---

Pointing: updated fit of central bearing problem

---

Since the discovery of the central bearing problem in spring of 1997 an increasingly more convincing worsening trend has been observed. Consequently an update to the empirical correction was installed on April 21, 1998 at 1 pm HST.

The pointing glitch is still well modeled with a double gaussian, although the detailed shape may have changed slightly. The results from a recent pointing run on 3C279 with Scuba are shown in the figure. Residual pointing errors with the new model are less than 1" even at low elevations.

The top plot shows the pointing observations without the empirical correction active (dots), the bottom one the residuals after fitting with several models (lines). The black line shows the empirical model in use at the time of the observations, the red line the better fit after increasing the overall amplitude by 20%. The adopted fit is shown in green: it combines a 22.5% increase in amplitude with a -0.04 degrees shift of the solution in Azimuth. The azimuth shift almost certainly results from the more accurate value of the average azimuth during the pointing observations available with SCUBA compared to heterodyne pointings. The installed empirical pointing corrections are based on the fit indicated by the green line.

The bottom plot also shows that pointing errors with the old solution do not exceed 2 arcsecs at an elevation of 30 degrees. Since these errors scale with cos(elevation), typically they they will have been more like 1 arcsec for most observations. Pointing errors close to the horizon may have amounted to approximately 2.3 arcsecs.

At this time we have chosen not to change the functional shape of the model since it fits the steep slope between the two extrema still very well. Residual pointing errors with the new model are less than 1" even at low elevations.

A more complete analysis of the original problem can be found on the here.

Please address any comments, suggestions or requests to:

Remo Tilanus