It is worth stressing that the final quality in MERLIN images is
usually determined by the quality of the calibration, except for very
extended sources where the image quality is limited by the
coverage. The quality of an image is often measured by its dynamic
range, defined as the ratio of the peak brightness in the image to
the rms fluctuations off-source
. Table 1.1 gives the instrumental noise levels
(which are a function of time on source and bandwidth); these are the
minimum map noise attainable under perfect conditions.
To first order, the dynamic range due to phase errors
radians on a timescale
for an observation with
antennas
for a total time
is:
| (3.9) |
For a typical observation using external phase calibration only, with
a residual phase-error scatter of typically 15
on a timescale of
5 minutes in a 12 hr observation, we may expect a dynamic range of
order 300 (or less for intrinsically faint sources). Under ideal
conditions with close phase-cals it may approach
1000. Self-calibration (see § 3.9.3) can improve the
dynamic range to several thousand for bright sources; beyond this the
map errors may be limited by the the accuracy in determining the
residual baseline-dependent errors (typically 0.1% in amplitude and
0.02
in phase). These can in principle be calibrated using the
target if they are assumed constant over the run
and dynamic ranges of
have been achieved in this way,
e.g. Fig. 3.7.
![]() |
Careful editing of the data is of paramount importance especially for faint sources where limited or no self-calibration is possible. At the other extreme `reverse phase-referencing' allows the accurate position of a strong target to be found using a weak phase-cal.