Re: poor agreement with Riess 99

From: Robert A. Knop Jr. (robert.a.knop@vanderbilt.edu)
Date: Wed May 07 2003 - 18:23:25 PDT

  • Next message: Greg Aldering: "Re: poor agreement with Riess 99"

    On Wed, May 07, 2003 at 06:14:36PM -0700, Greg Aldering wrote:
    >
    > Hi Rob,
    >
    > I started to investigate "what'sup" with SN1995bd and SN1996bo and
    > found that your B-band magnitudes differ, sometimes by a lot, from the
    > published Bmax tables from Table 3 of Riess 1999 (the 22 SNe paper).
    > Further investigation showed generally poor agreement. Can you
    > comment?

    Duuno. I do remember noticing that the template we have didn't fit
    several of the Riess SNe right at max that well-- there are a few Riess
    SNe which are "overluminous" at max. Some Hamuy's also. I'm not sure
    offhand which ones they are; I saw this fitting the full set, not just
    the set we use for cosmology. You can look at the snminuit output files
    to see how the set I have compares. If I get time, I will make a
    summary plot of all the low-z lightcurve fits so you can glance at them
    all at once.

    Note that the MX values are not K-corrected, and are not corrected for
    Galactic extinction. Those are the "raw" peak magnitudes from the fit.
    You need to take all that other stuff into account when comparing MX to
    anything. (Some of the riess SNe have substantial Galactic extinction.)
    I generally just don't use MX from the gersontables; by the time I'm
    reading that table, I'm using the mb value which has Galactic extinction
    factored into it.

    When all is said and done, *unless* there's a serious zero-offset
    missubtraction in the quoted riess data, the magnitudes I have fit are
    more likely to be the "right" ones to use with our high-z supernovae
    than the ones he has fit. We each used different templates and
    different methods. Moreover, the templates that I have did fit well
    after max, it was only at max that some of them had overluminous points.
    Since most of the statistical weight for most of the HST supernovae
    comes after max, that's where you want to have the fits comparing. *If*
    we believe the templates we have, then whatever the peak template from
    the fits after max yield are the right things to use.

    In any event, I'm not going to worry about this right now, because I
    simply can't... I'm trying to get enough things done so that I can
    dissapear tomorrow. Check to see how different the Hamuy magnitudes are
    from the P99 ones. If they aren't that different, then except for our
    two outliers the R98 fits can't be that wrong, because they line up
    fairly well on the same Hubble line.

    -Rob

    -- 
    --Prof. Robert Knop
      Department of Physics & Astronomy, Vanderbilt University
      robert.a.knop@vanderbilt.edu
    


    This archive was generated by hypermail 2.1.4 : Wed May 07 2003 - 18:23:26 PDT