From: Andy Howell (howell@astro.utoronto.ca)
Date: Thu Apr 29 2004 - 11:10:25 PDT
On 29 Apr 2004, Chris Lidman wrote:
> Hi Andy,
> OK, we'll leave it as unclassified.
>
> I do not want to use photometric information to help in the
> classification because the photometric data is very heterogeneous
> and the current analysis of the data is even more so.
Ok, but we should note that in some cases, such as this one, the
classification is borderline, and with a few extra pieces of
information, such as that provided by the lightcurve, the classification
could be tipped one way or the other.
> The classification scheme you use for the legacy survey has merit. I
> like it. However, I do not propose that we adopt it for the VLT paper.
>
> If I was to map what we have used for the VLT paper then
>
> a VLT Ia is somewhere between 1 and 2 on the legacy scheme of
> classification
>
> a VLT Ia? is somewhere between 2 and 3 on the legacy scheme
>
> a VLT - (unclassified) is not on the legacy scheme.
>
> Would you agree with this assessment?
Pretty much, but you have the scale backwards. I arbitrarily chose 5 as
the "certain Ia". But yes, you are interpreting it correctly. I usually
also do say Ia or Ia? Sometimes a 4 can be a "Ia" or a "Ia?" However, if
I were classifying the VLT SNe for our system, I would probably classify
the FBTs as 2-Probably not Ia, or 1-Not Ia, depending on how sure I was a
true spectrum of the transient (and did not have host contamination).
> Cheers, Chris.
>
> On Thu, 2004-04-29 at 10:02, Andy Howell wrote:
> > Chris,
> >
> > >From the spectrum alone I just would not say it reaches the level I would
> > call a "Ia?". So if these classifications are from spectra alone, I would
> > say no. I am just imagining someone from the other team, looking at that
> > figure with skeptical eyes, thinking (and saying loudly at every possible
> > collection of astronomer) that we have no idea what we are doing. However,
> > if you are willing to let other factors slip in, then that might change
> > the classification as long as we are very up front about that.
> >
> > For the Legacy survey, I have a 5 level classification scheme:
> > 5: Certain Ia
> > 4: Probable Ia (Ia?)
> > 3: Possible Ia, but no strong evidence
> > 2: Probably not a Ia
> > 1: Certainly not a Ia
> >
> > In this scheme, C02-028 is a "3".
> >
> > Do you have the light curve? I could fit it with our software here and
> > see how closely it matches a Ia. If the light curve fits a Ia, and we
> > can't rule it out as being a Ia from the spectrum, then I am willing to
> > classify it as something like Ia?* where the "*" means we used the
> > lightcurve and the fact that the spectrum was heavily contaminated by
> > galaxy light. Or Rob or Alex C. could probably run a quick SNMinuit fit
> > if you want to keep everything done with SCP tools.
> >
> > We should be very clear and well defined about what our classifications mean.
> > Because I have seen in the past that as soon as someone calls something a
> > "Ia?" for any reason it is lumped in with the "Ia"'s and used in every
> > plot imaginable. Then the reason it was classified that way gets
> > forgotten.
> >
> > -Andy
> >
> > On 28 Apr 2004, Chris Lidman wrote:
> >
> > > Hi Andy,
> > > Thanks for re-doing the fit. Do you think it is reasonable if we
> > > classify C02-028 as Ia?
> > >
> > > Cheers, Chris.
> > >
> > > On Wed, 2004-04-28 at 18:25, Andy Howell wrote:
> > > > Chris,
> > > > I re-ran C02-028 as you suggested, with no color mangling allowed.
> > > > You do have to subtract about half Sa galaxy to get the color to
> > > > match though. See attached comparison to a Ia at +5d.
> > > > I am still not convinced it is a Ia. It is not out of the question
> > > > though.
> > > >
> > > > -Andy
> > > >
> > > > On 26 Apr 2004, Chris Lidman wrote:
> > > >
> > > > > Hi Andy,
> > > > > I've been having a closer look at C02-028. This one is currently
> > > > > classified as "?".
> > > > >
> > > > > Could I ask you to try this one again, but with the extinction turned
> > > > > off. It will be difficult because it is dominated by the host. The
> > > > > reason why this could be a SN~Ia is that:
> > > > >
> > > > > i) The H and K absorption from the galaxy do not form a clean
> > > > > 4000 Angstrom break. There might be some light from the blue
> > > > > emission wing of the CaII feature of the SN.
> > > > >
> > > > > ii) A bump at 6600 Angstroms.
> > > > >
> > > > > iii) A reasonable looking light curve. See the attached plot. (You will
> > > > > see from the tiles why this was a difficult one to do spectrally.)
> > > > >
> > > > > The spectrum was taken on May 18th, about 4 days past Maximum light.
> > > > >
> > > > > Perhaps we can get a better match with these constraints.
> > > > >
> > > > > Cheers, Chris.
> > > > >
> > > > > On Mon, 2004-03-08 at 17:53, Andy Howell wrote:
> > > > > > Chris,
> > > > > >
> > > > > > I have mande several improvements to my program recently, most notably
> > > > > > better outlier rejection and better handling of error spectra, so it
> > > > > > certainly is worth rerunning objects for which we have questions. In these
> > > > > > plots, blue is the rebinned spectrum after host galaxy subtraction
> > > > > > (usually 5 A binning for these plots), black is the low-z template, and
> > > > > > green is the original data.
> > > > > >
> > > > > > I have attached 5 postscript plots of the best fits to the 4 SNe you sent.
> > > > > > > S01-004 - currently classified as ?
> > > > > >
> > > > > > Rerun with no host. The fit here shows a comparison to SN 1993J at +17d
> > > > > > at z=0.41 (z is not known from the host). The fit isn't that bad. Do we
> > > > > > know the date our spectrum was taken with respect to maximum light? I'm
> > > > > > sure +17d is too late, but these IIb's have different features come in at
> > > > > > different times depending on the relative thicknesses of the H and He
> > > > > > envelopes. I wouldn't say that it is either a II or Ib for certain, but I
> > > > > > think it is unlikely that it is a Ia. How about (IIb?) for the
> > > > > > classification?
> > > > > >
> > > > > > > S01-028 - currently classified as ?
> > > > > > > Significant host contamination. The percentage increase is only 27%.
> > > > > > > Previously, we had this one as Ia?
> > > > > >
> > > > > > As you say, the host contamination is pretty bad. Here I show a
> > > > > > comparison to SN 1999ee (Ia) at -8d after subtraction of an SB6 host
> > > > > > galaxy. I find it very unconvincing. CaII should be obvious but is
> > > > > > not. Still "?" as far as I'm concerned.
> > > > > >
> > > > > > > C02-028 - currently classified as ?
> > > > > > > Another one with significant host contamination. The percentage increase
> > > > > > > is only 13%.
> > > > > >
> > > > > > C02-028_spec.1.ps shows the comparison to SN 1999aa at -1d.
> > > > > > The blue side looks ok, and it is possible that the feature at 9000A is Si
> > > > > > 6150. The problem with this interpretation is that the SII "W" at rest
> > > > > > 5400A should be there, but it isn't! Neither is the "emission" (really
> > > > > > lack of absorption) just blueward of it at rest 5100A.
> > > > > >
> > > > > > C02-028_spec.18.ps shows a comparison to SN 1987K (II) at +7d after some
> > > > > > serious host galaxy subtraction and mangling of the color. I don't
> > > > > > believe it either, but it shows the difficulty in narrowing down the type.
> > > > > >
> > > > > > > SuF02-002 - currently classified as ?
> > > > > > > This one has some host contamination. Previously we had Ia?
> > > > > >
> > > > > > Ugh. This one is disgusting. About the only thing I can say is that it
> > > > > > is not inconsistent with a Ia. But on the other hand, there isn't
> > > > > > anything to make me believe it is a Ia either!
> > > > > >
> > > > > > -Andy
> > > > >
> > >
> >
> >
>
This archive was generated by hypermail 2.1.4 : Thu Apr 29 2004 - 11:10:38 PDT