Rosetta - Post Separation Ops at Comet 67P C-G, November 14, 2014 - Options

[Explorer1]

I think I heard it mentioned during the press conference today, (I can't find it now), about Rosetta itself possibly landing eventually, similar to what NEAR did at the end of the main mission at Eros? Since it's not like there's anywhere else to go with the remaining delta-v left by the end of 2015, and sunlight levels and activity starting to drop after perihelion, and the low gravity makes the difference between orbiting and 'landing' trivial. The whole thing would weigh a kilo or two, right?
Obviously there's a few more pressing concerns right now, but it's something to eventually think about.

[Gerald]

At some point Rosetta will run out of propellant for orbit corrections.

The mass of Rosetta at regular EOM should be 2900 kg - 660 kg - 1060 kg - 100 kg = 1080 kg.
(Start weight - propellant - oxidizer - mass of Philae, from here)
The surface gravity of the nucleus at 2 km distance from the center of mass should be
g = GM/r² = (6.672e-11 Nm²/kg² * 1e13 kg) / (2000 m)² = 1.668e-4 N/kg.
The weight of 1080 kg is F = m a = 1080 kg * 1.668e-4 N/kg = 0.18 N.
0.18 N is the force of a mass of 18.37 grams in 9.80665 m/s² gravity.
The actual weight of Rosetta would be a little less due to inertial forces by the rotation of the nucleus.

To be more precise: The centripetal force for a radius of 2000 m and a rotation period of 12.7 h = 45,720 s is m r 4 pi² / T² = 1080 kg * 2000 m * pi² / (45,720 s)² = 0.0102 N.

So we are at 0.18 N - 0.01 N = 0.17 N for Rosetta's EOM weight, corresponding to the weight of a little more than 17 grams on Earth.

That's a model estimate, and may differ, depending on the actual landing coordinates.

[MahFL]

They have already stated they want to follow the comet during 2016, as it enters the dormant state. They would need a mission extension for that.

[neo56]

NavCam mosaic taken on 6 November 2014, with Agilkia landing site on top of the top lobe:

[tolis]

Since it's not like there's anywhere else to go with the remaining delta-v left by the end of 2015, and sunlight levels and activity starting to drop after perihelion

Actually, it will go somewhere, after a fashion.

67P will approach Jupiter to within 0.4 AU in Nov 2018.

Whether Rosetta will be able to do anything about it, out of power and out of fuel, is another matter.

[astroman]

"rejoined" (forgot password) for this epic event- hopefully it will spawn a co-op esa-nasa cometary exploration program - especially good time being alive for following an historic and species first...


ADMIN: Post moved.

Attached thumbnail(s)

 

[SpaceScout]

Deutschlandfunk says the comet's surface remind of corals... Interesting comparison but might lead to some misunderstanding...

[katodomo]

"Coral reefs under the sea" is what Paolo Ferri supposedly said in a radio interview for HR-Info - which is where Deutschlandfunk took it from, among others. Interview isn't available online (yet?), so can't check it for certain.

Considering quite a lot of people aren't even aware corals are alive i think the extent of misunderstandings will be limited

[fredk]

I think I heard it mentioned during the press conference today, (I can't find it now), about Rosetta itself possibly landing eventually, similar to what NEAR did at the end of the main mission at Eros?

From this New Scientist story:

Rosetta itself could one day join Philae on comet 67P. The orbiter will run out of fuel at the end of 2016, and ESA must decide whether to put it into hibernation, or put it down on the surface.

A large flat area on the dark side of the comet was not an option for Philae, but it will be well illuminated by 2016. Rosetta could crash-land there, taking extreme close-up pictures of the comet and sniff its atmosphere on the way down. "If we are called to do something like this I would be pleased," says Accomazzo. "If you ask me personally, I wouldn't do anything else."

[Y Bar Ranch]

From this New Scientist story:

Obviously it'd take a lot of precise planning, but if in the end game it was possible to drift down, and then blip the thrusters for the next-to-the-last-time to clean off the overlaying dust from the icy underside...

[neo56]

NavCam mosaic of 17 November:

[Phil Stooke]

The long wait for an OSIRIS color image may be over:

https://agu.confex.com/agu/fm14/meetingapp.cgi#Paper/22395

The AGU abstracts are online and full of good stuff, and a very small number of them contain images, including this one.

Phil

[ugordan]

The color mottling in that image is very interesting (other than obvious color fringes caused by rotation between frames), I wonder if that's real or an instrument artifact.

[neo56]

Many talks will be given to present the results from the instruments of Rosetta.

[Malmer]

Made a new animation of navcams draped over my shapemodel. This time with the november 17 navcam quad.

My shapemodel is becoming better with each iteration. The match between features in the images and the shape of the model is surprisingly good if I may say so myself

http://mattias.malmer.nu/2014/11/navcam-november-17/

[Malmer]

The long wait for an OSIRIS color image may be over:

https://agu.confex.com/agu/fm14/meetingapp.cgi#Paper/22395

The AGU abstracts are online and full of good stuff, and a very small number of them contain images, including this one.

Phil

That is one very strange image indeed... the mottling is confined to a broad stripe of the image starting abruptly about 20% in from left and ending about 10% in from the right. It almost looks a little bit like it was added in post. (Scientific DRM?)

The "non mottled" areas look much nicer...

[climber]

I notice that Rosetta/Philae topics get (so far) about 1500 posts and 300.000 views. Wondering if this beat Curiosity around landing time?

[Spin0]

Newly published Navcam images of the region, 2.59 m/pixel: http://blogs.esa.int/rosetta/2014/11/24/co...ch-20-november/

ADMIN NOTE: Post moved from Philae topic to correct discussion.

[neo56]

Very nice animation Mattias, as always

NavCam mosaic taken on 20 November:


The shadow of the upper lobe is slightly visible on the coma.

[fredk]

the mottling is confined to a broad stripe of the image starting abruptly about 20% in from left and ending about 10% in from the right.

Very obvious if you extract the saturation channel and stretch:

Very sharp boundary between mottled area and unmottled to the right. The boundary's clearly unrelated to anything on the comet.

[Malmer]

Yes. And the mottling seems very regular. Like a Perlin noice function...

The parts outside the area are quite nice. One could perhaps work a little on the registration of the images to get less fringing. But you can clearly see the colour of the surface change on the different terrain types. (Especially if you remove the overall red tint)

[4th rock from th...]

Looks like noise / low resolution data. Perhaps some channels are of much lower resolution.

[ugordan]

Thing is, looking at the rotational state in each of the rgb components, it's suggestive that the source frames cover the entire body, not a case of a colorized greyscale image at the center. Very odd.

[Gerald]

A strongly hue and saturation stretched (hence false-color) version of the colored OSIRIS image:

My overall, somewhat subjective and qualitative impression is, that bluish spots are correlated to holes/depressions, and reddish spots to peaks/hills.
I can't say, whether this also means some correlation of hue to activity.

From the abstract of the paper I've some hope, that the OSIRIS team might have been able to quantify this presumption.

[Malmer]

My bet is that the image has been intentionally degraded to make sure that it is not used by anyone to do any science.

It is just way to much low frequency noise in the individual channels and there seem to be almost no correlation between surface type and color.
The images from OSIRIS we have seen so far are very very nice. there seem to be no noise pattern of any kind in those.

If one instead looks at the areas outside the noise stripe one is treated to a much higher quality.

[Michael Capobian...]

Is anyone interested in matching up the diagram of probable Philae landing zones with the recent Navcam images showing those parts of the comet?

[Floyd]

The zones are just on the other side of the dark grey ridge. Our point of view is 180 degrees off from what we would need to spot Philae.

[Malmer]

Very nice animation Mattias, as always

Thank you. I think it is a lot of fun to do. It makes the comet more like a physical object to see it like that. And I get to learn the surface very intimately. I have over 150 surface features that I match over the images. They each have a number. Its strange to have a private toponomy like that.

I go: -ooh look there's a nice image of twelve. Now lets see if i cant find seventeen in this picture. nope. At least I can see 119. that's enough for the camera solve...

[Phil Stooke]

The jets are really multiplying now! Also, I think this is the first time the dark side has been visible silhouetted against the background coma - we've seen it before with Halley and Hartley-2 but not here.

http://blogs.esa.int/rosetta/2014/11/26/co...ow-of-the-coma/

Phil

[SteveM]

Just noticed -- actually just realized what I've been seeing for some time -- that the jets are primarily pointing toward the Sun, while a comet's dust and ion tails point away from the Sun. Can Rosetta observe the process by which the trajectory changes as they become tails, or is Rosetta too close to see that effect?

[DFinfrock]

... the jets are primarily pointing toward the Sun, while a comet's dust and ion tails point away from the Sun. Can Rosetta observe the process by which the trajectory changes as they become tails, or is Rosetta too close to see that effect?

Absent any atmosphere, there is no air friction to slow down the jets at they stream outward from the surface. My guess is that as the comet's surface heats up (on the sun-ward side) the jets would be streaming directly away from the surface, or towards the sun. It would only be when the comet falls deeper into the sun's gravitational well, and picks up speed, that the tail is "left behind".

I await someone more knowledgeable about comets to correct me.

[Phil Stooke]

OK...

http://hubblesite.org/reference_desk/faq/a...cat=solarsystem

Phil

[fredk]

My understanding was that it was the pressure of the solar wind that forms the tail(s). So heating on the sun facing side leads to jets towards the sun, which eventually are deflected by the solar wind to form the tail.

If there were no solar wind, particles in a jet pointing towards the sun would presumably get ahead of the comet rather than be left behind (at least prior to perihelion).

Edit: Phil beat me by less than a minute!

[SteveM]

Thanks for the comments and for Phil's good link. The question remains, however, of what details Rosetta will be able to provide about the process changing jets emerging on the sunward side of the nucleus into tails streaming away from the Sun.

Steve M

[Gerald]

Small grains should be more susceptible to solar wind and radiation.
The GIADA instrument should be well-suited to investigate this question in very detail:

GIADA (Grain Impact Analyser and Dust Accumulator) will measure the number, mass, momentum and velocity distribution of dust grains in the near-comet environment. Giada will analyse both grains that travel directly from the nucleus to the spacecraft and those that arrive from other directions having had their ejection momentum altered by solar radiation pressure.

[SteveM]

GIADA should provide interesting velocity data. Look forward to more GIADA insights into dust particle trajectories since since this early detection of a few dust particles.

[neo56]

NavCam mosaic of 26 November. The rotation of the comet has been substantial in the twenty minutes that passed between the two lower images, resulting in artifacts in the stitching process.

[eliBonora]

1 December NAVCAM from 30.1 km

[ngunn]

Beautiful. Thanks for sharing that. It's amazing to pan around that view. What a place!

[Dan Delany]

Lovely mosaic eliBonora! The one on the ESA blog has a blurry seam, where the foreground and background limbs meet, that had me scratching my head until I saw yours

Here are a couple attempts at anaglyphs from the Dec 1st and 2nd NAVCAM images, made from the slices where the bottom two (foreground) images overlap. Some artistic liberties have been taken to account for areas which were shaded in one image and lit in the other.

Attached thumbnail(s)

 

[eliBonora]

Lovely mosaic eliBonora! The one on the ESA blog has a blurry seam, where the foreground and background limbs meet, that had me scratching my head until I saw yours

Thank you Dan. Nice anaglyphs.
I think the difference is that we combine these comets by hand and we don't use the software's blending option.

[neo56]

My take on NavCam mosaics from 30 November to 7 December:

[Phil Stooke]

The latest Navcam blog shows new - or apparently new - features in the neck. Here's a comparison between October and December images (December on the right). I would suggest a close examination of images would probably show more things like this.

Phil

[Phil Stooke]

Sure enough, here's a comparison between 24 September and 9 December with the new feature on 9 December noted. Differences in lighting and resolution are small enough that I think this is a real change.

Phil

(PS... must work on book... must work on book... aaaargh!)

[walfy]

...I think this is a real change.
Phil

Wow, those changes are amazing! Looking forward to the discussions about this.

[Gladstoner]

I would guess the pit is a low profile feature that produces shadows only when sunlight is at a very low angle. Otherwise, the lack of albedo differences in the smooth material renders it invisible otherwise.

Of course, the very recent formation of a collapse pit would be pretty cool....

[Phil Stooke]

I don't think the lighting is sufficiently different for the pit to not show up in the earlier image. What you describe is always a possibility in this type of comparison, but I don't think it applies here.

Phil

[nprev]

If it's all real, and it sure looks like it is, this will set some significant constraints on the life expectancy of comets in the inner Solar System. Could be a huge mission finding.

Well spotted, Phil!!!

[Phil Stooke]

I spotted the second example, but it was the Rosetta blog:

http://blogs.esa.int/rosetta/2014/12/11/co...tch-9-december/

which drew attention to the first one. I just illustrated that first one.

Phil

[Gladstoner]

If the pits did form very recently, then the smoothing process must have been recent as well.

[jmknapp]

Might the marks be the result of some fragment landing back on the comet after being wafted up by some outgassing activity elsewhere?

[Gerald]

Non-linear brightness stretched (about quadratic) and 8-fold saturation enhanced version of the official OSIRIS color image:


Brightness-stretching gets the grey value nearer to 50%, nonlinearity avoids more-than-necessary numerical instability near full saturation.

My first out-of-the-hip impression has been, that some of the color could be structural.

[eliBonora]

We have tried to pull out more color from the OSIRIS image, not because it 's what the human eye could see but because it's interesting to highlight the differences. Of course we can exasperate more and more but we have preferred to maintain a balance with a pleasant look.



Here's also a curious gif created with the three RGB channels splitted:
https://www.flickr.com/photos/lunexit/15820377179/sizes/o/

[Phil Stooke]

"Might the marks be the result of some fragment landing back on the comet after being wafted up by some outgassing activity elsewhere? "

That's exactly what I would expect for these specific changes.

Phil

[jmknapp]

Request to the comet gods: puff Philae out of its hole so it lands elsewhere, in the sun and upright of course.

[wildespace]

Great to finally see the comet in colour! The original OSIRIS composite has greenish cast (common to raw images, it seems, judging by the Mastcam and MAHLI raw images), so I've used "Auto color" in Photoshop to normalise the colours, and then decreased levels to reflect its dark albedo. This also brought out some colours, without the need for enhancing saturation manually:



The neck region seems definitely lighter and somewhat bluer, perhaps it's the un-darkened material being exposed by the comet's activity?

There is interesting golden-red colouration at the bottom of the larger lobe.

[Explorer1]

Request to the comet gods: puff Philae out of its hole so it lands elsewhere, in the sun and upright of course.

Or just sublimate and crumble the overhang above it while we're in the wishful thinking stage....

[neo56]

NavCam mosaic of 9 December:

[Gerald]

The original OSIRIS composite has greenish cast (common to raw images, it seems, judging by the Mastcam and MAHLI raw images), ...

OSIRIS used different color filters for the same CCD, not a Bayer pattern like MSL Mastcam or MAHLI (although Mastcam can also use color filters). So it's more a matter of how the OSIRIS team calibrated colors.

From this Rosetta blog:

A more detailed first analysis nevertheless reveals that the comet reflects red light slightly more efficiently than other wavelengths. This is a well-known phenomenon observed at many other small bodies in the Solar System and is due to the small size of the surface grains. That does not, however, mean that the comet would look red to the human eye. Natural sunlight peaks in the green part of the spectrum and the response of the human eye is similarly matched. Thus, overall, the comet would look rather grey to the human eye, as seen here.

[surbiton]

The neck region seems definitely lighter and somewhat bluer, perhaps it's the un-darkened material being exposed by the comet's activity?

Could it be the interior "ice" exposed ?

[Brian Lynch]

Could it also be that the neck region is occluded by the two lobes so that less dust and other material contributing to the dark colour is able to aggregate on the surface as the comet travels through interplanetary space? Admittedly, I have little expertise in the science of comet formation so that might be nonsensical (ie. is the surface colour even expected to be due to such aggregation?).

[neo56]

NavCam mosaic of December 10th:

[Hungry4info]

Focous on the surface change noted by Stooke in the 10 Dec 2014 Navcam image.

Attached thumbnail(s)

 

[Gerald]

Could it also be that the neck region is occluded by the two lobes so that less dust and other material contributing to the dark colour is able to aggregate on the surface as the comet travels through interplanetary space? Admittedly, I have little expertise in the science of comet formation so that might be nonsensical (ie. is the surface colour even expected to be due to such aggregation?).

I think, we've to go back to or even before the formation of our solar system to get most of the interstellar dust the comet is made of.
The net accumulation of interplanetary dust on the present orbit is probably much less than dust ejection. Therefore the currently accumulated dust on the surface - although originally material of the protoplanetary disk - has more likely been material of the comet than been collected recently from interplanetary space.

Exposure to radiation (solar and galactic) might be less in the neck region, leading to a different chemistry over long periods of time. But surface dust is probably reworked due to activity, such that this kind of radiation-induced differences should be hard to detect.
An approach based on activity and varying settlement of cometary dust will probably be more promising.

The dark surface color is (very likely) mostly due to carbon-rich compounds (somewhat similar to soot or asphalt), and to the fine-grained (almost fractal) surface structure (rough surfaces tend to be darker than smooth ones). Grain size influences color. Structural colors (in contrast to pigment colors) are caused by surface structures the size about the wavelength of visible light. Areas on the surface of the comet may vary in their preferred accumulation and ejection of grains of a certain size or structure, and vary in color by this mechanism. Pigmentation (chemical composition) may vary for a similar reason.

[Bill Harris]

My study of changes on the North Polar Plain 24 September '14 to 10 December '14.

http://univ.smugmug.com/Rosetta-Philae-Mis..._activity-L.png

http://univ.smugmug.com/Rosetta-Philae-Mis...arative-Series/

--Bill

[Habukaz]

Request to the comet gods: puff Philae out of its hole so it lands elsewhere, in the sun and upright of course.

Well, the original plans expected Philae to cook to death by March 2015. If it then instead wakes up from slumber; then, with some luck, the total number of months available for science on the surface could exceed the original plans.

IIRC, there was some talk about the possibility to spin up the flywheel as a last-ditch attempt at getting Philae away from its current position before the batteries ran out. If they could successfully pull that off when the insolation at the surrounding planes is habitable, they could possibly extend the lifespan even further.

So I think I'll sacrifice some dry ice for the comet gods in the hope that Philae stays put, for now.

[neo56]

NavCam mosaic of December 14th:

[Gerald]

That's a quick and very dirty try to get a 3d visualization of the first MIDAS dust grain scan as x-eyed stereo flicker, not really recommended, unless you need to know and are ready to risk some head-ache or sea-sickness, more an idea for someone who finds time do better:

raw image: ESA/Rosetta/MIDAS/Space Research Institute (IWF) in Graz, Austria/Mark Bentley
modified by "Gerald"

[Habukaz]

This cartoon might be the best rendering that we have of the 'dinosaur eggs' for a while...


(via @Geo_Miles)

[elakdawalla]

Just to be clear, though, the "dinosaur egg" pictures were not from Philae, they were from OSIRIS. I'll do my best to describe them: imagine a cliff face. Now imagine, instead of the usual kind of striations you'd see in a cliff face, that it was instead made entirely of gumballs. Or maybe peanut M&M's, since they didn't seem quite spherical, but rather slightly non-spherical, but still distinctly round. All of a uniform size. The scale bar on the image was 5m long, and the dinosaur eggs were somewhat smaller than that; Sierks said 2 or 3 meters. It was wacky-looking.

[Explorer1]

Yep; kinda wild that those images are being guarded at the same level of secrecy as surveillance satellites...
Presumably an image of Philae on the surface, once acquired, will be released earlier. Surrounding details can be easily cropped out after all.

[elakdawalla]

They said that images had been acquired in a sequence planned for when Philae should be in sunlight, but that the images weren't on Earth yet. I think.

[neo56]

A movie has been released with NavCam mosaics taken in November and December.
Informations on the making of this movie and files containing both the mosaics and the individual pictures are available here.
They also provide SPICE data for each pictures.
They invite amateurs to do their own mosaics and movie : challenge accepted !

[jmknapp]

In the image information file, how is "rotation phase" defined?

[neo56]

After some work of stitching, here is my animation of the rotation of comet Churyumov-Gerasimenko. Full resolution is available here.

Attached thumbnail(s)

 

[neo56]

I'm working on a smooth version of the NavCam animation.
I have included 7 mosaics yet and I'm pretty happy with the result! Pictures between each NavCam mosaic are interpolated using a morphing software: MorphX on Mac.

[aholub]

Very nice animation. Much better then ESA original. Thanks for creating and sharing.

[Jackbauer]

First NavCam image of 2015 :
http://www.esa.int/spaceinimages/Images/20...ary_2015_NavCam

[Phil Stooke]

Can't stop, rushing to class, but the newest image has lots of surface changes in it.

Phil

[Hungry4info]

Some surface changes I've identified comparing it with the 9 Dec 2014 NavCam image. It looks like some of the 'ripples' (I am sure there's a better term) that were first seen in that NavCam image have been erased by surface flow. Many new such features have appeared, however.

Attached thumbnail(s)

 

[Phil Stooke]

Here are two close-ups, same features as Hungry4info circled - if you look carefully there are others as well. Also a third post which is derived from my earlier one but adds a third time step. The pits shown with arrows look like the same feature until you check carefully, and they are different.

I have searched other areas on the nucleus in vain, looking for changes there.

Phil

[flug]

Some surface changes I've identified comparing it with the 9 Dec 2014 NavCam image. It looks like some of the 'ripples' (I am sure there's a better term) that were first seen in that NavCam image have been erased by surface flow. Many new such features have appeared, however.

I'm wondering to what degree those are new/changed features, and to what degree they are features that were there in the earlier images but revealed far more clearly by the sun angle/shadows in the later images.

For example, in the 24 Sept/2 Oct/9 Dec image, at first I thought the features indicated by the arrows in the 2 Oct & Dec were the same feature that had moved. But examining the 24 Sept & 2 Oct images more carefully, I think I see the same feature indicated by the arrow in the 9 Dec image. The lighting is just more directly from above in the 24 Sept & 2 Oct images, so there are no visible shadows. On 9 Dec, the sun angle is more oblique, giving far more obvious shadows.

Similarly for the object indicated with an arrow on the 2 Oct image--I can see those outlines in both 24 Sept & 9 Dec as well.

These **might** be pretty much identical features just seen under different lighting. Or they **might** have subtly changed between images--subsidence, perhaps? But I don't think they have **dramatically** changed or shifted.

FWIW this is somewhat informed by some hours spent observing the moon through a telescope. Features that are very obvious under oblique lighting that brings out the shadows will almost 100% disappear under more direct lighting. I'm not the most experienced observer by any means, but these features look in some ways similar to those lunar features that appear/disappear depending on sunlight angle..

I'll be curious what you all think!

[Hungry4info]

The illumination angles in these images appear to be similar, based on the shadows we see.

[Gerald]

I think, Phil is (one of?) the most experienced experts in this field, and identifies textural changes much better than me. So if I'm unsure, I'm attributing this to my lack of sufficient experience.

In this special case, the central pair of images shows rather similar illumination. Some subtle features look rather similar (unchanged) in both images, such that I strongly tend to agree with Phil, that some deepenings show up in the right image, exclusively.

But you're of course right, that changing illumination needs to be considered.

[Jackbauer]

Alléluia !! An image of OSIRIS !! Old (november 22) but spectacular :

http://www.esa.int/spaceinimages/Images/20...2_November_2014

[MahFL]

Looks to me the main jet is made up of about 14 jets all combining together, awesome.

[Jackbauer]

And now the daily NAVCAM image :
http://www.esa.int/spaceinimages/Images/20...ary_2015_NavCam

[Phil Stooke]

Look what happens when you stretch the contrast in the OSIRIS image: a beautiful silhouette at the bottom, and the shadow of the nucleus on the coma.

Phil

[elakdawalla]

Cool. Is it correct to say that the jets appear to curve because material within them is moving radially outward from the comet, while the comet (and hence the jet source regions) is spinning?

I am amused that the first phrase in the blog entry is "In the first OSIRIS image release of 2015..." We're supposed to be getting one per week -- even the ESA bloggers can no longer conceal their exasperation about that!

[fredk]

I suspect we're not seeing the "lawn sprinkler effect" here - to see curvature in the jets you'd need the speed of the jet material to be not too much more than the rotation speed. I don't know what the expected order of magnitude speed for the jets is.

To me it looks more like we're seeing different (more or less straight) jets in different directions superposed along the line of sight. So one jet in one direction in behind, and other jets pointing in a bit different directions in front. You'd expect jets from separate vents to travel in somewhat different directions.

[Gerald]

That's an annotated, brightness-stretched resized crop of this NavCam image:

The streak is probably the track of a dust grain, the red line as a reference line. The streak is curved. This indicates acceleration.
The cause may be drag by gas coming from a different direction, interaction with solar wind, or interaction with photons.
These possible mechnisms may also apply to dust jets.
Small charged grains may also interact with Sun's magnetic field (Lorenz force).

So we've several possible causes for the curvature of the jets.
See also this FAQ, or a little more detailed here. Wikipedia version.

[MarsInMyLifetime]

The change of direction seems rather more abrupt than gradual. Is it possible that frozen particles in a conglomerate might finish sublimating in a puff, sending their inert carrier particle in an opposite tangent?

[Gerald]

Break-up of some hydrated state? I couldn't rule this out, easily. But particles should be of very different size, otherwise we would see a fork.
I've also been considering electrostatic interaction (attraction / repulsion) with the spacecraft. But for this, the motion should be almost perfectly parabolic, which I doubt that it is.
Collisions with fast-moving tiny grains (less than 300 nm diameter to be invisible as individual particles) may be possible, too.

[Habukaz]

ESA's director general Jean-Jacques Dordain expresses his frustration and sympathies with the current image release policy of Rosetta (and other missions); but no signs of any changes coming soon:

"Even I've tried to get more data," Mr Dordain said. "I might be the DG but I'm also a fan of Rosetta and [its lander] Philae. It's a problem; I don't deny it's a problem. But it's a very difficult problem, too," he told BBC News.

"I understand the frustration of the public and the media, but, on the other hand, I understand the position of the principal investigators who have invented the mission."

Mr Dordain was speaking in Paris at his annual New Year breakfast with reporters.

[...]

"Maybe what we should do is distinguish better between data that would be considered absolutely key to making scientific discoveries and can be kept under wraps before publication [in journals], and the data that can be released to the public much sooner."

[...]

The DG said that current arrangements needed to be adhered to (on all sides), but added that polices as a whole should be reviewed.

[Gerald]

... I don't know what the expected order of magnitude speed for the jets is...

Vapor from sublimating ice should relax with about the speed of sound of the specific gas into vacuum.
Very small grains should be dragged the same speed.
Larger grains may be slower, down to zero for larger grains not able to escape from the comet.
With this approach I'd suggest a few hundred meters per second as velocity of the majority of the jets.

I thought I've read 400 m/s without explanation somewhere. The above approach is inspired by this vague memory.

I'm neglecting interaction with solar wind and uv radiation on the current level of activity and proximity to the nucleus.

[fredk]

Thanks, Gerald. For a 3 km radius, we have roughly 0.5 m/s rotation velocity, which is far smaller than hundreds of m/s. So indeed, for the near-thermal-velocity particles, it looks like we wouldn't expect to see curvature in the jets. Picture a sprinkler rotating extremely slowly...

[Sherbert]

I was wondering if interaction between the jets and the material of the inner comma may, by friction or ion interactions, tend to curve the tops of the jets as they are, initially anyway, tied to the rotation of the nucleus. The OSIRIS image from 22nd Nov seems to show some curving of the jets further from the nucleus. The very low densities involved would tend to make me think these effects would be rather small. The velocity of inner coma material relative to the surface would need to be known, Rosetta does have instruments to determine that, though.

[eliBonora]

Last NavCam mosaic: 10, 12 and 16 January

https://flic.kr/p/pTwGmz
https://flic.kr/p/qRfVAu
https://flic.kr/p/qRfVAu

This shows the same view of 2 November but under different lighting
(here in two versions https://flic.kr/p/pEDHFi https://flic.kr/p/pEY7qj)

[Habukaz]

As per the scientific results thread, 14 newly released OSIRIS images can be found here. Per Jonathan Amos, the dinosaur eggs mentioned at AGU are here referred to as 'goosebumps' (pic 13).

EDIT 2x: Some more context in this blog post.

Here is OSIRIS at full resolution from 8 km - almost like standing on the surface!

[Habukaz]

What the heck is this thing?



(from the 8 km)