On 28 September, a scientific paper was published in Nature, presenting a view on the formation scenario of Comet 67P/Churyumov-Gerasimenko, based on Rosetta OSIRIS images (read our news report here). The paper, led by Dr Matteo Massironi of the University of Padova, Italy, evaluated two possible models to explain the comet’s curious shape: the merging of two cometesimals or the erosion of a single object. Observational data and thorough analysis of the comet’s gravity field pointed towards the first of the two hypotheses: 67P/C-G seems to have originated from two separately formed comets that merged at low speed.

The topic has generated a lot of discussion in the comments thread of this blog, in particular by proponents of an alternative explanation put forward by our blog readers Marco and A. Cooper who suggest that the comet was once a single body that has since been stretched into two separate lobes.

We asked Matteo to share his opinion not only on this subject but also on the general topic of how planetary images are interpreted in order to arrive at a robust scientific theory.

Here is what he wrote:

Sorry for my late reply, but I wanted some time to go through the blog where the stretching hypothesis is described before saying something about the supporting geological proofs. As a scientist, I do not normally read works that have not been rigorously reviewed and published in scientific journals. Probably that is my fault and I want to thank ESA to have created this interesting blog allowing so many people to be an active part of Rosetta’s discoveries.

As a geologist, I would first like to point out some basic principles to follow when dealing with space images for geological interpretations. I believe these suggestions could be of some help to anyone who is going to submit his or her own work on cometary geology and structure.

These principles can be synthetized as follows:

  • Be a good field geologist
  • Be rigorous
  • Never fall in love with theories (particularly your own ones)
  • Be short

1) Be a good field geologist
Whenever we observe a geological landscape we filter it on the basis of our own experience. Becoming a good field geologist is a hard task because it requires the combination of a deep knowledge of geological processes and a large amount of direct experience. I normally spend half of my time working on Earth geology and doing science based on field geology. Indeed, working in the field deeply increases your skills in correctly recognising geological features (tectonic, volcanic, glacial etc.). If you are not able to recognize different geological features you might fall into common perceptive traps by, for example, perceiving continuous illusory lines linking different unrelated geological objects.

You might like to enhance your own geological experience by stepping back from your laptop or work-station, leave 67P/C-G images where they are, wear a good couple of boots, have a walk and reach a panoramic view where rocks are beautifully exposed. Since Comet 67P/C-G displays layered material, you should choose a good exposure of a well-stratified sedimentary succession or a heavily foliated metamorphic sequence. Once you have reached your observational point, sit down, take a breath and start drawing what you see.

You should at this point be able to recognise different features: the primary ones, which are strata and foliations, the secondary ones, that can be faults and fractures, and the morphology which in your sketch is the line separating the sky from the landscape. Most probably you will see that the morphology might in places follow the primary structures, in other parts the secondary structures and in many others it simply follows his own shape given by erosional processes. Any matching exercise in geology should take into account only primary and secondary structures, while the morphological ones, if not related to the previous ones, should be avoided.

For example, restoring a faulted and sheared geological section should be based in collimating the primary structure after having recognised the faults where shear took place. The surface morphology should be avoided because it is affected by later erosional processes, which in any case on Earth are often much less relevant with respect to the ones happening on comets.

In substance, we cannot rely on the morphology of a heavily changing body such as a comet, but only on the expressions of its internal structural features.

In any case, after your panoramic drawing is completed, please walk into your panorama and see if what you have interpreted is correct (it would be better to be accompanied by a field geologist). After this simple exercise, come back to your work-station and try again to interpret the comet images from Rosetta. You will probably find something different than what you first thought.

2) Be rigorous
Any time I work on planetary images, I need correctly scaled and calibrated data that can be overlapped onto Digital Elevation Models or Shape models. In this way, I know perfectly what the possible errors are on my geo-structural interpretation. The Rosetta-OSIRIS team is really smart on these procedures and many people from MPI, DLR and LAM are involved in image calibration and shape model realisation (applying different methodologies). Hence, for example, I know that the error in the estimate of angular deviation from perpendicular between the strata and the local gravity vector is on average of 2° and in the worst case of 4°. In any case, my point is that any promising observation deserves a thorough and detailed investigation to understand the possible errors.

My advice “be rigorous” also means to go deeply through any published works before critically judging it.

3) Never fall in love with theories (particularly your own)
In general, science works from observations to hypotheses to models to theories. An observation may trigger a hypothesis, which should be tested with different approaches trying to find evidences that might undermine your own hypothesis.

For example, neck-enhanced erosion and simple contact binary of two young cometisimals were the two favoured scenarios for Comet 67P/C-G’s formation because they do not present any problems to existing theories related to Solar System formation and evolution. On the contrary, a contact binary of two fully-formed comets with an ordered onion-like inner structure raises relevant issues with pre-existing models (already correctly highlighted by some readers of this blog in the comment section).

In my opinion, these issues can be synthesized as follows: i) understanding how strata can form in the primordial Solar System, ii) how a comet might avoid any catastrophic collision during their lives, and iii) how low collisional impacts can happen in the primordial proto-planetary disk. The only issue that was already solved is the formation and stitching of contact binaries through low velocity collisions (see the publication The shape and structure of cometary nuclei as a result of low-velocity accretion by Jutzi & Asphaug for more information).

Due to the controversial implications that the onion-like contact binary raises, we tried to find other lines of evidence that might undermine what was apparent from the former observations. This is why, from the best fitting planes, we passed to the geological sections and afterwards worked on the angular relationships between strata and the local gravity vectors. All these independent observations based on primary structures support the view in which the comet derives from a contact binary of two comets with an onion-like interior.

Our cometary mission is called Rosetta because it has the potential to decipher the origin and evolution of the Solar System. We should put aside our previous theories and models and try to study images without forcing our observations to a particular model or another. Theories and new models will come later (stay tuned!).

4) Be short
When presenting your theories it is important to be short and concise – more words do not necessarily make a scientific hypothesis stronger.


With those points in mind, and after spending some time familiarising myself with the stretch hypothesis of Mr Cooper, and with the comments posted on the original thread, I have the following remarks:

-It seems that most of Mr Cooper’s observations are based on surface morphology with many primary and secondary features not correctly recognized (see point 1 above for advice on geological interpretation)

-He begins explaining the stretching hypothesis, providing the observations that should support his theory only afterwards. In general, science works the other way around, from observations to hypotheses to models to theories (see point 3 above). On the contrary, his observations in support of the stretching hypothesis are only based on putative matching points recognised on the surface morphology. Something else is really needed to corroborate the hypothesis.

-Some other blog readers and commenters have asserted that matching is enough and probably the same was for Alfred Wegener’s description of continental drifting and the later plate tectonics theories. In fact, this is not correct: even Wegener provided many independent lines of evidence including geodetic, geologic, geophysics, paleontological and paleo-climatic arguments. And the same was for plate tectonic, with additional paleo-magnetic and geostructural evidences.

– Relevant suggestions from the general public are valuable to scientists, but please take your time to become familiar with the supporting scientific literature before immediately arriving at a different conclusion.

– If anyone wants to submit a scientific paper on the stretching hypothesis (or any other topic!) please keep in mind the points above about geology, rigour and providing convincing evidence in a concise way.

Best wishes,