One Cosmic Shoreline Becomes Two, and an Airless Valley Opens Between Them

Draw every known rocky planet on one chart, how hard its star bakes it, against how firmly its gravity holds on, and a line appears. Below the line, worlds keep an atmosphere. Above it, they are stripped down to rock. Astronomers call that line the cosmic shoreline, and it describes the solar system well enough that it became the default expectation everywhere else.

JWST has been returning planets on the wrong side of it, from both directions. Some of the hottest worlds known, orbiting their stars in less than a day, hot enough at the surface to be molten, show spectra that seem to demand a substantial atmosphere. Cooler rocky planets around small red stars, which should have held their air comfortably, keep coming back looking like bare rock.

A study published Tuesday in The Astrophysical Journal Letters argues that the line was never one line. Barron K. Nguyen and Laura K. Schaefer, planetary scientists at Stanford, and colleagues ran a model in which a planet's interior and its atmosphere evolve together over billions of years. It produces two boundaries: a hot one they call the cosmic sandbar, a cooler one that is the shoreline as previously understood, and between them a band they name the airless valley. It is a model built to explain other people's telescope data, and the authors present it as one. The modeling is peer-reviewed.

Those observations are mixed, and stronger on one side than the other. On the cool side, the case is clean. JWST measurements of the innermost TRAPPIST-1 planets show daysides too hot and too poorly stirred to sit beneath a thick atmosphere; TRAPPIST-1 b reads as essentially airless. On the hot side, the evidence is real but contested. 55 Cancri e is the archetype, a lava world whose JWST spectrum was read in 2024 as evidence of a secondary atmosphere, an interpretation other groups have since challenged. TOI-561 b is the other much-discussed case. The paper's own introduction calls these detections tentative.

The mechanism the model proposes for the sandbar is storage rather than resupply. A planet close enough to its star stays substantially molten, in some cases for billions of years, and a magma ocean is an excellent solvent. It holds most of the planet's water and carbon dissolved in the melt, leaving only a thin skin of gas above the surface for the star to work on. That skin is stripped efficiently. But it is a small target, and as it thins, the melt beneath keeps buffering what is left. In the authors' words, such a magma ocean "outgasses only minimally."

A cooler planet does the reverse. It solidifies, and solidifying is not neutral: it traps a large share of the volatiles in the deep solid mantle, out of circulation, while forcing the rest out into an atmosphere that is now fully exposed. Such a planet has less left to lose and loses it faster. That, in the model, is the airless valley.

Staying molten for that long takes help. In systems with several planets, neighbors tug one another's orbits into slight ellipses, and the flexing that follows heats the interior, the same process that keeps Io erupting. In the model's nominal TRAPPIST-1 case, that tidal heating stretches magma ocean lifetimes out to billions of years.

"Airless" is a definition here rather than a measurement. In the model, it means a residual atmosphere below 0.01 bar, a Mars-like benchmark the authors chose. And what sits in the valley may not be ordinary rock: the band may mark a graveyard of stripped sub-Neptune cores, the leftovers of planets that once carried thick hydrogen envelopes and lost them.

What lifts this above a nice picture is that it can be broken. Lava worlds with thick atmospheres should be unlikely around stars cooler than K-type unless extreme tidal or interior heating keeps them molten, so a thick atmosphere on a lava world orbiting a red dwarf would be a real problem for it. The authors also publish fitted curves for both boundaries across G, K and M stars. The fits run on a planet's volatile budget, mass, age and tidal heating, and are tabulated for direct use in choosing which planets to observe.

The model leaves out non-thermal escape processes such as ion pickup and sputtering, along with impact erosion, core formation and nitrogen. Its baseline is the same group's earlier work on 55 Cancri e, so no independent implementation has yet reproduced the two boundaries.

The observations that would test it are already scheduled. JWST's Rocky Worlds program, run on director's discretionary time, and continuing observations of the TRAPPIST-1 planets are aimed squarely at whether small worlds around red dwarfs hold any atmosphere at all. The sandbar makes a specific claim about exactly those planets, and it is the kind of claim a telescope can settle.