As I understand it CCD sensors are sensitive to the high level of electromagnetic radiation present in space, which will knock them out of commission in a short time period. Therefore they are not used in satellite applications, CMOS sensors are immune apparently. I think that some of the credit for the success of CMOS based cameras may be due to early R&D efforts from space based applications.
Virtually every Hubble Space Telescope image you've ever seen has been taken with a CCD or array of CCDs (there have been a couple of specialised non-CCD UV and IR imagers too, but their results rarely made the press). The WFPC2 workhorse camera had 4 CCDs and operated successfully in space for
15 years before it was removed to make room for its successor, the WF3 upgrade.
CCDs are highly efficient cosmic ray "detectors" (this is an unwanted property!), so above Earth's atmosphere the images become quickly peppered with spikes and streaks. A raw Hubble long exposure has so many hits that it almost looks obliterated. But we simply stack lots of them with statistical filters which operate on the principle of "lightning doesn't strike the same place twice" - given say 10 identical exposures, then at any given pixel position there will be at most 1 or 2 cosmic ray hits. Such pixels are easily detected and their signal is discarded from the stack.
Prolonged exposure to this radiation flux does have an "aging" effect on a CCD, increasing its average dark current rate. Since the CCD is massively cooled anyway, this is not much of an issue. Charge transfer efficiency also worsens somewhat. But it would not be true to say that they are knocked out of commission by any of this.
Ray
