Original NAVCAM frame taken on 12 May 2015. Credits: ESA/Rosetta/NavCam – CC BY-SA IGO 3.0
Today’s CometWatch entry shows a curious view of 67P/Churyumov-Gerasimenko that on first impression (see original image right) gives the illusion that the comet’s head is missing. The image was taken by Rosetta’s NAVCAM on 12 May from a distance of 166 km from the centre of the comet. The scale is 14 m/pixel and the 1024 x 1024 pixel frame measures 14.5 km across.
But bringing up the contrast, as we have in the image below, shows that the head is simply hidden in the shadow of the comet’s large lobe, its faint outline silhouetted against the brighter background activity.
The processing also better reveals the vertical shadow cast by the large lobe at its left edge in this orientation, and the beautiful interplay between shadows and activity in this portion of the image.
This processed version of the 12 May 2015 image reveals the comet’s small lobe in silhouette, along with details of the comet activity. Credits: ESA/Rosetta/NavCam – CC BY-SA IGO 3.0
Meanwhile on the surface of the comet, the elongate depression of Aten dominates the view, with the quasi-circular shaped feature immediately to the left part of Ash. Smooth Imhotep extends into the distance in the centre, with distinct jets of activity dancing along the horizon.
Discussion: 45 comments
I first found out that the whole small lobe can be in shadow from Mattias Malmer’s realtime Rosetta view simulation. The picture makes me wonder what the night-time sky from 67p must look like. Would one see stars and dust grains all mixed up? This was after all the navigation problem faced by Rosetta a couple of months back. Or would large parts of the sky be fuzzy like when we see the zodiacal light? If Philae wakes up (and we hope it happens now as hoped for), can it answer this kind of question?
“Would one see stars and dust grains all mixed up? This was after all the navigation problem faced by Rosetta a couple of months back.”
Sorry, Kamal, but I find it hard to believe that Rosetta’s highly sophisticated, fully programmed, navigation systems could, during the “14 February event” (https://blogs.esa.int/rosetta/2015/02/27/cometwatch-the-challenges-of-a-close-flyby/), have suddenly got confused to the point of no longer being able to distinguish “fixed stars” (https://en.wikipedia.org/wiki/Fixed_stars) from nearby lumps of 67P, of thus-far indeterminate composition, whizzing past it at relatively high speed (a simple question of parallax, as learnt in high-school…!) Many centuries (if not millennia) ago, our sea-going ancestors could already navigate by distinguishing real stars from shooting stars (and without the need to know that the latter were actually only the manifestation of the sort of “dust grains” you refer to, albeit in a very different context…).
In modern times, the problems our sophisticated navigation systems have are most often due to electromagnetic interference (EMI): https://en.wikipedia.org/wiki/Electromagnetic_interference. I confidently predict that once all the data about the 14th February navigation anomaly is finally released into the public domain, EMI will also be revealed as having been its exclusive cause. And obvious, more far-ranging conclusions will follow…
Finally, I think we can safely assume that this is also why Rosetta has since been parked at a very safe distance from the comet and is very unlikely to attempt to get much closer again over the coming months: not for fear of confusing fixed stars with shooting stars but to be absolutely sure of staying way out of range of the induced electromagnetic properties of the “jets” which nearly caused its downfall on 14th February.
A shooting star is a at max once or twice a minute event. It causes one( and only one ) temporary disstortion between the ‘expected’ night sky and the observed night sky. So the difference between the night sky w/ and /wout shooting start is less than 0,01%. However, a sky full of dust specs where the dust can make up more than 50% is a whole different story. You cannot compare both situations.
Thomas: One of the reasons why artificial intelligence and robotics are difficult is that what seems like common sense to us is hard to build into programs. My guess is that the Rosetta navigation software cannot “automatically” distinguish between stars and meteors, and that it was not designed to do so, quite likely with good reasons.
AFAIK, Rosetta has two star trackers. If one of them fails, the other tries to identify the necessary stars. If both star trackers fail, an inertial unit takes control. This inertial unit (gyro) works for some time. As this time is exceeded, the system switches into safe mode to avoid damage by wrong pointing (risk of instruments pointing to the Sun).
During a previous flyby, both star trackers failed, but the inertial unit bridged the time until the star trackers recovered. This kind of information is usually transmitted as houskeeping data.
With this load of grains the star trackers worked very well:
https://blogs.esa.int/rosetta/files/2015/01/ESA_Rosetta_Rotundi-et-al_OSIRIS.jpg
But at some point the limit of false positive stars is reached.
Kamal, the last images from Philae give the impression that it is more or less stuck in an awkward place on the comet. I imagine that the Philae teams are pondering what to do if it wakes up — it might be possible for it to dislodge itself by some combination of movements of its appendages and/or reaction wheels. Or it could get itself stuck deeper.
The next comet-encounter mission needs to include a rover. One that can move carefully enough not to jump off the surface in that low gravity. Or one that moves about by screwing one leg after another into the “ground” … Its design will be an interesting engineering challenge.
A similar idea isn’t quite new, although the preferred solution is a hopper, e.g.::
https://solarsystem.nasa.gov/docs/p395.pdf
I should maybe have added, that the Philae scientists never expected to finally land at such an interesting location as Philae now is. It’s a probably dust-free surface, probably with direct access to the ices, and without the usual surface regolith layer.
The challenge is the low temperature and the short solar illumination, maybe also the difficult access to the interesting stuff.
Hopefully Philae awakes soon, and establishes communication with Rosetta, to be able to continue his more than interesting exploration.
Hi Gerald. Thanks a lot for sketching the math of spin energy budget 🙂
https://blogs.esa.int/rosetta/2015/04/15/cometwatch-12-april/#comment-445141
Welcome!
Hi Gerald,
Same here. The calculations show how small the kinetic energy is in comparison to the outgassing, or insolation energy especially in terms of per mass or surface area.
Thus there is no *need* to postulate frictional energy sources unless there is other evidence that would need friction, like stretching or other deformation of the nucleus that would involve momentum transfers.
I would like to postulate the possibility of thermal exchange, whereby a temporary liquid solidifies due to cold temperatures, but warms up a nearby substance, which may have a different latent heat and might melt moving with gravity towards the interior of the comet. This could *convect* solar thermal energy to the interior of the comet..
I wouldn’t rule out a similar mechanism with sublimation and resublimation, considering the porosity of the nucleus.
Most liquids need some pressure to form. So a sublimation/resublimation mechanism would need less additional assumptions.
A thin impermeable surface crust with subsurface ices in a porous matrix would be sufficient. Subsurface ice would sublimate, and resublimate in a deeper layer, releasing heat to potentially sublimate supervolatiles.
The gas phase can diffuse or just flow through porous material. A pressure gradient would build up due to volume loss by resublimation in deeper layers. Diffusion would work to some degree even with supervolatiles migrating in the opposite direction (i.e. towards the surface), since partial pressure determines the direction of diffusion.
For liquids capillary forces would be much stronger than gravity, provided the pores are sufficiently small. But a substantial pressure would be needed, about 6 hPa for water, at least. CO2 needs a much higher pressure to become a liquid.
The point is, that the pressure gradient would tend to point outwards towards the vacuum of space sucking on any slight differential going the other way. This would limit the potential thermal transfer inwards with gases.
If there was an initial abundance of Ethane, a low density substance that has a very low pressure triple point, it would tend to melt, and both gravity and capillary action would conspire for it to flow deep below the surface, then be insulated from both the heat and the cold. Little Ethane would show up as outgassing compared to the volatiles you mention, even if there were huge subsurface Ethane lakes.
If – and that’s a big *IF* – liquids do exist internally, another effect will very probably come into play if there is a ‘free surface’..
There is a phenomenon known as Marangoni flow which is rarely important in ‘high’ (ie one) G environments, but can be dominant under low G conditions.
https://en.wikipedia.org/wiki/Marangoni_effect
Its caused by any temperature gradient causing a surface tension gradient which induces a flow. A related effect is ‘tears of wine’, the streaks you see on the side of a wine glass, where the surface tension gradient is caused by ethanol concentration gradients.
Fluid handling in spacecraft can be very strongly influenced, dominated, by Marangoni flow.
*IF* (BIG *IF*) 67P has internal fluids, temperature & composition gradients causing Marangoni flow is likely to be important. It does require a ‘free surface’ to operate; if it were a filled porous material, the effect is absent.
On the currently thin basis of knowledge of the interior of 67P/C-G that’s at least not obviously too far off the road, since abundant ethane has been detected in Hyakutake: https://www.sciencemag.org/content/272/5266/1310
The ethane partial pressure at 92 K would just be near 1.2 Pa, with ethane just above its freezing point.
But assumed, this scenario would withstand reality, I see just a possible one-way flow or diffusion of ethane to the interior, meaning the melting heat can be transported only once. How does convection come into play, as an efficient process to transport solar energy to the interior?
… physical properties of ethane taken from here:
https://encyclopedia.airliquide.com/Encyclopedia.asp?GasID=28
Once you assume the possibility of liquid lakes in the interior, it opens up a wide avenue of potential movements. Liquids would be more affected by tidal forces than would solids, especially now, at Equinox. Liquids close to the neck would migrate away from the centre of rotation towards the nearest lobe. The conservation of angular momentum would then slow the rotation of the comet nucleus. Capillary action would come into play which in some cases transfer liquids closer to warming surfaces. Organic nano sized particles would be transported in suspension. Water ice particles would float up. A low temperature crude oil like substance could permeate much of the interior. Lots of interesting things happen that would challenge core assumptions of comets, such as being pristine, or solid aggregates. We have to open our minds up to more complex possibilities.
Hi Marco,
that’s very creative, but no, that isn’t realistic.
Any liquid ethane, IF it exists, would form thin films over grains, or fill some pores.
Accumulation of free-floating lquid ethane may exist somewhere, e.g. on Saturn’s moon Titan, but not inside 67P.
Even in the presence of liquid ethane, the cometary material would feel dry, since the liquid would adhere to the matrix inside pores. Due to the presence of other organics, it may solve them and form some viscid solution or even freeze, or it would polymerize by loss of hydrogen.
There is no reasonable differentiation mechanism to overcome entropy.
Forget gravity or those minute tidal forces; those are far too tiny in comparison to molecular forces (van-der-Waals).
Hi Gerald, If there is a lot of porous ice, and I am assuming Ethane is hydrophobic, why would it stick to the pores? There’s no saying what ethane or any other hydrocarbons would do, but in their liquid state they are not likely to stay static with vapour pressure gradients, temperature gradients, and as you say, Van de Waal forces, etc. happening on a dynamic rotating body. The truth is we don’t know what is happening inside the comet because we can’t get in there, but just saying it has to be cold solid porous volatiles, because that makes for a simpler model is ludicrous.
Hi Marco,
the dust observed by GIADA is highly prous, almost fractal. So in some sense we can take a close look to at least some of the material the comet is made of.
Assuming a fundamentally different interior of the comet, regarding this non-volatile fraction, lacks any basis thus far.
About hydrophilic and hydrophobic liquids: Put a drop of olive oil onto a piece of paper; it will be adsorbed like a drop of water. It doesn’t form a lake inside the paper.
By heating it may move, but still no lake. Not even the 100,000-fold stronger gravity can do it. You would need kind of a distillation. That’s too much of unbased assumptions for me. And even then, the newly formed liquids would be adsorbed again by the porous substrate.
… meaning the 100,000-fold stronger gravity on Earth in comparison to the comet.
The dynamical forces on the same small order.
Tidal forces on the comet are another several orders of magnitude lower.
Assuming anything else than a cold brittle porous mix of dust and ices, maybe enclosed meteoritic pebbles, in the deep interior of the comet is pure speculation without any hint.
Lack of evidence doesn’t prove existence.
Hi Gerald,
While I agree that a mix of cold porous volatiles and dust and pebbles is the simplest solution that doesn’t immediately contradict observation, that does not make it the most likely, necessarily.
I look at the axiom of comets being “pristine” unprocessed remnants with the same skepticism as any unconfirmed hypothesis. The hypothesis that comets undergo significant solar processing once they reach even the outer planets of our system is well worth considering. The processing hypothesised is that a thick surface layer gets sintered and becomes volatile depleted and hydrocarbon rich. This surface layer, several 10s of metres thick begins to hold back and filter any internal pressure. Hydrocarbon rich layer is quite conductive and volatiles sublimate/evaporate from under this layer breaking through to the surface in geysers. The evaporation from the bottom of this crust cools down the underside of the crust setting up a steeper temperature gradient between the insulated surface and the interior, which helps the process of conduction into the interior. The interior gas pressure would travel in all directions including internally through the porous substrate, which would transport heat energy to other parts of the interior.
Of course, this hypothesis is not proven either, but it is worth looking at the differential evidence to see if it has merit..
That should be insolated surface, which is conductive, not insulative.
The “axiom” of comets being “prestine” is certainly not quite free of some wishful thinking. But comets are relatively prestine in comparison to planets or iron-nickel meteorites. So odds to study the early solar system are better for comets than for many other bodies.
But comets also have undergone some change, at least decay of short-lived isotopes, but also collisions and fragmentation.
From the high porosity, however, it’s rather clear, that they’ve never been part of a body of say a few 100 km or more in diameter, might be except some pebbles/meteorites from later impacts. And its clear, that they couldn’t have lived in the inner solar system for very long due to their high volatile abundance in low gravity.
A crust of 10s of meters thickness made of organics isolates thermally for years. Organic compounds are good thermal insulaters, in general, particularly if porous.
This constrains your scenarios of possible histories of the comet.
The other points in your latest post sound possible at first glance.
High Gerald,
while I am not to keen on the models so far presented for sublimation (that does not mean I discount it happening)
However in line with your model ”sublimation. re sublimation” is there room for a sort of water cycle particularly in the neck region.
If the dust is thrown in the air (does not matter which method) then is it possible that the dust that ends up in the shadow could have ice freeze on to it to form an icy coat. Then if the ice coated dust eventually lands back in the neck area and warms up again, then it should be free to sublimate. Its only a thought.
regards
Dave, that’s very close to what the VIRTIS team has been suspecting: Formation of surface ice (frost?) in the shadowed neck region, which sublimates shortly after it’s illuminated again by the Sun due to the rotation of the comet.
This seams to lead to a narrow seam of insolated surface ice along retracting shadows. The shadowed areas themselves, where most of the ice should be found, are hard to measure.
https://www.hou.usra.edu/meetings/lpsc2015/pdf/2021.pdf
These measurements are difficult, see e.g. this earlier paper:
https://arxiv.org/pdf/1503.08172.pdf
So I’d wait for further confirmation to be sure.
It’s also possible, that the surface ice is formed from subsurface sublimation of water ice while the subsurface is sufficiently warm after the previous insolation.
Saying farewell. Wishing the best. Been great people, living science 🙂
Logan, I can’t believe you’re gone…….
So, I personally like to analyse what we know, not make ‘confident predictions’ based on very little.
We know it lost its ‘star fix’; we know from a video they released that at one moment at least it was ‘jittering’ without a solid lock. We have two hypotheses, it’s induced by illuminated dust in the star tracker field of view, or it’s an electrical, EMC issue.
But first, how likely are we to get extra data? Don’t know, but maybe not that likely. The star tracker will for sure, as a minimum, provide a ‘data valid’ to the spacecraft; one bit. At the other end of the scale, lots of data about which stars it matched etc etc. In the middle, maybe accuracy estimates. How much of it is communicated to Rosetta, and how much is downloadable, we don’t know; but it might not be much; conspiracy theories loom.
We are told, without any reference to support it other than a general link on EMC, that most navigation system problems are EMC. Now that might even be true – but most navigation systems don’t depend utterly on optical data acquired through a cloud of dust; they depend on RF signals (GPS, VOR, DME, ILS, radar, …..) which are directly vulnerable to EMC issues; a star tracker has no RF input.
So let’s examine our two hypotheses.
Firstly, it’s optical. A star tracker depends on acquiring images of stars and matching them to a stored pattern. If you are looking through illuminated dust, it’s rather obvious this can confuse things. Both directly, dust particles looking like stars, but also because they may be very bright, and the PSF (point spread function) may well obscure nearly stars in the glare. Then close by dust will be out of focus, causing even bigger PSFs. They will also be moving. Remember this thing is twenty year old technology, (think of your phone, PC 15-20 years ago) designed for an unknown environment, and the star tracker is very probably largely an off-the-shelf design. It’s not as they say rocket science to see this as a very likely cause of trouble. (Incidentally grains of dust on the lens will not ‘obscure’ stars as has been suggested at times here, because they are totally de focused.) The video showing pointing jitter seems very consistent with a system which is basically working but confused by ‘false stars’.
Secondly, it’s an EMC upset.
Now spacecraft systems absolutely routinely have to concern themselves with EMC. They must not cause problems for the other systems, for the communications links, for the scientific instruments. They must survive being close to the spacecraft’s transmitters, and survive a possible lightening strike to the launch vehicle on he pad. Such systems are designed and extensively tested to deal with EMC issues. Why would *only* the star tracker fail? If the EMC environment was so severe it ‘crashed’ a device *with no signal input other than optical*, isn’t it kind of surprising other units didn’t suffer, especially those with antennas, booms etc. The CPU which correctly ‘safed’ Rosetta would have probably been more vulnerable than the tracker, many connections. Then of course, though we don’t know which were on, any instrument collecting ‘electrical data’ (plasma instruments, magnetic field etc) should be screaming it’s head off in such a violent location. One might also expect a lot of RF noise showing up as comms link issues, potentially even detectable at earth. And if it’s that violent, Rosetta has often been km, tens of km, from there, and the ‘signature’ should have been extremely dramatic; but has not been reported, An EMC ‘crash’ would be very likely to just totally take it out, not give pointing jitter, which we were shown.
So
– we might get more data, but maybe not.
– optical is straightforward and absolutely obvious as an explanation
– Yes, if violent enough I’m sure you could crash a start tracker with EMC. It just doesn’t fit just about anything we know about the incident.
Make your own deductions about the most probable explanation.
Emily, what has happened to this blog ? All of a sudden no new comments since the 20th May, 6 days ago, with many previously published comments disappearing . Infrequent updates from your team, none on any issues of fundamental importance. Is there a new policy in operation. Some enlightenment would be appreciated.
Hi originalJohn,
This problem has also been reported on another thread and as I posted there we have not been deleting any comments that are already approved. There is nothing obvious going on in the back-end of the blog but I’ll raise it again with our technical gurus.
We’re providing updates and images when we have them, as normal, so there is no new policy in place.
Apologies for any frustrations resulting from the disappearing comments.
Emily
Hi Emily,
To characterise the glitch, I copy my comment to my clipboard before verifying and submitting. About half the time, the comment disappears rather than going into the queue for moderation. But here is the kicker – if I repost the precise comment using paste from clipboard, and re submit, the comment is detected as a duplicate. If I add a dot at the end before resubmitting, it gets a chance at being properly put in the queue. It appears the disappeared comments are logged as submitted before they disappear.
Hi Emily, it’s been fantastic having the blog. However,I’ve noticed a similar shift in the blog the last few weeks. Am wondering if the blog is unfortunately just sort of going stale, accompanied by a general loss of interest. Some fresh content based on new scientific findings would of course help, or posts summarizing scientific papers written to date, what the teams are doing, etc etc. Just seems to be a general dearth of up to date, interesting, relevant information to liven things up and hold interest.
Hi, and thanks for the comment. We haven’t experienced a loss of interest, and we continue to report on results as we have them. As you already know, the scientific peer review process takes time, but more papers are becoming available as the months go on. Hope you continue to enjoy following the blog as we head towards perihelion.
Emily
Sorry can’t see where to post this post appropriately.
I think it unlikely convection would play a role. It’s one of those effects that starts when a dimensionless number (Grashof?? – can’t check, using an iPhone while travelling) exceeds a critical value (like Reynold’s number and turbulence.)
Now gravity will certainly come into that number ‘on the top line’ to some power, and of course it’s very low, even lower in the interior. So I’m dubious you’d ever get significant convection, but someone with better web access needs to look it up and plug numbers in.
Also
“How does convection come into play, as an efficient process to transport solar energy to the interior?”
Surely convection won’t do that anyway. If the interior is cold and you heat the outside, that’s convection stable, even if the critical xxxxx number is exceeded.
You have to hear the bottom of the saucepan, not the top surface, to start convection.
Ah whilst Gradhof is relevant, and so is Prandtl, I think the Rayleigh number is the important one. G is on the top line, unit power. Someone go figure.
Hi Gerald,
You say
“A crust of 10s of meters thickness made of organics isolates thermally for years. Organic compounds are good thermal insulaters, in general, particularly if porous.”
I thought so too, but a chemist that I have been talking to, thinks that the sorts of carbon compounds that would be formed due to the conditions in space, would be long chain and PAH’s etc. and not porous at the surface after processing with solar radiation etc. which I am told makes for a thermal conductor possibly better than metal. For instance, graphene, a special infinite form of PAH, has one of the highest conductivities known.
Hi Marco, there are a few such conducting carbon-bearing chemicals. Diamond is a very good thermal conductor.
There is some hype about graphene among chemists.
But either amorphous compounds or some polymers (“plastic”) are more likely.
Take fullerenes instead of graphene, and you get very good thermal insulators due to the random alignment.
Probably the chemistry is very heterogenious (look at the preliminary Ptolemy spectrum) on the nucleus, such that those highly ordered organic crystals needed for high conductivity appear rather unlikely.
Then, Harvey, Convection could occur from the bottom of the conductive PAH crust.
Marco, those PAHs usually are no good conductors. Take bitumen, plastic or soot as related types of chemical compounds.
Those (in comparison) few man-designed exceptions are too unlikely to form naturally outside the laboratory in the needed ordered way.
You may ask the chemist, how likely, he thinks, is the formation of those conductive organics in some random chemical experiment, not whether it’s possible under some very favorable assumptions.
Marco it really doesn’t matter what you make it of.
If the power input is ‘at the top’ (ie, at the lower gravitational potential’, it will not be prone to start convection. That is like heating the top surface of the water in the pan instead of the bottom. You form the less dense hot water on top -and it stays there.
But convection is driven by bouyancy, and hence involves gravity. The gravity on the comet is extremely low. It’s resisted by viscosity, and that is expressed via the Grashof /Rayleigh numbers. In such a low G environment I think it’s very unlikely convection would occur – but do the sums, it’s in Wiki.
(PAHs wouldn’t surprise me at all incidentally. But most will be solid at comet temperatures I think.)
Btw graphene is a very special case. (I’ve worked with it.) it does indeed have extraordinary thermal and electrical conductivity *in the plane of the one atom thick sheet’*. That is a very, very different matter to a bulk PAH; I don’t think it at all likely they will have very high thermal conductivities.
But in fact a *high* thermal conductivity would reduce the likelihood of convection, because it reduces temperature gradients for the same power flow.
Initiating convection requires an adequate temperature difference to drive the bouyancy.
I think Maragoni flow would be much more likely if we have a liquid in a wetted porous structure; it’s known to be important in low G.