How comets are born

This story is mirrored from the ESA Web Portal.

Comet 67P/C-G on 22 March 2015. Credits: ESA/Rosetta/NAVCAM – CC BY-SA IGO 3.0.

Comet 67P/C-G on 22 March 2015. Credits: ESA/Rosetta/NAVCAM – CC BY-SA IGO 3.0.

Detailed analysis of data collected by Rosetta show that comets are the ancient leftovers of early Solar System formation, and not younger fragments resulting from subsequent collisions between other, larger bodies.

Understanding how and when objects like Comet 67P/Churyumov–Gerasimenko took shape is of utmost importance in determining how exactly they can be used to interpret the formation and early evolution of our Solar System.

A new study addressing this question led by Björn Davidsson of the Jet Propulsion Laboratory, California Institute of Technology in Pasadena (USA), has been published in Astronomy & Astrophysics.

If comets are primordial, then they could help reveal the properties of the solar nebula from which the Sun, planets and small bodies condensed 4.6 billion years ago, and the processes that transformed our planetary system into the architecture we see today.

The alternative hypothesis is that they are younger fragments resulting from collisions between older ‘parent’ bodies such as icy trans-Neptunian objects (TNOs). They would then provide insight into the interior of such larger bodies, the collisions that disrupted them, and the process of building new bodies from the remains of older ones.

“Either way, comets have been witness to important Solar System evolution events, and this is why we have made these detailed measurements with Rosetta – along with observations of other comets – to find out which scenario is more likely,” says Matt Taylor, ESA’s Rosetta project scientist.

During its two-year sojourn at Comet 67P/Churyumov–Gerasimenko, Rosetta has revealed a picture of the comet as a low-density, high-porosity, double-lobed body with extensive layering, suggesting that the lobes accumulated material over time before they merged.


Evidence that Comet 67P/Churyumov–Gerasimenko is composed of ancient material preserved from the formation of the early Solar System and that came together under low speed. The evidence collected by Rosetta lies in the comet’s structural properties, the gases detected leaving the nucleus, and observations of surface features. Credits: Centre: ESA/Rosetta/NavCam – CC BY-SA IGO 3.0; Insets: ESA/Rosetta/MPS for OSIRIS Team MPS/UPD/LAM/IAA/SSO/INTA/UPM/DASP/IDA; ESA/Rosetta/MPS for COSIMA Team MPS/CSNSM/UNIBW/TUORLA/IWF/IAS/ESA/BUW/MPE/LPC2E/LCM/FMI/UTU/LISA/UOFC/vH&S

The unusually high porosity of the interior of the nucleus provides the first indication that this growth cannot have been via violent collisions, as these would have compacted the fragile material. Structures and features on different size scales observed by Rosetta’s cameras provide further information on how this growth may have taken place.

Earlier work showed that the head and body were originally separate objects, but the collision that merged them must have been at low speed in order not to destroy both of them. The fact that both parts have similar layering also tells us that they must have undergone similar evolutionary histories and that survival rates against catastrophic collision must have been high for a significant period of time.

Merging events may also have happened on smaller scales. For example, three spherical ‘caps’ have been identified in the Bastet region on the small comet lobe, and suggestions are that they are remnants of smaller cometesimals that are still partially preserved today.

At even smaller scales of just a few metres across, there are the so-called ‘goosebumps’ and ‘clod’ features, rough textures observed in numerous pits and exposed cliff walls in various locations on the comet.

While it is possible that this morphology might arise from fracturing alone, it is actually thought to represent an intrinsic ‘lumpiness’ of the comet’s constituents. That is, these ‘goosebumps’ could be showing the typical size of the smallest cometesimals that accumulated and merged to build up the comet, made visible again today through erosion due to sunlight.

According to theory, the speeds at which cometesimals collide and merge change during the growth process, with a peak when the lumps have sizes of a few metres. For this reason, metre-sized structures are expected to be the most compact and resilient, and it is particularly interesting that the comet material appears lumpy on that particular size scale.

Further lines of evidence include spectral analysis of the comet’s composition showing that the surface has experienced little or no in situ alteration by liquid water, and analysis of the gases ejected from sublimating ices buried deeper within the surface, which finds the comet to be rich in supervolatiles such as carbon monoxide, oxygen, nitrogen and argon.


The two main scenarios for the formation of comets: collisional (left) vs. primordial (right) rubble pile. Evidence collected by Rosetta strongly favours the primordial rubble pile hypothesis, namely that comets were built up slowly through low-speed accumulation of material into the shapes observed today. Credit: ESA

These observations imply that comets formed in extremely cold conditions and did not experience significant thermal processing during most of their lifetimes. Instead, to explain the low temperatures, survival of certain ices and retention of supervolatiles, they must have accumulated slowly over a significant time period.

“While larger TNOs in the outer reaches of the Solar System appear to have been heated by short-lived radioactive substances, comets don’t seem to show similar signs of thermal processing. We had to resolve this paradox by taking a detailed look at the time line of our current Solar System models, and consider new ideas,” says Björn.

Björn and colleagues propose that the larger members of the TNO population formed rapidly within the first one million years of the solar nebula, aided by turbulent gas streams that rapidly accelerated their growth to sizes of up to 400 km.

Around three million years into the Solar System’s history, gas had disappeared from the solar nebula, only leaving solid material behind. Then, over a much longer period of around 400 million years, the already massive TNOs slowly accreted further material and underwent compaction into layers, their ices melting and refreezing, for example. Some TNOs even grew into Pluto or Triton-sized objects.

Comets took a different path. After the rapid initial growth phase of the TNOs, leftover grains and ‘pebbles’ of icy material in the cold, outer parts of the solar nebula started to come together at low velocity, yielding comets roughly 5 km in size by the time gas has disappeared from the solar nebula. The low speeds at which the material accumulated led to objects with fragile nuclei with high porosity and low density.

This slow growth also allowed comets to preserve some of the oldest, volatile-rich material from the solar nebula, since they were able to release the energy generated by radioactive decay inside them without heating up too much.

The larger TNOs played a further role in the evolution of comets. By ‘stirring’ the cometary orbits, additional material was accreted at somewhat higher speed over the next 25 million years, forming the outer layers of comets. The stirring also made it possible for the few kilometre-sized objects in size to bump gently into each other, leading to the bi-lobed nature of some observed comets.

“Comets do not appear to display the characteristics expected for collisional rubble piles, which result from the smash-up of large objects like TNOs. Rather, we think they grew gently in the shadow of the TNOs, surviving essentially undamaged for 4.6 billion years,” concludes Björn.

“Our new model explains what we see in Rosetta’s detailed observations of its comet, and what had been hinted at by previous comet flyby missions.”

“Comets really are the treasure-troves of the Solar System,” adds Matt.

“They give us unparalleled insight into the processes that were important in the planetary construction yard at these early times and how they relate to the Solar System architecture that we see today.”




  • A. Cooper says:

    As I said ten months ago regarding the Massironi et al. contact binary paper, this is so so wrong:

    "The stirring also made it possible for the few kilometre-sized objects in size to bump gently into each other, leading to the bi-lobed nature of some observed comets."

    Scientists are now rewriting the theories of solar system evolution based on a misinterpretation of the comet's bilobed shape. It could hardly be a more serious situation, with implications for understanding the formation of the Earth and, as the mission scientists have stated, life on Earth.

    As I said in my reply to Dr. Massironi's follow-up post that rejected stretch theory, have you cut the neck out to see how neatly the head lobe fits to the body? That question was also posed ten months ago:

    The linked comment also addresses the other erroneous paradigms that are still being used ten months later, such as the specific kinetic energy anomaly between the so-called 'bicycle speed' collision model and 67P's supposed collision of two bodies. The modelled version is impeccable but has no applicability to 67P. The specific KE is 8 to 10 times higher for the slowest possible collision of the 67P lobes, implying high energy dissipation and significant reworking. And yet the model is still being relied upon heavily in the new work:

    "Earlier work showed that the head and body were originally separate objects, but the collision that merged them must have been at low speed in order not to destroy both of them."

    The link is to the Massironi et al. contact binary paper that relies so heavily on the 'bicycle speed' paper. At the time, logan made a point regarding the actual modelled speed of 0.375 metres per second: "baby crawling speed". 0.375 m/sec collision speed is a physical impossibility for 67P's mass and consequent gravitational parameter.

    Marco Parigi was admonished by Harvey nearly a year ago on his certainty that 67P had stretched. However, when you really have seen matching features, mirrored and translational, along with exquisite symmetries shared between head lobe and body lobe, all complementing each other via the strict constraints of the tensile force vectors brought about by stretch then yes, you can be certain.

    It definitely stretched and it would be remiss of me not to state it so unequivocally.

    Stretch theory has explained the different morphologies on every one of the named regions on the comet and done so with complete internal consistency, each region's morphology being consistent with its neighbours and the stretch force vectors running through them.

    The irony is that the clues to stretch are being observed and commented on in every paper that furthers the paradigm that it's a contact binary. It's only because stretch was never considered at the outset, as were contact binary and erosion models, that these observations aren't interpreted correctly. It's because stretch isn't even in the back of the minds of the scientists whilst CB theory and erosion always were. The observations are simply shoe-horned into the most appropriate model. This quote from the article is a case in point. It's a huge clue for stretch:

    "The fact that both parts have similar layering also tells us that they must have undergone similar evolutionary histories and that survival rates against catastrophic collision must have been high for a significant period of time."

    As Marco said simply and clearly in conclusion to his comment nailing his colours to the mast:

    "It stretched".

    • A. Cooper says:

      In my comment above, I wrote

      "The link is to the Massironi et al. contact binary paper that relies so heavily on the 'bicycle speed' paper."

      This was referring to the words "Earlier work" in the previous paragraph which was hypertexted as a link in the main article. I had copied and pasted it. It remained highlighted blue in the comment box but not in my comment. So it wasn't referring to the link further above, which was to my former comment.

      I've now read about half the Davidsson et al. paper, hopefully all the dynamical evolution parts. Whilst there is a reference to Massironi et. al and so an indirect reference to the "bicycle speed" collision model paper, that paper isn't explicitly mentioned. Instead, the modelled speeds are much higher: 40 m/sec or more, which would enhance the "reworking" I mentioned above somewhat to say the least.

      This paragraph from page 21 suggests the compacted outer layers may act as a defensive shield to allow it to maintain its shape when the two lobes collide:

      "An important aspect of the formation of these nuclei is that the relative velocities no longer are size dependent. In the solar nebula the gas drag prevented collisions among equal–sized cometesimals. The mild viscous stirring in the early gas-free primordial disk removes this limitation. Thus, the last step in comet formation could very well be the merger of two cometesimals of similar size. A large cometesimal with compacted outer layers may behave differently during a collision than a small unlayered one. The latter may break upon impact into a “talps” but the hardened surface of a layered cometesimal may protect it during collisions, allowing it to maintain its shape. We thus suggest that numerous bi-lobed nuclei formed during the first few ten Myr after solar nebula gas dispersal, and that the lobes are individually layered, due to viscous stirring by TNOs."
      But at 11,378* times the specific kinetic energy of Massironi et al's cited collision model is that really plausible?

      If the relative speeds are substantially lower, as some little-stirred, low inclination and eccentricity (i and e) orbits would allow, it means a much smaller chance of a collision as I've always maintained. I'm glad the Davidsson et al. paper mentions this relationship between stirring (orbit perturbation) leading to i and e changes thence to relative speeds and therefore the chances of a collision. As for the 1 metre per second, minimum possible speed for 67P, or even the bicycle speeds (which weren't actually implied by the collision model paper anyway) the orbits of the two would-be lobes would have to have virtually identical i and e. They would also need to have virtually identical a (semimajor axis) and w (argument of perihelion). So they'd have to be on essentially the same orbit and at exactly the same place in that orbit. Not gradually drifting together because that would necessitate their being on different orbits in the first place.

      Perhaps there is a tiny window of drift-together wiggle room: the baby crawling speed of 0.375 m/sec or similar. That would be the radiant speed outside the 67P gravity well that then accelerates it to ~1m/sec. This radiant speed implies a crossover angle of around 0.005° at these heliocentric radii and ~4-6 km/sec absolute speeds. A crossover of 0.005° is chance that's as good as zero.

      The Davidsson paper is entertaining far more realistic i and e variations which are the two variables (along with semimajor axis and argument of perihelion) that govern the relative speed or radiant approach vector. This higher variation gives the higher orbit crossover angles (bigger than 0.005° but still small) that increase the radiant approach vector. That's why the relative speeds are 40 m/sec, not 0.375 m/sec or 1 m/sec. So if "to bump gently" means at 40 m/sec and not 1 m/sec then I apologise for assuming the bicycle speed narrative was being invoked. The article cited the Massironi paper and reiterated the usual "low speed" needed in order not to destroy both lobes.

      40 m/sec is indeed slow by radiant speed standards and one really could say it was a bump, comparatively speaking, but 40 m/sec is a really mighty bump when trying to preserve the morphological features on a 70-80% porous comet. Even with the increased crust density suggested in the paper. I fear 67P wouldn't look quite so prettily layered as it does today after such a precipitate meeting of the lobes. And since layering is referred to in the paper as being the result of the accretion scenario in the first 25Myr epoch of the primordial disc, it can't relayer itself to repair the damage**.

      So contact binary theory is still struggling with the dynamical scenario of the 'fairy tale kiss' of the lobes that would allows 67P to keep its layers intact.

      *Bicycle speed collision model impact speed: 0.375 m/sec

      Collision speed 'u' used in the Davidsson paper: 40 m/sec or more

      Ratio of the two speeds: 106.66

      Ratio of the speeds squared: 11,378 (because kinetic energy depends on the square of the speed).

      Thus, 40 m/sec packs 11,378 times more energy into the collision than the actual modelled scenario in the collision model paper. By the same calculation, 40 m/sec puts 1600 times more energy into the collision than the minimum possible impact speed of 1 m/sec.

      **Marco has different ideas for the evolution of the layers. I agree with him. His scenario might imply that two collided lobes could actually resurface themselves. But they wouldn't do so in the manner we see the layers exhibited today as they are with no smooth layering rounding the head rim and extending down Hathor. It could only exhibit its present-day layering by evolving the layers as a single body, then stretching, then shearing. That's why Hathor looks cleaved, the head rim so sharply defined and parts of the rim flared out like a bell (because the head herniated from the body). Re the double-star footnote, I was referring above to standard theory for layering, not stretch-inspired theory.

    • Marco Parigi says:

      From my perspective, there is a lot to be said for studying actual and presumed ongoing changes, being that living with a comet through perihelion has given the opportunity to predict based on alternate theories.

      Armed with what we know about the corollary to the evidence of stretch - that matches should not be eroding their matches away, and that relative movement , rather than erosion would be evident - we have found areas of change from pre to post perihelion of just that nature. Changes on the neck, intriguingly lean towards evidence of ongoing stretch.

      I insist that as incumbent as they are, primordial theories are entirely unfalsifiable. As a case in point no physics model has come up with a solution of how molecular oxygen can get trapped into a clathrate in any modellable accretion scenario.

      Another intriguing thing is that the outgassing appears to happen from the interior rather than directly from the surface. If it is found that the overall volume of the comet reduces far less than the weight of the lost volatiles and dust, that the comet can substantially reduce its density over time. This could give a plausible solution to how a fragment of a larger TNO object could end up with such low density. This does not seem to be offered as a plausible solution of the various paradoxes.

      • logan says:

        " physics model has come up with a solution of how molecular oxygen can get trapped into a clathrate in any modellable accretion scenario."

        Maybe not at proto-disks, but at Mother Cloud. More of a 'coagulation' than accretion. ElectroSTATICS could become dynamically relevant at the sempiternal silence inside the heart of Mother Clouds. Also other VdW forces.

        [This is cometary fiction]. We need models, even if non-factual, in order to design future experiments.

        • Marco Parigi says:

          We should not believe in oxygen trapping clathrates by default. The question should remain open until there is some actual evidence or model solutions.

      • logan says:

        .Marco, I have not given up on the stretch model, yet. But at my mind it require not from a very energetic event, at just the precise amount of energy transfer. Quite the contrary. It requires 'unheard of' very long term stability, both kinetic and thermal.

    • A. Cooper says:

      This is how stretch theory explains the spherical caps that are thought possibly to be splatted cometesimals:

      As I said above, the explanations of 67P's features, according to stretch theory, are complementary to the morphological evolution of their neighbouring regions because all features in all regions are shaped by the tensile forces of stretch. The spherical caps are no exception in that they're intimately related to Ma'at and southern Bastet flanking them and Aker, below them.

  • Bill Harris says:

    Great article.


  • logan says:

    OK. We saw apparent scars of low speed impact at the 'shoulders' of Ducky, on that magnificent OSIRIS first public 3D shot. Then the 'cocoon' of the big lobe. [I have seen no 'cocoon' at small one -which weights in on them being 2 different objects].

    There remain a few clues that keep my mind refractory. First among them is the many, many shots suggesting of aligned ice matrices -in between lobes.

    Only admissible scenario to my little brain for shared latices is that both lobes where the innards of a former bigger object, for a very very long time.

    Big enough as to homogenize inside temperature orbit wise. Small enough as to gravity not to ruin 'perspiring' porosity.

  • logan says:

    Do you realize that this impact in between lobes had to occur in a non-rotational kinetic frame? Admitting the plausibility, and amazed at it. As for 'cold' kinetics talk 🙂

  • ianw16 says:

    The different densities of the two lobes would also appear to suggest that they were originally two separate objects, rather than an object that "stretched" or came back together after a parting of the ways.
    Supposed matches between the two lobes don't convince me, given that the surface has been losing many meters of material on every perihelion pass.
    What it looks like now will almost certainly be a fair bit different than what it looked like 4.5 Ga.

    • Marco Parigi says:

      Hi Ianw16,
      I assume that you have been noting (or looking for yourself) for changes over perihelion

      It is the "metres of sublimation" which is supposed and not apparent from the analysis of images, in all the areas which have been found to match.

      Presumed erosion is not happening. Boulders are moving, overhangs collapsing and so forth - matches are staying matches and revealing the secrets of what comets are actually doing rather than presumed by the rigid narratives you mention.

      • ianw16 says:

        You do know who you are starting to sound like, yes?
        Anyone who doesn't agree with me is wrong. Anyone who doesn't agree with me is guilty of rigid thinking, etc?

        Sorry, I'll assess the claims when I see them in a peer reviewed journal. Or evidence published in the aforementioned shows them to be wrong.

        • Marco Parigi says:

          Hi Ianw16,

          I am not sure who you mean - THOMAS? Albert Einstein?

          I do thank you for critiquing my ideas. Certainly, I do believe ideas can be discussed outside of peer reviewed articles.

          Gerald and Harvey are uncharacteristically quiet on this matter, whereas they had plenty to say regarding EU woo..

          I am not sure why you keep mentioning the 4.5 Gy. It is quite irrelevant to the stretch narrative. The stretch event is very recent. Certainly less than 1 Ky, and possibly even just a few decades ago. The comet bears little resemblance to what it did pre-stretch, so what it did before that cannot be deduced based on the comet material being pristine, which all peer reviewed papers do.

          There is a definite paradox between matches proving stretch and the assumption that the comet nucleus ( below surface) is unchanged from 4.5 Gy ago.

          They cannot both be right, and scientists are ignoring the stretch evidence precisely to avoid this paradox. Essentially, like with continental drift 100 years ago, it will have to wait for mid ocean ridges, or in this case, ongoing stretch to convince skeptical scientists who are invested in the status quo.

          I am that confident tat this evidence will materialise within my lifetime at least, that I wil patiently point out evidence piece by piece in the meantime.

    • A. Cooper says:

      Hi Ianw16

      You said, "Supposed matches between the two lobes don't convince me given that the surface has been losing many meters of material on every perihelion pass."

      It sounds as though you're basing your scepticism of the apparent matches on indirect countervailing evidence (the loss of large amounts of material) rather than investigating the matches themselves. Otherwise, why invoke the material loss argument rather than go to the nub of the issue: that you've seen our suggested matches and find them unconvincing?

      So are you referring simply to the supposition of mine and Marco's, which you find unconvincing as a theory or have you seen the apparent matches yourself and find them unconvincing?

      • ianw16 says:

        I've obviously not looked as assiduously as yourself.; however, I just don't find that particular line of reasoning to be a good one, regarding what may have happened billions of years ago. Both lobes will have lost plenty of material in that time. Both lobes, for presumably a very long time, have been subject to the same forces. That is, the same thermal stresses, and the same stresses induced by the spin of the comet. That they have similar appearances is not, in my view, particularly surprising.
        And the different densities of the two lobes does appear to be further evidence of a collisional history, as does the possible evidence of remnant impactors on the large lobe.

        • ianw16 says:

          Typo above: should be "remnant impactors on the SMALL lobe."

          • A. Cooper says:

            Hi ianw16

            Thanks for that. I can honestly see how you might casually refer to your being unconvinced by the supposed matches without saying you hadn't looked at our specific examples. However, I made the point because I wouldn't want other readers to be put off looking at the purported matches for themselves, thinking that other commenters appear to have looked and found them unconvincing already.

            Regarding the "billions of years ago" timeline. The asymmetrical outgassing required for a spin-up to 2-3 hours rotation period for stretch can be reasonably expected to happen only in the inner solar system, i.e. from the point when 67P became a JFC. This immediately cuts the spin-up/stretch time window down from billions of years ago to a few hundred thousand at most and, not unreasonably, a possible one thousand years ago or less. Here's a paper modelling 23K clones of 32 known Centaurs over 3Myr (forward and backward integrations in time with the present in the middle)


            The half life of the cloned Centaurs averaged a mere 2.7Myr. That's the half life time during which they reside in the solar system as Centaurs and may or may not become short period comets (SPC) including Jupiter Family Comets (JFC's). After that they get ejected. So straight away, the timeline for spin-up, stretch and match resilience is reduced to the low millions of years.

            On page 9 it says 34% of the clones became short period comets during the forward integration and 34% in the backwards integration. Table 8 shows the number of times an Earth crosser (EC) or Mars crosser (MC) might be expected to flip in and out between that particular designation and orbiting way further out in its 'standby' Centuar status: ~13 times for EC and ~14 times for MC.

            Page 9 also cites a paper that estimates the residence time for SPC's as ~6000 years before being ejected to Centaur status again. They likely return for another ~6000 year sojourn after another few hundred thousand years as Centaurs.

            Another paper is cited on the same page as giving the fade time for a SPC as 12,000 years. It doesn't distinguish between JFC's and much longer period (<200yrs) Halley-type.

            The above averages of 6,000 years and 12,000 years are cited below as if being rigid, planned visits and fade times. That's just for brevity and convenience so as to show that the average scenario points to something plausible for spin-up and match preservation.

            The fade time for a JFC such as 67P would, by the above reckoning, be at most 12,000 years and probably significantly less because it spends much more of its orbit at around 2-2.5 AU (last 400 years on JPL ephemeris) and 1.24-2.5 AU since 1959. These lowish perihelion values are the region where outgassing is much more marked as we very well know from Rosetta obs. That presumably decreases fade time dramatically because 67P spends 10 times longer in this region than an SPC with an orbital period ten times greater (65 years) and the same perihelion value. In fact, more than 10 times longer due to the larger semimajor axis of a 65-year orbit increasing the speed at the same perihelion distance.

            So if the fade time is 12,000 years and residence time is 6000 years, 67P may possibly have had a chance at spin up one 6000-year residence window ago with a period of a few hundred thousand years in between before returning. That's a lot less than the 4.5Gyr.

            If the fade time is 6-8000 years, the chances are much greater that the only opportunity for spin-up was this time round, i.e. in the last 6000 years. Since 67P is still quite active and nowhere near completely faded one might push the spin-up window to being just the last few thousand years.

            This iterative reduction of the spin-up window, measured in years before present, is the suggested maximum possible time during which spin-up could have occurred. So if the maximum window is 6000 years, there's a 17% chance the spin-up could have occurred in the last 1000 years. That's in the order of 120 orbits for an average 2-2.5 AU perihelion (informed guess, not calculated).

            Since there was selection bias for a comet with a low perihelion and lowish aphelion (JFC not Halley-type) there was also a selection bias for a comet with a shorter fade time and thus a shorter spin-up window. So we're looking at the spin-up lottery winner in terms of the chances of 67P having any chance at all at spin-up in the recent past.

            Seeing as in my and Marco's observations of the matches pre and post perihelion show no discernible change at the 0.8-metre resolution* for a 1.24 AU perihelion, it would seem that the maximum erosion at the matches over 1000 years would be less than 96 metres (120 x 0.8). But that's for the current 1.24 AU perihelion. For an average 2-2.5 AU it would be substantially less. 20-30 metres of erosion at the matches would leave all but the smallest ones still discernible as matches.

            So we're not far off an explanation for the resilience of the matches over time. If spin-up was 400 years ago (a 6.6% chance by the above reckoning) it would quite easily explain their resilience.

            *The 0.8 metre res comparisons are only for a few areas. Other areas are at maybe 2 metre res due to earlier photos being of that resolution. However it's clear that the predicted vast amounts of material loss simply didn't happen and certainly not around the matches. The very much more modest amounts that did occur (some of which have been found by me and Marco and referred on to the Rosetta team) were in the flat areas which are irrelevant to the matches. The matches would be spent, refractory crust according to stretch theory anyway.

          • A. Cooper says:

            And here's my typo correction:

            "...1.24-2.5 AU since 1959." Should read

            "...1.24 AU since 1959."

            It's been steady at around that value since 1959.

    • A. Cooper says:


      Has there been a paper written on the different densities of the two lobes? I've been hearing references to this fact for many months now but no mention of a paper.

      The most coherent reference to the density disparity that I heard was that the head was 10% less dense. Just like the Massironi layers paper, this is completely consistent with stretch. Material was drawn out of both head and body to form the neck, especially at Anuket and Bastet, which is why they look the same. Since the cross sectional area of extrusion was the same for head and body, there was the potential for both lobes to share 50:50 in their contribution to the neck material. But since the body is 2.4 times the size of the head, the head's density would be diminished to a greater degree than the body's density, hence its being less dense today.

      Stretch requires no special pleading, tying ourselves up in philosophical knots or overactive imaginations. However, it does require imagination on the part of the CB theorists to check whether all these amazing findings actually point to stretch rather than discount it. I've just described how the density issue points to stretch. With the Massironi et al. paper, it was the very fact that the head layers were at an angle to the body layers that pointed to stretch being the reason. This is because the head herniated from the body and thus today's head rim is angled down with respect to body layers but also has a flared 'bell shape' rim round the back at Serqet where it sheared last. The greatest flare is almost in line with the equator or rotation plane and exactly straddles the Paleo rotation plane as identified in stretch theory (not accepted by the Rosetta scientists).

      The most 'egregious' example of the head layers being angled to the body is the flat 'white' wall of Serqet sticking up like a sore thumb just above the said bell rim. It too straddles the paleo plane. Its anomalous angle to the body layers is a clear signature of stretch because it's a section of lower onion layer, cut either side by shear forces induced by the differential tensile forces of stretch rounding the proto-head. This isolated blocky chunk was then able to flip up and push right through the layers above it in an premature herniation attempt. This isn't speculation. The evidence is presented in Part 29 on the stretch blog with copiously observed movements of the layers delaminating around it. There's a wealth of other corroborating evidence for this in parts 22-28.

      Time and again, findings are being shoe-horned into one favoured narrative instead of stopping for long enough to think on whether stretch would really be as simple as a clean break with all layers and densities remaining the same.

      Stretch is proving to be a highly nuanced process. Anyone talking of head-body matches nowadays is I'm afraid way behind the curve. They've hardly been mentioned for over a year and 20 stretch blog parts. We're at least four stages past the matches which are nevertheless continually corroborated by other aspects of stretch anyway.

  • logan says:

    Wow! To those Souls waiting for "The Papers", this is the First Version of the Full Saga 😀

    Open Document.

  • logan says:

    "The key feature of hierarchical growth is that particles of different size have different stopping times (the time it takes for a velocity difference with respect to the gas to be damped by gas drag). Larger particles (with longer stopping times) need more time to adjust to changes in the turbulent gas environment than smaller particles, thus non-zero relative velocities are introduced between small and large grains. This is a necessary condition for collisions, thus possibility of growth. Similar-sized particles
    have small relative velocities (because they respond similarly to changes in the gas flow properties). Therefore, another signature feature of hierarchical growth is that large particles grow by accreting much smaller particles and not similar-sized ones."

    The "spry-paint" model. Gravity irrelevant at this scale. Great!

    • logan says:

      No bouncing barrier. 'Snowballs' doesn't bounce on snowballs. The surface of the big one similar to the projectiles.

    • logan says:

      The same way Amazon Basin is made fertile by the dust of the Saharan Desert my thinking goes for the "spry-paint" model: It is at the ISM embedding our Mother Stellar Cluster that its action is made relevant. So primigenial cometary material tell the History not only of our Star, but of their full family.

      And the idea is not mine. but don't remember where I read it. [Highly doubt any idea is mine].

      • logan says:

        Nebulas could have their own 'profound currents' and 'dead vortices'. And those their turbulent surfaces.

      • logan says:

        Another image allegory is that of bird flocks. 'Brotherhoods' are eventually decanted not by chem or phys, but by similar aerodynamical behavior.

      • logan says:

        Earth Late Heavy Bombardment being a close encounter with one of those cometary 'flocks'.

  • logan says:

    "...The km-sized bodies contain 1–100 m cavities..."


  • logan says:

    "...(a D = 2 km body has 0.3–0.7 km substructure). These subunits could have lived and aged for ∼10 5 yr prior
    to accretion."

    And the 'Great Walls" being the filling between those "subunits"?

  • logan says:

    "...Porosity variation among layers is expected because low and high-velocity assembly epochs may have alternated during the considered 25 Myr. Cometesimals impacting at unusually low speed may have survived intact."

    And just 'splattered' at high speed gas epochs. There is a common vision among the minds!

  • logan says:

    "The word 'primordial' does not necessarily require the survival of interstellar material, and does not exclude presence of heavily processed grains from the inner solar system."

    Once reaching the 100 meter scale, Inside 'climatology' starts being more relevant than Outside 'climate'.

    Same way that our cells are born and extinguished several times along our life, generations of grain could be born and extinguished inside those cometesimals.

  • logan says:

    "...if “accretion in reverse” pushes the size distribution back toward..."

    Yes, all the chat is about 'tidal' environments 😀

  • logan says:

    "...on top of the porous cores formed previously in the so-
    lar nebula..."

    Thanks for that, been always my presumption.

  • gilbert says:

    We spend a lot of time considering features we have seen. What about features we have not seen.
    For example, as an object with very low density, if it collided with a small but high density object, and not just a glancing blow, I would expect the high-density object to pass through and leave a “bullet” hole.
    But we don’t see any bullet holes in our sponge – just sinkholes.
    Has anyone looked on the opposite side of the comet to see if there is a matching entry or exit sinkhole there as well ?

    • logan says:

      "...But we don’t see any bullet holes in our sponge – just sinkholes."

      Hi Gilbert. Have not stopped disturbing image formation. And one of the reasons I suggest not discarding our Proto-Disk being gravitated from already predigested cometary material.

      How many failed attempts at proto-disks? How much cometary material was coagulated at Nebula itself?

      Where the maps end in incógnito. Seas of monsters and cascades into the void.

      But Authors are

  • logan says:

    Congratullations to ESA, The ROSETTA Endeavor Teams and B. J. R. Davidsson, H. Sierks, C. Güttler, F. Marzari and allies on the publication of "The primordial nucleus of comet 67P/Churyumov-Gerasimenko".

    Born a Classic. An indispensable Lecture to any Soul interested about Comets. This is the one to those who were asking for answers.

    Nowadays, the Document is presented as an invitation to new conversations on divergent lines of thought 🙂

    The Document is conclusive on primordialism, for 67P's family of Comets.

    [The math segments are beyond most of Us average aficionados, but prose very Outreaching].

    Also congratulating Claudia and the Out-Reach Team on Excellency at Journalism.

  • Kamal says:

    Dear Andy and Marco,

    Your careful notings of alignments are being undermined by the numerous recent pictures of cracks, boulders and shorelines which suggest that what happens at perihelion is crumbling, boulder and dust fall. I haven't really kept track of this but every putative change between pre-perihelion and post-perihelion may perhaps be seen as caused in this manner. Given the density and porosity of the nucleus these now appear to be likely outcomes.

    We now know that "show me the sublimation" is not a fruitful way of asking the question. Unquestionably large jet events take place, sublimation is hypothesized as their cause. We have evidence of icy substances lurking deep inside cave-like features.

    The effect on the broad nucleus surface that we see in the Navcam/Osiris images is through a chain of subsidiary events. It seems plausible that this can provide an explanation for the crumbling, although I would like to see
    an explicit connection being made between a jet event and some crack or boulder fall or shoreline. I think ESA have the data and it will be demonstrated when their analysis is more conclusive.

    But now the need for stretch disappears. Could there be a stretch at perijove then? We just do not have any evidence.

    Now the question for you is what then causes the noted alignments? Logan has also remarked on the crystalline appearance in some images. IanW16 seems to be saying that, given the kind of changes we have seen, no immediate cause is apparent. Could it be, for example, that
    the alignments are caused over lots of orbits by the fact that similarly angled places (with respect to the Sun) on both the smaller and bigger lobes go through similar cycles of daily activity, and hence face similar kinds of crumbling?

    So what you are calling "stretch" may be the manifestation of a long term (thousands of years as you say) stable regime of solar activity at perihelion and not a phenomenon resting upon some kind of gravitational perturbation. In particular one should expect to find it on other comet nuclei with similar composition. For example, on other bilobed comets there should be alignments between the two lobes like what you have seen on 67p.

    • logan says:

      "...We now know that "show me the sublimation" is not a fruitful way of asking the question..."

      Its the present image at my mind as the surface of a very fine screened car's radiator. But there are lots of nozzles, also. Recently fractured material next to cliffs also increases surface. 'Airborne' particulate is also sublimating, etc.

      • Kamal says:

        Logan: True, volatile as well as non-volatile sublimation has to be thought of.

    • A. Cooper says:

      Hi Kamal

      Though the surface is truly roughed-up, comparing the pre perihelion pictures to post perihelion shows virtually no change in the areas you refer to as undergoing crumbling, boulder and dust fall. This isn't to say they've never experienced crumbling, boulder and dust fall. They clearly have, but it's all consistent with the wrenching and tearing of stretch. Not just the general mayhem that you might expect with stretch but specifically aligned cliffs, lines of boulders, and translational matches of layers and tears, which betray delaminated layers. These specific alignments are all consistent with the tensile force vectors of stretch along the long axis of the comet.

      As for the erosion that has been found, it's all been on the flat areas that have sunk where there are no cliffs and almost no boulders. Marco and I certainly don't deny that sublimation is going on. In fact, stretch theory points to the possibility that it's carrying on at a deeper level than the near-surface sublimation so far detected. The subsidence on the smooth areas is almost certainly due to sublimation although I wonder if some of it is collapse after several perihelion cycles or perhaps many cycles of very gentle sub-surface sublimation.

      The Groussin et al. 2015 rates of material 'loss' at Imhotep were orders of magnitude greater than what is achievable in the models. I think most of that apparent loss was actually collapse of an already highly depleted layer. By that reckoning, the many small holes left by sublimated ices collapsed like compressing a sponge. Hence the even depth of the subsidence.

      The three signs of erosion you describe, cracks, boulders and shore lines, are attributable to stretching of the single body comet before the head sheared. As I usually mention, for readers only interested in the peer-reviewed literature, stretch theory isn't mainstream and isn't supported by the Rosetta scientists.

      As I said in a recent comment, stretch isn't restricted to the head lobe breaking away from the body and exhibiting the symmetries you mention as a result (and mirrored matches from head rim underside to the body shear line below). According to the stretch hypothesis the single body evolved from a potato shape to a diamond shape then to a diamond with a herniating head. Only then did the head shear. The diamond was therefore a quasi-ellipsoid. The tensile force vectors of stretch coupled with the resilience of the small chunks of crust at Aker and Apis explain why it formed a diamond shape and not an ellipsoid. That's why those two regions are uniquely smooth, the same width, at either end of the body and straddle the equator, which defines the stretch vector plane.

      Crucially, the crust yielded to the stretching core by delaminating its onion layers into thinner layers along the long axis with some outright tearing where it couldn't keep up. There were further coriolis delaminations once it was loosened.

      Virtually all the so-called erosion you see is as a result of those delaminations and tears. Little or no mass wasting occurred between these rifts because they simply opened up rather than being gouged out by erosion. The two classic 'eroded' depressions, Nut and Aten aren't eroded depressions at all. One appears to be a depression because an onion layer rifted up next to it (Nut) and one had an entire line of cliffs arrive on its doorstep one day, instantly creating a depression (the cliffs of Aten slid from the shear line at Babi).

      As for the shore lines, they're the signature of the delaminations and tears leaving material behind at the base of their cliffs as the cliffs slide. Sometimes there are multiple parallel lines of boulders betraying the movement of the cliffs e.g. in the U-shaped valley just above Philae's location. Philae is itself sitting in a ~150m wide x ~600m long tear that aligns with the direction of of the boulder lines. The boulder lines kiss the two ends of the tear betraying delamination above Philae and tearing below him, all attributable to one 150m movement. The two sides of the tear twist and turn in parallel along its length: it's a gigantic rift and not carved out by sublimation. The rift is perpendicular to the long-axis stretch vector as you'd expect. The Bastet side of the rift is V-shaped and nested to the same size v's at Anhur directly below.

      So all the roughed-up areas around Philae including the boulders, the cliffs, the chasms and the shorelines are adequately explained by a tearing, delaminating, stretching comet. There's no need to invoke sublimation although that is of course occurring at some level.

      Marco and I don't have to think any of this up on the spot in answer to the objections to stretch (which are welcome). Everything described above has been on the stretch blog for 6-18 months, totally internally consistent and being used daily by both of us as inputs to search out the movement of new ever-smaller, cliffs, rifts and shorelines.

      That movement was entirely due to stretch, some of it possibly/probably ongoing (see Marco's comment above). The matches along the cliff/rift perimeter tear lines and the shore lines between them have remained virtually unchanged throughout the process. The multitude of matches tell us that.

  • Marco Parigi says:

    Hi Kamal, You may not be aware of the blog post thread that has a collation of actual found changes.

    This as opposed to what I think you are suggesting - that this crumbling, cracking and so forth are unrelated to stretch.

    The first astounding change that I found, was a trio of boulders in an equilateral triangle on Anuket, becoming a bigger equilateral triangle slightly closer to the main crack.

    I am really curious for an explanation for it that doesn't invoke underlying lateral crust moving (ie. Stretch or tectonic plate style movement below the surface)

    Check it for yourself - Anuket is an amazing place!

      • ianw16 says:

        Re your blog post; it's vaguely possible that the red boulder has moved, but far from convincing. In the first image, I think that the red boulder appears closer to the other two boulders in the triangle due to foreshortening. That is, it appears to be close to the brow of an incline that leads down to the other two boulders. The second image shows it from a higher viewpoint. where the incline is more apparent.
        Also, when taking a (very) rough measurement from what appears to be a kink in the 'crack'. or whatever it is, there doesn't appear to be that much difference in distance between the red boulder and the large boulder on the other side of the crack.
        To illustrate what I mean, I've uploaded an image:

        So I guess what I'm saying, is that there may or may not be evidence for a slight movement of the red boulder (far from conclusive), but is that sufficient to invoke stretch as the cause?
        If somebody can prove to me that the comet's overall dimensions have changed in a certain direction, then it might start to get interesting.

        • Marco Parigi says:

          Hi Ianw16,
          Perhaps you should look at a few more "after and before" images. I looked at dozens as well as these:

          There is no doubt at all - I can now recognise whether an image that includes Anuket, whether it is a pre or post perihelion image. Even from images up to 60km away.

          • ianw16 says:

            What am I supposed to be seeing there? Is the 'crack' getting longer? Wider? Again, I'm not sure how this relates to 'stretch'.
            Is there anything there that wasn't covered in this paper?

          • Marco Parigi says:

            Hi Ianw16,

            No. The crack is neither permanently longer nor wider. What you need to see and confirm through observation of many pre and post perihelion images, is that the distance difference between the triangle of rocks in question is not an optical illusion due to foreshortening or other effects, but a considerable horizontal relative movement.
            This kind of horizontal movement is not covered in that paper you link to.

            Use the length of the wedge shaped rock that looks like a fallen obelisk (Tekhenu in ancient Egypt) and you can note from all angles the relative distances have shifted.

            I am not sure if you are still denying the change I am describing is real, claiming that it is perfectly explainable with sublimation induced changes, or that it is paradoxical, yet says nothing about stretch.

            I get the feeling that whatever the meaning or existence of the change, it cannot say anything about stretch because you believe stretch is impossible until proven otherwise.

            I would like to know some accurate physical measurements like the length of the comet (accurate to the resolution of images) and the width of the neck. Surely that is the easiest way to falsify ongoing stretch.(if calculations of the length stays the same or gets shorter, then ongoing stretch is falsified to everyone's satisfaction)

            Not sure why the length of the comet is still only quoted accurate to .1 of a kilometre.

          • ianw16 says:

            Nope, you've still lost me. If the stretch is occurring between the two back boulders and the 'red' boulder , then that would explain how the 'red' boulder is apparently further from the back boulders. It doesn't explain how the 'red' boulder is closer to the large boulder, given that everything on that side of the 'stretch line' will have moved in the same direction by an equivalent amount.

          • Marco Parigi says:

            Hi Ianw16,

            Am I right in saying that you think it is a paradox that doesn't necessarily say anything about stretch?

            Well before the "Call for contributions" post, I was trawling through Anuket images for just this kind of change/movement. It appears I was the first to find this change, which now seems obvious and significant. I was also at a loss to explain the reduction in distance with the wedge shaped boulder on the other side of the crack. At first, I imagined it to be a concertina style unfolding, where some distances shortened while others lengthened. I am also thinking that there is some tectonic style subduction at the crack. Thus the surface is stretching in sympathy with the subsurface, but at the crack, the surface slips underneath the surface where the wedge shaped Boulder is and the boulders end up closer....

          • ianw16 says:

            "Am I right in saying that you think it is a paradox that doesn't necessarily say anything about stretch?"

            No, I'm saying that I don't see anything, full stop. Certainly nothing that can't be explained by the fact that all of these 'boulders' are on what is quite obviously a sloping surface, and have been imaged from various angles and resolutions.
            If you're having to invoke 'stretch' in one direction, followed by 'subduction' to try to explain apparent changes, then it would appear, from my perspective, to be special pleading for something that I, for one, simply do not see, and/or can explain in perfectly reasonable ways (i.e. foreshortening).
            And, although it doesn't need to be invoked in this particular instance, one should always keep in mind that a 'smooth flow' on Tempel 1 was seen to recede by ~ 50m over the course of one orbit. And 'stretch' most certainly does not need to be invoked for the erosion of cliffs and scarps, due to loss of material from collapse.

          • Marco Parigi says:

            Hi Ianw16,

            The changes seen on Tempel 1 are far more prone to optical illusion and idle speculation than the changes on 67P. This is because there is only two images to compare..

            Are you saying that this change on Anuket is probably an optical illusion?

            Yes, one can look stupid if a purported change is shown to be nothing, but it can work the other way too, if the scientists insist that there is nothing and then have to admit the change is real...

          • Marco Parigi says:

            Hi Ianw16,

            I think this is "You can't prove that it is not an illusion"

            Versus "surely you can see that it is real"

            I really need to work out what the average person is seeing here. It is such an obvious and significant movement I cannot "unsee" it.

            I thought maybe people would see it but be in denial of its possibility. Or see it and not understanding the significance - But not seeing a difference at all is something I did not expect, especially if dozens of images were looked at by someone.

            One can rule out, one by one, all possible causes of illusion, but if one denies there is any hint of rock movement at all there is no illusion to explain.

            Are you saying there is no illusion to explain because they have not moved? Have you seen and understood A Coopers shadow length argument? How could the shadow end positions of the moving Boulder be explained if it hasn't moved?

          • ianw16 says:

            @A. Cooper,
            "Here you go chaps. This should iron it out for you:"

            No, not for me. I still say nothing has moved. I find it easier, from previous attempts at this sort of thing, to look for nearby fixed reference points as a better guide.


          • ianw16 says:

            @A. Cooper,
            And from a further recent image I found:

          • Marco Parigi says:

            Hi Ianw16,

            I too, tried to use those same bays nearby to give me some sort of reference, but the problem is knowing what is fixed. It is not a trivial question, because if the bays are possibly eroding or moving themselves, using them as reference points is far less than ideal. They are also too far away from the rocks in question to be of use because of their own parallax issues. Using Tekhenu as a yardstick is a far better proposition and for me it showe movement in the order of its length, but using shadows as guides to slope is a far more decisive proposition to rule out the very illusion you are invoking to convince yourself they haven't moved.

            I am sure that it is a matter of time and persistence. If you look at enough images yourself, and keep trying to prove to me/us that it hasn't moved, you will get the idea.

            I am quite prepared to eat humble pie if I am wrong, but I will warn you that I have been studying the intimate details of this part of Anuket for months, and the bits of new information and cross checking by A Cooper really rules out the kind of hand-wavy, has to be an illusion kind of argument that you are making against our strong evidence. I am hardly one to wait for official scientific peer reviewed confirmation because that is my nature, but I fully expect official scientific peer reviewed confirmation of this particular change, in due course....

          • ianw16 says:

            "....but I fully expect official scientific peer reviewed confirmation of this particular change, in due course...."

            And I would stake everything I own that you won't get it. I honestly don't think I'm the one being tricked by perspective and slopes here.
            If you follow the curve of 'Bay 2' (in the same way that you follow the curve of the plough to find Arcturus), it always intercepts the back boulder, nearest to the scarp.
            Images from 2014 & 2016 both show high angle images where the line through the back boulders intercepts the point I marked. So neither of them have moved. And following the line of the crack, as well as the ratio of the front boulder to the 'obelisk', and the 'obelisk' to the boulder behind it, shows that it simply hasn't gone anywhere.
            If I can see that, then I expect any peer reviewer would also see it.
            And the 'Bays' have gone nowhere. A little bit of erosion might deepen them, but there is nothing noticeable. And they certainly haven't moved laterally. They'd be closer to the crack (or further away). They aren't.
            I honestly believe that you're 'seeing' something that simply isn't there.
            Hopefully, future Osiris images will give a better view of the area. However, I am not going to be betting against myself in the meantime!

          • ianw16 says:

            "I am sure that it is a matter of time and persistence. If you look at enough images yourself, and keep trying to prove to me/us that it hasn't moved, you will get the idea."

            Not going to happen. I spent a couple of days going through EVERY single NAVCAM and Osiris image that has ever been posted to the image archive browsers. I downloaded EVERYTHING from the area in question. I studied them this way and that. I failed to see any evidence of anything moving.
            When I see a peer reviewed paper saying something has moved, then I'll believe it. Things may have moved. Just not in this instance.

          • Marco Parigi says:

            Hi Ianw16,
            I can see that you are missing something???
            There are very few suitable OSIRIS post PERIHELION images due to the embargo. I hope that you are not mistaking post landing with post perihelion. I only found 1 (one) suitable OSIRIS post perihelion image of the trio of rocks.
            One of the images you have posted also appears to be a mirror image but I didn't really understand your reasoning with the images you posted. They look to prove the point for me - I can tell whether they are pre or post perihelion, and whether the image is mirrored by the relative distance of the boulders - This is despite not having seen the particular image that you posted before.

          • ianw16 says:

            Maybe I'm just better at finding images than you are. Ever thought of that? You have obviously convinced nobody of anything so far; so, any chance you could post what you have, and I'll tell you if I've already seen it?
            Like I say, you have not convinced me, and I very much doubt that you would convince any peer reviewer. So, for the umpteenth time, what are you claiming?

          • A. Cooper says:

            Hi ianw16

            You wrote:

            "I still say nothing has moved. I find it easier, from previous attempts at this sort of thing, to look for nearby fixed reference points as a better guide."

            You can't get any more nearby than a reference point that's kissing a rock in one picture and marooned on the opposite side of the rock in the other. That's what I presented in the case of both moving rocks in Part 55 on the stretch blog (linked in my comment above and at the end of this comment).

            The following arguments are illustrated in my reannotation of your photos which are linked here (comment 8 at top of page):


            It would be a more thorough contestation if you were to point to the particular features the two rocks have moved past and explain why you think they haven't moved past them. Slopes and parallax can't account for something being physically on one side of a feature in one picture and physically the other side in another. It's a topological impossibility.

            I say "physically" to emphasise the actual physical movement of the rock over the top of the feature and on past it. That's in contrast to alignments drawn from other features which may make the rock appear to have moved either side of the drawn line due to sloping and parallax.

            Alignments are a perfectly good method to employ but it's a lot less accurate when producing the short lines between the rocks to features that are five times that distance away. Any error in the estimation of the line between the two rocks leads to an angle anomaly of a few metres, perhaps up to ten metres. That's then extrapolated along the produced line so that the error is then five times greater, approaching 50 metres.

            It also doesn't help if the lines also bend outwards slightly as they progress as yours do. Although, I do sympathise, it is very difficult to draw straight lines freehand as I do too. I spend hours getting them straight but I do use various geometry tools on a flat screen and pinpoint the end point very accurately on mega zoom. If dots dance in between a tiny bit it doesn't matter so much. Your two 2014 pictures showing non-alignments for the two back rocks fall into the trap of angular anomalies and slight bends in the lines as shown in the pictures linked above.

            Those are the two 2014 photos you use to show that supposed sloping ground in the vicinity of the two rocks is causing parallax. However, the irony is that the only sloping ground is 200 metres away from the two rocks, in the form of a tumbling scarp. And you've chosen to include that scarp in between the two rocks and your fiduciary points. That's introducing way more parallax than any sloping in the terrain local to the rocks (which remains resolutely flat from every angle according to the shadowing). It could hardly be more complex in comparison with my fiduciary points on top of which the rocks sat and past which they slid.

            Moreover, you can see in your photos the feature that I marked yellow in my photos. You can see that the lower, 'red' rock has moved from one side of it to the other side and moved to the degree that Marco and I claim.

            The upper-right 'blue' rock is also shown to have moved in your photos, in relation to its own truly 'nearby fixed reference point', the rectangle it physically sat on and then didn't. You can't get any more nearby than that.

            In this case, the rectangle serves as a proxy to the bright green line/crack reference point in my Part 55 link, which I'll link again at the bottom seeing as I'm referring to it a lot here. The bright green line is whited out in your photos but it forms the top edge of the rectangle which is quite chunky and obvious in other photos. The left side and front of the rectangle are discernible in your photos and I've annotated them in my versions.

            The blue rock in question is sitting on the rectangle in the before picture so it serves as its own reference point or area. The bright green line was the upper, short edge of the rectangle. In the after picture the rock has moved one rock width at least to sit outside the rectangle. It's now beyond the bright green line which defines the top of the rectangle and which we know is actually there from other photos.

            So your before/after photos prove that the two rocks in question did move, based purely on the nearby fixed reference points that I used. That is, my reference points that kiss their respective rocks and are visible in your photos.

            Regarding your proposal that the lower rock is between bay 1 and bay 2 in the 'before' and 'after' photos and therefore hasn't moved, your 'before' line points just inside bay 2 and your 'after' line points well into bay 1, proving that the rock has indeed moved. I checked the actual features (crack, rock, bays, imaginary parallel line construction) independently of your annotations and this confirmed that the rock had indeed moved between the exact points I marked in Part 55. That's also shown in the reannotated photos in the link at the top of this comment.

            The reannotations have keys showing what the colours mean.

            Regarding slopes and parallax, it's clear that the change in length of the triangle in relation to its base, whether apparent or real, is at or near a doubling. Put another way, the smaller (initial) triangle presents to us a base-to-vertex distance that's half the base-to-vertex distance of the longer triangle. If this is purely down to the parallax of a sloping terrain, then it's a simple trigonometry problem.

            I should just state that the parallax due to orbiter viewpoint angle is negligible between the two more overhead photos of yours which show a stretching triangle, one small and one longer in the respective photos. So the trigonometry problem can be applied to your sloping terrain issue alone and not corrected for viewpoint angle on top of that. The reason I say viewpoint isn't an issue is that the proportions of distances between other rocks (which we would assume not to have moved) are the same in the two photos:

            The distances I used, rock centre to rock centre were:

            1) top left blue rock in triangle to Tekhenu. Let's call this distance A
            2) the same top left blue rock to the obvious largeish rock below Tekhenu. We'll call this distance B.

            For the fuzzier overhead shot (after shot) A/B is 0.607
            For the clearer overhead shot (before shot) A/B is 0.608

            Measurement error is probably 3% so others may check and find it's nearer 0.6 or 0.62 but they will see they are almost identical. This similarity is a proxy for almost zero viewpoint angle change and therefore we can embark on the trigonometry problem with just your terrain issue in mind.

            If you tip the vertex of a triangle from flat on a table while letting the base stay put and swivel, the triangle will appear to contract its base-to-vertex distance making it fatter (when viewed from above). For it to present half its former length, it has to be tipped 60° (cos 60 is 0.5). Even if our 'before' triangle were 0.7 times as big as the 'after' triangle, it would have to be on a slope of 45° (cos 45° is 0.707). If the slope was 20° (which it clearly isn't) then the shortening would be a measly 6% to 0.94 of the longer version (cos 20 is 0.939). The triangle is stretching out of all proportion with any possible terrain slope/parrallax angle configuration that could explain it.

            Besides, the above paragraph, your third photo is looking from the south at a fairly low angle and no indication of any sloping terrain. My example in Part 55, used to drive home this point, is a lot lower still. You didn't comment on that argument just as with the fiduciary points the rocks sat on and then didn't sit on.

            Incidentally, did you find it odd that you were comparing rock movement to a pair of bays on one side of the triangle and then to a similarly sized pair of bays on the other side of the triangle? There's a third pair at the top of the scarp and all three align with each other, especially when you include the chunks that host them which also match in shape and size.

            The second pair is 200 metres exactly south of the first pair. There's a reason for this. The two long lines running up the neck are the 'ticker-tape' delaminations from two rocky massifs at the shear line below. The two rocky massifs delaminated from one massif south-to-north along the shear line before the head sheared away. They sheared south-north because they were being pulled along the shear line where the herniating head lobe was. They'd have preferred to go west-east but they were at the western end of the soon-to-be neck and trying to stretch around it and on down the sides. The only option was south-north at this end of the head lobe. The ticker tape effect up the neck from two similar massifs is why the lines running up the neck are parallel and share features at similar distances along their length. Like the three pairs of bays. The two lines meet with their shear-line source on the body below which exhibits the original delaminated sections of massif as clear as day. These are repeated on their matching south-north delaminations on the head rim underside 1000 metres above. They delaminated together when the head was clamped to the body. See these two tweets and their replies.



            Very little in this area of Anuket is a puzzle to us at the 100-metre scale. We're not casting around aimlessly for clues. We're completely aware of the mechanisms that have brought about what we see in this region. The mechanisms tell us where to look and so we're not very surprised when we find it.

            Here's the Part 55 post that's being discussed here.


    • Kamal says:

      Andy and Marco: My point was that earlier there was not any other decent explanation, now other possible explanations are emerging. So we need (like you have pointed out above) sharper boundaries regarding what seems impossible using other more conventional explanations.

      Hard to imagine how ices (ice slabs?) under a surface would move. The Consert data seems to suggest similar porosity all through, and I think this is the biggest stumbling block to stretch as you have outlined it. I find it easier to think of thicker material as we go in (so slabs are possible) and more porous stuff on the outside where a lot of long-distance transport might be happening. This was the kind of mental picture I formed from the Mupus drilling. If we find Philae and it is buried under rubble then we can estimate the amount of transport which took place during perihelion.

      • Marco Parigi says:

        Hi Kamal,

        I like to separate things that have direct evidence (eg. outgassing) with things that have indirect evidence (ice slabs) That is to say, that it is outgassing volatiles in their gas form is incontrovertible. I am not saying that there is no ice below the surface, but the idea that the sub surface is rigid is not evident, and there are a number of ways that the sub surface could be non-rigid in a way that allows for slab or layer movement below the surface akin to continental drift.

        So we should not be denying that there could be movement because we have a lack of imagination as to how it could happen given what we think we know is below the surface. We should be looking for evidence of movement and extrapolating backwards to see if the surface evidence matches the movement.

        There are a number of lines of evidence which indicate that the nucleus is not rigid, including the fact that there is no torque free precession in the rotation. ...

  • logan says:

    Maybe in the interests of Gerald:

    "Latest study of Tabby's star offers more weirdness".

    " Over the first ~1000 days, KIC 8462852 faded approximately linearly at a rate of 0.341 +/- 0.041 percent per year, for a total decline of 0.9%. KIC 8462852 then dimmed much more rapidly in the next ~200 days, with its flux dropping by more than 2%. For the final ~200 days of Kepler photometry the magnitude remained approximately constant..."


    "We examine whether the rapid decline could be caused by a cloud of transiting circumstellar material, finding while such a cloud could evade detection in sub-mm observations, the transit ingress and duration cannot be explained by a simple cloud model. Moreover, this model cannot account for the observed longer-term dimming..."

    From the hierarchical accretion model, as just described by B. J. R. Davidsson, H. Sierks, C. Güttler and allies up here. It could be attached to 'Tabby' study that each scale of the particulate is in a slightly different kinetic frame.

    So, IF 'Tabby' is just about to be 'photo-bombed' by a 'tiding' cloud THEN this updated model could explain the sub-mm observations.

    Maybe dust clouds are not kinetically [and thermally] that simple.

  • Bruno says:

    When I first heard about the Rosetta project, I was 30 and now Im 69.I've been a fan all that time. But ftom time to time it's difficult to follow uou guys :
    "The larger TNOs played a further role in the evolution of comets."
    I made a request on the letters TNO in capital on the article : all the Firefox system found was a part of "into", in lower case. Could anybody tell me what TNO means ? And could the writers give a translation of the abreviations they write ? Probably, all the searchers in planetology know, but I'm a woodworker.

    • Claudia says:

      Dear Bruno,
      many thanks for your comments and queries, and glad to hear you're a long time fan of Rosetta.

      I am sorry that your search couldn't find the explanation of the acronym TNOs - it's actually up there in the 5th paragraph, the first time the term is mentioned in this story (it's also explained in the long infographic).

      TNOs stands for trans-neptunian objects, which are minor bodies in the Solar System at large distances from the Sun, beyond the orbit of Neptune.

      Best wishes

Comments are closed.