Why gravitational waves from the early universe are a big deal

Today, the BICEP2 team announced the detection of what they claim is an imprint of long wavelength gravitational waves in the polarization of the cosmic microwave background.  If this holds up (a big if:  lots of exciting discoveries don’t hold up when some neglected systematic error turns up), it will be the most important discovery in cosmology since the first evidence for dark energy, and for physics in general I would rate it more important than the detection of the Higgs boson.

How far back in time we can see

Because light travels at a finite speed, when you see an object, you actually see it as it was in the past, rather than right now.  For objects in the same room, this delay is only nanoseconds, but for distant objects it can be big.  For example, light from the sun takes 8 minutes to reach us, so we see the sun as it was eight minutes ago; the star Sirius is 8.6 light years away, so we see it as it was 8.6 years ago.  So, looking far away is like having a time machine.  The universe is about 13.8 billion years old, so you might wonder, can we look at a region of the universe so far away that light takes 13.8 billion years to get from there to here, meaning we would see this spot as it was during the big bang?  Can we see the moment of creation itself?  The points of the universe that we should see as they were at the beginning of time form a vast sphere centered on the Earth called the “cosmological horizon” because it divides the part of the universe we can see (at some point in its history) from the rest that we can’t see at all.  (Note that every observer, not just Earth, is the center of its own cosmological horizon; no point is special.  The big bang happened everywhere.)

Unfortunately, we can’t see that far back, because if we go to early enough times, the universe is so hot and dense that light cannot travel freely.  Instead, the oldest light we see comes to us from the moment that space became transparent and light could for the first time travel freely.  This radiation is the cosmic microwave background, and when we look at it, we are seeing the universe as it was when it was about 400,000 years old.  We cannot see farther back, because we are essentially hitting a surface inside of which space is opaque, just as we can’t see deeper into the sun that the photosphere, where light stops scattering and becomes able to travel freely.

Other messengers

Is there a way to get around this limit?  In fact, we are able to test and confirm the big bang model back to about a second after the beginning, because if the process from weak interaction freeze-out at this time to the formation of primordial elements (especially most of the universe’s helium) a few minutes later were much different than believed, the abundances of these elements would not be predicted by the theory correctly as in fact it is.  A lot must have happened in that first second, though, and we’d like to know more.  The origin of the universe is a fascinating problem in itself, of course.  In addition, there’s the fact that as we study times closer and closer to the beginning (“t=0”), densities and temperatures get higher and higher (hypothetically being infinite at t=0).  At sufficiently early times, densities and temperatures were higher than the universe has since seen, even reaching levels inaccessible to particle accelerators.  Thus, the big bang is an opportunity to study the laws of physics, probing very high energies (which for quantum systems corresponds to studying very small scales).

One might ask how we were able to get around this problem for the sun.  How do we know what’s going on in the sun’s center?  The answer is to look at the sun through something other than light, something that is generated in the core and isn’t scattered while traveling through the sun but comes directly to us.  For the sun, that something is neutrino radiation.  Neutrinos interact with matter much more weakly than electromagnetic radiation, which makes them very hard to detect, but it also means that they can travel straight from the center of the sun or of a supernova explosion to our detectors.  There actually should be a (undetected) cosmic neutrino background, coming from when the universe was a little under a second old and became transparent to neutrinos.  However, t=1 second is already pretty well covered, so we need something even better.  And, luckily there is something.

Gravitational waves

Gravitational waves (note:  distinct from gravity waves like the ones you see on a lake) are wavelike distortions of spacetime generated by accelerating massive objects.  Like electromagnetic waves, they travel at the speed of light.  However, they interact with matter extremely weakly–only a black hole can block them.  In fact, we basically don’t have to worry about space ever having been opaque to gravitational waves (at least not until we’re so far back that the description of spacetime we have breaks down), so the only worry is whether or not a strong signal of these waves was ever made.  If it was, then it is still carrying information from these unexplored early times.

By the way, there are big hopes that gravitational waves will soon become a tool for astrophysics as well, once the technical challenges of detecting them are worked out.  Much of my own research is tied to these efforts.  The main sources of gravitational waves in the contemporary universe are various combinations of black holes and neutron stars smashing into each other.  Thus, from my point of view, gravitational waves are a tool for studying black holes and neutron stars.


A priori, there’s no reason to think we’d get lucky and find a big gravitational wave signal from the early universe.  Why should the early universe make one?  In fact, such a signal is predicted by the theory of inflation.  Inflation was originally proposed as a way to explain what seem to be unexplained coincidences in the standard big bang theory.  The basic idea is that at some point very early on, say 10^{-35} second after the beginning (we’re talking a lot earlier than t=1 second now!), the universe’s expansion went through an explosive exponential phase, growing by many orders of magnitude; then this phase ended, and the universe has since expanded as predicted by the regular big bang theory.  Why would this happen, and why would it stop?  Nobody really knows, so instead we’ve given the culprit a name, the “inflaton field”, to label our ignorance.  This field apparently acted somewhat like today’s dark energy (which is driving the universe into another exponential growth phase), albeit at a very different scale.  Then for some reason it went away.  How this idea solves problems for the big bang model is something I won’t go into.  Basically, postulating an enormous blow-up means that before that blow-up, the observable universe was smaller than we would have expected, and was better able to settle itself down in various ways.  What is important is that inflation would take small quantum fluctuations and blow them up to macroscopic scale.  This seeded the universe with small inhomogeneities in density which eventually grew to become the structure (stars, galaxies) in the universe we see today.  There should also be quantum fluctuations of gravitation seeded at the same time, meaning an ancient gravitational wave signal generated very, very shortly after t=0.

These gravitational waves would be tremendously difficult to detect directly. Fortunately, we don’t have to.  As they travel through matter, they distort it in various ways–squeezing in one direction, stretching in another.  This leaves a subtle imprint in the cosmic microwave background, in particular on the spatial pattern of the polarization of the light.  And this is what BICEP2 is claiming to have measured.  Because standard inflation models make distinct predictions about these gravitational waves, there is the hope that the theory of inflation can really be tested–its distinctive predictions rather than just its general post-dictions.

A peak at the highest energies we may ever be able to see

If the result holds up, then the BICEP team have found a signal that comes to us from a form of matter and energy that is utterly inaccessible to any conceivable experiment.  They’re talking about characteristic energies of order 10^{16} GeV.  For comparison, the LHC achieves collision energies of around 10^4 GeV.  The scale 10^{16} GeV has actually long been of interest to high-energy physicists as the proposed “grand unification scale”.  The idea is that, although in ordinary life, the electromagnetic, weak, and strong nuclear forces have very different strengths, these strengths change with energy scale, and if you extrapolate the curves 12 orders of magnitude beyond experiments (yes, I agree, this is absurd, but bear with me for a second), they all cross.  Does this mean anything?  I wouldn’t bet on it, but its been taken as a clue that this is the scale above which these forces cease to be distinct.  10^{16} GeV is also only a few orders of magnitude lower than the Planck energy, at which gravity also has the same strength and presumably unifies with the other forces.

Thus, gravitational waves from inflation would be our only signal from these exotic scales.  As far as we now know, it’s the only one we’ll ever have.

24 thoughts on “Why gravitational waves from the early universe are a big deal

  1. This is not by any means the first time that it has been claimed that gravity waves have been detected – it is something of a recurrent phenomenon. (I wrote a journalistic article on gravity waves in 1990, so naturally I am an expert…).

    One problem seemed to be that people knew exactly what they were looking for – this was merely ‘confirmation’ science, rather than any kind of test – which made it easier for researchers to convince themselves (and others) they had actually found it.

    Are this team competent? – more importantly are they absolutely HONEST? The prizes for pseudo-finding gravity waves include potentially vast levels of funding which tends to produce exactly the kind of *distortion* that they claim to have measured.

    Big Physics was probably the first field of science to be corrupted (being the first kind of Big Science) so I am not uncorking the (metaphorical) champagne quite yet.

    • I don’t think the history is that bad. The first claimed detection of a gravitational wave effect, the inspiral of the Hulse-Taylor pulsar binary system, has held up pretty well. The only claim for a direct detection that was taken seriously by the scientific community was from Joe Weber’s bar detectors in the 1970’s, but that claim was quickly rejected when other groups couldn’t reproduces Weber’s results. The Initial LIGO, Virgo, and GEO interferometric detectors were online for years with no claimed detections. They know they’re hunting for signals smaller than their noise, so they’re very careful in assessing statistical significance. This result may end up being like Weber’s, and we will find out soon as other experiments try to reproduce the detection.

      It is true that we can only detect a gravitational wave effect if we know ahead of time pretty much what the effect should be. This does mean that one can only claim detection to the extent that one can exclude any other possible source of the effect (CMB B-modes, little wiggles in detector arm lengths, vibrations in the bar, etc.). It also means that we have trouble detecting anything really unexpected. (The LIGO team does have a group working on the detection of unmodeled bursts; you can find things without precise search templates if the signal is big enough. If a supernova goes off in our own galaxy in a decade, we might detect it in gravitational waves, depending on the explosion mechanism.)

  2. The points of the universe that we should see as they were at the beginning of time form a vast sphere centered on the Earth called the “cosmological horizon” because it divides the part of the universe we can see (at some point in its history) from the rest that we can’t see at all. (Note that every observer, not just Earth, is the center of its own cosmological horizon; no point is special. The big bang happened everywhere.)

    I’ve never understood this. Probably I’ve heard explanations and then forgotten them serially. How am I supposed to visualize the expansion of the universe? If I visualize it like a regular explosion and I visualize us as living in the flying debris, then obviously there is nowhere we can look and see the explosion. The light from that flew past us right at the moment of the explosion never to return.

    • Visualize drawing points with an ink pen on a deflated balloon, then inflating the balloon. Each of the points on the surface of the balloon expand away from each other.

    • Hi Dr. Bill,

      We often talk about the big bang being an “explosion” but this isn’t really a very good way to describe it; it causes more confusion than it clears up. An explosion happens at a particular place, and then all the debris flies away from that place, expanding into some sort of empty outside region. Better to drop that idea and just think of the universe as having always (since the big bang, that is) been infinitely large and homogeneous (i.e. the same density and temperature everywhere).

      Imagine a line extending infinitely far in both directions, and on the line are a series of dots with uniform separation, say 1cm. There are an infinite number of dots in each direction, each with a neighbor dot 1cm to its left an another 1cm to its right. Each dot represents a region of space, which at the current phase of the universe’s evolution might be the location of a galaxy. Now, let us say our universe “expands”, meaning that over a period of time the separation between dots doubles to 2cm. Now any given dot will see the dot to the right being 2cm away when it used to be 1cm away, the dot to the right of that one being 4cm away when it used to be 2cm away, and so forth. To sum up, the dot sees every other dot flying away from it, and the farther away the other dot is, the faster it seems to be moving. But since every dot would see this same thing, this is still a homogeneous universe.

      How did the distances get bigger? One way would be for all of the dots to have moved around–a velocity proportional to position would have this separating effect. However, very distant regions of the universe are receding from us apparently faster than the speed of light. What’s really going on is that space itself is “growing”, so that even though galaxies aren’t moving much, the distance between them is getting bigger.

      Now, if the distance between dots is growing, then in the past it was smaller. Extrapolating backward in time, there might have been a point when the distance that is now 1cm approaches zero. Even right after the big bang (it’s subtle to talk about how things were right at t=0, because spacetime is singular there), the universe is infinitely large and uniform, with some points very far away from the point where we will someday be, and it’s only “exploding” in the sense that density and temperature are very high everywhere, and the space between regions of space is growing.

      To sum up, the big bang happened everywhere, so it’s a matter of when rather than where. If you could look at any point in space as it was 13.8 billion years ago, you would see the big bang happening. Looking into space, we can divide the universe into imaginary spherical shells centered on Earth. We see each shell at a different time, given by how long it takes light to get from there to here. Shells far enough away will be seen as they were when the universe was very young.

      • Thanks, that’s helpful. Because the big bang happened everywhere, there are always a lot of somewheres which are age-of-the-universe-in-years light years from me. So I always see the big bang, though increasingly far away with time (and red-shifted? increasingly red-shifted with time?).

        Suppose it is 364 days after the big bang on the line. There is dot one light year from me. I don’t see it because the light has not had time to get to me yet. It seems like I should be seeing it tomorrow (plus a little bit to account for the expansion that happens in a day). Is this right? In principle, do we get to see more stuff each year? And after that day plus epsilon, the big bang now appears to be behind the new dot?

      • Hi Dr. Bill,

        Yes, that’s the basic idea. The cosmological horizon expands as the universe ages. We get to see parts of the universe we couldn’t before, and right when a region becomes in-principle visible, it is in-principle visible as it was at the moment of the big bang.

        Because of the space-growing effect (expansion of the universe), which is now accelerating, this process is expected to asymptote. Eventually, the universe will be expanding (meaning the distance between galaxies will be increasing) nearly as fast as the horizon expands, and then we’ll have seen about as much of the universe as we ever will. That’s a detail, though.

      • Bonald,

        Is the universe actually infinitely large? Or did you write that to make your example more clear?

        Thanks for the post, very informative.

      • As far as we know, yes the universe is infinitely large. Of course, we’ll never really be able to prove it because we can’t see past the horizon. It is also possible for the universe to be finite while still not having edges if it has some sort of nontrivial topology. However, there’s no evidence of such a thing right now.

  3. Bonald,
    Doesn’t the idea of cosmic expansion depends on the assumption that our solar system is not located in a special place in the universe?
    Since all we observe are galaxies receding from US and assuming that we are not located in a special place, it implies that everyplace is receding from every other place.
    Is my summary right?

    Now, how is the basic premise, that we are not located in a special place, to be justified?
    Does it an empirical basis or an aesthetic basis or some other theoretical basis?

    • Recall that we don’t just see other distant galaxies receding from us; we also see that the ones farther away appear to be moving faster, with distance and velocity having a nearly linear relationship. For this sort of a flow, we can imagine ourselves in the place of another galaxy and ask what the flow would look like from its perspective. This is easier to visualize for the 1D example I gave Dr. Bill. You can easily convince yourself that any of these other galaxies should also see galaxies to its left moving to the left, galaxies to its right moving to the right, etc. We also can check that distant galaxies with a large recessional velocity from us do not see a doppler shift in the cosmic microwave background (CMB) spectrum. This is done by observing how CMB photons scatter off the hot gas in between galaxies in clusters, the so-called Sunyaev-Zel’dovich effect. (By the way, we ourselves do see a slight doppler effect on the CMB; this allows us to very precisely measure the solar system’s “peculiar velocity” relative to the Hubble flow of our neck of the universe.)

      So one certainly could declare that one galaxy’s perspective is to be privileged and regarded as more valid that others’, but the mathematics of the Hubble flow doesn’t need to assume anything about this one way or the other.

      • I am not declaring our “perspective” to be privileged. I am asking how does one exclude the possibility that we are in a special position in a real ontological sense.
        And assuming our special position, how does one goes about interpreting the experimental results of CMB and gravitational waves?

      • What do you mean by “special in an ontological sense”? Do you mean abnormal in some empirical but difficult to measure sense, such as that our region of the universe has a slightly different density or Hubble constant than the wider universe, or abnormal in some entirely non-empirical sense, analogous to the claim that there really is some absolute time even though the laws of physics take no account of it? (By the way, you’ll have noticed that in cosmology, there actually is a preferred time and preferred frame, given by the rest frame of the CMB.) In the second case, it’s not clear whether it’s relevant at all. The first case is tricky; it’s actually something that people have looked into to see how it affects the observed time evolution of the Hubble constant and anisotropies in the CMB. As I recall, the effects such specialness introduce are expected at large angular modes rather than the 1 degree scales that we’re talking about here.

    • If I am not mistaken, all measurements are taken from Earth. There exists no measurement that has been made from outside of the solar system.

      I wonder if cosmologists, intent in talk about the largest scales, forget this. Now I am observing universe from earth then how could I be sure I am not occupying a special place?
      Perhaps, the universe is expanding away from me, and not away from Andromeda galaxy?
      Are these concerns ruled out on some theoretical basis or ruled out on the basis that the universe could NOT have a special place.

  4. Pingback: Lightning Round – 2014/03/19 | Free Northerner

  5. I am reading Spengler’s Decline for the first time, so pardon me if this comment is of no interest. I see in this discussion a continuation of the dialogue between our Faustian, Gothic calling to explicate the prime symbol of infinite space, and our Magian, Arabian exploration of the magical world cave. These symbols seem linked in the concept of the time horizon, which rounds infinite space into a sort of cave.

    Spengler sees Faustian culture as beginning in the 11th or 12th centuries. But to me, St. Augustine seems very Faustian, with his enthusiasm, in the Confessions, for vastnesses of time and space, and in both the Confessions and the City of God, for detailed cosmic and mundane historicism, which are characteristically Faustian in Spengler’s typology.

    As a visitor to this site, it gives me pleasure to remind traditionalist Christians that it is our Faustian culture that has made the West and its imitators, for better or worse, and that we might be able to use an understanding of its true nature and of people’s misapplication of it to lead them back to the true path. Best wishes.

  6. National Geographic has an article about how gravitational waves suggest that we live in one universe among many in a multiverse. This seems to me like so much malarkey, not least because there’s no way to demonstrate or disprove the existence of other universes, but I am no cosmologist.

    Bonald, could you be so kind as to share your thoughts on the multiverse concept?

    • Well, this discovery has made me start to take such ideas more seriously. If inflation really happened, then the early universe did go through a phase transition, and breaking into distinct domains is a natural outcome. However, the claim about zillions of different values of physical constants being instantiated in different regions of space assumes more than I think we yet know about the low-energy landscape.

      It’s strange that it seems like only ten years ago or so that the string landscape was thought to be an embarrassment, and anthropic reasoning was strongly distrusted. Since then, and without any particular change in the arguments or the data (i.e. even before the BICEP results), it seems like everybody has been converted to the multiverse, and very few of us are still holding onto the gut feeling that there’s something wrong with it.

      An easier to grasp example of “multiverse” formation comes from ferromagnetism. Above a critical temperature, the substance (say, a block of iron) will not be ferromagnetic: the magnetic dipole moments of individual atoms will be pointing in random directions and average to zero. Below the critical temperature, the favored state is for atoms to align their magnetic moments, that is to all point in the same direction, because this is a lower energy state. So if you cool the iron below it’s critical temperature, the atoms start arranging themselves to point in the same direction as their neighbors. However, which direction they all decide on is completely random. In particular, we don’t expect different neighborhoods to all settle on the same choice, and in fact they don’t. The iron block divides into magnetic domains, with each one having its own magnetic field direction. This is the classic example of spontaneous symmetry breaking.

      The multiverse concept is kind of like this. (I dislike the name, by the way. These are not multiple universes. They’re multiple regions within one universe.) The first claim is that there are lots of equilibrium states for the fields that make up the universe (that is, lots of local energy minima with respect to the fields’ values) that pieces of the universe could fall into, and the configuration of our region is only one. This is not unlikely. The next claim is that these many energy minima correspond to universes with all sorts of different values for the mass of the Higgs, the strength of the cosmological constant, and other values that particle physicists have convinced themselves are “weird” in our universe. This claim is not crazy, but without a physical theory that applies outside the region of our own local energy minimum, I don’t see how we can say that this is even probable. The third claim is that the reason we observe weird values in this patch of space is because those are the values needed for intelligent observers. Thus, it’s no more weird that we see them than that we “just happen” to be one one of those rare planets with liquid water. This is the “anthropic argument”, and I suspect there’s something fishy about it, but I have trouble putting my finger on what.

      • Bonald,

        Extending your analogy with ferromagnetism further: magnetic domains will interact with one another, particularly with neighboring domains. According to the ‘multiverse’ theory, can the different regions also interact with one another?

    • Ian,

      In this model, there must be transition regions where some sort of interaction is possible. This is something that interests me too, although I don’t work in the field and have been surprised by the seeming lack of interest in this. On the other hand, in the standard inflationary theory, those regions where the inflaton field finds itself in just the right place in its potential undergo inflation and blow up exponentially fast, so that most of the interior (including, presumably, our neck of the universe) comes to be causally detached from anything outside.

  7. Pingback: Space-time ripples hint at physics beyond the big bangBig Online News | Big Online News


Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s