I wasn't challenging that paper's conclusion with respect to time dilation being confirmed, only their conclusion that this ruled out tired light hypotheses. I think the actual words they use are something like "tired light hypotheses or any other theories that exclude time dilation". If we accept that there could be tired light hypotheses that include time dilation (as you did earlier), then their wording is at least misleading. But as for the z range of the SNe they are using, I'm fine with that!

"as I did earlier."

Well, I suppose I did claim that "anything's possible." But as I said, no tired light has been proposed that produces time dilation, so this argument is really quite silly. Until one is proposed, claiming that tired light is disproven by the observation of time dilation is a perfectly valid claim.

__________________
"What do you care what other people think?" -- Richard Feynman
"For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." -- Feynman, at the conclusion of his Challenger report

I think the important thing is that there actually are not any tired light theories that have time dilation. So anyone trying to argue in favour of a tired light theory is either in conflict with the available observations or they are arguing for a (grossly) incomplete theory. There is no reason we should take the latter type of theories very seriously.

I don't think that there was any change. Cosmological redshift was predicted before it was observed and redshift has always been associated with the Robertson-Walker models and thus with the expansion of space. The equations in a RW spacetime that lead to the prediction of redshift themselves assume that there is time dilation.

Well, no, not so silly. Here's a more familiar example. Many astrophysicists accept that there is dark matter. No one has been able to pinpoint what that dark matter is precisely. Yet studies can still be done showing how it behaves, in colliding galaxies, for instance. Now someone could propose a candidate for dark matter and someone else could prove it's wrong, but that wouldn't disprove the dark matter hypothesis. Similarly, dark matter can't be proved till someone shows exactly what it is. We're in a somewhat similar situation with tired light (though fewer people are on this train). You can disprove a particular tired light mechanism, but that wouldn't disprove the concept itself. And we can still make statements in cosmology that use the tired light idea (dE/dt = - EH, where E is the initial photon energy and H the Hubble constant), even without knowing the precise mechanism. My gripe with the mainstream is that too often they try to say they've disproved tired light, when it's only one mechanism they've disproved.

As I mentioned in an earlier post, if we take a train of em waves, and suppose that it undergoes energy loss through any tired light process, you will see time dilation. The greater distance between the crests of the weaker waves ensures this.

It could have been that someone predicted time dilation in SNe using the Robertson-Walker model, but the first paper that actually predicted this time dilation used the simple Doppler shift equation:

Wilson, O.C., ApJ 90, 634 (1939)

It has little to do with it, but every measurement becomes less accurate with distance as spectra broaden. They point out in the paper there is no relativist trend in the local sample, it is too blurred.

It is an explosion, not a first order chemical reaction. LARGER events take longer to occur: If you looked at the spectra of two tons of TNT exploding two milliseconds after initiation, it looks very much like the explosion of twenty tons of TNT twenty milliseconds after the explosion. Ever look at explosive crater? It is hard to tell a firecracker crater from a pound of dynamite, unless you have some frame of reference. Why shouldn't supernova of various magnitudes present the same difficulty?

Supernova 2006gy was AT LEAST 1.5 magnitudes greater than your local garden-variety supernova events; and had similar spectral signature. Yes it was different, but how unique is this event when you compare it to the highly redshifted sample we happen to observe, which should be DOMINATED by over-magnitude events? What Blondin has really shown, is that either 2006gy was either a one-of-a-kind-never-observed in the most distant events, or that the most distant events we have observed truly look very much like the local sample, even though they are much bigger explosions. That is a reasonable conclusion.

We don't have any local examples of events like 2006gy - we cannot know the true relationship of the light-curve width (and the associated spectral signature) to magnitude until we have a more complete library of the spectra of overluminous events like 2006gy. Yes, it would be cruel of nature to conspire in a way that overluminous, distant events look very similar to smaller events we see locally, but it is more reasonable to conclude distant supernova are over-luminous than it is to conclude that, on the average, they are well-represented by fairly local events.

As for a comparison of the spectra of 2006gy and local events, there are more similarities than differences; and just as the general shape a large ocean liner is similar to a punt, we don't have the luxury of assuming the ultra-long light curve of 2006gy is a freak occurance. If 2006gy occurred at a redshift of ~0.3 and time dilation is a feature of space, we must find light-curves at redshifts of >0.6 that are at least twice as long as 2006gy. We haven't seen that. We can't assume we can see a single ocean liner near our dock, but only cruisers in the most distant arms of the ocean around us.

Again, there is no reason to anticipate larger events do not evolve in a similar manner to smaller events, only more slowly. Think about the Goldhaber and Perlmutter studies: They demonstrated NO selection effects occurred in the photometric light curves they studied. But we now know brighter supernova events do occur, and that they have much longer light-curves. Therefore, there should be obvious selection effects, why weren't they picked up? Why wasn't it obvious in the earlier studies that brighter events were being observed in the distance, brighter events with longer light-curves? Because the light-curves Goldhaber and Perlmutter studied were not longer, at least not after correction for time dilation! You can't say the earlier studies were flawed unless the studies based upon spectral signature could also be flawed for exactly the same reason: Both the spectra and photometric light curves of large supernova like 2006gy look very much like smaller ones.

Reddening is an important parameter for the same reason Malmquist bias is: More distant observations should be redden. Relativistic time correction creates an effect similar to reddening, and when you studied distant events, you must have reddening budget for both relativistic light dilation and dust reddening. Earlier studies glossed over this difficency - they did not have enough reddening effect to subtract out both relativistic and host galaxy reddening, so they assumed there was virtually no host-galaxy or intergalactic reddening in their observed sample. I think this is a most obvious error: Both Malquist bias and reddening should occur. If there is not enough budget for natural effects, something is unnatural about the data reduction.

[Their results depend] not at all on photometric properties, and the spectra they took were perfectly normal SN Ia. I don't see how your complaints are relevant, and I'm now even more convinced that Malmquist bias is unimportant for this measurement since the spectrum is so strongly dependent on the time evolution of the supernova.[/quote]

I would agree with you if I did not know 1) Malmquist bias is a natural condition: We are more likely to observe only the brightest of the most distant events we can detect. 2) much brighter supernova than your garden variety type Ia supernova can occur. 3) Supernova 2006gy was one of these events, and the spectral evolution over time was similar to a garden variety type Ia even though it was many times brighter. 4) We do not find nearly as many supernova at great distances as we expected to, based upon local observations 4) when you look at the light-curves of the brightest events we observe locally, they are longer than the light-curves of the most distant events AFTER the corrections are made for time dilation, which seems to indicate that the most distant events are not even as bright as the brightest of the local events. That is not natural, but if you do not make the correction for time dilation, the most distant events we observe are dominated by very long light-curves and therefore very luminous events - as the distant sample should be.

__________________
jwj

It's ok not to know. We should try harder to find out.

This is an unreasonable conclusion. We have one single extra-ordinary supernova, 2006gy, with its own unique spectral characteristics. This is inot enough to create any expectation that events like this will show up in a given sample.

Even if we expect supernovae like this to occasionally show up in distant samples, there are a number of cross-checks in the procedures to look out for outliers like this.
How is it reasonable, without any evidence, to suppose that there is a good chance of a conspiracy?
They demonstrated that significant selection effects influencing their results were unlikely, not that there were such effects.
Again, we know that there is the possibility, but we know nothing of the probability. It is not appropriate to simply assign a 50/50 chance here, especially given the kinds of cross-checks that are performed in multiple papers.
Please explain this.
Could you please point out where you find this motivation for not considering host galaxy reddening? The actual papers that I read don't seem to report this conclusion. Is there another conspiracy here?

While you're reviewing these papers, why don't you address their budget for Malmquist bias?

Because supernova type Ia have very distinct spectral evolution, due to radioactive decay, radiation pressure and gas dynamics? Duh.

Wrong, wrong, wrong. Look at Figure 4 of Smith et al. (2007). 2006gy looks nothing like a type Ia!

I may or may not reply to the rest of your comment (though Kwalish Kid did a good job), but you certainly got that one wrong! Why don't you download SNID yourself and try it out on some supernova spectra instead of just belly-aching about things you don't actually appear to understand?

Are you afraid that you might learn that supernova researchers can tell the difference between "anomalous" and "normal" events, thus destroying your most common complaint?

__________________
"What do you care what other people think?" -- Richard Feynman
"For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." -- Feynman, at the conclusion of his Challenger report

A comment on SN 2006gy. The extreme luminosity of this galaxy could simply be the result of an incorrect distance. This would not be surprising since the distance is calculated assuming Hubble distances with H0=72 ( Smith et al 2007). Smith et al adopt a distance modulus of 34.32.

The Perseus cluster to which the host galaxy (NGC 1260) belongs is at one one end of the Pisces filament. Tully&Pierce (2000) find a distance to Pisces of 60.0 Mpc or a distance modulus of 33.89 - which must be corrected by -0.06 mag to account for the final Freedman et al (2001) cepheid distances.

There are also a number of spirals in the Perseus cluster for which data is available for calculating K-band TFR distances. UGC 2736 has a redshift within 100 km s-1 of NGC 1260 and has a K-band TFR distance modulus of 33.50 using the 2MASS K-band magnitude and rotational velocity from Springob et al. This is 0.82 mag less than the Hubble distance modulus and would reduce the absolute magnitude of SN 2006gy from -22.00 to -21.18.

__________________
"The scientist who asks the right question reconnoiters a new patch of the unknown, and may, with luck, bring it within the constricted but expanding boundaries of the known."

~Timothy Ferris (The Red Limit) 1982

There is no way for us to cross check the magnitude of distant supernova; a feature article in the Universe Today is even more evidence of a giant boner:

Say what? That means, (if Driver et al is correct), the most distant supernova are also generally brighter than currently calculated! That means they are indeed over-bright relative to local type Ia, OR they are not nearly so far; throwing a complete curve in the Hubble distance ladder.

The 'cross checks' involve other magnitude/distance relationships, but none of them extend nearly so far as supernova...except maybe the Tully-Fisher relationship.

All nature needs to conspire is over-confidence by man. Whether we are building cosmic theories, cities near the ocean or dams near hidden fault lines, we will make mistakes.

But there should be selection effects! Now we know that supernove Ia occur that are much brighter than typical type Ia; we should question why their methodology did not pick this up.

If the cross checks really worked, they should have figured out that the universe is dimmed by dust. As I mentioned before, the supernova studies do not have enough reddening budget to account for both dust and GR reddening; so they assumed there is virtually no dust attentuation in the supernova population they studied. (I am speaking of the ~1996-2004 papers of Nugent, Perlmutter, Goldhaber and others.

http://xxx.lanl.gov/PS_cache/astro-p.../9812473v1.pdf


[quot=Smith]SN 2006gy is the first supernova for which we have good reason to suspect a pair-instability explosion. Based on a number of lines of evidence, we eliminate the hypothesis that SN 2006gy was a ``Type IIa'' event, that is, a white dwarf exploding inside a hydrogen envelope[/quote]

Extremely brilliant - Much more brilliant, and not a Type IIa event; and what mainly separates it from a Ia is the obvious hydrogen envelope. But what if two White dwarf without the envelope collided and produced an overluminous event more similar to a type Ia? What are the odds? No one knows. This paper claims Tired Light theories are dead with six sigma confidence. No one should have six sigma confidence we understand the magnitude limits of distant supernova events.

__________________
jwj

It's ok not to know. We should try harder to find out.

So why don't you calculate it?

But maybe the fact that all the anomalous events we've observed have been spectroscopically distinct means that such events are very rare, and that we can identify them spectroscopically. Can you produce an example of a non-type Ia in the local universe that was consistently misclassified as a type Ia? That seems to be what you're suggesting is occurring for all the supernova spectra used in this (and other) studies.

And, as Smith et al. point out in the caption to Figure 4, "Also plotted is a spectrum of the Type Ia SN 1991T at t = 35 d (Filippenko et al. 1992) for comparison with our day 36 spectrum of SN 2006gy; there is essentially no similarity between the two spectra." Just look at the spectra: "what mainly separates it from a Ia" is not only the presence of a hydrogen envelope, but all the other features of the spectrum as well. Pretty much all of section 3.3 discusses why it can't be a type Ia, approaching the problem from multiple directions. Do you just brush away their analysis because it is inconvenient for you?

And then there's the SCP Union Compilation, where the resulting cosmology is robust to various methods of outlier rejection, using ~300 distant supernovae from many different surveys...

__________________
"What do you care what other people think?" -- Richard Feynman
"For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." -- Feynman, at the conclusion of his Challenger report

I'll have to take a look at that. However, does this influence the question of the relative brightness of different SN Ia and their results?
Have you really haven't read through the papers involved? There are pages in each paper devoted to different means of measuring and reducing error. The various cross-checks on the strength of the results is not simply due to other means of checking distance.
Are you seriously saying that there is mental content in nature that actively works to change the physical nature of the universe based on its perception of human motivation? Or are you saying that in every case where scientists work out error bounds, they are grossly wrong in their calculations?
We have one example of a bright SN and it does not appear to be a SN Ia.
[quote]http://xxx.lanl.gov/PS_cache/astro-p.../9812473v1.pdf
So let's get this straight:
1. You make the claim that the SN Ia researchers ignore the possibility of reddening because they can't work it in to their error budget.
2. Your evidence of this is a passage from one paper where they explicitly take reddening into account and add to their error budget accordingly.

I feel the need to point out for other that this is a pattern that Jerry gets into. He makes claims and backs them up with papers, and sometimes even quotes, that say the opposite of his point.

There are a multitude of examples of type Ia that were later classified as type Ic.

What is more compelling to me is the distant observations that appeared to be type Ia, but then the light curve tailed-off way to soon, so they were ignored. Adam Reiss mentions one of these in his 2004 (?) paper. These are the supernovae that I contend are identical to local type Ia, but after correction for time dilation, they are too short. Why would a distant supernova have a type Ia spectrum, but be too short? We shouldn't stop observing such an event, because if the dust reddening is truely a factor of two greater, and type Ia are truely very tightly constrained in magnitude, Adam Reiss could have discovered this more than two years ago.