Results 1 to 16 of 16

Thread: Type 1a SN from White Dwarf Merger question

  1. #1
    Join Date
    Jul 2005
    Location
    Massachusetts, USA
    Posts
    19,915

    Type 1a SN from White Dwarf Merger question

    Finally, What is Friday without a Supernova paper?

    http://arxiv.org/abs/1201.2406How the merger of two white dwarfs depends on their mass ratio: orbital stability and detonations at contact

    Dan & co conclude there is a fairly broad range of merger masses that could achieve a supernova type Ia signiture.
    The above is from Jerry's entry today in the Fun Papers thread. The paper referenced is about simulations run on the inspiral of pair of white dwarfs becoming SN 1a's to see what biases there might be about mass ranges for the objects. There have been several 1a's studied recently showing that these particular ones could not have been the result of a near critical WD getting the last bit of mass needed from a bloated close companion (the normal model till very recently). The above paper does a nice job of looking at the behavior of various pairs of hypothetical masses between 0.2 and 1.2 solar masses. What it doesn't say, and what I've been trying to find out elsewhere, is this:

    Why do SN 1a's almost all (there are some obvious exceptions) produce exactly the same amount (within our ability to measure it) of Ni-56 (whose decay is the main driver for SN brightness for the weeks and months after the initial explosion), regardless of initial masses?
    Forming opinions as we speak

  2. #2
    Join Date
    Nov 2002
    Posts
    6,238
    Quote Originally Posted by antoniseb View Post
    Why do SN 1a's almost all (there are some obvious exceptions) produce exactly the same amount (within our ability to measure it) of Ni-56 (whose decay is the main driver for SN brightness for the weeks and months after the initial explosion), regardless of initial masses?
    I'm not sure this is the case for DD type SN1as. I've seen a couple of papers (and I'll have to look for them) where the Ni-56 mass was over 2 Solar masses. However this paper may be of use to you. There are several graphs on there where the C-O core of the white dwarf is rather consistent over the white dwarf initial mass, the main sequence initial mass, or the accretion rate. In addition, one of these authors also co-authored this paper which also takes rotation speed into account. I'll keep looking for the papers on the DD scenario papers.

  3. #3
    Join Date
    Jul 2005
    Location
    Massachusetts, USA
    Posts
    19,915
    Quote Originally Posted by Tensor View Post
    I'm not sure this is the case for DD type SN1as. ...
    Recently this appeared in the popular press.

    Aside from that there have been several such efforts in the last six months or so that have turned up either "no companion", or "no surrounding dust".
    I think the 2 solar mass one you referred to was considered way outside the norm IIRC. I think they usually produce about 0.5 solar masses of Ni-56.
    Forming opinions as we speak

  4. #4
    Join Date
    Nov 2002
    Posts
    6,238
    Quote Originally Posted by antoniseb View Post
    Recently this appeared in the popular press.

    Aside from that there have been several such efforts in the last six months or so that have turned up either "no companion", or "no surrounding dust".
    I think the 2 solar mass one you referred to was considered way outside the norm IIRC. I think they usually produce about 0.5 solar masses of Ni-56.
    I found one of the papers. Here it is , however, this was not the 2 solar mass supernova. This supernova (207if) contained 1.6 solar mass of nickel, which is still well outside the norm, since the nickel itself is above the Chandrasekhar limit. However, after going through several papers, I think I was remembering the total mass of the progenitors (this was 2.4 solar masses), not the Ni-56 mass. I've found the Nickel masses for 2003fg (1.29), 2006gz (1-1.2), and 2009dc(1.4-1.7). In a 2006gz paper there is a chart (page L20, the last in that paper) which shows the different Ni-56 masses for several different supernovae. It appears that the normal range for Ni-56 is from .4 to .9 solar masses.

  5. #5
    Join Date
    Mar 2004
    Location
    Ocean Shores, Wa
    Posts
    5,149
    Not sure how the Nickle mass could be titrated extremely tight. You are always looking at a spectra distributed over time with an estimate of the baseline. You can always get peak ratios, but somewhere, an assumption is made about the baseline/peak absolute height. In the early days, when the Chandrasaker mass was assumed; one could work backward.

    In todays reality, the 0.4 to 0.9 range should be held up with a suggestion, rather than an absolute: There are many more variables than anyone would like to deal with in a duel white dwarf situation.

  6. #6
    Join Date
    Jun 2010
    Posts
    652
    Quote Originally Posted by antoniseb View Post
    The above is from Jerry's entry today in the Fun Papers thread. The paper referenced is about simulations run on the inspiral of pair of white dwarfs becoming SN 1a's to see what biases there might be about mass ranges for the objects. There have been several 1a's studied recently showing that these particular ones could not have been the result of a near critical WD getting the last bit of mass needed from a bloated close companion (the normal model till very recently).
    Is that no longer the normal model?

    Quote Originally Posted by antoniseb View Post
    The above paper does a nice job of looking at the behavior of various pairs of hypothetical masses between 0.2 and 1.2 solar masses. What it doesn't say, and what I've been trying to find out elsewhere, is this:

    Why do SN 1a's almost all (there are some obvious exceptions) produce exactly the same amount (within our ability to measure it) of Ni-56 (whose decay is the main driver for SN brightness for the weeks and months after the initial explosion), regardless of initial masses?
    Well the obvious answer is the vast majority are produced by the (former) normal model and have the same initial mass.

  7. #7
    Join Date
    Jul 2005
    Location
    Massachusetts, USA
    Posts
    19,915
    Quote Originally Posted by Jerry View Post
    Not sure how the Nickle mass could be titrated extremely tight. You are always looking at a spectra distributed over time with an estimate of the baseline...
    In the case of Type 1a SN, after about 30 days, the brightness of the SN has the same half-life as Ni-56. Early, other species (such as Co-56) are turning into Ni-56. But it this quantity that makes 1a's so constant in brightness and behavior.
    Forming opinions as we speak

  8. #8
    Join Date
    Sep 2006
    Posts
    1,391
    Tiny nitpick: the decay in the lightcurve of a Type Ia SN is similar to the decay in energy emitted by Ni-56 once it enters into the late-time stages, which are more than 30 days after explosion -- more like 100 or 200 days. SN 2011fe in M101, for example, was decreasing in brightness at a rate ranging from 0.012 mag/day (in B-band) to 0.036 mag/day (in I-band) between days 60 and 130 after explosion. An object powered purely by Ni-56 (ignoring radiative transfer, blah blah blah) would decrease at a rate of 0.0098 mag/day.

  9. #9
    Join Date
    Mar 2004
    Location
    Ocean Shores, Wa
    Posts
    5,149
    From Tensor's Foley Paper: http://arxiv.org/abs/1202.0003 The Relation Between Ejecta Velocity, Intrinsic Color, and Host-Galaxy Mass for High-Redshift Type Ia Supernovae

    Quote Originally Posted by Foley
    The distributions of ejecta velocities for SNe Ia at low and high redshift are similar, indicating that current cosmological results should have little bias related to the velocity-color relation.
    First, it continues to erk me that as the data base of well-studied events increases, the number of internal correction factors increases, too; and generally with the same result: We have tightened the relationship. But a similar appoarch to any diverging data field could have the same result! (I actually saw this happen with the manufacturing of a particle size determination tool: every time the data seemed to diverge from expectations, the vendor would add another term to a a long line of polynomial corrections. We spent years on it before deciding they just didn't have a controlled system.)

    Second, Foley seems to be saying the distant sample is the same as the local sample; while other papers have found a divergence that correlates (very loosely) with galaxy type. Make it simple guys: look at the least-manipulated data and listen to what it is saying.

  10. #10
    Join Date
    Nov 2002
    Posts
    6,238
    Quote Originally Posted by Jerry View Post
    Quote Originally Posted by Foley
    Quote Originally Posted by Jerry View Post
    From Tensor's Foley Paper: http://arxiv.org/abs/1202.0003 The Relation Between Ejecta Velocity, Intrinsic Color, and Host-Galaxy Mass for High-Redshift Type Ia Supernovae
    The distributions of ejecta velocities for SNe Ia at low and high redshift are similar, indicating that current cosmological results should have little bias related to the velocity-color relation.
    First, it continues to erk me that as the data base of well-studied events increases, the number of internal correction factors increases, too; and generally with the same result: We have tightened the relationship. But a similar appoarch to any diverging data field could have the same result!
    Well, why don't you show us, using actual numbers, how such a data field would produce such a result? Instead of just complaining about the work of the people actually doing the papers.

    Quote Originally Posted by Jerry View Post
    (I actually saw this happen with the manufacturing of a particle size determination tool: every time the data seemed to diverge from expectations, the vendor would add another term to a a long line of polynomial corrections. We spent years on it before deciding they just didn't have a controlled system.)
    Nicely built strawman Jerry. This has zero to do with the paper, except maybe for some imaginary connection you think you found in the abstract.

    Quote Originally Posted by Jerry View Post
    Second, Foley seems to be saying the distant sample is the same as the local sample; while other papers have found a divergence that correlates (very loosely) with galaxy type. Make it simple guys: look at the least-manipulated data and listen to what it is saying.
    Yeah, Jerry, why don't you follow your own comment? You did read the entire paper, right? You aren't basing your comment on that one sentence you pulled out of the ABSTRACT, are you? How loosely with galaxy type Jerry? How closely does the distant sample match the near sample with respect to the ejecta velocity-color relationship? Which one has the highest sigma? I don't know why I bother. From past experience, I won't get the answers to those questions, just more words without actual numeric examples, complaining about the work actual scientists have done.

    What irks me is that you didn't even bother to read the paper before complaining about it. Foley is actually quite specific about what SNe 1a relations at low and high z have good correlation and sigma and where those relations are valid, which of those relations looks good, but needs more research, which of those relations don't have good correlation, and which of those relations are simply unknown at this time. Care to point out exactly what parts of his data the the reasoning behind that data, in the bode of the paper, you don't agree with?

  11. #11
    Join Date
    Mar 2004
    Location
    Ocean Shores, Wa
    Posts
    5,149
    Quote Originally Posted by Tensor View Post
    Well, why don't you show us, using actual numbers, how such a data field would produce such a result? Instead of just complaining about the work of the people actually doing the papers.
    Tensor, I have done that. In a paper posted in archives I wrote almost a decade ago, I demonstrated that I can reduce the scatter in Wang's Color method by removing the correction for time dilation. Why isn't this published? It is not a valid analytical technique. For the self-same reason that I can't use an unproven relationship to tighten one parameter, applying any scaler that can't be calibrated and thus reducing scatter and declare victory is flawed.

    Look at the stretch factor papers that came out around the turn of the century, and how tight of limits were imposed upon supernova magnitudes before we knew they are likely binary events! The simple stretch/magnitude relationship breaks down if the viewing angle, or duel or bimodal detonation mechanisms are likely (as the double-humped peak in many, but not all, Type Ia events may indicate).

    Bulk observational properties must be establish and trended first. Then and only then should more intricate relationships be investigated. It remains a curious bulk property that more distant events achieve brighter magnitudes in the ultraviolet (relative to other colors) than in the local sample - at least this is an internal, relative property that is likely tied to peak magnitude. This UV relationship runs counter to the conclusions drawn in the stretch papers, and remains a very important area of research.

    Foley may have a good foundation for what he is proposing; but reducing the apparent scatter in magnitude is not impressive to me if he cannot account for the excess found in the ultraviolete relative to local observations. It seems to me that in a binary event, the viewing agle is going to have a lot greater effect on the apparent velocity than the color relationship. Foley looks a dust, but does not discus relative geometry. Foley mentions the fact that event geometry is important, but then he states 'it is prudent to determine if SN Ia ejecta velocity is correlated with host-galaxy mass'. I don't get that - why would the host galaxy mass be more important than the geometry of the event? I would be concerned about a systematic error if my ejecta velocity correlated with host dust or host galaxy mass.

  12. #12
    Join Date
    Mar 2004
    Location
    Ocean Shores, Wa
    Posts
    5,149
    Quote Originally Posted by Tensor View Post
    Well, why don't you show us, using actual numbers, how such a data field would produce such a result? Instead of just complaining about the work of the people actually doing the papers.
    You got it: From the people actually doing the work:

    http://arxiv.org/pdf/1203.4832.pdf THE SPECTROSCOPIC DIVERSITY OF TYPE Ia SUPERNOVAE

    There is a large enough database now to sort-out spectral classifications with fairly local events. There are families, but there is also a continuem; and exceptions and nagging exceptions to the exceptions. All-in-all, as we increase the resolution of more distant events we will be able to statistically sort them as Blondin has in this paper. It is my opinion that we will only be able to properly judge distant events if we bury preconceptions and look at the supernova Ia family as truly revealing - and just like particle physics, not necessarily what we expected.

  13. #13
    Join Date
    Mar 2004
    Location
    Ocean Shores, Wa
    Posts
    5,149
    Quote Originally Posted by Tensor View Post
    Nicely built strawman Jerry. This has zero to do with the paper, except maybe for some imaginary connection you think you found in the abstract.
    Particle size distribution has everything to do with every distant observation. We know how badly local dust effects seeing, and how different dusty clusters and galaxies appear from more pristine environments. We have good reasons to suspect there are spinning, magnitized dust particles, and this will interfere with our ability to interpret microwave data. The best way we found to calibrate particle size is to shoot extremely thin streams particles with a laser and carefully study the defraction patterns. We can't do that in space - our reddening corrections are difficult to sort-out from relativistic corrections and likely remain the single largest source of error in all cosmic studies.

  14. #14
    Join Date
    Mar 2004
    Location
    Ocean Shores, Wa
    Posts
    5,149

    From Fun Papers in archives May 15

    *DD v. CD in SN1a*
    http://arxiv.org/abs/1205.2452
    Quote Originally Posted by Antoniseb
    This is just a simulation, but it looked at the frequencies of two white dwarfs merging (double degenerate) vs. a white dwarf getting critical mass from a host (core degenerate), and finds that DD should account for about 99% of the SN1a's that we see.
    Nice. More and more complex models are being developed that work with DDs but not SDs. The jury will remain out on this for some time - light echoes and background noise fuzz all of the lines) but the DD models win most of the time.

  15. #15
    Join Date
    Mar 2004
    Location
    Ocean Shores, Wa
    Posts
    5,149

    Another superluminous event

    http://arxiv.org/pdf/1205.6973v1.pdf PHOTOMETRIC OBSERVATIONS OF THE SUPERNOVA 2009nr

    While the existance of superluminous supernova Ia is well-established, this event throws another wrench in use of normal events to calibrate cosmic parameters. Prior to this event; a relationship appeared to be developing between the distance an event occurred from galactic centers and the magnitude of the event. 2009nr violates this relationship. More importantly, since it occurred in the parimeter of the host galaxy, it means some assumptions about the age of a stellar population required to spawn a super-occurance are suspect.

    Why is this important? The most distant supernova events are used to calibrate the 'dark energy' parameter. It is assumed these events are best represented by average supernova magnitudes. If they are superluminous, like this one, the 'dark energy' parameter is much greater than current theory suggest. (Actually too much greater than current theory allows.)

    Selection effects - the probability that more distant events we observe are skewed towards brighter objects should force us to predict that the most distant supernova type Ia we observe are over-luminous. Researches get around this inconvenience by assuming that in the early universe, super-luminal binary events are less likely to occur. The curve developed that ties luminosity to distance from the galactic center is consistent with this assumption. But now we have a superluminous event that is a clear exception to this rule; and more or less invalidates the age/event-type correlation. Without this relationship, it is more reasonable to assume the most distant events we observe are best represented by the superluminous class of events.

    There is even a deeper underlying assumption that falls amiss if the most distant events are indeed super luminous. There is a fairly well established relationship between light-curve width and supernova absolute magnitude. If the most distant events are brighter, they should also have longer light curves, but they don't - on average, the most distant events are observed to have average or shorter-than-average light curve length. If the most distant events are superluminous, the relationship between light curve width and luminosity breaks down. Or does it? We should be able to determine whether or not the most distant observations are superluminous by analysing the UV light curve. In local events, superluminous events have measurably brighter UV lightcurves, and so do the most distant events. So again, it is reasonable to assume the most distant events we observe are in all probability over-luminous. The 'deeper underlying assumption' I am refering to is the relativistic correction for time dilution. If this correction is too great, or should not be applied, the most distant events indeed have longer light curves than standard supernovae and are indeed superluminous.

    So how do we know that the relativistic correction is correct? We don't. The proof we have been using to establish that the universe is expanding in a time-dilated reality is that the most distant events have longer light curves before the correction for relativity. If they are truly superluminous supernova with extended lightcurves, this deeper assumption about the fundamental nature of the universe fails.

  16. #16
    Quote Originally Posted by Jerry View Post
    http://arxiv.org/pdf/1205.6973v1.pdf PHOTOMETRIC OBSERVATIONS OF THE SUPERNOVA 2009nr

    While the existance of superluminous supernova Ia is well-established, this event throws another wrench in use of normal events to calibrate cosmic parameters. ... [extra incorrect statements snipped]
    You've said this before. You were wrong the last time you said it, and you're wrong now.

    In case you didn't notice, in the paper you cite, they identify SN 2009nr as belonging to a particular class of luminous SNe Ia, identifiable from the light curve shape and color. Since we can identify the classes of SN Ia based on their light curve shapes, and correct the maximum luminosities using the "stretch" factor, this "problem" you keep mentioning doesn't exist. Your bringing it up again is just evidence that you never followed my reading suggestions in a number of previous threads, and thus still don't understand how supernova distance measurements are performed.

    But I assume you'll ignore or misunderstand me this time, as you have all the times in the past.

Similar Threads

  1. Replies: 0
    Last Post: 2012-Feb-13, 10:00 PM
  2. KOI 55, type sub dwarf B, supernova origins?
    By RogerG in forum Space/Astronomy Questions and Answers
    Replies: 1
    Last Post: 2011-Dec-26, 01:34 PM
  3. New Type of White Dwarf Stars Discovered
    By Fraser in forum Universe Today
    Replies: 15
    Last Post: 2008-May-02, 10:39 PM
  4. White Dwarf Question
    By kordic in forum Space/Astronomy Questions and Answers
    Replies: 9
    Last Post: 2007-Oct-17, 03:36 AM
  5. Urgent White Dwarf Question! :)
    By Jeff Root in forum Space/Astronomy Questions and Answers
    Replies: 10
    Last Post: 2007-Feb-19, 11:24 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
here
The forum is sponsored in-part by: