OWLS Pipeline Status Report

1
OWLS Pipeline Status Report

• STEP 2007 Workshop, JPL
• Mon, Aug 20, 2007
• MSS Gill, J. Young, P. Martini, K. Honscheid, D.
  Weinberg, C. Kochanek , E. Rozo OWLS Team
• ( OWLS OSU Weak Lensing Survey)

2
Outline

• Introduction to OWLS project
• Description of hybrid approach
• Our current results from this approach
• Cuts on various galaxy properties and their
  effects on results
• Plans

3
OWLS Project Goals

• Constraining cosmological parameters with
• cluster weak lensing

• Picture of Abell 611 taken with the LBT (largest
  effective area telescope in the world 11.9m)

4
LBT OWLS Projections

• Projections for LBT with several clusters by next
  Autumn

Assumes lensing using 50 richest clusters in
observed 250 degree squared field with all
background (lensed) galaxies at z0.6 Colors are
1,2,3 s contours Here s8 vs. Wm
(Plot J. Yoo)
5
Progress over the year..

• We initially started the project thinking we
  would be able to basically use the shapelets code
  out of the can to extract cluster shear
• But havent yet been able to get minimal KSB
  estimator to give us the known shears for STEP2
  fields we now know it has some instability (help
  from RJM and JB)
• So we came up with our own method to extract
  shear, which works decently well
• In hindsight, we have derived relations for
  ellipticity from shapelets which should be
  identical to sextractor values

(see backup slide for more detail)
6
Hybrid Approach

• A bit unorthodox, but we extract shear in 2
  steps
• For sextractor part, we need to make pstamp
  recompositions of each object from the shapelet
  decompositions
• Then separately sextract each of them, getting
  ellipticity for each object
• Recent work we can use shapelets coefficients to
  get this as well, and plan to show this
  equivalence in future work (see backup slide)

• Deconvolve image using shapelets
• Extract shear averages using sextractor

7
Deconvolving with Extracted PSF

• We used the STEP2 starfields for the 6 individual
  psfs
• Extract psf in the standard way pick stars,
  decompose them, fix fwhm and max decomposition
  order, decompose again
• Constant psf not assumed
• We then deconvolved the galaxy field with this
  psf model

psf
8
Example of Deconvolved Galaxy
(From STEP2 field psfD, file shear20, object 1)
9
From deconvolved fields to shear extraction

• Use standard approach of plotting the difference
  of extracted shear and the input shear, vs. the
  input shear, for a given field
• Fit this to a line, and extract output c
  (y-intercept) and m (slope) values
• Example (psfA)
• Ellipticity per object

Real (triangles) and Imag (diamonds) Shear Dif
vs. Input
Output-Input
Input
10
Our Current Results ADE Avg

• Average of PSFs A,D,E compared with STEP2
  groups
• Squares rotation-matched
• Triangles original only
• Red square our result -- only on 22 file pairs
  per psf

11
Our Current Results PSF A
m

• Squares rotation-matched,
• Triangles original only
• Red square our result

m
c
c
12
Our Current Results PSF D

• Squares rotation-matched,
• Triangles original only
• Red square our result

13
Our Current Results PSF E

• Squares rotation-matched,
• Triangles original only
• Red square our result

14
Variation with Cuts
Significant variation in resulting c and m With
varying cuts Here vary ab from 2 to 5 in steps
of 0.5 Interested in learning how other groups
came to their optimal sets of cuts
15
Plans

• Next go to Subaru A1689 images
• Then to LBT images currently have 2 clusters in
  hand
• Plan to get many more clusters from LBT time this
  coming year
• Ultimately extract cosmological parameters

16
BACKUP Slides
17
Results with Known PSF

• We initially used PSFs sent to us by J. Berge to
  get our technique working
• Example Plot of rotation-matched averaged c, m
  for D compared to paper, Red ours

18
Cut effects on results

• Made cuts on several quantities (sexarea,
  sexflux, ab, shapelets error) and saw results
  improve
• Example change in m as function of shapelets
  error cut

19
Ellipticities from Sextractor vs. from Shapelets
From Sextractor, second moments
Semimajor and Seminor axes
From shapelets, to lowest order
lt x2 -y2 gt (a_22 a_2,-2)/(a_00 2 a_40)
lt 2 x y gt -(a_22 - a_2,-2)/(a_00 2 a_40)
Where a_22 etc. are the shapelets coefficients
For power in that basis function