Self-Calibration is used to remove errors which remain after performing the regular calibration process. Regular calibration involves making occasional observations of a simple radio source of known structure to determine how the telescope responds to it. This then allows the astronomer to apply the telescope's response to the targeted source and turn its radio emissions into an image. The calibration process, however, is not perfect as it leaves behind errors in the image which can still be eliminated. The errors remain for two reasons.
First, the calibration source and the object under observation lie in separate parts of the sky. Consequently, the array is pointed at different parts of the atmosphere during calibration, then observation. Atmospheric conditions can vary considerably across the sky, affecting the incoming signals distinctly. Such directional differences typically show up as noise in the final image.
Second, it is not possible to observe the calibrator continuously as this would leave no time to observe the object of interest. However, the atmospheric conditions may fluctuate slightly while the object of study is being observed. If such fluctuations happen rapidly, the regular calibration process may not be able to correct for them adequately. Rapid atmospheric changes can add further noise to the final image.
Self-calibration offers a means to remove errors caused by spatial or temporal fluctuations in the Earth's atmosphere. The procedure is premised on the fact that such fluctuations affect the response of each individual receiving dish in the array uniquely. Every measurement from a baseline involving a specific antenna is related back to that dish's signal response. In effect, self-calibration uses some of the data gathered by each dish to more precisely determine its response; another part of this data is combined with signals collected by the other dishes in order to create the image.
After getting the first map, three rounds of self-calibration was performed using the task CALIB. First ones with phase only with solution interval 1min, second one with phase only with solution interval of '0' i.e. averaging over entire scan and the final one with both amplitude and phase and averaging over entire scan. The parameters used in the task CALIB to perform self-calibration were; SOLMODE='P'; APARM(9)=1; NCOMP=0; NMAPS=25 used for the first round. The input data was the UV-data obtained from the earlier, AIPS task SPLIT and the CLEAN model that was obtained after running the task IMAGR. NCOMP was kept to '0' i.e. all the clean components in all the fields were used for self calibration. NMAPS is the number of image files to use for model. When the solution fail or there is insufficient data, APARM(9)=1 preserves the previous calibration. SOLMODE is solution mode, which was phase only.
One more round of phase only self cal was done In the last round the parameters used are : SOLMODE='A&P'. In this iteration, the solution mode used was amplitude and phase.
Typically, three rounds of iterations of the task IMAGR and CALIB (for self calibration) were performed in order to obtain the final image.
The resulting set of facet images were put together to reconstruct the sky using the task FLATN. The FLATNed image was then primary beam corrected using the beam parameters for PBPRM.