Talk:Planetary formation and migration

It is important, also in a popular review such as this one, to maintain focus on the many ways in which current scenarios and theories fail to provide satisfactory explanations for many aspects of planet formation. The situation appears to often be characterized by comparisons of two (or more) scenarios, where the inadequacies of one is taken as evidence (or even "proof") in favor of another, while the possibility that none of them is correct is not considered seriously enough.

Or, a variant of this approach is to argue that "since we are here -- terrestrial, gas and ice planets and all, one of the considered scenarios must have worked, and since I can show it wasn't scenario A, it must have been B!". In the case of the Moon this is applied to 3-4 scenarios, but the method is the same; "since scenarios A, B, and C cannot have worked, it must be D".

Both of these fallacies preclude searching for new scenarios, which is perhaps what needs to be done. In many other branches of science one would conclude from evidence such as those at hand here that none of the existing scenarios is satisfactory, and the search would be on for a new (and correct!) one. That situation is likely to arise also with respect to planet formation, as new evidence (from exoplanet observatios, radionucleid studies of the solar system, etc.) keeps coming in.

This (the above) is criticism of the research area, more than of the current author, and when revising the text I have attempted to remove or alter text that either directly follows such a line of approach, or else refers to "widely held beliefs" that often have come about in more or less this manner (the "Big Whack" scenario for the creation of the Moon is a highly relevant example).

If for no other reason than to encourage young readers to be skeptical, and in the hope that some of them are even inspired to join the effort, it is important in this popular text to emphasize the "open" state of affairs, and to thus encourage creativeness, inventiveness, and new ideas. To gloss over the uncertainties and impossibilities of currently popular views is not in the best interest of science, or even in the best interest of individual, "innocent", public readers.

Author's response

I certainly agree that there are plenty of uncertainties in all aspects of planet formation, and that an article should reflect both the current best-guess ideas and also open questions. I've folded in most of the suggested changes proposed by Reviewer A (though with slightly more muted language). Specifically:

Planetesimal formation - added explicit discussion of the well-known issues with both of the suggested models for planetesimal formation, namely the sticking efficiency issue for pairwise collisions and the need for a grossly enhanced dust / gas ratio for Goldreich-Ward.

Terrestrial planet formation - added a time scale estimate for runaway growth (from Kokubo and Ida 1996, noting that their figures use scaled times), and the discussion of cosmochemical constraints (thanks... not mentioning that was an omission reflecting my astrophysical background!). For the Moon, my belief is that the largest problem for giant impacts is the composition similarity (I admit I haven't yet studied Pahlevan and Stevenson's proposed solution), so I added that as an example of what remains unknown. I also removed mention of there being an eccentricity problem in terrestrial planet formation - I heard Alessandro Morbidelli talk about this just this week and I'm largely persuaded that the too large eccentricities found previously were just because the earlier work didn't have enough small bodies. I also noted that because of the stochastic nature of accretion, comparison between the intricacies of the inner Solar System and simulations can only ever take us so far.

Giant planet formation - added the time scale for slow growth, and also a discussion of why this is poorly known (opacity issues in the envelope). Noted that the extrasolar planet correlation is only really a qualitative agreement with theory. Noted that the atmospheric composition of Jupiter presents problems also for core accretion (though my opinion is that this can of worms - much like the question of volatile delivery to Earth - is a first-order problem whereas the core mass is a zero-eth order challenge).

Reviewer A comments

Some of the language had become too "muted". Two examples:

1) The 'meter size barrier' for planetesimal formation was dismissed by simply stating that "The required rapid growth may occur as a result of ongoing pairwise collisions", while mentioning as a "substantial challenge" only that "For pairwise collisions to work fast enough, meter sized objects need to efficiently stick together upon collision rather than breaking up" -- a tautology (for this to work it must work), commented with "This has not been demonstrated". With the reference to human understandable speeds (car speeds) removed, the reader would have no way to understand how unlikely such a demonstration appears.

2) The discussion of the formation of the Moon used the formulation "requiring such a cataclysmic event to form the Moon is not a problem", followed by the mentioning of only one of several independently unlikely circumstances this requires: that only a small fraction of impact parameters leads to an orbiting moon. Another independently unlikely outcome -- the similarity of the compositions was mentioned as "One puzzle", rather than "Another puzzle". It is necessary, to be fair at this point, to mention that at least three (independently) unlikely choices of the impact parameter must be made to explain the Moon in this scenario; 1) only a small fraction of impact parameters leads to an orbiting solution for the ejecta, 2) only a small fraction of those solutions give the required similarity of compositions, and 3) only a small fraction of the solutions fulfilling 1 and 2 have (prograde) orbits as close to the ecliptic as that of the moon.

[ for some reason the last part of this comment was chopped off -- may have been lost in a copy & paste operation -- now resurrected ]

With the suggested changes in the main text, or changes to the same effect, the text is acceptable, and I'm therefore tentatively indicating it as such.