat?

  23. #23 idealist707
    January 8, 2012

    Although this discussion ended months ago, I can’t miss an occasion to post an opposing opinion.

    DCPetterson’s concept would seem to require the observer to be in a special position for all observed to be “behind, ie in the past”, thus being in the expansive phase.

    Science discarded that any special position exists a few centuries ago. Perhaps that’s why it is so easy to fall back into our anthropocentric mindset.

    All points being equal means that all observers will see a blue-shift; assuming that time remains the same—–which is not given.

    PS Recently had an exam question on just what did they discover giving them the Nobel Prize. Thus my interest long after this discussion.

  24. #24 Leonardo Rubino
    April 29, 2012


    in order to say my opinion on bangs and crunches, here is a naive paper on the Oscillating Universe:

    It’s also against the (unfound) dark matter, so somewhat in agreement with what just observed by astrophysicists in Chile.

    It’s naive, that’s why it’s worrying me.

    Thank you.

    Leonardo Rubino.

  25. #25 jayne
    October 2, 2012

    if the universe is expanding then im getting smaller