Using Resolved Calibrators

NRAO Home > CASA > CASA Cookbook and User Reference Manual

4.4.4.1 Using Resolved Calibrators

If the flux density calibrator is resolved, the assumption that it is a point source will cause solutions on outlying antennas to be biased in amplitude. In turn, the fluxscale step will be biased on these antennas as well. In general, it is best to use model for the calibrator, but if such a model is not available, it is important to limit the solution on the flux density calibrator to only the subset of antennas that have baselines short enough that the point-source assumption is valid. This can be done by using antenna and uvrange selection when solving for the flux density calibrator. For example, if antennas 1 through 8 are the antennas among which the baselines are short enough that the point-source assumption is valid, and we want to be sure to limit the solutions to the use of baselines shorter than 15000 wavelengths, then we can assemble properly scaled solutions for the other calibrator as follows (note: specifying both an antenna and a uvrange constraint prevents inclusion of antennas with only a small number of baselines within the specified uvrange from being included in the solution; such antennas will have poorly constrained solutions):

As an example, we first solve for gain solutions for the flux density calibrator (3C286 observed in field 0) using a subset of antennas

gaincal(vis=’data.ms’,
        caltable=’cal.G’,        # write solutions to cal.G
        field=’0’                # Select the flux density calibrator
        selectdata=True,         # Expand other selectors
        antenna=’0~7’,           #  antennas 0-7,
        uvrange=’0~15klambda’,   #  limit uvrange to 0-15klambda
        solint=90)               # on 90s timescales, write solutions
                                 # to table called cal.G

Now solve for other calibrator (0234+285 in field 1) using all antennas (implicitly) and append these solutions to the same table

gaincal(vis=’data.ms’,
        caltable=’cal.G’,        # write solutions to cal.G
        field=’1’,
        solint=90,
        append=T)                # Set up to write to the same table

Finally, run fluxscale to adjust scaling

fluxscale(vis=’data.ms’,
          caltable=’cal.G’,      # Input table with unscaled cal solutions
          fluxtable=’cal.Gflx’,  # Write scaled solutions to cal.Gflx
          reference=’3C286’,     # Use 3c286 as ref with limited uvrange
          transfer=’0234+285’)   # Transfer scaling to 0234+285

The fluxscale calculation will be performed using only the antennas common to both fields, but the result will be applied to all antennas on the transfer field. Note that one can nominally get by only with the uvrange selection, but you may find that you get strange effects from some antennas only having visibilities to a subset of the baselines and thus causing problems in the solving.

[next] [prev] [prev-tail] [front] [up]

More information about CASA may be found at the CASA web page

This code is available under the terms of the GNU General Public Lincense