Sara Seager has thought long and hard about the math: the odds that Earth harbors the only life in the universe are almost impossible. “The greatest discovery astronomers could possibly make is that we’re not alone,” writes the MIT astrophysicist in her new memoir The Smallest Lights in the Universe. “Humanity has searched the heavens for a reflection of ourselves for centuries; to see someone or something else, inhabiting another Earth—that’s the dream.”
A pioneer in the search for exoplanets, or planets orbiting other stars, she came up with the now-standard practice of studying the atmospheres of planets by analyzing the light that filters through them. Seager, who won a MacArthur Foundation “genius” grant, is the Class of 1941 Professor of Planetary Science and has appointments in the Departments of Physics and Aeronautics and Astronautics as well. She was also the deputy science director of the MIT-led NASA Explorer mission TESS (transiting exoplanet survey satellite) from 2016 to 2020, and a lead for Starshade Rendezvous, a feasibility study for a space-based mission to find and characterize Earth-like exoplanets. In her memoir, she shares her personal story of finding herself widowed at 40, a suddenly single mother of two young sons, while she explains the science of her search for other worlds.
This excerpt, drawn from different sections of her book, chronicles her work to develop ASTERIA. A satellite the size of a small suitcase, ASTERIA was designed to demonstrate the technology needed for a tiny telescope to search for exoplanets by detecting the minuscule dip in a star’s light when an orbiting planet passes in front of it. Seager initiated and developed ASTERIA at MIT, and later served as principal investigator while it was built and operated by the Jet Propulsion Laboratory from November 2017 until December 2019.
Searching for shadows to find other worlds
At its essence, astrophysics is the study of light. We know that there are stars other than the sun because we can see them shining. But light doesn’t just illuminate. Light pollutes. Light blinds. Little lights—exoplanets—have forever been washed out by the bigger lights of their stars, the way those stars are washed out by our sun. To find another Earth, we’d have to find the smallest lights in the universe.
If, for the moment at least, astronomers couldn’t fight the brightness of stars, maybe we could use their power to our advantage. Bodies in transit sometimes align. If we were lucky, a planet might pass between us and its star, creating something like a miniature eclipse. The moon looks giant when it blocks out the sun. The Transit Technique, as it would come to be called, applied the same principle to exoplanets. We would find them not by the light they emitted, but by the light they spoiled. Nothing stands out like a black spot.
In the fall of 1999, while I was a postdoctoral fellow at the Institute for Advanced Study in Princeton, the first transit of a known planet—HD 209458b, a “hot Jupiter”—was announced. It was absolutely fantastic news, in part because the discovery erased the last shred of doubt that exoplanets exist.
Studying starlight for signs of life
I had been turning over an idea—a genuinely original one—and the successful use of the Transit Technique gave it a greater urgency. A lot of science, especially pioneering science, relies on intuition. I didn’t have any evidence that my idea would work. But I was doubtless. I had realized that the technique might help reveal something more than the black silhouette of a planet. Immediately around that tiny partial eclipse, the same starlight that was being blocked by an exoplanet would pass through its atmosphere. The starlight would reach us, but not the way regular starlight reaches us. It would be filtered, like water running through a screen, or a flashlight’s beam struggling through a fog. If you look at a rainbow from a distance, its many colors form a perfect union. But if you look at a rainbow more closely, using an instrument called a spectrograph, you can see gaps in the light, minuscule breaks in each wavelength like missing teeth. Gases in the solar atmosphere and Earth’s own thin envelope interrupt the transmission of sunlight, the way power lines cause static in a radio signal. Certain gases interfere in telltale ways. One gas might take a bite out of indigo, while another gas might have an appetite for yellow or blue. Why couldn’t we use a spectrograph to look at the starlight passing through a transiting exoplanet’s atmosphere? That way we could determine what sorts of gases surround that exoplanet. We already knew that large amounts of certain gases are likely to exist only in the presence of life. We call them biosignature gases. Oxygen is one; methane is another. We could start with hot Jupiters, the planets we already know, and their more easily detectable atmospheres. Like a skunk’s spray, their traces of sodium and potassium would stand out amid the company of less potent atoms. I kept my idea to myself, because I knew it was great—I was the first to see the potential of the Transit Technique for studying atmospheres—and I knew, too, that great ideas get stolen. Dimitar Sasselov, my former PhD supervisor, was the only person I told about my theory, and he offered to help me bring it closer to practice. When we had worked out the details, I published a paper extolling what Dimitar and I called “transit transmission spectra”—reading the gaps in rainbows.
My paper received considerable attention. NASA was accepting proposals to use the Hubble Space Telescope; within a few months of publication, one team cited my work and won the rights to study the light that passed through the atmosphere of a transiting hot Jupiter. I was furious not to be included on that team, which chose an older male scientist over me.
Within two years, their work revealed the first exoplanet atmosphere. It didn’t surround another Earth, but my premise had worked. We had seen our first alien sky.
Spying on stars with tiny satellites
One of the great hurdles in looking for exoplanets is the time it takes to find them. The nearest and brightest sun-like stars are scattered all over the sky, which means that no telescope can take in more than a few at a time. But it’s prohibitively expensive, as well as nonsensical, to use something like Hubble or Spitzer to stare at a single star system waiting, hoping, to see the shadows of planets we’re not sure exist. Properly mapping a star system might take years.
I had been trying to make a long-term plan to find another Earth when I learned about what the community had taken to calling cubesats—tiny satellites designed to a standard form, which supposedly made them cheaper and easier to build and deliver into space. What if I made a constellation of cubesats, each assigned to look at only one star? I dreamed of space telescopes the size of a loaf of bread—not one, but an army, fanning out into orbit like so many advance scouts. Each could settle in and monitor its assigned sun-like star for however long I needed it to; each could be dedicated to learning everything possible about one single light. Hubble, Spitzer, Kepler—they each saw hugely. Maybe now we needed dozens or hundreds of narrower gazes, using the Transit Technique as the principal method of discovery. Cubesats wouldn’t see what larger space telescopes could see, but they would never need to blink.
I talked to David Miller, a colleague and engineering professor who was in charge of what would become one of my favorite classes: a design-and-build class for fourth-year undergraduates. It was revolutionary when it started, because it was so project-based; after a few introductory lectures, the students dived into the challenges of making an actual satellite. I asked David whether I could use his class to incubate my cubesat idea.
He was enthusiastic from the start. Maybe the best thing about MIT is that no matter how crazy your idea, nobody says it’s not going to work until it’s proved unworkable. And squeezing a space telescope inside something as small as a cubesat was a pretty crazy idea. The main challenge would be in making something small that was still stable enough to gather clear data—a tall order because smaller satellites, like smaller anything, get pushed around in space more easily than larger objects. To take precise brightness measurements of a star, we would need to be able to keep its center of brightness fixed to the same tiny fraction of a pixel, far finer than the width of a human hair. We would have to make something that was a hundred times better than anything that currently existed in the cubesat’s mass class. Imagine making a car engine that runs a hundred times better than today’s best car engine.
“Let’s do it,” David said.
Statistics and space hardware
Cubesats are much cheaper than regular satellites, because they’re smaller and easier to launch; they take up a lot less room in the hold of a rocket, and it costs $10,000 to send a pound of anything into space. Unfortunately, their cheap manufacture makes them prone to failure. Many of them never work. We use the same hopeless term for them that doctors use for patients they never got the chance to save: “DOA.”
One of our first hurdles, then, was a problem of statistics. (Every problem is a problem of statistics.) To make the cloud of cubesats that would come to be called ASTERIA, we had to figure out how many satellites we would need to give us a reasonable chance of finding another Earth-size planet. Thousands of bright, sun-like stars were worth monitoring, but we wouldn’t be able to build and manage thousands of satellites. We also knew that given the ephemeral nature of transits, the odds of an Earth-size planet transiting a sun-like star were only about 1 in 200. Some of our satellites would also no doubt fail or be lost. If we sent up only a few, we would have to be either very strategic or very lucky to find what we were looking for. There was some optimal number of satellites that, combined with a smart list of target stars, would keep our budget reasonable but still give us a good chance of success.
I was lucky to have a great group of graduate students and postdocs who I leaned on when my husband, Mike, got sick. I set one to work on ASTERIA’s optics, another on precision pointing, a third on communications. With their help, I’d made progress toward a prototype for my tiny satellites, inventing and testing precision-pointing hardware and software, and perfecting the design of the onboard telescope and its protective baffle. I worked hard to clear the rest of the path for ASTERIA to become real. After we’d laid the groundwork in the design-and-build class, my students and I were joined in our efforts by Draper Laboratory in Cambridge, where researchers work on things like missile guidance systems and submarine navigation. They also do a lot of work on space hardware. We had meetings every week, trying to solve the problems of small telescopes. We could build small enough components, and we could deploy the satellite and tell it what to do, but we still couldn’t figure out how to keep it as stable as we needed it to be. While we tried to solve that issue, I used my ongoing research on biosignature gases to determine what types of exoplanets deserved our focus. I thought we might be able to explore a hundred star systems or so in my lifetime; they had to be the right ones.
A test in the desert
Night fell, desert-hard and blacker than black as we huddled together on a big patch of concrete at an old missile site in the middle of New Mexico to test out a new component for ASTERIA. I was more and more certain of its value. It wasn’t Hubble or Spitzer or Kepler, and it might never be something so magnificent. But not every painting should or could be Starry Night. There is room in the universe for smaller work, a different kind of art. Kepler might find thousands of new worlds, but it wouldn’t reveal enough of any single one of them for us to know whether it was somebody’s home. It was sweeping its eye across star fields that were too far away for astronomers to make anything more than assumptions about places like Kepler-22b.
But if I could just make ASTERIA work, and then find a way to send up a fleet of satellites, it would combine the best outcomes of NASA’s Kepler space telescope, capable of finding smaller planets around sun-like stars, and the nascent TESS, with its more proximate search and sensitivity to red dwarf stars.
My team built a prototype for a possible camera, one that was promisingly stable and could operate at a warmer temperature than the detectors used in most satellites. (Most have to be cooled, which taxes the machine.) I just wasn’t sure that it would see what we needed it to see. I had a particularly bright and enthusiastic grad student at the time, named Mary Knapp; she had been an undergraduate in the first design-and-build class I taught. She encouraged us to test the camera outside, using it to look at real stars. Mary proposed the deserts of New Mexico as our proving ground. That April, there would be a new moon, casting the already clear desert sky an even pitcher black. That new moon also coincided with school break for my sons, Max and Alex, which meant that I could take them along. As much as I wanted to see the stars, I wanted to see them, too.
I had asked a local club of amateur astronomers where the best place to test our camera might be. That night they invited us to their star-viewing party, a celebration of the new moon. We arrived at dusk at the old missile site. I looked up at the stars and felt my childlike wonder return. I think the boys felt it too.
We set up the camera. We would have to wait until we were back at MIT to analyze our data, but our new type of detector, one not yet used for astronomy, seemed to do the trick. We knew at least that our experiment wasn’t a total failure.
A long-awaited launch
In August 2017, after years of work and hope and effort, SpaceX prepared to launch a Falcon 9 rocket into space. The rocket didn’t have a crew, but ASTERIA was on board.
It had been a difficult journey. The camera had made its way from my imagination to our design-and-build class, through drawings and prototypes and an old missile site in New Mexico. Then we’d run out of money at MIT, and Draper Laboratory had liked the technology better for other things. The Jet Propulsion Laboratory, which had always been interested in the possibilities of cubesats and ASTERIA in particular, picked up where MIT and Draper left off. Three MIT graduates there would play leading roles on the project; they took their work seriously, having seen firsthand how much it mattered. Their passion and expertise made sure that ASTERIA would become everything it could be, that it was built right and lovingly placed, at last, into the hold of a rocket, groaning on the launchpad on a beautiful late-summer day. The rocket would slice into the sky and rendezvous with the International Space Station. The astronauts there would set our little satellite free later in the fall. From a whisper in my dreams to space: I couldn’t believe that we were nearing the end of such a long reckoning.
I had planned on going to the ASTERIA launch, but it was delayed just long enough for travel and child-care plans to fall through. On the day of the launch, I took the train into Cambridge instead, walked to the Green Building, and took the elevator to my floor. I walked past the travel posters for distant worlds into my office, shut the door, and called up the online video stream. The launch was a big deal; all over the world, eyes were trained on that rocket, still waiting on the pad.
Every now and then I looked up from the cloudless Florida footage on my screen and out my windows, at my crystalline view of downtown Boston. There were clear skies everywhere I looked. I spent maybe 30 minutes in the quiet, writing thank-you emails to other members of the ASTERIA team. At the last second I decided not to send them. I know that superstition is unscientific. I understand that it doesn’t matter to the universe if a baseball player is wearing his lucky underwear—whether he gets a hit is mostly up to the pitcher and to him. But rockets are delicate, ill-tempered machines. Before the Russians launch rockets from the steppes of Kazakhstan into orbit, they summon an Orthodox priest to throw holy water at the boosters, his beard and cloak and the holy water carried sideways by the wind. I wasn’t going that far, but I wasn’t going to send a couple of emails until we were safely weightless. I was surprised by how nervous I was, watching the countdown clock tick down to launch.
The engines ignited with a great big ball of pure fire. The launch tower fell away, and the rocket eased its way off the pad, gained speed, and pushed its shining shoulders toward its future orbit. The onboard cameras recorded its arching flight as the sky around it went from blue to purple to black. The rocket had broken through into space. The boosters were jettisoned, and the remainder of the rocket continued its climb into the deepest possible night, the Earth blue and alight behind it, an impossible blackness ahead. It would take a little while for it to catch up with the space station, which was racing its own way through orbit at 17,000 miles an hour, about five miles every second. But the rocket, and our satellite, were well on their way.
Everything brave has to start somewhere, I thought.
Do I believe in other life in the universe?
Yes, I believe.
The better question: What does our search for it say about us? It says we’re curious. It says we’re hopeful. It says we’re capable of wonder and of wonderful things.
Adapted and reprinted from The Smallest Lights in the Universe. Copyright © 2020 by Sara Seager. Published by Crown, an imprint of the Random House Publishing Group, a division of Penguin Random House.
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.