skip to main content

The role of magnetic fields in the Moon's origin

Written by Jessica Raley

Since it formed roughly 4.5 billion years ago, the Moon has been Earth’s nearest neighbor and constant companion. Though it is the most familiar object in the night sky, the Moon’s origin remains in many ways mysterious. Researchers at the Illinois Center for Advanced Studies of the Universe (ICASU) are the first to examine the role of magnetic fields in the formation of Earth’s Moon, offering new insights into how and when the Moon may have formed.

Illinois Physics and Astronomy Professor Charles Gammie
Professor Charles Gammie

Illinois Astronomy graduate student Patrick Mullen and Illinois Physics and Astronomy Professor Charles Gammie introduced magnetic fields into computer simulations of the Moon’s origin. This research was inspired in part by the work of theorists specializing in neutron star mergers. Gammie says, “The collision that formed the Moon is physically analogous to the collision of neutron stars when they experience a merger driven by gravitational radiation. Two bodies with complicated, non-ideal equations of state collide off-center and then merge. Magnetic fields almost certainly govern the neutron star merger. We thought they might matter for the merger of Earth and its collision partner.”

Astronomy graduate student, Patrick Mullen
Astronomy graduate student, Patrick Mullen

According to Mullen, “For several decades, the leading hypothesis has been that the Moon formed shortly after the solar system itself came into being, while the planets were still in their final stages of formation. Many researchers agree that at this time, around 4.5 billion years ago, a planetary object roughly the size of Mars struck the proto-Earth at an oblique angle. The debris from this impact formed a disk around the proto-Earth and later coalesced into our Moon.” 


Scientists have studied several possible Moon-forming impact scenarios, but in any version of the giant impact hypothesis, the evolution of this debris disk—composed of liquid and vapor lunar rock—is not well understood. Mullen, who will complete his PhD next summer, says, “There is still much work that has to be done to bridge the gap between the giant impact and the formation of the Moon as it is today.”

Mullen notes, “Our simulations consider the collision between a magnetized proto-Earth and impactor. We find that magnetic field strengths are exponentially amplified by the giant impact. Our model suggests that the proto-Earth and the protolunar disk were strongly magnetized shortly after the giant impact. In fact, they might have had field strengths comparable to that inside an MRI machine. At that time, the magnetic fields were so strong that they were dynamically important, and they governed the evolution of the debris.”

Protolunar disk
The protolunar disk ~220 hours post-impact in a slice through the mid-plane.  Color denotes (log10) density; white lines represent magnetic field lines.  The field is disordered due to vigorous magnetized turbulence.  Field strengths are comparable to those inside a medical MRI machine.

This research offers a new explanation for how and on what time scale the material in the protolunar disk moved outside of the Roche limit, the radius beyond which tidal forces do not tear apart the material. Mullen explains, “In the prevailing hypothesis, the mechanism that makes material move outside this Roche limit is the gravitational instability. But in our simulation, the mechanism for transporting material outside this radius is magnetically driven turbulence.

“In this model, the disk spreading timescale is reduced from roughly 100 years to roughly 100 hours.”

Patrick Mullen, graduate student

“In this model, the disk spreading time scale is reduced from roughly 100 years to roughly 100 hours. The formation of the Moon may have happened earlier than researchers previously believed. At the same time, this magnetically driven turbulence also causes material to accrete onto the Earth. So, while the formation of the Moon is hastened in this model, it is also less efficient.”

The first ~48 hr in a magnetized, Moon-forming giant impact in a slice through the collision mid-plane. Color denotes (log10) density. At ~0-0.5 hours, strong shocks propagate through the colliding planetary bodies. At ~2-14 hours, a tidal arm of debris forms. Some mass falls back to the Earth while the remainder of the arm wraps around the proto-Earth, adding mass to the protolunar disk. At ~24-48 hours, the Earth, in distress, drives spiral waves into the protolunar disk.

Though the giant impact hypothesis is widely accepted, some scientists have discounted it, because previous models have failed to account for the similarities between samples from the Moon and the Earth.

According to Mullen, “Surface samples reveal that the Earth and Moon are anomalously similar in their isotope ratios. If you look at the oxygen isotopic ratio of the Earth and the Moon, you’ll see that they are almost identical. That’s strange, because we would expect to see a clear trace of the impactor in the lunar samples, but we do not.”

“ Previous simulations have been unable to demonstrate that there is enough mixing to reconcile geochemical analyses of lunar rock. By introducing magnetic turbulence, our model may offer an explanation for this mixing”

Patrick Mullen, graduate student

Because of the similarity between samples, Mullen explains, “The giant impact hypothesis relies on mixing between impactor and target-derived material. However, previous simulations have been unable to demonstrate that there is enough mixing to reconcile geochemical analyses of lunar rock. By introducing magnetic turbulence, our model may offer an explanation for this mixing.”

More work remains to be done in this area. Mullen says, “This is a puzzle that scientists have been trying to solve for many years. One of the big steppingstones was to find that magnetic fields do become dynamically important. But now we need to go back and find out how efficiently we can mix material to reconcile these lunar samples.”

These results have implications for potential future experiments when humans return to the Moon. Gammie comments, “This work strongly suggests that some of the rock that was incorporated into the Moon as it formed was heated to very high temperatures. This has implications for geochemists who want to understand the composition of lunar samples, and who are designing new experiments yet to be conducted.”

NASA’s Emerging Worlds program provided funding for this work. The simulations were done on supercomputers around the country, including Blue Waters at the National Center Supercomputing Applications (NCSA) at the University of Illinois, Stampede2 at the Texas Advanced Computing Center (TACC) at the University of Texas, and Pleaides at NASA. 

This research was published in the October 29, 2020 issue of The Astrophysical Journal Letters.