Revisiting the exoplanet radius valley with host stars from SWEET-Cat

¹ Department of Physics, Faculty of Science, Kyambogo University, PO Box 1, Kyambogo Uganda
² Max-Planck-Institut für Astrophysik, Karl-Schwarzschild-Str. 1, 85748 Garching Germany
³ Instituto de Astrofísica e Ciências do Espaço, Universidade do Porto, Rua das Estrelas, 4150-762 Porto Portugal
⁴ Departamento de Física e Astronomia, Faculdade de Ciências da Universidade do Porto, Rua do Campo Alegre, s/n, 4169-007 Porto Portugal
⁵ Max-Planck-Institut für Sonnensystemforschung, Justus-von-Liebig-Weg 3, 37077 Göttingen Germany
⁶ INAF-IAPS, Via del Fosso del Cavaliere 100, 00133 Roma Italy
⁷ Department of Physics, Mbarara University of Science and Technology, PO Box 1410, Mbarara Uganda

^★ Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Received: 3 October 2025
Accepted: 13 January 2026

Abstract

Context. The radius valley, a deficit in the number of planets with radii around 2 R_⊕, was observed among exoplanets that have sizes of ≲5 R and orbital periods of <100 days by NASA’s Kepler mission. This feature separates two distinct populations: super-Earths (rocky planets with radii ≲1.9 R_⊕) and sub-Neptunes (planets with substantial volatile envelopes and radii ≳2 R_⊕). The valley has been proposed to stem from either planet formation conditions or evolutionary atmospheric loss processes. Disentangling these mechanisms has led to numerous studies of population-level trends, although the resulting interpretations remain sensitive to sample selection and the robustness of host-star parameters.

Aims. Our aim is to re-examine the existence and depth of the radius valley, and how its location varies with orbital period, incident flux, stellar mass, and stellar age.

Methods. We derived robust fundamental stellar parameters of 1221 main-sequence stars (hosting 1405 confirmed planets) from the SWEET-Cat database using a grid-based machine-learning tool (MAISTEP), which incorporates effective temperatures and metallic-ities from spectroscopy, as well as Gaia-based luminosities. Our analysis covers stars with effective temperatures between 4400 and 7500 K (FGK spectral types) and estimated radii between 0.62 and 2.75 R_⊕. We attained a median uncertainty of 2% in both stellar radius and stellar mass. Combining the updated stellar radii with planet-to-star radius ratios from the NASA Exoplanet Archive, we recomputed the planetary radii, achieving a median uncertainty of 5%.

Results. Our findings confirm a partially filled planet radius valley near 2 R_⊕. The valley depends on the orbital period, incident flux, and stellar mass, with slopes of −0.12_−0.01^+0.02, 0.10_−0.03^+0.02, 0.19_−0.07^+0.09, respectively. We also find a stronger mass-dependent trend in average sizes of sub-Neptunes than super-Earths of slopes 0.17_−0.04^+0.04 and 0.11_−0.05^+0.05, respectively. With stellar age, the super-Earth/sub-Neptune number ratio increases from 0.51_−0.08^+0.11 (<3 Gyr) to 0.64_−0.11^+0.11 (≥3 Gyr). In addition, the valley becomes shallower and shifts to larger radii, indicating age-dependent evolution in planet sizes. A four-dimensional (planet radius, orbital period, stellar mass, and stellar age) linear fit of the valley in log-log space produces slopes in orbital period and stellar mass that are consistent with the results from the two-dimensional analyses, and a weaker slope of 0.07_−0.04^+0.03 in stellar age.

Conclusions. The valley’s shift and shallowing over gigayear timescales point to prolonged atmospheric loss, which is consistent with a core-powered mass-loss scenario. Our findings also highlight the importance of stellar age in the interpretation of exoplanet demographics and motivate improved age determinations, as is expected from future missions such as PLATO.