Bias offset determination in WFC images is performed by calculating
the bias level in selected columns of the leading physical overscan.
The calculated value is then subtracted to the entire active area
which show a slightly higher bias level. In principle, if the offset
between active area and overscan region is constant the full frame
bias subtraction should remove any residual difference. Unfortunately
this offset is not constant but shows random variations of the order
of 0.2-0.3 DN. This causes a small but noticeable jump in the middle
of each chip, in correspondence of the amplifier edge. We present a
study of the structure of the WFC bias frames, a statistics of all the
data acquired since ACS installation and the amplitude of the residual
offset. We studied the scientific implication of such residual and
present a possible approach that could mitigate the problem.
Back to publication list
ACS home page