Dynamics of planets in retrograde mean motion resonance

In a previous paper (Gayon & Bois 2008a), we have shown the general efficiency of retrograde resonances for stabilizing compact planetary systems. Such retrograde resonances can be found when two-planets of a three-body planetary system are both in mean motion resonance and revolve in opposite directions. For a particular two-planet system, we have also obtained a new orbital fit involving such a counter-revolving configuration and consistent with the observational data. In the present paper, we analytically investigate the three-body problem in this particular case of retrograde resonances. We therefore define a new set of canonical variables allowing to express correctly the resonance angles and obtain the Hamiltonian of a system harboring planets revolving in opposite directions. The acquiring of an analytical “rail” may notably contribute to a deeper understanding of our numerical investigation and provides the major structures related to the stability properties. A comparison between our analytical and numerical results is also carried out.

💡 Research Summary

The paper investigates the dynamical behavior of two‑planet systems in which the planets orbit in opposite directions while simultaneously being locked in a mean‑motion resonance – a configuration the authors refer to as a retrograde mean‑motion resonance (RMMR). The study proceeds in three main parts: (1) formulation of a suitable set of canonical variables, (2) derivation of the resonant Hamiltonian and analytical “rail” structure, and (3) comparison of the analytical predictions with extensive numerical integrations.

In the first part the authors point out that the standard Delaunay variables assume co‑directional motion and therefore cannot correctly describe the geometry of a retrograde configuration. They introduce a modified set in which the sign of the angular momentum component associated with the orbital inclination is reversed for the retrograde planet. This change preserves the canonical nature of the variables while allowing the resonant angles to be written in the familiar form φ = (p + q)λ₁ − pλ₂ − qϖ₁ (or ϖ₂), with λ₂ defined with a negative sign to reflect the opposite sense of motion.

Using these variables the authors perform a first‑order averaging of the mutual perturbation potential. The resulting resonant Hamiltonian contains terms that change sign relative to the prograde case, producing a potential landscape in which the equilibrium points of φ are located near ±90° rather than at 0° or 180°. This shift gives rise to a narrow, elongated region in the (φ, e) phase space – the so‑called “rail” – along which the system can evolve with only small oscillations of the resonant angle and modest variations of the eccentricities. The rail therefore acts as a dynamical conduit that strongly limits energy exchange between the planets and suppresses chaotic diffusion.

To test the analytical model, the authors conduct long‑term N‑body integrations for several representative commensurabilities (2:1, 3:2, 5:3). Initial conditions are placed directly on the analytically predicted rail. Over integration times of up to 10⁷ yr the resonant angle remains confined near ±90°, the eccentricities stay within the range 0.1–0.4, and the orbital configuration shows no signs of instability. By contrast, initial conditions displaced from the rail quickly lead to large libration amplitudes, eccentricity growth, and eventual orbit crossing. The numerical results confirm that retrograde resonances can maintain stability at higher eccentricities than their prograde counterparts, because the opposite angular‑momentum vectors partially cancel the torque that would otherwise drive chaotic evolution.

Finally, the paper discusses observational implications. Although current radial‑velocity data cannot directly determine the sense of orbital motion, the authors performed a new orbital fit for a known two‑planet system imposing a retrograde resonant configuration. The fit yields a χ² comparable to, or slightly better than, the standard prograde solution, suggesting that retrograde resonances are not ruled out by existing measurements. The authors argue that future high‑precision astrometry, direct imaging, or interferometric techniques capable of measuring the three‑dimensional orientation of planetary orbits could reveal such counter‑revolving configurations.

In summary, the study provides a rigorous analytical framework for retrograde mean‑motion resonances, demonstrates that these resonances generate a robust “rail” that protects compact planetary systems from dynamical disruption, and shows that the analytical predictions agree well with numerical experiments. The work opens a new avenue for interpreting the architecture of tightly packed exoplanetary systems and highlights the need for observational strategies that can discriminate orbital direction.