David A. Brain, Ofer Cohen, Thomas E. Cravens, Kevin France, Alex Glocer, Parker Hinton, Francois Leblanc, Yingjuan Ma, Akifumi Nakayama, Shotaro Sakai, Ryoya Sakata, Kanako Seki, Julián D. Alvarado-Gómez, Zachory Berta-Thompson, Eryn M. Cangi, Michael Chaffin, Jean-Yves Chaufray, Renata Frelikh, Yoshifumi Futaana, Katherine Garcia-Sage, Lukas Hanson, Mats Holmström, Bruce Jakosky, Riku Jarvinen, Ravi Kopparapu, Daniel R. Marsh, Aimee Merkel, Thomas Earle Moore, Yuta Notsu, Rachel A. Osten, William K. Peterson, Laura Peticolas, Robin Ramstad, Kevin B. Stevenson, Robert Strangeway, Wenyi Sun, Naoki Terada, Aline A. Vidotto
10 min
Abstract
Atmospheric escape is an important process that influences the evolution of planetary atmospheres. A variety of physical mechanisms can contribute to escape from an atmosphere, including thermal escape, ion escape, photochemical escape, and sputtering. Here we estimate escape rates via each of these processes for a hypothetical Mars-like exoplanet orbiting Barnard's star (an old, inactive M dwarf star). We place the planet at an orbital distance that receives the same total stellar flux as it does in our solar system. We use the measured stellar extreme ultraviolet (EUV) spectrum and assumptions on the star's magnetic field to determine both the high-energy radiation and the stellar wind environment around the planet. This information is used to model the response of the planet's thermosphere, exosphere and magnetosphere using a variety of models that have been validated against solar system observations. We find overall escape rates that are dominated by thermal processes and elevated by 2-5 orders of magnitude relative to present-day Mars, suggesting that a Mars-like planet orbiting Barnard's star would not retain a significant atmosphere for more than 10's of millions of years. Recently reported planets around Barnard's star should also not have retained significant atmospheres. By extension, Mars-like planets orbiting any M dwarf near the 'Habitable Zone' should not retain atmospheres for extended periods of time.
Alex: Okay, but isn't there more to loss than just heat? Like the star's wind picking off particles?
Sam: There is—the study tallies five escape routes together, which is rare. Beyond thermal, non-thermal processes kick in: ions get yanked away by electric fields in the star's wind, photochemistry creates fast-moving fragments from UV reactions, and sputtering blasts atoms loose via particle collisions. For ion escape, they ran three independent magnetohydrodynamic—or MHD—models. Think of MHD like simulating a flowing river of charged particles with magnetic fields twisting through it, slamming into the planet's weak shield. Models like BATS-R-US, REPPU, and MAESTRO all agree on oxygen and carbon ion loss rates around 100 times today's Mars.
Alex: So the wind's denser and magnetically stronger there too?
Sam: Correct—stellar wind density is roughly 15 times higher, magnetic field several times stronger, enabling those consistent pickup rates across models. Combined, these paths strip the atmosphere 100 to 100,000 times faster than now, explaining bare rocks around even old M-dwarfs. The paper notes uncertainties like eddy mixing, but the multi-model check adds confidence.
Alex: So those Jeans rates for heavier atoms like oxygen and carbon—how do they get high enough to matter, given Mars' gravity?
Sam: Intense far-ultraviolet light breaks apart CO2 molecules into oxygen, carbon, and CO—much more efficiently than on today's Mars. Barnard's Star emits strongly at wavelengths between 100 and 120 nanometers that zip through the thin upper air. This floods the thermosphere with those heavier atoms, which then get hot enough at 4000 Kelvin to have a good shot at escaping. The total Jeans rate across species ends up comparable to or higher than today's Mars, but for a wider range of atoms, not just hydrogen. Carbon and oxygen lead at around 10 to the 28th atoms per second.
Alex: Wait—carbon ions dominate up there? Why not oxygen, like on our Mars?
Sam: Oxygen ions deplete fast above 1000 kilometers through charge-exchange reactions, where they swap electric charge with neutral carbon atoms, turning C into C-plus. That reaction's rate is about 200 times faster than ways to neutralize oxygen ions at those temperatures, so C-plus takes over as the main ion.
Alex: And they checked if this turns into full-blown hydrodynamic flow?
Sam: Yes—they tested with PLANET-ITTR, a one-dimensional model that solves fluid equations for the upper atmosphere. The gas speeds up past sound speed around 16,000 kilometers and hits escape velocity by 20,000, suggesting bulk hydrodynamic outflow. The Jeans parameter of about 5 puts it in a transitional regime where both tail escape and whole-flow loss can happen together. But it's a simplified run without chemistry or full cooling, so it just flags the possibility—more work needed to confirm.
Alex: Okay, so ion pickup is huge, but what about photochemical escape—does that add much here?
Sam: Photochemical escape happens when UV light breaks molecules into charged pieces, then those ions grab electrons and shatter into super-fast atoms that can flee to space—like fireworks exploding from a spark. The higher UV makes it about 30 times stronger at first glance, but the upper air has fewer of those exact ions and more singles like C-plus, so models adjust it down to around 10 to the 26th oxygen atoms per second lost.
Alex: And sputtering—particles slamming into the air and knocking stuff loose?
Sam: Sputtering is when incoming ions crash into neutral atoms, handing over energy so some escape—like billiard balls knocking others off the table. Here, those ions don't speed up enough before hitting the puffed-up exobase, so the yield is under 1—meaning the planet actually gains more particles than it loses. It's near zero, unlike today's minor sputtering, because the stretched-out gas layer acts like a cushion.
Alex: Right—so all these together?
Sam: The paper sums five paths: thermal and possible hydrodynamic for heavies at 10 to the 28th or more, ions at 1 to 3 times 10 to the 28th, photochemical around 10 to the 26th, sputtering negligible. Total conservative oxygen loss hits about 6 times 10 to the 28th per second. For a thin CO2 air like today's Mars, that's gone in roughly 350,000 years; a thicker early one, about 50 million—far quicker than prior models predicted, even without flares. It fits JWST's bare rocks perfectly.
Alex: With escape so rapid, does the planet have any way to replenish its air from inside, like volcanoes pumping out gases?
Sam: Planets can release gases trapped deep inside through heat-driven processes, similar to how Earth's volcanoes belch out carbon dioxide from molten rock below the crust. Studies estimate Mars today might supply about 1 bar of CO2 total—roughly Earth's sea-level pressure. The paper notes this mostly happened in the first few hundred million years, so for an older M-dwarf system, escape would outpace it after that early burst.
Alex: So even if it outgassed a full atmosphere early on, it'd still vanish quickly?
Sam: Yes—for 1 bar, the modeled loss timescale is around 50 million years, conveniently after early outgassing wraps up. Gases might also hide in surface rocks or underground pockets, then release later, but that could only sustain an observable atmosphere for a tiny fraction of the star's life—like 0.5% for Barnard's Star.
Alex: And speaking of Barnard's—didn't they just find smaller planets there?
Sam: Recent observations spotted four sub-Earth worlds orbiting much closer in—2.3 to 4.6 times nearer than the modeled exo-Mars spot. Energy from EUV scales with the inverse square of distance, so they'd get 5 to 21 times more, driving faster escape despite their slightly higher gravity slowing it somewhat. The paper argues secondary atmospheres are unlikely there too.
Alex: Makes sense why JWST sees bare rocks. But these are rough estimates—any big caveats in the models?
Sam: Several—the paper highlights challenges like unmeasured EUV for Barnard's, estimated via a technique reconstructing it from nearby wavelengths, which can be off by factors of a few. Stellar wind inputs used a proxy star's map since none exists. One-dimensional thermosphere models assume even heating, ignoring tidal locking where one side always faces the star. Ion models vary by up to three-fold due to different setups.
Alex: So trends are clear—enhanced loss even from quiet stars—but details need refining.
Sam: Precisely. The multi-process tally and model checks build confidence in the overall rapid stripping, aligning with no atmospheres spotted yet on M-dwarf habitables. It suggests observing one would be rare timing.
Alex: A solid step in piecing together why habitability seems so tricky there. Thanks for breaking it down, Sam.
Sam: My pleasure, Alex.