Suggested Searches

Season 3, Episode 4: Why Do We Have a Moon? With Robin Canup

Season 3Episode 4May 9, 2019

Learn about how the Moon formed in this conversation with Robin Canup of the Southwest Research Institute.

Gravity Assist Logo

Learn about how the Moon formed in this conversation with Robin Canup of the Southwest Research Institute.

artist's rendition of what scientists think created the Earth-Moon system

Jim Green:One of the fantastic things about the Earth is that it has the Moon. The Moon is so important to us. Why do we even have a Moon? Let’s find out how we got it. I’m Jim Green, NASA’s chief scientist, and we’re here on “Gravity Assist” to talk about the Moon.

I’m here with Dr. Robin Canup, one of the top solar system dynamicists. And what a dynamicist does is really study the motion of things, how they collide, and come together. Robin also manages a major group of planetary scientists at Southwest Research Institute. And today we’re going to talk about the latest theories on the formation of the Moon. Welcome Robin.

Robin Canup:Thank you very much, Jim.

Jim Green: Well, it’s really great to have you here. Ever since planetary scientists looked up, saw the Moon, we were always wondering: how the heck did we ever get such a beautiful body? So what are some of the different theories about the Moon and how it was formed?

Robin Canup: This is, indeed, a very old question in planetary science. Some of the early ideas were known as “capture,” “co-formation,” and “fission.” So “capture” just proposed that the Moon formed independently from the Earth and that during a close flyby, enough energy with dissipated that it was captured into a bound orbit around the Earth. “Co-formation” imagined that as the Earth was accumulating material and growing, the Moon grew alongside it.

And then, finally, “fission” imagined that the Earth was once rotating so rapidly that it became unstable, and it became more and more oblate at its equator until the material that eventually formed the Moon was ejected from its equator regions. So those are traditional theories, but actually the theory that is favored now is a different one, one that was proposed in the mid-1970s, called the “Giant impact hypothesis.”

Robin Canup

Jim Green:So, sounds like the best theory going is a giant impact. So what did hit the Earth?

Robin Canup:So, we think it was another forming planet that collided with the Earth. We’ve since referred to this planet with a name, Theia, which is after the Greek goddess who was the mother of the Moon. So, we think it was another planet-sized object that collided with the Earth, near the near the end of Earth’s formation that gave rise to our Moon.

Jim Green:So, how do you tease that out? What did we know before and after the Apollo lunar missions?

Robin Canup:From remote sensing of the Moon, before we actually went there, you could tell things like the size of the Moon, its mass and, therefore, its bulk density. And we knew right away from the bulk density that the Moon, compositionally, was very unusual. Its density is very low, and that indicates that it lacks high density iron. So while the Earth’s iron rich core comprises about 30 percent of Earth’s mass, the Moon’s core is tiny; it contains only about a percent or so of its mass. So we knew that the Moon was iron depleted even before we went there.

Jim Green:Yeah, really anemic in that respect.

Robin Canup: Absolutely. We also knew the spin rate of the Earth, our 24 hour day, and we knew the length of the orbital period of the Moon around the Earth. And together those quantities constitute what we call the ‘Angular momentum’ of the Earth-Moon system.

Robin Canup: And we knew that as the Moon and Earth interacted over time, the Moon’s gravity had raised tides on the Earth; most notably in our oceans. And that the interaction of the Moon with those tides, had caused its orbit to spiral outward, over time, and the Earth’s spin too slow, over time.

Jim Green:So, we had some inkling the Moon must be moving away from us.

Robin Canup: Exactly. And we also knew, because of that, that when the Moon first formed long ago, it would have formed much closer to the Earth, and that the Earth would have been rapidly spinning with a day of only a few hours in length.

Jim Green: Yeah, and so like a ballerina spinning. As she puts her arms out, she slows down?

Robin Canup: Exactly and as she pulls them in, she speeds up. And in the same way, the Earth was rotating much more rapidly when the Moon was orbiting close to it.

Jim Green: So, that really gave us with our six Apollo lunar landers the concept of, “Well if we’re going to put some instruments out, one of them’s got to be a retro reflector.”

Robin Canup: Exactly. And that allowed us to confirm the predicted outward motion of the Moon, which is occurring at about a few centimeters per year, even today.

Jim Green: So Robin, does that mean that when the dinosaurs looked up, the Moon was bigger than it is today?

Robin Canup: No, the Moon was probably close to its current size then.

Jim Green: Yeah, ’cause–

Robin Canup: The reason is that when the Moon first evolved away from the Earth, it did so very rapidly. But the rate of that expansion slows very quickly as the Moon gets farther away from the Earth.

Jim Green:We brought back a lot of lunar material, I think we’ve got like 850 pounds of lunar material in our archive that we’ve been teasing out and looking at. And so that also told us a lot about the age of the Moon.

Robin Canup: Absolutely. In fact, it was those oldest samples that revealed that the Moon formed only about 60 million years after the oldest known grains in our solar system. And once we knew that, we knew that the Moon itself was a product of the same epoch that had produced the planets, that it was an ancient body, and that its origin was, likely, tied to that of the Earth itself.

Jim Green:So, 60 million years after those early grains were forming must mean then, where’s the Earth in that? Is the Earth already formed, or nearly formed, or did it form first?

Robin Canup: We build models to try to simulate the assembly of the planets in our inner solar system from a disk of material that existed around the young Sun. So those models track the material as solids collide and grow into, first, mini planets and then ever larger proto planet,s reaching, ultimately, Earth-size objects in the inner solar system.

Robin Canup: And when we follow that process, the initial phase, in which you form bodies similar in size to our Moon, takes place relatively quickly in only about a million years. And so at that stage, you have a system of mini planets, probably 100 or so of them in the solar system.

Jim Green: Protoplanets?

Robin Canup: Protoplanets, exactly. And, initially, they’re all in relatively circular orbits, but then on a longer timescale, their mutual gravitational interactions slowly cause their orbits to begin to cross, and once they cross, they can collide with each other.

And in these collisions, typically, there’s only one survivor; usually the planets that are colliding merge and you form a larger planet. So that process of these mutual collisions occurs over 10 to 100 million years, based on our numerical models of this process. And, interestingly, that phase of giant impacts coincides with the measured age of the Moon from the oldest lunar rocks.

Jim Green: That’s really neat. These protoplanets, they’ve already got their core, mantle, crust. They’ve already melted and the core is now concentrated in iron and some of the larger mass materials that we have.

Robin Canup: Absolutely. And by the time those proto planets are Moon-sized or larger, the melting that you mentioned has caused their highest density metallic components to concentrate in their centers into a core.

Jim Green: And that’s from radioactivity and also because of the crushing mass.

Robin Canup: Exactly. So there’s a constant heat source in the early solar system from the decay of unstable of isotopes. In addition, every time you have one of these big collisions, there’s a lot of heat deposited in the interiors of the colliding bodies; and that drives melting too.

Jim Green: So why do you think we should care about how the Moon was formed?

Robin Canup: Well, the big question of relevance here, of course, is how was the Earth formed and, more generally, how are Earth like planets formed in other solar systems as well? So the Moon is an amazing clue to answering that question.

So, our Earth, of course, is geologically active, it’s constantly evolving and, in that way, it tends to erase the record of its ancient past. The Moon, by comparison, is dormant and its highly cratered regions have been largely unchanged for billions of years, and they’ve preserved a record of the ancient past.

Jim Green: That is just really phenomenal. When you think about it, that history of the evolution and the origin of our solar system is all about studying the Moon and its place, how it came together, and the record that it has on its surface; this is a really fascinating topic to me, in particular.

Robin Canup: And not only is it ancient, but it’s our Moon.

Jim Green:Yeah.

Robin Canup: So it’s shared our same local history.

Jim Green: Right.

Robin Canup: And we think that its formation was linked directly to the final stages of formation of our planet through this giant impact event that we think produced the Moon.

Jim Green: The other part about that that’s fascinating to me, when I look at a planet, the first thing that comes to mind, the spherical nature of it, maybe gravity is uniformly distributed through that; and that’s just not true of anything out there. The Earth is, like, pear shaped, and the Moon has got huge mass differences, and with the Sun pulling on us, it’s amazing that our rotational axis isn’t being flipped around all over the place.

Robin Canup: One of the really interesting aspects of our Moon is that it has substantial gravitational effects on our planet’s rotational axis. So by that I mean, right now the tilt of our North Pole is about 23 and a half degrees and it’s that tilt, of course, that gives us our overall seasons.

Now with the Moon in place around the Earth, and with the Moon having its mass, the variation in that tilt is very small over long time scales; only about a degree or so. But within our solar system, if you take away the Moon, if you imagine the Earth without our Moon, or even the Earth with a Moon that was less massive by a factor of a few, because of the interactions of the Earth with the other planetary orbits, the tilt of our North Pole, our spin axis, would have varied largely over our history.

And by large, it depends somewhat on what you would assume the rotation rate of the Earth would have been without the Moon. But these variations would have been tens of degrees, perhaps up to 40 degrees, and so that would have had a profound effects on climate–

Jim Green: Climate, yeah, yeah.

Robin Canup: –On our planet, absolutely, and on how life may have been able to develop.

Jim Green: Yeah, the climate variations, alone, when you think about humans migrating, whole civilizations have been destroyed by climate, even during this time with our axis so stable over tens of thousands of years; even so recently as that.

Well you mentioned you’ve really been using computers and computer simulations to really figure out how the Moon was created and its relationship to the Earth. How do you do that, how do you approach that kind of problem?

Robin Canup: So we think the Moon formed by one of these giant collisions by a protoplanet with the forming Earth at the very end of its formation. So you had another planet-sized body colliding with the Earth. This is not an event we can simulate in the laboratory, obviously.

So instead, we build computer models and we put in all of the physical processes that we think will be important that we can include. And we use that computer simulation, in effect, as like an experiment to test, as a function of how large the impacting planet was, what angle it hit the Earth at, what speed it came in at. What are the impacts on our planet?

And, essentially, the main approach is to describe the colliding bodies by, typically, about a million individual parcels of material, whose individual evolutions you’re tracking during the course of the simulation. You’re tracking things like their position in space, their velocity, their temperature, and their phase; whether they’re a vapor, a melt, or combination of the two.

Jim Green: When did you just start doing that? When were computers good enough to enable you to do something that you could believe?

Robin Canup: Well the technique I just described was really initiated in the mid 1980s, before I was working on this problem. But at that time, it was extraordinarily difficult, computationally, to do these simulations. So, a single simulation with very low resolution could take several years of computer time.

Jim Green: Wow.

Robin Canup: So, initially, the number of these simulations we could do was very limited. And so with such groups, led by Willie Bends and Al Cameron did initial simulations that showed, yes, a collision like this could put debris into orbit around the Earth, so we could, potentially, form a Moon this way. But they weren’t able to identify the type of impact that could really give us our Earth-Moon system.

And so when I started working on the problem it was around 2000, and by that point we were able to do, say, 40 or 50 simulations within a few months of computational time. And with that kind of ability, you could narrow down the type of collision which we think to describe many features of the system would involve a, roughly, Mars-sized impactor.

Jim Green: Yeah. And in addition to that, I’m sure you’re looking for certain features that would come out, that would reproduce the current knowledge about our Earth/Moon system today. What were some of those important features that led you onto the idea that this must be the right way to go?

Robin Canup: So initially, we were trying to explain the very basic properties of the Earth-Moon system. So we were trying to explain why the Moon lacks iron, why it doesn’t have a large iron core. We were trying to explain, of course, the masses of the Earth and the Moon, and the Moon is quite massive relative to the Earth for a satellite; and so that was a key constraint.

And then, finally, we were trying to explain our current 24-hour day. In other words, we were looking at impacts that would have left the Earth rotating with a day of only several hours, so that as the Moon-Earth system tidally evolved, subsequently over 4.5 billion years, we would recover our current 24 hour day and the orbit of the Moon.

Jim Green: And that’s because the Moon would have to form close to the Earth from that impact.

Robin Canup: Absolutely, so these impacts eject some of the material into orbit around the Earth and it forms a disk around the Earth. Now it’s interesting because inside this critical distance, the Roche Limit, the planet’s gravity is strong enough that it keeps debris from accumulating and growing into a satellite.

And we know this, for example, from looking at the rings of Saturn. The rings of Saturn, those particles collide with each other all the time and yet they don’t coalesce into a Moon. And that’s because they’re orbiting within the Roche Limit at Saturn.

So, we know from this that when there was this massive hot disk around the Earth, produced by the giant impact, that the Moon would have coalesced, not at the disk inner edge, but beyond that distance at about three to four Earth radii from the center of the Earth.

Jim Green: Really close to us.

Robin Canup: Yes.

Jim Green: Wow, that’s fantastic. So it looks like the impactor, about the size of Mars, a fully formed large body, and I guess we call it ‘Theia’ today.

Robin Canup: That’s right. And, interestingly, the nature of this collision has really been debated over the past decade. And a collision by Mars-sized body, like we’ve talked about so far, does a great job of explaining the basic features: the spin rate, the masses, the Moon’s iron depletion.

It explains the Moon’s iron-depletion because when you have one of these collisions, the material that goes into orbit around the Earth from which the Moon forms comes, primarily, from the outer iron-depleted layers of the colliding bodies, rather than from their metallic cores. So the impact, naturally, produces an iron-depleted Moon.

But the really interesting part of this problem is what we know, solely, because we have samples of the Moon from Apollo. Now, interestingly, when we analyze those samples, even though we knew the Moon was very different from the Earth in terms of its bulk iron content, those samples show that, for example, when you look at the most abundant element in mantle rocks, oxygen, and you compared the relative abundance of the isotopes of oxygen in lunar rocks versus terrestrial rocks, they looked indistinguishable.

And yet when you look at a meteorite that either came from Mars, or came from a parent body in the asteroid belt, their distribution of oxygen isotopes and the relative abundances looks distinctly different from that of the Earth.

Jim Green: So the measurement of an isotope is really important. So, how do we define that, what is an isotope?

Robin Canup: So for an element like oxygen, most oxygen atoms will have eight neutrons in their nucleus, but a small percentage of oxygen atoms will have an extra neutron, and an even smaller fraction of oxygen atoms will have two extra neutrons. So these different types of oxygen atoms are known as ‘different isotopes of oxygen’.

And so what you can do is you can take a sample, from anywhere in the Earth actually, and you can measure the relative proportion of these different isotopes. And it’s by comparison of those relative abundances of the different isotopes that one can, essentially, see a fingerprint of the distribution of material that gave rise to the Earth, or to the Moon, or to meteorites, for example. So we use these isotopes to tell us about the original source material that went into a planetary object.

Jim Green: So it’s all about the number of neutrons.

Robin Canup: So from the Apollo samples we knew that, while the Moon lacked iron, in other ways it looked amazingly similar to the Earth’s mantle. So why is that important? So we did this first generation of simulations of giant impacts and showed that a Mars-sized impactor could explain the spin rate; the iron depletion.

But one of the things these impact simulations always find is that the material that goes into orbit around the Earth, in these types of impacts, comes from Theia; it comes from the outer layers of the impacting planet, rather than from the Earth. Now we don’t know what Theia’s isotopic composition would have been for certain, of course, that planet is now gone.

But if we look at meteorites and we say, “Well, Theia was probably as different from the Earth as, say, a meteorite from Mars is today.” And if we assume the pre-lunar disk came from Theia, we would expect there to be measurable differences between the Earth and Moon today in terms of this oxygen isotopes signature. And we don’t see any.

Perhaps Theia, in terms of its isotopic composition, just happened to be very Earth-like by virtue of having formed nearby from a common source of material, for example.

Jim Green:It had to be left alone long enough to be able to get as big as it was.

Robin Canup: Exactly. And it has to have avoided having a collision with the Earth for tens of millions of years based on the 60 million-year-old time before we know the Moon formed, based on the oldest lunar rocks. So what’s been happening is we’ve been looking at lots of different types of collisions to see if we can find collisions that can, in addition to explaining the mass and the iron depletion of the Moon, can also be consistent with this extreme isotopic similarity between the lunar rocks and the Earth’s mantle. And so there’s a variety of different types of processes that have now been proposed.

Jim Green: Another feature of the Moon is the crust thickness. We’re finding from GRAIL data that the crust on the far side of the Moon is thicker, by a considerable amount, than the near side of the Moon. Can we reproduce that in our computer simulations?

Robin Canup:Well, just after the impact, the Moon was at least molten in its outer few hundred kilometers, we know that; it may have been fully molten when it first formed. So, then, as the Moon cooled, it was also tidally interacting with the Earth, and its orbit was spiraling outward.

One of the interesting effects of the gravitational interaction between the Earth and the Moon is that as the Earth is tugging on the Moon, it tends to distort the shape of the Moon somewhat. It tends to make it slightly non-spherical, it tends to elongate it in the direction of the Earth.

And once that distortion forms, if the same face of the Moon isn’t always facing the Earth, if it’s rotating, then the Earth will tend to tug on that distorted shape that it’s created in the Moon. And that tugging dissipates energy until the Moon locks to having just one face towards the Earth, and so we think the Moon would have achieved that locked configuration with the Earth quite rapidly.

And so there are some models that associate that locking and the, subsequent distortion of the Moon by the Earth in that state while it still had an underlying melted region that we call a ‘Magma ocean’, with why the crust on the near side would have ended up being thinner, and the crust on the back side would have ended up being thicker.

There’s other models that propose that it may have been due to the fact that the side facing the Earth would have had a higher temperature due to radiation from the, then, still hot Earth from the giant impact. And that that may have led to additional deposition of some elements on the far side, leading to a thicker crust.

Jim Green:Yeah, so the Moon then would, literally, have an atmosphere, so to speak, and then the dynamics are so much different. I had also heard that another theory was that other objects, smaller, may be also accreting and then impacting the Moon on the far side, thickening that crust.

Robin Canup: Absolutely. So there is a theory by Jutzi and Asphaug that proposes that initially we accumulated one dominant Moon, our Moon, out of this disk produced by the giant impact. But that there was also a smaller companion Moon, and this is definitely an outcome you see and some of the numerical models of the Moon’s assembly after a giant impact.

Jim Green:Okay–

Robin Canup: So there are–

Jim Green: Okay, it forms objects?

Robin Canup: Yeah, so their idea was that that thickened far-side crust is due to a mini Moon, essentially, colliding with the Moon and depositing extra material just over, approximately, a hemisphere of the Moon’s surface.

Jim Green: Whenever we do these kind of simulations, we also find out some new things that we want to be able to see, or decide on new measurements that we want to make. Were there any things like that that were coming out of your models that allowed you then to say, “Oh, we have to check this out?”

Robin Canup: One of the interesting things has come from the models that look at how the Moon accumulates from this debris field produced by the giant impact. And as we talked about before, the Moon forms in the outer regions of the disk beyond this critical distance known as the Roche Limit.

So one of the things we found in the model is that the material in the outer disk often comes, more heavily, from the impacting planet Theia. While the material in the inner disk is derived, more strongly, from the Earth and, indeed, may be more likely to mix with material from the Earth. Material in the outer disk tends to come more from that impacting planet.

Robin Canup: Now, as we developed models of the Moon’s assembly, what we see is that material in the outer disk accumulates into an object with, say, half the Moon’s mass, very quickly in only a few months after the impact.

Jim Green: Wow, wow.

Robin Canup: Then over a longer time scale, as this hot inner disk near the Earth cools, condenses, and spreads outward, the last half of the Moon’s mass is then delivered from this inner region. So this has led to a prediction: that there are various signs that the Moon did not fully mix within its interior after it formed. And if that was the case, then we think this initial portion of the Moon that accumulated very rapidly, would be the most likely to be volatile rich, to be abundant in things like water, and to perhaps hold isotopic signatures of Theia, of that lost plant, that collided with the Earth.

In contrast, the final portions from the inner disk would be much more Earth-like, typically, and those would have been added on last. Those outer portions of the Moon would then be more likely to be Earth-like and dry, because they were delivered from this hot, personally vaporized region of the desk. So the prediction is that as we look at lunar samples, and we hope we’ll get many more of them in the near term–

Jim Green: That’s the plan.

Robin Canup: –That reflect the Moon’s composition at depth, these models would predict that they will tend to be more volatile rich, and they will tend to be more isotopically different from the Earth than those derived from the Moon’s outer or upper regions.

Jim Green: Just recently we announced with the LADEE mission the analysis of that over several months seemed to indicate that water is released during meteorite showers. You know, that little micrometeors hit the Moon and then they go down to 10 centimeters, is the estimate, it liberates water. This then bodes well for this part of the theory.

Robin Canup: Absolutely. And the whole existence of a water cycle on the Moon was something that we didn’t even know about a couple of decades ago.

Jim Green: Yeah.

Robin Canup: We now know that there are multiple processes associated with water on the Moon.

Jim Green: Wow. Who would have thought: A water cycle on the Moon?

Robin Canup:Exactly.

Jim Green: Wow, that’s a fantastic discovery. Well what’s left now to do with the simulations? What should we be up to?

Robin Canup: So what we’re trying to do right now, we have about a half dozen, different impact models that have been proposed; each of them, right now, appears viable. So we’re trying to push these models to the next level, and by the next level I mean, we’re trying to make them sophisticated enough so that they can predict observable properties of the Moon that would result in each of these different scenarios.

So the goal is, ultimately, to try to test these different theories against current and future data about the Moon’s physical properties. So what do I mean by that? We’re trying to develop models that tell us, as a function of the type of giant impact, what is the initial thermal state of the Moon? To what degree was the Moon, initially, molten?

Because it turns out there are a variety of properties of the Moon that suggest that it was not fully molten, and in its interior it was solid, but its outer layers were molten. We’re also trying to develop predictions for the abundance of different elements in the Moon as a function of these various theoretical models.

Jim Green:Wow, that’s really fascinating.

Robin Canup:Everything we’re debating about Moon origin today is due to the fact that we have samples-

Jim Green: Yeah.

Robin Canup: And that we can compare those samples to those of the Earth, and they tell us that it’s not as simple as what we thought. And so this is an amazing example in astrophysics where the knowledge of the chemistry, and the physical properties that you get from samples, changes the nature of the problem altogether.

Jim Green: Yeah, really good point. Well Robin, I always ask my guests to tell me what event, or activity, or place, or thing that happened to them in the past that so excited them, that they decided to change their direction and become the scientist they are today; I call that a ‘Gravity assist.’ So, Robin, what was your gravity assist?

Robin Canup: Well I think I had two. The first was when I was in middle school and two things happened. First, The Cosmos series with Carl Sagan-

Jim Green: Oh, okay.

Robin Canup: Was playing, I watched that, constantly; it was one of my favorite shows. The second thing that happened was Voyager 1 and then Voyager 2 flew past Saturn, and the images I saw from those encounters, I just found unbelievable.

Jim Green: Yeah.

Robin Canup:The intricacy, the beauty, the planet, the rings; so that kind of set the stage. Then years later in college I was a physics major, and I didn’t have any particular plan of what I wanted to do. But the best Professor I had was an astrophysicist, a professor by the name of John Kolena, who also taught at Duke where I went, and the North Carolina School of Science and Math; right across the town of Durham.

And he taught his class in a very different way from all my other science classes. He didn’t assign problems from a textbook, his tests were open book and you had 24 hours to complete them. Instead, he ran his class, and he as he designed his homework and tests as, in effect, mini research problems. He would tell you, “This is what we start with, here are types of observations we could potentially make. How could you learn something from these to answer a problem?”

Jim Green: Wow.

Robin Canup: So this was a completely different approach.

Jim Green: Yeah.

Robin Canup: Essentially, the focus was on how do we figure out new things rather than memorizing things?

Jim Green: Yeah.

Robin Canup: Or learning how to do things in a way that someone has already established? And so I loved this class and, in retrospect, it was because he ran it like you were working on a research project.

Jim Green: Right, right.

Robin Canup: And so that’s what really led me to graduate school in astrophysics and, ultimately, interplanetary science.

Jim Green: Well, I really want to thank you, it was just delightful to talk about a new way of understanding things by simulation computers. And what you guys have done in this particular area is just really astounding. So Robin, thanks so much.

Robin Canup:Thank you. It’s a pleasure.

Jim Green: Well, join me next time as we continue our exploration of the Moon. I’m Jim Green, and this is your “Gravity Assist.”

Lead producer:Elizabeth Landau

Audio Engineer:Eric Wisler