Reading alien air: why gases are astronomers' best bet for finding life

The Short Version

Astronomers look for life on distant planets by reading their air, because an atmosphere is the one part of such a world that can be measured from light years away [3]. Sara Seager, who has spent her career on the problem, says the task is "to look for gases that don't belong that we might attribute to life," and that the answer will come as a hint and never as certainty [1]. David Kipping explains why the classic sign, oxygen, can mislead in both directions, and why even the best telescope would need years on a single star [2]. Adam Frank describes what a stronger case looks like: gases that should have reacted away, an atmosphere far out of chemical balance, or an industrial compound nothing but technology makes [3]. A credible detection, on their combined account, is a specific gas or set of gases, on a planet whose conditions are understood, after every non-living source has been ruled out, confirmed by more than one team.

Table of contents

What They Said

Sara Seager: gases as the method, and its four problems

Seager, an MIT planetary scientist, frames the whole search in practical terms. We will have the capability to start answering whether we are alone within a decade or two, she says, and what it takes is space telescopes, or very sophisticated ground telescopes, able to study the atmospheres of other planets, look for water, "which is needed for life as we know it," and look for gases that don't belong [1]. She calls it "really nitty gritty astronomy" [1].

Why gases. The first assumption, she says, is that life everywhere uses chemistry, as it does on Earth: it eats, breathes and runs a metabolism that stores and uses energy, leaving waste products in the atmosphere [1]. She notes the assumption is not guaranteed; one could imagine life running on something like mechanical energy that never produces a gas [1]. The idea itself is old. At least almost a hundred years ago there is a record of someone proposing a gas as an indicator of life elsewhere [1].

The model case. Oxygen fills Earth's atmosphere to about 20 percent by volume, and without plants and photosynthesizing bacteria, which produce it as a waste product, we would have virtually none [1]. "Oxygen is a huge indicator," she says [1]. She imagines the test run in reverse. A civilization around a nearby star, with the kind of telescopes we are trying to build, looks at our Sun, sees Earth and sees oxygen. They would not be completely sure life was making it, but "if they go through all the possible scenarios, they'll be left with a pretty strong hint that there's life here" [1].

Problem one: the gas has to survive. The harder question, and the one her team studies, is what happens to a gas after life makes it [1]. Much of it is destroyed right away by ultraviolet radiation or by oxygen, which she calls "incredibly destructive" to many gases [1]. A gas that life produces in quantity may never build up enough to be seen.

Problem two: the gas has to be informative. Life might make what she calls a boring gas. Humans breathe out carbon dioxide, but Earth already has plenty of it, so it tells a distant observer nothing [1].

Problem three: something else might make it. Researchers have to rule out gases produced by non-living processes [1]. A big part of this growing field, she says, is "how to make sure that this gas isn't produced by another effect" [1].

Problem four: reach. There are hundreds of billions of stars in the galaxy, but in practice the search has to use stars quite close to Earth, a few light years to tens or maybe hundreds of light years away [1]. For detailed follow-up in this lifetime she counts perhaps a dozen or two dozen suitable stars [1]. And the planets themselves are brutally faint: an Earth is so small and so faint next to its star that finding one is "the proverbial needle in a haystack, but worse," a job for very sophisticated space telescopes costing billions of dollars [1].

Water as a second reading. We cannot see an ocean on a planet that is less than a pixel across. But on a small rocky planet, water vapor in the atmosphere suggests a liquid reservoir, because ultraviolet light breaks water apart and the hydrogen escapes to space, so the vapor must keep being replenished [1]. Water matters because life needs a liquid in which molecules can react, and water is the most abundant liquid and stays liquid over the widest range of temperatures and pressures [1].

Her equation. Seager reworked the Drake equation, with Frank Drake's blessing, to count something observable: the number of planets that show signs of life through gases in their atmospheres [1]. Its terms run from what can be measured to what cannot. First is the number of stars a specific survey can access; the TESS mission, for example, includes about 30,000 red dwarf stars [1]. Then the fraction of those stars that are quiet, since active stars brighten and dim and hide planets; then the fraction of planets we can observe, which for transiting planets is a matter of geometry; then the fraction in the habitable zone, which the Kepler telescope let astronomers estimate for certain types of star [1]. The two unknowns come last: the fraction of those planets that have life, and the fraction whose life produces a gas that accumulates and can be detected with spectroscopy [1]. She keeps them separate precisely because of the boring-gas problem [1].

Asked for her own estimate of that last fraction, she gives the honest range. "Like on the worst days, it's 1%," she says, and on the best days "it's like 80%" [1]. She calls the question almost her life's work, now and for the next 20 years [1].

What the answer will be. She likes Fridman's word, hint: "It's not going to be a 100% yay we found it." Future generations will have to do more careful work, perhaps even find a way to send a probe [1]. The field, she admits, relies on what she calls a "dream condition," that if the ingredients for life are present, life will form [1].

David Kipping: how the reading is taken, and how oxygen misleads

Kipping, an astronomer at Columbia, supplies the mechanics and the cautions [2].

How the measurement works. The main way to study a small planet is to watch it cross its star. An Earth crossing a Sun-like star dims it by only 84 parts per million, which he likens to a firefly flying in front of a giant stadium floodlight [2]. Smaller stars make detection easier, because the same planet blocks more of a small star's light [2]. TRAPPIST-1, with seven planets smaller than Earth, can be studied even from the ground, and TRAPPIST-1e, about 90 percent of Earth's size and 80 percent of its mass in the liquid-water zone, is a top target [2].

How long it takes. Detecting a marginal biosignature on TRAPPIST-1e with the James Webb Space Telescope would take on the order of 200 transits, roughly four years of staring at one star, which he says will never happen with a shared telescope whose time is oversubscribed many times over [2]. Earth-like orbits around Sun-like stars are harder still: such planets rarely line up, transit rarely, and need three or four years of watching just to be found; the Kepler mission ended after about 4.35 years and found zero true Earth twins around Sun-like stars [2].

Why oxygen is not enough. Oxygen was long seen as the key sign of life, since without photosynthesis Earth would lose its oxygen within about a million years [2]. Kipping gives two reasons it can fail. Most of Earth's history had life without oxygen, so a living planet could show none. And oxygen can appear without life, for example when ultraviolet light splits water vapor and the hydrogen escapes to space [2]. The second is the same process Seager uses to infer oceans [1]. So one mechanism can both point to water and counterfeit a sign of life.

The wider list. Researchers now also consider nitrous oxide, which microbes make, methane combined with other gases, and phosphine [2].

Two worked examples of ambiguity. Phosphine was claimed in the atmosphere of Venus about two years before his conversation, and may instead have been sulfur dioxide [2]. Methane seen on Mars could be biological or could be leaking seasonally from an underground reservoir [2]. Both are in our own solar system, where we can send spacecraft, and both remain unsettled.

A trap in the image. A future picture of an Earth-like world would blend planet and moon into one pixel. A Titan-like moon's methane, combined with oxygen from water-splitting on a lifeless planet, could fake a biosignature [2].

The stars that are easiest are also the roughest. Small red dwarfs can take a billion years or more to calm down, blasting their planets with X-rays and ultraviolet light that could strip atmospheres and water [2]. Even so, he predicts that any near-term biosignature detection will come from a red dwarf planet and not a true Earth twin [2].

The discipline. Kipping's rule for himself is "the more I want something to be true, the more I inherently doubt it." [2] When his own team found strong signs of a moon around a distant planet, they spent months trying to kill the signal, and when a journal asked them to call it a discovery he insisted on calling it evidence [2]. He warns that a claim of alien life would bring years of back-and-forth [2]. He cites Percival Lowell, who was convinced Mars was inhabited and saw canals that were not there [2]. Asked what would actually convince him of intelligence, he names something mathematical, such as a sequence of prime numbers, which no natural process produces [2].

Adam Frank: what a stronger signal looks like

Frank, an astrophysicist who leads NASA's first grant to search for technosignatures on exoplanets, explains why the search is organized around atmospheres and what kinds of evidence would be hardest to explain away [3].

Why atmospheres, and why the habitable zone. The habitable zone is the band of orbits where liquid water could pool on a planet's surface [3]. Frank agrees it is too strict as a definition of where life could be: Jupiter's moon Europa, far outside it, has a hundred-mile-deep ocean under about ten miles of ice [3]. But that is exactly the point for detection. He would not be able to see life on Europa from across light years, because it is under the ice [3]. Planets in the habitable zone are thought to have atmospheres, and "Atmospheres are the things we can characterize for across 10, 50 light years." [3]

What counts as a biosignature. He describes it as something that rock and air alone would not produce [3]. His standard example is a pair: oxygen together with methane, which react away quickly unless life keeps replenishing them [3].

Signatures that do not assume Earth's chemistry. Frank lists approaches he calls agnostic. One looks at the structure of the network of chemical reactions, since biology produces very different networks from rock and water [3]. Another applies information theory to the compounds in an atmosphere, asking whether there is too much information there for geology alone [3]. A third measures chemical disequilibrium using what is called the Gibbs free energy. A dead system relaxes toward equilibrium, so "if you see an atmosphere that is wildly out of equilibrium, that indicates that there's something happening on that planet," a biosphere or a technosphere pumping gases in and keeping the whole system from relaxing [3]. He notes that this works even for chemicals you have no idea are connected to life [3].

The cleanest case is industrial. Chlorofluorocarbons could be detected in an alien atmosphere across interstellar distances by passing the planet's light through a spectrograph and seeing their spectral fingerprint [3]. That would settle the matter in a way oxygen cannot, he argues, because "there's no other way to make chlorofluorocarbons except through some kind of industrial process" [3].

Where the instruments stand. The James Webb Space Telescope's real power, he says, is in spectra and not pretty pictures. Its data on the planet K2-18b, about 120 light years away, suggest a possible world with a hydrogen atmosphere over a warm ocean, with methane detected at five sigma [3]. He expects hard data that starts to put limits on how common life is within the next 10 to 30 years [3].

Why the planet has to be understood too. Frank's larger theme is that life and planet evolve together. He quotes colleagues: "Life doesn't happen on a planet. It happens to a planet." [3] Oxygen in Earth's air is the product of photosynthesis invented about two and a half billion years ago, and plate tectonics, volcanism and erosion shaped what life could do [3]. A gas reading therefore means little without a model of the world it comes from. Fridman draws the consequence for the search: if life and planet co-evolve, it is much harder to look out and work out which planets are producing life [3].

Why one detection would be enough to matter. We have one example of life, so we could be an accident. "Get one more example and now you're done," he says, because then you know it has happened more than once and can start reasoning that life is not hard to make [3].

The standard. Speaking of claims about alien technology, he says scientists are deliberately hard on each other about evidence, and that if his own team announced a technosignature they would expect to be shredded: "You don't get a pass because it's a really cool topic." [3]

Where They Agree, and Where They Don't

They agree on the method. Spectra of atmospheres, taken with space telescopes, are the realistic route to detecting life on planets around other stars [1][2][3].

They agree it will arrive as a probability. Seager says hint [1]. Kipping says evidence and expects years of dispute [2]. Frank talks about data that begins to put limits on how common life is [3].

They agree the main danger is a false positive. Each describes non-living sources as the central problem, and each gives examples [1][2][3].

They differ on how much weight oxygen can carry. Seager calls it a huge indicator, with the condition that other scenarios be worked through [1]. Kipping stresses that it can be absent on a living world and present on a dead one [2]. Frank treats it as convincing in combination with methane [3]. These are compatible once the conditions are read closely: oxygen alone is a lead, oxygen with a gas it should destroy is a much stronger one, and either needs the planet's context. The difference is one of emphasis, and it follows from their roles. Seager is describing why the approach works at all, Kipping is describing how it has already fooled people, and Frank is describing how to make it robust.

They differ on a single gas versus a pattern. Seager's work is on individual gases: which ones life makes, which survive, which are ambiguous [1]. Frank puts more hope in properties of the whole atmosphere, such as disequilibrium, that do not depend on guessing alien biochemistry [3]. Seager's caution that life might not use chemistry the way we expect is a reason to want Frank's agnostic measures, and her boring-gas problem is a reason they are hard: a subtle imbalance is harder to measure than one strong line [1][3].

They differ on where the first sign will come from. Seager's hope is an Earth-like planet around a Sun-like star with signs of water and gases that don't belong, and she accepts that planets around small red dwarfs are the easier near-term targets: "think like Earth cousin instead of Earth twin." [1] Kipping predicts the first detection will be at a red dwarf, while warning those stars may have stripped their planets [2]. Frank points to K2-18b, a kind of planet with no counterpart in our solar system [3]. All three are describing the same constraint, that the easiest planets to measure are the least like Earth.

They differ on timing. Seager says a decade or two to begin answering [1]. Frank says 10 to 30 years for hard limits [3]. Kipping's four-years-per-star arithmetic is the reason to take the long end seriously with current telescopes [2].

They differ on what they expect to find. Seager says she believes absolutely that there is life somewhere, given the size of the universe [1]. Frank's gut feeling is that simple life is common [3]. Kipping refuses to prefer any probability for life arising, because there is no rigorous basis for one [2]. The method does not depend on who is right, and Kipping's position explains why it must be applied without a thumb on the scale.

They differ on biology versus technology as the cleaner target. Frank argues an industrial gas would be less ambiguous than any biological one [3]. Kipping also finds a technological or mathematical signal the most convincing kind [2]. Seager still loves the search for signals and hopes someone is sending one [1]. But technosignatures require a civilization, and both Frank and Seager suspect simple life is far more common than that [1][3].

Why It Works

Light carries a chemical fingerprint. Passing a planet's light through a spectrograph shows which gases are present, each with its own spectral lines [3].

Atmospheres are global and visible. A biosphere that alters its planet's air advertises itself to anyone with a large enough telescope, as Earth's oxygen does [1]. Life hidden under ice does not [3].

Replenishment is the logic of the whole method. A gas that should be destroyed or should react away, yet persists, implies a source that keeps making it. Seager applies this to water vapor and to gases destroyed by ultraviolet light and oxygen; Frank applies it to oxygen with methane and generalizes it as disequilibrium [1][3]. Kipping's figure that Earth would lose its oxygen in about a million years without photosynthesis is the same argument with a number on it [2].

Uniqueness of source is what makes a gas decisive. Chlorofluorocarbons are useful as a sign because only industry makes them [3]. Oxygen is weaker because water-splitting makes it too [2]. A candidate gas is only as good as the list of other things that produce it is short.

One confirmed case changes the statistics. With a single example of life, nothing can be said about how common it is; with a second, independent one, it is known not to be a fluke [3].

Where It Applies, and Where It Doesn't

The method applies to planets with atmospheres, orbiting stars near enough and quiet enough to measure, that either transit or can be imaged directly [1][2][3].

It misses life that does not reach the atmosphere. Europa's ocean could be inhabited and no telescope would know [3]. Kipping suggests that for such places the route is instruments sent to work on site and report back [2].

It misses life that makes no distinctive gas, or one that is destroyed as fast as it is made, or one the planet already has in abundance [1]. It would miss life that does not run on chemistry at all [1].

It can be fooled by geology and photochemistry, by an unresolved moon, and by the observer's hopes [2]. Venus and Mars show that even nearby claims stay open for years [2].

It is limited by telescope time as much as by physics. Four years on one star is beyond what a shared observatory will grant, which is why Kipping hopes cheaper launches could give fields their own telescopes [2].

And it rests on an untested premise. Seager's dream condition, that ingredients lead to life, may be wrong, in which case a careful survey returns nothing and that result is itself the finding [1].

What To Do With It

Use this as a checklist when a headline says signs of life have been found on another planet.

1. Find out which gas, and how strongly it was detected. Frank cites methane on K2-18b at five sigma; Kipping's exomoon signal was also about five sigma and he still called it evidence [2][3]. Ask whether a second team has recovered it.

2. Ask what else makes that gas. Oxygen can come from water vapor split by ultraviolet light; methane on Mars may be geological; a claimed phosphine signal on Venus may have been sulfur dioxide [2]. A credible paper spends most of its length on this.

3. Look for a second gas that should not coexist with the first. Oxygen with methane is the example. A pair that would react away is much harder to explain without a continuing source [3].

4. Check whether the atmosphere as a whole is out of balance. Disequilibrium is a signature that does not depend on Earth's particular biology [3].

5. Ask about the star. Is it quiet? Red dwarfs are easier to observe and harder on their planets [1][2].

6. Ask about water. Water vapor on a small rocky planet suggests a surface reservoir, which supports the case that the world is habitable at all [1].

7. Ask whether a moon could be contributing. A single-pixel image blends planet and moon [2].

8. Note the word used. Seager says hint, Kipping says evidence [1][2]. Treat "discovery" in a headline as the publisher's word until the authors use it.

9. Expect a long argument, and count that as normal. Kipping predicts years of back-and-forth for any such claim, and Frank says a team making one should expect to be shredded [2][3].

10. Treat an industrial gas differently. If the claim is a compound with no known natural source, such as chlorofluorocarbons, the question shifts from what else could make it to whether it was really measured [3].

Sources

1. Sara Seager on the Search for Exoplanets, Signs of Life, and Finding Light After Loss · Lex Fridman Podcast
2. David Kipping on Exomoons, Habitable Worlds and Why We Must Stay Agnostic About Aliens · Lex Fridman Podcast
3. Alien Civilizations and the Search for Extraterrestrial Life · Lex Fridman Podcast