PDA

View Full Version : New Kind of Supernova Discovered



Fraser
2006-Sep-22, 09:37 PM
Astronomers used to believe that all Type 1a supernovae were essentially the same brightness. That's because they explode with the same amount of fuel. But now a supernova has been discovered that's twice as bright as all the other Type 1a supernovae. This is a problem, since this kind of supernovae are used as standard candles, to determine distances across the Universe. Most recently, these supernovae have been used to calculate the mysterious force called dark energy that seems to be accelerating the expansion of the Universe.

Read the full blog entry (http://www.universetoday.com/2006/09/22/new-kind-of-supernova-discovered/)

ngeo
2006-Sep-22, 11:53 PM
Well done Fraser. I was just now looking to see whether BAUT forum had this story, since the following story appeared on Space Daily today:

An international team of astronomers at the California Institute of Technology, University of Toronto, and Lawrence Berkeley National Laboratory have discovered a supernova more massive than previously believed possible. This has experts rethinking their basic understanding of how stars explode as supernovae, according to a paper to be published in Nature on September 21.
The lead author of the study, University of Toronto postdoctoral researcher Andy Howell, identified a Type Ia supernova, named SNLS-03D3bb, in a distant galaxy 4 billion light years away that originated from a dense evolved star, termed a "white dwarf," whose mass is far larger than any previous example. Type Ia supernovae are thermonuclear explosions that destroy white dwarfs when they accrete matter from a companion star.
The discovery was made possible through images taken as part of a long-term survey for distant supernovae with the Canada France Hawaii Telescope. Follow-up spectroscopy led by Richard Ellis, Steele Family Professor of Astronomy at Caltech, with the 10-meter Keck Telescope was key to determining the unusually high mass of the new event.
Researchers say the surprisingly high mass of SNLS-03D3bb has opened up a Pandora's box on the current understanding of Type Ia supernovae and, in particular, how well they might be used for future precision tests of the nature of the mysterious "dark energy" responsible for the acceleration of the cosmic expansion.
Current understanding is that Type Ia supernova explosions occur when the mass of a white dwarf approaches 1.4 solar masses, or the "Chandrasekhar limit." This important limit was calculated by Nobel laureate Subramanyan Chandrasekhar in 1930, and is founded on well-established physical laws. Decades of astrophysical research have been based upon the theory. Yet somehow the star that exploded as SNLS-03D3bb reached about two solar masses before exploding.
"It should not be possible to break this limit," says Howell, "but nature has found a way! So now we have to figure out how nature did it."
In a separate "News and Views" article on the research in the same issue of Nature, University of Oklahoma professor David Branch has dubbed this the "Champagne Supernova," since extreme explosions that offer new insight into the inner workings of supernovae are an obvious cause for celebration.
The team speculates that there are at least two possible explanations for how this white dwarf got so fat before it went supernova. One is that the original star was rotating so fast that centrifugal force kept gravity from crushing it at the usual limit. Another is that the blast was in fact the result of two white dwarfs merging, and that the body was only briefly more massive than the Chandrasekhar limit before exploding.
Since Type Ia supernovae usually have about the same brightness, they can be used to map distances in the universe. In 1998 they were used to make the surprising discovery that the expansion of the universe is accelerating. Although the authors are confident that the discovery of a supernova that doesn't follow the rules does not undermine this result, it will make them more cautious about using them to measure distance in the future.
Ellis summarizes: "This is a remarkable discovery that in no way detracts from the beautiful results obtained so far by many teams, which convincingly demonstrate the cosmic acceleration and hence the need for dark energy. However, what it does show is that we have much more to learn about supernovae if we want to use them with the necessary precision in the future. This study is an important step forward in this regard."
Peter Nugent, a staff scientist with the scientific computing group at Lawrence Berkeley National Laboratory, is a co-author of the Nature paper.

Then, taking this from the wikipedia entry on Dark Energy:

“However, it should be kept in mind that alternative models of type Ia supernovae exist. Noted gamma-ray astronomer John Middleditch favors the view that a Ia/c is produced from the merger of two degenerate cores of common envelope WR stars, or of two CO white dwarfs. The observed consistency in absolute magnitude of Ia would then be an artifact caused by selection bias; when the errors introduced by this bias are taken into account, there may be no cosmic acceleration effect in distant Sne.”

My comment is, it appears reports of cosmic acceleration are premature. And the question, what happens to the scheme for dark energy?

RussT
2006-Sep-23, 02:41 AM
My comment is, it appears reports of cosmic acceleration are premature. And the question, what happens to the scheme for dark energy?

Ah, we loose 75% of the universe...where did it go?

Kinda makes you wonder again, if DE/DM could be the same thing 'physically'.

What I would really like to know is, in addition to all the 1a/b and Type II supernova, where are all the 500 per day 1c's hiding??? That is almost 200,000 per year X 10 years, thats 2 million.
Especially considering that they have already declasified many 1c's back to 1a's and 1b's!

antoniseb
2006-Sep-23, 01:43 PM
There is something in this story that doesn't make sense to me:

In old, dead galaxies even the biggest stars are small, Nugent explains. The only kinds of Type Ia supernovae possible in these galaxies are likely to be the binary-system, mass-accreting, Chandrasekhar-mass type. But young star-forming galaxies produce massive objects and could be rich in white-dwarf plus white-dwarf binary systems, so-called “double-degenerate” systems.

To me it seems that only young star-forming galaxies could have any stars left in them that are large enough to drop a solar mass or more or accreted matter onto a closely orbiting white dwarf (note that RS Oph shows that a lot of that accreted matter doesn't stay put). Thus there would be a bias toward being a higher fraction of the spiralling white-dwarf SN1a events happening in older galaxies.

Even so, the ratio of normal SN1a and superchandrasekhar SN1a seems to be about a thousand to one (give or take a power of ten). It is important to weed these from the data, but they aren't that common.

antoniseb
2006-Sep-23, 02:34 PM
Just as a followup, this is from a short paper (http://www.arxiv.org/abs/astro-ph/0609616) to appear in the 21 Sept 2006 issue of Nature.

I often attribute such misunderstandings to the science journalist, but in this case it is the scientists who wrote the paper who are making the mistake.

ngeo
2006-Sep-23, 02:55 PM
Antoniseb isn’t the quote implying that supernovas in old galaxies will be very rare since there aren’t as many white dwarf binary systems in old galaxies, and the only way to make enough mass for a supernova is to slowly accrete matter? Whereas in young galaxies there are more white dwarf binaries and more chance for them to merge. Which in turn implies that more distant supernovas will be merger results rather than accretion results.

To add to this story there is a paper from Middleditch which I referred to in another thread,

http://arxiv.org/abs/astro-ph/0608386


The abstract reads:

“Type Ia Supernovae (SNe) have been used by many to argue for an accelerated expansion of the universe. However, high velocity and polarized features in all nearby SNe Ia, show that the paradigm for Type Ia SNe is drastically and catastrophically invalid. By now it is also clear that an extreme version of the axisymmetry seen in SN 1987A is the correct paradigm for SNe Ia and Ic. A Ia/c is produced from the merger of two degenerate cores of common envelope WR stars, or of two CO white dwarfs. Its polar blowouts produce the observed high velocity and polarized spectral features in Ia's, and its equatorial bulge is much brighter in Ia's, due to the greater fraction of 56 Ni contained within it. These become classified as Ia's when viewed from the merger equator, and Ic's when viewed from the poles. Thus cosmology determined strictly from Ia's alone is flawed at its very foundation: the local sample is selectively biased. The problem arose with the more distant supernovae, when the high velocity polar blowout features, which initially obscure part of the Ia/c equatorial bulge, expose a greater fraction of it, particularly when viewed off the equator, during the interval when Delta m_15 is measured, leading to a smaller decrease in observed luminosity. The width-luminosity correction was thus too small, and the result was a distant SN Ia which appeared to be too faint for its redshift. When the errors introduced by this process and others are taken into account, there may be no cosmic acceleration effect in distant Sne.”

Beside calling into question the whole idea of cosmic accleration, maybe that explains the Ic supernovas.

antoniseb
2006-Sep-23, 03:11 PM
Antoniseb isn’t the quote implying that supernovas in old galaxies will be very rare since there aren’t as many white dwarf binary systems in old galaxies, and the only way to make enough mass for a supernova is to slowly accrete matter? Whereas in young galaxies there are more white dwarf binaries and more chance for them to merge. Which in turn implies that more distant supernovas will be merger results rather than accretion results.

No. The quote clearly states:
The only kinds of Type Ia supernovae possible in these [elliptical] galaxies are likely to be the binary-system, mass-accreting, Chandrasekhar-mass type.
But that is exactly wrong. There can't be any stars massive enough in such a galaxy to provide that much accretion to a white dwarf to trigger the Chandrasekhar-type explosion. The only type possible in such a galaxy is the super-Chandrasekhar explosion from the merger of two white dwarfs.

Duane
2006-Sep-23, 05:40 PM
No. The quote clearly states:
The only kinds of Type Ia supernovae possible in these [elliptical] galaxies are likely to be the binary-system, mass-accreting, Chandrasekhar-mass type.
But that is exactly wrong. There can't be any stars massive enough in such a galaxy to provide that much accretion to a white dwarf to trigger the Chandrasekhar-type explosion. The only type possible in such a galaxy is the super-Chandrasekhar explosion from the merger of two white dwarfs.

Context. That there are going to be very few N. G and K type stars left in an evovled galaxy does not mean there are none.

The comment simply states that the only type of SNIa in such a galaxy would reaquire a "classical" white dwarf in orbit with another star.

Think of a white dwarf orbitting an old G or K type dwarf that finally reaches it's red giant stage.

antoniseb
2006-Sep-23, 05:49 PM
Context. That there are going to be very few N. G and K type stars left in an evovled galaxy does not mean there are none.

There should be plenty of G & K stars. There should be no O, B, or A stars, and an increasingly smaller supply of F stars. My understanding is that a normal white dwarf starts out at 0.6 to 0.7 Solar masses as a remnant from a 2 to 6 Solar Mass star. I think it is fair to assume that less than 30% of an expanding Red Giant's mass is going to be completely accreted onto a closely orbiting white dwarf. What size star is required to get up to a 1.4 Solar Mass Chandrasekhar explosion? I think it requires an A or better.

pantzov
2006-Sep-24, 08:41 AM
i love it! anything that puts into doubt peoples' certainty that that the age and size of the universe are known always makes me happy.

trinitree88
2006-Sep-25, 10:27 PM
Antoniseb isn’t the quote implying that supernovas in old galaxies will be very rare since there aren’t as many white dwarf binary systems in old galaxies, and the only way to make enough mass for a supernova is to slowly accrete matter? Whereas in young galaxies there are more white dwarf binaries and more chance for them to merge. Which in turn implies that more distant supernovas will be merger results rather than accretion results.

To add to this story there is a paper from Middleditch which I referred to in another thread,

http://arxiv.org/abs/astro-ph/0608386


The abstract reads:

“Type Ia Supernovae (SNe) have been used by many to argue for an accelerated expansion of the universe. However, high velocity and polarized features in all nearby SNe Ia, show that the paradigm for Type Ia SNe is drastically and catastrophically invalid. By now it is also clear that an extreme version of the axisymmetry seen in SN 1987A is the correct paradigm for SNe Ia and Ic. A Ia/c is produced from the merger of two degenerate cores of common envelope WR stars, or of two CO white dwarfs. Its polar blowouts produce the observed high velocity and polarized spectral features in Ia's, and its equatorial bulge is much brighter in Ia's, due to the greater fraction of 56 Ni contained within it. These become classified as Ia's when viewed from the merger equator, and Ic's when viewed from the poles. Thus cosmology determined strictly from Ia's alone is flawed at its very foundation: the local sample is selectively biased. The problem arose with the more distant supernovae, when the high velocity polar blowout features, which initially obscure part of the Ia/c equatorial bulge, expose a greater fraction of it, particularly when viewed off the equator, during the interval when Delta m_15 is measured, leading to a smaller decrease in observed luminosity. The width-luminosity correction was thus too small, and the result was a distant SN Ia which appeared to be too faint for its redshift. When the errors introduced by this process and others are taken into account, there may be no cosmic acceleration effect in distant Sne.”

Beside calling into question the whole idea of cosmic accleration, maybe that explains the Ic supernovas.

ngeo. Thank you.That the 1a's were suspect as standard candles was an issue when Peter Nissenson, Martina Karovska, and Costas Papaliolios determined that Mira pulsed asymmetrically, in a major/minor axis ratio of ~7/5, and that Nissenson et al determined that supernova 1987a had ejected a high velocity plasma cloud called in the press, Son of Supernova, and largely ridiculed as an artifact of image processing,...(later determined to one of two bipolar jets)...shortly after SN1987a catapulted to fame.(circa 1988)
The merits of using type 1a's as standard candles, when their luminosities were angle dependent was always questionable. Faraday rotation curves of remnants, to determine viewing angles can help ameliorate the inference from the asymmetry that they are strict inverse square law sources.
It took the papacy 400 years to apologize to Galileo. I wonder when Peter gets his just do? Kudos. Pete.

Ihopeso
2006-Sep-26, 01:28 AM
So what are the ramifications of this finding?

Is it possible we may be wrong elsewhere in our deduction of the goings on out there in the universe.

Getting more confused all the time. Just how far is the next galaxy over then?

:confused:

antoniseb
2006-Sep-26, 10:56 AM
So what are the ramifications of this finding?

Type 1a supernoae have recently been used as standard candles to determine distances to galaxies that are too far away to use Cepheid variables, or 'the largest red giant' method. But so far they haven't been usable beyond about z=2 because they aren't bright enough for our current equipment to measure properly at that distance or greater.

In this case, there is an indication that some percentage (apparently a small percentage) of SN1a's are brighter than the narrow range that most of them fall into. The one observed in this study was twice as bright, which implied that it was 30% closer than it really was.

The fact that this type of supernova is detectably different, and relatively rare means that no previous work is corrupted by the find, and while future studies will take a little more work to check on this possible detail, future work is more certain, not less.

VanderL
2006-Sep-26, 12:37 PM
Type 1a supernoae have recently been used as standard candles to determine distances to galaxies that are too far away to use Cepheid variables, or 'the largest red giant' method. But so far they haven't been usable beyond about z=2 because they aren't bright enough for our current equipment to measure properly at that distance or greater.

In this case, there is an indication that some percentage (apparently a small percentage) of SN1a's are brighter than the narrow range that most of them fall into. The one observed in this study was twice as bright, which implied that it was 30% closer than it really was.

The fact that this type of supernova is detectably different, and relatively rare means that no previous work is corrupted by the find, and while future studies will take a little more work to check on this possible detail, future work is more certain, not less.


No way Antoniseb! You can't just cherrypick the results that fit to already established findings. That would be a fundamental mistake. And how rare are these SNe really? How about the classification system, how do we know we didn't misclassify some or maybe many SNe?

Cheers.

antoniseb
2006-Sep-26, 01:14 PM
How about the classification system, how do we know we didn't misclassify some or maybe many SNe?

Going forward, we'll be testing all SN1a's to see if they are this variety. This will tell us just how rare or common they are. The fact that this one stands out on the graph alone as far as correlating redshift and SN1a brightness tells us it is pretty unusual.

VanderL
2006-Sep-26, 04:19 PM
Going forward, we'll be testing all SN1a's to see if they are this variety. This will tell us just how rare or common they are. The fact that this one stands out on the graph alone as far as correlating redshift and SN1a brightness tells us it is pretty unusual.

Ok, good approach, but until we have something more than this, it is a fundamental error to claim that this finding does not influence earlier conclusions. To me it seems obvious that supernovae are not standard candles, and anything based on that assumption must therefore be wrong too. The same thing is argued by Middleditch and Vishwakarma and probably some others, and most probably caution is also propagated by the same people that have been interpreting SNe as standard candles in the first place. The Chandrasekhar limit is fundamental in these interpretations, anything wrong with that and the house of cards comes tumbling down.

Cheers.

antoniseb
2006-Sep-26, 05:48 PM
The Chandrasekhar limit is fundamental in these interpretations, anything wrong with that and the house of cards comes tumbling down.

It still seems fundamental, and not contradictd by these findings. This particular supernova most likely was the result of two white dwarfs spiralling into each other, that collectively had 2 Solar masses. This should be a rare event.

The paper also cites the possibility that the supernova was a result of an extremely rapidly rotating white dwarf having the ability to be some percentage over the Chandrasekhar limit, but this hasn't been demonstrated.

I think you are wrong to characterize this as a house of cards.

VanderL
2006-Sep-26, 08:38 PM
It still seems fundamental, and not contradictd by these findings. This particular supernova most likely was the result of two white dwarfs spiralling into each other, that collectively had 2 Solar masses. This should be a rare event.

Most star systems are binaries, so why should this be a rare event? Plus I think you have it backwards here, the lightcurve can only be model-fit to a 2Msol star, the merger of 2 white dwarfs is a conjecture, not a fact.


The paper also cites the possibility that the supernova was a result of an extremely rapidly rotating white dwarf having the ability to be some percentage over the Chandrasekhar limit, but this hasn't been demonstrated.

Some? That would be more like 50%, right?


I think you are wrong to characterize this as a house of cards.

No, I think being forced to reconsider the Chandrasekhar limit and being forced to reconsider SN1a's as standard candles could mean exactly what I said.

Cheers.

antoniseb
2006-Sep-26, 08:53 PM
Most star systems are binaries, so why should this be a rare event? Plus I think you have it backwards here, the lightcurve can only be model-fit to a 2Msol star, the merger of 2 white dwarfs is a conjecture, not a fact.

The light curve only has to do with how much Nickel-56 is decaying. This is the right amount for the explosion to have happened with 2 solar masses of white-dwarf content. It says nothing about whether it was one star or two initially. Yes it is conjecture (to a degree), but it is the only conjecture so far that fits the data. This does not change the Chandrasekhar limit.

Concerning most stars being binaries, yes that is true, and binaries are required for any type 1a supernova. What makes this a rare event is that the binaries would have to both become white dwarfs, and then radiate enough orbital energy through gravity waves that they can merge. This is trickier than the normal type 1a sequence of events.

Duane
2006-Sep-26, 09:17 PM
I am conrfused as to what the problem is thought to be with Chandra's calculations. These are very high end math calculations true, but they've been around for 80 or so years now and I am not aware of any challenge to them. Would someone be so kind as to enlighten me?

antoniseb
2006-Sep-26, 09:26 PM
Would someone be so kind as to enlighten me?

My understanding is that Type 1a supernovae mostly depend on the Chandrasekhar limit. As near as I can tell VanderL is saying that this observation puts that limit in doubt somehow, not that it changes some aspect of the science, but rather that it may invalidate so much of the science that you can't point to one specific broken link in the chain.

As you know VanderL is one of our longest term supporters of alternative theories... not one in particular, but the idea that alternatives need to be explored. But as we know, I am mostly a supporter of the mainstream science.

My position is that the two-white-dwarf explanation is likely and explains everything here.

Duane
2006-Sep-26, 09:33 PM
Yes I agree with you antoniseb. The merger of two white dwarf stars would explain everything they saw here.

Ihopeso
2006-Sep-27, 12:01 AM
Yes I agree with you antoniseb. The merger of two white dwarf stars would explain everything they saw here.

>grunts, murmurs a lot of Mmmmm’s a nod and a wink from the gallery<

And it's settled then two white dwarfs it was, doing the death spiral.

So in other words this finding has little to bear on the standard candle model? :)

PS
become white dwarfs, and then radiate enough orbital energy through gravity waves that they can merge What's a gravity wave?

antoniseb
2006-Sep-27, 11:54 AM
What's a gravity wave?

Take a look at this (http://en.wikipedia.org/wiki/Gravitational_waves) Wiki article for a nice overview. In this case, it is important because it allows a pair of white dwarfs orbiting each other to lose enough orbital energy to eventually merge.

Ihopeso
2006-Sep-27, 12:39 PM
Take a look at this Wiki article for a nice overview. In this case, it is important because it allows a pair of white dwarfs orbiting each other to lose enough orbital energy to eventually merge.

thanks for the link antoniseb, but I'm still confused as the Wiki entry said they had not yet detected them. So two white darwfs spiraling into each other might just be a long shot?

antoniseb
2006-Sep-27, 01:53 PM
thanks for the link antoniseb, but I'm still confused as the Wiki entry said they had not yet detected them. So two white darwfs spiraling into each other might just be a long shot?

We haven't directly detected them, but the two orbiting pulsars have demonstrated that *something* is taking energy out of their orbits, and the amount that is being taken out agrees with it being gravitational waves within the limit of our ability to detect any difference.

Cougar
2006-Sep-27, 03:01 PM
i love it! anything that puts into doubt peoples' certainty that that the age and size of the universe are known always makes me happy.
Why?

Cougar
2006-Sep-27, 03:19 PM
I think being forced to reconsider the Chandrasekhar limit.... could mean exactly what I said.
I think you are the only one on the planet that thinks this finding might force reconsideration of the Chandrasekhar limit.

But now a supernova has been discovered that's twice as bright as all the other Type 1a supernovae.
It should be pointed out that the 1998 finding by the two supernova search teams that the expansion is accelerating relied on the observation that distant Type Ia supernovae reach maximum brightnesses approximately 25 percent fainter than the peak brightnesses they would attain if the expansion was not accelerating. It seems clear that this is why the reports are saying that this new finding does not invalidate that previous finding.

ngeo
2006-Sep-27, 03:44 PM
Antoniseb wrote that in an evolved galaxy, “There should be no O, B, or A stars, and an increasingly smaller supply of F stars.” Isn’t the Milky Way an evolved galaxy? According to wikipedia, such stars exist in the Milky Way galaxy, for example in the Orion Nebula.

I think there is some confusion (maybe mine too since I am way out of my depth) about “young stellar populations” versus “young star forming galaxies“. From the paper linked to by Antoniseb:

http://www.arxiv.org/PS_cache/astro-ph/pdf/0609/0609616.pdf

“Super-Chandrasekhar mass SNe Ia should bemore likely in a young stellar population, where
the most massive stars exist. The low mass, star forming host of SNLS-03D3bb is consistent
with this scenario (see SI). Thus, the apparent existence of super-Chandrasekhar mass SNe Ia may
explain why the most luminous SNe Ia only occur in young stellar environments. The standard
Chandrasekhar mass model offers no explanation for this behaviour, since the total amount of fuel
and triggering mechanism should be independent of the mass of the progenitor stars.
SNe such as SNLS-03D3bb will have to be screened out in cosmological studies. Since
younger stellar environments produce more luminous SNe, as the mean stellar age decreases with
redshift the mean properties of SNe Ia will change5. This can be calibrated if all SNe obey the
same stretch-luminosity relationship, but SNLS-03D3bb does not. Its peculiarity was so obvious
that it was excluded from the SNLS cosmological result, but less extreme objects could lurk in SN
samples. Future cosmology studies will have to carefully scrutinise SNe Ia from young populations
to see if they obey the same lightcurve shape-luminosity relationship as other SNe Ia.


Apart from that however, is the current understanding of the processes leading to a supernova well enough understood that data suggesting cosmic acceleration is beyond question? It seems to me the standard candle hypothesis is shaky if there are two kinds of supernovas, and perhaps not as rare as is assumed. It also seems the mass accretion scenario may be more variable than presumed, if rotation can influence the limit.

If science is supposed to be skeptical, at what point does the mainstream become skeptical of its own presumptions as observational power increases? (Maybe if things move faster now it will be quicker than 400 years Pete?)

ngeo
2006-Sep-27, 03:57 PM
Re Cougar's comment above, Cougar are you in a position to say that no super-Chandra mass supernovas are in the sample of the 1998 finding?

Also you have not commented yet on the Middleditch paper except to say he is in a minority.

Cougar
2006-Sep-27, 04:22 PM
....is the current understanding of the processes leading to a supernova well enough understood that data suggesting cosmic acceleration is beyond question?
I would say that nothing in science is ever beyond question.


It seems to me the standard candle hypothesis is shaky if there are two kinds of supernovas, and perhaps not as rare as is assumed. Well, Hubble mis-estimated the age of the universe by a factor of about 6 because it was not then known that there were two kinds of Cepheid variables, which were taken to be standard candles (and still are, but more accurately). The history of the development of Sne Ia as standard candles shows that this is not a simple matter. First of all, it must be determined that the supernova is a Ia, which depends on the presence of certain elements in its spectrum. Then there must be normalization adjustments made which depend on the shape of the light curve. I forget what else....


If science is supposed to be skeptical, at what point does the mainstream become skeptical of its own presumptions....
This occurs at t=0. As I recall, Alan Guth spent over a year trying to debunk his own theory of inflation before he submitted it for publication. Scientists know that once they publish, they are like ducks in a shooting gallery. You would be surprised at the extent that most scientists check their work, their observations, their assumptions, etc. This is why you don't see too many scientific papers that are overly sensationalistic.

Cougar
2006-Sep-27, 04:38 PM
Cougar are you in a position to say that no super-Chandra mass supernovas are in the sample of the 1998 finding?
Ha, no, I am certainly not in such a position. However, if there were any such super-Chandra mass supernovas in their sample, I believe that would mean that the universe is accelerating even faster than the supernova search teams reported. More likely, if they ran into one of these too-bright Sne's, they probably would have thrown it out of their sample because it didn't fit the "profile" of the Sne's they were using as standard candles. Come to think of it, it's very doubtful they ran into one of these new types, otherwise it would have stuck out like a sore thumb, and they would have made the discovery of this new type then rather than this international team making the discovery now.


Also you have not commented yet on the Middleditch paper except to say he is in a minority.
Yes, I'll have to review that more closely. Where should any comments be made?

Tim Thompson
2006-Sep-27, 07:37 PM
Cougar are you in a position to say that no super-Chandra mass supernovas are in the sample of the 1998 finding?

Ha, no, I am certainly not in such a position. ...
Cougar might not be in such a position, but I am. No, there were no super-chandra mass supernovae in the 1998 sample. In fact, there have not ever been any such supernovae in any of the samples used for cosmology, and there never will be. The reason is that the super-chandra supernovae are extreme outliers, they stick out like the proverbial sore thumb. The supernovae light curves used in these studies are fit to standard templates built from a library of observed light curves. The outliers are rejected before any cosmology is done, to make sure the sample is clean. This process unquestionably & always rejects anything that is even remotely like the SN reported in Fraser's original article.

The Real Deal is more subtle than that. This kind of super-chandra even is not & never will be a problem for cosmology because it is too obvious. However, what if there is a slightly-super-chandra SN, one that is only a little bit more massive than the norm? That's the kind of SN that can contaminate the database used for cosmology. That kind of contamination will not negate the conclusions of the cosmological studies, and will not negate the evidence for accelerated expansion. But it will make the error bars & uncertainties larger, or perhaps introduce a systematic bias. The point of the authors of the original study was that this is the kind of SN we have to look for & weed out.


The type Ia supernova SNLS-03D3bb from a super-Chandrasekhar-mass white dwarf star (http://cul.arxiv.org/abs/astro-ph/0609616), Howell, et al., Nature 443, p. 308, 21 Sep 2006

Cougar
2006-Sep-27, 08:37 PM
Well, Middleditch's paper On the Absence of Cosmic Acceleration is a fairly well-written paper. I don't know if it has been peer reviewed, so I don't know if someone who really knows the current state of the art in the field would throw it out because of factual or conceptual errors.

I note that Middleditch is a member of the Modeling, Algorithms, & Informatics Department of the Computer, Computational, and Statistical Sciences Division at Los Alamos National Laboratory, and I can find no record of his specific academic record, so I have a little problem with someone who is not a specialist in astrophysics telling the entire community of astrophysicists who are specialists that they have got it all wrong.

Middleditch has apparently studied Supernova 1987A in some depth, and he references this supernova often in this paper and, if I understand correctly, he attempts to carry over certain characteristics from 1987A onto all supernovas, including Type Ia's. However, Sne 1987A is a Type II, and the specialists will tell you that the mechanism causing Type IIs is drastically different than the mechnism for Type Ia's. They may not have all the details figured out, but I think they would tell you that they are not completely clueless about these things either.

There is also this little bit at the end of the paper where Middleditch says, "I would also like to thank Jerry Jensen for conversations and bringing this issue to my attention." Yes, that's our Jerry, :o whose reputation here at BAUT is not the most.... untarnished. But I mention this as more of a joke, because it is certainly no valid criticism.

But basically, as to the details, I don't think I'm really qualified to give this paper a proper, in-depth review. Middleditch claims Sne Ia's are double-degenerate merger-induced core-collapse Sne's of common envelope Wolf-Rayet stars viewed at different angles. I am, of course, skeptical, but offhand I don't really have the tools to contradict such a claim. A specialist may be able to say that's ridiculous and it cannot be, but I cannot.

The accelerating expansion is not a "paradigm" that I am married to, but being just an amateur, I will have to hear a consensus of specialists agree that there really is no acceleration before I seriously accept that position. One paper by a computer geek* at Los Alamos does not convince me that a significant paradigm shift is called for.

* This is intended to be descriptive, not derogatory!

RussT
2006-Sep-27, 09:11 PM
This is directly from the Middleditch abstract.



By now it is also clear that an extreme version of the axisymmetry seen in SN 1987A is the correct paradigm for SNe Ia and Ic. A Ia/c is produced from the merger of two degenerate cores of common envelope WR stars, or of two CO white dwarfs. Its polar blowouts produce the observed high velocity and polarized spectral features in Ia's, and its equatorial bulge is much brighter in Ia's, due to the greater fraction of 56 Ni contained within it. These become classified as Ia's when viewed from the merger equator, and Ic's when viewed from the poles.


[A Ia/c is produced from the merger of two degenerate cores of common envelope WR stars, or of two CO white dwarfs.]

Tim, isn't he totally missclasifying 1c's here?

VanderL
2006-Sep-27, 09:54 PM
I think you are the only one on the planet that thinks this finding might force reconsideration of the Chandrasekhar limit.


The point I was trying to make is nicely illustrated by Tim's post, the need for a clean sample of SN1a's, or standard candles means we discard the "rotten" cherries and only pick the ones that look nice. If we discard too many, we're exactly in the process of building a house of cards. This new supernova doesn't fit the mold, so we discard it and label it as a merger of white dwarfs, without any proof, because it "seems likely".

If I read Middleditch's paper correctly, there should be many hypernovae 1c, but they are being misclassified as 1a's. What about the other SNe, how many get discarded because they also don't fit?

Cheers.

antoniseb
2006-Sep-27, 10:07 PM
The point I was trying to make is nicely illustrated by Tim's post, the need for a clean sample of SN1a's, or standard candles means we discard the "rotten" cherries and only pick the ones that look nice. If we discard too many, we're exactly in the process of building a house of cards.

How many is too many. Here we are talking about discarding one in a thousand.

Concerning you reading Middleditch's paper correctly, it is possible that you are reading what he intended to write, but still not getting the mainstream view of sne1a's. He does disagree with most astronomers about the nature of these things.

Cougar
2006-Sep-27, 10:30 PM
Middleditch claims Sne Ia's are double-degenerate merger-induced core-collapse Sne's of common envelope Wolf-Rayet stars viewed at different angles....
Coincidentally, this is similar to the mechanism for Sne 1987A, the supernova that Middleditch has studied in depth. This Oxford Astrophysics Group site (http://www-astro.physics.ox.ac.uk/~sjustham/supernovae.html) has a good summary of fairly current research in these matters, at least at Oxford. As they say with respect to 1987A:

At the moment, these [various chemical] anomalies can be best explained if the progenitor originally was in a binary and has merged with its companion in the not-too-distant past.

Cougar
2006-Sep-27, 10:44 PM
....we discard the "rotten" cherries and only pick the ones that look nice.... we're exactly in the process of building a house of cards.
I don't think you get it. Say we want to find the average height of Australian men between the ages of 30 and 55. Now suppose we find out that a few people in our sample are women! Naturally we would drop them out of the sample. But you make this sound like it is scientifically suspect and a conspiracy of the mainstream to retain their funding! It's a silly position.

Ihopeso
2006-Sep-28, 01:29 AM
I don't think you get it. Say we want to find the average height of Australian men between the ages of 30 and 55. Now suppose we find out that a few people in our sample are women! Naturally we would drop them out of the sample. But you make this sound like it is scientifically suspect and a conspiracy of the mainstream to retain their funding! It's a silly position.

Funny is'nt, I'd have prolly asked why/how there were women in the sample. :whistle:

VanderL
2006-Sep-28, 09:09 AM
How many is too many. Here we are talking about discarding one in a thousand.

Concerning you reading Middleditch's paper correctly, it is possible that you are reading what he intended to write, but still not getting the mainstream view of sne1a's. He does disagree with most astronomers about the nature of these things.

The problem is in classifying them, many 1c's are spectrally close to 1a's, I'm not sure that the mainstream view of SNe's correctly classifies them.

Cheers.

VanderL
2006-Sep-28, 09:15 AM
I don't think you get it. Say we want to find the average height of Australian men between the ages of 30 and 55. Now suppose we find out that a few people in our sample are women! Naturally we would drop them out of the sample. But you make this sound like it is scientifically suspect and a conspiracy of the mainstream to retain their funding! It's a silly position.

No, the problem is not silly, you are not getting the point I try to make. It is scientifically unsound to discard SNe if we are not even sure that they are of a single class. And I never claimed anything about conspiracies of mainstreamers to retain funding, that's a strawman debating tactic.

Cheers.

antoniseb
2006-Sep-28, 11:38 AM
many 1c's are spectrally close to 1a's, I'm not sure that the mainstream view of SNe's correctly classifies them.

Can you point me to a link that shows that they are so similar that there is some risk (and no easy test) that 1c and 1a SNe can be mistaken for each other by people studying 1a's. My understanding is that this is completely not an issue right now.

Tim Thompson
2006-Sep-28, 02:50 PM
The problem is in classifying them, many 1c's are spectrally close to 1a's, I'm not sure that the mainstream view of SNe's correctly classifies them.
To the best of my knowledge, that is incorrect. Type Ic SNe do not exhibit silicon emission lines, whereas Type Ia SNe do so conspicuously. Do you have an example of a Type Ic SN with Si emission lines? Or perhaps a Type Ia without? I don't think it is so easy to confuse them.

Tim Thompson
2006-Sep-28, 02:57 PM
This is directly from the Middleditch abstract. [A Ia/c is produced from the merger of two degenerate cores of common envelope WR stars, or of two CO white dwarfs.] Tim, isn't he totally missclasifying 1c's here?
Well, he is certainly misclassifying Ia's, which have nothing to do with core mergers or core collapse at all. He may by screwing up class Ic as well, I am unsure. I know that Ic are core collapse events, but I don't know if they are specifically tied to merger events.

ngeo
2006-Sep-28, 10:22 PM
From the abstract contained in the Middleditch arXiv paper itself (differing somewhat from the abstract I quoted above):

“The unexpected faintness of distant Type Ia Supernovae (SNe) has been used by many to argue for an accelerated expansion of the universe. However, observations of nearby SNe show high velocity and polarized features in SNe Ia,
indicating that the single degenerate paradigm for many Type Ia SNe is drastically and catastrophically invalid. By now it appears that an extreme version of the axisymmetry seen in SN 1987A is the correct paradigm for many SNe Ia and
Ic, and that these are both core-collapse and thermonuclear objects, which leave weakly magnetized, rapidly spinning (~2 ms) pulsar remnants. In this paradigm, a Ia/c is produced from the merger of two degenerate cores of common envelope Wolf-Rayet stars, or of two CO white dwarfs. Thus the same explosive mechanism that underlay 10–15 M. in SN 1987A, underlies only 0–1.5 M. in Sne Ia/c. Its now visible polar blowout features produce the observed high velocity and polarized spectral features in Ia’s, and its equatorial bulge is much brighter in Ia’s, due to the greater fraction of 56Ni contained within it. Such merger Sne become classified as Ia’s when viewed from the merger equator, and sometimes Ic’s when viewed from the poles, where a hypernova signature and a gamma-ray burst will be observed for lines of sight close to the merger axis. Thus cosmology determined strictly from Ia’s alone is flawed at its very foundation: the local sample is selectively biased. The problem may be due to SNe from a class spanning a wide range of lower intrinsic luminosities, previously unaccounted for in the local sample, which were unknowingly included in the distant sample, and the result was distant SNe Ia which appeared to be too faint for their redshifts. When the errors introduced by this and other processes are taken into account, there may be no cosmic acceleration effect in distant Sne.”







From the Introduction:

“In spite of initial wild overestimates for the masses of some SNe Ic progenitors (SN 2002ap, 2003dh – Mazzali et al. 2002, 2003), no Ia or Ic progenitor has ever been identified (see, e.g., Maund, Smartt, & Schweizer 2005), thus they cannot be massive stars. The “usual” Ia paradigm, (gradual) accretion-induced collapse of a white dwarf, or single degenerate (SD), suffers from the absence of H and He which should have been advected from the mass donating companion star. There is also the dificulty in generating more than 0.1 M. of 56Ni from a single ignition source (Brown et al. 2005), though multiple sources have been discussed by Ropke et al. (2006). Finally, in addition to their high velocity and polarized features, Ia’s also show a broad range of diversity in their velocity gradients and correlations
with Si line ratios and (delta) m15(B) (Hachinger, Mazzali, & Benetti 2006; James et al. 2006), all unlikely side effects of simple thermonuclear disruption.”

Further:

“I have argued above that SNe Ia and Ic are the result of the same process, the DD merger-induced core-collapse of common envelope WR stars or CO-CO white dwarfs. I have further argued that many distant Ia’s in a continuous class will have (delta) m15’s within the acceptable limit, but low luminosities wrt the local sample, whether the WL relation is used or not. In addition, the high velocity of the small amount of matter in near-polar ejecta of Ia’s exposes a fraction of the Ia TNB to non-equatorial views during the interval when (delta)m15 is measured. These SNe were also frequently not included in the local sample for a number of reasons, including appearing to be too “Ic-ish.” Thus, yet another systematic bias entered the local sample of SNe in favor of those which dimmed too quickly due to an equatorial view of the DD merger. However the bias for the most distant, and hence low signal-to-noise SNe, is probably not the same, leading to less discriminating inclusion in the sample, skewing the results. A 7% Ibc contamination level is sufficient to produce (Omega lambda) = 0.7 from no effect (Homeier 2005).”

I think the whole paper is worth reading and commenting on by someone who understands it.

Cougar
2006-Sep-29, 05:30 PM
I think the whole paper is worth reading and commenting on by someone who understands it.

Well, I have some understanding of it. Wikipedia is accurate when it says....
"The most commonly accepted theory of [Type Ia] supernovae is that they are the result of a carbon-oxygen white dwarf [a single degenerate object] accreting matter from a nearby companion star, typically a red giant, until it nears the Chandrasekhar limit."
As you have cited, Middleditch bucks the commonly accepted theory and asserts....
a Ia/c is produced from the merger of two degenerate cores of common envelope Wolf-Rayet stars, or of two CO white dwarfs.
I don't really follow his first two supporting arguments. For example, Argument 2 states that if his assertion is correct, then the Polar Blowout Feature (PBF) will have a higher velocity and the ThermoNnuclear Ball (TNB) will be brighter. As I said, I don't follow how this offers his theory any support.

The main problem I see with Middleditch's theory is that Sne Type I's are classified Type I's because they have no Hydrogen Balmer lines, and Type II's are put in that category because they do have Hydrogen Balmer lines. Middleditch briefly mentions the presence and absence of these Hydrogen lines, but I don't think he adequately explains how his theory can result in both the presence AND absence of Hydrogen Balmer lines when (as I understand it) he claims ALL supernovas are either direct core collapse or double degenerate core collapse events, where either way Hydrogen Balmer lines are going to be present.

RussT
2006-Sep-29, 10:43 PM
The other BIG problem he has is saying that 1a's are viewed as bulge collapse and 1c's are viewed from the poles, in other words, beamed directly at us, which are hypernova's, which are GRB's and should not even be considered to be in this classification.

AFAIK, Hypernovas are in the 1c classification because they appear not to have any hydrogen or helium lines in their spectra, just like a white dwarf that has lost all of it's H, He, however, this could be a misconception on my part, in which case, some clarification of why Hypernova's are classified as 1c's would certainly be helpful.

Jerry
2006-Nov-06, 06:43 PM
This single event is vindictive of the assertions made by myself, Middleditch, and others, that using dark energy expansion derived from supernova Ia magnitude estimates is wishful thinking, not objective science.

http://arxiv.org/PS_cache/astro-ph/pdf/0611/0611125.pdf


It is a major concern for the use of SNeIa as cosmological probes that we do not yet have direct evidence of the nature of their progenitor systems.

What we do have evidence of, is considerable confusion about the differences between Ia and Ic supernova events. Since Ic are all over the mapm - in terms of magnitude, light curve length, and composition - if Ia are truly a unique and different type of event, there cannot be any Ia that fall out of the normal distribution. Otherwise, there is no justification for segregating Ic from Ia. This event breaks that rule.

If one local Ia is so much brighter, then in a cosmological sample, we should expect to find many brighter supernova. Nugent has argued that this would only make cosmic acceleration more evident. But it could just as easily - in fact, it is more likely - to indicate the attenuation factor derived from supernova used in assessing cosmic distance is too small.

Unless and until this event is understood, and we have a working understanding of supernova progenitor systems, and we can truly discriminate between progenitor types, cosmological parameterization based upon supernova observations is on permenant hold.

Jerry
2006-Nov-08, 02:43 PM
It is very common for a supernova to first be identified as a type Ic, then later identified as a type Ia when certain spectral lines appear. In Benetti et al, they are arguing that a peculiar type Ia, 2002ic, is actually a Ic or IIn. But if the spectral line distribution is so ambiguous, so similar, why is it still acceptable to assume the progenetor, the root mechanism, is completely different?

Given the current crop of supernova show much greater distribution in light curve widths and spectral features than expected, I don't see how the conclusion of 1998, that the rate of expansion of the universe is increasing, can be honestly supported: Whether or not the Middleditch alternative model is correct.

antoniseb
2006-Nov-08, 02:54 PM
I don't see how the conclusion of 1998, that the rate of expansion of the universe is increasing, can be honestly supported
That is misleading word salad. Take a look at the current graphs plotting SN brightness and redshift. There is no question anymore. Regardless of whether you say the cause is acceleration or some strange distribution of CREIL phenomena, the 1998 observation has been refined and confirmed, and in no way overturned.

Jerry
2006-Nov-08, 04:37 PM
Word salad? All of the conclusions drawn from supernova Ia research rely upon supernova Ia falling within a very narrow range of absolute magnitudes. Since the landmark papers, the error bars have been significantly widened by near-local events. Distant nova have been observed with all the signatures of local Ia, except that the light curves are too short, so they are omitted from the 'gold standard' templating.

Finally, we have this example of a local supernova with all the proper spectral signature of a Ia, but much too bright to fit the standard curve upon which the accelerated expansion was based. When Goldhaber and Perlmutter were looking at 'all possible scenarios'; this did not include the possibility that more distant supernova are much brighter than the local sample.

The supernova Hubble curve is based upon the assumption that events like this do not occur, but now we know that they do. How badly contaminated is the distant sample? Who knows? One local event does not give us enough clues. However, we do know the local sample is populated with hypernova type Ic - these events are brighter, and a handful of them have been observed locally, so there should be a statistical sample of these ultra-bright events recorded in the distant universe. They are just not there! So there has either been considerable evolution in the Ic process, or current supernova science underestimates the attenuation factors of space.

It is not intellectually honest to assume type Ic are evolving. Neither is it honest to assume that this local event is extremely rare in the redshifted universe. The spectral signatures of distant events are severely attenuated, not to mention the loss of spectral bandwidth due to redshifting.

The other evidentuary trails of cosmic acceleration (CMB and galaxy surveys), are of necessity, dependent upon the attenuation factors determined by supernova distance scaling. At this time, we do not have a secondary standard at cosmic distances. The only sane conclusion, is that current estimates of the supernova Hubble curve must be studded by a strong caviat.

Cougar
2006-Nov-11, 10:03 PM
This single event is vindictive of the assertions made by myself, Middleditch, and others, that using dark energy expansion derived from supernova Ia magnitude estimates is wishful thinking, not objective science. http://arxiv.org/PS_cache/astro-ph/pdf/0611/0611125.pdf
Of course that should be indicative, not "vindictive", although this slip could conceivably be considered Freudian. :)

But no, a "single event," regardless of how it is interpreted, is not going to overturn the finding of a clear pattern discerned from hundreds of "events."


What we do have evidence of, is considerable confusion about the differences between Ia and Ic supernova events.
Apparently not among those who know what the differences are.


Since Ic are all over the map - in terms of magnitude, light curve length, and composition... This event breaks that rule.
If it's noticeably different than a "normal" Ia, then it shouldn't be used as one of the standard candles in a study.


If one local Ia is so much brighter, then in a cosmological sample, we should expect to find many brighter supernova. Nugent has argued that this would only make cosmic acceleration more evident. But it could just as easily - in fact, it is more likely - to indicate the attenuation factor derived from supernova used in assessing cosmic distance is too small.
I don't see how this makes any sense. The cosmological samples contain Sne Ia's that are dimmer than they should be according to their redshift. But again, an odd outlier that doesn't fit the profile is not going to be considered a helpful indicator.


Unless and until this event is understood, and we have a working understanding of supernova progenitor systems, and we can truly discriminate between progenitor types, cosmological parameterization based upon supernova observations is on permenant hold.
Well, we do have a working understanding of supernova progenitor systems. Certainly research on these phenomena is ongoing, and the full variety of progenitors may not be fully understood, but the situation is not nearly as dire as you make it out to be.

mickal555
2006-Nov-12, 01:31 PM
Wow!

Jerry
2006-Nov-13, 10:30 PM
But no, a "single event," regardless of how it is interpreted, is not going to overturn the finding of a clear pattern discerned from hundreds of "events."

There are times when a single event nullifies all preconceptions to the contrary: The first manned flight, the first laser, first blazer, first high temperature superconductor, and so on. It is axionic to the use of supenova Ia as distance indicators that supernova with the signature of Ia fall within a very narrow magnitude range. One local event that violates this rule nullifies all science based upon this concept: Men can fly; and supernova 'Ia' can and do brighten to twice the previously known expected maximum brilliance. When we look at very distant supernova, we do not know how bright they are.




What we do have evidence of, is considerable confusion about the differences between Ia and Ic supernova events.

Apparently not among those who know what the differences are.

I get emails from the PI's in this area on a daily bases, and even though I cannot reprint them, I can assure you determining whether a supernova is a type 'Ia' or a 'Ic' is not easy. In this http://arxiv.org/PS_cache/astro-ph/pdf/0611/0611125.pdf paper, Bennetti argues a supernova with all the spectral signature of a 'Ia' is actually a 'Ic', but his reasoning is morphological.


If it's noticeably different than a "normal" Ia, then it shouldn't be used as one of the standard candles in a study.

Noticably different in what way? When we look at a very distant supernova, how do we know whether or not it is one of these ultra-bright dudes? That is the problem: We do not. True, this particular supernova had a relatively low velocity expansion envelope, but the reason behind this is unclear as well.



I don't see how this makes any sense. The cosmological samples contain Sne Ia's that are dimmer than they should be according to their redshift. But again, an odd outlier that doesn't fit the profile is not going to be considered a helpful indicator.

No, it is not helpful, any more than it is helpful to assume a killer who does not fit a standard FBI profile is innocent. Sorry: There are no standard killers, and there are no known standard cosmic-scale candles, at this time.



Well, we do have a working understanding of supernova progenitor systems. Certainly research on these phenomena is ongoing, and the full variety of progenitors may not be fully understood, but the situation is not nearly as dire as you make it out to be.
No, we do not - Please read the Bennetti article, where they state exactly the opposite. Also see Middleditch's introduction:

http://arxiv.org/PS_cache/astro-ph/pdf/0608/0608386.pdf



Type Ia supernovae (SNe Ia) have been used by at least two groups, and all without any explicit foreknowledge of their progenitors, to argue that the expansion of the universe is accelerating, and hence for the existence of “dark energy,” or a cosmological constant, This has the appearance of convenience, as it helps several other lines of inquiry, including the scale size of the fluctuations of the surface of last scattering of the cosmic microwave background (CMB), and measurements of the clustering mass on large scales, converge to a consistent set of parameters... However, at present SNe Ia represent the only firm, direct evidence for the existence of dark energy.

If dark energy has been convenient for cosmologists, it is certainly not so for the Standard Model, which allows no such subtle effect by 120 orders of magnitude.

(referenced omitted)

This lack off knowledge of the progenitors has always made supernova science iffy at best. About all we can say now, is they are likely the result of some binary process involving at least one fuel-starved star and many fusion processes. The fun part is, we get to keep looking.

antoniseb
2006-Nov-14, 12:44 PM
Here (http://arxiv.org/abs/astro-ph/0611354) is an interesting paper discussing several unusual supernovae that seem to be type 1a. The paper doesn't support Jerry's assertion that we can't tell which are which, but it does support his point that not all type 1a's are standard candles. It also gives us an idea what fraction of the 1a's can't be used as such.

Jerry
2006-Nov-15, 06:42 PM
Good catch, and here is another subluminous example.

http://arxiv.org/abs/astro-ph/0611295

At what point do the exceptions: Sub and superluminous events with the spectral signature of type 'Ia' become the rule? When do researchers say Gee, we really do not know with a high degree of certainty, that all of the events we have characterized at cosmic distances are 'normal type 'Ia' in fact are such, in this expanding sea of magnitudes? How many subluminous events, (or super luminous, for that matter) would it take contaminating the sample, and to drive false conclusions about the expansion rate, and/or the transparancy of the universe?

RussT
2006-Nov-18, 02:40 AM
This Toatally depends on whether any 1c's wound up being used in the 1998 SN1a data evaluations!

Jerry
2006-Nov-20, 07:46 PM
No, this is also depended upon the assumption that time dilation is a function of redshift: Remember, the only direct proof of this is in supernova studies which assume a constant magnitude/light curve relationship, and that there are no distance biased selection effects in the supernova sample. Since we are now certain that there are local events that violate both of these premise, all bets are off. If the 1998 SNIa data evaluations contained many of this type of supernova; (which should be easier to find at high redshift because they are brighter), they grossly underestimate comic attenuation factors.

antoniseb
2006-Nov-20, 08:03 PM
...all bets are off... they grossly underestimate comic attenuation factors.
Oh those comic attenuation factors. There's nothing funny about *them*.

More seriously, you are making it sound like there's no way to identify this fraction of a percent of the population of SN1a's that behave differently. The SN time dilation is increasingly supported by strong evidence. The SN1a's observed behind the LMC during the microlensing study have really sealed the deal. These were measured accurately every other day, and the light curves match the time-dilation model extremely well.

Jerry
2006-Nov-25, 12:45 AM
Reference please?

tnx

antoniseb
2006-Nov-25, 01:28 AM
Sorry Jerry, I'm having a little trouble tracking it down. I posted a message about it, including a link to the arXiv version of the paper. This was probably a month or two ago.

Essentially, there is a microlensing survey being done on the LMC, looking for dim objects between here and there. This survey has also turned up lots of variable stars, eclipsing binaries, and distant supernovae, including 22 type 1a's if I remember correctly, on the basis of one year of observing.

Jerry
2006-Nov-27, 06:18 PM
http://arxiv.org/abs/astro-ph/0509240

There has been an active debate on microlensing studies for almost a decade now:


Characterizing the nature and spatial distribution of the lensing objects that produce the previously measured microlensing optical depth toward the Large Magellanic Cloud (LMC) remains an open problem.

Using microlensing to resolve supernova issues, or visa versa, assumes there are no uncorrected errors, no incorrect assumptions, in one form of analysis or the other...or both.

One of the difficulties is determining whether a supernova that appears to be much brighter than it should, is microlensed or not. Since we now have a very local sample of a non-microlensed type Ia that is much too bright, magnitude limits used in interpreting micolensing limits are also suspect.

antoniseb
2006-Nov-27, 06:38 PM
http://arxiv.org/abs/astro-ph/0509240
Since we now have a very local sample of a non-microlensed type Ia that is much too bright, magnitude limits used in interpreting micolensing limits are also suspect.

Yes, it was a paper on SuperMACHO, but it wasn't this paper (The one I'm recalling was from this summer/fall). It is important to know that the type 1a supernovae in this paper were NOT microlensed. It is simply the case that the SuperMACHO study is looking for small variations in object brightness, and has discovered both microlensing events, and other transient events such as eclipsing binaries, Cepheids, novae, and background supernovae of all sorts. Lensing doesn't come into this as far as the supernovae are concerned. The study simply discovered them very pre-maximum, and has excellent light-curves for them. In most (or maybe all) of these, spectra were obtained using larger instruments to verify the type of SN.

antoniseb
2006-Nov-27, 06:51 PM
Here (http://arxiv.org/abs/astro-ph/0608639) it is.

Take a look Jerry. It is pretty solid, and more such SN will be seen each year for the next few years as this survey progresses.

Jerry
2006-Nov-28, 06:21 PM
Thanks - I know what a pain it is to chase down a lost resource. This paper is consistent with many of the trends in supernova data analysis that are fundamentally flawed: We cannot assume that an event observed locally is rare or exceptional in the extended universe.


Intrinsic diversity in the SN Ia family will also impact our estimate of the time-to-maximum from our composite lightcurve. To reduce the most gross impact of this effect, we remove the obvious fast riser and decliner SM-2004-LMC-1060 from our composite lightcurve.

The sample is redshifted, and corrected for time dilation. If the time dilation correction is wrong, this may not a fast riser, but a 'fairly normal' riser and the balance of the sample are 'slow risers'. This is not as outlandish as it sounds: Longer light curves are usually the norm for brighter supernova, as they concede in the next paragraph:


To account for the effect of variation between the remaining SNe, we add a free “stretch” parameter, s, for each of the SNe in the fit following Goldhaber et al. (2001).
Using the stretch factor requires four assumptions: 1) A single parameter can be used to adjust light curve magnitude to width: there is a strong correlation between lightcurve width and magnitude. 2) The redshift time dilation factor is correct. 3) There is no distance selection effect. 4) All of the supernova in this distance sample are similar to those identified as the local norm.

Think about this: We have a local supernova that is twice as bright as the local sample with a very similar light curve. When we look further into the past, we are more likely to observe brighter events, therefore in this more distant sample, we should be more likely to spot one or more of these brighter events. How badly skewed is the distant sample by these brighter events? No one knows.


Using C-MINUIT we perform a multiparameter fit to the composite lightcurve and fix the stretch parameter for one of the SNe in the composite to 1, no stretch. Effectively, the other SNe in the composite are normalized to the shape of the unstretched SN. We choose SM-2004-LMC-944 as our fiducial SNe, because it has the median width of the 3 SNe in the
composite.
But is this median event more like the brighter or dimmer local type Ia with the same size light curve? And in either case, is the redshift time dilation correction appropriate? All of our distant supernova science is based upon the assumption that ultra-bright supernova type Ia events do not occur in the distant universe; or we have not observed them, even though we know that they should, and we should.

There are other problems inherent in normalizing a sample before making comparisons: Any trends in the sample that run opposite each other are cancelled out and lost in the normalization process. A better approach would be to analyse the rise times before applying the stretch factors: Is there a trend towards longer and slower rise times? There should be, if, as they have assumed, brighter supernova, on average, exhibit longer light curves. Normal distance selection effects should favor our observation of brigher events with increasing distance. (This is what bit Hubble, in his analysis of Cephied variables, and created a distance modulus that was much too large.)

If we cannot look at the distant sample and identify a trend towards brighter supernova that we know should exist, something is wrong with our processing of the data. In the Goldhaber paper of 2001, they emphasize the fact that the normalized data contains no evidence of distance magnitude bias. That is odd, because the obvious conclusion is that the normalizing process allows some systemic bias to cancel out distance magnitude trends that we know we should see. We should not have confidence in a methodology that violates this fundamental rule.

antoniseb
2006-Nov-28, 06:49 PM
Think about this: We have a local supernova that is twice as bright as the local sample with a very similar light curve. When we look further into the past, we are more likely to observe brighter events, therefore in this more distant sample, we should be more likely to spot one or more of these brighter events. How badly skewed is the distant sample by these brighter events? No one knows.

Which local overbright supernova are you referring to? Can you show its lightcurve compared to the *very* similar light curves in this study, and still tell me it doesn't stand out?

Jerry
2006-Nov-29, 03:02 AM
???

The supernova featured in this article...which I have assumed is fairly local
(yes/no?) If not, the argument still holds: This supernova must be at least twice as bright as the local standard, how can we assume there is not a broad continuum of supernova brilliance that is independent of the light curve width?


Some variations have been observed in Type Ia supernovae, but these are mostly reconcilable. Brighter Type Ia’s take longer to rise to maximum brightness and longer to decline. When the time-scales of individual light curves are stretched to fit the norm, and brightness is scaled according to the stretch, Type Ia light curves match.

Brightness differences could be due to differing ratios of carbon and oxygen in the progenitors, resulting in differing final amounts of nickel in the explosion. The radioactive decay of nickel to cobalt and then iron powers the optical and near-infrared light curves of Type Ia supernovae. Differences in apparent brightness could also be products of asymmetry; an explosion viewed from one angle may be slightly dimmer than from another.

None of these possible differences are enough to explain supernova SNLS-03D3bb’s extreme brightness — which is much too bright for its light-curve “stretch.”

What is different, is not the light curve, but a lower-than-average expansion velocity. But think what this means, when we observe a supernova in redshifted space if the assumed time dilation factor is wrong: The correction for time dilation will speed up the expansion and reduce the lightcurve width, underestimating the magitude of the supernova. If this is happening to any degree, (and the lack of an observable distance/magnitude relationship indicates to me that it is), attenuation factors are grossly under-estimated.

Poor attenuation estimates might help explain why we see so many gamma rays that we expect to be associated with supernovae (like 1987A) occurring in places where no supernovae are found.

Michael Noonan
2006-Nov-29, 03:05 PM
How much of an effect could there be to the light if it had to enter the gravity lines at a certain trajectory for it to be able to be seen. If the galaxy is in a hyper light transition wouldn't the light have to conform to the medium it joined or if it could enter from any angle how far bent would it get before we registered it from the direction we received it from.

Second thought if the view we have of the temperature variance in the background cosmic radiation is the same in all directions could that be more to do with our own galaxy acting as less than a perfect lens. If that was corrected for shape would those tiny variations if the inverse of the pattern of heat was applied tell us more about our galaxy than the universe as a whole. Where the agreed basic temperature should give more of an indication of the state of the universe when corrected for heat transfer due to the galaxy being an open system to the universe.

Cheers

antoniseb
2006-Nov-29, 04:07 PM
1. How much of an effect could there be to the light if it had to enter the gravity lines at a certain trajectory for it to be able to be seen.
2. If the galaxy is in a hyper light transition wouldn't the light have to conform to the medium it joined or if it could enter from any angle how far bent would it get before we registered it from the direction we received it from.

1. Are you referring to the eleven supernova discussed in the SuperMACHO paper linked above? I'm guessing not. I think you're just trying out astro-grummelot.
2. More evidence that this is just grummelot.

Michael Noonan
2006-Nov-29, 04:09 PM
Forgive me I am probably just nit picking at a small thought that has already been considered.
The idea was that I was sitting in a very twisted glass bubble that had very minor imperfections in it that could be corrected by something like reverse engineering shadows like a graphics card might so that the initial math would already exist if all the points of light and its trajectory were known.

Also I heard that more raw computing power is devoted to animating characters for the film industry than is reserved for most scientific calculations, if that could help animate the real position of what we receive it may give better understanding to how to interpret the whole picture of the universe.

These sort of minor optic corrections most likely have already been taken into account and so it was probably just an idle thought.

Cheers

antoniseb
2006-Nov-29, 04:38 PM
How much are you imagining that the path of light is bent on its way to us? Even in the most extreme cases, the light is not bent more than half a minute of arc. In the cases we're looking at in this paper it is doubtful that they are more than a few milli-arcseconds away from their 'true' direction. Is there something you think can be gained by trying to sort that out?

Hamilcar
2006-Nov-29, 04:54 PM
Astronomers used to believe that all Type 1a supernovae were essentially the same brightness. That's because they explode with the same amount of fuel. But now a supernova has been discovered that's twice as bright as all the other Type 1a supernovae.

1991bg (underluminous) and 1991T (overluminous) Type Ia SNe have been known since, well, 1991. The sometimes wilful ignorance of the people doing cosmology with supernovae regarding stelalr evolution and death is sometimes staggering.

Note e.g. this paper:
http://arxiv.org/abs/astro-ph/0608324

Jerry
2006-Dec-18, 07:11 PM
http://arxiv.org/PS_cache/astro-ph/pdf/0612/0612408.pdf


With a peak absolute magnitude of V ≈ −22.2, SN 2006gy is probably the brightest SN ever observed. The gradual brightening, the peak brightness and the presence of Si II and Ha lines lead us to suggest that SN 2006gy may be a hybrid IIn/Ia SN (also known as type-IIa)...


This is a stunning development. At high redshift with limited spectral information, a IIn/Ia would be extremely difficult to discern from a typical Ia. Even more important, a supernova this bright would be much more likely to be detected than your run-of-the-mill -19.4 magnitude SNIa, so we would be much more likely to find this type of event in our search for the most distant supernovae.



Thus, if indeed SN 2006gy is a Ia event then the energetics may require a super-Chandrasekhar-mass white dwarf progenitor (e.g., Howell et al. 2006) or a double degenerate merger progenitor.

Which is in pretty darn close agreement with what Middleditch has proposed. Every time one of these ultra-bright events occurs locally, the probability that the high redshift sample is over-represented by these events increases exponentially - completely throwing cosmology based upon a standard light curve magnitudes out-the-window, dark energy, lock, stock and barrel.