The thrill of exoplanet characterization
As told by Sara Seager

From my first encounter with the dark night sky at age ten, I was mesmerized. On a camping trip in Ontario, Canada, I stumbled out of the tent late one night and looked up. My heart skipped a beat. The sky, ablaze with countless stars, seemed at once within reach and impossibly distant, igniting a fascination with the mysteries of the Universe.

Yet, my childhood was not full of wonder, but instead a living nightmare because of a tyrannical stepfather who mentally tortured me and my siblings at every opportunity. Weekend escapes with my father, who earnestly entertained even the most outlandish ideas, from reincarnation to the power of positive thinking, meant that at a young age I poured through countless library books in order to form my own opinions. Thus, my childhood formative experiences instilled in me a spirit of independent thought, openness to unconventional ideas, deep resilience, a skeptical view of authority, and a disdain for the status quo, which, unbeknownst to me at the time, are critical characteristics for pushing the boundaries of science.

After high school in inner city Toronto, I pursued a degree in math and physics at the University of Toronto. During two summers at the David Dunlap Observatory, I cataloged variable stars and managed a nightly observing program. I moved to the USA for graduate studies at Harvard in 1994, and soon after the first exoplanet orbiting a Sun-like star was announced. 51 Peg b shattered the paradigm of planetary system architectures by its extremely close proximity to its star—seven times closer to its star than Mercury is to our Sun. By the summer of 1996, just as I was exploring potential Ph.D. thesis topics, a handful of these so-called “hot Jupiters” had been identified. My research supervisor, Dimitar Sasselov, entrusted me with the pioneering project of computer modeling hot Jupiter atmospheres. These planets, with orbital separations of about 0.05 AU or less, are subjected to temperatures of 1000–2000 K or higher, presenting a new frontier in planetary science.

"It seemed as though every astronomer I told about my thesis topic advised me against it."

Theoretical models and descriptive predictions
Study of exoplanetary atmospheres was considered risky in the mid 1990s because at the time most astronomers were skeptical of the new planets’ very existence, since the signals were products of indirect detection techniques. In theory, the exoplanet signatures could have been illusions caused by phenomena like stellar variability. It seemed as though every astronomer I told about my thesis topic advised me against it. Even those who believed that exoplanets were real assured me that exoplanetary atmospheres would never be observable. Always a risk-taker—and in no way committed to a career in science—I had nothing to lose, so I forged ahead. My Ph.D. thesis was one of the first on exoplanets, and ultimately provided the first predictive descriptions of hot Jupiters under strong stellar irradiation.

By September 1999, I was a newly minted Ph.D. headed for a postdoctoral fellowship at the Institute for Advanced Study (IAS) in Princeton, New Jersey (USA). Surrounded by those who had never met an exoplanet researcher before, I was persistently asked, ‘‘What’s the next big thing?’’ It was crystal clear to me that any day a transiting planet could be discovered—and that such a discovery would change everything. A planet going in front and behind its star as seen from Earth would open up possibilities for studying the planet’s atmosphere, as well as many other physical properties. Starlight would filter through the transiting planet’s atmosphere, imprinting the spectral signatures of atmospheric gases in its passage.

The resulting signal promised to be large enough for measurement with current instruments, so I drafted a paper on my idea of observing exoplanetary atmospheric transmission spectra. As soon as the first transiting planet was announced a couple of months later, I dropped everything to complete my paper. My theoretical work set the stage for the first observation of an exoplanet’s atmosphere, emphasizing sodium as a detectable gas at visible wavelengths. An observing group’s proposal and observation of sodium gas in HD 209458b’s transmission spectrum with the Hubble Space Telescope were successful — ushering in the birth of the brand-new field of exoplanet atmospheres.

I continued to develop theoretical models and apply them to descriptive predictions for many exoplanet characterization techniques, nearly all being the first ever on the topic: exoplanetary scattered light illumination phase curves and polarization signatures; estimating a planet’s rotation rate from a projected oblateness measurement; atmospheric photochemistry; Earth-like exoplanet cloud and rotation rate characterization by diurnal photometric variability; atmosphereless planet identification by the hotspot of its thermal phase curve; and more. I also worked with observational teams for telescope time proposals and subsequent interpretation for the limited cases where this was possible—one exciting example was the first detection of a hot Jupiter secondary eclipse to yield an atmospheric temperature.

At the time, it felt like so few others really cared. Even so-called experts thought either the field would soon dead end for lack of bright targets or that a legit observational field was too far off in the future to warrant much attention. Fortunately, at the IAS, our legendary astrophysics leader, John Bahcall, had a visionary and bold philosophy. He believed that as long as the underlying physics is sound and well-developed, and the phenomenon conceivably detectable within one’s lifetime, an astrophysics topic is worth pursuing. I am grateful for this extraordinary enabling attitude.

"My style is to pioneer new ideas, not to perpetuate established topics."

Innovated methods
My approach to pioneering new techniques in exoplanet characterization involves nurturing nascent ideas over extended periods, sometimes spanning months or even years, until the time is right for execution. One favorite example involved elucidating the ambiguous nature of mini-Neptunes—planets sized between Earth and Neptune, characterized by their nebulous average densities. I had already initiated and been collaborating on this project with my then-graduate student Leslie Rogers. We were able to swiftly publish our findings following the initial discovery of this novel and subsequently prevalent planet type, one which lacks a solar system counterpart.

Alongside other graduate students, I developed innovative methods, such as with Julien de Wit, where we proposed to determine a planet's mass from its transmission spectra by analyzing atmospheric “scale height” which includes surface gravity. My student Nikku Madhusudhan and I introduced the concept of “atmospheric retrieval”, an inverse method to quantify physical properties from observational data. Combined with transmission spectra, this now serves as a cornerstone in contemporary exoplanet atmospheric studies. Despite today's students preferring to avoid risky projects, it is these challenges that lead to major breakthroughs and advance our understanding.

Post-traumatic growth
Backing up to my career timeline, after my postdoc at the IAS (1999-2002), I spent a few years on staff (2002-2006) at the Carnegie Institute of Science—one of the only places hiring in exoplanets during those early years. MIT recruited me once it was clear exoplanets were here to stay and a leader was needed to initiate this now burgeoning field university-wide. A few years later, my husband fell ill with late-stage cancer. We found ourselves asking a harrowing question. Which is worse: being the one forced to come to terms with a terminal illness and an early death, or the one who survives, left to pick up the pieces and manage the crushing, debilitating grief? Widowed at age 40 and left to raise our two young children alone, it took many years to find my own answer to this deeply distressing question. Moreover, his death plunged me into a deep professional crisis by a catastrophic shift in my sense of purpose.

Intensifying my professional crisis, the field of exoplanet research was rapidly advancing, becoming increasingly crowded with duplicated studies. My style is to pioneer new ideas, not to perpetuate established topics. Thus, I needed to focus my research portfolio. I chose the most potentially impactful topic I could envision: discovering another Earth, one marked by signs of life through atmospheric biosignature gases. I redoubled my efforts on two of my major research initiatives: compiling an exhaustive list of potential biosignature gases and their detectability with futuristic telescopes, and the quest to find an Earth analog: an Earth-size planet in an Earth-like orbit about a Sun-like star.

One concept I've come to appreciate is post-traumatic growth. I often recalled my late father's affinity for the power of positive thinking. Every day, I would repeat to myself: “It will get better. I may find a new best friend, and someday, I will be even happier than before.” With this mindset, many amazing opportunities unfolded.

One bright wintry day, I accidentally met Melissa at our local kids’ sledding hill and through her joined a newly formed group of young widows in our town. My kids’ babysitters and my housekeeper became extended family, so instrumental in helping me survive those dark times. At a large gathering in Thunder Bay, Canada, I spotted a tall, dark, handsome man across the room. It was love at first sight—straight out of a movie, only it happened at an amateur astronomy conference. Charles and I married soon after, and he adopted my two boys.

Studying the Starshade
On the professional front, a pivotal growth moment in my career came with an invitation to lead a new NASA Mission Concept study on Starshade. Starshade is a large, specially-shaped screen designed to fly tens of thousands of kilometers from a space telescope, blocking starlight so that only planet light can enter the telescope. This innovation allows the use of standard space telescopes without special stability features. Although conceived in the 1960s and revived nearly every decade since, Starshade was still deemed unfeasible as of 2012. My team pushed it into a viable mission concept, leading to significant technological advancements and lab-demonstrated planet-star flux contrast meeting the 10⁻¹⁰ requirement at small planet-star separations.

Currently, NASA is prioritizing the coronagraph, a more efficient planet-finding device that suppresses the starlight inside the telescope. If and when the coronagraph can meet the necessary planet-star contrast and exquisite telescope wavefront stability, its low throughput and narrow spectral bands will still make it far less efficient for exoplanet atmosphere characterization compared to Starshade. With its high throughput at all wavelengths, simultaneous broad spectral range, and independence from super-stable telescope requirements, Starshade remains the most capable tool for characterizing terrestrial exoplanet spectra. It is now being advanced by non-mainstream groups, poised for future revitalization.

"discovering an Earth analog is incredibly challenging due to its proximity to the bright host star."

Without question, discovering an Earth analog is incredibly challenging due to its proximity to the bright host star. Earth is 10,000 times smaller in area, 300,000 times less massive, and 10 billion times fainter than the Sun. I am dedicating my unwavering efforts to this pursuit, doggedly pursuing each of these threads. For planet area (transits), I have spearheaded the development of a telescope constellation, each unit focusing on a single bright Sun-like star. This strategy overcomes the limitations of a single telescope tracking multiple distant stars simultaneously. My MIT team in collaboration with Draper Laboratory solved the key technical challenges for achieving precision pointing—about 100 times better than previous satellites of similar mass. A NASA Jet Propulsion Laboratory team (which included some of my newly graduated MIT students) took over the project and built and successfully operated the constellation prototype, “ASTERIA”, on orbit from 2017 to 2019. For planet mass, my ERC Synergy Team REVEAL is tackling the impact of stellar variability on mass measurements, utilizing advanced stellar models. For planet faintness, I continue to develop and support both ground- and space-based direct imaging missions that suppress starlight to isolate exoplanetary signals.

The search for life on other planets
While we await further technological advancements for discovering an Earth analog, many of us in the astronomy community are exploring different types of planets for signs of life using today's James Webb Space Telescope and its transmission spectroscopy capabilities. These are the “Earth cousins,” planets orbiting small red dwarf stars—far more observationally accessible than an Earth analog in nearly every aspect. I am actively working to overcome challenges such as contamination from the host stars, since M dwarf stars are notoriously magnetically active, with star spots and other surface inhomogeneities that interfere with accurate measurements.

My decades-long biosignature gas work has yielded a list of options, one favorite being phosphine. On Earth phosphine is highly thermodynamically disfavored and is only associated with life. Because of this, my group joined Prof. Jane Greaves’ team to search for phosphine as a potential biosignature in the Venus atmosphere. Our subsequent report of the detection of phosphine gas at part-per-billion levels could not be explained by any known non-biological processes, such as volcanism or meteoritic activity, thus leaving open the possibility of microbial life in the Venusian clouds. Yet, this reported discovery was attacked on every level. Is the signal real? If real, is the signal attributed to the right gas? If real and properly attributed, is the gas presence due to life or a non-life mechanism? This “reality-check” experience was a prescient wake up call for our future of exoplanet biosignature gas searches. If we struggle to agree on signal detection, attribution, and source for a planet right next door, how can we have any hope to have a clear, robust finding on an exoplanet where we will know next to nothing about the detailed atmosphere context, surface, or mantle and interior conditions?

While setting aside this conundrum to mull over, I have gone full force on the search for life on Venus, one that echoes the early days of exoplanets, with many questioning its pursuit. For context, while the Venus surface is too hot for life, the temperature drops with increasing altitude above the surface, just as on Earth. High above the surface, at 48 to 60 km in Venus’ permanent cloud deck, temperatures are conducive to life. However, these clouds consist not of water, but of a toxic chemical: concentrated sulfuric acid.

My core team (Janusz Petkowski, Max Seager, and I) conducts experiments on the stability of biological molecules in sulfuric acid, yielding astonishing results. We aim to demonstrate that an informational polymer is stable in sulfuric acid, to motivate astrobiology-focused space missions to Venus. To that end I lead the Morningstar Missions to Venus, which additionally aim to revolutionize our approach to space missions through more frequent, focused, and cost-effective explorations.

The shifting line between crazy and mainstream
My new work is still often rejected as unorthodox, while most of my offbeat work that led to this Kavli Prize is now as mainstream as it gets. Hence my maxim, “In exoplanets and the search for life beyond Earth, the line between what is considered mainstream and what is crazy is constantly shifting.” I hope this inspires scientists to spend some part of their time pursuing bold new ideas, especiallythose that are joyful and thatmay lead to the unparalleled thrill of discovery.