Wiki » History » Version 21
Version 20 (Shantanu Desai, 01/06/2013 04:07 PM) → Version 21/88 (Shantanu Desai, 01/06/2013 04:07 PM)
h1. Wiki
This can be divided into 3 steps
h2. Step 1 run modelfitting photometry
you can find the scripts at /big/users/shantanu/VLTmfp/catalogs/runmfp_b.sh (for b band and so on)
note that since initial catalogs will be uncalibrated , use -mag_zeropoint for b,I and z bands.
The zeropoints for HST bands is :
F606W 26.493mag, F814W 25.947mag.
the sextractor param file used in the scripts is
/big/des_home/stable/etc/sex.param_sersic
you can edit to remove params you don't need.
h2. Step 2 Produce a matched catalog across all 4 bands
This is the most tedious step. Unfortunately you cannot run sextractor in dual image mode
since coadds are of different size.
The way I did it was to run a modified version of a code
used in DESDM called coadd_grabmatches. I first ran it to match R vs I band and then used
that match catalog to match against z band and so on.
A code to do this is at /home/moon/shantanu/desdm/devel/ImageProc/trunk/src/coadd_grabmatches_singleband.c
There is one line in the code (below # 266) which needs to be changed depending upon number of coluns
in input catalog.
Once this is done you need to make a file with the matched magnitudes in all four bands
(its possible to automate this script with some effort so that it does not need to be changed)
h3. Step 3 Doing the stellar locus
One you have the catalogs all you need to do is to do a fit by eye or quantitative to SED fits provided by Joerg.
Here is my 2nd order polynomial curve fit done to these relations
1) b-R vs R-I
b -R = -0.2904*x^2+1.319x +0.2941
(where x=R-I)
2) R-I vs I -z
R-I = -1.016*x^2+ 3.532*x +0.1294
where x=I-z
3) b-R vs I-z
R-I = -2.20378*x^2+ 3.96368*x +0.386878
You can either use (1) & (2) or (2) & (3) to get the zeropoints.
I did the plotting in ROOT, but probably it can be done in any software.
h4. Step 4 Rerun sextractor with the new zeropoints
Same as (1) but use the new zeropoints now.
h3. Direct calibration of I-band
Joerg pointed out that I band can be directly calibrated using
I_BESS = a0 + a1 * F814W + a2 * (F606W - F814W)
with
a0 = -1.009
a1 = 1.067
a2 = -0.001
!{width: 10%}Izeropoint.png!
This allows us to get direct calibration of I-zeropoint= 28.37+/- 0.02 (where error = sigma of fit/sqrt(no of points))
We then recalibrated (b-R) vs (R-I)
!{width: 10%}plotslrfit.png!
Also a histogram wrt stellar locus
!{width: 10%}slrhisto.png!
the zeropoint for b band = 26.88
h1. Results for other clusters
h2. SPT-CLJ0102-4915
I band zero-point = 28.52 +/- 0.0137
b band zp = 26.87 +/- 0.016
This can be divided into 3 steps
h2. Step 1 run modelfitting photometry
you can find the scripts at /big/users/shantanu/VLTmfp/catalogs/runmfp_b.sh (for b band and so on)
note that since initial catalogs will be uncalibrated , use -mag_zeropoint for b,I and z bands.
The zeropoints for HST bands is :
F606W 26.493mag, F814W 25.947mag.
the sextractor param file used in the scripts is
/big/des_home/stable/etc/sex.param_sersic
you can edit to remove params you don't need.
h2. Step 2 Produce a matched catalog across all 4 bands
This is the most tedious step. Unfortunately you cannot run sextractor in dual image mode
since coadds are of different size.
The way I did it was to run a modified version of a code
used in DESDM called coadd_grabmatches. I first ran it to match R vs I band and then used
that match catalog to match against z band and so on.
A code to do this is at /home/moon/shantanu/desdm/devel/ImageProc/trunk/src/coadd_grabmatches_singleband.c
There is one line in the code (below # 266) which needs to be changed depending upon number of coluns
in input catalog.
Once this is done you need to make a file with the matched magnitudes in all four bands
(its possible to automate this script with some effort so that it does not need to be changed)
h3. Step 3 Doing the stellar locus
One you have the catalogs all you need to do is to do a fit by eye or quantitative to SED fits provided by Joerg.
Here is my 2nd order polynomial curve fit done to these relations
1) b-R vs R-I
b -R = -0.2904*x^2+1.319x +0.2941
(where x=R-I)
2) R-I vs I -z
R-I = -1.016*x^2+ 3.532*x +0.1294
where x=I-z
3) b-R vs I-z
R-I = -2.20378*x^2+ 3.96368*x +0.386878
You can either use (1) & (2) or (2) & (3) to get the zeropoints.
I did the plotting in ROOT, but probably it can be done in any software.
h4. Step 4 Rerun sextractor with the new zeropoints
Same as (1) but use the new zeropoints now.
h3. Direct calibration of I-band
Joerg pointed out that I band can be directly calibrated using
I_BESS = a0 + a1 * F814W + a2 * (F606W - F814W)
with
a0 = -1.009
a1 = 1.067
a2 = -0.001
!{width: 10%}Izeropoint.png!
This allows us to get direct calibration of I-zeropoint= 28.37+/- 0.02 (where error = sigma of fit/sqrt(no of points))
We then recalibrated (b-R) vs (R-I)
!{width: 10%}plotslrfit.png!
Also a histogram wrt stellar locus
!{width: 10%}slrhisto.png!
the zeropoint for b band = 26.88
h1. Results for other clusters
h2. SPT-CLJ0102-4915
I band zero-point = 28.52 +/- 0.0137
b band zp = 26.87 +/- 0.016