From: Alex Kim (agkim@lbl.gov)
Date: Mon May 17 2004 - 23:57:17 PDT
All,
Here is my understanding of Chris' point which is correct.
Since we are presenting Johnson-Cousins photometry, we need to use
Johnson-Cousins zeropoints and uncertainties in the K-correction. Since
Vega has the best calibrated spectrophotometry, the J-C zeropoints are
most reliably determined via Vega. Vega zeropoints are corrected for
the magnitude of Vega in the J-C system to give the J-C zeropoints. The
equation Chris gives is correct for the relation between J-C and Vega
K-corrections.
There are two sources of error; from the Vega magnitude in the J-C
system and from the Vega zeropoints. Both should be included in the
error budget. It should be noted that both errors are completely
correlated between supernovae that are K-corrected with the same filter
pair. Also, the Vega zeropoint errors in different filters are also
likely to be correlated.
Alex
Chris Lidman wrote:
>Hi Serena,
> I've though a bit more about the issue of cross filter k-corrections
>and I attach a short report. Alex, I'd like you to read this and to give
>us your opinion.
>
>Cheers, Chris.
>
>IR to Optical k-corrections
>===========================
>
>The cross filter k-correction is defined in Kim, Goobar and Perlmutter
>(1996)
>and is applied to apparent magnitudes according to
>
>m_y = m_x + K_xy(z)
>
>where x and y are different filters.
>
>Let's assume that z=0 and that Z=F. Let's further assume that we are
>using
>the Kurucz model for Vega. In this case K_xy(0) = 0 and, hence,
>
>m_y = m_x
>
>However this is not true for Vega for all x and y (i.e. all filters)
>
>m_V=0.026 +/- 0.008 (Bohlin and Gilliland, 2004)
>m_I=0.031 +/- 0.009 (Bessell, Castelli and Plez, 1998)
>
>m_J=-0.001 +/ -0.005 (Cohen, Wheaton and Megeath, 2003)
>
>where J is on the 2MASS system.
>
>If we set x to I and y to J
>
>Hence m_J=m_I-0.032.
>
>or m_I - m_J = 0.032 = -K_IJ
>
>I propose that we add an extra term to the cross filter k-correction as
>defined by Alex which reflects the colours of Vega. Note that Alex's
>definition is completely correct for an object in which all colours are
>zero.
>
>The new k correction is then
>
>K'_xy(z) = K_xy(z) - (x-y),
>
>where (x-y) is the colour of Vega and K_xy(z) is the k-correction
>defined by Alex.
>
>For the I band paper, we added this correction to the J band photometry.
>I think we should change this by changing the way we do the
>k-correction,
>as I have described above.
>
>Apart from the uncertainties described here, there is a systematic
>uncertainty
>in how well the Kurucz spectrum represents Vega. Bohlin and Gilliland
>quote an uncertainty of 2% from optical to IR wavelengths. We should add
>this systematic uncertainty to all IR to optical K corrections.
>
>Bessell, M. S., Castelli, F. and Plez, B. 1998, AA, 333, 231
>Bohlin, R. C. and Gilliland R. L. 2004, astro-ph/0403712
>Cohen, M., Wheaton, WM. A. and Megeath, AJ, 2003, 126,1090
>Kim, A, Goobar, A, and Perlmutter, S. 1996, PASP, 108, 190
>
>
>On Fri, 2004-05-07 at 19:59, Chris Lidman wrote:
>
>
>>Hi Serena,
>> I'd like to draw your attention to some nice work that has been done
>>on the absolute flux calibration of the Vega SED, which is central
>>to the way we compute cross-filter k-corrections and, in particular,
>>the corrections from IR to optical filters. See Bohlin and Gilliland
>>(astro-ph/040371) and Cohen et al. (AJ 126, 1090).
>>
>> The core of their results is that Vega has V-J = 0.026 +/- 0.008.
>>It is not 100% clear to me what this implies for IR -> optical
>>k-corrections, but I'll read these papers carefully over the
>>next week and I'll let you know what I think. It is interesting to note
>>that this number is not too far away from the ensemble of A0 stars
>>from the Hipparcos and 2MASS catalogs that were used to compute
>>the offset between IR and optical photometric systems - V_J-J_2MASS =
>>0.043
>>
>>Cheers, Chris.
>>
>>
>>
>>
>
>
>
This archive was generated by hypermail 2.1.4 : Mon May 17 2004 - 23:57:27 PDT