Miriad's Calibration Model

Miriad's suite of calibration tasks generally produce or use items which contain the relevant calibration tables and parameters. The tasks to manipulate these calibration tables generally have names start with gp, though there are a few exceptions such as selfcal and mfcal. The calibration parameters are antenna-based ones, with the calibration data being used to derive a model of the response of each antenna. The response of an antenna to radiation is modelled by four sets of parameters:

Ignoring leakage (which is discussed in too much detail in Chapter 12), we model the composite gain function of an antenna (representing both atmospheric and instrumental terms) as

\begin{displaymath}
g(t)g_{P1}(\nu) \exp(i2\pi\tau(t)(\nu-\nu_0))
\end{displaymath}

Here g(t) is the frequency-independent part of the antenna gain, $g_P(\nu)$ is the bandpass function, $\tau(t)$ is a delay term. The delay is calculated with respect to a reference frequency, $\nu_0$. For dual polarisation systems the two polarisation bands are assumed to have independent gains and band passes, although the delay is common.

Note that Miriad datasets cannot contain more than one set of calibration tables - you cannot have multiple versions of calibration tables. Running a calibration task twice will result in the first calibration table being overwritten. This may be inconvenient if the dataset has two sets of data that need to be calibrated separately.

Miriad manager
2016-06-21