Next: Step 1: Amplitude-based flagging Up: Automated Editing of Radio Previous: Introduction

How Pieflag works

Pieflag reads in data from Miriad data sets and tries to identify RFI-affected sections of data by comparing the visibility amplitudes of each frequency channel to a channel in the spectrum which is known to contain little or no bad data (the reference channel). Consequently, the data must be bandpass-calibrated before Pieflag is used. The hypothesis used here is that data are affected mostly by man-made RFI, which typically is limited to very narrow frequency bands (see for example the spectra measured at the ATCA³). This type of RFI thus can be detected when data from different channels are compared. The amplitudes of atmospheric noise and the astronomical signal usually vary between baselines and coordinates at which the telescopes are directed (a ``pointing''), hence the comparison must be done separately for each baseline and pointing.

The minimum number of parameters Pieflag needs to work is the name of a directory with a Miriad data set and the number of the reference channel. Pieflag works in four steps to generate flags, the first two of which are used to determine which data are deemed to be RFI-affected, and the last two of which transfer the flags to adjacent data which are likely to be affected as well. These four steps are described in the following sections.

Subsections

• Step 1: Amplitude-based flagging
• Step 2: Rms-based flagging
• Step 3: Postprocessing to find small clusters of bad points
• Step 4: Postprocessing to find larger clusters of bad points

Next: Step 1: Amplitude-based flagging Up: Automated Editing of Radio Previous: Introduction