Available color data for Europa is so severely limited that reprocessing the Voyager data is still rewarding after all these years. Here is my version of the Voyager 2 color mosaic from 240,000 km. It's a mosaic of 5 color frames, each of which is composed from orange, blue and violet images:
Click to view attachmentThis is actually the *only* hi-res Voyager color observation of Europa (that is, if 2.4 km/pixel qualifies as "hi-res"). The second highest resolution Voyager color observation is this one by Voyager 2 from 1.2 million km - note the color difference between the dark terrain at left and the dark terrain at right:
Click to view attachmentBoth of these images are shown slightly bigger than their original size to compensate for the slight loss of resolution due to resampling in some of the image processing steps. They have also been sharpened a bit for the same reason.
These images highlight the really poor color coverage of Europa (in particular, both Voyagers obtained better color coverage of both Ganymede and Callisto). Galileo obtained a 1.4 km/pixel IR756-GR-VI mosaic during its E14 orbit showing largely the same terrain as the Voyager mosaic. There are several versions of the E14 mosaic in circulation; the best known one is probably Ted's version. The Galileo E14 data is almost global but has some gaps. There is also a partial 1.6 km/pixel color image from G1 and a very noisy 1.4 km/pixel color image from E12. In addition there is near-global coverage at much lower resolutions (by far the best one is the 7 km/pixel global G2 image). In addition there are some high resolution color samples but these are extremely limited in coverage.
Some notes on the color processing in the mosaic above:
I used the calibrated and geometrically corrected Voyager images available at the PDS Rings Node as source data. The flatfielding in these images is exquisite; this is particularly important when mosaicking images showing rough terrain near the terminator. The only drawback is that in my opinion the reseau removal could be better in some cases - sometimes an excessive number of pixels gets smoothed but this is easy to fix in Photoshop once the color channels have been aligned.
Voyager 2 was imaging Europa using the orange, blue, violet and ultraviolet filters. At first glance this may not seem ideal since there are no green filter images but things are not that simple. The effective wavelength of the Voyager filters is OR=589 nm, GR=564, BL=475, VI=402 and UV=332. One important thing to note is that there are only 25 nm between the OR and GR filters (and lots of overlap.) This makes it rather difficult to create synthetic red from OR and GR by extrapolating - the results are usually bad (it should be noted though that the OR and GR filters are pretty different even though 25 nm isn't a big number). I usually get much better red images by using OR and BL (or even VI when BL isn't available). Luckily, Voyager 2 acquired test images of Jupiter using all of the filters at the start of its approach phase and these can be used for checking various filter combinations. C1838155 is the first one of these images. I used these as 'ground truth' when testing various ways of making synthetic images (red etc.) since I have a pretty good idea of what Jupiter's color should look like. Once I had acceptable color of Jupiter from OBV images I used identical color processing for Europa. I'm fairly happy with the resulting color - in particular I get very similar color when processing the Galileo E14 stuff even though the filters used are different. I also experimented with interpolating the data to get the entire visible spectrum and then converting it to sRGB. The results were similar (slightly stronger green color component though).
For the interested, the synthetic colors were created using these formulas:
R = 1.526315789 x OR - 0.52631579 x BL
G = 1.01 x (0.719298246 x OR + 0.280701754 x BL)
B = 1.07 x (0.726027397 x BL + 0.273972603 x VI)
This results in fairly realistic images of Jupiter with whitish zones etc. - the 1.01 and 1.07 values are there mainly to make the zones more whitish. The 'target wavelengths' here are R=649 nm (the Cassini red filter), G=557 nm (the Galileo green filter) and B=455 nm (the Cassini blue filter). The Cassini green filter is 569 nm which is too close to yellow for my taste. This approach might work fairly well for the entire Voyager 2 Jupiter data since the source data is calibrated but I need to check that more carefully. In particular, Io is a likely exception because of its weird spectrum.
The original data had some gaps but they aren't big and in all cases two colors are available, making it possible to fill the gaps using synthetic data. The location of the missing data can be seen in this ancient version that was released back in 1979:
http://photojournal.jpl.nasa.gov/catalog/PIA00366