Pointing data with SCUBA from UT 20021205 are shown below :

They show a very strong trend between the azimuth residual and elevation (central panel in upper row), and a strong trend of elevation resdiual with time, which also revealed as a trend against mean antenna leg temperature. A full fit of all 7 parameters was made and a new SCUBA model created. On this model the data should look like this :

. . . or this (TPOINT version). The azimuth residuals have a rms of 1.2" (excellent) while the elevation residuals now show a discontinuity between first and second shifts. The TPOINT analysis provided a new model, which was installed on 05 Dec 2002, but not in time for use that night (HST). That night's data showed no such del -vs- mean leg temperature trend (!)

This analysis seemed to demand that the SCUBA and heterodyne models be different and I distributed instructions for toggling between models and matching collimation files if necessary. (I was working from the UK without full graphics capability). Nick Rees subsequently analyzed the 4 nights of SCUBA data following its return to service and installed an single updated model with a single collimation file that he thinks should work without the need for toggling. So far this is working fine.