A new Stanford-led model offers an explanation for how some lava-covered exoplanets can retain thick atmospheres for billions of years despite orbiting close enough to their stars for intense radiation to steadily strip gas into space.

The key is not that the planets avoid atmospheric loss. Instead, their molten interiors can release new gas slowly enough to replenish what is being lost, according to the Stanford Doerr School of Sustainability.

The study, led by Barron Nguyen and published in The Astrophysical Journal Letters, combines the evolution of a planet's interior with atmospheric escape rather than treating the atmosphere separately from the molten world beneath it.

Why the lava changes atmospheric survival

Researchers modeled the exchange of gas between an atmosphere and a molten surface while also accounting for atmospheric escape, planetary cooling and the eventual solidification of the magma ocean.

The model indicates that long-lived magma oceans can keep much of a planet's volatile material dissolved in molten rock. That means only a smaller fraction of the total gas supply is exposed in the atmosphere at any one time.

As stellar radiation removes atmospheric gas, additional material can gradually escape from the molten interior through outgassing. On sufficiently hot worlds, the rate of replenishment can balance atmospheric escape closely enough for a substantial atmosphere to persist for billions of years in the model.

This creates a regime the researchers call the cosmic sandbar. It extends the existing "cosmic shoreline" framework, which describes a boundary between rocky planets able to retain atmospheres and those expected to lose them.

The new work argues that there is not one atmospheric-survival boundary. Instead, there can be a hot, outgassing-controlled sandbar and a cooler cosmic shoreline, with an airless valley between them.

Why some cooler planets can lose their air

The model produces a counterintuitive result. A planet slightly farther from its star can sometimes be less able to maintain an atmosphere than a hotter lava world.

That happens because the cooler planet can solidify more quickly. Once gases become trapped in its deeper solid mantle, the planet may no longer outgas rapidly enough to replace the atmosphere being removed by stellar radiation.

By contrast, a hotter and more massive lava world can remain molten for longer and continue supplying gas to its atmosphere.

The study therefore shifts part of the question from simply how much radiation a planet receives to how atmospheric escape interacts with planetary mass, internal heat, volatile supply and the rate at which the interior cools.

55 Cancri e and TOI-561 b show why the question matters

One of the clearest examples motivating the work is 55 Cancri e, a super-Earth that orbits extremely close to its star. James Webb Space Telescope observations reported in 2024 provided evidence that the planet possesses a substantial atmosphere despite conditions that should promote severe atmospheric loss.

More recently, observations have added other unusually hot rocky worlds to the discussion. NASA reported that Webb measurements of TOI-561 b are best explained by a thick, volatile-rich atmosphere above a magma ocean rather than a bare rocky surface.

TOI-561 b completes an orbit in less than 11 hours and receives extreme stellar radiation. Its apparent atmosphere is therefore another example of the type of close-in lava world that atmospheric-loss models must explain.

The Stanford study does not establish that every such planet preserves its atmosphere through exactly the same process. Its result is a theoretical model showing a physically plausible route by which molten interiors can regulate atmospheric supply over very long periods.

The model makes testable predictions

The paper calculates how the proposed boundaries change with stellar type, planetary mass, age, volatile inventory and tidal heating.

One of its more specific predictions is that lava planets with thick atmospheres should be unlikely around stars cooler than K-type stars unless exceptionally strong tidal heating or another source of internal heat keeps their interiors molten.

Tidal heating could matter particularly in multi-planet systems, where gravitational interactions can help maintain orbital eccentricity and continuously generate heat inside a planet.

These predictions give astronomers something that future observations can test. If surveys find atmosphere-bearing lava worlds where the model predicts sustained outgassing, while planets in the proposed airless valley are predominantly bare, the cosmic sandbar framework would gain observational support.

For now, the confirmed result is narrower: the researchers have demonstrated through coupled atmosphere-and-interior modeling that extreme stellar irradiation does not automatically require a lava planet to become airless. Under the right conditions, a molten interior can keep supplying gas for billions of years even while the star continuously removes it.