Does the radius valley form in 100 Myr or over billions of years?
In plain words
Photoevaporation should finish stripping planets within about 100 million years, while core-powered loss takes about a billion. Comparing young and old planet populations can separate them.
Precise statement
Measure the ratio of sub-Neptunes to super-Earths and the gap location in clusters, associations and field stars with ages $10^7-10^10\ \mathrm{yr}$. Photoevaporation predicts the ratio settles by $\sim 10^8\ \mathrm{yr}$; core-powered mass loss predicts evolution until $\sim 1-2x10^9\ \mathrm{yr}$. The answer is the measured time dependence with enough planets per age bin to distinguish the two.
What would settle it
Ages from cluster membership, gyrochronology or asteroseismology for several hundred small planets with precise radii.
Status in the literature
Analyses of 2020-2023 (Berger et al. 2020; David et al. 2021; Christiansen et al. 2023) report the small-planet population changing over Gyr timescales, which favors core-powered mass loss; samples of planets younger than $10^{8}\,\mathrm{yr}$ remain small.
Related problems
- Special case of What produces the gap between super-Earths and sub-Neptunes?