PDA

View Full Version : Type 1a supernova...slow to decline but not overluminous in the optical



trinitree88
2011-Jun-22, 02:37 PM
Everybody likes a good mystery story. Type 1a's that decline extremely slowly are thought to be overluminous in the optical, but not necessarily so. It's possible the effect is due to gamma trapping in the interior layers of the ejecta. One has to remember, this is not a homogeneous population of progenitors, considering the metallicities of their environments SEE:http://arxiv.org/PS_cache/arxiv/pdf/1106/1106.3968v2.pdf

Jerry
2011-Jun-28, 03:33 AM
Another new paper, another new supernova data reduction pipeline.

Ok. A slooow decliner. They exist; this one is pretty local, fairly average velocity. Fine structure uncovers some spectral differences between this one and other local events, but what about more distant events?

In the mid-1990's Leibengut and others use the decline rate of distant supernova as the first proof positive of relativistic time dilation in distant supernova events. Now we have the first proof-positive that a garden variety local event can also be a sloooow decliner which cannot be evidence of time dilation. Egro the noisy spectral signatures of distant supernova events can no longer be considered proof positive of time dilation.

It gets better.

There are two selection biases at play here: natural (type II) Malmquist bias, which is the propensity to find brighter objects with increasing distance; and the fact that many many supernova searchers looking for time dilated events have scheduled searchs so that they would be more likely to see slow decliners in the past; even to the point of rejecting events from further study that do not have a sufficiently long light curve to be time dilated events.

No matter how you bake it; slow decliners in local space nullify prior claims that more distant events clearly demonstrate time dilation. Nobody knows what is happening out there: Slooow local decliners mean Slooow distant decliners don't prove time dilation. We would have to find a distant sloooooooooooW decliner; and no one has seen one yet. This is soooooo true that it is no longer a tenable argument to state otherwise. If anything, the data in hand suggest the theory of relativist expansion is wrong.

Shaula
2011-Jun-28, 05:39 AM
Can you present your evidence that shows that researchers reject 'normal' duration supernovae at greater distances? What relevance does the brightness have on the length of time the light curve decays over? You seem to be taking a very, very large leap from one data point. Almost like you are applying a powerful selection bias to this piece of evidence because it supports a pre-existing ATM point of view...

Tensor
2011-Jun-28, 05:51 AM
Another new paper,

snip...

We would have to find a distant sloooooooooooW decliner; and no one has seen one yet. This is soooooo true that it is no longer a tenable argument to state otherwise. If anything, the data in hand suggest the theory of relativist expansion is wrong.

Jerry, how about instead of just firing shots at any old paper that shows up, you actually produce a paper that can show, with all the math, all the assumptions you'll be making, all the observations, etc, exactly what is it you're claiming? It's easy to take pot shots, much harder to actually produce something that would make some sense. Why is it that you continue to take pot shots, but can't produce some actual conclusions that have some actual observations to back up the conclusions? You can't even honestly refute the papers you comment on. Just some wild claims, with no actual evidence. Just your interpretations, which we know are biased and from previous history here, without actual support.

Jerry
2011-Jun-29, 04:17 AM
Can you present your evidence that shows that researchers reject 'normal' duration supernovae at greater distances? What relevance does the brightness have on the length of time the light curve decays over? You seem to be taking a very, very large leap from one data point. Almost like you are applying a powerful selection bias to this piece of evidence because it supports a pre-existing ATM point of view...

There have been a half dozen or so events where supernovae with a basic Ia signature demonstrate extended light curves in fairly local space. This is the first one I am aware of that decays very slowly and has (as near as can be estimated) an average absolute magnitude.

What is missing in the spectrum of supernova events, is a very distant event of an average or greater magnitude with a light curve that matches the slow local light curve length of these local events. That is what is missing from the data base.

The search algorythms used for very high redshift events rule-out the detection of supernova that, after the correction for time dilation, would appear very small. There is one paper I am aware( of (Reis ~2004) where they specifically state follow-up study of a ultra-short very distant event was curtailed.

It only takes a single local event with a very long light curve to nullify the supposition that all distant events have time dilated light curves-just as it only takes a single close pass of a Boeing 747 to prove all jet planes are not the size of 737s.


On the other hand, if the assumption that distant events are time dilated is false, as the number of distant events observed increases, it will become painfully obvious that after correction for time dilation, the light curve appears to be shorter. This is already true for gamma rays,.

Shaula
2011-Jun-29, 05:47 AM
It only takes a single local event with a very long light curve to nullify the supposition that all distant events have time dilated light curves-just as it only takes a single close pass of a Boeing 747 to prove all jet planes are not the size of 737s.
But the proposition is not that ALL the distant events have a time dilated light curve. It is that on average these supernovae in the past have a longer light curve. That is to say you take the length of the light curve and plot against distance for supernovae with reasonably well determined distances. You see that there is a correlation between distance and length of curve. Sure there is scatter and the odd outlier but to invalidate the idea then this correlation would have to be shown to be statistically insignificant. Can you do that? Or are you just flinging mud and hoping enough sticks?

BTW the distant slow events are probably there but hidden in the noise. And now people know they exist they can go looking for them.

And similarly the rejection of one ultra-short event doesn't mean that short events are rejected. Note the phrase ultra-short there. Not just short but ultra-short.

Source for your assertion on gamma rays?

As I said before a sober approach would be to see this as a single data point with all the uncertainties associated with that. It proves nothing beyond the fact that long duration events can occur. We need more data to establish their frequency and any defining features. Then if discrepancies pop up between the distribution of supernovae lengths now and far away (compared to the current model) we need to investigate. But so far the evidence is strong - far too strong to be demolished by one event.

Jerry
2011-Jun-30, 04:48 AM
But the proposition is not that ALL the distant events have a time dilated light curve. It is that on average these supernovae in the past have a longer light curve. That is to say you take the length of the light curve and plot against distance for supernovae with reasonably well determined distances. You see that there is a correlation between distance and length of curve. Sure there is scatter and the odd outlier but to invalidate the idea then this correlation would have to be shown to be statistically insignificant. Can you do that? Or are you just flinging mud and hoping enough sticks?
There is a significant number of overly-bright very long light-curve events in fairly local space. These are generally classified as "Type Ic". (Although in a recent, professional challenge, no one could correctly split synthetic 'type Ia' and 'type Ic' spectra when they are are aged to cosmological distances.)

So there has been a significant number of long light-curve events in local space; but to the best of my knowledge, there have been ZERO very long light curves found in deeply redshifted space. This includes both 'Type Ia' and 'Type Ic' events. At what point does this become statistically significant? From the first evidence that light-curves can age very slowly in local space, there should have been huge caution flags raised: Blanket statements saying slowly evolving light curves (or spectra) in distant events are clear evidence of time dilation completely ignore the possibility that distant events are more like the slowly evolving local events - which are generally, also the brightest.



BTW the distant slow events are probably there but hidden in the noise. And now people know they exist they can go looking for them.
They have always been looking for them. We have very bright local events that take 30 days to lose one magnitude in blue. A similar event at a red shift of one should take 60 days to lose one magnitude, if time dilation is a physical reality. No one should be comfortable with time dilation until such an event is found.

And similarly the rejection of one ultra-short event doesn't mean that short events are rejected. Note the phrase ultra-short there. Not just short but ultra-short.
Remember there is a pretty good correlation between light-curve width and absolute magnitude. Brighter events have longer light curves.

The most curious statistic to date is that the distribution of supernova types found at high redshift is so much tighter than the local distribution. As currently interpreted, both over-bright slow, and fast, low magnitude events are missing. The fast, lower magnitude events may be missed due to selection, but the over-bright slow evolvers shouldn't all be hiding.


Source for your assertion on gamma rays?I think it was a paper by Friel, but let me dig on that one for a few days. The most distant events are super-short, they are even shorter if you remove the time-dilation factor; which pretty much proves there is an inverse correlation between UHEGR luminosity and curve length...at least within the admittedly small sample set we have today.


As I said before a sober approach would be to see this as a single data point with all the uncertainties associated with that. It proves nothing beyond the fact that long duration events can occur. We need more data to establish their frequency and any defining features. Then if discrepancies pop up between the distribution of supernovae lengths now and far away (compared to the current model) we need to investigate. But so far the evidence is strong - far too strong to be demolished by one event.
Yes - the distribution will eventually rule; if it doesn't already. Remember this is not one event - there are at least a half dozen slow burners on the 'local' table. Where are the distant slow burners? Why can't we find them?

Jerry
2011-Jun-30, 05:04 AM
Jerry, how about instead of just firing shots at any old paper that shows up, you actually produce a paper that can show, with all the math, all the assumptions you'll be making, all the observations, etc, exactly what is it you're claiming? It's easy to take pot shots, much harder to actually produce something that would make some sense. Why is it that you continue to take pot shots, but can't produce some actual conclusions that have some actual observations to back up the conclusions? You can't even honestly refute the papers you comment on. Just some wild claims, with no actual evidence. Just your interpretations, which we know are biased and from previous history here, without actual support.

Such a paper would challenge long standing foundational physics and require, according to todays rules, exceptional claims require exceptional evidence. I don't have exceptional evidence. I have tantalizing tidbits: Missing over-bright supernovae in redshifted space, funky spectra on Titan, Martian probes that invariably overshoot their landing targets, weird Love numbers for the Martian interior, strange mass/density correlations.

It will take focused, funded research to tease out results that demonstrate spectacular errors in established physics. This is not a one man project.

The failures of so many modern experiments should turn more heads. There are a lot of icebergs out there, but the Titanic steams full speed ahead.

Shaula
2011-Jun-30, 06:59 AM
I await your references. You make quite a lot of very firm statements here that contradict anything I have read. As you note the burden of proof is on the researchers looking at this. And I am not convinced it is going to be that expensive a project. One person with the data and a good statistical background could do it as an armchair project. The fact that no one doing that has found anything significant is telling IMO.

My point about distant slow burners remains the same - we probably see them, they are the noise on our curves. Or they are misjudged as to the range. But to claim that all the ones we see are slow burn defies the laws of probability. I don't believe in coins that only shows heads at home and tails outside.

PaulLogan
2011-Jun-30, 09:04 AM
I have tantalizing tidbits: Missing over-bright supernovae in redshifted space, funky spectra on Titan, Martian probes that invariably overshoot their landing targets, weird Love numbers for the Martian interior, strange mass/density correlations.

i would be interested in learning more about the items you mentioned here. i did some searching but nothing much came up.
could you please provide some info or links on these topics?

i am no practicing physicist or astronomer but i have long had a gut feeling that some of our assumptions (e.g. redshift distance measurements) are way off. i can't explain it but that gut feeling is quite nagging (and they are often right), so i am honestly interested in what you have found so far.
i like your titanic analogy!

p.s. i'm happy with a pm if you prefer but more people might be interested...

parejkoj
2011-Jul-01, 04:26 AM
Jerry's at (http://www.bautforum.com/showthread.php/73365-Supernova-time-dilation-confirmed-%28again%29?p=1412742#post1412742) it (http://www.bautforum.com/showthread.php/73365-Supernova-time-dilation-confirmed-(again)?p=1414548#post1414548) again (http://www.bautforum.com/showthread.php/89672-Score-one-for-Arp?p=1531333#post1531333).

Cite some papers without understanding them, make vague insinuations that cosmology is about to fall apart based on misunderstandings of recent research results, claim that we obviously understand nothing at all about supernova, suggest that he really understands this stuff better than people who work on it regularly, without ever producing a model for his suppositions.

Tensor's post #4 (http://www.bautforum.com/showthread.php/116987-Type-!a-supernova...slow-to-decline-but-not-overluminous-in-the-optical?p=1906216#post1906216) is correct, and not really answered. Anything that Jerry doesn't understand he puts down to our understanding of fundamental physics being wrong, rather than the more likely possibility that he is misinterpreting what he is reading... again.

Though I probably shouldn't bother, I really have to ask for a citation regarding this quote:


...no one could correctly split synthetic 'type Ia' and 'type Ic' spectra when they are are aged to cosmological distances.

Jerry
2011-Jul-01, 05:06 AM
Jerry's at (http://www.bautforum.com/showthread.php/73365-Supernova-time-dilation-confirmed-%28again%29?p=1412742#post1412742) it (http://www.bautforum.com/showthread.php/73365-Supernova-time-dilation-confirmed-(again)?p=1414548#post1414548) again (http://www.bautforum.com/showthread.php/89672-Score-one-for-Arp?p=1531333#post1531333).

Cite some papers without understanding them, make vague insinuations that cosmology is about to fall apart based on misunderstandings of recent research results, claim that we obviously understand nothing at all about supernova, suggest that he really understands this stuff better than people who work on it regularly, without ever producing a model for his suppositions.

Tensor's post #4 (http://www.bautforum.com/showthread.php/116987-Type-!a-supernova...slow-to-decline-but-not-overluminous-in-the-optical?p=1906216#post1906216) is correct, and not really answered. Anything that Jerry doesn't understand he puts down to our understanding of fundamental physics being wrong, rather than the more likely possibility that he is misinterpreting what he is reading... again.

Though I probably shouldn't bother, I really have to ask for a citation regarding this quote:

No, I am fundamentally correct about supernova type Ia variability; and this has been completely conceded within the cosmological community: Type Ia and type Ic cannot be differentiated at high redshift; except possibly in the UV, but we don't have a good sample of local UV data yet, at least not yet published. If there are super slow burners found at high redshift, this has also not yet been published either; but if such evidence ever is published; even I would concede the evidence for time dilation is strong. Without distant observations of light-curves that are comparably longer after red-shift correction, than the big bright more local slow burners. Currently popular physical interpretations are in limbo.

It is silly to suggest that one person could put the physical world on solid ground; assuming it is not.

But we all get to watch and listen: Let's enjoy the science at Mercury as it comes in from Messenger. And the Dawn asteroid probe; as Emily Lakadawalla has observed: Why do the asteriods look so much like some of the moons of Saturn?

There is much to learn. There is much to unlearn.

Tensor
2011-Jul-01, 08:36 PM
Such a paper would challenge long standing foundational physics and require, according to todays rules, exceptional claims require exceptional evidence. I don't have exceptional evidence.

So, in the meantime, you just make underhanded comments about current science without having to provide any kind of support for your assertions. Insinuate the people working in science are purposely publishing wrong data. Don't you think that is just flat out intellectually dishonest on your part?


I have tantalizing tidbits: Missing over-bright supernovae in redshifted space, funky spectra on Titan, Martian probes that invariably overshoot their landing targets, weird Love numbers for the Martian interior, strange mass/density correlations.

How many data points? At what sigma? Have you run monte carlo simulations on these? Are these tidbits something you've only noticed? If so, have you considered that it might be your interpretations that are wrong? Taking potshots at the scientists that publish these papers, without providing some kind of support is also intellectually dishonest, don't you think?


It will take focused, funded research to tease out results that demonstrate spectacular errors in established physics. This is not a one man project.

Arp was able to produce results, mostly on his own. Most of your claims are very similar to Arp's in that they are based on data that it currently available in papers. What is the difference?


The failures of so many modern experiments should turn more heads.

Data Jerry, data. What data and experiments definitively show a failure of modern experiments and how does the data show this? Not sly insinuations. Why does your interpretation trump others interpretations?


There are a lot of icebergs out there, but the Titanic steams full speed ahead.

Yeah, but you're trying to claim the icebergs are 1 km in size, while it's obvious, based on the actual data, that the icebergs are only a meter across.

Tensor
2011-Jul-01, 10:13 PM
No, I am fundamentally correct about supernova type Ia variability;and this has been completely conceded within the cosmological community:

No, you're not correct. At what wavelengths Jerry? Have you forgotten that you provided a paper that showed that in the IR, there isn't variability? In the optical, there are various causes, THAT ARE KNOWN AND CAN BE TAKEN INTO ACCOUNT. I notice you missed including this.


Type Ia and type Ic cannot be differentiated at high redshift; except possibly in the UV, but we don't have a good sample of local UV data yet, at least not yet published.

Oh yeah? Care to support this one? Your inference here is that we can differentiate them at lower redshift, right? So how does the process used to differentiate them change at high redshift? I know how the major one possibly changes, but what about the rest of the distinctions?


If there are super slow burners found at high redshift, this has also not yet been published either; but if such evidence ever is published; even I would concede the evidence for time dilation is strong. Without distant observations of light-curves that are comparably longer after red-shift correction, than the big bright more local slow burners.

What data do you have or know of that shows that slow burner spectra are not different enough to identify at higher redshift? Or is this just more dumping on the mainstream, without, as I suspect, any hard evidence?


Currently popular physical interpretations are in limbo.

And your evidence for this is? The papers I've seen don't seem to indicate this. There are several different models, FOR DIFFERENT OBSERVATIONS. But that hardly rates as being in limbo.


It is silly to suggest that one person could put the physical world on solid ground; assuming it is not.

But, that person seems to feel they can claim what they want, put some link to a paper, that turns out to refute the specific claim, right?


But we all get to watch and listen: Let's enjoy the science at Mercury as it comes in from Messenger.

If that's how you feel, then why keep making comments as though you think the scientists don't know what they are doing?


And the Dawn asteroid probe; as Emily Lakadawalla has observed: Why do the asteriods look so much like some of the moons of Saturn? There is much to learn. There is much to unlearn.

Yes there is. So work on providing us with actual data and actual facts. Not insinuation or vague claims or ideas.

Jerry
2011-Jul-03, 05:48 AM
So, in the meantime, you just make underhanded comments about current science without having to provide any kind of support for your assertions. Insinuate the people working in science are purposely publishing wrong data. Don't you think that is just flat out intellectually dishonest on your part?
Wrong data? The supernova data stream is incredibly rich. But two things are wrong: 1) Coming up with new data reduction schemes every six months. This was my primary gripe in the first post. It is hard to compare apples with apples when every crate is handmade.


How many data points? At what sigma? Have you run monte carlo simulations on these?
Orbital gravitation data is increbibly complex - but anyone can look at gravitational maps of the moon, Mars and Venus and look for trends or clues that something is amiss...and we all know this is not the place to discuss this.


Are these tidbits something you've only noticed? If so, have you considered that it might be your interpretations that are wrong? Taking potshots at the scientists that publish these papers, without providing some kind of support is also intellectually dishonest, don't you think?
One of the most frustrating aspects of aerospace engineering of the last three decades is that the analytical instrumentation has changed so dramatically, you can't compare, for example, particle size data from the two decades ago with data produced today - and engineers gripe about that. They aren't calling their scientist bad, but they are frustrated that they have to reinvent the wheel.

In astronomy, the changes in instrumentation have been just as overwhelming. It makes it hard to compare data, so we gripe.


Arp was able to produce results, mostly on his own. Most of your claims are very similar to Arp's in that they are based on data that it currently available in papers. What is the difference?
We would all like science to be clear, but the edges ARE murky. I don't think anybody dreamed the severity of thunderstorms could be related to solar storms a few decades ago. Now we have the tools explore this question and get real answers. Likewise with supernovae, we are getting enough data to possibly refute older theories - calling for clarification is not calling the previous generation bad scientists.


Data Jerry, data. What data and experiments definitively show a failure of modern experiments and how does the data show this? The gravity B probe was a 750 million dollar space anchor. I wouldn't call every gravity antenna ever built a failure: Some of them are incredible instruments; but they have all failed to detect gravitational waves. That doesn't mean the theorists who predicted that we would see them by now bad theorists, either. But they were definitely wrong. It is time to question some of the fundamental assumption use in making the predictions.


Yeah, but you're trying to claim the icebergs are 1 km in size, while it's obvious, based on the actual data, that the icebergs are only a meter across. Supernovae vary much more in magnitude and light-curve length than researchers thought as little as a decade ago; and it is more difficult to classify distant events than any of the principle investigators surmised. Many of the papers published before ~2004 were optomistically based upon wrong suppositions - just like the gravity antennas.

Jerry
2011-Jul-03, 06:24 AM
No, you're not correct. At what wavelengths Jerry? Have you forgotten that you provided a paper that showed that in the IR, there isn't variability? In the optical, there are various causes, THAT ARE KNOWN AND CAN BE TAKEN INTO ACCOUNT. I notice you missed including this. That is a silly thing to say on this thread. This thread is about a supernova of average magnitude with a very long light curve in all wavelengths. And we can't compare IR spectra for the most distant events with local events - they are too redshifted. (We have very little local UV spectra because they are not reshifted enough!)


Oh yeah? Care to support this one? Your inference here is that we can differentiate them at lower redshift, right? So how does the process used to differentiate them change at high redshift? I know how the major one possibly changes, but what about the rest of the distinctions? High redshift spectra is very noisy and missing a lot of detail. So even if it is argued that there are discriminaters at low redshift, many of them are washed out at high redshift. I didn't set up the challenge to see if type Ia and Ic could be properly seperated at high redshift - it was a supernova community challenge.




What data do you have or know of that shows that slow burner spectra are not different enough to identify at higher redshift? Or is this just more dumping on the mainstream, without, as I suspect, any hard evidence?
Results from the Supernova Photometric Classification Challenge (Kessler)

http://xxx.lanl.gov/abs/1008.1024


CONCLUSION
We have presented results from the SN classification challenge that finished 2010 June 1. Among the four
basic strategies that were used in the SNPhotCC (3), three strategies show comparable results for the entries
with the highest figure of merit. Therefore no particular strategy was notably superior. For all of the entries, the
classification performance was significantly better for the spectroscopic training subset than for the unconfirmed
sample. The degraded performance on the unconfirmed sample was in part due to participants not accounting
for the bias in the spectroscopic training sample.

Shaula
2011-Jul-03, 07:14 AM
On the other hand, if the assumption that distant events are time dilated is false, as the number of distant events observed increases, it will become painfully obvious that after correction for time dilation, the light curve appears to be shorter. This is already true for gamma rays,.
You started the spectral component debate. If it is valid to argue this for for gamma sources why not IR?

Tensor
2011-Jul-03, 11:25 PM
Wrong data? The supernova data stream is incredibly rich. But two things are wrong: 1) Coming up with new data reduction schemes every six months. This was my primary gripe in the first post. It is hard to compare apples with apples when every crate is handmade.

And you can't do your own, Why? All the data is out there. Dave Russell seemed to be able to do his own research. But somehow, you can't. Right.


Orbital gravitation data is increbibly complex - but anyone can look at gravitational maps of the moon, Mars and Venus and look for trends or clues that something is amiss...and we all know this is not the place to discuss this.

Yeah, it's not the place. But you keep bringing it up, in ways you don't have to support it, to use it to take shots at the mainstream. For instance:


Such a paper would challenge long standing foundational physics and require, according to todays rules, exceptional claims require exceptional evidence. I don't have exceptional evidence. I have tantalizing tidbits: Missing over-bright supernovae in redshifted space, funky spectra on Titan, Martian probes that invariably overshoot their landing targets, weird Love numbers for the Martian interior, strange mass/density correlations.

Why is that Jerry? Why not explain, in detail, with evidence and facts to back it up? Instead of vague claims? Is it because you know you can't support it?


In astronomy, the changes in instrumentation have been just as overwhelming. It makes it hard to compare data, so we gripe.

Gripes I can deal with. Making accusations that researchers are fudging or manipulating their data is what I can't deal with.



We would all like science to be clear, but the edges ARE murky. I don't think anybody dreamed the severity of thunderstorms could be related to solar storms a few decades ago. Now we have the tools explore this question and get real answers. Likewise with supernovae, we are getting enough data to possibly refute older theories - calling for clarification is not calling the previous generation bad scientists.

But you are not calling for clarification. You're flat out saying the scientists don't know what they are doing:


There are two selection biases at play here: natural (type II) Malmquist bias, which is the propensity to find brighter objects with increasing distance; ...

So, you're saying the researchers are not aware of the Malmquist bias? That they ignore it? That they don't correct for it?

(section pulled out to be used below)


No matter how you bake it; slow decliners in local space nullify prior claims that more distant events clearly demonstrate time dilation.

No it doesn't Jerry. Can you show the local slow decliners have the same spectra as the more distant events?


Nobody knows what is happening out there: Slooow local decliners mean Slooow distant decliners don't prove time dilation.

Actually, slow local decliners that match distance slow decliners would actually support time dilation. Local slow decliners with spectra that match distant spectra slow decliners would be against dilation. Do you have evidence of this?


We would have to find a distant sloooooooooooW decliner; and no one has seen one yet. This is soooooo true that it is no longer a tenable argument to state otherwise.

Why. Provide the evidence for this. How is it that we can match the spectra of local events to the dilated spectra of the distant events? How does that make it a no longer tenable argument. If no one has seen a distant slow decliner, then how can you be sure that the nearby normal decliners spectra don't match?


If anything, the data in hand suggest the theory of relativist expansion is wrong.

This doesn't sound to me as if you are calling for clarification. You're flat out saying that current theories are wrong. Of course, you don't present any kind of support for that, other than your spin on it. Numbers Jerry. Can you show that we can't differentiate between the slow decliners and the normal Sne 1a? Or, is this just another way to cast doubt on the mainstream theories, without actually having to present any kind of evidence?

Or, what is worse to mine mind, you are claiming the researchers are falsifying their data, by limiting their searches:


and the fact that many many supernova searchers looking for time dilated events have scheduled searchs so that they would be more likely to see slow decliners in the past; even to the point of rejecting events from further study that do not have a sufficiently long light curve to be time dilated events.

To me, you are effectively calling the researchers liars. Is that what you mean Jerry? That the researchers are intentionally lying about their data? If not, what exactly does that statement in bold mean?


The gravity B probe was a 750 million dollar space anchor. I wouldn't call every gravity antenna ever built a failure: Some of them are incredible instruments; but they have all failed to detect gravitational waves.

We know you don't like the data from GPB. But, can you show that the data they are getting from it, using the methods they are using, are somehow wrong? Not that you don't like it, but that it is wrong.


That doesn't mean the theorists who predicted that we would see them by now bad theorists, either. But they were definitely wrong. It is time to question some of the fundamental assumption use in making the predictions.

Jerry, please provide a quote from anywhere, that claims that there would be a definite (not we think, not there's a good chance, not it's a good bet), but a DEFINITE prediction of detecting gravity waves with current equipment.


Supernovae vary much more in magnitude and light-curve length than researchers thought as little as a decade ago;

How much more? In what wavelengths? In what bands?


and it is more difficult to classify distant events than any of the principle investigators surmised. Many of the papers published before ~2004 were optomistically based upon wrong suppositions - just like the gravity antennas.

Jerry, please provide a quantitive difference between papers prior to 2004 and current papers, as far as error rate is concerned. I'm not saying there's no difference. But you are making it sound as if those papers prior to 2004 couldn't even get within three or four gigaparsecs of distance. Again, it would be nice that, instead of vague handwaving claims and accusations, you would actually produce something that can be checked for an actual number.

Tensor
2011-Jul-04, 12:31 AM
That is a silly thing to say on this thread. This thread is about a supernova of average magnitude with a very long light curve in all wavelengths.

Hey, you brought up the variability, not me. I was just pointing out that at IR wavelengths, SNe1a are even more of a standard candle that optically. I noticed you totally ignored the methods for identifying type 1a, or the methods for identifying the actual peak magnitude in the presence of dust or other processes.


And we can't compare IR spectra for the most distant events with local events - they are too redshifted. (We have very little local UV spectra because they are not reshifted enough!)

And, so? You are still totally ignoring that we can differentiate between type 1a from other types by spectra.


High redshift spectra is very noisy and missing a lot of detail. So even if it is argued that there are discriminaters at low redshift, many of them are washed out at high redshift.

Support for this? That's why I asked you to provide some kind of support as for not being able to differentiate between 1a and 1c, BY SPECTRA. Not your own claim. Speaking of which:


I didn't set up the challenge to see if type Ia and Ic could be properly seperated at high redshift - it was a supernova community challenge.

Results from the Supernova Photometric Classification Challenge (Kessler)

http://xxx.lanl.gov/abs/1008.1024

Yeah, you might not have set it up, but as usual, it appears that you either didn't actually read the paper or misunderstood what the paper was about. The comment you posted was for those supernova that were spectroscopically unconfirmed. That means, that the supernova were classified based on photometric details, not their spectra.

While less than half have been spectroscopically identified, it is those specifically that have been used for cosmological comparison and time dilation studies. For instance, the paper you presented states:


In short, cosmological parameter estimates from the much larger recent surveys are based solely on spectroscopically confirmed SNe Ia .

Remember, I asked you for evidence that type 1a and 1c cannot be differentiated at high z by spectra? Well, according to the paper you provided, it appears that they can tell the difference with no problem using spectra, since they use the spectra to verify other methods. And, it seems, a large majority of the time, they can even do it using photometric methods. Not saying they're perfect with the photometric methods, but....

Basically Jerry, the that paper is being used to


(1) learn the relative strengths and weaknesses of the different classification algorithms, (2) improve the algorithms, (3) understand what spectroscopically confirmed subsets are needed to properly train these algorithms, and (4) improve the simulations.

In other words Jerry, they're looking for ways to improve the selection process using Photometrics. They used Spectra to verify the photometric method. You've again provided support for the refutation of your contention. So, again, do you have anything that provides support for your contention that the SPECTRA of type 1a and type 1c cannot be differentiated at high z?

Jerry
2011-Jul-04, 04:44 AM
You started the spectral component debate. If it is valid to argue this for for gamma sources why not IR? We don't have complete IR spectra for highly redshifted events - they are redshifted out of the detectable, resolvable wavebands. Likewise, since our atmosphere is opaque to most UV wavelenghts, we have only limited UV observations of local events. (In highly redshifted spectra, the UV is redshifted into visible bands, and this is way we have more UV information about more distant sup

So we can't see IR spectra in distant events, and from the surface of the earth, we can't see Local UV spectra.

parejkoj
2011-Jul-06, 02:32 AM
So we can't see IR spectra in distant events, and from the surface of the earth, we can't see Local UV spectra.

And your point is?

Shaula
2011-Jul-06, 05:02 AM
And when you say IR do you near NIR, SWIR, MWIR or LWIR? For NIR to be shifted out of the LWIR atmospheric window you need redshift of about 15 IIRC. You should be able to grab some data on NIR And SWIR out to quite large distances, especially from high dry sites.

Jerry
2011-Jul-08, 03:34 AM
And when you say IR do you near NIR, SWIR, MWIR or LWIR? For NIR to be shifted out of the LWIR atmospheric window you need redshift of about 15 IIRC. You should be able to grab some data on NIR And SWIR out to quite large distances, especially from high dry sites.If and only if your scope has this bandwidth. Remember, most of our supernova observations are in the optical bandwidths...or UV redshifted into optical.

Jerry
2011-Jul-08, 03:37 AM
And your point is?

We have diffferent spectral windows available for local and highly redshifted supernova events, k corrections notwithstanding. In the bandwidths that over-lap, there is ambiguity between type Ia and type Ic.

Jerry
2011-Jul-08, 04:01 AM
And you can't do your own, Why? All the data is out there. Dave Russell seemed to be able to do his own research. But somehow, you can't. Right. I work a scheduled 86 hour week...


Originally Posted by Jerry

No matter how you bake it; slow decliners in local space nullify prior claims that more distant events clearly demonstrate time dilation.

No it doesn't Jerry. Can you show the local slow decliners have the same spectra as the more distant events?
The slow decliner is a local, spectragraphically confirmed type Ia. No, I can't confirm the distant events have identical spectra to this one slow decliner, or any of the other local events, but neither can anyone else assume that the distant events must be the same as local, normal decliners:

If you assume the most distant events are average in light-curve width and magnitude and time-diluted, then you must also assume we have not yet seen any highly redshiftedf slowly declining type Ia or type Ic. Since Ic are brighter than Ia, why haven't we found any time-dilated type Ic? There is no good answer to this question.

Jerry
2011-Jul-08, 04:24 AM
But you are not calling for clarification. You're flat out saying the scientists don't know what they are doing: The author stated "There was significantly better results with the training set." Why not state the obvious: "The methods did not work, and they did not work because all four groups tried to get the same distribution with the test sample as they did with the training sample: They tweaked the results until the results matched the distribution they expected - they guess that the test sample was the same as the training sample. That is a freshman mistake.



Originally Posted by Jerry
There are two selection biases at play here: natural (type II) Malmquist bias, which is the propensity to find brighter objects with increasing distance; ...
So, you're saying the researchers are not aware of the Malmquist bias? That they ignore it? That they don't correct for it?

The irony here is that I think the distant distribution should mirror the local distribution - I think it is the only reasonable assumption - So when none of the most distant events have long enough light curves after correction for time dilation, to be equal in length to the brightest local event - this screams to me that there has not been a proper correction for Malmquist bias. I have argued for a decade now, that the distance modulus is wrong.

Jerry
2011-Jul-08, 04:52 AM
Hey, you brought up the variability, not me. I was just pointing out that at IR wavelengths, SNe1a are even more of a standard candle that optically. I noticed you totally ignored the methods for identifying type 1a, or the methods for identifying the actual peak magnitude in the presence of dust or other processes.
Again, I am not aware of good IR discriminators between type Ia and type Ic - especially at high redshift distance. I do know that the studies ~ 2000 - 2002 (Goldhaber, Perlmutter) that assumed there was little or no dust interference in the most distant spectra they observed were probably incorrect.


And, so? You are still totally ignoring that we can differentiate between type 1a from other types by spectra. You continue to ignore the obvious: none of the more distant events are classified as subtype Ic, which are very difficult, locally, to separate from type Ia.

Shaula
2011-Jul-08, 05:15 AM
If and only if your scope has this bandwidth. Remember, most of our supernova observations are in the optical bandwidths...or UV redshifted into optical.
Not hugely relevant. You were arguing that IR is shifted outside the atmospheric transmission windows. I was highlighting the range of redshifts for which this for which this was the case. You've apparently shifted from "It is impossible" to "Well, OK, it is possible.... But there aren't enough telescopes!". If I gave a list of relevant ones are you going to shift stance again and talk about contention or something?

parejkoj
2011-Jul-08, 04:45 PM
We have diffferent spectral windows available for local and highly redshifted supernova events, k corrections notwithstanding. In the bandwidths that over-lap, there is ambiguity between type Ia and type Ic.

What do K-corrections have to do with spectra? (we've been through this before, and I suspect you didn't learn anything from that conversation).

Also, according to the paper you linked (Kessler et al. 2010), we unequivocally can tell the difference between Ia and Ic spectroscopically, and we can identify Ia from Ic with ~80% accuracy with only photometric information and no redshift whatsoever. That doesn't sound like ambiguity to me.

Jerry
2011-Jul-10, 05:00 AM
What do K-corrections have to do with spectra? (we've been through this before, and I suspect you didn't learn anything from that conversation).
Yes we did. K-corrections vary in meaning with application. Hogg did a formalization of supernova K corrections, lumping spectral and lens corrections:


Technically, the K correction described here includes a slight generalization from the original
conception: The observed and emitted-frame bandpasses are permitted to have arbitrarily different
shapes and positions in frequency space (as they are in, e.g., Kim et al 1996). In addition, the
equations below permit the different bandpasses to be calibrated to different standard sources.


Also, according to the paper you linked (Kessler et al. 2010), we unequivocally can tell the difference between Ia and Ic spectroscopically, and we can identify Ia from Ic with ~80% accuracy with only photometric information and no redshift whatsoever. That doesn't sound like ambiguity to me.
I don't have to agree with Kessler's summation, especially when trying interpret highly redshifted, background loaded low signal/noise events. In any case, whether or not type Ia and Ic can be discriminated is almost a moot point: Here we have a very slowly expanding event that is a type Ia, not a type Ic.

parejkoj
2011-Jul-10, 01:09 PM
Yes we did. K-corrections vary in meaning with application. Hogg did a formalization of supernova K corrections, lumping spectral and lens corrections:


I said this to you in the other thread. The K-correction has precisely one meaning, and it has nothing to do with spectroscopy. Full stop, end of story, do not pass go, that's all she wrote.

Since repetition is the spice of life, I'll say this again: you K-correct photometry, not spectroscopy.

You really don't know what you're talking about on this one. If I were you, I'd back down, study Hogg (2002) for a while and implement your own K-correction code (or learn how the NYU version works), before you make any more false claims about it. As I guessed, you didn't learn anything from the last time we talked about this.


I don't have to agree with Kessler's summation, especially when trying interpret highly redshifted, background loaded low signal/noise events.

Waitaminute... So, you don't agree with the results of the paper you cited because it contradicts your interpretation of their results? Did I get that right?


In any case, whether or not type Ia and Ic can be discriminated is almost a moot point: Here we have a very slowly expanding event that is a type Ia, not a type Ic.

You brought up our ability to distinguish Ia/Ic as part of your general bellyaching early in the thread, not me:



There is a significant number of overly-bright very long light-curve events in fairly local space. These are generally classified as "Type Ic". (Although in a recent, professional challenge, no one could correctly split synthetic 'type Ia' and 'type Ic' spectra when they are are aged to cosmological distances.)

Kessler et al. (2010) found that we damn well can distinguish between Ia and Ic spectroscopically, and we can do very well photometrically, with no redshift information provided. That's pretty dang impressive, and completely at odds with what you are claiming.

Though since we're apparently arguing with your interpretation of the results of that challenge, you should probably take this to ATM, and then spell out how exactly you arrived at your interpretation...

tusenfem
2011-Jul-11, 07:13 AM
Okay, Jerry seems at his often discussed ATM once more. He only understands and mainstream is wrong.
I fail to see why this should continue here and certainly not in S&T.
Closing the thread, if anyone feels the need to discuss this topic with Jerry , report this post and I can move it to ATM.