≡ Menu

Inconstant Moons: A New Lunar Origin Scenario

A recent snowfall followed by warming temperatures produced a foggy night recently, one in which I was out for my usual walk and noticed a beautiful Moon trying to break through the fog layers. The scene was silvery, almost surreal, the kind of thing my wife would write a poem about. For my part, I was thinking about the effect of the Moon on life, and the theory that a large single moon might have an effect on our planet’s habitability. Perhaps its presence helps to keep Earth’s obliquity within tolerable grounds, allowing for a more stable climate.

But that assumes we’ve had a single moon all along, or at least since the ‘big whack’ the Earth sustained from a Mars-sized protoplanet that may have caused the Moon’s formation. Is it possible the Earth has had more than one moon in its past? It’s an intriguing question, as witness a new paper in Nature Geoscience from researchers at the Technion-Israel Institute of Technology and the Weizmann Institute of Science. The paper suggests the Moon we see today is the last of a series of moons that once orbited the Earth.

“Our model suggests that the ancient Earth once hosted a series of moons, each one formed from a different collision with the proto-Earth,” says co-author Assistant Prof. Perets (Technion). “It’s likely that such moonlets were later ejected, or collided with the Earth or with each other to form bigger moons.”

To explore alternatives to giant impact theories, the researchers have produced simulations of early Earth impacts, varying the values for the impactor’s velocity, mass, angle of impact and the initial rotation of the target. The process that emerges involves multiple impacts that would produce small moons, whose gravitational interactions would eventually cause collisions and mergers, to produce the Moon we see today. Here’s how the paper describes the process:

… we consider a multi-impact hypothesis for the Moon’s formation. In this scenario, the proto-Earth experiences a sequence of collisions by medium- to large-size bodies (0.01–0.1M). Small satellites form from the impact-generated disks and migrate outward controlled by tidal interactions, faster at first, and slower as the body retreats away from the proto-Earth. The slowing migration causes the satellites to enter their mutual Hill radii and eventually coalesce to form the final Moon. In this fashion, the Moon forms as a consequence of a variety of multiple impacts in contrast to a more precisely tuned single impact.

Here’s a graphic from the paper (listed as Figure 1) that shows the process at work:


Image (click to enlarge): a,b, Moon- to Mars-sized bodies impact the proto-Earth (a) forming a debris disk (b). c, Due to tidal interaction, accreted moonlets migrate outward. d,e, Moonlets reach distant orbits before the next collision (d) and the subsequent debris disk generation (e). As the moonlet–proto-Earth distance grows, the tidal acceleration slows and moonlets enter their mutual Hill radii. f, The moonlet interactions can eventually lead to moonlet loss or merger. The timescale between these stages is estimated from previous works.

The Hill radius mentioned above describes the gravitational sphere of influence of an object; in this case, meshing Hill radii can produce interactions that sometimes lead to mergers. The paper notes that in head-on impacts, the rotation of the planet is important because the disk needs angular momentum resulting from the rotation to stay stable. With increased rates of rotation, the angular momentum of the disks increases. Moons like ours emerge from many of the simulations:

We find that debris disks resulting from medium- to large-size impactors (0.01–0.1M) have sufficient angular momentum and mass to accrete a sub-lunar-size moonlet. We performed 1,000 Monte Carlo simulations of sequences of N = 10, 20 and 30 impacts each, to estimate the ability of multiple impacts to produce a Moon-like satellite. The impact parameters were drawn from distributions previously found in terrestrial formation dynamical studies. With perfect accretionary mergers, approximately half the simulations result in a moon mass that grows to its present value after ~20 impacts.

If the multi-moon hypothesis proves credible, how would it affect the larger astrobiology question? In Ward and Brownlee’s Rare Earth (Copernicus, 2000), after a discussion of obliquity and the Moon’s effect on the Earth’s early history, the authors say this:

If the Earth’s formation could be replayed 100 times, how many times would it have such a large moon? If the great impactor had resulted in a retrograde orbit, it would have decayed. It has been suggested that this may have happened for Venus and may explain that planet’s slow rotation and lack of any moon. If the great impact had occurred at a later stage in Earth’s formation, the higher mass and gravity of the planet would not have allowed enough mass to be ejected to form a large moon. If the impact had occurred earlier, much of the debris would have been lost to space, and the resulting moon would have been too small to stabilize the obliquity of Earth’s spin axis. If the giant impact had not occurred at all, the Earth might have retained a much higher inventory of water, carbon and nitrogen, perhaps leading to a Runaway Greenhouse atmosphere.

The idea of a series of impacts eventually leading to a larger moon significantly muddies the waters here. It is true that in our Solar System, the inner planets are nearly devoid of moons, but we have no way of extending this situation to exoplanets without collecting the necessary data, which will begin with our first exomoon detections. Certainly if numerous collisions in an early planetary system can produce a large moon, as this paper argues, then we can expect similar collisional scenarios in many systems, making such moons a frequent outcome.

The paper is Rufu, Oharonson & Perets, “A Multiple Impact Hypothesis for Moon Formation,” published online by Nature Geoscience 9 January 2017 (abstract).


Comments on this entry are closed.

  • James Stilwell January 13, 2017, 14:38

    Learning to safely live on a lunar research station is far more practical than one on Mars. Six days to reach safety…This may also be a factor in advancing human civilization…having a large moon as a stepping stone to the planets…assuming of course that Warp One is a fantasy…

  • LocalFluff January 13, 2017, 15:29

    It is great that new ideas about the Moon’s formation are still being formulated! It is not so easy to challenge an established dogma like the one-hit Moon creation idea (which is a bit suspect in its unique proposal). There was probably just a gradual difference between one big and several medium impacts.

    Mars’ moon Phobos will be tidally destroyed already in about 0.01 billion years, forming a ring and raining down on Mars. Isn’t that a coincidence? And Venus’ was resurfaced by some unknown calamity about 0.3 billion years ago. Maybe Venus was much more habitable for 4 billion years before, and then it too had a moon that crashed into it? At least the existence Phobos challenges the idea that the Solar system has been frozen as it is for almost 4 billion years. Our large Luna, and its formation process, maybe swept up all the Phobos that were gravitationally programmed to kill us today. Maybe moons are destroying most terrestrial planets even billions of years after their formation.

    • Ashley Baldwin January 14, 2017, 5:40

      There are two further moon/exomoon papers out this week published by Amy Barr et al and available through arXiv . Well worth a read as they take this article’s central premis to further bounds and one in particular gives a delightful overview of moon formation taking the moons of of the solar system alan example and looking at their likely mode of formation in turn. Including Phobos , which is particularly interesting and with unexpected findings . A very readable introduction . Some technical stuff like Roche limit and Hill radius but these are absolutely key factors to moon formation ( and planets too ) and are well introduced here.

      As to the formation of moons via impact and subsequent “co accretion ” from the resultant debris cloud – the authors also look at how the impact theory on moon creation ,which though sound , has been limited historically by simulations ( as in the article here ) having been based around the perceived parameters of our Moon’s formation only . What happened if the impactor had been larger or smaller or travelling quicker and at a different angle ? All would have a large influence .

      Yet it hasn’t been done amazingly so the final possibilities are unknown despite the fact that it appears to be the case that impacts in early stellar systems and their circumstellar disks likely occurr very frequently, and at all distances from the star .( Pluto/Charon are likely to have been formed in the same way ) .

      These articles are now appearing I would guess because after decades it’s finally been accepted near universally the way in which the moon formed and thanks to missions like Cassini and New Horizons the wider aspects of moon formation are being understood . All fuelled by the looming shadow of “hunt for exomoons with Kepler” . As David Kipping increasingly refines and adapts the cutting edge investigative processes required I’m sure it will soon bare fruit , possibly even this year . Once we’ve confirmed all the candidate planets in the data , the work will start all over with the moons .

      • Alex Tolley January 14, 2017, 22:10

        Thank you for the refs. I will read them this weekend.

      • Alex Tolley January 15, 2017, 14:27

        The Barr article “On the Origin of Earth’s Moon” is interesting in that it explains the differences in the types of impact and accretion models and the assumptions used. Bottom line is that a single impactor can explain all the features of Earth’s Moon. Whether this is the case, or multiple impacts are a better model probably needs some physical evidence to determine which case was responsible for our Moon.

  • LocalFluff January 13, 2017, 15:33

    Mercury has had its crust removed. Venus has been resurface recently. Mars has a north-south dichotomy and will soon be pounded again by Phobos. Earth has an oddly large moon. It seems hard to be a terrestrial planet in this physics.

    • Ashley Baldwin January 14, 2017, 13:55

      It is at the start of a stellar systems life. All these impacts occur within the first few hundred million years of planetary formation , often not long after the dissipation of the gas component of the circumstellar disk , indeed before hand with the gas giants . Impact followed by co-accretion into a moon of the resultant debris disk that lies between the inner boundary represented by the planetary Roche lobe ( where gravitational tides prevent accretion and draw any material back down to the planet ) and the outer boundary Hill radius ( at which the planetary gravity is outweighed by that of its star also preventing accretion ) . How much debris lies within this depends on the mass of both the planet and impactor , their relative collision velocity and its angle . Modelled in detail for the Earth /Moon as here , but not ironically for other variants of these parameters . Moons formed this way have tell tale characteristics , namely a prograde , low inclination orbit . This would suggest that Phobos which has a prograde orbit nearly Coplanar with Mars equator is in fact created this way rather than being ( as often believed ) captured asteroid . Must have been acused by a low velocity , low mass impactor that left a debris cloud perilously near the Roche lobe. Given the powerful gravitational influence of nearby Jupiter , Mars’ Hill radius would also be much smaller than say that of ( also ) bigger Earth ,hence its close in moons .

      • LocalFluff January 16, 2017, 7:44

        The conclusion that all major impacts happened very early on is based on asteroid cratering history (on the Moon), right? That doesn’t say anything about collisions with moons

        I suppose that Mars’ dichotomy is well known to have been formed early on. But Phobos is going down soon and Venus was resurfaced only about ~0.3 bln years ago.

  • David H January 13, 2017, 16:37

    If this is supposed to have relevance to the question of life’s origin, we should first figure out whether life first formed in deep sea “black smokers” – a theory which, if true, has huge consequences for xenobiology. If much of evolution happened in the oceans and not on land, the issue of whether you need a big moon to stabilize your axis seems less important.

    • Alex Tolley January 13, 2017, 17:07

      I agree. And while a stable rotation is convenient, terrestrial plants and animals can disperse to follow the climatic changes. Oceanic organisms have done so in the past, and even survived massive environmental changes that resulted in mass extinction events. “Life finds a way” as the fictional Ian Malcolm says.

      • Ashley Baldwin January 14, 2017, 14:16

        I think the issue is that the “Rare Earth ” theory posits that intelligent life in the universe is rare rather than life as a whole. To this end Earth’s stable obliquity ( fluctuating by no more than 1.5 degree over several tens of thousands of years ) is mooted as being due to the stabilising presence of its unlikely large moon which reduces the dramatic variations in obliquity and hence climate that would otherwise occur ( as with Mars which whose obliquity has varied by tens of degrees during its geological history ) making the development of advanced life much less likely if not impossible .

        This article by suggesting Earth has been impacted on numerous occasions and had several moons infers that the formation of large moons around Earth like planets is likely to be far more common than previously thought debunking one of the central tenets of Rare Earth in the process. Mars is too near Jupiter’s gravitational influence to have a large moon ( also explaining its low mass ). Venus ,like Earth ,may well have had an impact related moon but this was subsequently lost during a further impact or sequence of impacts . It’s high axial tilt and retrograde rotation might be evidence of this.

        That’s why the Hunt for Exomoons with Kepler ,HEK, is so key as so far we only have one planetary system of moons to base all this conjecture on. Indeed most of the work to date is only based around the formation of just our own moon ! No one really knows just what the proto planetary environment of the early solar system was like so looking at other systems with moons should give a better insight.

        • Alex Tolley January 14, 2017, 18:11

          I think the issue is that the “Rare Earth ” theory posits that intelligent life in the universe is rare rather than life as a whole.

          Full Title: Rare Earth: Why Complex Life is Uncommon in the Universe

          I think it is pretty definitive from reading the title and the preface that Ward is talking about multicellular life as a whole, not intelligent life. Intelligence may be rare anyway for other reasons.

          • Ashley Baldwin January 15, 2017, 6:15

            Subtle . In stating that multicellular life is rare the indirect inference is that “intelligence ” or sentient life is rarer still , indeed limited only to Earth in both space and time. There are many that disagree with both this inference and even more so the rationale behind it . It’s worth doing a quick search on the authors’ publications and who they collaborate with .

  • Alex Tolley January 13, 2017, 17:02

    I’m not sure what I make of this paper.

    On the one hand, the multiple impacts remove the problem of a single hit resulting “just so” in a stripping off enough material to form the Moon of just the right composition. But on the other, we now need tens of impacts of Moon (0.01Me) to Mars-sized(0.1Me) objects to form the Moon. Where are all these huge bodies coming from and why was just Earth delivered a moon, not Mars or Venus?

    I’m not clear how we might test these hypotheses? Would statistical observations of exoplanet moons really distinguish the models?

    • Ashley Baldwin January 14, 2017, 5:47

      Good questions . I would read the excellent and readable articles on moons/ exomoons by Amy Barr et al (who I asked many of these self same questions ) also published this week and fully available on arXiv . They give a wider overview ( including of all the solar system moons )and help address what needs to be done . Professor Barr is very keen and approachable and answered the questions almost immediately ,with relevant references .

  • DJ Kaplan January 13, 2017, 17:08

    David, the Rare Earth theory – among other things – suggests that our large-ish moon helped to make conditions for life on Earth possible:

    It’s kinda based on a small sample size and glancing around about what’s different about the Earth compared with the handful of other planets in our system.

    But by integrating incoming objects into its own mass, we can think that it protected us, but not from the Chicxulub impact.

    • Ashley Baldwin January 15, 2017, 17:32

      A bit maybe by standing in the way , but it’s gravity is small and along with its proximity doesn’t help attract many potential impactors. Jupiter has the biggest influence on the inner solar system. Acting as shield protecting the inner planets from impacts was another tenet of Rare Earth hypothesis. Current thinking is Jupiter’s effect at best is neutral. It certainly has helped protect against long period comets ( big, fast and very deadly especially as their arrival cannot be anticipated with a long run in time ) as evidenced by Shoemaker-Levy and originating from the outer regions of the solar system.

      However at the same time disturbing asteroids from the main belt and surrounding area into the inner solar system ( one of which probably caused the KT impact ) . If it has had any beneficial effect it may have been by ejecting lots of Kuiper belt and main belt objects during its “grand tack” 3.85 billion years ago (though even then leading to the late heavy bombardment but at least long before vulnerable multicellular life needed to worry about such things . )

  • Michael January 14, 2017, 15:19

    Slightly off topic, there is a local astronomy club in Welwyn garden city and Prof Cathie Clarke will be giving a talk on proto-planetary discs and planet formation if anyone is interested in the U.K. It would be interesting to ask Professor Clarke how likely moons are to form around planets in these ‘discs’.

    If the skies are clear there are normally some telescopes that can be used and there is a very good bar available.



    • Ashley Baldwin January 15, 2017, 6:24

      Sounds good . Surprisingly little work has been done on this important area even from a solar system point of view. Dependent on many factors . Wish I could go . If you want to swat up ahead of attendance , the most comprehensive study ,with simulations, was done by Morishima et al in 2010. “From planetesimals to terrestrial planets” …….Published as a preprint on arXiv in July 2010.

  • Michael Fidler January 15, 2017, 2:10

    It would be interesting to see a mission to the moon designed specifically to find the ages of all the large craters on the moon. Could this be done from remote sensing or would it require possible miniature probes that could be sent so that they have a deep impact and could measure the radioactive dating? I know that a
    technique called as Crater Counting is the standard method now, but a dedicated mission might give us more clues to the impact history and future of impacts on the earth! Since the moon is a living fossil of this information it would seem to be the perfect place for this kind of research.

    • Ashley Baldwin January 15, 2017, 17:44

      I’m no expert on this but I understand this has largely been done. First by working out a map of what lies on top of what ( such as lava flows ) and combining this with accurate radio metric dating ( involving various isotopes with different half lives ) of samples returned by the Apollo and Luna missions from across the Moon which together gives reasonable coverage .

      An extension of this compared to that of Earth’s crust first supported the idea of an impact followed by co-accretion origin of the moon sometime between 50 and 100 million years after the formation of Earth .

      • Michael Fidler January 17, 2017, 9:09

        Yes, but I’m not talking about the Moons origin but the later impacts up to the current recent history and its relation to impacts on earth. Much of earths history of impacts has been destroyed or are in the oceans, so seeing if there are any correlations to known impacts on earth could give us an idea of the chances of a similar comet type impact as took place on Jupiter in 1994. The moon is like Antarctica and the meteorites found there, it has a fossil history of bombardments. I would venture to say that our moon has the most pristine record of impacts of any object in the solar system. Like they say ” You don’t know what you’ve got till it’s gone”.

        • Michael Fidler January 17, 2017, 9:18

          Latest #SETITalks: The Late Veneer and Earth’s habitability – Norm Sleep, Stanford University
          Asteroid impacts were a hazard to any life on the Hadean Earth. A traditional approach to geochemical models of the asteroid impactors uses the concentration of highly siderophile elements including the Pt-group in the silicate Earth. These elements occur in roughly chondritic relative ratios, but with absolute concentrations less than 1% chondrite. This veneer component implies addition of chondrite-like material with 0.3-0.7% mass of the Earth’s mantle or an equivalent planet-wide thickness of 5-20 km. The veneer thickness, 200-300 m, within the lunar crust and mantle is much less. The accretion of a large number of small bodies would provide comparable thicknesses to both bodies, as the effect of gravity is modest.
          Watch here: http://buff.ly/2jpzmSB

  • Michael Fidler January 15, 2017, 10:27

    One of the most fascinating impacts on the Moon is Mare Orientale, its beautiful structure was not well understood until the space age. The unusual aspect of this impact is that it is a bulls-eye on the forward facing orbit of the moon. Could this of been a Trojan moon at the L4 location that was slowly destabilized by solar gravitational influences?

  • Patient Observer January 16, 2017, 14:44

    Truly fascinating information. Just from memory, there was a paper years ago purporting the possibility that the impactor created two moons. Eventually, the smaller of the two made a relatively low velocity impact with the larger body on the far side of the moon. Rather than vaporizing or blasting away the lunar surface, the material simply distributed itself in some manner over the original surface. The paper suggested that this process would explain the significant difference in the topology of the near and far sides. Presumably this impact happened after the heavy bombardment that creates the lunar “seas” common on the near side.

    • Bruce Mayfield January 16, 2017, 21:01

      But there is no large mare or crater on the Moon’s far side. Even a “relatively low velocity impact” should have left a big crater if the impactor was big enough to have produced much of the topology on the far side.

      • Mark Zambelli February 15, 2017, 7:15

        The South-polar-Aitkin Basin… one of the largest basins in the solar system.

  • Geoffrey Hillend January 18, 2017, 18:06

    Quote by Ashley Baldwin: “An extension of this compared to that of Earth’s crust first supported the idea of an impact followed by co-accretion origin of the moon sometime between 50 and 100 million years after the formation of Earth .” This is the giant impact hypothesis. Our astronauts also found rocks that show the entire crust of the Moon was molten and chemically composed with the same material as Earth’s mantle which was entirely blasted off by the impact and re combined in space to form the Moon. It is made of anorthosite or plagioclase feldspar.

    I can’t agree with the idea that there was more than one body that combined to form our Moon for the following reasons: 1) The odds of that happening to too low. There was an early and late bombardment periods by meteoroids and asteroids in our solar system but these are mostly smaller bodies and there is more of those so they have a higher probability of collision. These were cleared out of the solar system or became planets which themselves were already the result of the accretion theory that shows that all rocky planets formed by collisions one at a time. In other words, two or more moon sized bodies colliding with the Earth at different times and not being separate orbits or recombining into one orbit seems highly improbable.

    2) These collisions would have to be very close or relatively close in time since the Roche limit tends to want to tidally tear an object apart because that is where the tidal forces are greater than the internal gravity which hold together the moon. The body has to outside of the Roche limit or near the end of it to stay together. The moon does not stay in one place but slowly moves away from Earth at 3.78 centimeters per year due to the transfer of rotational angular momentum of the Earth to the orbital momentum of the Moon. The tidal forces of the Moon on the sea and the Earth slow down the rotation of the Earth by 1 and a half millisecond per one hundred years so the day gets longer over time. By the law of conservation of angular momentum, the energy lost must be transferred to the orbital velocity of the Moon. Also the sea is a gravitating body and causes a force on the Moon and pushes it ahead in its orbit.

    I don’t see how several collisions can combine if the orbiting bodies being large don’t stay in close or exact orbits over time. They would have to re combine into a single object. We would have several moons with that idea and we don’t. It seems more likely that it was one Mars size body that broke up after the collision and recombined to form the Moon and then slowly moved away from the Earth over time.

  • Geoffrey Hillend January 18, 2017, 19:00

    The moons all have to have the same velocity, angle or the right mass for the shattered debris to end up in the same orbit. It’s not impossible but I think less probable.

  • ljk March 8, 2017, 12:04

    So apparently Earth can be and has been hit by giant space rocks and ice balls at any old time:


    The solution: Beefed up 24/7 vigilance. Same goes for SETI.