Unrealistic theoretical noise estimate
My Rapthor run has stopped due to thinking it diverged (see part of the log below), however the images look fine and instead from the logs the theoretical noise estimate is quite off. It estimates 13.4 uJy/beam as the theoretical level. This is a wider issue that I also noticed with the noise calculators like LUCI in general, which documentation mentions that it follows the calculation from, as that does not take into account elevation, for example. The calculated value is about ~5x lower than what we see empirically from LoTSS.
I didn't record the exact commit I'm using, but I installed Rapthor freshly from git like about a week ago.
INFO - rapthor:image_3 - <-- Operation image_3 started
INFO - rapthor:image_3 - --> Operation image_3 completed
INFO - rapthor:image_3 - Diagnostics for sector_1:
INFO - rapthor:image_3 - Min RMS noise = 129.6 uJy/beam (non-PB-corrected), 172.5 uJy/beam (PB-corrected), 13.4 uJy/beam (theoretical)
INFO - rapthor:image_3 - Median RMS noise = 183.2 uJy/beam (non-PB-corrected), 727.1 uJy/beam (PB-corrected)
INFO - rapthor:image_3 - Dynamic range = 7.1e+03 (non-PB-corrected), 1.3e+04 (PB-corrected)
INFO - rapthor:image_3 - Number of sources found by PyBDSF = 7098
INFO - rapthor:image_3 - Reference frequency = 144.6 MHz
INFO - rapthor:image_3 - Beam = 8.7" x 4.8", PA = 77.7 deg
INFO - rapthor:image_3 - Fraction of unflagged data = 0.68
INFO - rapthor:image_3 - LOFAR/TGSS flux ratio = 0.9 +/- 0.1
INFO - rapthor:image_3 - LOFAR-TGSS RA offset = -0.2" +/- 1.6"
INFO - rapthor:image_3 - LOFAR-TGSS Dec offset = 0.0" +/- 1.6"
INFO - rapthor:mosaic_3 - <-- Operation mosaic_3 started
INFO - rapthor:mosaic_3 - --> Operation mosaic_3 completed
INFO - rapthor - Checking selfcal convergence...
INFO - rapthor:field - Ratio of current median image noise (non-PB-corrected) to previous image noise for sector_1 = 0.99
INFO - rapthor:field - Ratio of current median image noise (non-PB-corrected) to theorectical minimum image noise for sector_1 = 13.71
INFO - rapthor:field - Ratio of current image dynamic range (non-PB-corrected) to previous image dynamic range for sector_1 = 1.21
INFO - rapthor:field - Ratio of current number of sources to previous number of sources for sector_1 = 1.14
WARNING - rapthor - Selfcal has failed due to high noise (ratio of current image noise to theoretical value is > 10.0)
INFO - rapthor - Stopping selfcal at iteration 3 of 8
WARNING - rapthor - Selfcal diverged or failed, so skipping final iteration (with a data fraction of 1.00)
INFO - rapthor - Rapthor has finished :)
Edited by Frits Sweijen