Transcript Optimization of DELPHI for Weak Lensing Measurements: A

Using DELPHI for
Weak Lensing Measurements:
Science Return and Mirror Size
Jes Ford, JPL, UNR
SURF 2007
8/21/07
Mentor: Jason Rhodes
Co-mentor: David Johnston
DELPHI: Background
Originally a midex mission planned by Jason Rhodes









Orbit: 600 km Sun Synchronous, 97.79°
Estimated observatory mass (spacecraft
plus instruments): 205 kg
Estimated payload power consumption:
< 50 W
Mission duration and launch
constraints: 2 years / Pegasus
Sky coverage: 21,000 deg2 over two
years
Frequency: Visible
Temperature: Telescope – ambient,
Detectors – 170 K
Pointing requirements: ~
milliarcseconds
Data rate to ground: 54 GB/day
TRADEOFFS:
 Orbit Selection
 L2 vs. Sun-Synchronous
 Thermally stable orbits
 Telecommunications requirements
increase subsytem mass for L2
mission
 Pegasus does not have the
performance to place a s/c in an
L2 halo orbit
 Scanning Strategy
 Drifting vs. Step-and-Stare
 Drifting strategy works best with
L2 orbit
 Combination of integration time
and sun-synchronous orbit require
step-and-stare scanning
DELPHI: Trade Studies
 Telescope Design
 Mirror diameter
 0.5 m, 0.75 m
 Three-mirror anastigmat vs. Cassegrain
 Plate scale and focal length
 15 m, 20 m
 Detector / Pixel Sizes
 NIR HgCdTe Hawaii 2RG
 E2V visible, frame transfer CCDs
 Buses
 Ball Aerospace
 STP-IV
 Orbital Science Corp.
 MicroStar
MIRROR SIZE IS A COST DRIVER!
DELPHI: Current Status
 NASA recently announced small midex (SMEX) mission
opportunity - not MIDEX
 DELPHI cannot fit tight budget constraints
 However, since Mirror size is main factor in the cost of a
telescope, it is important to know how small of a mirror is
still worthwhile to launch
 MY PROJECT: what is the minimum mirror size
that can recover weak lensing data reliably?
Image Simulation Parameters





Created using Shapelets
Pixels: 4096 x 4096 pix
Optical Filter: Wide filter centered on I-band
Input Shear:   [0,0] , no shear
PSF shape: roughly circular PSF, based on SNAP’s
telescope design
 PSF size: 2 pixels per FWHM

 Throughput: peak throughput ~70%
Image Variations
 Mirror Sizes: range from 20 cm - 2.4 m in diameter,
in 20 cm increments
 2 sets: - constant exposure time (1500s)
- constant photon flux
(varying exposure times, 1500s at 1.2 m)
 Separate Galaxy and Stellar images created
 Total of 23 star/galaxy image pairs
Sample Images
2.0 m mirror, 1500s exposure
40 cm mirror, 1500s exposure
Steps of Analysis
 Objects detected and catalogue created using
Source Extractor
 Object moments recalculated using RRG method
 Stellar images used to measure the PSF moments
 PSF is removed from the galaxy images (RRG)
 Bad galaxies are cut based on: moments, ellipticity,
size compared to PSF size, signal-to-noise ratio (RRG)
 Shear and shear error are measured from the galaxy
images (RRG)
 Plots created to analyze number of useful galaxies
(those that make the cuts) as a function of mirror size
 Plots created to analyze measured shear and error as a
function of mirror size
RESULTS 1:
Number of useful galaxies as a function of mirror size
Diamonds:
constant exposure
time simulations
Crosses:
constant flux
simulations


Useful galaxies are those that survive the cuts and are used to measure the shear
Number of galaxies has been normalized to number per square arcminute of sky
RESULTS 2:
Measured Shear as a function of Mirror size

Continuing Research
 Currently processing set of 143 simulations with
non-zero input shear:   1,  2 
-  1 = 0,  2 = -5, -3, -1, 0, 1, 3, 5 %
-  2= 0,  1 = -5, -3, -1, 0, 1, 3, 5 %
- Mirror 
Sizes: 0.4 m - 2.4 m in 40 cm increments
- one set at constant exposure time (1500s)
 - one set at constant flux

 Images need to be analyzed by others using methods
other than RRG… contact Jason Rhodes.
Acknowledgements
Many many thanks to:
 Dr. Jason Rhodes, my mentor
 Dr. David Johnston, co-mentor
 Dr. Richard Massey, writer of Shapelets
simulation pipeline
Questions?