From: Robert A. Knop Jr. (robert.a.knop@vanderbilt.edu)
Date: Tue May 20 2003 - 06:36:49 PDT
I've attached a plot that shows the old and new K-corrections, and how
much the K-corrections change given an E(B-V) of +-0.2 (an extreme
outside error on "reasonable" E(B-V) values-- all the HST ones are at
least a factor of 2 better than this; 15 of the P99 SNe are this bad or
worse (a couple much worse).
At z=0.3, the EBV uncertainty this large would put a +-0.1 mag
uncertainty on the K-correction. AT z=0.6, it's about +-0.08 mag. This
is still much smaller than the 0.17 intrinsic dispersion (and hugely
smaller than the propogated magnitude uncertainty from the EBV
dispersion itself), although it's about at the point where we should be
smarting a bit for not including it for those worst SNe. For the HST
SNe, it's not a worry.
The only real issue is the really bad ones, like 9739. Probably the
right thing to have done would be to have thrown them out as having "no
reasonable color measurment", as I did with 976; a cut at d(R-I)<0.25
would throw out 5 more of the P99 SNe, including 9739. On the other
hand, it means redoing everything yet again... and I suspect they aren't
making that much difference in the final cosmology. If we can know
internally what's going on and be aware of it, but leave them in the
paper, that's easiest. If we want to make this color-error cut to get
ride of the ones with really bad colors, though, I can do it.
As for the systematic offset of 0.04 magnitudes at z=0.7 between the old
and new K-corrections, I'm not too worried about that. We should get
Peter back in this conversation, as we had that one a month ago; I think
we really do believe the new K-corrections are better.
-Rob
-- --Prof. Robert Knop Department of Physics & Astronomy, Vanderbilt University robert.a.knop@vanderbilt.edu
This archive was generated by hypermail 2.1.4 : Tue May 20 2003 - 06:36:51 PDT