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

Acellerating expansion? Bah!

  1. Nov 17, 2009 #1
    I'm sure it's just me since I've never seen this opinion anywhere, but has it occurred to anyone else that when type 1a supernovæ are dimmer than you expect, the proper response is "Gee, let's find out why the type 1a supernovæ are dimmer than we expect", rather than "Let's explain it by inventing a mysterious invisible entity about which we know nothing"? They've been doing THAT for thousands of years.

    Dark energy is more of a problem than what it (supposedly) explains. It's like saying "the big bang was caused by an invisible god". Not only do we then have the far larger problem of explaining where the invisible"god" came from, but far worse, that answer doesn't actually tell us how the big bang happened, either.

    Same with dark energy. I predict that when the source of the 1a attenuation is found, "dark energy" will be next to "Eoanthropus Dawsoni" in the footnotes of science history.

    --faye kane homeless brain

    Don't EVEN get me started on "dark matter"!
  2. jcsd
  3. Nov 17, 2009 #2


    User Avatar
    Science Advisor

    There are physicists looking for other possible explanations for both dark matter and dark energy. So far no one has come up with viable alternatives, although MOND has some followers as an alternative to dark matter.

    Physical theories (in contrast to religion or economics) are not dogma. All theories must be consistent with observed data or else they are discarded.
  4. Nov 17, 2009 #3


    User Avatar
    Gold Member

    Faye, as you may or may not know, there is a large body of peer-reviewed, published research showing that low-redshift bodies are shown to be physically associated with bodies of higher redshifts. You might want to Google on Arp, the Burbidges (both Geoffery and Margaret), Jayant Narlikar and others. Their work (and research by myself and my collaborators, drawn from publicly-available databases, and published after peer-review) suggests that there may be a component to redshift that is intrinsic. In other words, cosmological expansion, peculiar motion, and gravitational redshift may not be entirely responsible for the total redshift of an astronomical object.

    What is the significance of this concept? Lets consider: in which objects are the redshifts most dominated by the Hubble/distance relationship and the least contaminated by any other possible contributor? They are the objects that are the farthest away from us. In light of this, SN1ae data should have been calibrated on the most distant galaxies, since they are the sample that is least contaminated with peculiar motions and other contributions to total redshift.

    How does this pertain to the dark energy problem? If the SN1ae in distant galaxies are taken as our calibration standards, then the question should not be "Why are distant SN1ae too faint?" but "Why are nearby SN1ae too bright?". The simplest answer is that we have over-estimated the distances to nearby galaxies. If Arp, the Burbidges, Narlikar, and others are right, their oft-derided view of intrinsic redshift can neatly explain why nearby SN1ae are brighter than more distant ones. In this view, we have overestimated the distance to nearby galaxies because we have no way to separate the effects of intrinsic redshift from the total redshift of a galaxy. This is not a popular concept, but it is a realistic application of observational astronomy as a test of cosmology.

    In cosmology, we know little and assume much. There are times when cosmology can be tested by observation. I believe that this is such a time, and the opportunity has been largely overlooked.
    Last edited: Nov 17, 2009
  5. Nov 18, 2009 #4
    I'm not understanding this. The important thing about SN1a data is that there is very little calibration necessary. As far as we can tell, all SN1a have the same intrinsic brightness and are standard candles.

    But unless I'm wrong the calibrations for the SNIa bypass redshift all together. The galaxies in which we see nearby SNIa also have cepheid data, and so redshifts are irrelevant, since they aren't included in the calibration at all.

    The reason that supernova Ia are important is that you considerably reduce the assumptions. The one big assumption (which people are testing both theoretically and observationally) is that supernova Ia don't evolve in brightness. However, in the original paper IIRC they mentioned that you had to have very large amounts of differences in order remove dark energy.
  6. Nov 18, 2009 #5
    Take a look at Section 5.1 in the original paper.....


    The basic answer if you try to explain this as an evolution effect, you have to explain why 1) there seems to be no evolution effect in computer simulations of type Ia 2) why there doesn't seem to be any correlation between Sn Ia brightness and whether the galaxy is elliptical or spiral and 3) why you have to explain why it's not only dimmer but dimmer in such a way that the whole light curve shifts down.

    And you are doing this anyone by invoking some mysterious invisible entity that makes SnIa's act differently.

    Sure. Come up with a better explanation?

    It's not that bad because you can put in the amount of dark energy, and get out predictions for other things that seems to make sense. It's the difference between saying that the big bang was caused by an invisible god, and the big bang was caused by an invisible God whose name was Fred and was bigger than a breadbox but smaller than an SUV and likes to eat brocolli and snails, and we know his name is Fred because he signed his name in the Andromeda galaxy.

    We are at the point that if dark energy does exist, that we can say thing about it.

    And you know this how? The original paper went through all of the obvious sources of attenuation and gave reasons why it couldn't be

    1) Evolution
    2) Extinction
    3) Selection bias
    4) Effect of local voids
    5) Weak graviational lensing
    6) Light curve filtering method
    7) Sample contamination

    Now what type of bias are you suggesting? Saying that "dark energy" is causing the universe to expand is much less mysterious than talking about some "unknown supernova Ia attentuation factor" because "dark energy" is falsifible, whereas some weird SNIa attentution field isn't until you give some more details.
    Last edited: Nov 18, 2009
  7. Nov 18, 2009 #6


    User Avatar
    Science Advisor

    Yes. But the problem is that the brightness of supernovae is set by the physics of how the stars collapse, so there isn't a whole lot of room for there to be a systematic change in brightness with redshift.

    No, it's actually not even a close comparison. You see, the various dark energy proposals are specific proposals with specific observational predictions. This is vastly different from proposing a nebulous entity whose definition is so vague that you can't ever possibly test for it.

    Anyway, the thing is that by now, we don't even need to use supernovae at all to measure the accelerated expansion. There are a number of other observations that show the same thing. Perhaps the most interesting of which is the baryon acoustic oscillation measurement: in the early universe normal matter (which cosmologists call baryons) was in a plasma state, and sound waves through this plasma propagated throughout our universe. We see the imprint of these sound waves on the CMB. These sound waves have a characteristic size. We can then compare the average distances between nearby galaxies, and correlate it with this length scale set by the CMB. This gives us a "standard ruler". And when we use this "standard ruler" to measure the expansion rate with time, we get the same accelerated expansion rate (to within the errors) as with supernovae.
  8. Nov 18, 2009 #7
    There are several factors that go into Chandaraska's equation that determines the brightness of a 1a SN - M and G may have behaved differently in early epochs - even though the MG product may be temporally constant (as is established by experiment) - G may have been larger and inertia may have been less - consequently earlier SN would be dimmer and the width of the luminosity profile would have been narrower when the brightness is calibrated against closer SN. Nonetheless, it is much more likely that the acceleration interpretation is correct.
  9. Nov 18, 2009 #8
    Last edited by a moderator: Apr 24, 2017
  10. Nov 18, 2009 #9


    User Avatar
    Science Advisor

    Sorry, but that thread's topic is quite different from this thread's. While they are similar, one is addressing how we talk about the expansion. This one, on the other hand, is addressing how, if we agree on how we want to talk about the expansion, said expansion is progressing.
    Last edited by a moderator: Apr 24, 2017
  11. Nov 18, 2009 #10
    Won't work. If you change G and M enough to change the brightness of supernova Ia, then you change the brightness of all the other stars enough so that people will notice. Now if you can come up with a clever explanation of why a change in G and M would affect *only* SN Ia and nothing else (or maybe it's affecting something that we haven't measured) well...... Let's see the explanation first....

    I should point out that this is one problem with popular accounts of science, in that you end up with people thinking "scientists are stupid because they didn't think of this obvious explanation." What invariably happens is that people actually *do* think of obvious explanations and there are reasons why the obvious explanation won't work.
  12. Nov 18, 2009 #11


    User Avatar
    Science Advisor

    This is one of the things that really struck home for me when I was a student at the university. I found, very quickly, that when tackling a currently-unknown problem, the scientific community as a whole is just extraordinarily resourceful at finding possible alternative explanations. Some of them are obvious. Some are extremely creative and insightful. But in any case the sheer scope and variety I found to be really impressive.

    Of course, as with many people interested in this stuff, I also had my own ideas of possible alternative explanations. What I invariably found was that either my own alternative explanation was just flatly and obviously wrong (this was the case most of the time), or somebody else had already thought of it and had developed the idea far, far more than I ever did. Now that I am a trained scientist, I would still say that 99% of my creative ideas turn out to be flat wrong (which I usually realize within a few seconds of making the idea).
  13. Nov 18, 2009 #12


    User Avatar
    Science Advisor

    I know exactly what you mean Chalnoth! One of the tricky things in being a scientist is to get the right balance between self confidence and humility. On the one hand you can place too much confidence in your ideas without carefully checking the literature to see who has already thought of them and who has shown why they don't work. On the other hand you can get too used to assuming your idea won't work for some reason or mustn't be novel, which can hold you back. I've seen that in myself and colleagues at times. The balance is really quite tricky!

    I spent about 6 months when I was a PhD student trying to work out why a result I was getting was wrong, since it contradicted a previous work by a very well known and successful researcher. I was originally hoping to extend this previous work, and re-producing it was going to be the first step. Eventually it dawned on me that just maybe I was actually right, and fortunately the 6 months I spent working out why I was wrong turned into an exhaustive proof of my result! After 2 years and several journal articles back and forth between me and my supervisors and the other group we've finally agreed that I was right all along, very gratifying!

    That being said, most of my ideas do turn out to be wrong, the hard part is being able to both be confident when you really think you're onto something and also honest enough to relent when the evidence turns against you. That's the lesson Halton Arp never learnt (although he's not alone there!).
  14. Nov 18, 2009 #13
    At which point you think some more. When I tell someone "changing G won't work because you can't obvious change the magnitude of SN Ia without changing other things." That's a challenge. If you think for a while and come up with a way so that you *can* change SN Ia by changing G without changing anything else, GREAT!!!!! That's worth an letter to Ap. J.

    If you think that you can explain dark matter by SN Ia differences, then you should work at it. Here are all of the papers on SN Ia, read them for a few weeks. If you can find a gap in logic or you can think of something original then GREAT!!!! Again, a letter to Ap.J. If you can mention how to fix the gap, then this is progress. Don't just say "I think there might be something weird about SN Ia that may cause distances to be off". Use your imagination and knowledge to figure out what it might be, and then come up with some specific observations that can address the weirdness, write a paper, and the upload it to the Los Alamos server.

    OK, intrinsic redshifts might throw off the accelerating universe. Quantify the effect. Come up with experiments that would show what that effect is.

    Yes, it's hard, but you'll be doing science.

    The neat thing is that when the web, the sum total of what has been tried before is now online, so it becomes possible for people to do this sort of thinking without needing access to a university library.

    That's why it's important to come up with 100 ideas. Also from a theorist point of view, the goal is not to be right. It's up to Mother Nature to tell you whether you are right or not. The goal is to be *original* and *interesting*. If you come up with an new idea then most of them die pretty quickly. Occasionally, you'll come up with a new approach or a new way of looking at things, which is publishable because it adds to the conversation, and you get progress even if you end up being wrong.

    As a theorist, you can't avoid being dead wrong, since it's nature that decides right and wrong in science. You can at least be wrong in interesting and productive ways.
  15. Nov 18, 2009 #14


    User Avatar
    Science Advisor

    Well, right, and this has been tried. I think the general consensus in the supernova community right now is that we actually know that there are some biases here, such that the measured expansion rate isn't quite right, but they're not large enough to change the overall picture.
  16. Nov 19, 2009 #15
    You do know that people have been detecting dark matter particles in laboratories for over 50 years now? So could you please "start on it". :)
  17. Nov 19, 2009 #16
    I don't want to beat this idea to death - but to answer the above - in Chandrasekar's limit - G and the mass of the star M do not appear as a product MG which determines the simple force relationship of gravity - therefore it is not proper to conclude that a diminished G would have the same affect on the brightness of other objects (those governed by the simple MG product).

    In short - the changing G idea is probably of no merit - but not for the reason given above
  18. Nov 19, 2009 #17
    That's the consensus, but there are enough gaps in our understanding of Sn Ia that it wouldn't hurt for someone to think some more about these issues. Also, once you get past the overall picture, then there is a lot of room for pinning down the exact numbers. Variations in SN Ia may not be enough to kill dark energy, but it is a good idea to get an exact number for how much difference it does make because it may matter for other things like the exact nature of dark matter.
  19. Nov 19, 2009 #18
    Except that they the main sequence line isn't governed by the simple MG product. If you change G, then you change the pressures/temperatures in stellar cores, which changes fusion rates which changes just about everything else.

    Whoa here. Suppose you were to get around that objection, then why wouldn't changing G be of no merit?
  20. Nov 19, 2009 #19


    User Avatar
    Science Advisor

    There is an industry of people thinking more about those issues. Google for "Nearby supernovae factory" for an example of an obervational effort and "snob ssupernovae simulation" for an example of the theory. These are just tips of the iceberg.

    The consensus view at present is that SN have hit the 'systematic limit', meaning that we can't really learn anything more about cosmology by simply getting more high z SN data of the quality we can get with current instruments (that includes the HST). The systematic errors (from possible evolution amongst other things) are now greater than the statisical ones and more data just beats down the statistics not the systematics.

    The way out of this is to either make some theoretical breakthrough, so that the empirical relationships (such as stretch correction) could be put on a firmer basis, get many many more low z supernovae (as the NSF is doing) or preferably to get shiny new instrument, such as the JDEM (Joint Dark Energy Mission) proposed by NASA/DOE and/or the Euclid mission proposed by ESO. When the James Webb Space Telescope comes along this will probably get some good SN data, although the wavelength coverage isn't prefectly suited for SN1A, so a dedicated mission would be better.

    Edit: On the varying G issue, there are plenty of papers looking at this as well. You can't just take the standard equations and simply change G -> G(t) though, since that would violate (in general) Lorentz invariance. To do it properly you need to add an extra scalar field to Einsteins Equations, and in general G will vary with time and space, although the spatial perturbations are relatively small (at least in models obeying observational constraints). Saying 'varying G' then becomes equivalent to a class of modified gravity models containing extra scalar fields, and these are being looked at in significant detail at present.

    Note that it's a strawman to try focus on SN alone though, the evidence for DE is strong, and comes at least as much from structure formation than from SN. It is in structure formations that you get the tighest constraints of gravity variation, and those constraints mean that the internal physics of SN explosions will be unaffected compared to the standard model for gravity models obeying observational constraints. You can't simply look at one data set in isolation.
    Last edited: Nov 19, 2009
  21. Nov 19, 2009 #20


    User Avatar
    Science Advisor

    The supernova community is very much interested in these issues, and they're continuing to work on them. There really isn't a problem there.

    Naturally, and one way of doing that is to make use of different measures of expansion entirely. When they start to disagree (which they really haven't yet), then we can say there's something interesting to say about cosmology from improving on the systematic errors.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook