Perlmutter & Supernovae: Debunking the Myth of Accelerating Galaxies

[twofish-quant]

Perhaps I am mistaken, as I don't have a good grasp of the math of all this, but isn't the accelerating universe model the "best fit" to the data?

I don't think this is a good way of thinking about the problem. Since "best fit" really means nothing. The problem is that the reasoning is circular. In order to have a "best fit" you have to have a model of the problem which is a problem if you don't understand what is going on. If you don't have a model for what is going on, then how can you tell if one fit is "better" than another?

One reason I'm arguing with RUTA is that I do think we would have a serious problem if cosmologists were doing what he thinks they are doing, but they aren't.

What is better is to look at the data, go through all of the possible explanations, and then see which ones are excluded and which ones are allowed. As you get more data, the number of viable explanations goes down.

Would assuming that DL=(1+z)Dp is true only for small Dp be a less reasonable assumption than assuming it is true for all values?

Doesn't matter. The problem is that when you are dealing with unexpected data, there is no basis for figuring out what is a "reasonable assumption." So what you do is to assume that you've got the relationship wrong, and then see what happens.

In fact, what happens is that you end up with a Taylor expansion, and for small z, the first term is (1+z)Dp.

Do we have any real reason for believing that?

GR says that DL *isn't* (1+z)Dp for curved spacetime. However GR also puts some limits into what the relationship between DL and Dp can be.

 

[Drakkith]

What is better is to look at the data, go through all of the possible explanations, and then see which ones are excluded and which ones are allowed. As you get more data, the number of viable explanations goes down.

I'm not sure I see the difference between this and what I said. Is that not a "best fit"? Or am I missing a key point between the two?

 

[twofish-quant]

I have not seen an alternative to accelerated expansion that fits the data as well as the concordance model (LambdaCDM).

I have. Void models seem to work, but they have other problems. Also, as a supernova geek, I'm *really* worried about the assumption that SN Ia are standard candles, but fortunately people are reproducing the data with other distance measures.

One other problem is what does the "fit" tell you. For example, I can take the data, and draw a line through it, but that tells me nothing.

Whether someone would consider alternatives to the assumptions required to render an accelerated expansion depends on their particular motivations.

I don't think it really does have much to do with motivations.

If you keep all the assumptions that lead to accelerated expansion, then you're left having to explain the acceleration. So, why close the door on alternative assumptions motivated by other ideas for new physics that lead to decelerated expansion?

1) First of all, you eliminate the low lying fruit first. There are a *LOT* of possible explanations for the data that involve no new physics at all. As we get more and more data, those explanations are less and less plausible.

2) Second of all, no one else closing the door on new physics that lead to decelerated expansion. The trouble is that no one has come up with one that fits the data. There is an entire industry of people working on alternative gravity models.

But, when the community says they've discovered the accelerated expansion of the universe, that's exactly what they're doing.

What people are saying is that we've spent ten years trying to come up with explanations, and none of them seem to work. If saying "we've discovered accelerated expansion" is stopping people from looking into modified gravity models, that's a bad thing, but I see no evidence that this is the case. It's the reverse, the modified gravity people are telling us that they've tried to come up with alternative explanations, and none of them seem to work.

They might come up with something tomorrow, but if you look for Bigfoot and can't find him, them maybe it's because he isn't there.

If in, say, 20 years we have a robust unified picture of physics and it points to and explains accelerated expansion, I will be on board. I'm not arguing *against* accelerated expansion. I'm arguing for skepticism.

What I'm saying is that if you assume X, Y, and Z you get acceleration. I then go through through X, Y, and Z and then explain the current state of research for X, Y, and Z.

Also part of the reason the point here is to figure out what we need to research next.

 

[RUTA]

Twofish, you and I could continue to discuss technical details associated with alternative assumptions, but unless I can keep that conversation centered on a published paper, such a discussion would violate forum rules (rightfully so, it’s too speculative). Thankfully, that discussion is not essential to the point at hand.

You have presented arguments for your claim that the accelerated expansion of the universe has been directly measured, where by “directly measured” you mean in a sense equivalent to measuring the acceleration of a car on the street or a ball rolling down an inclined plane. I was very interested in these arguments because if I could be convinced that the acceleration was “directly measured,” I would certainly accept it as “fact” and forgo any attempt to explore decelerating alternatives. While you have failed to convince me that we have “discovered the accelerating expansion of the universe,” i.e., that we have indeed “directly measured” accelerating expansion, this discussion allows readers to see why such a claim is made and why it is challenged. They can now make a more informed decision as to whether to believe or remain skeptical.

 

[RUTA]

I'm not sure I see the difference between this and what I said. Is that not a "best fit"? Or am I missing a key point between the two?

Twofish, you say you've seen fits to the SN data that match LCDM. [I'm inferring that these are decelerating models given the context in which you made that statement.] Of course, you can't share them here if they're not published, but do you have any published examples? The decelerating fits I've seen are all discernibly weaker than LCDM at large z.

 

[Drakkith]

Twofish, you say you've seen fits to the SN data that match LCDM. [I'm inferring that these are decelerating models given the context in which you made that statement.] Of course, you can't share them here if they're not published, but do you have any published examples? The decelerating fits I've seen are all discernibly weaker than LCDM at large z.

Are you meaning to quote me and talk to twofish, or is that just a mistake?

 

[RUTA]

Are you meaning to quote me and talk to twofish?

Yep.

 

[RUTA]

Here is an example of what I'm talking about:
Figure 2 in arXiv:gr-qc/0605088v2 (published in Class. Quant. Grav.). You can see the two curves (m vs z) diverging at z = 0.8. The figure stops at z = 1, but if the divergence continues at this rate, the fit would be terrible at z = 1.4 (end of Union2 Compilation, for example).

Here is another example:
http://www.physorg.com/newman/gfx/news/2011/supernovaelight2.jpg (published in Mon. Not. R. Ast. Soc.). He doesn't show the LCDM fit, but I've done this fit (mu vs log(z)) with the SN data in this range and LCDM is discernibly better at large z than this fit.

Anyway, Twofish, if you know of any decelerating models that fit the SN data at large z as well as LCDM, please let me know.

 

[RUTA]

Here is an example from the Supernova Cosmology Project website showing the difference between accelerating and decelerating cases being determined at large z (0.2 and up):

http://supernova.lbl.gov/PDFs/HubbleDiagramPhysicsToday.pdf

I'm working on a sum of squares error for Annila's version of mu (Mon. Not. R. Astron. Soc. 416, 2944–2948 (2011)) using linearized Union2 data from the SCP website. Then we can see how it compares to LCDM's 1.79 and the flat, dust-filled model's 2.68 posted earlier in this thread. Annila shows a fit of mu vs log(z) using data from the SCP website (Fig 3 of his paper), but he does not provide an SSE. Annila's mu = 5log(z*c*T*sqrt(1+z)/10 pc) (obtained via Eq 4 of his paper) so it has only one fitting parameter, T, age of the universe. In that same paper, he has DL = c*T*z/(1+z), so I notice he is not using mu = 5log(DL/10 pc). I'm hoping Twofish will have something to say about that. Anyway, his mu vs z with T = 14Gy maps roughly (eyeball) to LCDM with Ho = 65 km/s/Mpc, OmegaM = 0.24 and OmegaL = 0.76 (best fit for mu vs log(z) in Fig 4 of arXiv:astro-ph/9805201v1 which appeared in Ap. J.). Here is that "eyeball fit"

http://users.etown.edu/s/stuckeym/Plot 15Gy.pdf

Green curve is Annila and red is LCDM. In Fig 3 of his paper, his "best fit" uses T = 13.7Gy, but it looks weak at larger z. I'll let you know what I find.

 

[RUTA]

Oops, that comparison of Annila with LCDM was using T = 15Gy, not 14Gy. Here's the comparison using T = 14Gy:

http://users.etown.edu/s/stuckeym/Plot 14Gy.pdf

In this figure you can see Annila is a bit lower than LCDM at high z, which is consistent with the curve in his Fig 3 looking like it's a bit low at high z using T = 13.7Gy per his figure caption.

http://users.etown.edu/s/stuckeym/Annila Figure 3.jpg

The best fit for Annila gave SSE = 1.95 (same as best fit line) using T = 14.9Gy. For T = 13.7Gy (per his caption) I have SSE = 2.69 (same as best fit flat, dust-filled model). To remind you, I had SSE = 1.79 for LCDM using Ho = 69.2 km/s/Mpc, OmegaM = 0.29 and OmegaL = 0.71. So, Annila's model isn't as good as LCDM, but it's an improvement over LCDM without Lambda.

 

[RUTA]

Here is a paper that was just accepted at Class. Quant. Grav. I couldn't say anything about it before since it wasn't yet accepted, but the Union2 Compilation data is fit with a decelerating universe just as well as LambdaCDM. It's a flat, matter-dominated universe and as far as I can tell, it shouldn't have any problems with WMAP either, although I'd be interested in comments in that regard.

 

[RUTA]

Our essay (http://users.etown.edu/s/STUCKEYM/GRFessay2012.pdf) “Explaining the Supernova Data without Accelerating Expansion” won Honorable Mention in the Gravity Research Foundation 2012 Awards for Essays on Gravitation.

http://www.gravityresearchfoundation.org/announcements.html

There's a nice quote in the essay from Yousaf Butt at the Harvard-Smithsonian Center for Astrophysics:

Various alternatives to an accelerating universe have also been proposed (see,
for example, C. Tsagas, Phys. Rev. D 84, 063503 (2011)). Whether such
alternatives are viable remains to be seen, but the Nobel Committee for
Physics has perhaps acted somewhat prematurely by selecting a preferred
interpretation of the supernova projects’ data. The effect, intentional or not,
is to bully the skeptics into silence, self-censorship, or ridicule, whereas good
science proceeds with a healthy dose of skepticism and with open minds.

There were some big names in the Honorable Mention list to include Jacob D. Bekenstein, Carlo Rovelli, and Ted Jacobson, so we were indeed “honored” to be “mentioned” in that list :-)

The essay is based on our March 2012 paper (see previous post) in Classical and Quantum Gravity (http://arxiv.org/abs/1110.3973) where we fit the supernova data without accelerating expansion or dark energy by suggesting a correction to GR. The idea for proposing such a correction to GR comes from our interpretation of quantum mechanics as described most recently in our April 2012 paper in Foundations of Physics (http://arxiv.org/abs/1108.2261).

 

[twofish-quant]

The problem with the papers is that not obvious that the universe ends up decelerating. The data presentation doesn't include the standard delta distance modulus diagram. Also it didn't include a comparison of a(t) evolution over time. They assert that their model ends up with a decelerating universe but nowhere did I see the graphs to *show* that their model ends up with a decelerating universe.

The problem with the paper was that it was trying to do two things at once. It ended up with a new theory of gravity and then tried to show that it results in a decelerating universe. I would have liked to see a graph of a(t) using their best parameters versus a graph of a(t) in the standard cosmology. Also more discussion about where the differences comes from. They do two things, the change the DL<->DM factor and then they also have a new evolution equation for a(t). Which one causes the universe to decelerate?

 

[twofish-quant]

One thing that bothers me is that it looks like an example of "tired light" and there are reasons to rule out those models...

http://en.wikipedia.org/wiki/Tired_light
http://en.wikipedia.org/wiki/Tolman_surface_brightness_test

One thing that wasn't clear from the papers was how much change in the modulus was needed to eliminate acceleration. One way of doing this would be graph DP_GR versus DP_new_model. Once you have that number, then show it to some observers, and I'm pretty sure they'll consider the amount of darkening you need to be out of bounds.

 

[twofish-quant]

The more I think about it, the more the paper looks like a weak tired light model.

"Tired light" was a class of cosmological models that assumed that GR did something weird to light so that the universe was not expanding. In this situation, the assertion is that the universe is expanding, but that GR is doing something to light to make it look like the universe is expanding more quickly than it really is. At that point, the experiment evidence against "tired light" becomes important.

So my guess is that if someone goes through the papers on tired light, they will find one or more experiments that kill the idea. I'm guessing that someone already did this, but it's not publishable to find that yes, "weak tired light" doesn't work.

Now if it turns out that the experimental evidence doesn't rule out "weak tired light" then you've got a paper.

The general way of presenting unconventional results is to present the paper as something that will confirm the conventional result. If you find something that supports "weak tired light" claim it's an anomaly that requires further investigation and that if you find X that will support the prevailing theory. Of course, you may be of the opinion that people will find not-X, at which point you act surprised.

The problem with the paper as written is that it's a quantum gravity paper and not an observational astronomy paper. The parameter that causes false acceleration to be observed is a free parameter, and I know that if I twist the parameters hard enough, I can get whatever result that I want. The question that I'd be interested in is "how hard to you have to twist the parameters" and are there any observational blocks to twisting those parameters.

 

[RUTA]

Thanks for your response, twofish. The evolution of the modified equations follows a(t) for Einstein-deSitter very precisely all the way back to the stop point, so it is in fact decelerating. Also, there is no mechanism causing light to redshift in transit as in tired light. We are proposing a different mechanism altogether for the coupling of EM sources in an FRW universe.

Do we believe astrophysicists should be exploring such a proposed change to general relativity? No. The proposed modification has serious consequences for many other things that work well, i.e., all those associated with the Schwarzschild solution. Until those ramifications are fleshed out, the idea is largely worthless for astrophysics. We are working on the Schwarzschild modifications now and that issue will be resolved in the next year or two.

The reason the paper is published in CQG is, as you point out, because it’s a paper on quantum gravity rather than astrophysics. The reason I posted it here is, as I argued earlier, because I believe the Nobel citation claiming “the discovery of the accelerating expansion of the universe” is premature. Is accelerating expansion the best explanation of the data as of now? Yes, but who knows what the future holds. The Nobel committee decided to award the prize for a particular interpretation of the data, rather than for acquiring the data itself (which I think is worthy). As Dr. Butt said, “The effect, intentional or not, is to bully the skeptics into silence, self-censorship, or ridicule, whereas good science proceeds with a healthy dose of skepticism and with open minds.”

 

[twofish-quant]

Also, there is no mechanism causing light to redshift in transit as in tired light. We are proposing a different mechanism altogether for the coupling of EM sources in an FRW universe.

I'm less interested in the specific mechanisms than the observational tests. By changing the DL / DP relationship, that may causes changes that act "as if" it were tired light (even though it isn't). What I'm interested is if the observational evidence against tired light also constrains the DL / DP. One reason I'm interested in this is that there *aren't* observational constraints against the DL / DP relationship, this is a "hole" in Perlmutter's paper, and it's something that should be patched up.

One problem is that because I'm not physically on a university, I don't have easy access to people that I can ask about this. There were some people that I'd ask about this in the past, and sometimes they come up in five minutes with a reason why this won't work. If they think about it for two weeks, and they can't come up with anything, then it's a paper.

The SN papers did a pretty good job at "patching holes", and I don't recall anyone mentioning variations on the DL/DP relationship. This could either be because it's so obviously wrong that no one bothered mentioning it. Or it could be because no one thought this as an issue.

Do we believe astrophysicists should be exploring such a proposed change to general relativity? No.

But it could be that the even if the quantum gravity theory is wrong, if that something else is changing the DL/DP relationship, that's still very interesting.

The reason I posted it here is, as I argued earlier, because I believe the Nobel citation claiming “the discovery of the accelerating expansion of the universe” is premature.

On the other hand, it's clear that Perlmutter discovered *something big*. If it turns out that the universe is decelerating and GR is wrong, that's even more earth-shattering than an accelerating universe, and worth a Nobel.

The Nobel committee decided to award the prize for a particular interpretation of the data, rather than for acquiring the data itself (which I think is worthy).

I don't think that's quite true. "Dark energy" is the "least weird" explanation for the Perlmutter's results. There are other explanations but they are all weirder.

“The effect, intentional or not, is to bully the skeptics into silence, self-censorship, or ridicule, whereas good science proceeds with a healthy dose of skepticism and with open minds.”

I don't think this is worse than any other "dominant paradigm" and I don't think it's too bad in astrophysics.

Also there are social tricks that get around this. There's the "Columbo strategy." If I thought that the world were flat, I wouldn't publish a paper saying the "The World Is Flat", I'd publish a paper saying "Observational Constraints on the Roundness of the Earth." Here are some observational tests that you can do to show that the world is round, oh wait, you did those tests and then didn't work, well... that's surprising... Hmmm... Well since we all know the world is round, why don't you try doing this... Oh... Your coming up with odd answers... Well... What do *you* think is going on?

I don't think astrophysics is ossified. I do have very little respect for academic finance and economics, but that's something else.

 

[Chronos]

Knowledge is power and power is jealous. The ancients clearly understood this and defended their knowledge from prying eyes. Sacrificing 'heretics' was a well received and popular tactic dating back thousands of years.

 

[twofish-quant]

Knowledge is power and power is jealous. The ancients clearly understood this and defended their knowledge from prying eyes. Sacrificing 'heretics' was a well received and popular tactic dating back thousands of years.

In science sometimes the lunatics end up running the asylum.

Also, there are a surprising number of people with power that have very unconventional views. I know of at least two Nobel prize winners that are convinced that black holes don't exist, and I know of a former president of the American Astronomical Society that has extremely unconventional views on galactic jets.