Photometry with technique of Template Subtraction: personal experience.

1
Photometry with technique of Template
Subtraction personal experience.
Stefano Valenti Physics Department of Ferrara
INAF Astronomical Observatory of
Capodimonte Padova-Asiago Supernova Group
(http//web.pd.astro.it/supern/)
2
Other techniques

• Aperture Photometry
• Isolated object
• Photometry with PSF Fitting
• Not isolated object
• measure the PSF
• Fit the object with the PSF

Mark A
sn2004gt
3
PSF
--------------------------------------------------
--------- on GRAPHIC display gt a lt to accept
star, gt d lt to delete on IMAGE display gt f
lt to fit PSF, gt w lt to write PSF fit, gt q lt
to quit ------------------------------------------
--------------------- Computing PSF for image
temp_SN97ab_281299_R 9 stars read from
temp_SN97ab_281299_R.mag Star 1 rejected by
user Star 2 has been added to the PSF star list
X 380.00 Y 166.00 Mag -7.644 Dmin
6078.78 Dmax 36255.29 ... Star 9 has been
added to the PSF star list X 395.00 Y
827.00 Mag -4.913 Dmin 6067.815 Dmax
8494.79 8 PSF stars read from temp_SN97ab_281299_
R.mag Fitting function gauss norm scatter
0.02782232 Analytic PSF fit Function gauss
X 375. Y 537.5 Height 135507.7 Psfmag
-7.644 Par1 2.375538 Par2
1.847269 Computed 1 lookup table(s) Writing PSF
image temp_SN97ab_281299_R.psf.fits Writing
output PSF star list temp_SN97ab_281299_R.pst Writ
ing output PSF star group file temp_SN97ab_281299_
R.psg
PSF (noao.digiphot.daophot.psf)
allstar (noao.digiphot.daophot.allstar)
PSF, Magnitude of field stars, Aperture correction
ecpsf (package of SNOOPY) Cappellaro Patat
Ledan software
4
PSF PHOTOMETRY OF SN
allstar (noao.digiphot.daophot.allstar)
PSF used for the fit
Difference image
Residual image
gtgtgt input file ? (sn04gt_030105_R) gtgtgt
reference FWHM ? (10.) gtgtgt aperture correction
(ap-fit) ? (0.023) background
fit gtgtgt mark corners of
bg-region with gtblt, exit gtqlt Type b again to
draw box. gtgtgt Order of function in x for bg fit ?
(7) gtgtgt Order of function in y for bg fit ?
(7) gtgtgt background fit OK ? (yes) gtgtgt
recentering for targets ? (yes) gtgtgt Iterate
on background ? (no) gtgtgt Not yet happy ? Wish
to adjust manually stellar peak ? (no) gtgtgt
Errors estimate (through artificial star
experiment ?) (no)
id x_ori
y_ori x y ap_ori ap_bgsub fit_mag
err_art err_fit SN1 395.617 737.999 52.617
51.999 -9.708 -8.809 -8.849 0
0.009

sn2004gt
SN 2000cf
Original image
Residual image
PSF used for the fit
ecsn (package of SNOOPY) Cappellaro Patat
Ledan software
5
When we use template subtraction ?

• Template subtraction
• Not isolated object
• Faint object
• Scientific image, Scientific image without
  object, tool to make difference image, PSF,
  aperture correction

SN1997ab
SN1997ab
HII region
HII region
?
SN1997ab
01-02-1998
12-12-1999 (filter R)
12-12-1999 (filter B)
6
Result
Host Galaxy SN1997ab 15-03-2004 Ekar
SN1997ab 28-12-1999 Danish

• Aperture photometry
• Photometry with PSF fitting

• Geometric Transformation
• To be compare the PSF of the two images
• To be compare the Flux scale of the two images

difference image
7
Subtraction with ISIS (or DIA)
ISIS C. Alard http//www2.iap.fr/users/alard/pa
ckage.html
DIA P.R. Wozniak http//www.astro.princeton.edu
/wozniak/dia/
Ref Reference Image (best seeing) Im Program
Image Ker Kernel image (to be found)
the flux scaling factor is the sum of the local
kernel (C. Alard, 2000)
the result is on the scale of the program image.
I divide it by the norm of the kernel and express
the flux in the units of the reference image(P.
Wozniak, private communication)
the result is on the scale of the program image.
I divide it by the norm of the kernel and express
the flux in the units of the reference image(P.
Wozniak, private communication)
the result is on the scale of the program image.
I divide it by the norm of the kernel and express
the flux in the units of the reference image(P.
Wozniak, private communication)
(C. Alard 2000 P.R. Wozniak 2001)
8
Subtraction between two images ISIS or DIA

• Program image

• Reference image

• Difference image

(best seeing)

• Geometric scale
• Scale the list of program images to the reference
  one
• PSF
• Flux Scale

The difference images have the PSF of the program
images

• Flux Scale

• Flux Scale

9
Subtraction between two images for SNe with SVSUB

• Difference image

• Artificial stars experiment I

• Geometric scale
• (immatch)

• PSF (daophot)

• Make difference mrjphot (ISIS)

• Artificial stars experiment II

• Photometry (daophot)

• Target (image with SN)
• Reference
• (best seeing)

Flux Scale
PSF Photometry on difference image
Error of subtraction

• Template
• (image without SN)

• Target (image with SN)

PSF Photometry on difference image
Error of subtraction

• Template
• (image without SN)
• Reference
• (best seeing)

FSF
?
?
10
Working in progress mrjphot aga(DIA)

• Difference image

• Artificial stars experiment I

• Geometric scale
• (immatch)

• PSF (daophot)

• Make difference mrjphot (ISIS)

• Artificial stars experiment II

• Photometry (daophot)

Flux Scale

• Target
• (image with SN)
• Reference

PSF Photometry on difference image
Error of subtraction

• Template
• (image without SN)

• Target
• (image with SN)

PSF Photometry on difference image

• Template
• (image without SN)
• Reference

Error of subtraction
FSF
Its possible to select witch star use to make
difference
We know the FSF
11
Scaling Factor for images of one telescope
B
FSF
FSF
12
Scaling Factor for difference telescope
HG
B
FSF
FSF
FSF
In the subtraction between two images we lose the
information about the color term
13
Little Example
SN2001ge
SN2001gf
SN2001ge
SN2001gf
14
Geometric scale
14-02-2000 Danish
15-03-2004 Ekar 1.82m
Geomap Compute geometric transforms using
coordinate lists (cl.images.immatch)
Gregister Register 1-D or 2-D images using the
geomap transforms (cl.images.immatch)
15-03-2004 Ekar 1.82m after Geometric scale
Target Template
15-03-2004 Ekar 1.82m
14-02-2000 Danish
ISIS (C. Alard 2000) DIA (P.R. Wozniak 2000)
use the same technique without find stars
manually.

• Time

• PSF of the two images (noao.digiphot.daophot.psf)
  
• Photometry of sequence stars (noao.digiphot.daoph
  ot.allstar)

15
mrjphot (ISIS)
PARAMETERS nstamps_x 9 / Number of
stamps along X axis / nstamps_y 9
/ Number of stamps along Y axis/ sub_x 1
/ Number of sub_division of the image along X
axis / sub_y 1 / Number of sub_division
of the image along Y axis / half_mesh_size
10 / Half kernel size / half_stamp_size
15 / Half stamp size / deg_bg 2 /
degree to fit differential bakground variations
/ saturation 60000 / degree to fit
background varaitions / pix_min 5. /
Minimum vaue of the pixels to fit /
min_stamp_center 130 / Minimum value for
object to enter kernel fit / ngauss 3 /
Number of Gaussians / deg_gauss1 6
/ Degree associated with 1st Gaussian /
deg_gauss2 4 / Degree associated with
2nd Gaussian / deg_gauss3 2 /
Degree associated with 3rd Gaussian /
sigma_gauss1 0.7 / Sigma of 1st Gaussian
/ sigma_gauss2 1.5 /
Sigma of 2nd Gaussian /
sigma_gauss3 2.5 / Sigma of 3rd Gaussian
/ deg_spatial 2 /
Degree of the fit of the spatial variations of
the Kernel /
stamp
sub_x
The difference image have the flux scale of the
reference
16
Artificial experiment I
Target artif. stars
Target
--------------------------------------------------
-------- gtgtgt Give position for artificial stars
for calibration(gt x lt then gt q lt) artificial star
magnitude for calibration (-5.) OUTPUT IMAGE
_targart.fits Added Star 1 - X 686.00 Y
201.00 Mag -5.000 Added Star 2 - X
706.00 Y 255.00 Mag -5.000 Added Star 3
- X 560.00 Y 274.00 Mag -5.000 Added
Star 4 - X 546.00 Y 277.00 Mag -5.000
Added Star 5 - X 434.00 Y 148.00 Mag
-5.000
-4.880 -4.869 -4.888 -4.884 -4.890 -
4.886 -4.892 gtgtgt zero point mag difference
0.12 gtgtgt error 0.00

.... min-max rejection .... -4.892 -4.869
-4.8856 0.0038470768111717 5 gtgtgt zero point
mag difference 0.11 gtgtgt
error 0.00

Difference image
addstar (noao.digiphot.daophot.addstar)
17
Artificial experiment II
--------------------------------------------------
-------- gtgtgt radius of circle inside of which
background is not calculated in unit of FWHM ?
(3.) gtgtgt half-lenght of square to cut in unit
of FWHM ? (6) gtgtgt Give position for artificial
stars for error estimates(gt x lt then gt q lt) zero
magnitude difference between target and template
(0.1144) artificial star magnitude for error
estimate (-1.3158180721099) OUTPUT IMAGE
_targart.fits Added Star 1 - X 682.00 Y
493.00 Mag -1.430 Added Star 2 - X
517.00 Y 361.00 Mag -1.430
.
-1.3538180721099 -1.5208180721099
. gtgtgt average magnitude -1.40 gtgtgt
error 0.07
.... min-max
rejection .... -1.3876180721099 0.03703646851414
5 gtgtgt average magnitude -1.39 gtgtgt
error 0.04
!!! zero 0.11/-
0.04, err 0.04 id ap_original ap_bgsub
fit_mag err_art err_fit SN_FIT SN1
-1.3734 -1.3984 -1.4302180721099
0.03703646851414 0.059
Target artif. stars
Difference image
addstar (noao.digiphot.daophot.addstar)
18
the flux scaling factor is the sum of the local
kernel (C. Alard, 2000) the result is on the
scale of the program image. I divide it by the
norm of the kernel and express the flux in the
units of the reference image(P. Wozniak, private
letters)
ISIS (C. Alard 2000) DIA (P.R. Wozniak 2000)
Maximize a cross correlation function
19

• Register the images
• Calculate the PSF
• Subtract the images
• Measure the strumental magnitude on the
  difference image
• Apparent magnitude
• Estimate error

Choice the reference image (image with best PSF)
photometric scale reference image
geometric scale target
Difference image photometry scale of
reference Geometric scale target
20
Subtraction between two images

• Program image

• Reference image

• Geometric scale

• Difference image

• PSF

• Photometric Constants
• Exposure time
• Airmass

• Photometric Constants
• Exposure time
• Airmass

?
21
Subtraction between two images for SNe

• Difference image

• Photometric Constants
• Exposure time
• Airmass

• Geometric scale
• (immatch)

• PSF
• ( daophot)

• Target (image with SN)
• Reference

• Template (image without SN)

• Target (image with SN)

• Template (image without SN)
• Reference

• I prefer that the difference image have the same
  characteristics of the target
• I have to control that the FSF is that what I
  want.

Why I want to know the FSF ?
22
Working in progress
mrjphot (ISIS)
aga (DIA)

• ISIS program is difficult to read.
• If we know the FSF we have to use an experiment
  of artificial stars.

• We know the FSF in DIA.
• DIA program is written well and its possible to
  do easily some change

23
Subtraction with ISIS (or DIA)
Ref Reference Image (best seeing) Im Program
Image Ker Kernel image (to be found)
the flux scaling factor is the sum of the local
kernel (C. Alard, 2000)
the result is on the scale of the program image.
I divide it by the norm of the kernel and express
the flux in the units of the reference image(P.
Wozniak, private communication)
the result is on the scale of the program image.
I divide it by the norm of the kernel and express
the flux in the units of the reference image(P.
Wozniak, private communication)
the result is on the scale of the program image.
I divide it by the norm of the kernel and express
the flux in the units of the reference image(P.
Wozniak, private communication)
(C. Alard 2000 P.R. Wozniak 2001)
ISIS http//www2.iap.fr/users/alard/package.ht
ml
DIA http//www.astro.princeton.edu/wozniak/dia
/