Validation of HLA Source Lists

1
Validation of HLA Source Lists

• Feb. 4, 2008
• Brad Whitmore

• Overview
• Plots
• Summary

2
Overview
A number of validation checks have been made of
the the HLA DAOPHOT source lists. Fewer
validation checks have been made of the
SExtractor source lists, hence they have been
designated as Beta products. This document
collects plots and tables from various
presentations that have been made during the
past several months (e.g., to the STScI Users
Committee). A more formal set of documents will
be available in the future (e.g., PASP paper,
Instrument Science Reports)

3
Comparison of four different visits

• 47TUC - Stellar Field
• NOTE The offset at the bright end is due to
  saturation in the HST image.
• DAOPHOT and SExtractor match to better than 0.02
  mag when comparing aperture photometry for F435W
  image.
• 4 different visits (fields A - D) match to
  better than 0.02 mag.
• Only 1 star is not saturated in F555W
  observations, but that matches with Stetson to
  -0.008 mag.

4
Comparison with M31 Catalog

• M31 Halo (Tom Brown) vs. HLA_9453_33
• Bottom Panel
• Circled points predicted by HLA to be saturated.
  Good prediction.
• Offsets and RMS within design goals.(i.e., 0.1
  offset, 0.3 RMS)
• Middle Panel
• Find correlations between residuals and
  Concentration Index (CI)
• Top Panel
• Use correlation to select small CI and improve
  offset and RMS
• NOTE
• Tom Brown adds 100 orbits to get to 31 mag
• HLA is single orbit so only reaches 27 mag

5
Comparison with M31 Catalog - deeper

• M31 Halo (Tom Brown) vs. HLA_9453_33
• .
• Middle Panel
• Correlations between residuals and CI less well
  defined since lower S/N
• Smaller residuals for small CI, as might expect
• Bottom and Top Panels
• Similar to brighter comparison with RMS
  increasing for fainter objects, as expect.
• Offsets and RMS still well within design goals
  even for faintest objects in HLA catalog.

6
Comparison between HRC and WFC for slightly
Resolved Star Clusters

• HRC in crowded part of Antennae vs. WFC
• Conclusions
• Good photometric offset and RMS comparisons,
  especially for bright objects
• Astrometry for HRC is not very good (i.e. 0.789
  arcsec offset) as expected since small field of
  view and crowded region means no standard stars
  to compare with.

7
Comparison with SDSS

• Comparison of random HLA source list vs. Sloan
  Digital Sky Survey (SDSS) catalog
• Conclude
• Photometric offset and RMS well within design
  goals for stellar like sources
• Astrometry well within design goal (lt 0.3 arcsec)
• Not many objects to compare with generally
• Several other comparisons give similar results.

8
Comparison of SExtractor source list with SDSS

• Random HLA field vs. SDSS - extended objects
• Conclusions
• Photometric offsets and RMS barely within design
  goals.
• Several reasons likely
• Extended nature of objects.
• SDSS phot is ground- based, so some mismatch due
  to how things look from space and from the ground
  (i.e., different resolution).
• Transformations from SDSS filters to HLA.

9
Comparisons for stellar field, slightly extended
objects, and faint galaxies

• M87 - slightly extended, globular clusters
• Extended objects have offset since using stellar
  aperture corrections.
• Correction made in top panel, but not general,
  so flag extended sources with 999 for total mag
  in data files

• 47 TUC - stellar field
• Comparison with Stetson ground-based B is
  excellent
• Shows multidrizzle maintains photometric
  integrity for ACS images

• UDF - faint galaxies
• SExtractor source lists look promising, but very
  limiting testing so far.

10
DAOPHOT Source List Artifacts (Early Data
Release, Data Release 1)

NOTE 1. Conclude The Data Release 1 (Feb.
2008) source lists are much improved over Early
Data Release (July 2007) 2. The HLA design goal
is lt 20 artifacts. These source lists have room
to spare. Some observing strategies (e.g., N2
hence impossible to remove all cosmic rays)
result in poorer quality source lists, hence the
20 is still relevant. 3. No longer make source
list for 47_tuc WFC, since all N1. The HRC
catalog is quality1 .
11
Summary

• All tests of the DAOphot source list to date have
  shown them to be well within the design goals of
  0.10 absolute astrometry, and 0.3 mag RMS.
• The existing tests of the Sextractor source lists
  are also promising, but have not been as
  extensive, hence we are defining these as BETA
  products for now.
• The number of artifacts has been improved
  dramatically since the Early Data Release, but
  there will always be artifacts at some level
  since some observing strategies make it difficult
  to remove cosmic rays (e.g., N 2), hot pixels
  (undithered), or impose other limitations (e.g.,
  saturated, very crowded, ). We have defined lt 20
  artifacts as our design goal.
• It should be kept in mind that these are meant to
  be general use source lists. In most cases it
  will be necessary to make your own source lists
  focused on your specific science goals (e.g.,
  going as deep as possible).