Exoplanet Biosignatures

The idea: a planet's air can be a tell

Imagine you could sniff the atmosphere of a world orbiting another star. On a dead, geologically quiet planet you would expect the air to settle into chemical equilibrium — every reactive molecule eventually meets its partner and the mixture goes inert. Earth's atmosphere does the opposite. It is wildly out of equilibrium: 21% oxygen sitting next to methane, nitrous oxide, and a trace of other reactive gases that, left alone, would react away in years to decades. Something is constantly resupplying them. That something is the biosphere.

A biosignature is any observable that is much easier to explain with life than without it. For exoplanets the most powerful biosignatures are gaseous, because gases mix into the whole atmosphere and imprint themselves on the starlight we can actually measure. The single most compelling case is not one gas but a pair held in mutual disequilibrium — oxygen and methane — because no known geology can keep both topped up at once. James Lovelock and Joshua Lederberg made exactly this argument in the mid-1960s while NASA was designing life-detection experiments for Mars: don't look for a specific organism, look for an atmosphere that thermodynamics says shouldn't exist.

Why disequilibrium is the real signal

Oxygen and methane are chemically incompatible. In an atmosphere with sunlight and water vapour they react via the net reaction

CH₄ + 2 O₂  →  CO₂ + 2 H₂O

On Earth this destroys atmospheric methane on a timescale of roughly a decade. Yet methane sits steadily at about 1.9 parts per million while oxygen holds at 21% — a coexistence that demands continuous replenishment. The biological fluxes that do this are large: photosynthesis releases of order 10¹³ molecules of O₂ per cm² per second globally averaged, and biogenic methane sources (methanogens in wetlands, guts, sediments) supply on the order of 10¹¹ CH₄ molecules per cm² per second. Remove life and the methane is gone within decades; the oxygen, lacking its biological source, drifts back into rocks and the ocean over millions of years.

The strength of the disequilibrium can be quantified as the Gibbs free energy you could in principle extract by letting the atmosphere relax to equilibrium. Krissansen-Totton and colleagues computed this for Solar System atmospheres: Earth's atmosphere–ocean system carries of order a few thousand joules per mole of available free energy, dominated by the N₂–O₂–liquid-water system, and is a striking outlier compared with Mars, Venus, or the giant planets. They argued that the more uniquely biological signal is the simultaneous presence of oxygen and methane — a pairing whose coexistence, rather than its raw free energy, is the tell. That excess free energy is itself a biosignature — an "agnostic" one that does not assume any particular metabolism.

How we read the air: transmission spectroscopy

We never image these molecules directly for a small planet. Instead we watch the planet cross the face of its star. During this transit, a thin annulus of the planet's atmosphere — a few scale heights thick — is backlit by the star. Molecules in that ring absorb at their characteristic wavelengths, so the planet blocks slightly more light at those wavelengths and the transit looks deeper. Plotting transit depth versus wavelength gives a transmission spectrum whose bumps are molecular fingerprints.

The size of the signal is set by the atmospheric scale height H = k_B T / (μ g), where μ is the mean molecular mass and g surface gravity. The extra absorbing area is roughly an annulus of width ~5–10 H around the planet, so the spectral feature amplitude is

δ ≈ 2 n H R_p / R_star²        (n ≈ 5–10 scale heights)

For an Earth-twin (H ≈ 8 km, R_p = 6371 km) transiting a Sun-like star (R_star = 696,000 km), one scale height changes the transit depth by only about 0.2 parts per million, and even integrating the full ~5–10-scale-height absorbing annulus the strongest features reach only about 1–2 ppm. That is the brutal number behind the whole field: you must measure a roughly one-part-per-million dip-within-a-dip in starlight, repeatedly, to build signal-to-noise. It is far easier around small stars, where R_star is smaller and the same atmosphere blocks a larger fraction — which is why M-dwarf planets dominate the target lists.

Candidate biosignature gases

No single gas is a slam dunk; each has biotic and abiotic stories. The community evaluates them by detectability, the strength of the biological case, and the severity of false positives.

Gas	Biological source	Key abiotic source (false positive)	IR feature	Strength of case
Oxygen (O₂)	Oxygenic photosynthesis	Water photolysis + H escape	0.76 µm (A-band)	Strong only with context
Ozone (O₃)	Photochemical proxy for O₂	Same as O₂	9.6 µm	Strong proxy, saturates
Methane (CH₄)	Methanogenesis	Serpentinization, volcanism	3.3, 7.7 µm	Weak alone, strong with O₂
Nitrous oxide (N₂O)	Denitrifying bacteria	Lightning, stellar particles	7.8, 17 µm	Strong, few abiotic sources
Dimethyl sulfide (DMS)	Marine phytoplankton	Cometary delivery (minor)	3.4, 9–13 µm	Promising, under study
Phosphine (PH₃)	Anaerobic microbes (claimed)	Lightning, deep convection	4.3, 8.9 µm	Contested (Venus 2020)
Chloromethane (CH₃Cl)	Marine algae, fungi	Negligible	9.7, 13.7 µm	Clean but faint

The pattern is clear: gases with weak or absent abiotic sources (N₂O, CH₃Cl) are cleaner in principle but faint and hard to detect; gases that are easy to detect (O₂, CH₄) have real abiotic mimics. This is why the field has shifted from hunting one "magic" molecule to demanding a self-consistent set of gases interpreted in planetary context.

The quantified reality of the search

Concrete figures sharpen why this is at the edge of the possible:

• Signal depth. Earth-twin around the Sun: ~0.2 ppm per scale height, or roughly 1–2 ppm for the strongest integrated features. The same atmosphere around an M dwarf like TRAPPIST-1 (R_star ≈ 0.12 R_☉) gives a signal roughly (1/0.12)² ≈ 70× larger in fractional terms.
• Photon budget. Reaching a few-ppm detection of CO₂ or CH₄ on a TRAPPIST-1 planet with JWST takes on the order of tens of transits stacked — dozens of hours of one of the most oversubscribed telescopes in history.
• Habitable-zone flux. The classic Kasting habitable zone for the Sun runs from about 0.95 to 1.67 AU; the conservative inner edge (moist greenhouse) sits near 0.99 AU. For an M dwarf it shrinks to ~0.02–0.06 AU, with orbital periods of days.
• Disequilibrium energy. Earth's atmosphere–ocean free-energy disequilibrium is of order 10³ J per mole, dominated by the N₂–O₂–water system, versus near-zero for Mars and Venus; the diagnostic biological marker is the coexisting O₂–CH₄ pair rather than its modest energy.
• Timescales. Atmospheric methane lifetime against O₂: ~10 years. Oxygen drawdown without a source: ~10⁶–10⁷ years. Abiotic O₂ build-up by hydrogen escape: tens to hundreds of millions of years on a desiccating world.
• Targets. The TRAPPIST-1 system (about 40.5 light-years) holds seven Earth-sized planets, three or four in the habitable zone; K2-18b is 124 light-years away at 8.6 Earth masses.

False positives — the field's central problem

The hardest scientific work is not detecting a gas but proving life made it. Each candidate has a non-biological pathway that must be ruled out:

• Abiotic oxygen from water loss. A planet undergoing a runaway greenhouse photolyses water; the freed hydrogen, being light, escapes to space (Jeans and hydrodynamic escape), leaving oxygen behind. Hundreds of bars of abiotic O₂ can build up. This is the leading worry for hot rocky planets around active M dwarfs.
• Oxygen from CO₂ photolysis. On a dry, CO₂-rich world, UV can split CO₂ into CO and O, and O₂ accumulates if there is nothing to recombine it. The tell is co-detected carbon monoxide, which a biosphere would scavenge — so a CO-rich O₂ atmosphere argues against life.
• Geological methane. Serpentinization — water reacting with iron-rich ultramafic rock — produces hydrogen that drives methane formation. Volcanic and hydrothermal outgassing add more. Methane alone is therefore a weak biosignature.
• Stellar contamination. Star spots and faculae imprint their own spectral features on a transmission spectrum, mimicking or masking planetary signals — a major systematic for active M dwarfs.

The accepted defense is the "biosignature in context" framework: never interpret one molecule in isolation. Demand a chemically inconsistent set of gases (O₂ + CH₄, or O₂ without the accompanying CO of an abiotic world), check the planet's mass, irradiation, and evolutionary history, and assign a confidence level rather than a binary yes/no. NASA's 2018 biosignature reports and the proposed "CoLD" (Confidence of Life Detection) scale formalize this caution.

Where the search stands: real targets

• K2-18b. A 8.6-Earth-mass sub-Neptune in the habitable zone of an M dwarf, 124 light-years away. JWST detected CH₄ and CO₂ (2023) consistent with a possible "Hycean" ocean world, plus a tentative, disputed hint of dimethyl sulfide strengthened but not confirmed in 2025 data. It is the most-watched biosignature target — and a case study in how hard significance thresholds are.
• TRAPPIST-1 planets. Seven Earth-sized worlds around an ultracool dwarf about 40.5 light-years away, several in the habitable zone. JWST is testing whether the inner planets even retain atmospheres; early results for TRAPPIST-1 b and c suggest little to no thick atmosphere, a sobering data point for M-dwarf habitability.
• Venus phosphine (a cautionary tale). The 2020 claim of phosphine in Venus's clouds — proposed as a possible biosignature — was challenged on data-reduction grounds and remains contested. It illustrates how a single-line detection at the edge of the noise can ignite, and then deflate, an entire debate.
• Earth as the calibration target. Earthshine spectroscopy and observations of Earth transiting the Sun (seen from spacecraft) provide the only ground truth we have: an unambiguous O₂–O₃–CH₄–N₂O–H₂O biosphere fingerprint that future missions aim to recognize elsewhere.

The instruments doing the work

Biosignature science is instrument-limited, and the roadmap is explicit:

• JWST (operating since 2022). Transmission and emission spectroscopy from 0.6 to 28 µm. It can reach CO₂, CH₄, and H₂O on favourable sub-Neptunes and is straining to detect anything on rocky M-dwarf planets. Not designed to detect O₂ at 0.76 µm with the sensitivity needed for a true twin-Earth.
• Ground-based ELTs (late 2020s). The 39-m Extremely Large Telescope and the GMT/TMT will use high-resolution cross-correlation spectroscopy to chase O₂ at 0.76 µm on the nearest M-dwarf planets — a path that beats the few-ppm photon problem by spectrally resolving thousands of individual lines.
• Habitable Worlds Observatory (2040s, proposed). The flagship recommended by Astro2020. A ~6-m UV/optical/IR space telescope with an internal coronagraph or starshade to directly image reflected light from ~25 potentially habitable Earth-sized planets and search them for O₂, O₃, H₂O, and CH₄. This is the mission designed from the ground up to find a biosignature, not stumble on one.
• LIFE (study concept). A proposed mid-IR space interferometer to capture thermal emission (the 9.6 µm O₃ and 7.8 µm N₂O bands) and measure a planet's temperature directly.

Common misconceptions and edge cases

• "Finding oxygen means finding life." No. Oxygen is the most famous biosignature precisely because it is also the most famous false positive. Abiotic O₂ from water loss or CO₂ photolysis is a real and expected outcome on some planets. Oxygen is only persuasive with corroborating evidence.
• "A biosignature is a smoking gun." Modern practice treats detection as a confidence spectrum, not a yes/no. A first claim of life will almost certainly arrive as a contested, several-sigma anomaly that takes years and independent confirmation to settle — exactly what is playing out for K2-18b.
• "We'll image the aliens' cities or vegetation." For the foreseeable future these planets are unresolved points of light. The "vegetation red edge" is a real surface biosignature, but it is a few-percent step in disk-integrated reflectance, not a picture of forests, and alien photosynthesis may peak at a different wavelength.
• "Methane on Mars means Martian life." The variable methane Curiosity detects in Gale crater is more likely geological or seasonal than biological; methane is the textbook example of a gas with strong abiotic sources.
• "The same gases work for any planet." Biosignature interpretation is host-star and planet specific. On an M dwarf, the high UV flux changes the photochemistry so that O₂ and CH₄ can coexist abiotically in ways they cannot around a Sun-like star — the context, not the molecule list, decides.
• "Equilibrium chemistry can be ignored once a gas is detected." The opposite: the whole inferential power comes from chemistry that should not coexist. A detection that turns out to be thermodynamically comfortable is, by definition, a weak biosignature.