Over time the SXT CCD collects some contaminant, which is believed to be water ice. This is still happening, apparently at about the same rate, after more than 2 years in orbit! The problem is first apparent in optical images as noise in the form of short horizontal ``dashes'' randomly sprinkled over the image. Somewhat later, dark absorption ``freckles'' appear in the thin filter images at the same location as the ``dashes''.
The contamination is removed by heating the CCD to +20 C about every 3 months. Sometimes we have waited too long and these contamination artifacts are evident in the images. Although the effect of this contamination on observability of solar features is minimal, we have not yet quantitatively evaluated effects on things like measurement of temperature, emission measure, errors in intensity measurements, etc.
The periods of CCD bakeout are recorded in the appendix of the
Reference Guide and can be obtained with the program PR_DATES_WARM.
A sample call is:
IDL> pr_dates_warm