When you plot the sizes of all known exoplanets — thousands of them now, catalogued from Kepler, K2, and TESS transit surveys — against how many planets exist at each size, something odd appears. The distribution is not smooth. There are abundant planets smaller than about 1.5 Earth radii. There are abundant planets larger than about 2 Earth radii. And between those two populations, there is a gap: a statistical valley where planets are conspicuously rare. This is not a selection effect or observational bias. It is real, reproducible across independent data sets, and it is telling us something fundamental about how planets lose their atmospheres.
The gap was first clearly identified in 2017, when Caltech astronomers Benjamin Fulton and Erik Petigura analyzed a sample of 1,300 Kepler planets with precisely measured stellar radii — the latter being critical because planet size as measured by transit depth is given in stellar radii, so errors in the star's size propagate directly into errors in the planet's size. With a clean sample and accurate stellar radii, the bimodal distribution became unambiguous: a peak of rocky planets below 1.5 Earth radii, a gap, and then a peak of larger planets at 2 to 3 Earth radii that likely retain substantial hydrogen and helium envelopes. The valley sits at roughly 1.5 to 2 Earth radii, and it is mostly empty.
Photoevaporation: the leading mechanism
The explanation that has gathered the most observational support involves high-energy radiation from host stars stripping planetary atmospheres away. Early in a planetary system's history — the first hundred million years or so — stars emit intense extreme ultraviolet and X-ray radiation that can heat atmospheric gas to escape velocity and drive it off into space. For small planets with weak gravity, this process, called photoevaporation, can strip away the entire hydrogen envelope over relatively short timescales, leaving behind a bare rocky core. For larger planets with stronger gravity, the atmosphere can resist this stripping and the planet retains its gaseous envelope.
The radius gap falls naturally out of this mechanism. A planet with a small rocky core accumulates a thin hydrogen envelope during formation. If it sits close enough to its star to receive substantial high-energy radiation, photoevaporation strips the envelope on a timescale of tens to hundreds of millions of years, leaving a rock slightly larger than Earth. If the same planet had a bigger core — enough gravity to resist atmospheric escape — it retains its envelope and appears as a sub-Neptune or mini-Neptune with a radius above two Earth radii. The gap between these two outcomes is the radius valley.
The competing explanation: core-powered mass loss
A second mechanism, core-powered mass loss, produces similar outcomes but operates on different timescales and through different physics. In this model, the heat trapped inside a planet's rocky core from formation — residual thermal energy from accretion and differentiation — slowly diffuses outward over billions of years and powers a steady outflow of atmospheric gas. Unlike photoevaporation, which is driven by the star and operates most strongly early in the system's history, core-powered mass loss operates over longer timescales and is driven by the planet's own internal heat.
Distinguishing between these two mechanisms observationally has proved challenging because they produce similar outcomes: a depleted population of planets in the radius gap. But they make subtly different predictions about how the gap position should depend on stellar mass, orbital period, and system age. Observational surveys comparing gap positions in young stellar clusters versus old field stars are testing these predictions.
What JWST is finding
The James Webb Space Telescope is now directly probing the atmospheres — or lack thereof — of planets in and near the radius gap. Transmission spectroscopy, where starlight filtered through a planet's atmosphere during transit is compared to starlight outside transit, can detect the chemical fingerprints of atmospheric gases. Planets with no atmosphere show flat, featureless transmission spectra; planets with substantial atmospheres show absorption features from water vapor, methane, carbon dioxide, or hydrogen.
Early JWST results on planets near the radius gap are consistent with the atmospheric stripping picture: planets at the smaller end of the gap show featureless spectra suggestive of bare or nearly bare rocky surfaces, while larger planets with similar orbital periods show atmospheric features. This is not yet definitive — sample sizes are still small — but the trend is what the photoevaporation and core-powered mass loss models would predict.
Implications for Earth-like planets
The radius gap has a direct implication for the frequency of truly Earth-like worlds. Earth sits below the gap — it is a rocky planet without a massive hydrogen envelope. The existence of the gap suggests that planetary formation naturally produces both rocky planets and sub-Neptunes, and that the boundary between these populations is set by atmospheric physics rather than by the initial composition of the protoplanetary disk.
Every planet in the radius gap is either a world that recently lost its atmosphere or one that is in the process of losing it. The gap is not a gap in the population — it is a gap in stability. Planets at that size don't persist. They either lose their envelopes and become rocky planets, or they don't lose them and become sub-Neptunes. The middle ground, cosmically speaking, is not a place that worlds tend to stay.
