95 SCUBA data taken over the last 3 nights (and since a datum-ing of the TMU) were collected for analysis by TPOINT. Their raw rms's in (dS,dZ) or (daz,del) were (2.2",1.7") :

After a fit for all 7 TPOINT parameters the corresponding expected performance of the model should be (1.7",1.8") :

The new model was installed at about noon on 18 Feb 2003.

NB : 1 : Using HESA as an 8^th parameter in the model reduces the rms's again to (1.4",1.6") with a suggested value for HESA of 3.3"+0.5", consistent with other determinations so far this year.

NB : 2 : The plot below shows the TPOINT plot 'g s n' (azimuth error on the sky -vs- index number - akin to 'time') for all SCUBA data since 20030210 : there is no sign of the TMU-datum-ing event on the 14th, which occurs about one third of the way through this dataset.

20030219
Subsequent pointing with SCUBA produces possibly the best raw performance seen to date :

HESA ?
The new data are not improved by the use of the TPOINT term HESA; with a formal optimal value of 0.3+0.7".