Reducing ATCA millimetre observations
With the new observing bands at 12 and 3 mm, ATCA observers have an opportunity to exploit a unique instrument in the southern hemisphere. Coupled with this are new steps that need to be considered in the data reduction process. MIRIAD has a number of tasks and options that help with these steps. These are documented as new chapters in the current version of the MIRIAD Users Guide (available on the web in postscript or html formats). This article sketches some new post-observation considerations. Note also that different observing strategies and calibrations are needed at the shorter wavelengths: these are documented in the web-based observing guides to these new systems.
New monitoring data: The RPFITS file produced by an observation now contains weather measurements, antenna-pointing centre (in addition to delay centre), and a measure of the tracking error. In blustery conditions or when there are antenna drive problems, the tracking error may be a significant fraction of the 3-mm beam size. In this case, an observer can flag data based on excessive tracking error.
Opacity correction: At 3 and 12-mm wavelengths, the atmosphere can no longer be approximated as transparent. The current 3-mm method of determining the effective system temperature (a modified "chopper-wheel method") accounts for atmospheric opacity on-line. This is not the case, however, at 12-mm. An option has been added to the miriad task, atlod, to make a first-order opacity correction based on weather data and a model of the atmosphere. This, along with a calibrator observed at a similar elevation, will largely correct for opacity.
3-mm initial fixes: The data measured at 3 mm still reflects some immaturity in our systems at these wavelengths. A MIRIAD task is available to apply a number of "standard" corrections. These include antenna-location correction; shadowing correction; gain/elevation-curve correction; and the application of the system temperature.
Figure 1: Flux density of 1934-638 at 12 mm. The dashed line shows the model of 1934-638's flux density determined from centimetre wavelengths. The solid line shows the new model to fit the observed 12-mm data.
Absolute flux calibration: At short wavelengths, it is harder to find compact sources that can be used as absolute flux calibrators. At 12 mm, 1934-638 is still reasonably strong, and has proven to be constant over at least six months. Consequently it can be used as an absolute flux calibrator. 1934-638's flux density over the 12-mm band has been determined and is now built into the calibration software (see Figure 1).
At 3 mm, 1934-638 is far too weak: either Mars or Uranus serves as the best absolute flux calibrator. A number of tasks within MIRIAD assist in picking the appropriate planet and then bootstrapping with it. Note that planets can be completely resolved out in some arrays, and so checking is needed before the observation.
Additionally, the 3-mm flux densities of a few calibrators will be monitored during the winter. Recent measurements of their flux densities could be used for absolute flux calibration purposes. Recent flux-density measurements are available on the ATCA-calibrator web pages.
Bob Sault
(Bob.Sault@csiro.au)