--- title: Atmospheric escape canonical_url: https://paxabyssi.com/wiki/Atmospheric_escape markdown_url: https://paxabyssi.com/wiki/Atmospheric_escape.md type: wiki-page revision_id: 650 revision_view: stable last_updated: 2026-10-06 license: CC BY-SA 4.0 license_url: https://creativecommons.org/licenses/by-sa/4.0/ science_status: - observed - model - sim summary: The loss of a planet's or moon's air to space, gas molecule by gas molecule or in a bulk wind driven by starlight. It decides which worlds keep an atmosphere, and it carves the radius valley and the hot Neptune desert. categories: - Planets - Atmospheric science - Planet formation - Exoplanets aliases: - Atmospheric loss - Atmospheric mass loss - Atmospheric stripping - Jeans escape - Jeans parameter - Hydrodynamic escape - Energy-limited escape - Photoevaporation - Core-powered mass loss - Boil-off - Exobase - Non-thermal escape infobox: type: physics_concept name: Atmospheric escape image: File:Gliese_436b_hydrogen_tail_ESAHubble.jpg caption: "Artist's concept: the cloud of hydrogen escaping the warm Neptune GJ 436 b, measured by Hubble" sim_use: "Every mini-Neptune the generator makes has its escape worked out: XUV flux from its star's luminosity, type and age, energy-limited and core-powered rates, a Roche-lobe check and an envelope lifetime. The fastest losers become evaporating mini-Neptunes (MNH-EV); stripped giant cores are chthonian planets." symbols: M, M_p planet mass; m molecular mass; T exobase temperature; r exobase radius; v_0 = sqrt(2kT/m); F_XUV X-ray and extreme-ultraviolet flux; eta heating efficiency (of order 0.1); K tidal correction for the star's pull (1 or less) formulae: - "Jeans parameter at the exobase: lambda = G M m / (k T r) = (v_esc / v_0)^2" - "Energy-limited escape: dM/dt = eta pi F_XUV R_p^3 / (G M_p K)" validity: Jeans escape applies when lambda is large and the gas leaks slowly; the energy-limited formula is an upper estimate for strongly irradiated hydrogen atmospheres, and becomes recombination-limited at the highest fluxes definition: The loss of gas from the top of an atmosphere to space, either thermally (Jeans escape, molecule by molecule; hydrodynamic escape, as a bulk wind) or by non-thermal processes driven by charged particles, photochemistry and impacts. last_verified: 2026-10-07 key_quantities: Escape velocity (Earth 11.2 km/s, Mars 5.0, Titan 2.6); exobase temperature; X-ray output, about a thousandth of a young Sun-like star's total light during its first 100 million years worked_example: "Earth's exobase, about 500 km up at about 1,000 K: lambda is about 7 for hydrogen atoms and about 110 for oxygen atoms, so hydrogen leaks away slowly and oxygen stays." observed_examples: - Mars (MAVEN) - Venus (deuterium-to-hydrogen ratio) - HD 209458 b - GJ 436 b - WASP-107 b (helium) literature_schemes: - Jeans (thermal) - hydrodynamic and energy-limited - recombination-limited - core-powered mass loss - boil-off - Roche-lobe overflow - non-thermal (sputtering - ion pickup - charge exchange - photochemical escape) - impact erosion related: - https://paxabyssi.com/wiki/Hot_Neptune_desert.md - https://paxabyssi.com/wiki/Radius_valley.md - https://paxabyssi.com/wiki/Greenhouse_world.md - https://paxabyssi.com/wiki/Mini-Neptune.md - https://paxabyssi.com/wiki/Super-Earth.md - https://paxabyssi.com/wiki/Lava_world.md --- # Atmospheric escape > Source: https://paxabyssi.com/wiki/Atmospheric_escape > > Licence: [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/). Text by Pax Abyssi Wiki contributors; history at https://paxabyssi.com/wiki/Atmospheric_escape/history > > Revision 650, 6 October 2026 **Atmospheric escape** is the loss of a planet's or moon's air to space. Every atmosphere leaks from the top, where the gas is so thin that a fast molecule heading upward can leave without hitting anything on the way. For Earth the leak is slow and mostly hydrogen. For a planet close to a young, active star it can be a wind that carries away a whole envelope of gas in a few hundred million years. Escape explains why Mars has thin air and Venus has almost no water, why some rocky exoplanets are bare, and why two of the clearest patterns in the planet census exist: the [radius valley](https://paxabyssi.com/wiki/Radius_valley.md) among small planets and the [hot Neptune desert](https://paxabyssi.com/wiki/Hot_Neptune_desert.md) close to their stars. In Pax Abyssi the generator works out escape for every mini-Neptune it makes, and the planets losing their air fastest form a type of their own. ![An illustration of a planet in front of a red star, trailing a vast pale cloud of escaping hydrogen.](https://media.paxabyssi.com/public/d4787069cda7e39ef7cfaebb72bf497f25663650c588793e067706a720da39f1/2560.webp "Artist's concept: the cloud of hydrogen escaping the warm Neptune GJ 436 b, measured by Hubble. The cloud is far larger than the planet and trails it like a comet's tail.") *Figure 1.* Artist's concept: the cloud of hydrogen escaping the warm Neptune GJ 436 b, measured by Hubble. The cloud is far larger than the planet and trails it like a comet's tail. Credit: NASA, ESA, STScI, and G. Bacon. Licence: CC BY 4.0. ## Where gas leaves Low in an atmosphere, molecules collide constantly and none gets far. Higher up the air thins until, at a level called the **exobase**, a molecule is more likely to fly off on a ballistic path than to hit another one. Earth's exobase is about 500 km up, in the thermosphere, where sunlight in the extreme ultraviolet heats the thin gas to around 1,000 K, more in years of high solar activity [1]. Whether a molecule at the exobase escapes depends on how its speed compares with the planet's escape velocity there. ## Thermal escape ### Jeans escape At any temperature the molecules of a gas have a spread of speeds, and a few in the fast tail of the distribution always exceed escape velocity. Their steady trickle away is **Jeans escape**, after James Jeans, who worked it out early in the twentieth century. Its strength is set by one number, the **Jeans parameter**, the ratio of a molecule's gravitational binding energy at the exobase to its thermal energy: $$ \lambda = \frac{G M m}{k T r} = \left(\frac{v_\mathrm{esc}}{v_0}\right)^2 , $$ where $M$ is the planet's mass, $m$ the mass of the molecule, $T$ the exobase temperature, $r$ the exobase radius and $v_0 = \sqrt{2kT/m}$ the most probable thermal speed. The escaping flux falls roughly as $(1+\lambda)\,e^{-\lambda}$, so a modest change in $\lambda$ makes an enormous difference [1]. At Earth's exobase, $\lambda$ is about 7 for hydrogen atoms and about 110 for oxygen atoms. Hydrogen therefore leaks away, while oxygen, sixteen times heavier, is held for the life of the planet. Earth loses about 3 kg of hydrogen and some 50 g of helium every second, through Jeans escape and the non-thermal routes described below [2]. The hydrogen comes mostly from water vapour and methane broken apart high in the atmosphere, and its loss is limited less by the exobase than by how fast hydrogen-bearing molecules can diffuse up to it, the **diffusion limit** [3]. ### Hydrodynamic escape If the upper atmosphere is heated hard enough, $\lambda$ falls toward a few and escape changes character. The gas no longer leaks molecule by molecule but expands outward as a bulk flow, accelerating through the speed of sound like the solar wind. This is **hydrodynamic escape** [4]. A strong enough outflow of hydrogen drags heavier atoms with it, and because lighter species are carried more easily, it leaves the remaining gas enriched in heavy isotopes, a fingerprint that survives billions of years [5]. The energy for the flow comes from the star's X-ray and extreme-ultraviolet light (together, XUV), absorbed high in the atmosphere. If a fixed fraction of that energy goes into lifting gas out of the planet's gravity well, the mass-loss rate is $$ \dot{M} \approx \frac{\eta\,\pi F_\mathrm{XUV} R_p^3}{G M_p K}, $$ the **energy-limited** rate, where $F_\mathrm{XUV}$ is the XUV flux at the planet, $R_p$ and $M_p$ its radius and mass, $\eta$ a heating efficiency of order 0.1 and $K$, at most 1, a correction for the star's tidal pull, which helps gas over the top [4] [6]. At the highest fluxes the formula overestimates: much of the absorbed energy is radiated away by hydrogen recombining with electrons, and the rate grows only as the square root of the flux, the **recombination-limited** regime [7] [8]. ### Starlight that fades The XUV output of a star changes with age. A young Sun-like star spins fast, and its magnetic activity holds its X-ray output near a ceiling of about a thousandth of its total light [9]. It stays there for about its first 100 million years, then fades as the star's spin slows [10]. Low-mass red dwarfs spin down more slowly and stay at that ceiling for longer, which makes their planets the hardest places to keep an atmosphere [9]. Most escape driven by starlight therefore happens early, and a planet's atmosphere today is the record of what survived its star's youth. ## Stripping whole envelopes Close-in planets with hydrogen-helium envelopes are the most vulnerable, because hydrogen is light and a puffy planet presents a large target with weak gravity at its top. Several processes act on them, at different times. - **Boil-off.** A planet that forms inside its star's gas disc is hot and swollen. When the disc disperses, the pressure holding its outer envelope in place disappears, and a young planet can lose a large share of its envelope in a rapid episode as it contracts [11]. - **Photoevaporation.** Over the next few hundred million years, XUV heating drives an energy-limited wind. Owen and Wu (2013) and Lopez and Fortney (2013) showed that it should strip the envelopes from small cores on close orbits and spare heavier cores or more distant planets, predicting a gap in planet sizes before it was found [12] [13] [14]. - **Core-powered mass loss.** The heat left from a planet's formation also escapes through its envelope, and the star's ordinary light keeps the outer layers warm. Together they can drive the envelope away over about a billion years with no help from XUV at all [15] [16]. - **Roche-lobe overflow.** A planet so close to its star that its atmosphere reaches the point where the star's gravity takes over loses gas directly across that boundary. The hot Jupiter WASP-12 b is distorted by tides and probably overflowing in this way [17]. The first two predict almost the same split of small planets into stripped rocky cores and survivors with envelopes, the [radius valley](https://paxabyssi.com/wiki/Radius_valley.md). For larger planets, the same erosion helps empty the [hot Neptune desert](https://paxabyssi.com/wiki/Hot_Neptune_desert.md). ## Seeing it happen Escaping gas is too thin to image, but it can be seen in silhouette. During a transit, a planet with an extended cloud of escaping hydrogen absorbs the star's ultraviolet light in the Lyman-alpha line over an area far larger than the planet itself. The first detection, around the hot Jupiter HD 209458 b in 2003, showed hydrogen extending beyond the planet's Roche lobe [18]. The warm Neptune GJ 436 b trails a hydrogen cloud large enough to block about half the star's Lyman-alpha light, in a tail that follows the planet like a comet's [19]. Lyman-alpha light is largely absorbed by interstellar hydrogen, so only nearby stars can be studied this way. In 2018 Oklopčić and Hirata proposed a line of helium at 1,083 nanometres, in the near-infrared, which passes through interstellar space and can be observed from the ground [20]. Helium escaping from the super-Neptune WASP-107 b was detected with it the same year [21]. Because hydrogen escapes more readily than helium, slow escape could leave some warm Neptunes with atmospheres rich in helium [22]; see [Exotic worlds](https://paxabyssi.com/wiki/Exotic_worlds.md). ## Non-thermal escape Atoms and molecules can also be given escape speed without being heated as a gas. Sunlight and particles from the solar wind ionise the upper atmosphere, and ions can be picked up by the passing wind's magnetic field and swept away (**ion pickup**), or hit neutral atoms hard enough to knock them out (**sputtering**). A fast ion can steal an electron from a slow atom and leave as a fast neutral (**charge exchange**). Molecules broken apart by sunlight can leave fragments with enough energy to escape (**photochemical escape**) [1]. These processes dominate on small, cool worlds. NASA's MAVEN orbiter has measured these processes at Mars, which loses about 2 to 3 kg of gas a second today; extrapolated over the Sun's history, escape removed at least 0.8 bar of carbon dioxide and enough water to cover the planet 23 metres deep [23]. See [Arid world](https://paxabyssi.com/wiki/Arid_world.md). Whether a global magnetic field protects an atmosphere is less clear than often assumed. A field deflects the solar wind, but it also channels energy into the polar regions and opens other escape routes, and Earth, Mars and Venus lose ions at comparable rates today [24]. ## Who keeps an atmosphere Across the Solar System, the bodies with substantial atmospheres are those whose escape velocity is high for the light they receive. Zahnle and Catling traced a **cosmic shoreline** separating them from airless bodies, in which the insolation needed to strip a body rises roughly as the fourth power of its escape velocity [25]. Mercury and the Moon fall on the airless side; [Titan](https://paxabyssi.com/wiki/Titan.md), with a lower escape velocity than Mercury's but far from the Sun, keeps a thick one. Venus kept its carbon dioxide but lost its water: the ratio of deuterium to ordinary hydrogen in its atmosphere is about 100 times Earth's, the heavy isotope left behind as hydrogen escaped [26]. See [Greenhouse world](https://paxabyssi.com/wiki/Greenhouse_world.md). The shoreline is now being tested on rocky exoplanets, most of them around red dwarfs. The hot rocky planet LHS 3844 b has no thick atmosphere [27], and the James Webb Space Telescope found that TRAPPIST-1 b's dayside is as hot as a bare rock's [28]. See [Barren rock world](https://paxabyssi.com/wiki/Barren_rock_world.md). > **In Pax Abyssi** > > **IN THE GAME.** Every mini-Neptune in the game's generated star systems was made with its escape worked out. The generator estimates the XUV flux at the planet from its star's luminosity, type and age; computes the energy-limited photoevaporation rate, with the tidal correction for the star's pull, and the core-powered rate from the planet's own cooling; keeps the stronger of the two; and checks whether the planet overfills its Roche lobe. From the rate it derives how long the envelope will last and how the loss would look from afar: how far the hydrogen cloud reaches, how long a helium tail trails behind, and whether a comet-like tail and a bow shock form. > > A hot mini-Neptune that is losing more than a tenth of its mass every billion years, or overflowing its Roche lobe, or hotter than 1,200 K, becomes an **evaporating mini-Neptune** (MNH-EV). A census of the game's generated systems on 7 October 2026 counts 87 of them. Giants stripped down to their cores are **chthonian planets**, and each carries the process that stripped it, chosen from photoevaporation, core-powered mass loss, Roche-lobe overflow and a giant impact; the same census counts 272. See [Mini-Neptune](https://paxabyssi.com/wiki/Mini-Neptune.md) and [Exotic worlds](https://paxabyssi.com/wiki/Exotic_worlds.md). ## See also - [Radius valley](https://paxabyssi.com/wiki/Radius_valley.md) - [Hot Neptune desert](https://paxabyssi.com/wiki/Hot_Neptune_desert.md) - [Mini-Neptune](https://paxabyssi.com/wiki/Mini-Neptune.md) - [Hot Jupiter](https://paxabyssi.com/wiki/Hot_Jupiter.md) - [Exotic worlds](https://paxabyssi.com/wiki/Exotic_worlds.md) - [Barren rock world](https://paxabyssi.com/wiki/Barren_rock_world.md) - [Arid world](https://paxabyssi.com/wiki/Arid_world.md) - [Greenhouse world](https://paxabyssi.com/wiki/Greenhouse_world.md) - [Titan](https://paxabyssi.com/wiki/Titan.md) - [Red dwarf](https://paxabyssi.com/wiki/Red_dwarf.md) ## References 1. Catling, D. C. and Kasting, J. F. (2017). Atmospheric Evolution on Inhabited and Lifeless Worlds. Cambridge University Press. 2. Catling, D. C. and Zahnle, K. J. (2009). The Planetary Air Leak. Scientific American 300, 36-43. 3. Hunten, D. M. (1973). The escape of light gases from planetary atmospheres. Journal of the Atmospheric Sciences 30, 1481-1494. [https://doi.org/10.1175/1520-0469(1973)030<1481:TEOLGF>2.0.CO;2](https://doi.org/10.1175/1520-0469\(1973\)030<1481:TEOLGF>2.0.CO;2) 4. Watson, A. J., Donahue, T. M. and Walker, J. C. G. (1981). The dynamics of a rapidly escaping atmosphere: applications to the evolution of Earth and Venus. Icarus 48, 150-166. 5. Hunten, D. M., Pepin, R. O. and Walker, J. C. G. (1987). Mass fractionation in hydrodynamic escape. Icarus 69, 532-549. 6. Erkaev, N. V. et al. (2007). Roche lobe effects on the atmospheric loss from "Hot Jupiters". Astronomy & Astrophysics 472, 329-334. 7. Murray-Clay, R. A., Chiang, E. I. and Murray, N. (2009). Atmospheric escape from hot Jupiters. The Astrophysical Journal 693, 23-42. 8. Owen, J. E. and Alvarez, M. A. (2016). UV driven evaporation of close-in planets: energy-limited, recombination-limited, and photon-limited flows. The Astrophysical Journal 816, 34. 9. Wright, N. J. et al. (2011). The stellar-activity-rotation relationship and the evolution of stellar dynamos. The Astrophysical Journal 743, 48. 10. Jackson, A. P., Davis, T. A. and Wheatley, P. J. (2012). The coronal X-ray-age relation and its implications for the evaporation of exoplanets. Monthly Notices of the Royal Astronomical Society 422, 2024-2043. 11. Owen, J. E. and Wu, Y. (2016). Atmospheres of low-mass planets: the "boil-off". The Astrophysical Journal 817, 107. 12. Owen, J. E. and Wu, Y. (2013). Kepler planets: a tale of evaporation. The Astrophysical Journal 775, 105. 13. Lopez, E. D. and Fortney, J. J. (2013). The role of core mass in controlling evaporation: the Kepler radius distribution and the Kepler-36 density dichotomy. The Astrophysical Journal 776, 2. 14. Owen, J. E. and Wu, Y. (2017). The evaporation valley in the Kepler planets. The Astrophysical Journal 847, 29. 15. Ginzburg, S., Schlichting, H. E. and Sari, R. (2018). Core-powered mass-loss and the radius distribution of small exoplanets. Monthly Notices of the Royal Astronomical Society 476, 759-765. 16. Gupta, A. and Schlichting, H. E. (2019). Sculpting the valley in the radius distribution of small exoplanets as a by-product of planet formation: the core-powered mass-loss mechanism. Monthly Notices of the Royal Astronomical Society 487, 24-33. 17. Li, S. L. et al. (2010). WASP-12b as a prolate, inflated and disrupting planet from tidal dissipation. Nature 463, 1054-1056. 18. Vidal-Madjar, A. et al. (2003). An extended upper atmosphere around the extrasolar planet HD209458b. Nature 422, 143-146. 19. Ehrenreich, D. et al. (2015). A giant comet-like cloud of hydrogen escaping the warm Neptune-mass exoplanet GJ 436b. Nature 522, 459-461. 20. Oklopčić, A. and Hirata, C. M. (2018). A New Window into Escaping Exoplanet Atmospheres: 10830 Å Line of Helium. The Astrophysical Journal Letters 855, L11. 21. Spake, J. J. et al. (2018). Helium in the eroding atmosphere of an exoplanet. Nature 557, 68-70. 22. Hu, R., Seager, S. and Yung, Y. L. (2015). Helium Atmospheres on Warm Neptune- and Sub-Neptune-sized Exoplanets and Applications to GJ 436b. The Astrophysical Journal 807, 8. 23. Jakosky, B. et al. (2018). Loss of the Martian atmosphere to space: Present-day loss rates determined from MAVEN observations and integrated loss through time. Icarus 315, 146-157. 24. Gunell, H. et al. (2018). Why an intrinsic magnetic field does not protect a planet against atmospheric escape. Astronomy & Astrophysics 614, L3. 25. Zahnle, K. J. and Catling, D. C. (2017). The Cosmic Shoreline: The Evidence that Escape Determines which Planets Have Atmospheres, and what this May Mean for Proxima Centauri B. The Astrophysical Journal 843, 122. 26. Donahue, T. M. et al. (1982). Venus Was Wet: A Measurement of the Ratio of Deuterium to Hydrogen. Science 216, 630-633. 27. Kreidberg, L. et al. (2019). Absence of a thick atmosphere on the terrestrial exoplanet LHS 3844b. Nature 573, 87-90. 28. Greene, T. P. et al. (2023). Thermal emission from the Earth-sized exoplanet TRAPPIST-1 b using JWST. Nature 618, 39-42. ## Infobox (physics concept) | Field | Value | | --- | --- | | Name | Atmospheric escape | | Image | File:Gliese_436b_hydrogen_tail_ESAHubble.jpg | | Caption | Artist's concept: the cloud of hydrogen escaping the warm Neptune GJ 436 b, measured by Hubble | | Sim use | Every mini-Neptune the generator makes has its escape worked out: XUV flux from its star's luminosity, type and age, energy-limited and core-powered rates, a Roche-lobe check and an envelope lifetime. The fastest losers become evaporating mini-Neptunes (MNH-EV); stripped giant cores are chthonian planets. | | Symbols | M, M_p planet mass; m molecular mass; T exobase temperature; r exobase radius; v_0 = sqrt(2kT/m); F_XUV X-ray and extreme-ultraviolet flux; eta heating efficiency (of order 0.1); K tidal correction for the star's pull (1 or less) | | Formulae | Jeans parameter at the exobase: lambda = G M m / (k T r) = (v_esc / v_0)^2, Energy-limited escape: dM/dt = eta pi F_XUV R_p^3 / (G M_p K) | | Validity | Jeans escape applies when lambda is large and the gas leaks slowly; the energy-limited formula is an upper estimate for strongly irradiated hydrogen atmospheres, and becomes recombination-limited at the highest fluxes | | Definition | The loss of gas from the top of an atmosphere to space, either thermally (Jeans escape, molecule by molecule; hydrodynamic escape, as a bulk wind) or by non-thermal processes driven by charged particles, photochemistry and impacts. | | Last verified | 2026-10-07 | | Key quantities | Escape velocity (Earth 11.2 km/s, Mars 5.0, Titan 2.6); exobase temperature; X-ray output, about a thousandth of a young Sun-like star's total light during its first 100 million years | | Worked example | Earth's exobase, about 500 km up at about 1,000 K: lambda is about 7 for hydrogen atoms and about 110 for oxygen atoms, so hydrogen leaks away slowly and oxygen stays. | | Observed examples | Mars (MAVEN), Venus (deuterium-to-hydrogen ratio), HD 209458 b, GJ 436 b, WASP-107 b (helium) | | Literature schemes | Jeans (thermal), hydrodynamic and energy-limited, recombination-limited, core-powered mass loss, boil-off, Roche-lobe overflow, non-thermal (sputtering, ion pickup, charge exchange, photochemical escape), impact erosion | ## Related pages - [Hot Neptune desert](https://paxabyssi.com/wiki/Hot_Neptune_desert.md): A scarcity of planets of roughly Neptune's size and mass on orbits of a few days or less, where hot Jupiters and stripped rocky planets are common. Starlight that strips envelopes and the way giant planets migrate inward are the leading explanations. - [Radius valley](https://paxabyssi.com/wiki/Radius_valley.md): A shortage of planets between about 1.5 and 2 Earth radii that splits small close-in planets into rocky super-Earths and gas-wrapped sub-Neptunes. - [Greenhouse world](https://paxabyssi.com/wiki/Greenhouse_world.md): A rocky planet smothered by a thick carbon dioxide atmosphere whose greenhouse effect bakes the surface far above the boiling point of water. Venus is the Solar System's example, a near-twin of Earth in size. - [Mini-Neptune](https://paxabyssi.com/wiki/Mini-Neptune.md): A planet between about 1.7 and 4 times Earth's radius with a rocky or icy core wrapped in a thin envelope of hydrogen and helium. The most common kind of planet Kepler found, and one the Solar System lacks. - [Super-Earth](https://paxabyssi.com/wiki/Super-Earth.md): A planet more massive than Earth but lighter than Neptune, usually rocky, with no counterpart in the Solar System. Super-Earths are among the most common planets found around other stars. - [Lava world](https://paxabyssi.com/wiki/Lava_world.md): A rocky planet so close to its star that its dayside rock is molten, often a magma ocean under a thin atmosphere of vaporised rock. Dozens are known, most on orbits shorter than a day. Categories: [Planets](https://paxabyssi.com/wiki/Category:Planets.md), [Atmospheric science](https://paxabyssi.com/wiki/Category:Atmospheric_science.md), [Planet formation](https://paxabyssi.com/wiki/Category:Planet_formation.md), [Exoplanets](https://paxabyssi.com/wiki/Category:Exoplanets.md)