Single-exposure WPHot-derived vs Flux-weighted centroid source positions (v5)

Below we summarize differences between the positions of (mostly compact/point-) sources derived from PSF-fit photometry and then independently derived using flux-weighted centroids in the x-y system of single exposure frames. The goal is to explore the accuracy of the PSF positioning for profile-fitting and prior w1-to-w2 band offsets for use in astrometric calibration across these two bands.

Two periods in the post-cryo phase of the mission were explored: the first few days (~ Oct 4, 2010) and the last few days (~ Jan 27, 2011). Supporting plots for each period are below. The following statistics represent trimmed-medians with uncertainties derived from half the 84^th - 16^th percentile range per axis divided by √N.

W1 [EARLY post-cryo]:
  52047 sources 
  Dx = 0.01517 +/- 0.00275 arcsec
  Dy = 0.00509 +/- 0.00273 arcsec

W1 [LATE post-cryo]:
  11748 sources
  Dx = 0.03223 +/- 0.00348 arcsec
  Dy = 0.03628 +/- 0.00351 arcsec
__________________________________

W2 [EARLY post-cryo]:
  44929 sources 
  Dx = -0.01871 +/- 0.00288 arcsec
  Dy = -0.00900 +/- 0.00275 arcsec

W2 [LATE post-cryo]:
  5768 sources
  Dx = -0.02201 +/- 0.00601 arcsec
  Dy = -0.00307 +/- 0.00581 arcsec

Figure 1a - 2D density plot of W1 'WPHot - flux weighted centroid positions' in early post-cryo.	Figure 1b - 2D density plot of W1 'WPHot - flux weighted centroid positions' in late post-cryo.
Figure 1c - 2D density plot of W2 'WPHot - flux weighted centroid positions' in early post-cryo.	Figure 1d - 2D density plot of W2 'WPHot - flux weighted centroid positions' in late post-cryo.