The plot below shows the pointing data taken with RxA on UT Oct 01,02,03:

There is such a strong trend of the azimuth residuals with elevation (or zenith distance; centre-left panel) that a TPOINT analysis was performed. By changing the parameters (AN,AW,NPAE,CA,IA,IE,TF) by (-0.4",-0.8",+0.3",+4.4",-6.7",+0.2",-3.9") the rms residuals in (dS,dZ) of (2.2",3.5") are reduced to (1.5", 3.4"):

None of the changes to the parameters is formally significant, but the dramatic overall improvement in the azimuth residuals is hard to dismiss as unworthy of application. The elevation errors are more subject to anomalous atmospheric effects and ought not to exhibit an improvement if such effects are indeed dominant, but the improvement in the azimuth residuals is much more likely to be truly structural.

The pointing model was updated at 14:15 on 03 Oct 2005, and the next night's RxA data seems to exhibit an improvement:

05 Oct
The model corrections improved the azimuth performance nicely but introduced an unrealistic flexure term, TF. The data over the next two nights (UT Oct 04,05) reveal that in the centre panel below:

A new model with TF corrected by 10" (and other parameters adjusted slightly all round) would have produced these residuals:

The new model was installed at 12:00 noon Oct 05 2005, and . . .

06 Oct
. . . RxA pointing the next night was v.acceptable:

. . with raw rms residuals in (daz,del) of (1.2",1.6").