up to PS1_GPC1_Rotator_Measurements

2008.09.30

I've performed my analysis of the fitted vs observed boresite positions for the 20080915 data that JT was discussing a bit today. I've done the same fits as before: modeling the path of the boresite offset as an ellipse. The boresite offset is defined to be the distance in pixels that the boresite coordinate is shifted relative to the center of the camera (midway between ota 33 cell 00 pixel 00 and ota 44 cell 00 pixel 00). I've cleaned up the plot to show in one graph the path in X,Y coordinates as well as the following values vs position angle: the X-offset, the Y-offset, the X-offset residual, and the Y-offset residual. The residuals are the difference between the observed offset and the fit to the elliptical model. Per confusion from Jeff, I've adjusted plot of the path to have an aspect ratio of 1 and a fixed scale for all of the attached plots (220x220 pixels). I've also added a mark to represent on the path the position for which the rotator was closest to -115.

Attached are plots for the g & r sequences from 20080915 and the i sequence from 20080827. The quality of the astrometry solutions was worse for 20080915: there were fewer stars in these fields, and I've had to reject a few points that were seriously outside the observed trends. Below are the fitted ellipse parameters for each data set. My initial impressions are:

1) g & i both seem to confirm the trend that the boresite error in the X direction gets larger for lower altitude, but that the in the Y direction it is generally constant. for the altitudes 70, 50, 30: RX(g) = 52, 77, 92; RX(i) = 75, 108, 145; RY(g) = 39,45,55; RY(i) = 53, 59, 67. So, some growth in the Y-direction, but much less than in X.

2) the r-band data, taken the same night as the g-band data does not agree well with these trends: RX(r) = 45, 43, 60; RY(r) = 48, 59, 88. It seems that for this r-band sequence, the trend is flipped (X nearly constant, Y grows with decreasing altitude).

any one else have other thoughts? this does not seem to be a simple trend of the offset as a function of rotator angle and altitude...

gene

==> rotseq.30g.pars <== # fitted values: Xo: 18.828287 Yo: -71.378357 RX: -92.772423 RY: 55.671108 P0: 0.660415 T0: -0.036934

==> rotseq.50g.pars <== # fitted values: Xo: -24.885120 Yo: -52.015652 RX: -77.185303 RY: 45.916832 P0: -2.276575 T0: 3.251292

==> rotseq.70g.pars <== # fitted values: Xo: -64.563072 Yo: -4.641876 RX: -52.212910 RY: 39.014824 P0: -1.914627 T0: 3.266521

==> rotseq.30r.pars <== # fitted values: Xo: -71.361954 Yo: 33.617271 RX: -45.239864 RY: 88.274780 P0: 1.339565 T0: 0.634835

==> rotseq.50r.pars <== # fitted values: Xo: -48.113983 Yo: 27.144241 RX: -43.854301 RY: 59.156593 P0: 0.679057 T0: -0.143105

==> rotseq.70r.pars <== # fitted values: Xo: -16.860861 Yo: 80.069687 RX: -60.495838 RY: 48.157955 P0: 0.831692 T0: -0.456550

==> rotseq.30i.pars <== # fitted values: Xo: 45.276226 Yo: -9.302369 RX: 145.009399 RY: 67.481323 P0: 0.646368 T0: 0.357436

==> rotseq.50i.pars <== # fitted values: Xo: 21.140253 Yo: -14.027385 RX: 108.704620 RY: 59.530045 P0: 0.587284 T0: 0.475980

==> rotseq.70i.pars <== # fitted values: Xo: 10.799455 Yo: -14.946228 RX: 75.923737 RY: 53.394173 P0: 1.152103 T0: 0.182733