Exoplanet atmospheres: transmission spectroscopy, the habitable zone, and the search for biosignatures

Intuition Beginner

For more than twenty-five years, astronomers have known that planets orbit other stars. These exoplanets number in the thousands. But knowing they exist is only the start. We want to know what they are made of, whether they hold atmospheres, and whether any might support life. The key tool is to study starlight as a planet crosses in front of its host star.

When a planet transits, a sliver of the star's light skims through the planet's atmosphere on its way to us. Atoms and molecules in that atmosphere absorb specific wavelengths, like colored filters in front of a lamp. Plot the slight extra dimming at each wavelength and you get a transmission spectrum: a graph of which wavelengths got absorbed. Sodium (2002), water (2014), and carbon dioxide (2022) have all been detected this way.

The biggest goal is biosignatures: gases that life produces and maintains. Earth's air carries oxygen, ozone, methane, and nitrous oxide, all kept out of chemical balance by living things. If a similar cocktail showed up in the spectrum of an Earth-like exoplanet, that would be strong evidence for life beyond the Earth. JWST and the next generation of giant telescopes are built for exactly this search.

Visual Beginner

The diagram shows the four observational techniques used to characterize an exoplanet atmosphere. Transmission spectroscopy plots the transit depth against wavelength, peaking where molecules absorb. Emission spectroscopy measures the planet's dayside at secondary eclipse. Phase curves track brightness through the orbit, revealing day-night heat transport. Direct imaging spatially resolves the planet from its star. The habitable zone is marked with the Kasting moist-greenhouse inner edge and the maximum-greenhouse outer edge, scaled for a Sun-like star and for a cool M dwarf.

Each technique measures a different aspect of the same atmosphere and each requires different signal-to-noise thresholds to succeed.

Worked example Beginner

WASP-39b is a Saturn-mass planet about 700 light-years away. It orbits close to its host star with a year of just four days, so its atmosphere is heated to roughly 1100 Kelvin. In 2022 the JWST Early Release Science program observed WASP-39b transiting its star using the NIRSpec instrument.

Step 1. JWST measured the transit depth at hundreds of near-infrared wavelengths. A planet with no atmosphere produces a flat line: the depth is the same at every wavelength. An atmosphere produces a curved line with extra dimming at wavelengths where atmospheric molecules absorb.

Step 2. The WASP-39b spectrum showed a clear absorption feature at 4.3 microns, exactly where carbon dioxide absorbs. This was the first unambiguous CO2 detection on any exoplanet. Water, sulfur dioxide, sodium, and carbon monoxide also appeared. The sulfur dioxide detection is striking: making SO2 requires ultraviolet light from the star driving photochemistry.

Step 3. The CO2 feature amplitude is about 100 parts per million of extra dimming, indicating a carbon-rich atmosphere. The carbon-to-oxygen ratio inferred from the spectrum constrains how the planet formed: a value near solar suggests WASP-39b accreted substantial solid material beyond the water-ice line.

What this tells us: JWST can characterize exoplanet atmospheres at the precision needed to constrain atmospheric chemistry and formation history, the prerequisite for every future biosignature search on rocky planets.

Check your understanding Beginner

Formal definition Intermediate+

Three formal objects anchor exoplanet atmospheric science: the transit-depth spectrum, the atmospheric scale height, and the habitable-zone boundaries. The definitions below follow Seager ^{[Seager2010 Princeton]} and Wordsworth-Kreidberg ^{[Wordsworth-Kreidberg 2022]}.

Definition (transit depth and transmission spectrum). A transit occurs when a planet crosses the line of sight between observer and host star. The transit depth at wavelength $\lambda$ is

$$\delta (\lambda )=\frac{F_{\mathrm{out}}(\lambda )-F_{\mathrm{in}}(\lambda )}{F_{\mathrm{out}}(\lambda )},$$

where $F_{\mathrm{out}}$ is the out-of-transit stellar flux and $F_{\mathrm{in}}$ is the in-transit flux. For a planet with no atmosphere, $\delta =(R_p/R_{\ast }{)}^2$. With an atmosphere, $\delta (\lambda )$ increases at wavelengths where atmospheric molecules absorb. The function $\delta (\lambda )$ is the transmission spectrum.

Definition (atmospheric scale height). For an isothermal atmosphere in hydrostatic equilibrium with temperature $T$, mean molecular mass $\mu m_H$, and surface gravity $g$, the scale height is

$$H=\frac{k_{B}T}{\mu m_{H}g},$$

the vertical distance over which the atmospheric density falls by a factor of $e$. The scale height sets the characteristic thickness of the absorbing annulus probed in transmission.

Definition (habitable zone, Kasting-Whitmire-Reynolds 1993). The habitable zone around a star of luminosity $L_{\ast }$ is the range of orbital distances over which a rocky planet with a CO2-H2O atmosphere can maintain liquid water on its surface. The inner edge — the moist greenhouse limit — occurs where water vapour saturates the upper atmosphere, photodissociates, and hydrogen escapes. The outer edge — the maximum greenhouse limit — occurs where CO2 condenses and its clouds cease to provide greenhouse warming. For the present Sun, $d_{\mathrm{inner}}\approx 0.95$ AU and $d_{\mathrm{outer}}\approx 1.67$ AU ^{[Kasting1993 Icarus 101:108]}. Both edges scale as $d\propto \sqrt{L_{\ast }}$, so an M dwarf with $L_{\ast }\sim 10^{-2}{L}_{\odot }$ has a habitable zone near $0.1$ AU.

Definition (biosignature gas). A biosignature gas is an atmospheric species whose observed abundance requires a continuous biological source to maintain against photochemical or geochemical loss. Following Seager ^[Seager2013], the principal biosignature gases for an Earth-like metabolism are O2, O3 (its photochemical derivative), CH4, N2O, and H2O. The simultaneous detection of O2 (or O3) together with CH4 or N2O indicates thermodynamic disequilibrium: the species react with each other on geological timescales, so their coexistence requires active production.

Counterexamples to common slips Intermediate+

• "A planet in the habitable zone is habitable." No. The habitable zone marks where surface liquid water is physically possible given a suitable atmosphere, not where it actually exists. Venus and Mars both lie within the Sun's habitable zone today; neither currently supports surface life. Atmospheric mass, magnetic shielding, volatile inventory, stellar ultraviolet flux, and impact history all enter.

• "Oxygen implies life." No. Abiotic O2 forms readily via CO2 photolysis followed by H escape (Mars, Venus) and via H2O photolysis during the prolonged pre-main-sequence phase of M dwarfs. Distinguishing biotic from abiotic O2 requires measuring co-absorbers (CH4, N2O, CO), the stellar ultraviolet flux, and the atmospheric redox state.

• "Transmission spectroscopy works for every exoplanet." No. Only transiting planets — those whose orbital geometry carries them across the stellar disc — are accessible. For an Earth analogue orbiting a Sun-like star the transit probability is roughly $R_{\ast }/d\approx 0.5\%$. Direct imaging and high-resolution Doppler spectroscopy cover the non-transiting population.

• "A single biosignature gas is sufficient evidence of life." No. Every biosignature gas has abiotic sources. The inferential strategy is to combine multiple co-absorbers in thermodynamic disequilibrium, rule out abiotic pathways using the stellar and planetary context, and require consistency across multiple epochs and instruments.

• "Exoplanet atmospheres are uniform." No. Phase curves of WASP-12b show diurnal CO and CH4 variations of hundreds of parts per million; clouds and hazes suppress molecular features on GJ 1214b and HAT-P-11b; alkali metals and titanium oxide shape the spectra of the hottest planets. The 1-D isothermal model used in retrieval is a parameterisation, not a description.

Key result: JWST WASP-39b CO2 and the era of exoplanet atmospheric characterization Intermediate+

The detection of carbon dioxide in the atmosphere of the hot Jupiter WASP-39b by JWST's NIRSpec instrument in 2022 established transmission spectroscopy as a quantitative tool for atmospheric chemistry on exoplanets.

Result (WASP-39b CO2 detection, JWST ERS 2022). The JWST transmission spectrum of WASP-39b shows an unambiguous CO2 absorption feature at $4.3\mu \mathrm{m}$ with amplitude approximately $1.4\times 10^{-3}$ above the surrounding continuum, plus a simultaneous SO2 feature at $4.0\mu \mathrm{m}$ produced by UV-driven photochemistry ^{[JWSTERS2022 Nature 614:649]}. The CO2 detection constrains the atmospheric metallicity to within $\sim 0.2$ dex of the host-star value and the carbon-to-oxygen ratio to near-solar, indicating accretion of solid material beyond the water-ice line during the planet's formation.

Defense (signal-strength argument). A Saturn-mass hot Jupiter at $T\approx 1100$ K with a hydrogen-helium envelope has scale height $H\approx 550$ km. The transit-depth signal from an atmospheric absorption feature with slant optical depth ${\tau }_{\mathrm{slant}}$ is

$$\Delta \delta \approx \frac{2R_{p}H}{R_{\ast }^2}{\tau }_{\mathrm{slant}}\approx \frac{2(9.0\times 10^7\mathrm{m})(5.5\times 10^5\mathrm{m})}{(7.0\times 10^8\mathrm{m}{)}^2}{\tau }_{\mathrm{slant}}\approx 2\times 10^{-4}{\tau }_{\mathrm{slant}}.$$

A slant optical depth of order unity through the CO2 band yields $\Delta \delta \sim 10^{-4}$ to $10^{-3}$, comfortably above JWST's photon-noise floor for a single transit. By contrast, the same feature on a TRAPPIST-1 rocky planet — ten times smaller in radius, with $H\sim 7$ km — produces a signal of order $10^{-6}$, two orders of magnitude weaker and at the edge of JWST sensitivity even with many transits stacked. $\square$

Bridge. The WASP-39b detection builds toward the TRAPPIST-1 system, where JWST is currently characterizing the atmospheres of three Earth-sized planets in the habitable zone, and appears again in 28.10.02 astrobiology, where the same transmission-spectroscopy framework extends to disequilibrium biosignature detection on rocky exoplanets. The foundational reason the technique works is that an isothermal atmosphere produces a scale-height-thick absorbing annulus whose spectral signature scales linearly with composition, and this is exactly the bridge between the bulk-transit detection of 28.05.01 exoplanet survey and the chemical-characterization regime that defines the field. Putting these together identifies the transmission spectrum with a calibrated chemical probe of the planet's atmosphere, and the pattern generalises from the Hubble-era hot Jupiters to the rocky-planet targets that will dominate the JWST and ELT-era observing programs.

Exercises Intermediate+

Advanced results Master

Theorem 1 (Charbonneau 2002: first exoplanet atmospheric detection). Charbonneau, Brown, Noyes, and Gilliland ^{[Charbonneau2002 ApJ 568:377]} detected atomic sodium in the atmosphere of the hot Jupiter HD 209458b using the Hubble Space Telescope Imaging Spectrograph. The extra transit depth at the Na D doublet (589.0 and 589.6 nm) was approximately $2.3\times 10^{-4}$ above the surrounding continuum, the first detection of any atomic or molecular species in an exoplanet atmosphere.

Theorem 2 (Deming-Seager 2005: first secondary eclipse). Deming, Seager, Richardson, and Harrington ^{[Deming2005 Nature 434:740]} detected the secondary eclipse of HD 209458b at $24\mu \mathrm{m}$ with the Spitzer MIPS instrument, measuring the planet's dayside thermal emission directly. The eclipse depth of $\sim 0.3\%$ constrained the dayside brightness temperature to $\sim 1100$ K and opened emission spectroscopy as a technique complementary to transmission.

Theorem 3 (Kasting-Whitmire-Reynolds 1993: habitable-zone limits). Kasting, Whitmire, and Reynolds ^{[Kasting1993 Icarus 101:108]} computed the radiative-convective equilibria of rocky-planet CO2-H2O-N2 atmospheres around main-sequence stars of spectral types F, G, K, and M. The moist-greenhouse inner limit at solar flux $1.1$ times solar and the maximum-greenhouse outer limit at $0.36$ times solar define the canonical habitable-zone boundaries; both scale as $d\propto \sqrt{L_{\ast }}$ with stellar luminosity.

Theorem 4 (Madhusudhan-Seager 2009: phase-curve retrieval). Madhusudhan and Seager ^{[Madhusudhan2009 ApJ 707:24]} formulated the Bayesian retrieval formalism for exoplanet atmospheres: given observed spectra, the posterior distribution over temperature-pressure profiles and mixing ratios is sampled via Markov-chain Monte Carlo. Applied to phase curves, the formalism recovers the longitudinal brightness map and heat-redistribution efficiency.

Theorem 5 (Kreidberg 2014: water on a non-Jupiter exoplanet). Kreidberg et al. ^{[Kreidberg2014 Nature 505:69]} detected robust water-vapour absorption in the Hubble transmission spectrum of the Neptune-mass exoplanet HAT-P-11b. The result established that hydrogen-rich envelopes extend below Jupiter mass and validated retrieval pipelines in the presence of clouds and hazes.

Theorem 6 (Seager 2013: biosignature gas framework). Seager, Bains, and Hu ^[Seager2013] systematised the search for biosignature gases along three axes: Earth-like biosignatures (O2, O3, CH4, N2O, H2O), thermodynamic disequilibrium indicators, and non-Earth biosignature gases from hypothetical metabolisms. The framework is the operational basis for JWST and ELT target selection.

Theorem 7 (JWST ERS 2022-2024: WASP-39b, WASP-12b, K2-18b). The JWST Early Release Science program delivered the first CO2 and SO2 detections on WASP-39b ^{[JWSTERS2022 Nature 614:649]}, the diurnal CO and CH4 phase-curve variations on WASP-12b, and a contested dimethyl-sulfide feature on the Hycean-planet candidate K2-18b ^{[Madhusudhan2024]}. The detections validate JWST's characterisation capability and demonstrate simultaneously the power of transmission spectroscopy and the systematic-error pitfalls at the threshold of rocky-planet characterization.

Theorem 8 (Wordsworth-Kreidberg 2022: review). Wordsworth and Kreidberg ^{[Wordsworth-Kreidberg 2022]} synthesise the state of exoplanet atmospheres as of 2022: the Hubble era of hot-Jupiter characterisation, the JWST era of chemical-precision spectroscopy, and the ELT-era path to direct imaging of Earth analogs. The review formalises the signal-to-noise scalings that connect transit, emission, phase-curve, and direct-imaging methods.

Synthesis. The WASP-39b detection is the foundational reason that exoplanet atmospheric characterisation has shifted from speculative inference to quantitative measurement, and the central insight is that the transit-depth signal in a molecular band scales linearly with the atmospheric scale height and the slant optical depth, making the technique predictively quantitative rather than discovery-limited. This is exactly the bridge between the bulk-transit detection of 28.05.01 exoplanet survey, where only the planet's radius is measured, and the chemical-characterisation regime of 28.10.02 astrobiology, where individual molecular abundances are retrieved and disequilibrium combinations tested for biogenicity.

The pattern generalises from the Hubble-era hot-Jupiter detections of sodium (Charbonneau 2002) and water (Kreidberg 2014) through the JWST-era CO2 on WASP-39b, the diurnal variations on WASP-12b, and the contested DMS on K2-18b, and appears again in 28.09.02 pending adaptive optics, where extreme high-contrast instruments on the ELT, GMT, and TMT of the 2030s will spatially resolve Earth analogs around nearby Sun-like stars in a regime wholly inaccessible to transit methods. Putting these together identifies the transmission spectrum as a calibrated chemical probe whose sensitivity is bounded by stellar photon noise and by atmospheric cloud and haze opacity, and the bridge is from the present era of biosignature-detection attempts on M-dwarf rocky planets to the future ELT-era direct spectroscopic characterisation of true Earth twins.

Full proof set Master

Proposition (transit-depth signal in the Lambert approximation). For a planet of radius $R_p$ with an isothermal atmosphere of scale height $H\ll R_p$, orbiting a star of radius $R_{\ast }$, the additional transit depth at a wavelength where the slant optical depth at the reference radius is ${\tau }_{\mathrm{slant}}(\lambda )$ satisfies, in the optically thin regime,

$$\Delta \delta (\lambda )\approx \frac{2R_{p}H}{R_{\ast }^2}{\tau }_{\mathrm{slant}}(\lambda ).$$

Proof. Adopt cylindrical coordinates centred on the planet, with impact parameter $b$ measured from the planet's centre and chord coordinate $s$ along the line of sight. For $b$ just larger than $R_p$, the height of the chord above the reference radius at position $s$ along the chord is

$$h(b,s)=b\sqrt{1-s^2/b^2}-R_p\cdot \sqrt{1-s^2/b^2}{|}_{\text{expanded for }b=R_p+x}\approx x+\frac{s^2}{2R_p},$$

with $x:=b-R_p\ll R_p$. The atmospheric number density along the chord, by hydrostatic equilibrium in plane-parallel geometry, is

$$n(b,s)=n_0\exp \left(-\frac{h(b,s)}{H}\right)=n_0\exp \left(-\frac{x}{H}\right)\exp \left(-\frac{s^2}{2R_{p}H}\right),$$

where $n_0$ is the density at the reference radius $R_p$. The slant optical depth at impact parameter $b=R_p+x$ is

$$\tau (b,\lambda )={\int }_{-\infty }^{\infty }\sigma (\lambda )n(b,s)ds=\sigma (\lambda )n_0\exp \left(-\frac{x}{H}\right)\sqrt{2\pi R_{p}H}.$$

Defining ${\tau }_{\mathrm{slant}}(\lambda ):=\tau (R_p,\lambda )=\sigma (\lambda )n_0\sqrt{2\pi R_{p}H}$, the additional blocked area is the annulus integral

$$A_{\mathrm{atm}}(\lambda )={\int }_{R_p}^{\infty }2\pi b(1-e^{-\tau (b,\lambda )})db.$$

For optically thin absorption, $1-e^{-\tau }\approx \tau$. Substituting $b=R_p+x$ with $2\pi bdb\approx 2\pi R_{p}dx$ (valid since $x\ll R_p$),

$$A_{\mathrm{atm}}(\lambda )\approx 2\pi R_p{\tau }_{\mathrm{slant}}(\lambda ){\int }_0^{\infty }{e}^{-x/H}dx=2\pi R_{p}H{\tau }_{\mathrm{slant}}(\lambda ).$$

The additional transit depth is $A_{\mathrm{atm}}(\lambda )/(\pi R_{\ast }^2)$, giving

$$\Delta \delta (\lambda )=\frac{2R_{p}H}{R_{\ast }^2}{\tau }_{\mathrm{slant}}(\lambda ).\square$$

Proposition (habitable-zone scaling with stellar luminosity). The inner and outer edges of the Kasting habitable zone scale with stellar luminosity as $d_{\mathrm{HZ}}=d_{\mathrm{HZ},\odot }\sqrt{L_{\ast }/L_{\odot }}$, with fixed flux thresholds.

Proof. Surface liquid water requires the equilibrium temperature $T_{\mathrm{eq}}$ to lie between the freezing point and the moist-greenhouse threshold. The equilibrium temperature satisfies the absorbed-flux equals emitted-flux balance

$$\frac{(1-A)L_{\ast }}{16\pi d^2}=\sigma T_{\mathrm{eq}}^4,$$

where $A$ is the Bond albedo. Solving for the orbital distance at fixed $T_{\mathrm{eq}}$ and fixed $A$,

$$d=\sqrt{\frac{(1-A)L_{\ast }}{16\pi \sigma T_{\mathrm{eq}}^4}}\propto \sqrt{L_{\ast }}.$$

The inner and outer habitable-zone edges correspond to fixed values of $T_{\mathrm{eq}}$ (the moist-greenhouse and maximum-greenhouse thresholds from Kasting, Whitmire, and Reynolds ^{[Kasting1993 Icarus 101:108]}), so both edges scale as $\sqrt{L_{\ast }}$. For an M dwarf with $L_{\ast }\approx 10^{-2}{L}_{\odot }$, the habitable zone sits at $0.1$ times the solar value, near $0.1$ AU. $\square$

Connections Master

• Exoplanet survey 28.05.01. This unit deepens the bulk-detection framework of 28.05.01 by deriving the transmission-spectroscopy signal and the Kasting habitable-zone limits in physical detail, and by extending the chemical scope into biosignature gases. The survey-unit discussion of transit photometry, radial velocity, and the habitable-zone concept provides the observational foundation on which the present atmospheric-characterization results are built.

• Astrobiology and biosignatures 28.10.02. The biosignature-gas framework developed here — O2/CH4 disequilibrium, abiotic O2 false positives, the multi-gas inferential strategy — extends in 28.10.02 to the broader question of life detection across solar-system and extrasolar environments. The astrobiology unit supplies the biological and chemical context that the spectroscopic technique here is designed to probe.

• Adaptive optics and high-resolution astronomy 28.09.02 pending. Transit spectroscopy is one of two complementary paths to exoplanet atmospheric characterisation; the other is direct imaging with extreme-adaptive-optics coronagraphs on the ELT, GMT, and TMT in the 2030s. 28.09.02 pending develops the high-contrast imaging technology that will enable direct spectroscopy of Earth analogs around Sun-like stars — a regime wholly inaccessible to transit methods.

• Climate change 27.07.01. The radiative-convective equilibrium models that define the moist-greenhouse and maximum-greenhouse habitable-zone boundaries are direct descendants of Earth climate models. 27.07.01 develops the greenhouse-effect physics, CO2-H2O feedbacks, and runaway processes that reappear here in the exoplanetary context: the same equation of radiative transfer governs both, and the runaway greenhouse of 27.07.01 is the inner-edge boundary condition of the habitable zone treated here.

Historical & philosophical context Master

David Charbonneau, Timothy Brown, Ronald Noyes, and Ron Gilliland ^{[Charbonneau2002 ApJ 568:377]} in 2002 made the first detection of an atmospheric species on an extrasolar planet, observing sodium in transmission through the hot Jupiter HD 209458b with the Hubble Space Telescope Imaging Spectrograph. The modern theoretical framework for the habitable zone was set by James Kasting, Daniel Whitmire, and Ray Reynolds ^{[Kasting1993 Icarus 101:108]} in their 1993 radiative-convective calculation of the moist-greenhouse and maximum-greenhouse boundaries, which fixed the canonical solar habitable-zone edges at $0.95$ and $1.67$ AU. Drake Deming, Sara Seager, and colleagues ^{[Deming2005 Nature 434:740]} in 2005 made the first secondary-eclipse detection of an exoplanet's dayside thermal emission with the Spitzer MIPS instrument, opening emission spectroscopy as a technique complementary to transmission.

Sara Seager's monograph Exoplanet Atmospheres ^{[Seager2010 Princeton]} in 2010 crystallised the field's methodology, and her 2013 framework paper ^[Seager2013] systematised the search for biosignature gases along the disequilibrium axis that now organises JWST target selection. Nikku Madhusudhan and Seager ^{[Madhusudhan2009 ApJ 707:24]} in 2009 developed the Bayesian retrieval formalism that converts observed spectra into posterior distributions over atmospheric parameters. Laura Kreidberg and colleagues ^{[Kreidberg2014 Nature 505:69]} in 2014 detected water vapour in the atmosphere of the Neptune-mass HAT-P-11b, the first robust atmospheric characterisation below Jupiter mass. The JWST Early Release Science program ^{[JWSTERS2022 Nature 614:649]} in 2022 detected CO2 and SO2 on WASP-39b — the first unambiguous CO2 on any exoplanet and the operational validation of JWST atmospheric characterisation — and the K2-18b Hycean-atmosphere candidate ^{[Madhusudhan2024]} inaugurated the contested-biosignature era. Wordsworth and Kreidberg ^{[Wordsworth-Kreidberg 2022]} synthesise the field as of 2022.

Bibliography Master

@article{Charbonneau2002,
  author  = {Charbonneau, David and Brown, Timothy M. and Noyes, Ronald W. and Gilliland, Ronald L.},
  title   = {Detection of an Extrasolar Planet Atmosphere},
  journal = {Astrophysical Journal},
  volume  = {568},
  pages   = {377--384},
  year    = {2002},
  doi     = {10.1086/338770},
}

@article{Kasting1993,
  author  = {Kasting, James F. and Whitmire, Daniel P. and Reynolds, Ray T.},
  title   = {Habitable Zones around Main Sequence Stars},
  journal = {Icarus},
  volume  = {101},
  pages   = {108--128},
  year    = {1993},
  doi     = {10.1006/icar.1993.1010},
}

@article{Deming2005,
  author  = {Deming, Drake and Seager, Sara and Richardson, L. Jeremy and Harrington, Joseph},
  title   = {Infrared thermal emission from an extrasolar planet},
  journal = {Nature},
  volume  = {434},
  pages   = {740--743},
  year    = {2005},
  doi     = {10.1038/nature03507},
}

@article{Madhusudhan2009,
  author  = {Madhusudhan, Nikku and Seager, Sara},
  title   = {A Temperature and Abundance Retrieval Method for Exoplanet Atmospheres},
  journal = {Astrophysical Journal},
  volume  = {707},
  pages   = {24--39},
  year    = {2009},
  doi     = {10.1088/0004-637X/707/1/24},
}

@article{Kreidberg2014,
  author  = {Kreidberg, Laura and Bean, Jacob L. and D{\'e}sert, Jean-Michel and others},
  title   = {Clouds in the atmosphere of the super-Earth exoplanet {GJ 1214b}},
  journal = {Nature},
  volume  = {505},
  pages   = {69--72},
  year    = {2014},
  doi     = {10.1038/nature12888},
}

@book{Seager2010,
  author    = {Seager, Sara},
  title     = {Exoplanet Atmospheres: Physical Processes},
  publisher = {Princeton University Press},
  year      = {2010},
}

@article{Seager2013,
  author  = {Seager, Sara and Bains, William and Hu, Renyu},
  title   = {A Biomass-Based Model to Estimate the Plausibility of Exoplanet Biosignature Gases},
  journal = {Astrophysical Journal},
  volume  = {775},
  pages   = {104--116},
  year    = {2013},
  doi     = {10.1088/0004-637X/775/2/104},
}

@article{JWSTERS2022,
  author  = {{JWST Transiting Exoplanet Community Early Release Science Team}},
  title   = {Identification of carbon dioxide in an exoplanet atmosphere},
  journal = {Nature},
  volume  = {614},
  pages   = {649--652},
  year    = {2023},
  doi     = {10.1038/s41586-022-05269-w},
}

@article{Madhusudhan2024,
  author  = {Madhusudhan, Nikku and Constantinou, Savvas and Holmberg, Mads and others},
  title   = {Carbon-bearing molecules in a possible Hycean atmosphere},
  journal = {Astrophysical Journal Letters},
  volume  = {956},
  pages   = {L13},
  year    = {2023},
  doi     = {10.3847/2041-8213/acf140},
}

@article{Lecavelier2008,
  author  = {Lecavelier des Etangs, Alain},
  title   = {A diagram to determine the evaporation status of extrasolar planets},
  journal = {Astronomy \& Astrophysics},
  volume  = {481},
  pages   = {L53--L56},
  year    = {2008},
  doi     = {10.1051/0004-6361:20078949},
}

@article{WordsworthKreidberg2022,
  author  = {Wordsworth, Robin and Kreidberg, Laura},
  title   = {Atmospheres of Exoplanets},
  journal = {Annual Review of Astronomy and Astrophysics},
  volume  = {60},
  pages   = {111--158},
  year    = {2022},
  doi     = {10.1146/annurev-astro-052920-125333},
}