Flux calibration

When clicking on FLUX (in the CLIC menu; Fig. ) a widget similar to the one shown in Fig. is opened. The flux calibration is an iterative process in which the known flux of one or more calibrators is fixed to determine the efficiencies (Jy/K) of the antennas, which are then used to estimate the flux of the other calibrators. When a flux is fixed, SOLVE derives efficiencies and fluxes. GET RESULT, STORE, and PLOT store the solved flux densities and plot the amplitudes scaled by the derived fluxes (in K/Jy). These scaled amplitudes correspond to the inverse of the antenna efficiencies, that, in an ideal project, should remain constant and equal to their nominal values.

The amplitudes obtained for a calibrator often vary along the track due to effects of a changing atmosphere or instrumental problems. Antenna efficiencies should be estimated by considering the best data ranges. Observational glitches, data obtained with bad pointings or focus measurements, or limited intervals of bad data should not be considered. If for an observed polarization amplitude oscillations on a 24 hour scale are observed, this is often an indication of that the emission is polarized. Having H and V polarizations is needed to confirm the presence of polarized emission. (The degree of polarization of the PdB phase calibrators is evaluated and archived at PdB, and so can be checked by your local contact if needed.) Anyway, there is nothing to do in the flux calibration with respect to this. The Scan List option permits to select the scan ranges to be considered in the calculation of fluxes and efficiencies. After PLOT, scan numbers can be determined by using the command ``cursor'' and clicking on the display (an example is shown in Figure ).

All the calibrators observed during the track are shown in this widget. Currently the main flux calibrator is MWC349, which is observed in most of the projects; when included, a flux is proposed to be fixed, as we can see in Fig. . Its use must be however considered by checking the quality of the observed correlations: i.e. correlations on MWC349 showing a big scatter in amplitude should not be considered unless they are representative of the track observing conditions. The other calibrators, the one used to calibrate the RF and also the phase calibrators, can be used in this process. Their flux may be known from other tracks observed close in time. Note also that we monitor the flux of the brightest calibrators. Your local contact can provide you this information, and also an estimate of the right efficiencies with a reasonable accuracy. For example, for a track observed in good conditions the expected efficiencies for the different antennas should range from 20 to 25 Jy/K, from 25 to 32 Jy/K, and from 32 to 45 Jy/K for receiver bands 1, 2, and 3 respectively. Values much larger than those expected should be well explained by the observational conditions. Efficiencies significantly smaller are not possible.

Figure: Flux Calibration widget, showing on top the solved antenna efficiencies. Note that these are characteristics of the antennas, and the different values obtained between antennas should be well understood by the AoD at the time of the observations and likely also by your local contact. For instance, the low efficiency of A6 at the moment of these observations was due to a polluted injected LO signal; a new LO box was installed in A6 just after these observations. In the second row we define the scan intervals to be considered in the flux calibration.

By properly using this method, the absolute calibration obtained is typically precise to less than 10$\%$ at 3mm and $\sim$ 20$\%$ at 1mm. Special attention should be payed to the relative flux calibration between different tracks; normally the different tracks provide complementary information in the uv-plane, and so a wrong relative flux calibration can be interpreted as source structure.

Figure: Flux Calibration plot, showing the amplitudes scaled by the solved fluxes. Note that the amplitudes from both Narrow correlator inputs are averaged in this plot.

As mentioned in Sect. the green lines at the bottom of the plots show, when being above zero, the regions in which the atmospheric phase correction is applied as resulted from PhCor.

Note finally that the option CHECK, at the top of the FLUX calibration widget (Fig. ), permits to obtain a solution by fixing a reference calibrator and ignoring scans of quality below a certain threshold. This may be used as a first try, a second iteration is often needed (after storing the first one). A solution is stored with GET RESULT and STORE, and a new PLOT should be created accordingly.

The flux calibration is performed by averaging the amplitudes from all the spectral units, i.e. from all the correlator inputs⁸, H and V polarizations. Delays (see Sect. and ) result in flux losses due to that phases are frequency averaged. To correct for remaining delays you should follow the instructions given in Sect. . Also, as mentioned in Sect. , differences in the (frequency averaged) phases from H and V polarization receivers introduce amplitude losses in the flux calibration for the RF and flux calibrators. See Sect. to correct for this effect. Note anyway that either the presence of delays or polarization differences are rare, since the standard PdBI observing procedures correct for them at the very beginning of each track.

If flagged data are masked, such masks should be reset before the flux calibration. Since fluxes are solved by averaging the data of all the spectral units, we may not be able to identify problems coming from some flagged data from, for example, one of the narrows.

Figure: Example of the standard amplitude calibration, performed independently for the V and H polarization receivers. In this example the calibrator emission is found to be polarized. The option ``let do_avpol yes'' will allow repeating the amplitude calibration by averaging the calibrator emission for both polarizations, as shown in Figure .