The origin of the Moon is a long-standing problem in planetary science. Reconstructing complex events in the distant past is difficult and requires both knowledge and imagination. The facts to be explained are relatively straightforward. The Moon’s overall density (about 3.3 grams per cubic centimeter) and bulk chemical composition are about the same as that of the mantle of the Earth, suggesting a possible relationship between the two. The idea that Earth and Moon are compositionally related is supported by the ratio of isotopes of oxygen in the lunar samples, which indicate that Earth and Moon are made from matter derived from the same region of the solar nebula (material that is compositionally distinct from that making up the various meteorite groups). Finally, the Earth and Moon collectively have a very high angular momentum, mostly as a consequence of the high spin rate of Earth and the relatively large mass of our Moon compared to its primary planet.
Prior to the Apollo missions, three different models (capture, fission, binary accretion) vied for acceptance among the lunar science community. The capture model proposed that the Moon formed elsewhere in the Solar System before a close, chance encounter resulted in the Earth capturing the Moon into orbit. The fission model proposed that a large mass of molten material spun off a rapidly spinning early Earth, was thrown into orbit and over time, coalesced into the Moon. The binary accretion model suggested that Earth and Moon assembled themselves independently as two distinct and separate bodies from the beginning. None of these models seemed able to account for all the “constraints” mentioned above, but no one had any better ideas.
About 30 years ago, the problem of lunar origin was widely considered “solved” with the general acceptance of the Giant Impact model. In this concept, four and a half billion years ago, the proto-Earth shared its orbit around the Sun with an object about the size of the planet Mars (dubbed Theia, in Greek mythology, the titan who gave birth to Selene, goddess of the Moon). A chance encounter between these two planetoids resulted in their merging as the Earth-Moon system. It was thought that a grazing (low angle) impact would serve to both spin up the Terra-Luna system, resulting in its relatively high angular momentum, and hurl vaporized mantle material from Theia into orbit around the Earth. The disk of orbiting debris quickly coalesced into the Moon and this rapid accumulation resulted in the release of large amounts of heat, which proceeded to melt at least the outer few hundred kilometers of the Moon, creating an “ocean” of molten rock, or magma.
The Giant Impact model seemed to nicely account for most of the properties of the Moon. But like many big ideas in science, the closer and longer we look at it, the more issues seem to arise. It was long assumed that the Moon was made of material derived mostly from mantle of the impacting planet (Theia); in this view, the Giant Impact was really just a variant of the capture model. As such, it did not explain either the chemical similarity of the Moon to the mantle of the Earth, nor their identical oxygen isotope compositions. This objection was usually brushed away with the admonition that complications might be expected from planet-scale impacts.
A new set of computer models has looked at the consequences of a slightly more head-on planetary collision. In contrast to the traditional oblique (few degrees) off-center Big Whack, researchers modeled the effects of an impact at about 30° incidence and relatively high velocity (about 1.3 times escape velocity, or roughly 14 km/sec). They find that in this case, most of the material from which the Moon forms comes not from the impactor Theia, but from the mantle of the Earth. This result might better explain the compositional attributes of the Earth-Moon system. In fact, several models were run (slightly varying these conditions) and while none perfectly fit the chemical and dynamical constraints, this one matched them most closely.
While this modeling was underway, another group was analyzing the composition of isotopes of titanium in samples from the Earth, the Moon and meteorites. The work has established that the chemical fingerprints that relate Earth and Moon are not merely close – they are virtually identical (to the best precision of the measurements). The authors of this study claim that this result creates problems for the Giant Impact model, as that idea had called for most of the Moon to be derived from the mantle of the impacting planet Theia. However, with the results of the new computer models of giant impacts discussed above demonstrating that the parameters of the collision can be adjusted to match the constraints on lunar origin, perhaps this is not such a problem for the Giant Impact model after all.
These developments should probably give lunar scientists pause. After all, the Giant Impact model became popular because the earlier, traditional three models (capture, fission, binary accretion) were all inadequate and their boundaries and defining parameters had to be adjusted to permit their (barely acceptable) viability. In other words, the models were stretched to fit any inconvenient facts or problem observations. Now it appears that the same thing is happening to the new, “explains-it-all” Giant Impact model. A scientific idea that can be stretched to fit any observable fact is not very useful as an explanatory principle – it is simply a glorified “Just So” story. The late Karl Popper argued that often in science, an idea cannot be shown to be true, but it can always be shown to be wrong – that is, “falsified.” If a hypothesis cannot be falsified, Popper argued, then it was not scientific. We need a mechanism in science to enable us to dismiss useless or irrelevant concepts and falsification is one way to do that.
So where does such philosophy leave the origin of the Moon? Perhaps more knowledge and imagination is needed before we can pronounce lunar genesis a “solved problem.”