The Shaw Prize in astrophysics has been awarded to Saul Perlmutter, Adam Riess, and Brian Schmidt, for discovering the acceleration of the universe by measuring the Hubble diagram using Type Ia supernovae. The Shaw Prize is relatively new, having first been given in 2004, and is awarded in three areas: Astronomy, Mathematical Sciences, and Life Sciences and Medicine. It comes with a total of US$1 million, split between the three recipients. Competitive with, although not quite as much as, the Nobel prize…
Brian Schmidt and Saul Perlmutter come to blows over whose universe is accelerating faster
Brian was the leader of the High-Z Supernova Search Team and Adam was lead author on their paper; Saul was the leader of the Supernova Cosmology Project and also lead author on their paper. (Get some more inside scoop from Rob Knop.) To most of us, their finding was a complete surprise, as we were all quite familiar with the fine-tuning problems associated with the cosmological constant (the most straightforward explanation for the acceleration). But by 1998, it had become impossible to deny that something fishy was going on — the universe was not the simple matter-dominated flat Einstein-de Sitter cosmology of the standard Cold Dark Matter model. In late 1997 I was asked to give a review talk at a CMB conference in Santa Barbara, on the topic of “every way to measure the cosmological parameters other than the CMB.” In assembling the talk the overall message came through loud and clear, from considerations of the age of the universe, direct measures of the mass density, and properties of large-scale structure. There were plenty of ideas floating in the air, including an open universe (the most obvious choice), warm dark matter, a mix of hot and cold dark matter, or some dramatic features in the primordial power spectrum, as well as the old standby cosmological constant. But only the last of these solved all of the problems with one fell swoop. So when the two supernova groups announced in 1998 that they had direct evidence that there really was a cosmological constant, in the form of an accelerating universe, the community was primed to believe them, which they did fairly quickly. Soon thereafter, of course, improved measurements of the CMB anisotropies indicated that the universe was spatially flat, in perfect accord with the combined supernova and matter-density measurements. If you were to plot the inferred density of both matter and cosmological constant, the constraints from the three different techniques — supernovae, matter dynamics (clusters or large-scale structure) and the CMB — the allowed regions overlapped in perfect harmony.
A preposterous universe, maybe, but I like it.
It does appear preposterous. Why should someone like it?
Reasons to like it:
(1) There are major things in it that we don’t know. Physicists always love it when there are big and important and interesting things left to find out. Even the nature of regular dark matter is a mystery, although it’s a mystery we have hopes of solving (or making great strides towards solving) experimentally in the next several years. But Dark Energy : who ordered that? (OK, bad reference, becuase as Sean says, it wasn’t completely out of the blue the way the Muon was; indeed, the seminal review article all about Dark Energy (before it had that name) was a Carroll & somebody ARA&A paper in 1992, six years before the accelerating Universe discovery).
(2) Because it all hangs together. Cosmology got a bad name among astronomers because until the mid-90’s, it was chock full of internal contradictions. Different teams would measure the Hubble constant (the current expansion rate of the Universe), and report results with 10% error bars… but get answers that were different by a factor of two. Cosmological estimates for the age of the Universe including everything that was reasonable, either with a theoretical bias (Universe is flat) or an observational bias (Universe is open, with just the dark matter we’ve seen from galaxy and cluster dynamics) were several billion years LESS than the age estimates for the oldest star clusters in our galaxy. Oops.
Now, with the acceleration, the Universe is comfortably older than the oldest cluters… but only by a little bit, so those clusters are indeed plausibly formed in the early stages of the assembly of galaxies.
What’s more, there are multiple different observations that all contribute to the same picture and ALL are consistent, even beyond the three that all intersect in Sean’s plot above.
Even though I know some older or more traditional astronomers who would still stick to the bias that “cosmology is the worst sort of astronomy” (a quote I received during my job interview at Vanderbilt a mere 5 years ago), and that cosmology is borderline pseudoscience, cosmology now has an overall theory and picture of what’s going on that’s approaching the consistency and sensibility of Biology’s evolution.
(3) Because it’s just cool. The Universe’s expansion is ACCELERATING. The Universe is filled with something truly bizarre that has NEGATIVE PRESSURE. What’s not to like? Dude, I like Tom Stoppard plays too.
-Rob
One more thing : the Universe that we came up with not only satisfied the pre-existing expectations of the theorists (the Unvierse has a total energy density very close to the critical density), but it also matched the pre-existing observations (the total MATTER density of the Universe is only something like 1/3 the critical density).
How often can you find a solution to two seemingly irreconcilable views that fits with what everybody needs to make all of their other things work?
Even though I know some older or more traditional astronomers who would still stick to the bias that “cosmology is the worst sort of astronomy” (a quote I received during my job interview at Vanderbilt a mere 5 years ago), and that cosmology is borderline pseudoscience, cosmology now has an overall theory and picture of what’s going on that’s approaching the consistency and sensibility of Biology’s evolution.
At DAMOP back in May, Eric Cornell joked during his Nobel Symposium talk that cosmology has gotten so good that “now we call it physics, not astronomy.”
Could someone explain what Omega_Lambda and Omega_M (the axes on the figure) are? I was quite enjoying the cogent explaination of the result up to that point…
Ralph, those are the “inferred density of matter and cosmological constant” referred to in the post. “M” for matter, “Lambda” is the cosmological constant. The density is normalized so that a total of 1 corresponds to a spatially-flat universe; greater than one would be positively curved, and less than one would be negatively curved. Omega — the sum of Omega_M and Omega_Lambda — is just the density in units of the density a flat universe would have. (Note that it is, indeed, about one in the real universe.)
Omega_Lambda is the ratio of the energy density contained in the cosmological constant to the critical energy density of the universe. Omega_M is the ratio of the energy density contained in matter (baryonic and dark) to the critical energy density of the universe. (The critical energy density is that which would yield a spatially flat universe)
If the universe is spatially flat (as it seems to be), and GR is correct, then these are just the fractions of the total energy budget of the universe in these two components respectively. i.e. Omega_Lambda =0.7 would mean that 70% of the energy density of the universe, on cosmological scales, is provided by the cosmological constant.
While the SN teams deserve major recognition, one still must ask:
1) Is it a complete theory if it relies on inferred energies? What is known about dark energy? Is it wave or particle, constant or changing? Why is it 71%, not more or less? All we have are a divergence of possible solutions to keep cosmologists guessing. Please state the overriding principle behind this “cosmology”.
2) Does it really hang together? CMB data allows a variety of cosmologies with total density Omega = 1. In Sean’s chart, the centroid of SN data does not appear to coincide with the centroid of CMB data. Supernova data tells us very little about Omega-matter, and the CMB tells us nothing about acceleration.
3) Cool is cool, but the evidence of an accelerating Universe, whether from SN or LSS, relies ENTIRELY on redshifts. As someone assured me, “A varying speed of light might reproduce the high redshift supernova diagram.” An oversight in redshift measurements could be very negative.
4) While we are all converging on a density of Omega =1, does completing the pie diagram with inferred energies suggest ether or epicycles? While more SN data is needed, it is premature to proclaim a complete picture of the Universe. Please state why Omega = 1 and not something else.
I’m not a physicist, but a programmer who likes to read physics stuff in order to get confused. (It’s one recreational activity that requires neither money nor illegal substances)
So this may be a slightly dumb question.
known: Particles that have any mass at all, when emitted at relativistic velocities, increase in mass according to Einstien’s equations. The increase can be substantial when the particle is travelling very close to light speed.
Does the increased mass of all the radiation of all the stars & evaporating black holes in all the galaxies represent a significant amount?
If so, is it taken into account when searching for dark/missing mass?
PS: about how significant is it?
Sean, Mark: thanks for the explanation.
At DAMOP back in May, Eric Cornell joked during his Nobel Symposium talk that cosmology has gotten so good that “now we call it physics, not astronomy.
Yes, that’s why many astronomers hate cosmolgy now. And, I can see how one would get grouchy when a whole bunch of particle physicists who know how to and are used to working in big groups swoop in and grab large amounts of astronomy funding to do things like LSST and such that are based strongly in cosmology….
I was at a consortium board meeting for a group of universities that run a bunch of small (1m-class) research-grade telescopes. At one point, somebody made the comment that we’re all small telescope users, so we’re all “nicer” — the actual quote was, “there are no asshole cosmologists here.”
Of course, everybody can be an elitist, even if you think you’re being anti-elitist in your elitism….
The one most fundamental thing I’ve learned from academia is that “anybody not working on what I am working on is wasting their time on an unimportant and uninteresting subject.” This is perhaps the one thing that everybody in a Physics and Astronomy department could agree upon.
-Rob
“A varying speed of light might reproduce the high redshift supernova diagram.”
If you varied the speed of light, you’d also be varying other things like the structure of atom, etc., and galaxies out at a redshift of z=0.5 (where much of the statistical weight of the original accelerating Universe lies) would look very different from the nearby ones… yet they don’t.
You can always pick some parameter and tweak it to get the answer you want on one plot, but when it’s something as fundamental as the speed of light, you’re almost certainly going to screw up a whole host of other observations.
In other words, what your friend assured you, while technically correct, is very wrong– yes, you could probably reproduce that plot, but not in the context of the rest of the world.
Is it a complete theory if it relies on inferred energies?
Is Darwinian-style evolution a complete theory if it relies on biogenesis outside the theory before it can do evolution of species? Creationists always tell us “no”, but they’re wrong.
The term “complete” is always a scary one. The fact is that Einstein’s equations from GR are completely consistent with matter that has dark energy like properties. There’s no real tooth fairy here, just stuff out there that we don’t know a lot about and haven’t found except through its affect on the Universe as a whole. Dark matter is similar.
In Sean’s chart, the centroid of SN data does not appear to coincide with the centroid of CMB data.
Irrelevant, because the size of the areas matter just as much as their centers. The areas are “confidence intervals”; the inner area is “we are 68% sure that the answer is somewhere in here”, and the outer area is “we are 95% sure that the answer is in here”. (I think; I don’t remember the actual number for the outer answer.) The actual centroid of the supernova data isn’t really significant, because the (effective) error bars are big enough that the concordance point is within one error bar. Within the uncertainties, everything is remarkably consistent.
Please state why Omega = 1 and not something else.
Because of the green bits on the plot above.
None of the three areas on the plot assume Omega=1, but allow Omega_M and Omega_Lambda to vary independently. The results of the CMB data indicate, however, that Omega_total is very close to 1, which is why the pie diagram that Sean (and others) show use that value.
-Rob Knop
Particles that have any mass at all, when emitted at relativistic velocities, increase in mass according to Einstien’s equations.
Eh, not really. That’s sort of a retro way of looking at it.
Momentum increases as velocity increases, but when velocity becomes an appreciable fraction of the speed of light, a similar “momentum equals mass times velocity” relationship no longer applies. One way you can deal with this is to keep the linear relationship, but say that mass increases for something moving faster. Anther way is to just use “rest mass” for mass, and say that the momentum equation is nonlinear (momentum starts to go up faster than the linear relationship once v is an appreciable fraction of c).
In some sense, it’s a semantic argument, but most of the time now when a physicist says “mass”, he just means “rest mass”.
To answer the other part of your question: all of this is based on GR, which is General Relativity, so it naturally subsumes all relativistic effects. As for what happens to stuff flowing into and Hawingly radiated out of Black Holes, all of that is in the “fine grain” that doesn’t matter any more than individual stars or galaxies on the very largest scales that we’re describing when we talk about the expansion of the Universe. (An analogy would be considering the heights of buildings when talking about the shape of the Earth; the differences are so small compared to Earth’s radius that we can safely neglect them and still get very good answers.)
-Rob
The one most fundamental thing I’ve learned from academia is that “anybody not working on what I am working on is wasting their time on an unimportant and uninteresting subject.” This is perhaps the one thing that everybody in a Physics and Astronomy department could agree upon.
I don’t agree.
I don’t agree.
See if I vote for you when you’re up for tenure!!
“The one most fundamental thing I’ve learned from academia is that “anybody not working on what I am working on is wasting their time on an unimportant and uninteresting subject.””
The one most fundamental thing *I’ve* learned is that a great many people have massive inferiority complexes, and that this can express itself in a variety of grotesque ways which have one thing in common: they make the sufferer appear very much stupider than he [probably] really is. See eg Phil Anderson’s amazingly dim-witted comments on how cosmology would work if he had been present at the Creation, or R. Laughlin’s astounding book, which takes vicarious embarrassment to levels hitherto undreamed-of.
Obligatory addendum: from a particle physicist’s viewpoint, the dark energy is very ugly, because it happens on an energy scale which is much, much smaller than all effects that produce vacuum energy in particle physics models that seem to have something to do with reality. Enormous amounts of time and energy have been wasted so far in attempting to put together models which produce anything like the right answer.
Also: These nice stripes on the graph rely on a number of assumptions, some of which are currently not well testable – for example the nature of the dark matter that makes up Omega_M, where most standard assumptions do not work very well for small scale structures. Or an assumption that the local Universe is sufficiently homogeneous to extrapolate out to large distances. What if the acceleration is only happening in our cosmic back yard, as it were, due to a local blip in the distribution of matter – in which case it doesn’t tell us about the underlying background evolution.
The current picture looks self-consistent, but that doesn’t mean there are no serious discrepancies left in cosmology. There is a chance that it could be self-consistently wrong, and each of the assumptions has to be further tested.
Oh, and ‘varying speed of light’? Don’t get me started. The speed of light in SI units is *defined* to be a constant number, and you can always make a choice of units so that this remains so. (Relativity looks best in units where c=1.) There is no physical content in ‘varying’ speed of light, only in dimensionless constants like the fine structure constant.
Rob Knop –
Anther way is to just use “rest mass” for mass, and say that the momentum equation is nonlinear (momentum starts to go up faster than the linear relationship once v is an appreciable fraction of c).
Thank you. This explanation is clearer (and more complete) than I’ve read elsewhere, and makes my admittedly amateur understanding of the subject much better. That is, it makes sense.
Having worked with radar, laser ring gyros, and instrumentation in aircraft, “nonlinear” is definitely a relationship that I comprehend.
It is nice to be able to see how GR works on such a cosmological scale. 🙂
“Energy density” then of course throws in some questions for me, about the dynamical nature of our universe.
Are the “quantum perceptions” ready then, to accept the dynamical nature QM and GR in our cosmos as having now shown themself together?
Interesting ideas. If c changes but the quantity hc remains constant, then h is increasing and GR-QM are linked. The nice stripes on the graph do seem to rely on many assumptions.
If Earth is “defined” as the centre of the Universe, you can always make a choice of reference frames so that this remains so.
Just as anyone can observe retrograde motion of planets, the CMB indicates that the Universe expanded faster than the present value of c. One must infer a repulsive inflaton field (Epicycle #1). Supernova and other data indicate that relationship v/c is accelerating, so one must infer another repulsive dark energy (Epicycle #2). We would be fortunate to live in the centre of these forces.
Varying c indeed changes observations of galaxies at z=0.5, it affects their luminosity and redshift so that they fit the curve precisely. If a prediction fits the graph of many data points without inferred energies, what does it mean?
We would still need corroborating evidence for a “c change.” More coming soon.
http://www.lns.cornell.edu/spr/2005-06/msg0069818.html
Isn’t there a paper out using Gamma Ray Bursts (GRBs) instead of SNs to do the same sort of thing? The GRB data is public, even if there aren’t alot of them, more data comes in more or less daily. GRBs are more energetic, so if this kind of data can come from them, then a deeper look back into time and space can be achieved.
Stephen, yes, we talked about it here:
http://blogs.discovermagazine.com/cosmicvariance/2006/01/11/evolving-dark-energy/
GRB’s are very bright, but much harder to “standardize” than are Type Ia supernovae.
Well of all things I came to ressurect, and it had already been done with the point of GRB’s. 🙂
More on name. I am rushed this morning so I had to do this quickly. Layman thoughts forming.
WMAP distribution? Can it cause some problems in regards to the “information from measure” that are being recieved in regards to how we think of the universe’s shape?
Okay, Yet Another Programmer With Too Much Time on My Hands.
Does it not follow that if the farther away an object is, the faster it appears to move…that, as an object gets farther away, it appears to move faster?
In effect, there is an apparent accelaration due to distance? Even my limited calculus derives that for an observer on the circumference of a simple circle with an expanding radius. I’m not sure of where the variables fit in such a simple model, for if the expanding radius were “time-like”, and the expanding circumference was “space-like” we would have an explanation for the passage of time that we all observe. If the “world line” a particle follows was defined by something equally simple like perhaps a spiral, it would also account for time dliation.
Just curious…