Two Central bearing measurements

The last measurements ( 981213 and 990120 ), showed the empirical correction to be overestimating the current size of the defect. These previous measures had peculiar problems (like weakness of source and hence increased noise) - so today's measures were both made using Mars, and with the empirical correction disabled, so as to determine anew the current size and profile of the defect.

Mars was tracked using the SCUBA map16 method, and with the empirical corrections disabled, from :

1. (az,el) = (107/17) at 01:28 HST to (113/30) at 02:25, and
2. (az,el) = (120/38) at 03:05 HST to (133/50) at 04:06.

The rms scatters in the elevation residuals are ~1.0" and 0.3", respectively; the first dataset containing some low elevation data that are particularly noisy. Taking account of the baseline trend in daz in the second dataset, and assuming none in the first dataset, the azimuth residuals show peak-to-peak errors of 8" centered at (109/22) and 5.3" centered at (126/44), giving full amplitudes at the horizon of 8.6" and 7.3", respectively, confirming the most recent previous estimates. (The current full amplitude of tel_empirical is 14.3").

The sense of the profiles is different - with the first event showing a -ve residual followed by a +ve, and the second event going +ve then -ve.

The location and width are within 0.2 degrees of those expected.

This reversal of sense is a total surprise, and if the characteristics of the c.b. defect have changed so radically then an even more comprehensive set of data (tracking Mars throughout second shift) must be contemplated.

Note added later : The reversal of sense was due to the empirical correction actually being left on during the first event.