Landing site search narrows

The approximate locations of the five regions are marked on these OSIRIS narrow-angle camera images taken on 16 August from a distance of about 100 km. Credits: ESA/Rosetta/MPS for OSIRIS Team MPS/UPD/LAM/IAA/SSO/INTA/UPM/DASP/IDA

The approximate locations of the five candidate regions are marked on these OSIRIS narrow-angle camera images taken on 16 August from a distance of about 100 km. Credits: ESA/Rosetta/MPS for OSIRIS Team MPS/UPD/LAM/IAA/SSO/INTA/UPM/DASP/IDA


Using detailed information collected by Rosetta during its first two weeks at Comet 67P/Churyumov-Gerasimenko, five locations have been identified as candidate sites to set down the Philae lander in November.

The Landing Site Selection Group (comprising engineers and scientists from Philae’s Science, Operations and Navigation Centre at CNES, the Lander Control Centre at DLR, scientists representing the Philae Lander instruments and ESA’s Rosetta team) made the decision this weekend at CNES, Toulouse.

An overview of the five sites is shown above. The sites were assigned a letter from an original pre-selection of 10 possible sites, which does not signify any ranking. Three sites (B, I and J) are located on the smaller of the two lobes of the comet and two sites (A and C) are located on the larger lobe.

Read the full story and view close up images of the individual sites in the announcement just posted on the ESA Portal: Rosetta: Landing site search narrows



  • Andrew R Brown says:

    Thank you so much for this incredible update Emily.

    Very prompt too, also very grateful for that too. 🙂

    Even without the Philae sites, these are astonishing images.

    Interesting to see that the prime sites are on differing terrain, more diverse than I expected.

    My favourite is site C. Looks like a vent opening or an impact crater, either way exposing a sub surface.

    Excellent stuff. 😀

  • Andrew R Brown says:

    To be honest I am also quiet taken by candidate B too. The crater on the top of the ‘head’ has always looked interesting.

    So it has to be B or C. Perhaps now B for Philae and C if the it is decided to land the Rosetta spacecraft at the end of the mission, returning data and images during a slow descent.

  • Hilmar says:

    Doesn’t it pose any problems that gravity is only about half of the estimated value, if you land on “inside” parts of the comet?

  • cosmo says:

    Was it really essential that the resolution of the three overview images of the Philae candidate landing sites had to be downgraded? 🙄

  • Peter G says:

    What I like so much about this mission is that the clock is always ticking. Do or die within the time available, that’s it, without the option to insert (asymptotical-) timeline-increases in pursuit of lower risk factors.

    Go Philea! Go Rosetta!

  • Derek says:

    I think A would have the best view. Many of the other sites wouldn’t allow you to to take in the stunning geometry of the comet, whereas at A you could get fantastic photos of the bulbous, alien mountain towering above.

    • Ales H. says:

      I agree. Site A is also my favorite (because of nice view from this spot).

    • Hokai2 says:

      Also agree on site A. Looks like a deeper layer near the protruded lobe. (smaller lobe due to a possible massive loss after a hit) in any case a place to gather more information about the comet. If landing performs well it will be an outstanding achievement in space study history. Congratulations ESA ()

  • While I understand everyone’s interest in Site A for the spectacular view of the secondary component, keep in mind that it doesn’t seem the lander’s cameras can look “up.” There’ll be a spectacular view, yes, just not one Philae can see.

  • Dave Higgins says:

    The full description of how and why the landing sites have been selected is great, however the last paragraph ends with a passage that could have been written 10 years or ago or even 10 years before the launch.
    The great thing about the mission has been the anomalies that have been found that run contrary to the 100 year old theory of the comet formation, for example,
    – No mention of ice at the surface (despite it should be a dirty snowball
    -Temperature of comet higher than expected.
    -Density higher than expected
    -First dust particles collected seem bigger than was originally expected to big for theorized interstellar dust left over from the solar system birth.
    -Seems to be no out gassing jets visible at the surface even though there is a coma.
    -crater shapes (often hexagonal, with steep thin walls do not look like impact craters, despite that you might expect from impacts from steady accretion over 4 billion years.

    These are all exiting differences which will require lots of analysis to determine exactly where and how the comet was formed.
    Also your data readings probably have more anomalies that the public has not seen yet.

    So for the last paragraph to begin ‘Comets are time capsules containing primitive materials left over from the epoch when the sun and planets formed’ seem just a tad unimaginative, when the data you already have suggests a paradigm shift from the current theory.

    • Andrew R Brown says:

      Hi Dave,

      H2O I am sure has been detected in the coma. Also the density is less than expected.

      The dust certainly is in greater abundance than expected for sure, but this does not counter any current ideas, 67P has made several crossings of the Asteroid Belt, would not surprise me if there is sulphur captured from the volcanoes from the Jupiter moon Io during ‘close’ Jupiter passes.

      The comet has seen a lot of action, but the warmer than expected dusty coating still does not rule out a primarily icy body underneath.

      We are going to learn a huge amount for sure.

      • Dave Higgins says:

        Thanks Andrew, it just looks like a lack of ambition, which is probably not how the team feel.

    • emily says:

      Dave, this standard text appears at the end of all formal releases as a way to provide a very short summary of the mission in question and the partners involved; it is not intended as a conclusion to this specific story. We’re looking forward to covering the many science topics you listed in much more detail in the fullness of time, once the data have been collected and analysed, and can be used to draw meaningful conclusions.

  • Alan Mehew says:

    Can’t wait for closer views of this beautiful object. some of the features look like layers. I assume it is a chunk of something bigger? If so, what?

  • rob knaap says:

    Great pictures but can we have a SCALE BAR on all pictures please.suxleai

    • emily says:

      Scale info is available if you click through to the image captions. For context, the comet is about 4 km across.

  • Robin Kouba says:

    I vote A, C 🙂

  • Dave Robinson says:

    I am curious to know how the surface, with its clearly visible craters and smooth areas, compares to that of the moon and other planets (when photographed from a similar distance)? Are the craters similarly sized and distributed, for example?

  • Clive Hartland says:

    Has radiation from Alpha, Beta or Gamma been measured?


  • Kamal Lodaya says:

    My guess is one should be on a site where one does not expect outgassing where the lander can do its sampling of the soil, and from where one does expect to have views of sites where one does expect outgassing to take place. Do we have any data (or will get over the next month) as to where the outgassing is (will) take place from?

  • Haring says:

    A safe landing is the first prerequisite: to me, “B” and “C” appear to have clear ground, with no obstacles, easier places to be approached. They are also distant from the neck that restricts the view and could see channelized particles flying into, when the comet would start delivering material.
    Just my silly opinion.

  • Michael Wiggins says:

    Congratulations ESA on this AMAZING achevement … and the exploration hasn’t even STARTED!

    I, too, am a proponent of Site A. While the view Philae will have there will make for great pictures, I have other reasons as well.

    1. By being able to see more of a percentage of the comet from its vantage point, Philae will be able to see more of the outgassing from other parts of the comet. The other sites (IMHO) are only at the “top” or “bottom” of the comet. This would limit the total science Philae might be able to return.

    2. Flat landing surface.

    3. By having a horizon, this can help with navigation (I know: Rosetta is already there), establishing visibility factors as the comet begins to outgas. For example, if you know where the sun is going to rise at any given time you can point the camera(s) toward that location (even if it’s not visible) and “follow” it with the idea of determining the amount of particulates should the sun happen to shine through. (I hope that made sense. Sorry if it didn’t.)

    Anyway, I can’t wait for the days and weeks to come! Go Rosetta and Philae!!

  • logan says:

    Please consider risk of fluidization of regolith.

  • Vincent Pinto says:

    Let’s assume that the lander lands exactly as planned and after landing, everything on it functions as planned. Is the camera spatial resolution on Rosetta high enough for it to be planned to take pictures of the lander on the surface?

    If not, how large would the lander have to have been for it to be possible, or conversely, how low in orbit would Rosetta need to be to make this possible?

    Finally, I suppose the lander has a means to take visuals of the surface, etc. Can the lander camera be pivoted to take a picture of Rosetta?

    How cool will all this be!

    • Vincent Pinto says:

      Meant to add this earlier…

      Could someone in the technical team please let us know if there is a plan to do this? Thanks.

  • Roberto Nesci says:

    The resolution of the images with the landing sites marked is about 2 m/pixel, which is the approx size of Philae itself!
    The circles drawn on the selected areas are about 400 m diameter

  • Frank Desmond says:

    When Philae finally does leave Rosetta,
    how detailed will Rosetta’s pictures of Philae be ?

    Actually is ‘Philae’ written on the craft and
    when it actually lands on 67P would people back on Earth
    actually be able to read that using Rosetta’s cameras ?

    And vice versa : can cameras on Philae look back up at Rosetta ?

    Advances in modern science : selfies from a comet.

  • logan says:

    Amazing place!
    This is going to rewrite a lot of science.

  • A.Cooper says:

    It seems the head of the ‘duck’ fits onto the body so I checked all the photos released, taken from many different angles, and have found the following:

    I’ve matched contours along the rim of the head to the same contours around the main body perimeter. I’ve found humps and depressions on the underside of the head (the cliff) that fit closely with their mirror counterparts on the body. 

    I’m therefore near certain 67P/C-G was once a single rock and that the head was stretched away from the body. That would mean 67P/C-G is not a contact binary or a scooped-out rock. 

    The only scenario I can propose for the stretching is that 67P/C-G underwent a close pass to Jupiter, under the Roche limit (240,000km), at some point in the distant past. Jupiter perturbs 67P regularly so the MOID could have been even smaller in the past. 

    Here is a tentative suggested scenario: the spin (z) axis of 67P is  roughly at right angles to its 1959 Jupiter-centric radiant leading up to closest approach (references: Lowry (2012)- axis longitude “78 degrees”; JPL Horizons approach over 25 days: 23H30M to 23H50M /+8 to -65 deg declination). Therefore the current end-over-end rotation is in line with 67P’s approach along that radiant. 

    At closest approach the 1959 radiant swung from left to right of Jupiter’s prograde motion, sweeping across its bow and under its south pole. 67P was perturbed but it was too distant for a Roche pass. However, The 1959 approach radiant would be in line with the exit radiant of the much earlier Roche pass so that would have been a Roche pass across Jupiter’s bow from left to right and under the pole as well. The spin axis and oriention neatly corroborate this as being the path of that purported, ancient Roche pass. 

    Roche passes can stretch asteroids into dumbell shapes that tumble end over end along the radiant, just like 67P/C-G.

    I should think the pass altitude was between 150,000km and 220,000km. Any lower and 67P could be shredded (like SL-9); any higher and it wouldn’t stretch or undergo spin-up. 

    I presume this has implications for outgassing- the comet has been ripped in two and its guts are just beneath the dust on the ‘neck’. It might explain 67P/C-G’s observed high activity. 

    I am hoping to do some Roche pass simulations with a colleague soon. I hope all this is useful- and I can share my photo observations.

    • Michael Wiggins says:

      Interesting analysis, A, Cooper. While I didn’t understand the minutiae of the Jovian perturbations, I was pleased to learn something new: that Jupiter affects the orbit of 67P/C-G the way it does. Based on what you described, the comet sounds like one of the ones that got away (from Jupiter, that is).

    • Marco says:

      I can propose other scenarios for the stretching. Internal processes, spin up and spin down due to jets. Centrifugal forces are not enough at the current spin rate, but the spin has got the right kind of axis. Measurements over a heliocentric orbital cycle will show whether the lobes are moving further apart due to the combination of spin, internal forces and evaporation of volatiles. We could them extrapolate to the future when the lobes would separate completely and to the past, when the comet was an ovoidal singular lobe. We could also see if the same thing is happening with Hartley.

      • A. Cooper says:

        Hi Marco,

        Yes, spin up and spin down is a possibility too. I was trying to incorporate the fact that the end-over-end tumbling is in line with the jovicentric radiant, which is rather a coincidence.

        It occured to me that now we have the mass of 67P, we don’t need to go by the nominal 1mm/sec^2 gravitational force they used for simulations. I worked out that it would be around 0.3mm/sec^2 at the extremities of the original lump before stretching. That force would have to be overcome before the comet started to stretch, break or build up tensile forces. To overcome it, I calculated it would be a 1.5 to 2 hour rotation rate or alternatively a sub 135,000km Roche pass.

        I referred in my above comment to Shoemaker-Levy 9 breaking up at 140,000 km but I now think that must have been the estimated threshold below which it could break up notwithstanding its tensile strength, so it may have gone a bit lower.

        After more scrutinising of photos I’m convinced 67P was one original lump that split in two. Furthermore, I can see that three vast slabs along the side of the main body were dislodged by the uplifting head but didn’t travel up with it. Their perimeter is delineated by the main rift line from which the head lifted and the curved areas further down the body which look like fractured marble. They cover the whole of one side, describing two large arcs sandwiching a much smaller arc. Landing site A is one of these arcs but I think the rift extends even further round from there, anticlockwise to a particular angled ridge that had joined to a matching ridge on the head. All these slabs are gone but in a spin-up or Roche pass, they would have drifted away into slightly different orbits. However, there is some slab-like material wedged against the perimeter of site A in a place that is consistent with the likely force direction for both spin-up or Roche pass scenarios.

        • Marco says:

          Thanks for the reply, A. Cooper. I was convinced, the first time that I saw come Hartley, that separation of lobes is something that comets do as a rule. Close ups of C-P just confirmed it to me. I think something more fundamental is the plasticity it implies for the nucleus as a whole. Putty, plastic or rubber rather than ice and dust, as a model.

          • A. Cooper says:


            I too have looked at Hartley- and Halley as well. And asteroids such as Itokawa and Kleopatra. All peanut shaped. Hartley and Itokawa have flung their boulders to each end leaving a smooth saddle in the centre. Ten percent of NEO’s are peanut shaped so there must be a common mechanism afoot. It was this NEO proportion that got me and my collaborator into modelling Roche passes because they could have come very close indeed in the distant past. Even after significant Roche perturbation they still cross our orbit at the same node and remain as NEO’s. Some of our modelling results produced shapes almost indistinguishable from Kleopatra. But for comets I agree that the outgassing must have a net torque effect, especially if they’re not the Jupiter-crossing type.

            I agree 67P/C-G is probably like putty or maybe plasticine. Although it looks like conventional rock, it is about 10 times less dense and (if there is little molecular cohesion) it is subject to only minute amounts of gravitational cohesion. A friend said it didn’t look like a rubble pile but solid rock. I wonder if it is a pebble pile, grit pile, compacted dust pile or all three. There are some strange rock-like stratifications which are very puzzling to me but it would suggest sub-micron, dust compacted to some extent, like silt stone. So it would have some integrity but still be stretchable enough to morph and set in a new shape. If compacted though, it would have had to have been part of a bigger body aeons ago.

          • Marco says:

            The internal structure must necessarily be different from the external structure. CONSERT experiment should elucidate on that. Primary evidence for that is that material ejected is noticably incompatible with the surface composition. Jets from within powered by volatiles not evident on the surface. Material ejected not black. We should be expecting a honeycomb layer below the surface, and volumes of mud- fines mixed with liquid water deeper down. Considerable variation in density was found on Ikotawa between lobes, and the same should be with C-P. Anything liquid may flow towards the lobe ends.

          • Marco says:

            Re: a pebble pile, grit pile, compacted dust pile or all three.
            Why would any of these be black and like putty? Crude oil, tar, vulcanised rubber are black, would stretch but still have some integrity at the temperatures and pressures on the surface of comets. With a surface like that, what is beneath the surface wouldn’t stay a rubble pile.

  • A. Cooper says:

    @ Michael Wiggins

    Hi Michael . If the Roche pass theory is correct then yes, 67P/C-G would have been one that got away. It would’ve bent round fairly drastically but not enough to go into an orbital ellipse. It’s called a hyperbolic orbit. This all happens as the planet and comet orbit the sun so the jovicentric perturbation is stretched out into a graceful kink that changes the comet’s orbit.

    It occurred to me that a test of the theory might be for Philae to look for minute traces of dust from Jupiter’s tenuous gossamer rings. 67P would almost certainly have flown through them in the hypothesised Roche pass because they extend to 226,000km and I guessed an upper limit of 220,000km for the pass (no coincidence there because the rings are there due to the Roche limit). They are 8400 km thick and the comet would take several minutes to go through at relative speeds of perhaps 10-15 km/sec. At that speed the dynamic pressure of the tenuous dust would be greatly augmented and perhaps embed itself in the existing cometary dust.

    I did a rough calculation using info from the Wiki Jupiter rings page. If 67P had absorbed all the ring dust (unlikely) it would have had a concentration in the order of of 0.3 to 30 micrograms per square metre- but only on the leading edge that faced the flow, presumably. If it happened 10,000 years ago, it would be a push to find it, to say the least.

  • Pete Williams says:

    What is the true colour of the comet. Does Rosetta carry any colour cameras?

  • Francis Fitz. says:

    I wonder is they could turn the camera and have a look back at Mother Earth.
    Long days and Pleasant Nights

  • logan says:

    Looks like ‘landing zone A’ is contributing to erosion of ‘neck’.

  • logan says:

    May I suggest to your attention this Great Collapse at the Base of the duck? ‘Canyons’ have always been interesting to science.

    First frame of

  • logan says:

    Starting to see along the comments the forbidden word: mud. (Frozen dry)

    • Marco says:

      Why frozen? Why dry?

      Frozen dry mud is just dust.

      So thawed mud requires some gas pressure and temperature above zero degrees. Why is it forbidden or deemed impossible?

      • logan says:

        Hi Marco, thanks for the reply. Just think of coffee dust and coffee crystals.

      • logan says:

        You are absolutely right about what is needed to form mud.

  • The CNES report about the 5 candidates highlights the special qualification of ‘I’ over the other four – “la zone I fait l’unanimité des ingénieurs et des scientifiques. Les zones C, J, A, B sont les autres sites candidats mais ils ont tous leur défaut” – but I don’t see that mentioned in the concurrent ESA, DLR or MPS releases. Why?

Comments are closed.