It came from outer space.
Four and a half billion years ago, when the solar system was a young, violent and messy place, a hot and nasty planet hailing from somewhere between Mars and Venus took a detour. Traveling at 11 kilometers per second, it screamed toward one of its neighbors, another hot and nasty planet much larger than itself: earth.
The marauder clipped the top of the earth, shearing away layers of mantle and spewing a jet of vaporized rock, liquid metal and pulverized debris. Then, drawn back by the gravitational pull of the now-deformed earth, the planet slammed down again, splashing even more rubble into space and destroying itself. Within thirty hours, the dust began to settle. The earth was left spinning at the accelerated rate of five hours a day, and the remaining chunks of the errant planet slowly merged into its molten surface.
In orbit, a cloud of gas, powder and rubble began to form, like a hotter version of Saturn's rings. In time, much of this debris fell toward the proto-earth, while the rest cooled, bumped around and merged into larger and larger clumps until finally, only a single satellite remained. And this is how the moon was formed.
In February 2001, Robin Canup, an intense, enthusiastic and occasionally goofy 33-year-old astrophysicist, sat at her desk in Boulder, contemplating that scenario and preparing to give a talk about the origins of the moon. She'd been working in the field for only eight years, but already she'd become one of its leading theorists. She found the scientific mysteries "romantic" and the complex equations "centering."
For centuries, the world's brightest minds had pondered this puzzle. A mid-'70s theory about a collision made the most sense, but no one could make all of the pieces fit. If something hit the earth, then how big was it? How fast was it moving? What was the angle of impact? Would there be enough debris in orbit to form the moon? And if so, how did the moon actually clump together?
As she sat in her office that winter afternoon, studying results from forty computer simulations of the collision, Canup began to doubt.
"Gosh," she thought. "This is frustrating. None of these is right."
She checked one chart, then another.
"You know, we're looking at this the wrong way."
She ran a finger down a graph.
"If we go by the earth and moon system we have today, what different impactor size would that imply?"
She scribbled an equation on the bottom of her speech notes.
She checked her calculations for accuracy.
"That can't be right. Surely someone must have looked at that. Something must be off. But I'll just start a simulation to see what happens."
She punched the numbers into her computer, hit "Enter" and launched a massive three-dimensional representation of how the crash may have happened.
A day and a half later, her computer finished the program. Canup reviewed the results.
She rang William Ward, her colleague, friend and trusted sounding board.
"You're not going to believe this."
Taylor's Axiom: "The best models for lunar origin are the testable ones."
Taylor's Corollary: "The testable models for lunar origin are wrong."
-- Scientist paraphrasing fellow scientist Stuart Ross Taylor at 1984 conference (from Origin of the Moon)
Humans have always gazed up at the night sky and contemplated the silvery face of the moon. Ancient cultures ranging from Mesopotamian to Egyptian to Chinese had theories about what it was and where it had come from. Some simply proclaimed it a god.
Through scientific reasoning, philosophical discourse and flat-out guesswork, the Greeks determined that the darker portions of the moon's surface were oceans and the lighter parts were land. They thought people lived there, too.
In 1610, science took a great leap forward when Galileo peered through a telescope, made a few detailed illustrations and pronounced that the darker and lighter regions of the moon weren't land, water or even cheese, but mountains and valleys.
Nearly two centuries later, scientist Pierre-Simon Laplace offered the proposition that the solar system evolved from a rotating nebula of gas and dust, which implied that the earth and moon formed simultaneously and stayed bound by gravity.
That made sense to many scientists until the late 1870s, when George Howard Darwin, son of the evolutionist, came up with another theory: During its formative years, the largely molten earth was spinning so fast that a large, unstable lump formed at the equator; eventually, it was ripped away by the sun's gravity and tossed into space, where it became the moon. Four years later, geophysicist Osmond Fisher bolstered this concept by suggesting that the Pacific Basin was actually a scar left from the errant lump.
In 1909, however, Thomas Jefferson Jackson See rejected the lump idea, suggesting that the moon was a wandering planet from another part of the solar system that had been captured by the earth's gravity.
None of those theories, which came to be known as the Big Three, was entirely convincing, but in the absence of better explanations, all three remained dominant well into the twentieth century.
Gradually, scientists learned more and more about the moon, including its size, mass and density. They discovered that the moon is larger in relation to the earth than other satellites and their planets, with the exception of Pluto's Charon. They learned that the early rotations of the earth and moon were faster than those of their counterparts. And they determined that the moon has less iron than they'd expected. Most rocky planets, including the earth, are about 30 percent iron, concentrated mostly at the core; they estimated the moon's center at less than 5 percent iron.
But the more they learned, the more they wondered.
Then came Sputnik, the Mercury astronauts, Gemini and the Apollo missions, which were supposed to settle the debate once and for all. From 1969 to 1972, a dozen astronauts kicked around the moon's powdery surface and rumbled along in lunar buggies, collecting about 850 pounds of dust and rock. Instead of solving the mystery, though, these specimens raised more questions.
For one thing, compared to the earth's crust, moon rocks contained an extremely low concentration of volatiles that evaporate easily, such as water and potassium. This suggested that the moon had undergone some thermal process that the earth had not. The lunar crust also contained twice as many refractory elements as the earth, including aluminum, calcium oxide and titanium, and those ingredients are hard to vaporize.
But there were similarities, too. Moon rocks showed that the earth and the moon were approximately the same age, about 4.5 billion years, and had matching oxygen isotopes, which indicated that they had formed in the same neighborhood.
The moon specimens both confirmed and contradicted earlier theories. The Big Three began to topple.
In 1975, William Hartmann and Don Davis of Tucson's Planetary Science Institute introduced the idea that roaming "planetesimals" could have struck the earth, sending enough debris into orbit to create the moon. A year later, working independently of Hartmann and Davis, Alastair Cameron and William Ward of Harvard University suggested that the earth may have been hit by a rogue planet the size of Mars.
And the Giant Impact theory -- or the Big Whack, as it came to be known -- was born.
On the surface, this hypothesis seemed to explain many of the moon's weird attributes rather nicely. A collision could have blasted enough rocky mantle from both the earth and the rogue planet into orbit to form a low-density moon. A collision could have produced hellish temperatures of several thousand degrees Kelvin, which would have been enough to vaporize the volatile elements from moon rocks. And a collision, especially at a slight grazing angle, could have left the young earth spinning faster, too.
Yet the reaction to this theory was "underwhelming," Hartmann recalls. "It certainly did not make a splash."
At the time, many scientists, particularly geologists, frowned on any notion of cataclysmic events explaining the mysteries of the solar system. Traditional views held that the planets had formed through a slow and steady accumulation of cosmic rubble, not through hot and nasty objects smacking into one another. The process, they believed, was more like a waltz than a slam dance.
"People thought it was just too ad hoc," Hartmann says. "People thought, 'Oh, you guys are just inventing this stuff out of the blue rather than using existing phenomena that we already know about. What evidence do you have?' People just weren't willing to accept random events out of the sky having major consequences for the history of earth and life on earth."
And so the theory "sort of languished," he says.
In October 1984, scientists gathered in Kona, Hawaii, for a conference devoted to the moon's origin. For two and a half days, scientists from seven countries compared, contrasted and critiqued the leading theories as 62 authors offered 58 papers. But once all the paperwork settled, the Big Three had been laid flat, and one hypothesis stood tall: the Big Whack.
"People went home and thought, 'Hey, we'd better do some theoretical studies of this," Hartmann says.
Al Cameron remembers a bandwagon forming and "nearly everyone jumping on."
At Kona, the Big Whack theory received a boost from the Carnegie Institution's George Wetherill, who had studied the creation of the solar system. His work convinced many scientists that large collisions may indeed have played a crucial role in the final phases of planetary formation. The cosmic spheres may have accumulated gradually, but as they grew larger, they began to jockey for position around the sun. A collision or two was certainly possible.
Unfortunately, such collisions weren't easily duplicated in laboratories, and the calculations were too complex for early computers. But Kona changed that, too. Before the conference, Cameron had done basic calculations on IBM clones. After the conference, supercomputers arrived.
In November 1984, at Los Alamos National Laboratory to deliver a talk on the moon's origins, Cameron was approached by Wayne Slattery, a former post-doctoral student of his, and Willy Benz, a Swiss post-doctoral student in astronomy. Slattery worked in the lab's weapons division; Benz had been tinkering with a sophisticated code he'd modified called "smooth particle hydrodynamics," which divides matter into mathematical particles and tracks their interactions individually. Benz thought the technique might work well on the moon-origin problem. Slattery agreed. So the three scientists tried it out on the lab's classified supercomputers.
At roughly the same time, at Sandia National Laboratories in Albuquerque, Jay Melosh, an expert on "how metals blow up," was sitting down to lunch with computer analyst Marlin Kipp. Initially, Melosh had been a Big Whack basher; after talking with colleagues who'd attended the Kona conference, he set out to write a paper proving that the theory "was all just bunk," he says. But as he started his research, he saw the light. Over lunch, he told Kipp about his conversion, and the two decided to perform their own simulations on Sandia's Cray supercomputers using codes originally developed for nuclear-bomb explosions.
By the mid- to late '80s, both teams had developed movies and slides demonstrating, in a general way, how a collision may have worked. But in subsequent years, as they continued to perform tests, they ran into one complication after another. No matter what they tried, the scientists couldn't hit on a single scenario that would explain the present-day earth and moon.
At night in the suburbs of Poughkeepsie, New York, six-year-old Robin Canup would snuggle under the covers while her dad reached for her favorite bedtime book: a text on the formation of the earth. While she drifted off to sleep, he recited passages about magma, explosions and poisonous gas.
"Oh, I loved it," she says. "The idea that the early earth was so vastly different than it is today. The whole process of life coming out of such an environment. I had him read it to me over and over."
Her choice of bedtime reading wasn't unusual in the Canup home. Robin's father was a "very applied" and "very practical" Texaco physicist studying engine development. One of her four half-brothers became a NASA technician.
"Our house was a pretty technical place," she remembers. "There were always things being built or taken apart, usually electronic in nature. It was not uncommon to find an oscilloscope and a soldering iron on the kitchen table. I had no idea that was strange."
In school, she translated family discussions about "how life may have originated" into straight A's in math and science. She even performed basic calculations on how long it would take before she shuttled off to Mars: She figured she'd be there by the age of 32.
"It was fun," Canup says. "I had no idea I'd be able to make a living doing this type of thing. It was just something I loved."
But at the time, she loved something even more: ballet. At age six, she took up dance on the advice of a doctor, who thought it might improve her asthma. As a teen, she trained daily, visited New York City several times a week for lessons and performed with the Poughkeepsie ballet.
"That was my focus," she says. "Ballet had this very demanding and challenging technical structure, but with this emotional and romantic, creative appeal superimposed on it. I took to it immediately."
In high school, though, Canup began to recognize the demands of the profession and the limitations of her body. So she enrolled at Duke University as a biology major. An enthusiastic professor inspired her switch to physics, which pleased her dad, and then another enthusiastic professor inspired her to switch to astrophysics. Her childhood passion for "big questions" was reignited: "Big questions like, 'Where did it all come from?' 'How did it all begin?'"
Canup is a polite, friendly and talkative woman with straight brown hair and bright eyes. She listens intently, speaks rapidly and chooses her words precisely. She likes to laugh and laughs a lot, often nodding her head and rolling her eyes at even the smallest of jokes. But when she's talking about her work, serious work, her hands tremble with excitement.
After graduating from Duke in 1990, Canup decided to pursue a Ph.D. in astrophysics at the University of Colorado at Boulder. Soon she and professor Larry Esposito developed the first computer models demonstrating how the tiny particles and small moons in Saturn's rings may have formed.
In 1993, Canup presented their findings at an annual meeting of the American Astronomical Society's Division of Planetary Sciences, held in Boulder. After her talk, a tan, lanky scientist from Arizona approached and said she might consider applying the Saturn models to the origin of the moon. He was William Hartmann, co-founder of the Big Whack theory.
"People are doing all these models on the impact, and all these models seem to be showing that hot gas and dust and debris are thrown into orbit, but then -- click -- that's the end," Hartmann said. "Someone ought to figure out what happened to the debris. Does it really form a moon?'"
Canup didn't know anything about the moon, she told him. And even if she did, certainly someone must have already studied the problem. No, Hartmann replied, no one had. He'd developed a few simulations himself, but he didn't have the computational tools for a more sophisticated analysis. Canup did, though, because of her work on Saturn.
Intrigued, she read everything she could about the Big Whack theory and discovered that Hartmann was right. So Canup approached her thesis advisor, and he gave her the green light to launch her moon project.
"She just picked up the ball and ran with it," Hartmann says.
It didn't take long for Canup to make her mark.
In 1996, for the second part of her dissertation, she and Esposito ran the first computer models showing how the moon may have formed from the impact rubble: First, they found that the blast would have had to force two moons' worth of debris into space. Second, they found that at least half of that debris would have to have been blasted beyond the point at which it would be pulled down by gravity. In the end, however, they were unable to re-create the conditions to produce a single large moon.
A year later, however, Canup, CU's Glen Stewart and Shigeru Ida of the Tokyo Institute of Technology took those findings a step further and found a way to show how one big satellite could coalesce from the debris. They also established a fairly good estimate of the amount and type of debris that would be needed to make the moon.
Canup's work had caught the attention of scientists worldwide, including Cameron. In 1997, they began collaborating. With Cameron studying the impact and Canup studying the formation, they hoped to paint a more complete and consistent picture of the Big Whack theory.
"I do the impact calculations, which end with a lot of material orbiting close to Earth," Cameron told the Harvard University Gazette in September 1997. "She plugs these results into another computer model to determine how that material coalesced into the moon."
But the task wasn't nearly that simple. Cameron and his collaborators already had run dozens of computer models using dozens of variables. They'd come close to the answer and made important discoveries, but they hadn't found a single scenario that fit completely, Canup says.
"He'd get several of the parameters right, but not all of them," she says. "There would be some piece of his system that wouldn't match current earth-moon conditions."
Cameron had begun to believe that a rogue planet twice as big as Mars had struck the earth. That could put enough debris into orbit to form the moon, but it would also leave the earth spinning so fast that a second collision would be required to slow it down. So he scaled back his parameters and began to think that a smaller, half-formed earth might work. That would solve the spin problem, but it raised another issue: How could the earth grow to its full size without the moon also growing and collecting too much debris, such as iron, in the process? The more complicated the explanations became, the more Canup began to wonder.
"The more you lengthen the series of events that would have to occur for something to happen, the more you tend to question that it's really how it happened," Canup explains. "In my perception, when things seem incredibly complicated, it's not usually that nature is incredibly complicated; it's just that we haven't seen our way through to the answer yet."
Canup suggested that they run more tests, particularly using different sizes of rogue planets, but Cameron resisted, she recalls. He was an icon, a man Hartmann called "a great god of the field," a brilliant, gruff and eccentric scientist who preferred PCs over supercomputers so that he could "sit and watch calculations evolve with time." And Canup, as bright and promising as she was, felt awkward questioning him. After all, she was only a post-doctoral research associate at CU.
"I think he was convinced he knew how changing all the impact variables would change things, and that he had stumbled upon the only way, which was to have the earth half-formed," Canup says. "At the time, he saw no reason to change his thinking."
On the other hand, she remembers being "frustrated out of my mind. I was dependent on the results from his impact simulations. They were fundamental to my work. But I had these very basic concerns. The early-earth scenario just had so many restrictions and problems. It was a mess. I couldn't understand how you could grow the earth and not contaminate the moon. I just didn't believe it."
So she "rolled up her sleeves," she says, and started from scratch.
In March 1998, Canup became a senior research scientist with the Southwest Research Institute, an independent, nonprofit organization that studies such matters as planetary science under public and private grants from NASA, the National Science Foundation and other groups. Although the institute is based in Texas, Canup moved into an office near the Pearl Street Mall; she didn't want to leave Boulder.
After taking up ballet again in graduate school, she'd been performing with the Boulder Ballet for several years. She had other local ties, too: She'd married Richard Mihran, an electrical engineering Ph.D., lecturer and adjunct professor at CU, in the mid-'90s. And she loved the mountains.
Several months into her new job, Canup contacted Eric Asphaug, assistant professor of earth sciences at the University of California at Santa Cruz; Asphaug had performed computer studies of asteroid and comet impacts using a derivative of the smooth particle hydrodynamics code that Cameron and Benz had used over a decade before. Asphaug agreed to help her develop a new computer model to advance moon research.
Next, Canup and William Ward, an institute colleague who'd worked with Cameron on the invention of the Big Whack theory back in 1976, began working with Cameron to analyze, plot and examine trends from about forty of Cameron's models. Together they developed a fresh way to compare, contrast and display the results. Then they searched for patterns.
"We said, 'Wait a minute. What are these results really saying?'" Canup recalls. "`Surely there's some impact that could more closely get us to the earth-and-moon system of today. And if there's not, that's significant.'"
In December 2000, after dabbling in other projects, Canup and Asphaug ran their first computer trials of possible moon-forming collisions. Since they used the data from Cameron's tests, their results were similar. Canup still doubted the parameters, but at least she knew the new computer program worked, or at least that "they were not working all the same way," she says.
Two months later, Canup was preparing for a talk at the University of Arizona on the origins of the moon. She planned to discuss the latest research, offer an update of her own work and then end with a discussion of the ongoing problem of finding the right scenario to validate the Big Whack.
As she read the analysis she'd recently published with Ward and Cameron, she had an epiphany about what the right rogue-planet size might be. Canup dashed off a calculation, and the answer came back: the size of Mars. She ran a computer simulation, and it worked. Then she called Ward into her office and walked him through her reasoning.
"Makes sense," he said. "But that's the first thing they would have tried. Surely someone would have caught it."
Canup hit the books, and what she discovered was this: Although Cameron and Ward had proposed a Mars-sized rogue planet back in 1976, scientists had not tested it on a computer until the mid-'80s. And when they did, the theory was quickly dismissed.
Scientists had reasoned that a collision with that small a rogue planet would require a more severe impact angle to blast enough rubble into orbit and to leave the earth spinning at the correct rate. But the more severe the angle, the greater the proportion of iron from the rogue planet's core that would get spewed into space -- and that would leave too much iron in orbit to form a low-density moon. Time and again, their results were consistent. So without looking back, Cameron and other scientists had begun testing larger rogue-planet models. And that's the direction moon research had taken for over a decade.
But as she read the reports, Canup began to think that the early simulations weren't detailed enough to make reliable predictions about the quantity of iron in space. Computers were too slow back then, and a single simulation took nine months to complete. Cameron and his colleagues could run only a handful at a time. Not only that, but the entire collision was represented with only 3,000 particles -- compared with the 30,000 particles used in 2001 models. The fewer particles you have, the less detail you have. And with an event as complex as the Big Whack, detail is crucial.
Canup believes that Cameron, too, would likely have settled on a Mars-sized rogue planet if his technology had been advanced enough to run more simulations with greater resolutions.
"I'm sure they would have found it," she says. "They're all smart guys. But Cameron formed his thinking about how these impacts behaved back when he could only get results from one or two simulations per year. I had the benefit of looking at the results of forty simulations at once."
After double-checking the results of her model using a Mars-sized rogue planet, Canup e-mailed Cameron. Within five minutes, she received his reply. "I don't have time to read your paper right now, but I can tell you straight away that it won't work," he said. "We've already tried it. It produces too much iron."
Canup was disappointed by what she calls Cameron's "reflexive response," but she wasn't discouraged. She ran forty more of her own computer models using Asphaug's codes and the following basic parameters: a Mars-sized rogue planet traveling at 11 kilometers per second and slamming into a full-sized earth at a 50-degree impact angle.
Canup and Asphaug announced their results in the August 16, 2001, edition of Nature.
"It was the first time the simplicity and elegance of the original theory had been demonstrated," Canup says. "We have one impact that gives us the right spin rate of the earth, an iron-depleted moon, the right mass for the earth and the right mass for the moon. It was luck. A complete accident. It was extremely exciting."
It also inspired an immediate response.
According to Hartmann, currently a senior scientist at the Planetary Science Institute, the simulations done by Canup and Asphaug make the original Big Whack theory "look better and better."
"Their models are saying that this could work," he adds. "You still have people who say the theory is crazy, but no one is quibbling with [Canup's] calculations. People agree that her work is very solid. Combined with the growing belief that there were big interplanetary bodies roaming around the solar system at the time planets formed, this makes it very credible that this could have happened."
Jay Melosh, a professor of theoretical geophysics and planetary surfaces at the University of Arizona's Lunar and Planetary Lab, was also encouraged. For that same issue of Nature, he wrote a commentary proclaiming that Canup and Asphaug had ushered in a new era of computer simulations.
"It has taken squadrons of physicists in the United States, Russia and elsewhere nearly fifty years to come up with the computers and three-dimensional computer codes that can adequately treat the effects of impacts and explosions," he wrote. "Thanks to the coevolution of modern computers and simulation algorithms, the means to model events of this kind can now be found on the desk of nearly any scientist."
He did add a qualifier, however: Since Canup and Asphaug did not use the most sophisticated equations distinguishing solid, molten and vaporized rock, "as encouraging as these results are, they are not the final word."
Although Melosh credits Canup with "checking up" on Cameron and with capitalizing on technological and mathematical advances, her work, like that of any other scientist, also needs to be checked. And since the number of people studying the moon's origin is small and "incestuous," Melosh hopes that verification comes soon.
"This is not a field in which only one or two people need to be involved," he says.
Cameron is even more skeptical -- mostly because Canup's results disagree with his own. Until he runs more detailed computer simulations later this year, he says, he'll view her models cautiously. "Obviously, the onus is on me," he adds.
Now a research professor at the Lunar and Planetary Lab, Cameron questions whether Canup and Asphaug do a good enough job showing the fraction of debris that actually forms the moon. Their model ends thirty hours after the collision, he points out, when the demonstration should run at least a few weeks.
And despite Canup's doubts that the earth could have been half formed when the Big Whack occurred, Cameron still stands firmly behind that possibility. He wouldn't be too quick to dismiss his mid-'80s simulations, either. Although his models may not have been detailed, he believes a Mars-sized rogue planet might indeed blast too much iron into orbit. "I don't know if that result is necessarily wrong," he says. "It just needs to be confirmed."
But while he questions her results, Cameron still tips his hat to Canup, "an energetic young lady who is pushing at the state of the art.
"And good luck to her," he adds. "We're fiddling with the details. We'll be fiddling with the details for a long time to come. As we do, the details will only become more detailed."
Canup acknowledges the skeptics with a nod and a smile. She fully expects someone to check up on her work. In fact, several scientists, possibly even Willy Benz, are preparing to do just that.
"What we said in our paper is, look, the impact we have is by definition the simplest case," she says. "It's the least restrictive. That's not to say that a more complicated scenario isn't possible; it's just that this is the simplest one. And from a strategic perspective, it's the best one to look into. The response I've gotten from the field has been, 'That's it. This makes sense. It's the best one.'"
Canup is also quite willing to give credit where it's due. Advances in computers and mathematics certainly played a role in her work, as Melosh suggests, but she was still able to recognize patterns in Cameron's data that had not been seen before and to predict the right size errant planet, which she believes is the key to the puzzle. "Once I recognized the trends, it only took a simple calculation to determine what impact would work," she says. "Within a few simulations, I had it."
And without Cameron's work, she notes, none of it would have happened.
"I'm extremely fortunate that I had his data," Canup says. "It was Cameron, after all, who got everyone thinking for the first time that it was a Mars-sized impactor. Even though his simulations led the field away from that in the 1990s, it was his work that in fact originated the belief. He's also extremely supportive of younger colleagues and has always encouraged my interest in this problem."
In part because of Cameron's criticisms, Canup is ironing out the kinks in the 2001 models. She and Asphaug are designing codes that will extend their simulations as long as Cameron suggests. And they're working with Melosh to update the equations predicting how solid, molten and vaporized rock may have behaved during the Big Whack.
Using simulations of up to 60,000 particles, Canup's examining other things, too. Whether the moon formed from a cloud of smaller particles or larger chunks. How the debris heated and cooled when it clumped together. How the debris was gravitationally and thermodynamically affected when other rubble smacked into it.
So far, her model based on a Mars-sized rogue planet continues to hold up. Canup believes she and her colleagues "have the basics okay." This summer, they plan to pre-sent a paper on their updated findings.
"We're not contradicting the earlier results," she says. "And that's the place you want to be."
Today Canup toils away in her Boulder office beside a chalkboard crammed with calculations on the satellites of Jupiter, her current obsession. Although she's still studying the origins of the moon, she hopes her models might help scientists solve other riddles, including why Uranus rotates as if it were tipped on its side, why Mercury has so much iron and so little mantle, and how Pluto's moon got so large. Someday her work might even help scientists discover the chain of events that sparked life on earth.
"They're all big-picture things," she says.
On the little-picture front, Canup recently hung up her ballet slippers. On her nightstand, the earth-formation textbook has been replaced by Atlas Shrugged. And while she missed her deadline of reaching Mars by the age of 32, these days she'd prefer to devote her free time to solving a conundrum closer to home: how to work with her husband on a bathroom remodeling.
"Don't say that," she jokes. "He'll kill me."
Canup is now a principal investigator with NASA's Origins of the Solar Systems program and the National Science Foundation's Planetary Astronomy program. The girl who never thought she'd make a living studying space is a young woman trying to answer some of science's greatest questions.
But even after a day filled with calculations and simulations and skeptics, at night she still regards the heavens with "appreciation and wonder," she says. And then, just as she did so many years ago, she falls asleep under the light of a mysterious moon.
Get the Weekly Newsletter
Our weekly feature stories, movie reviews, calendar picks and more - minus the newsprint and sent directly to your inbox.