Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Perlmutter & Supernovae

  1. Oct 11, 2011 #1
    I've been an amateur quantum physicists for most of my life, and ever since 1998 I've been wondering about this issue, but I figured someone would address it. Now they've given the Nobel to the guy and I still don't understand something.

    Perlmutter says that galaxies are accelerating away from each other. He bases this on the fact that things that are farther away from us are moving faster than things which are closer. The problem is that we see things farther away from us as they were farther in the past. So quasars at the edge of the visible universe were traveling at .9c 13.7 billion years ago. Galaxies half as far away were travelling half that speed 7 billion years ago, or whatever the numbers are. Andromeda is actually moving closer to us and that is still 2.5 million years ago. The evidence seems to me to indicate, not that things are accelerating, but that they are slowing down. We have no idea what speed those distant galaxies are moving at right now. They could be moving closer to us by this point.

    Can anyone show me where I'm wrong on this?
     
  2. jcsd
  3. Oct 11, 2011 #2
    Perhaps the answer lies in relativity itself. The measurements seem to show that the farthest objects that can be seen in the universe are accelerating away from us. But, since motion is relative, it also means that we are also accelerating in our motion through spacetime. We cannot measure our own motion so easily as we can very distant objects, so we need to use indirect inferrence.

    By the way, it is said that we should regard the expansion rate of the universe as generating true kinematic motion, not just a "stretching" of spacetime. The idea here is that we shall avoid confusion this way. So, it is perfectly proper to speak of the expansion of the universe as if it gave rise to actual motion.
     
  4. Oct 11, 2011 #3

    Drakkith

    User Avatar

    Staff: Mentor

    Gork, all I can say is that everything supports our observations. Einsteins General Relativity actually PREDICTED this possibility before we ever observed it. I would recommend visiting wikipedia and hitting up its articles on expansion of the universe, hubbles law, and general relativity if you are actually interested in learning about this.
     
  5. Oct 11, 2011 #4

    marcus

    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Well your first statement is wrong. Perlmutter et al do not base the inferred accel. on that fact.

    That fact is contained already in Hubble law which we have had for over 60 years.

    Think about the standard statement of Hubble law v = HD. The standard formulation does not involve redshift directly because no simple relation between current expansion rate and observed redshift (only approximate for small distances.)
    D is a distance now---what you would measure if you could stop expansion at this moment and use some conventional means like radar.
    v is the rate that distance is expanding now.

    That law is implicit in the definition of H. It is a'/a as comes up in the Friedman equation.
    This is all really classical! Friedman equation goes back to 1920s! You've seen v = HD hundreds of times I expect.

    v = HD already says the rate of increase is proportional to length. The longer the distance D is, the faster it increases. That goes for any moment in time on the Friedman model clock (the universe time the cosmologist's model runs on.)

    Perlmutter et al result was that a' is increasing. The time derivative of the scalefactor a(t) is increasing.

    Your statement is incorrect because they did NOT base this conclusion on the fact (known for over 60 years) that longer distances grow faster.
    ===========================

    You might want to back up and ask what DID they base it on? What slight difference from what they expected? What slight adjustment in the model was required to fit the new observations? Various people might want to lend a hand in explaining.

    BTW the law v = HD is exactly true, in that form, in an ideally uniform universe that obeys General Rel. Uniform in the sense of homog. and isotr. It is not just an empirical law based on observation only. And the universe so far seems to be remarkably close to uniform at large scales. One reason that Hubble law is interesting. In its standard formulation, confirmed by observation, it is evidence for the correctness of Gen Rel.

    ======================

    Gork, just an afterthought. In the largescale pattern of distance expansion which we are witnessing, nobody gets anywhere. So it is significantly different from ordinary motion (in some local frame of reference) that we are familiar with--e.g. motion with a definable destination that is approached.

    So the language you use ("farther things moving faster") is a bit misleading. It will tend to confuse you if you use that figure of speech too much. It is better to simply say "longer distances increase faster". That is more neutral and more in keeping with the math picture of the universe cosmologists normally use.
     
    Last edited: Oct 11, 2011
  6. Oct 11, 2011 #5

    Drakkith

    User Avatar

    Staff: Mentor

    Gary I'm not sure this is correct. For one I don't think our own motion through spacetime can be said to be accelerating or decelerating.

    Unfortunately I don't have time to look up why at the moment as I have delicious tacos to go eat!!
     
  7. Oct 12, 2011 #6
    v=HD is still based on Hubble's redhsift observations, which still has the same fundamental problem as Perlmutter's, in that we only know the velocity and distance of galaxies, and that galaxies at longer distances were, in the past, moving faster than closer galaxies in the more recent past. This is, incidentally, exactly what the universe would look like if it expanded at near-light speed at the time the first galaxies were forming and slowed down over time. You can't say that just because v=HD is old that it must be right. You can't measure acceleration, a function of time, by taking a snapshot, particularly one that is 13 billion years old.
     
  8. Oct 12, 2011 #7

    Drakkith

    User Avatar

    Staff: Mentor

    Are you certain of that gork? If it were that easy then the current model would not have developed as it did.
     
  9. Oct 12, 2011 #8
    Well, you know that deferring to research isn't a valid argument. I've looked through every article I've found and nothing can explain what I'm talking about, so if anyone on this forum can, then I'd be glad to accept it as someone who knows more than I do explaining the issue, but so far people have just said "that's the way it is" or "Hubble Hubble Hubble".

    If I see two cars in front of me, one is going 10 mph and is 100 ft away and another is going 90 mph and is 900 ft away, and I take a photograph of them and write their speeds at the time of the photograph, does that photograph prove the farthest car is accelerating faster than the nearer car? See, in my example, velocity is a function of distance, but only by coincidence. The nearer car could easily be accelerating and the farther car breaking, but my instantaneous snapshot will not record that data.

    You would need to take data from many galaxies over a period of years indicating that the redshift was increasing over time, and that the amount of increase was proportional to their distance from us.

    Can anyone tell me how I'm wrong about this without simply invoking Hubble?
     
  10. Oct 12, 2011 #9

    phinds

    User Avatar
    Gold Member
    2016 Award

    But that is EXACTLY what has been done and it has nothing to do with Hubble. What you seem to not be taking into account is that the observed objects are NOT just farther away, they also represent different points in time, so we have a large SET of snapshots taken at lots of different times.
     
  11. Oct 12, 2011 #10

    George Jones

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I think gork wants to look at individual objects over a period of time in order to see the redshifts of individual objects increase. We can't do that yet. From

    http://arxiv.org/abs/0802.1532:
    Also, redshifts of individual objects don't necessarily increase with time. Figure 1 from the above paper plots redshift versus time. The three red curves are for objects in our universe. As we watch (over many years) a distant, high redshift object, A, we will see the object's redshift decrease, reach a minimum, and then increase. If we watch a much closer, lower redshift object, B, we see the object's redshift only increase.

    Roughly, when light left A, the universe was in a decelerating matter-dominated phase, and when light left B, the universe was in the accelerating dark energy-dominated phase.
     
  12. Oct 12, 2011 #11

    RUTA

    User Avatar
    Science Advisor

    If you go to the Supernova Cosmology Project website, you'll find actual data:

    http://supernova.lbl.gov/Union/figures/SCPUnion2_mu_vs_z.txt

    The data are redshifts and distance moduli for type Ia supernovae. Redshifts are relatively easy to measure, given our assumption that we know what the emission spectra look like is true. Distance moduli (called mu) for large redshifts (z) are what we find so valuable. For that you measure how bright an object appears (apparent magnitude, m) and assume you know its absolute magnitude M (how bright it would appear if you were only 10 parsecs away, i.e., the app mag at 10 pc). The trick is to know M for type Ia supernovae. That's what they figured out how to do (that number is at top of data, ~ -19). Anyway, if you know how bright the object would appear at 10 pc and you know how bright it appears now, then you can calculate how far away it is (although, you have to assume a particular cosmology model). Ok, so you plot m - M (mu) vs z and try to find a cosmology model to fit that data. The best fit cosmology model for that data is a spatially flat general relativity model filled with pressureless dust with a cosmological constant (Lambda) where the total energy density is divided 70% Lambda and 30% matter. According to that model, the universe started with a decelerating expansion rate when matter dominated and switched to an accelerated expansion rate once Lambda dominated. Thus, you read that dark energy (Lambda) is driving an accelarated expansion rate for the universe. Of course, that's not what they actually measured, i.e., they didn't measure Lambda nor did they measure an accelerating expansion rate, they measured mu vs z, Lambda and accelarated expansion are artifacts of the best fit cosmology model. There are attempts to fit the data without dark energy or accelarated expansion. One was just published last month in PRD, the reference was on Physics Forums, sorry I don't have it on this computer. Another is inhomogeneous spacetime. Still others are modifications of general relativity (f(R) gravity I believe is one such attempt). Hope this helps.
     
  13. Oct 12, 2011 #12

    marcus

    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Good idea! It's great to see the actual SNe data that had to go on. Basically you could say that model fitting, at the simplest level, just involves adjusting 3 numbers (today's matter fraction, cosm. const. fraction, and Hubble rate).

    And you are showing me hundreds of numbers. So I have to adjust 3 parameters to get the best fit. It is a good lesson.

    People don't realize how tightly constrained the whole thing is, since the model derives from the GR equation that has been checked in so many ways and passed so many tests with exquisite precision. As long as a GR model with only 3 adjustable parameters can fit data nicely people are not going to be too interested in alternatives.

    (Present company excepted I mean, if some guy were to offer a model with only 2 parameters, now, and it could fit the data as well! :biggrin:)
     
  14. Oct 12, 2011 #13

    RUTA

    User Avatar
    Science Advisor

    Yes, if you can beat this model (called LambdaCDM), then you might get some attention (depending on how you did it). There are really only two parameters in LCDM, since the matter and Lambda contributions must sum to the critical density (which makes the model spatially flat). So, you'd have to beat a 2-parameter model. And, there is data besides the supernova data that would have to be accounted for. I don't think there is any other model that can do all that. The alternatives I told you about all have some difficulty with fitting everything.
     
  15. Oct 12, 2011 #14

    marcus

    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    That's right. there is a lot of evidence that the U is SPATIALLY not curved or very nearly so. So if we accept that, then there are just two parameters. The current Hubble of course, 71, and the other two matter density .27 and Lambda density .73 forced to add up to one so really only 2 free parameters.

    http://supernova.lbl.gov/Union/figures/Union2_Hubble_slide.pdf

    So their data amounts to relation between two MEASURED quantities. Hundreds of datapoints. Each datapoint is a pair of numbers (z, mu) where z is observed redshift and mu encodes the distance

    Like, looking at the plot there, z = .2 corresponds to around mu=39.9 which by convention means about 2.6 billion lightyears ( "now" distance)

    And z = .4 looks like it corresponds to about mu=41.7
    which by convention is 5.1 billion lightyears.

    And the present LambdaCDM model fits that very nicely with the three (or two) parameters just mentioned adjusted as we said: .27, .73, 71.

    The nice thing from my very personal perspective is that gauging the luminosity and allowing for the observed wavelength stretch gives you by the inverse square law just the distance you want! Namely the freezeframe proper distance NOW which you would measure if you could stop the expansion process at this moment and use any conventional means. It does not give some confusing pseudodistance like light travel time. It gives the actual present day distance distance. And we are DIRECTLY MEASURING. I love it. Others may disagree :biggrin:

    proper distance now = 32.6*10^(mu/5)/(1+z) lightyears.

    Great to see all that direct measurment data, RUTA!

    This gets nailed to a virtual wall somewhere:
    http://supernova.lbl.gov/Union/figures/SCPUnion2_mu_vs_z.txt

    They say their presentation is dated 2010, but it looks rather similar to some of the figures and tables in this 2008 preprint, so I offer that to provide discussion if anyone wants:
    http://arxiv.org/abs/0804.4142
     
    Last edited: Oct 12, 2011
  16. Oct 12, 2011 #15

    Ken G

    User Avatar
    Gold Member

    This is all well and good, but I think gork is actually asking a much simpler question. He is asking why the Hubble law is not interpreted as deceleration, period-- that question really doesn't have anything to do with acceleration or Lambda CDM! It is true that the longer ago we see some object, the faster it is moving away from us. Why doesn't that mean the expansion is decelerating?

    The answer is called "the cosmological principle." This principle is the key unifying and simplifying factor behind all cosmological models-- it is the thing that allows us to take individual "snapshot" of different eras of our past, at different distances from us, and cobble them together into a global description of the history of our universe. The cosmological principle asserts that the universe is everywhere the same, except that it changes with time-- it ages, everywhere the same.

    So when we look at galaxies at different redshifts, two things are happening-- we are seeing different times in the past, and we are seeing different distances away. gork is focusing only on the former issue-- if everything in the universe was moving away from us at thesamespeedateach age (myspacebarisn'tworking), thenhe'dberight--butthatwouldn'tobeythecosmologicalprinciplebecausethenwe'dbe inaspecialplace.
     
  17. Oct 12, 2011 #16

    marcus

    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    KenG, here is Gork's question in post #1. I replied to it and explained where he was wrong. This was what he was asking for. But he did not get back to me
    In post #4 I replied and pointed out that the Perlmutter et al result was NOT based simply on larger distances increase at (proportionally) larger rates. That by itself does NOT imply acceleration.
    We've known that for generations. Gork's first mistake was to make a kind of straw man of Perlmutter et al and suggest that they claimed acceleration based on what would be incompetent reasoning which everybody knows does not imply it.

    Gork has not yet acknowledged the error in #1. So the discussion does not go anywhere.

    My advice to Gork would be to make a fresh start and say: "OK I was wrong. Perlmutter et al result was NOT based on longer distances growing faster. That by itself does NOT imply accel and nobody claimed it did. Now how did Perlmutter et al deduce that the parameter Lambda in the model is positive?"

    I suggest he ask a real question like: "How DID they figure that Lambda > 0?"

    Note that acceleration is a circumstance that goes along with positive Lambda as one of the ways Lambda manifests itself. The public does not know or care about Lambda. But the public can understand acceleration, so acceleration was the exciting way they publicized the finding. It communicates. And they also got some attention when somebody thought to call Lambda "dark energy". That is mainly in the realm of PR. Lambda is a curvature and if you drag it over to the other side of the equation (not a natural place for it IMHO) it acquires units of energy. Any constant curvature on the LHS of the Einstein equation would do likewise.

    What we are really talking about is an important constant of nature, like Newton G or Planck h. It is the other constant in the GR equation---our prevailing well-tested law of gravity. We are at the next stage of refining today's law of gravity.

    It does not help anybody to insist on staying down in the level of PR and pop-sci.
     
    Last edited: Oct 13, 2011
  18. Oct 13, 2011 #17

    Ken G

    User Avatar
    Gold Member

    Yes, and you were quite correct. But it just means that gork did not present his question clearly. His subsequent posts show that his real issue was not with the acceleration, nor the Hubble law, but the fact that deceleration is not inferred from the Hubble law. My point is that if we did not have the cosmological principle, but instead a model where the Earth was at the center and everything in the universe moved at the same time-varying speed away from us, then we would indeed conclude that the universe was decelerating. All we need to not conclude that is the cosmological principle, which immediately creates a different interpretation of the Hubble law (along the lines of what you explained quite correctly but I don't think gork would get the punchline without the cosmological principle).

    I agree, but first we must establish the more basic aspect of the question, which has nothing to do with Perlmutter. It's crucial that gork understand why the Hubble law is not interpreted the way he imagines. I'm not saying anything in your answer was wrong, only that I think the question is at that more basic level.
    That's all true, I just think gork's incredulity about acceleration stems from a more basic issue about what the Hubble law means. There is a situation where his interpretation could be perfectly correct, so we need to look at why that situation is not the one that gets used-- it's because it wouldn't satisfy the cosmological principle, it would put the Earth at a very special place in the universe, and wouldn't obey GR with or without Lambda. But everything you said is certainly both correct and useful for understanding the modern state of affairs, I'm not criticizing.
     
  19. Oct 13, 2011 #18
    George and RUTA, thank you very much for your explanations. I understand now.

    Marcus, quit trying to get an apology from me. You never explained anything, just told me I was wrong because someone smarter than me already figured it out. You weren't helpful at all, so stop pretending you were. I am smart enough to understand it once it was explained. You never tried. So I won't say I was wrong, because I wasn't. I was asking a question that you didn't answer multiple times. George and RUTA answered it thoroughly and were very helpful. And without being pedantic. You could learn a lot other than physics here.
     
  20. Oct 13, 2011 #19

    Ken G

    User Avatar
    Gold Member

    Still, I wonder if you understand now, gork, that none of the things you find in those detailed explanations of how the redshift observations are interpreted would be possible without the cosmological principle, and why it is the cosmological principle, not those details, that answer why a Hubble law does not mean the expansion is decelerating. That's the only issue here that isn't "black boxy", that isn't "this is what the people get when they make detailed models", which isn't all that satisfying by itself. You also have to bear in mind that marcus was probably reacting to all the people who come on here and say "I think cosmologists are wrong because they overlooked some really basic point", which is not what you did (you framed it as "what is wrong with this reasoning", and the answer is, "it doesn't satisfy the cosmological principle"). Also, bear in mind that everything he said was a correct explanation, and that it took him time to lay out that correct explanation, so you should still thank him even if it rubbed you the wrong way that he wanted you to recognize the misconception behind your question. Often, we find it is important to "kill the misconception" in order to achieve real learning, but that can sometimes come out sounding like "kill the questioner", leading to "kill the answerer."
     
  21. Oct 13, 2011 #20
    I should have explained my question more clearly, but I didn't realize that people still thought of Earth as being the center of the Universe. I didn't know the specific jargon, but I already realized that the Earth is moving away from everything else the same way it is moving away from us. I like the example of drawing points on the surface of a balloon and then inflating the balloon, but then imagining the surface of the balloon is 3-dimensional. But yes, I recognize that to an observer in a galaxy redshifted from our point of view the Milky Way would be equally redshifted. That was never the issue. The issue was mostly that I didn't recognize that Perlmutter was taking that snapshot I was talking about and figuring out that the Universe is bigger than it would've been had the Universe not accelerated. That's something that none of the articles I read on the topic actually mentioned.

    And I understand that most people who post questions like this on here might not understand that answer were it given to them. I actually started studying astrophysics when I was 12, but I went to college to be an art teacher, and kept physics as a hobby that I take pretty seriously.

    Thanks again for the help!
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Perlmutter & Supernovae
  1. Supernova in m101 (Replies: 19)

  2. Supernova energy (Replies: 2)

  3. Supernova shockwave (Replies: 6)

  4. Energy in supernovae (Replies: 2)

  5. Supernova in M82 (Replies: 18)

Loading...