Biosignatures

The idea: chemistry that shouldn't be there

Imagine you could not visit a planet, could not photograph its surface, could not land a probe — but you could split the thin sliver of starlight that grazes its atmosphere into a spectrum. Could you tell whether anything lives there? The startling answer is: sometimes, yes. Life is a chemical engine. It eats some molecules and excretes others, and if it does this on a planetary scale for long enough, it tips the whole atmosphere out of the balance that lifeless physics would otherwise impose. A biosignature is the imprint of that imbalance — a gas, a surface colour, or a seasonal cycle whose abundance can only plausibly be explained by a living source.

The single most powerful idea in the field is chemical disequilibrium. A dead planet's atmosphere settles toward thermodynamic equilibrium: reactive gases consume each other until what remains is stable. Earth's air refuses to do this. It holds 20.9% molecular oxygen — an aggressively reactive gas — alongside 1.9 parts per million of methane, a gas that oxygen destroys. These two should not coexist. The only reason they do is that Earth's biosphere replenishes both, every single day, faster than the atmosphere can burn them away. That ongoing fight against equilibrium is the loudest thing about our planet, and it is what we hope to hear from another.

The mechanism: thermodynamic disequilibrium

The rigorous way to state "this chemistry shouldn't be there" is to compute the atmosphere's distance from chemical equilibrium. You take the observed mixing ratios, find the equilibrium composition the same atoms would reach at the same temperature and pressure, and measure the Gibbs free energy released in going from the observed state to equilibrium. For Earth, that available free energy is dominated by the O₂–CH₄ couple, and it is enormous compared to other Solar System planets. The governing reaction is simply combustion:

CH₄ + 2 O₂  →  CO₂ + 2 H₂O      ΔG ≈ −818 kJ/mol

Left to itself this reaction runs to completion: methane is photochemically oxidised in Earth's atmosphere on a timescale of roughly

τ_CH₄ ≈ 8–10 years

So if methane were not being resupplied, it would fall below detectability within a human lifetime. To hold it steady at 1.9 ppm against that loss requires a global source flux of order 500–600 megatonnes per year — on Earth, supplied by methanogenic microbes, wetlands, and (now) agriculture. The key insight from Lovelock (1965) and later Hitchcock & Lovelock (1967) is that simultaneous abundances of mutually destructive gases imply continuous production; the larger and more reactive the disequilibrium, the harder it is to sustain without biology.

The detectable surrogate for O₂ in the near-infrared is often ozone (O₃), which is a nonlinear tracer of oxygen — even a few percent of Earth's O₂ level produces a nearly full-strength ozone layer, making O₃ a sensitive but saturating proxy. Together, an O₂/O₃ + reduced-gas (CH₄ or N₂O) combination remains the most defensible atmospheric biosignature we know how to look for.

The key numbers

Biosignature science lives and dies on small numbers. Here are the ones that matter:

• Earth's O₂: 20.9% by volume — built up over the 2.4-billion-year-ago Great Oxidation Event and a second rise ~0.8 Gyr ago.
• Earth's CH₄: ~1.9 ppm today (pre-industrial ~0.7 ppm), with a ~10-year lifetime.
• Earth's N₂O: ~0.33 ppm — a microbial denitrification product and an underrated biosignature with little abiotic background.
• Key spectral bands: O₂ A-band at 0.76 µm; O₂ B-band at 0.69 µm; O₃ Hartley–Huggins UV band and the 9.6 µm thermal band; CH₄ near 3.3 and 7.7 µm; CO₂ at 4.3 and 15 µm; H₂O across the near-IR.
• Red-edge contrast: vegetation reflectance jumps roughly five-fold between ~680 nm and ~750 nm.
• Transit signal size: the atmospheric annulus of an Earth twin around a Sun-like star changes the transit depth by only ~1 ppm — the fundamental reason Earth analogues are so hard.
• Scale height: H = kT/(µg) ≈ 8.5 km for Earth — the vertical thickness of the absorbing annulus, and the lever that sets the signal.

How biosignatures are observed

The dominant technique is transmission spectroscopy. When a planet transits its star, a ring of starlight passes through the planet's upper atmosphere on its way to us. Gases absorb at their signature wavelengths, so the planet's effective radius — and therefore the transit depth — increases at exactly those colours. The extra transit depth from one atmospheric scale height is approximately

Δδ ≈ 2 R_p H / R_star²        (signal per scale height)
H  = k T / (µ g)               (atmospheric scale height)

where R_p is planet radius, R_star the stellar radius, T the temperature, µ the mean molecular mass, and g surface gravity. The smaller and cooler the host star, the larger R_p/R_star and the deeper the signal — which is why the first rocky-planet biosignature searches target Earth-sized worlds around M dwarfs (the TRAPPIST-1 system) and small temperate planets around K dwarfs.

For planets that do not transit, the future lies in direct imaging: block the star with a coronagraph or a free-flying starshade, isolate the planet's faint reflected light (contrast ~10⁻¹⁰ for an Earth at visible wavelengths), and take its reflection or thermal-emission spectrum. That is the strategy of the Habitable Worlds Observatory. Nulling interferometry in the mid-infrared, where the O₃ 9.6 µm and CH₄ 7.7 µm bands sit, is the long-discussed alternative (the cancelled Darwin and Terrestrial Planet Finder concepts; the ESA-studied LIFE mission).

Comparison: the major biosignature classes

Biosignature	Where	Spectral handle	Strength	Main false positive
O₂ + CH₄ disequilibrium	Atmosphere	0.76 µm + 3.3/7.7 µm	Strongest known	Hard to fake jointly
O₂ / O₃ alone	Atmosphere	0.76 µm; 9.6 µm O₃	Strong, needs context	Ocean loss on M dwarf
Nitrous oxide (N₂O)	Atmosphere	7.8, 8.5, 17 µm	Low abiotic background	Lightning, stellar UV
Methane alone	Atmosphere	3.3, 7.7 µm	Weak by itself	Serpentinization, volcanism
Dimethyl sulfide (DMS)	Atmosphere	~3.4, 9.6 µm	Speculative, no abiotic source known	Photochemical haze confusion
Red edge	Surface	~700 nm reflectance rise	Direct but model-dependent	Mineral / scattering slopes
Seasonal CO₂/CH₄ cycle	Temporal	Band depth vs. orbital phase	Compelling if seen	Volatile sublimation cycles
Glint / specular ocean	Surface	Phase-dependent brightening	Habitability indicator	Specular ice, cloud tops

No single row in this table is decisive on its own. The discipline's hard-won lesson is that a credible detection is a system-level argument: the right gas, in the right context (liquid water, the right host star, no obvious abiotic source), surviving every false-positive vetting test you can throw at it.

Worked example: detecting Earth's oxygen from afar

Suppose an alien astronomer watches Earth transit the Sun and wants to detect our O₂. How big is the signal? Earth's atmospheric scale height is

H = kT/(µg)
  = (1.38×10⁻²³ × 255 K) / (4.8×10⁻²⁶ kg × 9.81 m/s²)
  ≈ 7.5 km            (using T = 255 K, µ = 29 amu)

The transit-depth boost across, say, ~5 scale heights of O₂ A-band absorption is

Δδ ≈ 2 R_⊕ (5H) / R_☉²
   = 2 × (6.37×10⁶ m) × (3.75×10⁴ m) / (6.96×10⁸ m)²
   ≈ 9.9×10⁻⁷  ≈ 1 ppm

One part per million. For comparison, JWST's best-demonstrated transit-spectroscopy precision is tens of ppm per spectral bin, so a single Earth transit is hopelessly faint — and Earth transits the Sun only once a year, lasting ~13 hours. You would need to co-add of order 100 transits, i.e. a century of observing, to build the signal-to-noise — which is precisely why detecting O₂ on a true Earth twin is reserved for a dedicated reflected-light mission rather than transit spectroscopy. By contrast, an Earth-sized planet in the habitable zone of the M dwarf TRAPPIST-1 (R_star ≈ 0.12 R_☉) yields a transit signal ~70× larger, ~70 ppm — within reach, which is why those systems are observed first even though their false-positive risks are higher.

History: from Lovelock to JWST

The conceptual foundation was laid in the 1960s. While consulting for NASA on how to detect life on Mars, James Lovelock argued in 1965 that the surest signature would be atmospheric chemical disequilibrium — and noted that Mars's air, near chemical equilibrium and dominated by CO₂, looked dead, a prediction the Viking landers (1976) broadly confirmed. The disequilibrium idea, formalised with Dian Hitchcock in 1967, became a cornerstone of astrobiology and seeded Lovelock's later Gaia hypothesis.

The modern observational era began with the first transiting-exoplanet atmosphere detection (sodium on HD 209458 b, Charbonneau et al. 2002 with Hubble). The 1990s spawned ambitious mission concepts — NASA's Terrestrial Planet Finder and ESA's Darwin — both eventually cancelled. The James Webb Space Telescope (launched 25 December 2021) transformed the field: it measured CO₂ on the hot Saturn WASP-39 b in 2022, and in 2023 detected CO₂ and CH₄ — plus a contested, low-significance hint of dimethyl sulfide — on the temperate sub-Neptune K2-18 b (Madhusudhan et al.), the first time a possible biosignature gas was discussed for a habitable-zone world. The 2020 US Astrophysics Decadal Survey then recommended the flagship Habitable Worlds Observatory (HWO), a ~6-metre UV/optical/IR telescope explicitly designed to image and characterise ~25 potentially Earth-like planets and search them for O₂, O₃, and water — the mission that could finally close the loop on oxygen.

Variants and related signatures

• Surface biosignatures. The vegetation red edge near 700 nm and pigment colours from photosynthesis or photoprotective compounds — spatially unresolved "colour of a living world."
• Temporal biosignatures. Seasonal modulation of CH₄, CO₂, or O₂ band depths as a planet's hemispheres tilt through their year — Earth's CO₂ "breathing" via the Keeling-curve sawtooth is the archetype.
• Antibiosignatures. Gases whose abundance argues against life, e.g. high CO together with abundant O₂, which a biosphere would normally scrub — used to down-weight tentative claims.
• Agnostic biosignatures. Signs of life-as-we-don't-know-it: complexity, unexpected chemical patterning, or molecular assembly indices that flag non-random chemistry without assuming Earth biochemistry.
• Technosignatures. Evidence of technology rather than biology — radio leakage, lasers, industrial CFCs and NO₂, megastructures, or waste heat. A biosignature implies life; a technosignature implies a civilisation.
• In-situ biosignatures. Within the Solar System — isotopic fractionation, lipid biomarkers, chirality, or morphological fossils sought by Mars rovers and proposed Europa/Enceladus missions, where you can sample directly rather than read a spectrum.

Common misconceptions and subtleties

• "Find oxygen, find life." No single gas is proof. Abiotic O₂ can build up when an M-dwarf planet loses its oceans to a runaway greenhouse: water is photodissociated and hydrogen escapes, leaving hundreds of bars of O₂ behind. Context — host star, water, the full gas suite — is mandatory.
• "Methane means microbes." Methane has strong abiotic sources: volcanic outgassing and serpentinization (water reacting with olivine-rich rock) both make CH₄. Methane alone is at best a weak hint; it is the company it keeps (oxygen) that matters.
• "Equilibrium tells you nothing." It tells you a great deal — Mars, Venus, and Titan all sit far closer to equilibrium than Earth, and that contrast is itself the diagnostic. Disequilibrium magnitude must be weighed against plausible abiotic drivers (Titan's CH₄–H₂ disequilibrium is photochemical, not biological).
• "JWST will image continents." JWST sees planets only as a single unresolved point; "imaging" an exoplanet surface is beyond any planned instrument. All atmospheric biosignatures are disk-integrated, one pixel of light.
• "One detection settles it." A biosignature claim is a statistical, multi-year argument. The K2-18 b DMS hint is a cautionary tale: low statistical significance, alternative non-biological interpretations, and active debate over whether the planet is even a water world or a mini-Neptune with no surface. Extraordinary claims demand reproducible spectra and exhausted false positives.