Caltech, April 2026: The Astronomer Who Found 47 Exoplanets. None Like Earth.

On a Tuesday afternoon in late April, Dr. Jessie Christiansen sits in her office at Caltech's Infrared Processing and Analysis Center with three computer monitors arranged in a shallow arc before her. The leftmost screen displays a scatter plot of 5,437 confirmed exoplanets—worlds orbiting stars other than our Sun—colour-coded by detection method. The middle monitor shows raw spectroscopic data from the James Webb Space Telescope, downloaded six hours earlier from the Space Telescope Science Institute in Baltimore. The right screen has her email open. She ignores it.

Christiansen is the project scientist for NASA's Exoplanet Archive, the world's authoritative catalogue of planets beyond our solar system. She has personally validated 47 of them. On her desk, beneath a stack of peer-review drafts and a coffee mug from the 2019 American Astronomical Society meeting, lies a printout of a paper she published in The Astrophysical Journal three weeks ago. The title: "Atmospheric Biosignatures on TRAPPIST-1e: Non-Detection and Implications for Habitability Assessment." It is, she tells me, the most depressing paper she has ever written.

"We thought TRAPPIST-1e was it," she says. "Earth-sized. Rocky. In the habitable zone. We pointed Webb at it for 27 hours of observation time—do you know how expensive that is in telescope hours?—and we found nothing. No oxygen. No methane. No water vapour we could confirm. Just... nothing."

She clicks a tab on the middle monitor. A spectrum appears: a graph plotting wavelength against brightness, the fingerprint of light filtered through an alien atmosphere 40 light-years away. There are dips and peaks, absorption lines where molecules should have left their mark. But the lines are shallow, ambiguous, contaminated by stellar noise from TRAPPIST-1 itself—a red dwarf star prone to flares.

"People don't understand what 'habitable zone' means," Christiansen says. "It just means the planet is at the right distance from its star for liquid water to exist on the surface—if it has an atmosphere, if it has water, if the star hasn't stripped it all away. It's a necessary condition. Not a sufficient one."

The Catalogue

Christiansen arrived at Caltech in 2012, the year after NASA's Kepler Space Telescope began flooding astronomers with planetary candidates. Kepler worked by staring at a single patch of sky containing 150,000 stars and watching for the tiny dimming that occurs when a planet crosses in front of its host star—a transit. The telescope found thousands of these dips. Christiansen's job was to determine which were real planets and which were instrumental noise, eclipsing binary stars, or stellar variability masquerading as planets.

She pulls up a file labelled "KOI-7016.01"—Kepler Object of Interest 7016, candidate 01. It was flagged by the Kepler pipeline software in March 2014. The light curve shows a periodic dip every 19.3 days, consistent with a planet roughly 1.6 times Earth's radius orbiting a Sun-like star 1,200 light-years away in the constellation Lyra. Christiansen spent six weeks on it. She checked the pixel data to rule out background stars. She cross-referenced Gaia parallax measurements to confirm the star's distance. She ran statistical validation tests to calculate the false-positive probability. In June 2014, she certified it as Kepler-452b.

NASA held a press conference. News outlets called it "Earth's cousin." The discovery was published in The Astronomical Journal. Christiansen was listed as third author. "That was the high point," she says. "We thought we were on the verge of finding another Earth. We thought it was just a matter of time and telescope hours."

The archive Christiansen maintains is a living document, updated daily as new observations come in from ground-based surveys, the Transiting Exoplanet Survey Satellite (TESS), and the James Webb Space Telescope. Each entry includes orbital period, planetary radius, equilibrium temperature, detection method, and discovery date. The database is public. High school students use it for science fair projects. Graduate students mine it for doctoral theses. Christiansen uses it to watch the field she helped build struggle with a problem no one anticipated a decade ago.

"We found the planets," she says. "Thousands of them. We just can't study them in the detail we need. Not with current technology."

The Problem of Red Dwarfs

Most of the potentially habitable planets in Christiansen's archive orbit red dwarf stars—M-type stars smaller and cooler than the Sun. This is not an accident. Red dwarfs make up 75 per cent of the stars in the Milky Way. They are dim, which means their habitable zones are close in—planets orbit every few days or weeks, not years—and close-in planets are easier to detect. They transit more frequently. Their gravitational tug on the host star is stronger, making them easier to find with radial velocity measurements.

But red dwarfs are violent. They flare unpredictably, releasing bursts of ultraviolet and X-ray radiation that can strip away planetary atmospheres. Proxima Centauri b, the nearest exoplanet to Earth at 4.24 light-years, orbits a red dwarf that flares every few days with energy equivalent to a solar superflare. In March 2023, a team led by Dr. Meredith MacGregor at the University of Colorado published observations showing Proxima b likely lost any atmosphere it once had billions of years ago.

Christiansen's March 2026 paper on TRAPPIST-1e was the culmination of a four-year observational campaign targeting the TRAPPIST-1 system, a red dwarf 40 light-years away with seven Earth-sized planets, three of which orbit in the habitable zone. The system was discovered in 2017 by a Belgian-led team using the TRAPPIST telescope in Chile. It became the poster child for exoplanet habitability. NASA commissioned an artist's rendering showing the view from the surface of TRAPPIST-1e: a red sun low on the horizon, sibling planets hanging in the sky.

James Webb observed TRAPPIST-1e in November 2025, using its Near-Infrared Spectrograph to analyse the light filtering through the planet's atmosphere during transit. The team, which included Christiansen, expected to find water vapour at minimum. The planet is the right size. It receives roughly the same amount of starlight as Earth. Thermal models suggested it should have retained an atmosphere.

The spectrum came back flat. No absorption features. Either the planet has no atmosphere, or it has an atmosphere so thin and cloud-covered that Webb cannot detect its composition. Follow-up observations in January 2026 confirmed the result. The paper was submitted to The Astrophysical Journal in February, peer-reviewed in three weeks, and published on April 2. Christiansen was lead author.

"The reviewers asked if we'd considered instrumental systematics," she says. "We had. We spent six months modelling every possible source of noise. The non-detection is real."

Looking in the Wrong Place

On a Thursday morning in mid-April, I meet Dr. Eliza Kempton in her office at the University of Maryland, 2,600 miles east of Caltech. Kempton is a theoretical astrophysicist who models exoplanet atmospheres. She has spent the past decade trying to understand what James Webb is seeing—and what it is not seeing.

"We optimised our search for the wrong kind of planet," she says. "Kepler and TESS are transit surveys. Transit surveys are biased toward short-period planets around small stars. Short-period planets around small stars are tidally locked—one side always faces the star—and tidally locked planets around active stars lose their atmospheres. We built a detection pipeline that led us directly to worlds we can't study."

Kempton is part of a working group convened by NASA's Exoplanet Exploration Program to reassess habitability criteria. The group's preliminary report, circulated internally in February 2026, argues that the field has conflated detectability with habitability. The easiest planets to find—small worlds around nearby red dwarfs—are precisely the planets least likely to support life. The planets most likely to support life—Earth-sized worlds around Sun-like stars—are too far away and orbit too slowly for current instruments to characterise.

"It's a selection effect," Kempton says. "We designed a method that was extraordinarily good at finding one kind of planet, and then we spent fifteen years being surprised that those planets don't look like Earth."

The problem is technological and economic. Detecting an Earth analogue—a planet the size of Earth, orbiting a Sun-like star at Earth's distance—requires observing at least one full orbit. For Earth, that is 365 days. For most Sun-like stars visible to current telescopes, it means multi-year campaigns with space-based observatories that cost hundreds of millions of dollars to operate. Kepler found a handful of such planets, but they are hundreds or thousands of light-years away, too faint for Webb to study in detail.

Meanwhile, the red dwarf planets keep arriving. In January 2026, the TESS mission announced the discovery of TOI-5205 b, an Earth-sized planet orbiting an M-dwarf 280 light-years away. In February, a team at the European Southern Observatory reported LP 890-9c, a rocky planet in the habitable zone of a red dwarf 105 light-years distant. Both are candidates for Webb follow-up. Neither is likely to have an atmosphere.

The Next Telescope

In March 2026, NASA convened a workshop at the Goddard Space Flight Center in Greenbelt, Maryland, to discuss the design parameters for the Habitable Worlds Observatory—a proposed space telescope intended to succeed James Webb in the search for life beyond Earth. The mission, still in the planning phase, has a projected launch date of 2039 and an estimated cost of $11 billion.

Christiansen attended remotely from Pasadena. The discussion centred on mirror diameter and wavelength coverage. To directly image an Earth-sized planet around a Sun-like star—to see the planet as a distinct point of light rather than inferring its presence from a transit or radial velocity wobble—requires a mirror at least six metres across and a coronagraph capable of blocking the star's light by a factor of ten billion. James Webb's mirror is 6.5 metres, but it observes in the infrared. Habitable Worlds would observe in optical and ultraviolet wavelengths, where biosignatures like oxygen are more easily detected.

The Goddard workshop produced a 47-page report, delivered to NASA headquarters on April 15. The report's executive summary includes a single-sentence recommendation: "A mission capable of detecting and characterising Earth analogues around Sun-like stars cannot be accomplished with current or near-term technology within existing budget constraints." The report is not public. Christiansen forwarded me a copy on the condition I not publish it in full.

"They're telling us we need to wait," she says. "Another decade of planning. Another decade of construction. Another decade of commissioning. I'll be retired before Habitable Worlds launches. The planets will still be there. We just won't know what's on them."

What the Archive Contains

Back in Pasadena, Christiansen shows me a query she runs every Monday morning. She filters the Exoplanet Archive for planets meeting four criteria: radius between 0.8 and 1.25 Earth radii, equilibrium temperature between 200 and 320 Kelvin, host star within 100 light-years, and confirmed detection. The query returns 23 planets. She has requested Webb observation time for six of them. She has been awarded time for two.

One of those two is LTT 1445Ac, a rocky planet orbiting a red dwarf 22 light-years away in the constellation Eridanus. Webb will observe it in November 2026, weather and instrument health permitting. The observation will take 18 hours of telescope time. The data will arrive in Pasadena three days later. Christiansen will spend six months analysing it. If she finds water vapour, it will be the first confirmed detection of water on a terrestrial exoplanet in the habitable zone. If she finds oxygen, it will change the field overnight.

"I'm not optimistic," she says. "LTT 1445A is an M-dwarf. Active. We're going to see the same thing we saw at TRAPPIST-1. But we have to check. That's what the archive is for—so the next generation knows what we already looked at."

She clicks back to the leftmost monitor, the scatter plot of 5,437 exoplanets. Most are gas giants, hot Jupiters orbiting close to their stars. A smaller cluster represents the rocky planets, colour-coded green if they are in the habitable zone. She zooms in. The green dots are scattered across the plot, spanning dozens of stellar types and hundreds of light-years. None have confirmed atmospheres. None have biosignatures.

"This is what we have," Christiansen says. "Five thousand planets. A handful we can study. None that look like home."

Tuesday Afternoon in Pasadena

It is nearly five o'clock when I leave Christiansen's office. Outside, the San Gabriel Mountains are backlit by a sky turning pink and orange. The Infrared Processing and Analysis Center sits on the northwest corner of Caltech's campus, a low building with narrow windows and a white façade that reflects the late afternoon sun.

Before I go, I ask Christiansen if she ever looks up at the night sky and tries to pick out the stars she has studied. She laughs. "Not anymore," she says. "Most of them are too faint. You need a telescope just to see the star, let alone the planet. And the ones I can see—Proxima Centauri, TRAPPIST-1, LTT 1445—they're in the southern hemisphere. I can't see them from here."

She walks me to the door, then pauses. "Do you know what the hardest part is?" she asks. "It's not that we haven't found life. It's that we've found all these worlds, and we can't answer the simplest question about them: Do they have air? Do they have water? Are they alive or dead? We built this incredible catalogue, and it turns out a catalogue isn't enough. You need to go there. Or at least get close enough to really see."

I ask when she thinks that will happen—when we will have the technology to definitively characterise an Earth-like planet around a Sun-like star.

She considers the question for what feels like a long time. "2045," she says finally. "Maybe 2050. If the funding comes through. If the engineering works. If we don't lose another generation to budget cuts and mission delays." She shrugs. "I'll be sixty-eight. I'll probably still be here, updating the archive. Adding planets we can see but not touch."

The printout of her TRAPPIST-1e paper is still on her desk when I leave, half-buried under new drafts and observation proposals. The abstract is visible: "We report non-detection of atmospheric water vapour on TRAPPIST-1e using JWST/NIRSpec. These results suggest that terrestrial planets in the habitable zones of M-dwarf stars may lack the conditions necessary for long-term atmospheric retention." The paper has been cited fourteen times in three weeks. It will be cited hundreds more. It is, Christiansen told me, the kind of null result that changes how an entire field thinks about its future.

On the scatter plot still glowing on her monitor, 5,437 exoplanets hang in digital space, catalogued, confirmed, and—for now—unreachable. The archive will be updated again tomorrow. The number will tick upward. The green dots will multiply. And somewhere, 22 or 40 or 280 light-years away, a world no one has ever seen continues its orbit, indifferent to whether anyone is watching.