The radius valley — a deficit of planets between roughly 1.5 and 2 Earth radii — was supposed to mark a clean boundary: rocky planets below, volatile-rich planets above. Photoevaporation or core-powered mass loss strips smaller planets of their atmospheres, leaving bare rocky cores, while larger planets retain their gas envelopes. The picture predicts two populations with a gap between them.
Palethorpe and colleagues measure masses and radii across six multi-planet systems with precisions better than 15% and 5% respectively, using CHEOPS, K2, TESS, and ground-based radial velocities. What they find is not two populations but a continuum. Rocky worlds sit at the valley's lower edge. Above them, intermediate-density planets require volatile envelopes to explain their bulk densities — but the envelopes are not uniform. Some are consistent with thin hydrogen-helium atmospheres. Others require water-rich compositions — “steam worlds” with volatile fractions too large for accreted gas alone. Still others are low-density water-worlds that challenge both the photoevaporation and core-powered mass loss mechanisms.
Note (2026-11-15; read against the paper's full text): The paper reports compositions for six systems (K2-79, K2-106, K2-111, K2-222, K2-263 and TOI-1634): rocky planets near the radius valley, intermediate-density planets needing steam-rich or mixed volatile envelopes, and low-density planets consistent with gas dwarfs or water-worlds. I found no statement that the radius valley is a continuum or that the earlier clean dichotomy was an artifact of measurement precision (the words continuum, artifact and dichotomy do not appear), and six systems could not establish that. The 50 percent mass-uncertainty comparison and the claim that the categories dissolve at higher resolution are the essay's own.
Several systems show radial-velocity evidence for additional planets that don't transit — companions detectable gravitationally but invisible photometrically, whose compositions remain unconstrained. The observed population is already a continuum; the unobserved population could fill it further.
The clean dichotomy was an artifact of measurement precision. With 50% mass uncertainties, a rocky planet and a water-world look the same. At 15%, they separate — but they separate into a spectrum, not into two bins. The categories were real at the resolution where they were defined. Higher resolution didn't sharpen the boundary; it dissolved it.