A Land of Ice & Dust

A three-colour image of the surface of Comet 67P/Churyumov-Gerasimenko from 17km as seen by @Rosetta_OSIRIS NAC on 12 March 2016.

Credit: ESA/Rosetta/MPS for OSIRIS team/Mark McCaughrean

Data & processing details below. 1/

#RosettaForever

Three OSIRIS NAC images at 880.0nm (F41), 648.5nm (F22), & 481.0nm (F24) were combined to make this RGB image. The original Level 3 processed & calibrated images were downloaded from ESA's Planetary Science Archive @esapsanews. /2

Original filenames in time order:

N20160312T213008533ID30F22
N20160312T213020274ID30F24
N20160312T213032416ID20F41

3/

From the filenames, you can see that the images were taken within 24 seconds of each other. The exposure times were 0.400 secs (F22), 0.5760 secs (F24), & 1.7280 secs (F41).

4/

The images have been radiometrically calibrated by the OSIRIS team & were all in units of W/m2/sr/nm. I converted the original 32-bit FITS files to 32-bit TIFFs for further processing.

5/

The three images were initially read into Adobe LightRoom to bring up the shadows – only a few parts of the comet at top-left were in direct sunlight at the time. The same exposure adjustment was applied to all three: no colour stretch or white balance adjustments were made.

6/

The images were then exported to Adobe PhotoShop as layers & aligned using the Auto-Align tool with geometric distortion on. This is very good, but not strictly perfect, as the comet rotated slightly in 24 secs & the Sun illuminated slightly more of the top-left peaks.

7/

That meant that the peaks were "bigger" in the red (F41) filter than in the other two, leading the bright red edges when the three images were combined as R, G, & B channels in a new image. Thus, purely for the aesthetics, I desaturated the peaks to make them white/grey.

8/

The only other editing was to remove quite a few red, green, & blue "blotches", i.e. things seen just in one image. Some were cosmic rays, but others were likely dust & ice in space between Rosetta & the comet, moving between exposures. Removing them seems reasonable.

9/

I then re-exported back to LightRoom for final tweaks. Note: no changes to the vibrance, saturation, or white balance were made, & no texture or clarity processing. Just an increase in exposure & a bit of added contrast to the overall image.

10/

Does that mean that these are the colours the human eye would see looking at the comet? Not really, as you'd need to adjust from flat W/m2/sr/nm to compensate for human eye sensitivity, & besides, 880nm is beyond the human cutoff at 740nm anyway.

11/

But at the same time, the colours haven't been pushed, stretched, or enhanced: I was careful not to do that. So, although Comet 67P/C-G is pretty grey overall, there are real differences across the surface of the comet. Why? Here comes the science finally.

12/

Each time the comet comes closer to the Sun (every ~6.5 years), it gets heated by the Sun & the H2O+CO2+CO ice near the surface sublimates, making the coma & tail of the comet. These gasses also lift dust from the surface into the coma & tail.

13/

As the reddish dust is lifted away, it makes the coma appear redder, while the freshly-exposed & as yet unsublimated ices on the surface are somewhat bluer, as seen here, making the nucleus turn bluer on average as a result.

14/

Then, as the comet moves away from the Sun again, some of the dust released from the surface & the sublimated ice matrix slowly falls back to the surface, recoating it & making it redder. As it does, the coma has less dust in it & turns bluer again.

15/

That coma & surface colour-changing process was directly observed by Rosetta as it followed the Comet 67P/C-G to & from its closest approach to the Sun in 2015. It's described in a paper in Nature by Filaccchione et al. & in this @esascience article.

16/

https://sci.esa.int/web/rosetta/-/rosetta-and-the-chameleon-comet

So, to recap, the colours you're seeing in the original image are the red dust that has been deposited on the surface after previous close approaches to the Sun, & bluer icy material, freshly revealed by the heat of the Sun around the 2015 perihelion passage. They're real

17/

Coda: this process of gradually depleting the volatiles (ices etc.) from a comet while coating its surface with the dust component ultimately leads to the dust preventing the Sun's heat from reaching the deeper ice, so that cometary activity stops. Comets effectively die.

18/

Indeed, it's thought that there are many "dead comets" in the solar system, now inactive. Dave Jewitt & others have worked on this topic, looking for dead & almost-dead comets for years. For the technically-minded, here's a 2003 review by Dave.

19/

http://ifa.hawaii.edu/~meech/a740/2004/fall/papers/J2004C.pdf

Anyway, that's it for now. Back to the work I'm supposed to be doing, although as that includes reviewing recent science papers from #Rosetta, this this sort of counts. Still, the old adage applies: don't go to the Rosetta archives unless you have time on your hands

20/20

Ah, forgot to include this detail: at a distance of 17km from the surface of 67P/C-G, each OSIRIS NAC pixel covers 32cm. This slightly cropped image thus covers ~650x650 metres, perhaps a bit larger than you might initially think.

21/

And if you'd like to download a better-quality version of this Rosetta OSIRIS colour image of Comet 67P/Churyumov-Gerasimenko, I've put it on Flickr.

I've made it CC-BY, so do as you will with it, just as long as you give credit as in the original tweet. https://flic.kr/p/2j6KkiU