Analysis of 1998 SCUBA pointing data for signs of az-track errors

SUMMARY
All large SCUBA pointing datasets from 1998 are combined and analysed for the effects of inclinometry spikes. A weak link is evident.

All the SCUBA pointing datasets with more than 10 logged entries were added together, with alignment of mean values of daz and del. The data were stripped of any particularly odd data as assessed on a nightly basis, leaving 4004 pointing measures. The data are shown below :

The data were binned at a resolution of 1-degree in azimuth. For each bin the mean and rms scatter of the contributing data were calculated, with sensible defaults (0.0" means, and 1.5" rms errors) allocated to those bins with 0 or 1 entries. The results for daz and del are shown below left and right, respectively :

The spikes in the inclinometry - as seen for instance in the data reported on 981107 - suggest that pointing at certain azimuths should be systematically different from that expected from the telescope model, at least some of the time. If the spikes are real and variable in strength, as it appears, and the inclinometry is taken often enough to combat these effects, then we would see no pointing errors at all. Since the inclinometry lags any changes in the reality, the pointing will be affected. In any event, this should show up here either as systematic offsets from zero of dazmean or delmean, or higher than normal values of dazerr or delerr.

Such structures are seen :

• There are broad systematic structures of size 2" or so in both daz(mean) and del(mean) as a function of azimuth. However, these structures are considerably wider than the inclinometry effects sought and indicate errors in the telescope model or previously unknown large scale effects in the inclinometry and track model.
• Note also that the largest (5") and narrowest excursions of dazmean and delmean are due to single data and do not imply systematic errors of this size in either model.
• Small sized (2") 5-degree wide structures (i.e. matching the inclinometry spike profile) exist most clearly in dazerr at azimuths 120, 150, 220, 240, 300 and in delerr at 250, which match some but not all of the inclinometry spikes in, for instance, the 981107 data.

The correlation between pointing errors and inclinometry spikes is weak.