Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Cataclysmic collision of proto-earth and planetisimal

  1. Sep 3, 2009 #1
    For cataclysmic collision of proto-earth and planetesimal, how long would it take for the liquid cores to combine after collision of 2 bodies' surfaces; how fast would proto-earth's surface radially expand?
  2. jcsd
  3. Sep 3, 2009 #2


    User Avatar

    Impact of the proto-earth and moon likely resulted in both objects becoming essentially molten within an hour. The metallic cores combined quickly, but it probably took several weeks for most ejected material to condense into the moon or collide with earth. Notice that the density of the moon is similar to the earths crust and mantle and it has no metallic core.

    Meanwhile, the earths surface was completely molten for an extended period (millions of years) and the original atmosphere lost.
  4. Sep 3, 2009 #3
    There are some concerns about that hypothesis.

  5. Sep 3, 2009 #4


    User Avatar
    Science Advisor

    You need to read that paper again. You appear to have completely misunderstood what it is about. The models to which the extract you have quoted refer are models for how material coalesces after the impact; NOT alternatives to the notion of the giant impact.

    The title is "Isotopic Equilibration of Earth’s Mantle and the Moon subsequent to the Giant Impact?", by A. Zindler and S. B. Jacobsen.

    Even the title should give the clue that this is not expressing any doubts about the giant impact model for formation of the Moon. The paper does not present or even consider any other model; it is taken for granted throughout as the working basis for the study. The question is about how isotopic equilibration arose. In particular, the question is whether the Moon is formed mostly from material of the impactor, or from the mantle of the proto-Earth thrown up in the impact. The paper argues for the latter conclusion.

    Here's an extract from the article giving the proposal that the paper is criticizing. Quoted text in blue; bold as in the article itself:
    A Novel Approach: Pahlevan and Stevenson [8] explored a very different potential solution to this problem: that the Earth and protolunar disk, largely molten but isotopically dissimilar in the immediate aftermath of the giant impact, were able to achieve oxygen isotopic equilibrium via exchange of oxygen through the shared, hot, dense, silicate vapor atmosphere that prevailed for a short time between the impact and lunar accretion [5].

    The paper goes on to consider whether this exchange is able to achieve equilibrium. They argue that it is not.

    So in brief, carry on. Whatever concerns you might have Andre, they don't appear to be the same as those in the paper. This becomes clear when reading the parts of the paper that actually speak about what the concerns are and how they are addressed.

    The paper focuses on a consideration of what happens to the disk of molten material from which the Moon formed, and how isotopic similarities arise. The the final paragraph gives the paper's conclusion:
    We conclude, therefore, that the liquid-vapour exchange subsequent to the giant impact does not represent a compelling explanation for the isotopic similarities of the Earth and Moon in the cocntext of current lunar formation theories. [...] Ringwood's [11] assertion that the Moon formed from the Earth's mantle, remains a difficult hypothesis to reject out of hand.

    Xnn's comment describes the likely time scales of events from this impact.

    Cheers -- sylas
    Last edited: Sep 3, 2009
  6. Sep 3, 2009 #5


    User Avatar

    Yeah; I don't see much in the paper that out right contradicts my original post; however it does discuss a number of interesting details that are being worked out.

    First; SPH models. Had to look this up; it refers to Smoothed Particle Hydrodynamics. Obviously a class of computer models used to study the formation of planets and moons. I believe these are the models that are in need of significant revision.

    Second; Apparently there are problems with assuming that the impactor formed at the exact same distance from the sun as the earth is. However, the paper doesn't state what the problems are or how much further or closer to the sun the impactor may have formed. IMO, the paper may be splitting hairs as it would only make sense for the impactor to have formed somewhere between Venus and Mars.

    Third; Not only did the two bodies become molten, but there may have been significant vaporization of rocks (silicates) which existed as a shared atmosphere for a while. However, the paper concludes that this does not represent a compelling explanation for the similarities between composition of the earth crust and mantel and the moon. So, while it's an interesting point, I don't see much that can be concluded about this aspect.

    Lastly, getting back to the original question about how long it took the two cores to combine. A simple calculation shows that the moon currently travels thru an earth diameter in about 4 hours. The impactor was probably traveling at a greater velocity than the moon is currently traveling. So, the original impact probably occurred within this time scale. However, since the moon is so much smaller than the earth, it's not clear to me that the impactor necessarily had much of a core to begin with. In other words, the impact involved the crust and mantle of the earth, but had little to do with the core. Never the less, it provided enough additional energy to melt the surfaces which included vaporization of some rocky materials.

    After the impact, there was a large mass that settled into an orbit similar to that of the moon. There were numerous subsequent collisions with other ejected material for many years, but the predominate mass should have been discernible after the first revolution (a month).
  7. Sep 4, 2009 #6
    Right I thought already that one of us needed to read something again and that one of us completely misunderstood what it is about. And I knew it wasn't me :tongue:

    concern - noun

    or does anyone see perhaps...
    One might want to be careful with fallacies like strawman tactics. Once exposed, those tend to turn against the originator in an intelligent society like this.
    Last edited: Sep 4, 2009
  8. Sep 4, 2009 #7


    User Avatar
    Science Advisor

    In the context of this discussion, to express "concern" about "that hypothesis" would lead a normal person to think you mean there is some reason to be skeptical of the hypothesis actually being discussed, of the giant impact itself. It remains the case that based on what you had written that you truly do appear to have completely misunderstood the paper. Which is what I said.

    I regret to say that I am no longer willing to simply assume good faith here by default; but I do grant that the "appearance" of misunderstanding might be itself misleading. Perhaps you actually did understand what the paper was about, and chose for reasons unknown to distract the thread with this irrelevant side issue to the discussion, expressed in minimal terms that would normally suggest some kind of actual difficulty with the hypothesis of a giant impact and the merging of proto-Earth and an impacting planetesimal, as described in the thread. If anyone thinks this possibility makes ME look bad, I am amused, but I don't really care very much.

    The main point I want to make is a clarification about the content of the paper by Zindler and Jacobsen linked in msg #3.

    • The paper doesn't actually say anything about how long it took for the cores of the impactor and the proto-Earth to merge.
    • The paper doesn't give any reason to doubt the hypothesis that the Moon was formed from a giant impact between proto-Earth and another large planetesimal.
    • The paper argues that the Moon is mostly formed by material thrown up from the proto-Earth's mantle in the impact.
    • The "current models in need of revision" mentioned in Andre's quoted extract is a reference to models of how the material from the impact distributes between Earth and Moon.

    It appeared to me from your post that you might not have understood these points. But maybe you did. In either case, I still think this is a relevant clarification, given the thread topic. Andre; you will go up in my estimation substantially if you can agree with these points as relevant to the actual question of the thread.

    Cheers -- sylas
    Last edited: Sep 4, 2009
  9. Sep 4, 2009 #8


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Hmm...were there any posts deleted in this thread?
  10. Sep 5, 2009 #9


    User Avatar

    Well that was cheery.

    Ease up guys; can't we all be friends?

    As an outsider, I see two things in this paper and have a few questions:

    one must conclude either that significant aspects of current models are in need of revision, or attribute important aspects of the Earth-Moon system to a rather large coincidence.​

    Ringwood’s [11] assertion that the Moon formed from the Earth’s mantle, remains a difficult hypothesis to reject out of hand.​

    ...What? First you say the models are horrible, then you say the hypothesis is "difficult to reject?" This just doesn't make sense.

    Second, this stuff looks like horrible speculation to me.

    You wanna know what I thought when I was reading this explanation of how the Moon came to be?: Did you just pull that out of your ***? The certainty of knowledge expressed is so fantastical. It took an hour to become molten, and then it was molten for millions of years? You don't say!

    Boy you scientists are smart! It's like you know everything. I do not mean to attack you personally at all Xnn.

    I mean, WHERE did that idea come from? Is this supposed to be science we're talking about? Let's think back to what science is:

    How TESTABLE is this?
    Is this FALSIFIABLE?

    This "giant impact hypothesis" is almost entirely in the realm of speculation. During planetary formation the moon could have just been floating around and Earth held it in orbit. Maybe the moon just happened to form next to the Earth. The time scale we are talking about is so massive and beyond human understanding that we know there are huge factors that we couldn't even imagine or possibly ever know. For instance, this "impactor"— we seem to will it into being. Where did it come from? Where did it go?

    This hypothesis is crazy. The entire earth was a magma ocean for millions of years? Is there any evidence of that?

    Last edited: Sep 5, 2009
  11. Sep 5, 2009 #10


    User Avatar

    Hi MK;

    Yeah; you're correct about the magma ocean on earth.
    Not much evidence that it was global. My mistake.
    I could go back and edit that post, but will leave it be for all to see the blunder.
    Still, the collision had to have taken place within an hour and there would have been lots of molten or vaporized material as a result.

    Anyhow, a giant impact is still the leading hypothesis. The capture hypothesis that you mentioned however has serious problems. Namely, what would cause the moon to slow down enough to be captured? The earth would need a huge extended atmosphere or there were tidal forces strong enough to nearly rip the earth apart. Also, the tilt of the earths axis points to something less benign occurring it the past.

    More about the impactor: I don't understand why that paper has a problem with it forming at 1 AU from the sun. I would think it formed in about the same orbit as earth at a Lagrange point. Eventually, as it grew larger and larger (from accretion), its orbit would have become unstable and it would have been gravitationally attracted to earth.

    Research work is being done to "prove" this hypothesis by developing physic based computer models that are consistent with observations, but I don't believe there are any other hypothesis that are being seriously considered by the experts right now with the possible exception that there were obviously many impacts over time that were sufficient to throw material into orbit around the earth. So, maybe there were several "big" impacts.
  12. Sep 5, 2009 #11


    User Avatar
    Science Advisor

    It makes sense just fine if you read the paper more thoroughly, in my opinion; and that was the intent of my original comment. I don't mean any insult by that, by the way. It's a serious recommendation, given in all honesty.

    The phrase "significant aspects of current models" in the quoted extract refers to the models of how material is thrown up and coalesces in the giant impact. It is not a reference to the very idea of a giant impact itself. Once this is recognized -- and it is quite clear in the paper itself -- the perception of any conflict between these quotes is gone.

    This, by the way, is a good reflection of the state of knowledge in science for origins of the Moon. Science never gives final proof to put conclusions about the real world forever beyond question. But we still get pretty darned certain about some things. And the formation of the Moon from a giant impact is by now pretty much stands alone as the only model able to account for all the evidence available. Other proposals have fallen by the wayside, as inconsistent with the available evidence. All the argument in the paper takes the impact for granted.

    The paper is defending the idea that the Moon is formed mostly out of material from the proto-Earth's mantle, as proposed by Ringwood. This is contrasted with other models that suggest about 80% of the material of the Moon is from the impactor. The paper argues that those models will need to be revised.

    The paper takes the impact pretty much for granted. It does describe it as a "hypothesis" -- the leading hypothesis for formation of the Moon. Nothing in the paper at all is presenting any reason to doubt that leading hypothesis, and no mention is made of alternatives, or of any other way the Moon might have formed other than this impact. The whole question in the paper is about the scientific models for what happens to colliding planets. How material is impacted and distributed between the impactor and proto-Earth; and the newly formed Moon and the resulting Earth including the impactor.

    Now I gather you personally have some issues with the giant impact model as science. But the cited paper has no such concerns. It applies conventional scientific method to the formation of the Moon in the same way that scientists do for any study of events in the past.

    Our scientific theories about past events are tested in the light of the traces left behind that we can observe today. The theories can be falsified, like any other scientific theory, by repeatable observations -- in this case, repeatable observations of the evidence left behind. We don't need to see the event first hand to take a scientific approach to sorting out those events... this is the whole basis of forensic science also.

    It is simply incorrect to think that science includes any requirement for the events of study themselves to be repeatable.

    This is getting a bit into philosophy of science. When the phrase "repeatable" is used in science, it means repeatable observations; always. People study the isotopic composition of the Earth and the Moon. We can repeatably observe and measure isotopic compositions. That evidence is potentially able to falsify ideas about the formation of the Moon and the Earth. If anyone doubts the evidence itself, they also can go out and collect data and measurements for themselves, as a check.

    Events of the past... ice ages, drifting continents, super volcanoes, bolide impacts, formation of the solar system, and much else besides, are perfectly legitimate as objects of scientific study. There's no requirement that the event being studied and explained must be repeated. It's not even true in general that observations must be completely repeatable; philosophy of science does get a bit more subtle than this; but for formation of the Moon this is not an issue. All the evidence we actually use in this case is available for anyone to go out and measure or collect again, independently, repeatably, as often as you like. (Getting more Moon rocks takes some doing; but you can still repeat tests on material still available from the expeditions of the seventies.)

    Cheers -- sylas

    PS. Lisa, I don't think any posts have been deleted. Everything I've seen seems to be still here.
    Last edited: Sep 5, 2009
  13. Sep 5, 2009 #12


    User Avatar
    Science Advisor

    As I understand it, the paper starts out by suggesting that "current models" propose that material going into the Earth samples a large region of the early nebula. The "problem" with an impactor forming at 1 AU from the Sun would (in this case) be that it becomes a coincidence. You are basically saying that the material of the impactor would sample a much narrower region... and this would count as a "revision of current models", in the parlance of the paper; and it would help resolve the problems considered. Here's the extract from the paper which I read as making this argument:

    However, current models suggest that terrestrial planet formation culminates with a period of major impacts between growing planets and planetary embryos, thought to sample a large radial zone of the nebula extending to beyond the radius of Mars [6, 7]. The giant impactor that formed the Moon, therefore, is unlikely to have originated at
    one AU, or to have had isotopic characteristics indistinguishable from the proto-Earth.

    If it is presumed that there's no good reason for the impactor and proto-Earth to have similar isotopic composition, then isotopic similarities might be explained by having the Moon formed mainly from material thrown up from proto-Earth. This, however, runs into problems with current "SPH" models (smooth particle hydrodynamics) which would have the Moon formed 80% from the material of the impactor.

    Suggestions that the Moon formed from material ejected from the Earth’s mantle by the impactor [11], or from mass-relative proportions of Earth and impactor, are incompatible with SPH models which overwhelmingly predict that 80% or more of the protolunar material originates from the impactor [see 5 and references therein].

    Here also, are models that might need to be revised (according to the paper).

    In view of these considerations, one must conclude either that significant aspects of current models are in need of revision, or attribute important aspects of the Earth-Moon system to a rather large coincidence.

    The three extracts I have given above are continguous in the introduction to the paper. Hence, the "aspects of models that are in need of revision", according to the paper, are either for the range of the planetary nebula that is sampled to make up the impactor and proto-Earth, or the SPH models for how the impact occurs.

    By the way... the giant impact model involves a global magma ocean on the Moon; not on the Earth, since it has the Moon formed from molten material thrown up in the impact. This makes sense of a number of features of the Moon and its composition.

    Cheers -- sylas
  14. Sep 5, 2009 #13


    User Avatar


    I appreciate your comments, but I still don't see what's wrong with the impactor initially accreting at a Lagrange point. That is, having it formed from basically the same material as earth. In addition, as the earth is accreting at this time, there were many smaller collisions that also threw material into space. Some of this material then falls to the impactor. Likewise, as the impactor is also accreting from collisions, it is throwing material into space that lands on earth. Since both the earth and impactor are sweeping areas near 1 AU, as long as most impacts are small, they will assume nearly identical materials on their surfaces.

    Over time the impactors orbit become unstable and it eventually collides with earth. However, what is different is that the impactor has enough momentum that some of it along with material ejected from earth settles into an orbit around the earth. Of course accretion is not over and both the earth and moon continue to gain mass on their surface from many other smaller impacts, but it is all essentially the same material that is being swept from 1 AU. The only difference being gravitational strengths which allows earth to retain gaseous matter.
  15. Sep 5, 2009 #14


    User Avatar
    Science Advisor

    I don't see what's wrong with the impact accreting at a Lagrange point either (or what's right with it). This is not a topic I know a great deal about.

    I'm not in any kind of position to judge between competing ideas on the source of material accreting into planetesimals, or where they came from. I would think that the Lagrange point is a consequence of the location of the large mass of the Earth, and so I am not even confident that there would be a well defined Lagrange point when planetesimals were accreting.

    I am only saying what the paper sees as the problem, since I had taken that to be the question. The paper says that "current models" indicate that the Earth and the impactor would be taking in material from much more than simply the general region of 1 AU; in which case it would be rather a co-incidence for them to have very similar isotopic compositions. This is seen in the extract I quoted.

    You are saying instead that the Earth and impactor would be taking in material from around 1 AU. So it looks to me that what you are proposing might correspond to one of the ways in which "current models" could be modified to account for isotopic similarities.

    To go into any further detail on where material for planetesimals originates, one would have to go beyond this paper. They cited references 6 and 7 (both by a J.E. Chambers) for this matter. You probably know more about it than I do.

    Cheers -- sylas
  16. Sep 5, 2009 #15
  17. Sep 5, 2009 #16


    User Avatar
    Science Advisor

    I had a look. There's a lot of detail in that reference; very useful. (Title: "Where Did The Moon Come From?", by Edward Belbruno and J. Richard Gott III) And in particular it speaks directly to my initial skepticism of the existence of a stable Lagrange point during the accumulation phase.

    If folks were interested in comparing with the models used in the Zindler and Jacobsen, then it would also be worth looking at their reference number 7:
    • Cambers, J.E. (2004) http://www.geosc.psu.edu/~kasting/Meteo_466/Readings/Chambers_EPSL_04.pdf [Broken] in Earth and Planetary Science Letters Vol 223 pp 241–252.

    The two papers may even be consistent; I don't know. The Chambers references goes right through all the various stages of the accumulation of material from the initial nebula, and it might all be entirely consistent with simultaneous accumulation of Theia (I didn't know the impactor had been given a name; thanks!) and Earth both at about 1 AU but each one drawing in material from much further afield.

    Cheers -- sylas
    Last edited by a moderator: May 4, 2017
  18. Sep 24, 2009 #17
    I think some interesting data you can find in two articles:

    George Q. Chen, Thomas J. Ahrens, Erosion of terrestrial planet atmosphere by surface motion after a large impact, Physics of the Earth and Planetary Interiors, 100 (1997) 21-26; Information about surface velocity on p.23, Fig.2 and Fig.3.

    Jach, K., Leliwa-Kopystynski, J., Mroczkowski, M.,
    Swierczynski, R. and Wolanski, P., 1994. Free particle modeling
    of hypervelocity asteroid collisions with the Earth.
    Planet. Space Sci., 42: 1123-1137. Here you are the information (p.1133, Fig.9) about motion of surface in the antipodal point.

    Information about deformation, pressure and temperature eg.
    of outer and inner core you can find in my Ph. D. dissertation (1994) or in our monograph from 2001 (but both in Polish).

  19. Sep 28, 2009 #18
    On the subject of the inclination of Earth's axis - does anyone know of models (built upon current orbital and axial parameters of the Earth-Moon system) which show that Earth's inclination was caused by a collision?
  20. Jan 29, 2010 #19


    User Avatar

    Hey! An article came out two days ago in the MIT Technology Review. Meijer and Westrenen propose an alternative hypothesis for the origin of the Moon, they propose a... nuclear explosion.
    R.J. de Meijer & W. van Westrenen, "An alternative hypothesis for the origin of the Moon":
    Also, thanks for the post sylas. :smile:
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook