The Lava Planet That Shouldn't Have a Sky
JWST's 37-hour observation of TOI-561 b reveals a silicate-cloud atmosphere reflecting 75% of starlight above a magma ocean -- a finding that upends expectations for irradiated close-in rocky worlds.
Most models of close-in rocky planets around their host stars expect the atmosphere to be gone. Billions of years of high-energy radiation, thermal escape, and sputtering should strip the volatiles down to nothing, leaving a bare, airless rock orbiting a star that’s been trying to boil it off since the system formed. TOI-561 b, a lava world 280 light-years from Earth, apparently disagrees with that assumption.
The James Webb Space Telescope spent 37 hours watching TOI-561 b across roughly four of its orbits, and the resulting observations show the planet reflecting approximately 75% of the starlight that reaches it – a brightness signature that demands an extended, scattering atmosphere rather than a bare surface. For context, Mercury, with its tenuous exosphere, reflects a tiny fraction of incident sunlight. A value that high means light is bouncing off a substantial layer of particulates suspended above the planet’s surface, not off a dark, airless rock.
The composition the team inferred points to silicate hazes: clouds of silicon dioxide (SiO₂) and magnesium silicate (MgSiO₃) floating above a global lava ocean. In that architecture the clouds function less like weather and more like a protective layer – a mineral sunscreen sitting on top of a planet that is, at its core, an ocean of molten rock. The silicate particles are light enough to be lofted and suspended by thermal escape from the magma below, stable enough to persist as a coherent reflective layer, and chemically robust enough that the star’s radiation scatters off them rather than breaking them apart on every pass through the planet’s thermosphere.
The harder question is what keeps the atmosphere alive. A one-time outgassing event would be exhausted on geological timescales, and the irradiation environment around TOI-561 b’s star should accelerate that loss. The Birmingham team’s working hypothesis is that the atmosphere is being continuously replenished by the magma ocean below, in a cycle analogous to the SO₂ cycle on Io or the volcanic resurfacing that has maintained Venus’s atmosphere for billions of years. If the lava ocean is still convecting, still exsolving volatiles, then the atmospheric loss to space is matched by a steady source from below, and the planet can maintain its silicate shroud indefinitely.
An independent analysis of the same JWST dataset, reported on October 27, 2026, reached a complementary conclusion from the thermal side. Lisa Dang at the University of Waterloo led a phase-curve analysis of TOI-561 b’s reflected and thermal emission, with Anjali Piette of the University of Birmingham as a co-author, and the result was a dayside cooler than standard irradiated-planet models predict and a nightside too cold to account for the observed temperature difference on its own. That mismatch is exactly the fingerprint of a thick, reflective upper layer: the silicate clouds intercept incoming starlight and scatter it back before it can heat the surface, keeping the dayside cooler it would be without an atmosphere, while the nightside temperature reflects the limited heat the surface can radiate away.
What makes the finding genuinely difficult to reconcile with existing models is the timescale. TOI-561 b is not a young, freshly volcanic world with a transient plume that will disperse in a few million years. It has been orbiting its star for billions of years, absorbing a constant flux of extreme ultraviolet and X-ray radiation that should, by every standard escape-rate calculation, have stripped even a massive atmosphere long ago. The surviving silicate layer is therefore evidence either that the replenishment mechanism is more efficient than current models allow, or that the loss processes operating on a lava planet with a magma-ocean source are qualitatively different from those that govern rocky worlds with solid surfaces and thin volatile atmospheres.
Neither of those resolutions is trivial. If the first is correct, the atmospheric lifetimes predicted for the many other close-in rocky planets already observed or soon to be observed by JWST need to be revised upward substantially. If the second is correct, then the physics of volatile transport from a convecting magma ocean into a radiating, irradiated upper atmosphere is not yet adequately captured in the circulation and escape codes that the field currently uses. Either way, a small, bright, mineral world 280 light-years away has just put a constraint on models the community did not expect to need.