28th of November 2023 |
---|
ASKAP Update for November |
The
ASKAP
update for November reports
on survey progress, the latest ASKAPsoft pipeline
software release and plans to improve ASKAP’s astrometry.
It has long been known that individual ASKAP observations can exhibit
larger than usual astrometry errors of up 6 arcsec in extreme cases,
but usually better than 2 arcsec. ASKAP is susceptible to astrometry
offsets because we do not conduct any phase referencing during an
observation due to the large field of view. Extensive internal
discussion has converged on a theory that the dominant cause of
astrometry offsets in ASKAP is the fact that our bandpass calibration
observations (which provide the fundamental phase reference) are
spread out sequentially for each beam over a time span of more than 2
hours. The ionosphere and other environmental factors do not remain
fixed during this time, so the astrometry of each beam can be
inconsistent.
Planning is underway to test this theory using a local sky model of a
specifically chosen field to phase calibrate all beams
simultaneously. This would use much less observing time than bandpass
calibration and could therefore be done more often.
The figure above shows astrometry offsets (in arcseconds) measured in RACS-low2 fields by comparing the positions of sources visible in two beams simultaneously where they overlap. Different colours represent different survey fields, and the vector origins are placed between the beams from which they were derived, on a plane representing the beam footprint. This shows that neighbouring beams can have different astrometry, and that there are trends following the sequential pattern in which the bandpass calibration is performed. The fact that the pattern is similar for different fields suggests that the initial phase calibration derived from the bandpass observation is more significant than the contents of the science field itself. (Image credit: Emil Lenc) |