Study using TESS data independently detects Neptunian ridge

Amadeo Castro-González’s analysis of TESS data independently detected the Neptunian ridge, a concentration of exoplanets at orbital periods of roughly three to six days, in a study posted to arXiv on Nov. 2, 2026. The ridge sits between the short-period Neptunian desert and the moderately populated Neptunian savanna at longer periods. Classified under Earth and Planetary Astrophysics, the paper was submitted as version 1 at 18:58:36 UTC.

Castro-González et al. (2024a) originally identified the ridge at periods of 3.2 to 5.7 days among planets of 5.5 to 8.5 Earth radii. That work separated the short-period desert from the savanna at longer periods, and the period range resembles that of the hot Jupiter pileup. The new analysis builds on those earlier Kepler findings and offers insight into the architecture of close-in exoplanet systems.

For planets of about five to eight Earth radii, occurrence density on the ridge is double that of the denser adjacent region, a contrast the study quantifies as J=2.05+0.52−0.44. The researchers compared six occurrence profiles through leave-one-out cross-validation and Bayesian evidence, methods that weigh competing shapes for how planets are distributed across the desert, ridge, and savanna.

The ridge exceeds both neighboring regions in 6,953 out of 6,960 cases, giving p+=99.899% and Z_eq=3.09. The analysis presents this as strong posterior support for a local occurrence excess among planets of 5.02 to 7.96 Earth radii in the TESS data.

The work draws on TESS occurrence-rate posteriors from Cui et al. (2026a), a combined sample of FGK stars, or Sun-like stars spanning the F, G and K classes. The archive holds 6,960 posterior samples on a 10-by-10 grid covering orbital periods of 0.5 to 16 days and planetary radii of two to 20 Earth radii.

Model comparison favors a lognormal profile over a symmetric Gaussian, implying a narrower inner rise and a more extended outer decline, with only weak evidence for asymmetry (ΔlnZ_LN,G=1.90). The lognormal peaks at 4.35+0.37-0.30 days, inside the ridge, and improves on the Gaussian by 4.50±1.33 in ELPD, a predictive-accuracy measure, and on the broken power law by 1.14±0.41.

Bayesian evidence strongly favors both the broken power law and the lognormal over a simple power law, with ΔlnZ values of 6.41 and 8.15 respectively. The broken-power-law fit places its break at 3.73+0.73-0.61 days, within the ridge, with a rising inner slope of 2.30+1.18-0.84 and a falling outer slope of -1.30+0.53-0.74.

The researchers split the planet selection into 5.02 to 6.32 Earth radii and 6.32 to 7.96 Earth radii, analyzing the intervals separately and then together. Separately, the contrasts are J=1.62+0.71-0.51 and J=2.54+1.14-0.77, but neither interval individually reaches 3σ posterior support.

For the full 5.02 to 7.96 Earth radius selection, occurrence on the ridge beats the power-law prediction, with K=2.38+0.57-0.45 and K above 1 across all 6,960 samples. Separate posteriors for F, G, and K stars show no significant differences in ridge contrast.

The same period window also contains a hot Jupiter pileup at roughly 10 to 16 Earth radii, with a contrast matching the ridge. At the smaller end, the analysis excludes an analogous excess for sub-Neptunes of roughly two to four Earth radii, with 3.4σ posterior support, indicating that the ridge is specific to Neptunian-sized planets rather than appearing across all small planet populations in the same period range.

The study links the desert and its neighboring overdensities to high-eccentricity migration, in which gravitational perturbations drive planets onto highly eccentric orbits and tides near periastron determine whether they are disrupted or circularized. A sharper inner edge could result from tidal disruption, while a more gradual outer decline could stem from variations in circularization efficiency. Kepler-derived central contrasts of 1.6 to 2.2 match the TESS result within reported uncertainties.

Fukuda’s team reported on June 12, 2026, that TOI-1883 b, a super-Neptune or Neptunian-sized planet, has a radius of 5.65 Earth radii and a mass of about 13.7 Earth masses, giving a mean density of 0.4 g/cm3. The planet orbits an M-dwarf star, a small cool star, 383 light-years away every 4.5 days, placing its 4.5-day orbit within the subclass’s 3.2-to-5.7-day period range. Among super-Neptunes around M dwarfs, it has the lowest density recorded so far and is likely extremely inflated and puffy. Its host star, TIC 348755728, is about half the size and mass of the Sun.