What this research found
Beta Pictoris became a genuine multi-planet dynamical problem once a third giant, the roughly 2.4 Jupiter-mass Beta Pic d, was directly imaged. K-Dense built a self-consistent N-body model of all three known giants around the 1.789 solar-mass host, confirmed the configuration is stable with no orbit crossings or ejections, then ran 2,574 four-body integrations to ask where an undetected Neptune-mass planet could still be hiding. Three coherent safe havens emerged, the widest and most forgiving of them lying beyond 31 AU where 78.2% of tested orbits survived.
- The nominal three-planet configuration is dynamically stable across the full 100,000-year integration: no orbit crossings, no ejections, net fractional drift in semi-major axis of 6 times 10 to the -5 or less, and energy conserved to about 1.5 times 10 to the -14 with angular momentum to about 1.1 times 10 to the -15, so the result is round-off limited rather than an artefact of integration error.
- The giants sit near low-order mean-motion resonances without being locked in them. The b-to-c period ratio of 7.197 lies 2.81% wide of 7:1 and the d-to-b ratio of 4.159 lies 3.98% wide of 4:1, while the non-adjacent d-to-c ratio of 29.932 falls just 0.23% from a high-order 30:1 commensurability. Adjacent pairs are separated by 7.5 and 6.5 mutual Hill radii, well above the classical Hill-stability boundary near 3.46.
- Eccentricities librate within secure bounds — 0.14 to 0.31 for planet c, 0.00 to 0.20 for b, and 0.06 to 0.31 for d — while inclinations stay tightly clustered near 89 degrees, confirming that the near-coplanar, disk-embedded geometry is internally consistent and self-maintaining.
- Of the 2,574 four-body integrations testing a 17.07 Earth-mass fourth planet, 896 or 34.8% survived the full 10,000 years. The rest were dominated by orbit-crossing and close-encounter outcomes at 1,652 cells, with only 23 outright ejections, 3 star-grazing collisions, and no planet-planet collisions.
- Surviving orbits cluster into three coherent havens rather than scattering at random: an inner warm zone from 1.0 to 2.7 AU interior to planet c with 38.5% survival, a narrow island at 18 to 20 AU inside the b-to-d gap with 27.9% survival against just 8.3% across the broader 12 to 21 AU range, and a broad outer zone beyond 31 AU with 78.2% survival, inward of the debris-disk edge near 50 AU.
How it was done
Masses and orbital elements were taken from the recent dynamical-characterisation literature and cross-checked against the NASA Exoplanet Archive: 8.70 Jupiter masses at 2.700 AU for planet c, 11.73 at 10.070 AU for b, and 2.40 at 26.000 AU for d, all near 89 degrees inclination around a 1.789 solar-mass star of roughly 20 million years. The nominal configuration was integrated for 100,000 years, about 1,010 orbits of the outermost planet, with the fifteenth-order adaptive Gauss-Radau integrator IAS15. For the hidden-planet search a Neptune-mass test planet was inserted into the fully interacting system and the four-body problem integrated for 10,000 years per cell with the hybrid symplectic MERCURIUS scheme, across 99 semi-major axis values from 1 to 50 AU crossed with 26 eccentricities from 0 to 0.5. Runs halted on ejection beyond 200 AU, on pericentre inside 0.05 AU, or on a close encounter within one Hill radius of a giant, while purely geometric orbit crossings were recorded without halting so that phase-protected orbits were not rejected.
Data sources
- NASA Exoplanet Archive (Akeson et al., PASP 125:989, 2013) — published orbital and physical parameters for all three planets
- Brandt et al., Astronomical Journal 161:179 (2021) — precise dynamical masses and orbital fits for Beta Pic b and c
- Lacour et al., Astronomy & Astrophysics 654:L2 (2021) — mass of Beta Pic c from b's orbital motion
- Nowak et al., Astronomy & Astrophysics 642:L2 (2020) — VLTI/GRAVITY direct confirmation of Beta Pic c
- Lacquement et al., Astronomy & Astrophysics 694:A236 (2025) — dynamics of the Beta Pictoris system and the possibility of an additional planet
Limitations
The eccentricity of Beta Pic d is the least constrained input and, because that planet defines the inner boundary of the widest haven, it directly controls where the outer safe zone begins and how wide the fragile b-to-d island is, so those survival fractions are model-dependent. The integration spans robustly capture fast Hill-sphere and resonance-overlap instabilities but not slow secular or diffusive ones acting over millions to billions of years, and the test planet was inserted at fixed orbital angles and a single inclination rather than marginalised over phase and mutual inclination.
How this research was produced
K-Dense Web planned and ran this astrophysics investigation end to end — gathering the sources, carrying out the analysis, producing the figures, and drafting the report. The full session transcript, including every intermediate step, is available to view.
