Calibrators

All observations must ultimately be related to a compact source in the southern sky whose flux density is constant, unpolarised and known. At present, at cm wavelengths and 12mm we are using PKS B1934$-$638/J1939$-$6342 as this primary amplitude calibrator. This calibrator should be observed during each observing session. Note that in mid-1994 accurate measurements of 1934$-$638 led to a revised flux scale for this source at all ATCA frequencies. The changes were of order 10% at all frequencies. The new flux densities, which are written on the white-board in the control room, are used in the online calibration software (CACAL) and local AIPS and MIRIAD implementations. A separate primary calibration must be done at each observing frequency, even if the observing frequencies differ by as little as the minimum 1MHz increment. This is required to calibrate the phases, which are wound by the electronics more than the atmosphere over short frequency intervals.

You must also periodically look at a strong compact source as close as possible to the target source: The secondary calibrator. These observations are needed to correct for changes in gain and phase caused by the receiver and the local oscillator as well as atmospheric instabilities during the synthesis. In summer, especially during the daytime, phase stability can be poor. Observations requiring maximum phase stability (e.g., at mm wavelengths) are best made at night or during winter. Secondary calibrator observations should be a few minutes long and occur about once per hour at low frequency, or as often as every 10 minutes for higher frequencies (depending on conditions and phase stability).

Most compact sources are variable in amplitude on longer times scales, so they need to be calibrated for the observation against 1934$-$638. If you are concerned with accurate positions, you will need to choose a secondary calibrator whose position is accurately known, and which is close to your source. Note that any observations prior to 31 October 1994 which used B1950 calibrator positions will have positional errors up to 0.3" due to neglect of the $e$-terms. Refer to ATCA Bug Reports

To find the best secondary calibrator for your project use the calibrator page . Using this webpage it is possible to find the calibrators close to your source, and view the accumulated information on them.

A spectral line observation will normally require a bandpass calibrator. At low frequencies the primary calibrator 1934$-$638 is sufficient, but at 3cm and higher frequencies, a stronger source such as 1921$-$293 may be required. The ATCA bandpasses are stable, so a single bandpass calibration is sufficient unless very high dynamic range is required.

When selecting calibrators, be careful to find sources that are unresolved and sufficiently strong when observed with the array you are using. At 20cm, calibrators with amplitudes of at least a few jansky should be selected to minimise the problem of confusion. However, the troposphere has little effect on 20cm waves, so your secondary calibrator can safely be up to 20° away from the program source.