Systematics summary - update

From: Isobel Hook (imh@astro.ox.ac.uk)
Date: Sun Mar 09 2003 - 06:31:09 PST

  • Next message: Robert A. Knop Jr.: "parsing weird Riess parameters"

    Hi,

    Here is an updated summary of systematics following the telecon last
    Friday. Note that we all generally agree with Ariel's comment that we
    should concentrate on effects new to this paper, especially effects on
    colours. The expectation is that most of the list below will generate a
    few sentences in the paper saying that the effect is similar to that in
    P99 - but we do need to check that is correct for each item below. Also it
    would be good to advertise our other recent papers such as the rise-time
    paper and the work on host galaxy types (Sullivan et al).

    Isobel.

    ----------------------------------------------------------------

    Summary of systematic effects v2. Updated with comments from telecon
    on 07/03/03.

    4.1 Extragalactic extinction
      - thouroughly covered in HST paper

    4.2 Malmquist bias
      - What matters is the difference in Malmquist bias between low and
        high-z sets. HST paper uses new data for both low- and high-z SNe.

        High-z: P99 estimated size of effect by considering completeness
                and associating this with SN number counts as a fn of mag
                (actually it's not clear how this was done!). Also took
                into account non-perfect correlation between discovery mag
                and peak mag.

                07/03/03: Greg will recalculate correlation, and will look
                into method from last time. Need to check how we used the
                efficiency measurements in P99 and whether we need to redo
                that. [Greg]

        Low-z: P99 assumed similar effect but correlation between peak and
                discovery mags is stronger.

                07/03/03 Greg: Reiss sample probably less malmquist biased
                because included heavily reddened ones. Sullivan et al
                paper claimed Riess sample are fainter than Hamuy sample.

       Other malmquist-bias related checks
           Include comparison of stretch distributions as in P99 [Rob]
           Redo cosmology fits with and without new R99 [Rob]

    4.3 Gravitational Lensing

         07/03/03: Greg states that uncertainties in the details of the
         model for the form of lensing (machos, empty beam etc) dominates
         over the small change in z range of HST SNe compared to P99.
         Need a sentence in the text saying this [Greg?]

    4.4 SN evolution
       
       - Spectra consistent at low & high-z (same argument as P99)
     
      - lightcurve rise-time & shape similar (same argument as P99 but
         update based on Gerson's paper) > constraint on changes in
         width-luminosity relation. [Gerson/Peter/Greg?]

         07/03/03: simplest test is to check the current HST ones are
         consistent with template from recent paper [Don?]
         Also, can we state that the current set does not constrain
         rise-times compared to previous paper?? CHECK [who???].

       - Small evolutionary effects on rest-frame U-band (cf Hofflich's
         paper) could affect colour & k-correction. P99 argument needs
         updating since more of the HST data sample the
         U-band. [Greg/Rob/Peter?]

         07/03/03: Metallicity effect on selection: cf Hamuy et al (host
         paper from 2001). Also cite Nomoto paper limit on metallicity.
         [Greg].

       - Mention constraints from Sullivan et al. paper: e.g. constraints
         on variation of width-luminosity relation with environment
         [Mark?] Note we don't have host types for these ones
         (yet). Someone needs to work out what we can say about
         this. [Mark?]

    4.5 Further checks

      - Sensitivity to width-luminosity relation. Could redo fit without
        using stretch correction? (cf fit D in P99) or at least show
        stretch distribution. [Rob]

     - Sensitivity to non-Ia contamination.
      
       07/03/03 There is one SN from P99 that we now think is not a Ia and
       4(5?) more we don't have good enough info. Throw these out now
       for main result, and do cross check keeping them in [Rob]. Need a
       sentence explaining more conservative strategy compared to P99 now
       that we have enough data to do this [Saul?]

     - sensitivity to galactic extinction model: P99 re-did fits using
       Burstein & Heiles extinction model. Not necessary now? (or just
       re-quote numbers from P99?)

       07/03/03: Greg says that if anything, HST ones have lower galactic
       extinction because better field selection (higher resolution maps
       were available). Need a sentence about this in the paper [Greg]

    -------------------------------------------------------------------------



    This archive was generated by hypermail 2.1.4 : Sun Mar 09 2003 - 06:31:13 PST