The image you have carefully nurtured through a variety of complex procedures requires a finishing touch. Because each telescope in the synthesis array has its own beam response (the so called `primary' as opposed to `synthesised' beam), the image you have produced is really the sky (hopefully) multiplied by the primary beam. To correct for this, so that flux densities that you measure are not biased low, you must divide the image by the primary beam. In doing this, we assume that the primary beam is the same for each telescope, and that it is circularly symmetric. I refer you to the ATNF memo by Mark Wieringa (1992) if you wish to examine how good these assumptions are and what the true shapes of the ATCA primary beams are.
Note that you should not make the primary beam correction on a dirty image and then try to deconvolve it. You should also be aware that the noise, as well as any source will be amplified by the primary beam correction.
The AIPS task PBCOR has knowledge of the ATCA primary beam as well as the VLA primary beam. It works out which telescope to use by looking in the header, in which the TELESCOP keyword should be either `ATCA' or `VLA'. Use PUTHEAD if it doesn't say what it should (use IMHEAD to view the header). PBCOR also uses the frequency axis of the image to work out what the parameters of the primary beam correction should be, so make sure the header correctly reflects the frequency as well.
| PBCOR | |
| inname,inclass | Image to correct |
| inseq,indisk | |
| outname,outclass | Corrected image |
| outseq,outdisk | |
| blc =0 | Do entire |
| trc =0 | image |
| dparm =0 | Defaults OK |
| gpos =0 | Defaults OK |
Last update : 27/11/93
1