Title: Pan-STARRS Image Processing Pipeline
1Pan-STARRS Image Processing Pipeline
Astrometry and Photometry
IFA Pan-STARRS Seminar 735
October 14, 2004
2Summary of Topics
- Astrometry and Photometry Precision Requirements
- Summary of the AP Survey
- Achieving the Photometry Goals
- Achieving the Astrometry Goals
3- Precision Requirements for Pan-STARRS (PS-4)
- 30 milliarcsec relative astrometry
- 100 milliarcsec absolute astrometry
- 5 millimag relative photometry
- 10 millimag absolute photometry (internal system)
These goals can only be efficiently met after we
have produced a Pan-STARRS Astrometric and
Photometric reference catalog
4- PS-1 AP Survey Parameters
- gizy bright sweep, 1 x 5 seconds, 2 x 30
seconds - r bright sweep, 1 x 5 seconds, 6 x 30
seconds (over 6 months) - 1 photometry 19.5 magnitudes ( 1 x 109
stars) - saturation 14 magnitude (5 seconds), 8 magnitude
(sweep) - 50 overlap dither pattern
- 2 of survey time to calibrations
- 12 standards observations per night per filter
(40 min)
5- Photometry Science Motivations
- galactic stellar populations
- galaxy cluster evolution
- rare object searches
- high-z QSOs
- extremely red galaxies
- low-mass objects (L T dwarfs)
- YSOs
6(No Transcript)
7- Photometry Analysis Overview
- linearize detector flux
- apply shutter correction
- flatten images
- photometer objects
- apply image zero point, color correction, airmass
correction
8- Photometry Detector Linearization
- stable light source variable exposure time?
- calibrated light source?
9- Photometry Shutter Correction
- measure shutter fly-over times dt(x,y)
- apply to flat-field or science images
- probably small for 30 second exposures (0.1 30
ms jitter - may be significant for 5 second exposures...
10- Photometry Flat-Fielding Issues Illumination
Source Options - twilight-flat
- pros continuum source, spatially uniform
(usually), bright - cons very blue, limited availability, cirrus
issues - dome-flat
- pros continuum source, available anytime,
repeatable (?) - cons spatial structures, low count rates,
emission line dangers - night-sky flat
- pros obtained 'automatically'
- cons low count rates, stellar contaminations,
- spatial structures unknown, emission line
source
11- Photometry Flat-Fielding Issues Flat-field
Corrections - correct for geometrical distortion scattered
light - stability time scale?
12- Photometry Color Corrections
- chip-to-chip color terms (possibly linear, small)
- internal system vs external system
- external transformations are often ambiguous
Mginst -2.5 log (counts / sec) Mgsys Mginst
Cg Kg(1-z) Fg(color) Mgcal Mgsys
Qg(color)
13- Photometry Absolute Photometry
- use relative photometry with reference overlaps
- measure zero points atmospheric corrections,
apply - combination method (relphot uniphot)
ref 1
ref 1
ref 2
ref 2
14- Photometry Zero-point Stability (long-term
trends) - system zero-points vary 0.1 mag on timescales
100 days
15- Photometry Atmospheric Stability (short-term
trends) - variations in time Cf(t)
- variations in space Cf(ra,dec)
- these variations are NOT strongly correlated
16Photometry Atmospheric Stability (short-term
trends)
17- Photometry Atmospheric Stability conclusions
- photometric conditions exist at lt1 level
- sometimes there is haze Cf(t)
- sometimes there is thin cirrus Cf (x,y)
- haze is apparently more common...
- make use of external indicators of transparency
conditions - SkyProbe
- NIR Camera
18- Photometry Bright Stars Flux Calibrations
- OTA guide stars tie 30 sec exposures to 10 msec
exposures - OTA 'sweep' can yield survey of stars 8 - 14 mag
- Bright stars provide flux calibration
(spectrophotometric standards) - SkyProbe A will provide atm transmission function
19- Astrometry Science Motivations
- proper motions / baseline
- high-quality grid for weak-lensing (starting
point) - stellar matching in crowded fields
20- Astrometry Basic Concepts
- RA,DEC lt-gt X,Y
- linear fit
- RA RAo XdRdX YdRdY, etc
- 100 mas _at_ 36 arcsec FOV
- projection fit
- RA,DEC -gt P,Q
- P,Q f(X,Y)
R,D
P,Q
21- Astrometry Mosaic Astrometry
- boresite projection (RAo, DECo, ?)
- distortion Nth order polynomial L,M f(P,Q)
- chip coordinates Xo, Yo, ?
- watch for stability issues
R,D
projection
P,Q
optical distortion
L,M
L,M
chip locations
X,Y
22- Astrometry Mosaic Astrometry Stability
- chip coordinates distortion are fairly
degenerate - direct fitting is a large, multiparameter,
non-linear problem - use local gradients instead
- fit L,M assuming no distortion
- fit chip parameters from L,M
- measure L,M residuals (?L, ?M)
- measure local gradients (d?L/dP, d?L/dQ, d?M/dP,
d?M/dQ) - fit local gradients to Nth order polynomial
- resulting terms are coefficients of L,M vs P,Q
(N1)th order fit - fit is insensitive to chip positions, boresite
center
23- Astrometry Example from MegaPrime
- chips are probably NOT flat!
24- Astrometry Example from MegaPrime
- chips are probably NOT flat!
25- Astrometry Calibrating the AP Survey
- what is stability of model components?
- boresite changes with every exposure
- optical distortion long-term stability expected
- chip warps short-term stability? temperature
dependence? - chip positions gravity vector dependence?
- regularly measure model components track
changes - tie to ICRS with Guide Stars (Tycho)
- atmosphere may introduce 50 mas scatter model in
overlaps
USNO-B deep, dense (20 mag), 150 - 250 mas
scatter, large-scale errors UCAC modest (16
mag), 20 mas scatter proper motion Tycho
shallow (11.5 mag), 10 mas scatter proper
motion Hipparchos very shallow (7.3 mag), 1 mas
scatter proper motion