TRANSMISSION // 2611.21436v1

Gaia links outer companions with small planets

Source published

A Gaia astrometry study finds more outer companions around stars with super-Earths and sub-Neptunes, while leaving the proposed formation pathway uncertain.

AI-generated conceptual illustration for Gaia links outer companions with small planets
AI-generated conceptual illustration; not documentary evidence.

Do distant giant planets and brown dwarfs hinder the birth of smaller worlds, or can both kinds of planets share a system? A study using Gaia measurements finds an excess of outer companions around stars with inner super-Earths and sub-Neptunes. The result is a population-level association, not proof of how or where the planets formed.

The question: do outer companions suppress small planets?

A massive planet or brown dwarf orbiting far from a star can reshape the disk of gas and dust from which planets form. By opening a gap, it may limit the inward flow of solid material, potentially making it harder for small planets to form closer in. But outer companions and inner planets might also form in the same system without one preventing the other.

A team led by Alexandra Rochon used Gaia astrometry to test whether stars with known inner small planets are more or less likely than comparable stars to host outer companions. The study focuses on super-Earths and sub-Neptunes, categories defined here by planet radius rather than direct measurements of composition. The authors define outer companions as objects between 1 and 80 Jupiter masses, orbiting at 1–20 astronomical units. The data are not equally sensitive across that entire range.

How the comparison worked

The analysis began with 7,321 nearby FGKM dwarf stars: 678 stars with known transiting small planets and 6,643 comparison stars without known planets. The researchers separated planet-host samples into systems containing only super-Earths, only sub-Neptunes, or both. They matched comparison stars to planet hosts using six properties—including stellar mass, brightness, distance, and measures of astrometric data quality—to make the groups more comparable in their ability to reveal companions.

To search for companions, the team fitted Gaia DR2 and DR3 astrometric measurements with the orbit-modelling program Octofitter. Astrometry measures a star’s apparent motion across the sky; a companion can tug the star away from a simple straight-line path. Rather than label every star as a definite detection or non-detection, the researchers combined uncertain results across each sample using a Bayesian hierarchical model. This approach can incorporate signals too weak to establish a companion around an individual star.

The central comparison is a relative occurrence rate: the inferred frequency of outer companions among planet hosts divided by the frequency among matched comparison stars. It is not an estimate of the absolute fraction of stars with companions.

What the study found

Across the full sample, the estimated relative occurrence was 1.77 ± 0.49. In other words, the model indicates more outer companions in systems with known inner small planets than in the matched comparison population. The estimated ratios were similar for the two planet categories: 2.02 ± 0.52 for super-Earth systems and 1.97 ± 0.50 for sub-Neptune systems. The authors report that the ratios exceed one with 99.3% and 98.2% confidence, respectively, under their analysis.

The study therefore finds no sign that the measured outer-companion population is less common around either category of inner planet. It also finds no significant difference between the super-Earth and sub-Neptune estimates. This does not show that outer companions have no influence on planet formation; it means the results do not support a simple picture in which they prevent inner small planets from existing.

The estimated excess appeared somewhat stronger around M dwarfs than around FGK stars, by a factor of 1.41 ± 0.78. That uncertainty is substantial, and several planet-host subgroups contain fewer than 100 systems, so the apparent difference should be treated cautiously. The researchers also found no strong dependence on stellar metallicity in their relative comparison. That result is limited by mixed sources of metallicity estimates, sizeable measurement uncertainties, and possible mixing between the study’s metal-rich and metal-poor groups.

What the result does—and does not—say

The authors interpret the similar super-Earth and sub-Neptune results as support for a scenario in which small planets form farther out and migrate inward before an outer companion opens a disk gap. They propose Type I migration, in which a planet moves through its interaction with the surrounding disk. This is a possible explanation, not a direct observation of planets forming or migrating. Other histories may also be consistent with the measured association.

There are important limits. Gaia DR2 and DR3 provide a relatively short astrometric baseline for this search, and the study excludes Hipparcos observations to keep its dataset consistent. The authors say their analysis is most sensitive to companions above roughly 10 Jupiter masses, so it is especially informative about brown dwarfs and other massive companions—not ordinary gas giants across the full stated range. A sensitivity mask excludes regions where results are strongly driven by model assumptions, but the effective coverage is narrower than the nominal 1–20 AU and 1–80 Jupiter-mass bounds.

The planet sample also includes vetted transit candidates, which may still contain false positives, and reference-star matching is not perfect in every subgroup. The metallicity and stellar-mass comparisons are less secure than the overall result. Future Gaia data could improve sensitivity and test these patterns more precisely.

The source is a November 2026 arXiv preprint (version 1). The supplied material does not establish its peer-review status. As with any single analysis, its findings remain open to further scrutiny and testing with additional data. Read the source.

Automatically matched by shared topic keywords. These links do not imply that the sources agree or establish a consensus.