A dedicated pointing was performed with SCUBA :

Coverage in (az,el)-space is good - although more high-elevation data would have been useful to explore the apparent turnover in azimuth residuals above 70^o (central panel in upper row) : otherwise, a very strong trend between the azimuth residual and elevation is seen. Large datasets were also obtained in the week or so prior, and these pointing data were added and the sum is shown below :

The daz -vs- elevation trend is confirmed as present throughout this period, and the turnover may be tentatively established at 70^o. Additionally, the data show moderately strong (3 to 4 sigma) correlations between the elevation residual and various antenna temperatures (T_front-T_back, 0.5*(T_front+T_back), and Tair), but none reduce the rms scatter in del by more than 0.1", so no thermal corrections are adopted.

A full fit of all 7 parameters was made and a new SCUBA model created. On this model the data should look like this :

The residuals now look perfectly 'normal' (even the highest-elevation data don't seem greatly discrepant), but the rms scatters in (daz,del) are, disappointingly, at or worse than 2.0". The TPOINT analysis was also performed and the model installed by Russell Kackley on 14 Jan is confirmed. Since then only 2 large-ish (N>10) SCUBA datasets have been obtained, and the observed performance is as expected from the above.