Skip to main content

Surprise! Earth and the moon aren't made of exactly the same stuff.

This composite image of the moon was constructed using data gathered by NASA's Clementine mission in 1994.
This composite image of the moon was constructed using data gathered by NASA's Clementine mission in 1994.
(Image: © NASA)

The moon and Earth may be more different than previously thought, challenging existing models for how the moon formed, a new study finds.

Earth originated about 4.5 billion years ago, and previous research suggested that the moon arose a short time after that. For the past three decades, the prevailing explanation for the moon's origin was that it resulted from the collision of two protoplanets, or embryonic worlds. One of those was the newborn Earth, and the other was a Mars-size rock nicknamed Theia, after the mother of the moon in Greek myth. "Once the dust settled, two bodies were left — Earth and the moon," new study co-author Zachary Sharp, a planetary scientist at the University of New Mexico in Albuquerque, told Space.com.

This "giant impact hypothesis" seemed to explain many details about Earth and the moon, such as the large size of the moon compared with Earth and the rates of rotation of the two bodies. However, in the past 20 or so years, evidence has emerged to challenge that hypothesis and suggest a multitude of alternatives.

Related: How the moon formed: 5 wild lunar theories

Computer models of the giant-impact scenario often say that 70% to 90% of the moon should be made of material from Theia. The problem is that most bodies in the solar system have unique chemical makeups, and so the Earth, Theia — and therefore the moon — should too. However, rock samples that the Apollo missions returned from the moon show that the natural satellite's composition is uncannily similar to Earth's, much more similar than such models would predict for versions of elements called isotopes. (Isotopes of an element each have different numbers of neutrons in their atomic nuclei.)

This extreme similarity in isotopes of elements such as oxygen has raised great challenges for the giant-impact scenario. One possibility is that the proto-Earth and Theia were nearly identical to start with when it came to oxygen isotopes, which seems unlikely. Another is that the proto-Earth and Theia's oxygen isotopes were fully mixed in the aftermath of the collision, perhaps due to an impact so violent that it vaporized a large portion of the early Earth, with the moon emerging from the resulting, doughnut-shaped mass called a synestia. But this and other scenarios may require unlikely impact conditions, scientists have said.

In the new study, researchers conducted new high-precision measurements of oxygen isotope levels in a range of lunar samples. The researchers expanded on previous work by focusing on a wide variety of types of moon rock.

The scientists found that there were subtle but regular differences in oxygen isotopic composition depending on the kind of lunar rock tested, Sharp said. This suggested that prior work that averaged together lunar isotope data while ignoring differences in rock type might not have given an accurate picture of the differences between Earth and the moon.

"Going into this project, it was expected that our results would likely mirror that of previous studies," study lead author Erick Cano, a stable-isotope geochemist at the University of New Mexico, told Space.com. "The most surprising part of our results was finding the amount of variation that we did between the individual lunar samples."

Related: How the moon was made: Lunar evolution explained (infographic)

To explain these findings, the researchers suggested that the giant collision between proto-Earth and Theia did indeed lead to mixing between the bodies. Still, the resulting moon and Earth had distinct compositions, albeit very similar ones, Sharp said.

Later, in the first 1,000 or so years after the impact, vaporized rock from the disk of debris left behind by the impact likely led "to lava raining down on the moon for hundreds of years," Sharp said. Complex physical and chemical interactions between this lava rain and the ocean of magma that covered the newborn moon could then have led to an oxygen isotopic composition in the uppermost lunar rocks that was more similar to Earth's. In contrast, samples that came from the deep lunar mantle had the most different oxygen isotopic composition of the lunar rocks tested when compared to Earth.

The most important implication from these findings is that giant-impact models no longer have to account for virtually indistinguishable oxygen isotopic compositions between Earth and the moon, Cano said. "I think this will open the door for an entirely new range of impact scenarios," he added.

Future research can expand on this new study by analyzing other lunar samples, Cano said. "The obstacles for this future research may be the limited quantities of material that we have from the Apollo missions," he said. "Some of these lunar rock types were only brought back in very small quantities and can be very difficult to obtain for study."

Cano, Sharp and study co-author Chip Shearer, a lunar scientist also based at the University of New Mexico, detailed their findings online Monday (March 9) in the journal Nature Geoscience.

Follow Charles Q. Choi on Twitter @cqchoi. Follow us on Twitter @Spacedotcom and on Facebook.

OFFER: Save at least 56% with our latest magazine deal!

All About Space magazine takes you on an awe-inspiring journey through our solar system and beyond, from the amazing technology and spacecraft that enables humanity to venture into orbit, to the complexities of space science.View Deal

Join our Space Forums to keep talking space on the latest missions, night sky and more! And if you have a news tip, correction or comment, let us know at: community@space.com.

  • rod
    Admin said:
    The moon and Earth may be more different than long thought, challenging existing models for how the moon formed, a new study finds.

    Surprise! Earth and the moon aren't made of exactly the same stuff. : Read more

    Other reports are out on this topic too. Apollo Rock Samples Heat Up Moon Formation Debate, "A new study suggests there are key differences between the compositions of Earth and its natural satellite, with significant implications for lunar history" "A new study published in Nature Geoscience may resolve some of the issues. Erick Cano of the University of New Mexico and his colleagues examined samples of the lunar surface collected by the Apollo missions and found that the deeper under that surface you go, the more different the moon looks from Earth. This result suggests that the moon and our planet are not as identical in composition as once thought,...The new results, though, show there is still much to be learned about the moon’s composition, and it may be some time before scientists can agree on a single theory as to how the satellite formed. “A lot of people are really interested in getting to know how it was made,” Thiemens says. But if a smoking gun for Theia does exist beneath the surface, it could help us finally work out where this impactor came from and how it led to the creation of our celestial neighbor."

    Okay, using my telescopes, I never observed Theia :). Theia is a product of computer models using different masses, composition and densities for the protoplanetary disk and accretion rate(s). There are plenty of reports showing very different disk masses and dust disk masses, scattered all around in stars today that make Theia and giant impact modeling for the origin of the Moon, challenging.
    Reply
  • Torbjorn Larsson
    Admin said:
    The moon and Earth may be more different than long thought, challenging existing models for how the moon formed, a new study finds.
    Surprise! Earth and the moon aren't made of exactly the same stuff. : Read more

    This opening up of the parameter space for the Moon forming impactor is interesting in the context of the proposed ages of Earth and Mars, since Earth may have formed quickly by pebble showers at a system age of ~ 5 Myrs https://www.space.com/meteorite-iron-shows-earth-formed-fast.html ] while Mars growth was cut off at ~ 10 Myrs age by the gas giant migration https://www.space.com/early-mars-formed-slow-ancient-collisions-show.html , https://www.sciencemag.org/news/2020/01/cataclysmic-bashing-giant-planets-occurred-early-our-solar-systems-history ]. The new result would not immediately tell us the mass of the impactor, but its late arrival at ~50 Myrs system age may be caused by migration from far out instead of being Mars massed and related to Mars growth zone.

    The migration phase itself takes ~ 10 Myrs to reformat the disk https://arxiv.org/pdf/1912.10879.pdf ]. So reasonably we would see something like 30 Myrs of system age for an impactor migrating inwards reaching the inner system. That seems close enough in terms of back-of-the-envelope estimates. So the new result is not only interesting but promising in better matches between different observations, perhaps even naturally resolving some of the residual tensions. The amount of finetuning is multiply lowered by the new find. I peeked at the figures (paywalled paper) and the result - if not the model - seemed pretty forwardly tied to identifying types of mineral grains and their origination depth within the Moon.
    Reply
  • CSirolli
    Theorize that a giant body collided with earth, where the debris managed to consolidate into what we now call the moon, and just so happened to be placed in orbit at just the right position where even after billions of years of it moving away it just so happens that it is positioned so that we can see full eclipses. And it consolidated into the perfect size too for us to see those eclipses as well.

    Also, after billions of years, the moon should be further out than it is, but if we assume it is only 6000 years old then the moon would have barely moved at all, so there is no issue with its position.

    When considering these two theories, keep in mind Occam's razor.
    Reply
  • rod
    Here is an observation I have. The original composition of Theia, the proto-earth, and proto-moon before and after the giant impact event used in computer models to create the Moon we see today - seems very difficult to define and test as this new report on oxygen isotopes demonstrates that offers some hope perhaps. We have a very real problem with the Earth-Moon system and that is the angular momentum problem. "...But scientists conclude the Moon could not have been receding at this rate throughout its history, because projecting its progress linearly back in time would put the Moon inside the Earth only 1.4 billion years ago.", Ancient shell shows days were half-hour shorter 70 million years ago The angular momentum problem has been known since at least the 1960s, before Apollo missions. Various calculations are performed to solve the issue including looking for fossil record evidence to show Earth had a shorter day during various geologic ages. An 18 hour day was reported for Earth, 1.4 billion years ago, Thank the moon for Earth's lengthening day
    What was the original length of day for the proto-earth after Theia hit it? That is much fun to study and attempt to show the evolution of the Moon's orbit over some 4.5 billion years or more. New theory explains how the moon got there
    The giant impact event with Theia, the proto-moon forms near 3 earth radii compared to the present mean some 60.27 earth radii today. The proto-moon and evolving Moon over long time integration, has a very different orbital period too compared to the present.
    Reply
  • Torbjorn Larsson
    rod said:
    The original composition of Theia, the proto-earth, and proto-moon before and after the giant impact event used in computer models to create the Moon we see today - seems very difficult to define and test as this new report on oxygen isotopes demonstrates that offers some hope perhaps. We have a very real problem with the Earth-Moon system and that is the angular momentum problem.

    The giant impact hypothesis was partly - mostly, perhaps - inspired by that it can replicate the angular momentum of the two orbiting bodies. I assume the other large impact binary of Pluto and Charon does the same. Tidal forces are - somewhat unpredictable, c.f. the problems of solving for Enceladus global ocean - responsible for the historical slowing, so conversely I don't think they are considered part of the problem. (Until you want to constrain impact models.)

    It would be interesting if someone takes one or more - or preferably tries all - of the newer study results and tries to make an impact model. Pre-impact Earth could be iron core from initial accretion with chondrite mantle from a pebble rain. The impact body could be a Kuiper Belt Object for all I know, modeled on Triton perhaps (but I haven't read the paper) https://en.wikipedia.org/wiki/Triton_(moon) ] - I like the timing for that, and Triton showed the necessary migration happened at least once. Or perhaps a shed gas giant moon, modeled on Titan perhaps https://en.wikipedia.org/wiki/Titan_(moon) ]. Or perhaps another "happened at least once", a Ceres analog of late planetesimal migration https://en.wikipedia.org/wiki/Ceres_(dwarf_planet)#Origin_and_evolution ].
    Reply
  • rod
    Torbjorn Larsson said:
    The giant impact hypothesis was partly - mostly, perhaps - inspired by that it can replicate the angular momentum of the two orbiting bodies. I assume the other large impact binary of Pluto and Charon does the same. Tidal forces are - somewhat unpredictable, c.f. the problems of solving for Enceladus global ocean - responsible for the historical slowing, so conversely I don't think they are considered part of the problem. (Until you want to constrain impact models.)

    It would be interesting if someone takes one or more - or preferably tries all - of the newer study results and tries to make an impact model. Pre-impact Earth could be iron core from initial accretion with chondrite mantle from a pebble rain. The impact body could be a Kuiper Belt Object for all I know, modeled on Triton perhaps (but I haven't read the paper) https://en.wikipedia.org/wiki/Triton_(moon) ] - I like the timing for that, and Triton showed the necessary migration happened at least once. Or perhaps a shed gas giant moon, modeled on Titan perhaps https://en.wikipedia.org/wiki/Titan_(moon) ]. Or perhaps another "happened at least once", a Ceres analog of late planetesimal migration https://en.wikipedia.org/wiki/Ceres_(dwarf_planet)#Origin_and_evolution ].

    "The giant impact hypothesis was partly - mostly, perhaps - inspired by that it can replicate the angular momentum of the two orbiting bodies."

    My observation. After the Apollo missions, in 1975 the giant impact model was proposed or at least more widely studied because of the known issue with angular momentum. In order to replicate the angular momentum of the two bodies, a proto-earth and proto-moon initial rotation rate is needed and then continued to evolve via some type of accretion into present day masses and orbits, no easy job. Ancient eclipses and the Earth's rotation
    Using ancient eclipse records from Assyria and Babylon, there is some 2800 years of astronomical observation including telescope measured total solar eclipses documented (since George Darwin in 1880s), to model the tidal dissipation rate parameter for the present angular momentum of the Earth and Moon system and rate of lunar recession as well as Apollo lunar laser ranging experiments. Establishing the initial angular momentum of the system is the big problem and verifying this. Different giant impact models result in different initial spin rates for the proto-earth when the Moon evolved. Some claim a 2 hour day, some 5 hour day, and before the giant impact event, a very slow rotation for early Earth. The problem is widespread in the solar system using accretion and giant impacts, including Mars rotation and spin period today observed today vs. what a proto-mars may have had. So here is a problem. There is about 2800 years of documented measurement for the tidal dissipation rate parameter today, extrapolating back billions of years runs into problems. The giant impact model offers so hope here. However, the pre-impact angular momentum and post-impact angular momentum and evolution, still limited in defining constraints and testing in the model - my opinion.
    Reply
  • Torbjorn Larsson
    rod said:
    The giant impact model offers so hope here.

    "No" hope (s & n not QWERTY adjacent) or "So much" hope!?

    I agree with the latter, since it is the consensus model. By the way, Mars modeling also need/are undergoing a do over, since the moons' orbital dynamics implies they formed from ejecta as well. (And there are 1 or more equatorial very flat infall craters that fit that too.)
    Reply