From: Greg Aldering (aldering@panisse.lbl.gov)
Date: Sat May 03 2003 - 19:17:52 PDT
Hi Rob,
I agree with you that there isn't an obvious pattern in the
ground-based R-band lightcurve residual that would point to the
calibration. In parallel with your study, I examined which nights went
into calibrating each SN using calibnights and tracker. There are cases
where the same or similar calibration was used on SNe which are
well-fit and those that are not; thus logic would suggest calibration
can't cause the main problem. For completeness, I present this work
below.
Below I have assembled the zeropoints of all the R-band calibration in
the calibration database. I list the zeropoint for each BTC chip, and
list the UT date (not evening date) of the calibration. (You will
notice the large change in zeropoint for BTC chip's 2 and 3 between
1998 and 1999 - this is real as the CCD's were changed.) I then
examined which SNe were observed on which chips on nights that were
calibrated. The chip number is listed under each SN for each
calibration night.
98as, 98as, and 98ax were considered as having good fits, based on my
earlier "by eye" analysis of the R-band lightcurves. Therefore one of
things I looked for was evidence that those SNe came from different
calibrations than the SNe with bad fits. For those, I placed an "*"
next to the chip number. However, if a calibration was used for both a
good- and bad-fit SN, I placed an "o" next to the chip numbers.
You can see that the bad-fit SNe, 97eq, 98be, and 97ek, use calibration
that was also used to calibrate SNe with good fits. This suggests that
calibration can't be the main issue.
Zeropoints by BTC CCD <- good fits -> <---- "bad R-band fits ---->
1 2 3 4 UT date 98as 98aw 98ax 97eq 98be 98ay 98bi 97ez 97ek
-----------------------------------------------------------------------------
26.54 26.27 26.69 26.72 2 10 97 0 0 0 0 0 0 0 0 0
26.53 26.27 26.69 26.72 2 11 97 0 0 0 0 0 0 0 0 0
26.46 26.15 26.57 26.62 3 06 97 0 0 0 0 0 0 0 0 0
26.49 26.17 26.59 26.67 3 10 97 0 0 0 0 0 0 0 0 0
26.67 26.12 26.72 26.57 11 28 97 0 0 0 4* 0 0 0 3* 1*
26.59 26.18 26.57 26.56 3 01 98 4 4 3 0 4o 2* 2* 0 0
INDEF 26.68 INDEF 26.64 1 15 99 4 0 0 4o 0 2* 0 (3) 4o
26.60 26.52 26.90 26.64 4 12 99 0 4 3 0 0 1* 0 0 0
Thus, my more limited study agrees with your conclusion that ground-based
calibration isn't the solution. Since you noted a clear correlation with
redshift, that seems to be the place to focus.
Even so, somewhere in the paper we need to clearly demonstrate that the
ground and space calibrations are consistent. This would be expected
even if there were no apparent R-band mis-fits. Can you run a
lightcurve for even one star that is measurable on both the ground and
HST images? If that lightcurve is flat, that would be quite convincing!
Cheers,
Greg
This archive was generated by hypermail 2.1.4 : Sat May 03 2003 - 19:17:53 PDT