From: Isobel Hook (imh@astro.ox.ac.uk)
Date: Mon Feb 24 2003 - 10:08:03 PST
Hi Rob,
Here are my comments on the HST paper. Generally I think it's pretty
well-advanced and there's a lot of good stuff in there, so don't be put
off by the number of comments below! Many are just suggestions.
Talk to you on Thursday.
Cheers,
Isobel.
-------------------------------------------------------------------
Abstract: I think the first sentence is background info that
should go in the introduction. Suggest starting abstract with "We
present results from 11 SNe...and make the point that these give an
independent check..
Here and elsewhere it is a bit confusing when you switch from
discussing the non-matter component of the universe as Omega_Lambda to
w. In the abstract I would take out the part in brackets about
Omega_lambda because it's covered in the next sentence anyway (I
realise this comment is inconsistent with Saul's!).
Intro:
You say the previous results ruled out a flat, matter dominated
universe at extremely high confidence - it depends what you call
"dominated". I would say here "a flat universe with Omega_M=1".
Section on WFPC2 photometry: you discuss errors on redshift and then
say these are "dwarfed by photmetric errors". It looks a bit as if you
are saying photometric errors affect redshift determination! I suggest
you say that the uncertainties on redshift are not a significant
contribution to the uncertainties in the final cosmological
measurements (photometric uncertainties being dominant).
Text after Equation 1: suggest you spell out what the free parameters
are, i.e. x_0i, y0i, f_0i, zi (or new name for this as Saul
suggested). By the way, why is x0, y0 able to vary between images?
More general comment: Although each individual section of this paper
is well-explained, I think there is a slight lack of continuity
between the sections. A sentence spelling out what you actually do
with these fitted parameters would help - e.g. the fitted f_0i values
are then used in the lightcurve fits in section 2.2.
"A single Tiny Tim PSF.. was used for all images of a given band" -
again spell out what you mean, i.e. is this an assumed functional form
for psf(u,v) in equation 1?
Section 2.2 and end of section 2.1 - ground based data. Could we have
a table of these points too, or a statement about how many there are
reletive to HST points?
Section 2.2 first sentence: typo "first" > "fits"
Section 2.2 general - For much of this section it appears that you are
fitting lightcurves without any k-corrections at all - e.g. it says on
p7 "..A B-band template was fit to the R-band lightcurve". Insert
"observed frame", "k-corrected" and "colour-corrected" where
appropriate (referring to section 2.3 if necessary). In fact the first
couple of paragraphs of section 2.3 (which explain the need for k- and
colour corections) could perhaps be moved earlier, before you start
mentioning lightcurve fits.
Section 2.3: I wonder whether the details of determining intrinsic
colours could go in an appendix, to aid the flow of the main theme of
the paper (i.e. cosmological results). At a minimum I think it needs
its own section.
Para starting "The intrinsic colors of the supernova were determined
with the BVRI spectra range.." (couple of typos in there): I got quite
confused at this point. I wasn't even sure whether you were talking
about distant or low-z SNe!
I agree with Saul that an example figure (at least for us) might help
explain the ridge-line fitting.
Table 3: The column labels or notes need more explanation. e.g. m_X is
the PEAK magnitude in band X? (or the value at t_Bmax?). Don't
understand note 3 "This is just the measurement uncertainty" - isn't
this a measured colour? Note 5: This column is labelled
E(B-V)_host. The note says its the observed R-I colour.. isn't there a
factor R between the two? In the text is says these are the E(B-V)
values needed to reproduce the observed R-I colours (which makes more
sense).
Para starting "Once a template spectrum ..has been produced". I think
you could start a new section here (or return from the
appendix). Again, make it clear which SNe you are talking about now.
Section 3 "only with new lightcurve fits" > "except with new.." You
mention H99 (should be R99?).
Table 7, note: "This includes outliers as well as supernovae with
E(B-V)>3sig" - aren't these the same thing?
End of section 3.1 - this ends rather abruptly and could leave the
reader thinking that the P99 analysis was all wrong (it talked of a
"Bad E(U-B) determination" etc). I suggest adding a comment on the
possible size of the effect on the cosmological measurements in P99
(or just refer to fig 1 which shows it isn't that big an effect?).
Section 3.2. Suggest giving the equation for the lightcurve fit again
(as in P99). Define curly-M properly (be more quantitative than saying
it "contains the combined effect of H0 and the abs mag of a SNIa"). In
fig 3 you mention a value for M so it needs to be explained.
discussion of other uncertainties: "finally, only for fits which
explicitly included host galaxy extinction..." you mention how you
derive uncertainties in colours but it is not clear how these are
propagated into errors in peak mag.
Captions to Figs 5 and 6: Looks like some text is missing after a %
symbol.
Table 8 : Are the omitted SNe with E(B-V)>3sig from P99 or from this
paper?
End of section 3.2 - "The fits with extinction corrections applied
confirm and strengthen the conclusions of P99: The high-redshift
supernova data ... require Omega_Lambda >0 to >99% confidence". Which
figure does this statement correspond to? (fig 6 left panel?) If you
look at the right panel (which includes P99) or Fig 7 then the
confidence is only ~90%.
Section 4.1 - tidy up, e.g. "With and without host galaxy extinction"
need the word "correction" here and in several other places.
Fig 8 caption - this is pretty difficult to follow! State that the
last sentence is referring to shaded contours. Can you give an
idication of which is our best estimate?
Summary and conclusions: point 3 - evidence for Lambda at "very high
confidence". This is only true for a flat universe. Otherwise I think
"very" high is an overstatement (it's a ~2-4 sigma result depending on
which plot you look at).
This archive was generated by hypermail 2.1.4 : Mon Feb 24 2003 - 10:08:11 PST