SMU chop update bug in jiggle mapping/photometry

On 27th July, we discovered a bug in the SCUBA observing software related to (ra,dec), or LO, chopping. The correct behaviour is that the components of the chop in the SMU (az,el) co-ordinate frame (shown on the status screen as NSamp and EWamp) should slowly but continuously update as the parallactic angle rotates on the sky. Unfortunately, the commands to perform this "chop tracking" had been left out of the SCUBA observing scripts. This means that at the beginning of each new observation, the chopper would be started with the updated (az,el) components corresponding to the requested (ra,dec) chop throw, but these would then remain fixed during the observation.

This bug will have affected all observing modes apart from raster mapping (which is performed by a different, correct script). In particular jiggle mapping and photometry performed with LO chop throws will have suffered from this bug. Two-bolometer chopping will also have been affected. The problem was identified and cured on 28th July 1998.

This problem does NOT affect programmes where azimuth chopping was employed; the major programmes affected will be the deep surveys where long integrations (~1 hour at a time) and small LO chops (in-field chopping using 30-45 arcsec throws) were adopted. Some galactic programmes may have also been affected for which LO chopping was used to avoid off-beam signal contamination. Although larger chop throws (e.g. 120 arcsec) were generally employed in these cases, the integration times would have usually been shorter, reducing the severity of the effect (see below).

However, we hope that this bug will not have had a major impact on existing data. The effect will basically have been to smear the off-beam in an arc from the correct position. Assuming a reasonably large region of blank sky had been chosen for the off position, the only effect should be that a "different" piece of blank sky would have been used. Obviously if a source was in a confused region, and long individual integrations were performed, there is a possibility that the off-beam may have moved across a region of emission. Again, this will only have been a significant effect if sources were observed, with long individual observations, near transit.

The magnitude of the effect depends upon the declination of the source and the hour angle of the observation. For sources above dec = 60, or below dec = -20, the rate of rotation varies from ~10 - 20 deg/hour. For sources which transit at higher elevations, the rate of rotation starts off lower, but increases rapidly as the source transits. Formulae for calculating the rate of rotation are available from the JCMT on request.

In order to demonstrate the magnitude of the effect, and ensure that the fixed observing scripts were working correctly, test observations were carried out using the bright blazar 0528+134. The blazar was observed before transit for 45 mins with chop tracking disabled, and similarly immediately after transit with the new, corrected scripts. Figure 1 shows the case when chop tracking was disabled, and the two off-beams are seen to smear out into an arc around the central source. Figure 2 is the observation of the same same with chop tracking enabled after transit - the off-beams are now well-defined. This example is also somewhat of a "worse-case scenario" since the source declination of +13 degrees means that the rate of rotation of the parallactic angle is quite extreme near transit.

Please feel free to contact Richard Prestage or Wayne Holland at the JCMT if you need any further information.

Richard Prestage, Firmin Oliveira, John Lightfoot, Wayne Holland
31st July 1998