How to check for excessive light from faculae on the dwarf planet Ceres?

Question

The Dawn space probe took a lot of photos of the dwarf planet Ceres in the asteroid belt, and discovered the strange looking faculae (white spots) in the Occator crater that seem to glow in the photos, se photo below. The scientists that analyzed the pictures determined that the pixels showing the faculae are not saturated. That is, the CCD chip has not taken in excessive light in those pixels that image the faculae. That may very well be true, but the problem is that near saturation the charge accumulation becomes non-linear. Photons come in but, the charge in the "wells" doesn't grow linearly.

I strongly suspect that the faculae actively emit light, like a hot, white glowing object just out of a furnace. From what I understand the following should occur if the "wells" that collect electric charge in proportion to incident light, have received lots of light, that is close to the maximum; the accumulation of charge becomes "non-linear". The curve is no longer linear, but concave down and lies under the straight line. Is there any way that scientists can actually determine where on the curve they are? If you are close to saturation?

The chip in the Dawn camera, was called TH788X. It was apparently used in astronomical cameras in the 80's and also security cameras. Can something be said about it's non-linear properties near saturation? If you want a camera with a similar chip, what camera should you choose?

Update: I wonder if anyone has information on if and how the non-linear characteristics mentioned above, change when temperatures are very low, such as in space? Perhaps this is an issue that no one is really interested in? I mean, let's say you take a picture and then look at it and discover some overexposed areas; the normal procedure then is just to reduce light intake and take another picture, until you get a reasonable greyscale or colour scale.

Generally, are CCD-sensors in space probes heated, or are they just a few degrees above absolute zero?

Answer 1

Presumably, the people processing the images know the characteristics of the CCD sensor, thus, be able to ascertain where in the curve the bright information falls.

CCDs generally have linear characteristics up to a certain well capacity. Some are fairly linear up to saturation, others will deviate some amount from linear near full-well. Since no streaking (blooming) is evident, excessive saturation has not been reached.

The albedo of Ceres is 0.09 making it a fairly dark object. The white areas are presumed to be salts which exhibit a much higher albedo which will give the contrast you see in the image..

Answer 2

I had once done an experiment studying the response of my CMOS camera and it was very linear from bottom to top except noise variation. Sources claiming that response is linear: 1 2 3

But if you resize an image where the the photon noise causes some pixels to clip at white value you will effectively get some kind of non-linear response near the brightest values because clipped pixels will be averaged with darker neighbours. Same applies if you are averaging over an area of pixels some of which are blown out - the result of averaging will be very linear except when some pixels get clipped because of photon noise.