The influence of strong sources in a map

During the development of the automatic reduction package we have had several problems. Another big problem occurred in fields with bright sources (e.g. 3C sources with a peak flux density Jy). Near these strong sources the dynamic range of the maps was poor. As a result the local noise in the neighborhood of 3C sources increased by a factor of two to three (see figure 3.4). Visual inspection of the maps before the source extraction could already show areas where the software would encounter problems. Since the first algorithm we used only calculated a global noise estimate, the source list was contaminated with noise peaks. By cutting the strong sources out of the map and redo the extraction process we got a new list without contamination. The strong source had to be measured interactively and later added to the list. This was a very inefficient step in the extraction process and errors were easily made.

Since the noise estimate of the map was based on a global measurement , high noise levels in one part of the map also increased the estimate for the noise in other parts. This resulted in a too large noise estimate and therefore incompleteness of the source list at low flux densities.