Reports of odd azimuth pointing prompted the examination of recent SCUBA pointing data. Data between UT 20020315 and 20020319 are collected together below :

Note the strong systematic variations of daz with both time (upper right), particularly during first shift, and -- despite the paucity of data from quadrants 1 & 5 -- with azimuth (upper left). The daz-vs-az plot is displayed below using TPOINT :

Running the data through FIT7 on the summit VAX yields the following best fit :

with the following improvement in daz -vs- azimuth from TPOINT :

The resulting expected rms of 2.2" in the azimuth residual is not great and the residuals are not normally distributed in azimuth yet, but at least the new model suppresses the systematic to near-tolerable levels.

The new model was installed at ~11:00 HST 19 Mar 2002.