Marc on the zero-point issue:
The photometric zero-points of the INT images (Harris-B and Harris-R filters) were calculated on the basis of the SDSS ugriz photometry of 120 non-saturated stars scattered across the field of A963, and 140 non-saturated stars scattered across the field of A2192.
From the INT images, curves-of-growth (CoGs) were constructed out ot 15“ radius using ELLINT in the GIPSY package, using the same parameters for the B- and R-band images. Each stellar CoG was inspected visually and the radial range representative of the sky background was determined. Instrumental magnitudes (B_inst, R_inst) for all the stars were calculated from the background-subtracted integrated CoGs.
For the SDSS photometry, the psf-magnitudes of the stars were used (pfsMag_ugriz). These SDSS magnitudes (AB mags) of the stars were converted to B- and R-magnitudes using the transformation formula of Lupton (2005) as stated at http://www.sdss.org/dr5/algorithms/sdssUBVRITransform.html meant to match Peter Stetson's published photometry which is tied to the Landolt UBVRI standard stars:
B_J(Vega) = g + 0.3130*(g-r) + 0.2271 = B_sdss (Johnson B in the Vega system)
R_C(Vega) = r - 0.2936*(r-i) - 0.1439 = R_sdss (Cousins R in the Vega system)
We calculated the zero-points as:
B_zp = B_sdss - B_inst
R_zp = R_sdss - R_inst
and plot these versus instrumental colors (click on plot for larger version):
This photometry of the stars yields the Vega zero-points of:
k1_B = 31.969 and k1_R = 31.739 for Abell 963
k1_B = 32.078 and k1_R = 31.721 for Abell 2192
Note that we have ignored the color-terms, i.e. the slopes in these correlations.
Instrumental INT magnitudes of the galaxies were derived from the B/R_FLUX_AUTO provided by SExtractor according to
B_inst = -2.5Log(B_FLUX_AUTO) → m(B) = B_inst + k1_B
R_inst = -2.5Log(R_FLUX_AUTO) → m(R) = R_inst + k1_R
These are thus galaxy magnitudes (Vega) within a Kron-like elliptical aperture.
INT zero-point issue (Yara):
Because the INT images were collected with the B and R filters during 3 non-photometric observing runs, we have to photometrically calibrate the data after the fact. For this, we are using the total de-reddened SDSS magnitudes of all the spectroscopically confirmed galaxies and find that actually its not so bad (see plots)
The way we do it is to convert SDSS dered (AB) mags into B and R (still AB):
R=dered_r-0.0576-0.3718*((dered_r-dered_i)-0.2589)
B=dered_g+0.2354+0.3915*((dered_g-dered_r)-0.6102)
I took these formulas from Blanton & Roeis 2007
then convert to Vega:
sdss_R_vega=sdss_R_AB-0.21
sdss_B_vega=sdss_B_AB+0.09
and compare with INT mB and mR (that are in Vega):
Delta_R = sdss_R_vega - mR
Delta_R = sdss_B_vega - mB
In these plots the INT mB and mR were NOT corrected for extinction in our Galaxy (whilst the SDSS mags WERE) but even if they are not, the correction is tiny… The Galactic extinction values from Schlegel at el as reported by NED are:
- A963: A_B=0.064 and A_R=0.039
- A2192: A_B=0.046 and A_R=0.028
Thanking all of this into consideration, the final magic formulas are:
R_sdss,vega,dered = r - 0.0576 - 0.3718 * [(r-i) - 0.2589] - A_R - 0.21
B_sdss,vega,dered = g + 0.2354 + 0.3915 * [(g-r) - 0.6102] - A_B + 0.09
R_int,vega,dered,corrected_zp = R_int,vega + 0.058 - A_R
B_int,vega,dered,corrected_zp = B_int,vega + 0.101 - A_B
— Yara Jaffe 2012/08/22 16:01