Bright source effect

[Joshuaalbert]

Steps

1. Simulate 1Jy/beam point source at source_offset=f*PB, e.g. [1.5deg, 2deg, ..., 15deg], with antenna pointing offsets with pointing_stdev in [0', 0.5', 1', 1.5', 2'], and calibration errors of 2deg, and no noise. Snapshot 1.5s, full band --> 10.3min doesn't give much.
2. Grid the PB 17k x 17k at 0.87" pixel size, center on phase center.
3. tile FOV into n x n patches, e.g. 5x5
4. For each patch compute RMS --> RMS(source_offset, pointing_stdev)
5. We know thermal noise is 2 muJy/beam in 10.3 min, so 1.1 thermal is 2.2 mu Jy/beam.

@varosero
7. Take our bright source sky model, S(ra, dec), and our pointings (ra,dec) and compute the offset to all neighbours within 15 deg.
8. We want to translate each S(ra, dec) into an effective RMS inside the pointing. We do this by selecting the the closest RMS in offset, and multiply by S(ra, dec). Effectively, S(ra, dec) * RMS(|dist to source|, pointing_stdev) --> RMS'(ra, dec).
9. Then determine which pointings have an RMS'(ra, dec) > 2.2 mu Jy/beam.
10. Should < 95% of all.