A pointing run of 25 measurements was done w/RxA at the end of the night of UT 28 Oct. This followed a night of rather poor pointing when it was clear that a dedicated run was needed to reclaim the nominal pointing accuracy. Ben W. was TSS, and reported that "the weather was not as good as I would have liked, and the coverage was poor in quadrant 1/5". The data are nonetheless displayed below using TPOINT after the removal of collimation offsets:

Elevation/zenith-distance errors do not seem to be normally distributed, and the left-right error (dS) turns negative at high elevations - a likely result of the expected azimuth encoder zero-point change that often follows central bearing adjustments. A new pointing model was derived using TPOINT allowing all 7 parameters of the model to be optimized. The changes amounted to a -30" change in IA (the encoder zero-point), with some compensatory adjustment to NPAE. AW,AN, TF remained essentially unchanged. The resulting errors are distributed as shown below:

The systematics are removed; although the lack of data at the highest elevations leaves some iteration ahead. Scatters are reduced only to (4.3", 3.0") rms in (dS,dZ).

The new model was installed at ~15:00 HST 28 Oct 2006, with a plea for a repeat in better conditions.

20061029 UT
22 pointings were made (serendipidously) the next night. The weather remained poor (tau_225 = ~0.2); residuals are presented below:

The collimations implied by the model installed yesterday hold well, with observed mean (CA,IE) of (1.3",2.6"), and the observed rms scatters in (dS,dZ) of (5.0",1.8") (giving a total value on the sky of 5.3" rms) are as poor as predicted above. However, there is better coverage at higher elevations (see the systematics in the centre-left box) and these data allow a further update to the model. Optimization of all 7 parameters of the model results in further change to the azimuth encoder zero point of - 43"; the errors from the new model are distributed as shown below:

RMS scatters are reduced to (3.5",1.7") (3.9" rms in overall vector on the sky) - still not nominal . . . but better.

The new model was installed at ~14:40 HST 29 Oct 2006, and 15 serendipitous data taken on another wet night gave (dS,dZ) = (2.0",3.9"), hinting at some progress.

We may still need a dedicated run under better conditions.