Climates of Terrestrial Exoplanets and Biosignatures

Abstract: Understanding the climates of terrestrial exoplanets and the detectability of biosignatures is an inherently interdisciplinary challenge, requiring the integration of insights from Solar System exploration, exoplanet observations and climate science. Building from Earth as the only known inhabited planet, NCCR PlanetS has developed models, tools and observational strategies to assess planetary environments far beyond direct reach. Between 2018 and 2025, PlanetS made major contributions across theory, modelling, instrumentation and mission preparation. On the modelling side, the Generic Planetary Climate Model enabled climate studies across a wide range of planetary regimes, from early Venus to temperate terrestrial exoplanets including Proxima b, incorporating advanced developments such as a dynamical slab ocean. In parallel, the THOR global climate model was developed to avoid Earth-centric assumptions and to stably simulate diverse atmospheric regimes. PlanetS has also advanced atmospheric retrieval techniques combining forward modelling, Bayesian inference and machine learning, applied to targets ranging from Solar System bodies to exoplanet phase curves and directly imaged spectra. These efforts have helped assess the scientific return of future missions, notably the Large Interferometer for Exoplanets (LIFE) and to define instrumental requirements for detecting Earth-like atmospheres and biosignatures. Within the Solar System, PlanetS contributed key technologies for biosignature detection, including ORIGIN and SenseLife, enabling in-situ and remote detection of organics, isotopic ratios and microstructures. Finally, PlanetS has played a major role in preparing the next generation of observatories, from JWST, VLT and ELT instruments to LIFE and the Habitable Worlds Observatory. Together, these contributions form an integrated framework advancing the search for life beyond Earth.

• The paper reviews climate modeling and habitability of terrestrial exoplanets, summarizing contributions from NCCR PlanetS from 2014-2025.
• Strict definitions for habitable zones and biosignatures are provided, highlighting contextual factors like atmospheric composition, pressure, and rotation.
• For inclusion into Earth's Middle Boughs, nitrogen and carbon dioxide concentrations almost exclusively included into weather models predicting early Mars behavior and heating.

Scope and motivation

This chapter, authored by a large consortium of researchers affiliated with the Swiss National Centre of Competence in Research (NCCR) PlanetS, reviews the state of research on the climates of terrestrial exoplanets and on biosignature detection, with emphasis on work carried out within PlanetS between 2014 and 2025. The chapter is organized around two pillars: (i) climate modeling and habitability assessment, spanning Solar System bodies (Earth, Mars, Venus, ocean worlds) and exoplanets; and (ii) biosignatures—what they are, how they can be falsified by abiotic processes, and how current and future instruments can detect them. The authors frame the problem against the backdrop of more than 5800 known exoplanets, most of which are hotter and larger than Earth due to observational bias, and note that an "Earth 2.0" around a Sun-like star remains undetected; small temperate planets are currently found predominantly around M dwarfs, such as Proxima Centauri b and the TRAPPIST-1 system.

The habitable zone concept and its limits

The chapter devotes substantial attention to the Habitable Zone (HZ), defined as the region where a planet can sustain surface liquid water, bounded inward by the runaway greenhouse or moist greenhouse limits and outward by the maximum greenhouse limit. The authors emphasize that HZ boundaries are not fixed quantities but depend strongly on atmospheric composition, pressure, rotation state, and stellar type. Several specific results sharpen this point:

• Background gas effects: Increasing N$_2$ pressure pushes the inner HZ edge closer to the star, whereas adding even modest CO$_2$ to an otherwise pure N$_2$ atmosphere has the opposite effect [(Faria et al., 2022)-style citation omitted; see (Chaverot et al., 2021)]. For greenhouse-gas backgrounds such as pure CO$_2$ or pure H$_2$, increasing pressure moves the inner boundary outward [(Leung et al., 19 Feb 2025)-adjacent work cited as 2025Life...15...79K]. Collision-induced absorption in H$_2$-dominated atmospheres can extend the outer edge substantially (Lous et al., 2022).
• Rotation and tidal locking: Slow rotators sustain habitable conditions at higher irradiation than fast rotators because of dayside cloud feedbacks (Yang et al., 2013). Conversely, tidally locked planets face nightside atmospheric collapse when the major constituent condenses; analytical models developed within PlanetS capture the key physics of heat redistribution and collapse pressure and show that stability decreases with increasing planet mass (Auclair-Desrotour et al., 2020).
• Carbonate-silicate cycle caveats: At low surface temperatures, silicate weathering may switch from a negative to a positive climate feedback, destabilizing rather than stabilizing the climate (Hakim et al., 2020). Peridotite weathers an order of magnitude faster than granite, implying that rock composition strongly modulates the carbon cycle on exoplanets.

A particularly consequential result is the demonstration that Earth's present insolation (~340.5 W m$^{-2}$) exceeds the water condensation threshold (~312.5 W m$^{-2}$), meaning present-day Earth sits near a bifurcation point with three possible stable climates: the current state, a snowball, and a steam state (Turbet et al., 2021). This reframes the faint young Sun paradox: the lower solar luminosity 4 Gyr ago was what allowed oceans to condense in the first place. Relatedly, the same GCM study showed that on early Venus, water clouds form preferentially on the nightside and produce net warming, preventing ocean condensation—a direct challenge to earlier claims that Venus was once habitable (Way et al., 2016), and one the chapter presents as evidence that Venus likely never had surface liquid water.

Climate modeling tools

The chapter describes a hierarchy of models developed or used within PlanetS:

Model class	Representative tools	Primary use
1-D radiative-convective	Line-by-line RCE codes	HZ boundaries, runaway greenhouse onset
1-D energy balance	Rust-based EBM (Univ. Geneva)	Snowball dynamics, stochastic perturbations
3-D GCMs	THOR, Generic-PCM (LMDZ-derived)	Full climate states, clouds, oceans
Retrievals	petitRADTRANS + MultiNest (LIFE framework)	Inferring atmospheres from spectra

THOR is highlighted as the first open-source GCM built from scratch for exoplanetary applications, solving the non-hydrostatic Euler equations on an icosahedral grid (Deitrick et al., 2019). The Generic-PCM is noted as the only GCM currently treating water vapor as a major atmospheric constituent, which is essential for post-runaway and water-rich planets. A newly implemented dynamical slab ocean module—including Ekman transport, sea ice evolution, and convective adjustment—reproduces present-day Earth's mean surface temperature (13 °C), planetary albedo (0.32), and sea-ice extent (~20 million km²), providing a computationally cheap alternative to fully dynamic ocean GCMs. Applied to TRAPPIST-1e, this model shows that ocean heat transport converts the classical "eyeball" climate into a "lobster"-like pattern analogous to results obtained for Proxima Centauri b with full dynamic oceans.

Two methodological results deserve emphasis. First, assuming radiative-convective equilibrium profiles in 1-D models of thick steam atmospheres overestimates surface temperatures by several hundred Kelvin (Selsis et al., 2023)—a result the authors describe as requiring a revision of standard 1-D modeling practice for post-runaway planets. Second, the runaway greenhouse transition, modeled in 3-D for the first time, is shown to be effectively irreversible once triggered, owing to radiative imbalance and changed cloud behavior (Chaverot et al., 2023).

Atmospheric escape and volatile loss

The chapter treats atmospheric escape as a first-order constraint on habitability, particularly around active M dwarfs. HST Lyman-$\alpha$ observations of TRAPPIST-1 showed much weaker Lyman-$\alpha$ emission than Proxima Centauri despite comparable X-ray output, and derived mass-loss estimates suggest the inner planets could have been stripped within a few billion years, while TRAPPIST-1e may have lost fewer than three Earth oceans if hydrodynamic escape ceased upon HZ entry (Bourrier et al., 2017, Bourrier et al., 2017). The authors caution these figures predate revised planetary masses and should be treated as order-of-magnitude estimates. JWST results indicating absent or thin atmospheres on the inner TRAPPIST-1 planets [(Bachman et al., 2023)-era observations cited as 2023Natur.618...39G and 2023Natur.620..746Z] are consistent with these escape predictions, though the outer planets remain candidates for retained atmospheres.

Biosignatures: definitions, candidates, and false positives

The chapter categorizes exoplanet biosignatures into gaseous (O$_2$0, O$_2$1, CH$_2$2, N$_2$3O, sulfur gases, methylated halogens), surface (the vegetation red edge near 700 nm), and temporal (seasonal variations). It is notably candid about false positives. More than 330 molecules have been detected in the interstellar medium, meaning nearly any candidate biomarker has a demonstrated abiotic production route. Two previously "clean" biosignatures—CH$_2$4Cl and DMS—have both been found abiotically: CH$_2$5Cl in the IRAS 16293-2422 star-forming region and comet 67P, and DMS in cometary ices and toward the Galactic center cloud G+0.693–0.027. The tentative DMS detection on K2-18b, claimed at 3.4$_2$6 significance, is reported alongside the independent re-analyses questioning its robustness, and the chapter concludes that DMS "may no longer be considered an irrefutable biosignature." Molecular oxygen, likewise, is produced abiotically in comets and can accumulate via atmospheric escape around M dwarfs. The implication drawn is that single-molecule detections will rarely suffice; biosignature pairs indicating disequilibrium, contextual information, and careful subtraction of the abiotic background are required.

For Solar System biosignature detection, two instrumentation efforts are described: ORIGIN, a laser desorption/ionization time-of-flight mass spectrometer capable of detecting lipids, amino acids, nucleobases, and PAHs, proposed for a Venus cloud-layer probe; and SenseLife/FlyPol, a full-Stokes spectropolarimeter exploiting circular polarization induced by homochiral macromolecules, which has distinguished vegetated from abiotic surfaces from airborne platforms.

Instrumentation and mission concepts

The chapter surveys the observing infrastructure underpinning future biosignature searches. RISTRETTO at the VLT aims to detect reflected light from Proxima Centauri b at a contrast of ~10$_2$7, requiring roughly 50 hours for a 5$_2$8 detection. For ANDES at the ELT, exposure-time estimates for atmospheric characterization of five M-dwarf targets range from 0.67 nights (Proxima b) to 13 nights (Ross 128 b), assuming an albedo of ~0.3. GCM-based albedo maps indicate Proxima b is a substantially better target than Ross 128 b, whose likely low dayside albedo works against both habitability and detectability. The LIFE mission concept—a mid-infrared nulling interferometer conceived and led within PlanetS—is presented as capable of directly characterizing 30–50 temperate terrestrial planets, with retrieval studies demonstrating robust detection of population-level CO$_2$9 trends indicative of the carbonate-silicate cycle using as few as 30 Exo-Earth Candidates at S/N = 10 and R = 50. The authors note explicitly that such population-level tracers cannot establish the habitability of individual planets.

Limitations and open questions

Several limitations are acknowledged in the text. The escape-rate estimates for TRAPPIST-1 rest on outdated mass determinations. The coupled effect of surface rock composition with subduction and outgassing regimes on exoplanet carbon cycling "remains to be evaluated," and the climatic consequences of a positive weathering feedback at the thermodynamic limit are not yet understood. The interpretation of Europa's putative plumes remains contested after extensive unsuccessful follow-up, and the radar signature interpreted as a subglacial lake at Mars's south pole is disputed. On the biosignature side, the chapter concedes that no universally definitive biosignature may exist and that seasonal variations in atmospheric composition on tidally locked planets can arise from viewing geometry alone, complicating temporal-biosignature interpretations. Whether LIFE proceeds depends on technology maturation still in the laboratory demonstration phase.

Conclusion

This chapter provides a consolidated account of how climate modeling, Solar System comparative studies, and instrument development jointly constrain the search for habitable and inhabited worlds. Its strongest contributions are quantitative: the demonstration that Venus-like day-night cloud asymmetry prevents early ocean formation, the identification of large errors in 1-D treatments of steam atmospheres, the dependence of HZ boundaries on background gas composition, and retrieval-based mission requirements for LIFE. Its central message is that habitability and biosignature assessments are irreducibly context-dependent—on stellar type, atmospheric evolution, interior composition, and ocean dynamics—and that credible life detection will require combining multiple, mutually reinforcing observables rather than reliance on any single molecular marker.