Re: comments on May 7th draft - round 1

From: Greg Aldering (aldering@panisse.lbl.gov)
Date: Wed May 14 2003 - 10:02:05 PDT

  • Next message: Greg Aldering: "f(OM,OL,w) for galaxy redshift distortion"

    >On Tue, May 13, 2003 at 11:15:03AM -0700, Greg Aldering wrote:
    >> a) Follow Phillips 1999 in throwing out very red objects even though
    >> you are going to extinction correct. I believe that Phillips
    >> threw-out objects with Bmax-Vmax > 0.2 mag. (Talk about dirty
    >> laundry!)
    >
    >It's not clear to me that he did this when finding his RB values. He
    >did do that for a bunch of the analysis, but for the RB stuff
    >specifically he only talks about throwing out the SNe without a point
    >near max. I believe that his value *is* with the most reddened SNe (he
    >certainly plots those), and as such throwing out the most reddened SNe
    >from our set isn't a matter of saying "just as Phillips"....

    Correct. Ok, then just throw out the red ones, as Phillips did in his
    analysis and use the regular R_B. From my fitting, I find that when the
    reddest objects (E(B-V) > 0.2) are thrown out, our SNe like R_B =
    3.3+/-0.5. The chi^2 is only worse by 2.5 when R_B = 4.1 is used. This
    means that we should get a good fit using the canonical R_B if we
    exclude the red objects. In this case, I believe we can simply cite
    Phillips 1999. (By the way, you can see in figure 10 of Phillips that
    his red objects systematically prefer a smaller R_B.)

    Note that there is an old and ugly history of low R_B's having been
    suggested for SNe, prior to the advent of width-luminosity relations.
    So using the canonical R_B is definitely the safest thing to do,
    otherwise older reviewers could really light into us. Challenging the
    canonical R_B requires a very high level of proof that all the other
    steps (like determination of the fiducical color) are unassailable. On
    a related note, Lifan says that the CMAGIC analysis likes R_B = 4.1.

    >The expression I'm using is:
    >
    > B-V = -0.076 + 0.0446 * (1/s - 1)
    >
    >I think the numbers you used in P99 were -0.11 and 0.243 respectively.
    >This is a redder central color and a milder slope, but looking at the
    >fits I couldn't justify the bluer color, and the slope wasn't as clear
    >as all that. (Indeed, the uncertainty on the slope is 0.037, nearly the
    >full size of the slope.)

    Concerning your shallow color-versus stretch relation, have you compared
    it to equation 23 of Phillips 1999? He finds:

      (Bmax-Vmax)_0 = (Bmax-Vmax) - 0.114*(dm15 - 1.1)

    Using equation 5 of P97 (dm15 = -1.96*(1/s-1) + 1.07), this becomes:

      (Bmax-Vmax)_0 = (Bmax-Vmax) - 0.114*(1.96*(1/s-1) + 1.07 - 1.1)
                      (Bmax-Vmax) - 0.114*(1.96*(1/s-1) + 1.96*0.114*1.07 - 0.114*1.1)
                      (Bmax-Vmax) - 0.223*(1/s - 1) - 0.11

    That is, Phillips 99 gets a relation almost identical to what we used
    in P99. I know that you are using (B-V)@Bmax and not Bmax-Vmax, but
    still, I'm surprised that the dependence of color on stretch basically
    vanishes in your analysis.

    To make some progress here, can't you plot your measured B-V vs stretch
    for SNe in E/S0 hosts, and overlay your color-stretch relation to look for
    consistency? I think that would settle this issue.

    - Greg



    This archive was generated by hypermail 2.1.4 : Wed May 14 2003 - 10:02:06 PDT