The orbits of the inner planets are chaotic, and over billions of years chaotic diffusion can drive Mercury's perihelion precession frequency g1g_1 into the destabilizing secular resonance g1−g5g_1-g_5, eventually producing close encounters or collisions. Recent realistic secular models show that the probability distribution of the time of the first such destabilization is well reproduced by a one-dimensional Wiener process with a reflecting barrier, but only with fitted parameters, and later work shows that a purely diffusive model badly underpredicts the rare instabilities within the remaining 5 Gyr of the Solar System's lifetime. The open problem is to derive this effective stochastic dynamics from first principles: starting from the planetary Hamiltonian, identify the slow stochastic variable(s), derive the transport coefficients (diffusion or subdiffusion parameters) from the resonant structure of the secular Hamiltonian rather than by fitting, account for the dynamical constraints that make the destabilization so slow, and thereby explain the observed instability statistics and reconcile them with the much shorter timescales of simplified Mercury models.

Background

When the planetary equations of motion are averaged over the fast orbital (mean-longitude) motions, one obtains the secular dynamics: the slow precession of the perihelia and nodes of the planetary orbits. Linearized about nearly circular, coplanar orbits, this dynamics is quasi-periodic, decomposable into fundamental precession frequencies g1,…,g8g_1,\dots,g_8 (perihelia) and s1,…,s8s_1,\dots,s_8 (nodes); the first four of each are dominated by Mercury, Venus, Earth, and Mars. A secular resonance is a near-commensurability among integer combinations of these frequencies; its small denominator invalidates the quasi-periodic representation and can generate chaos. That the inner planets' orbits are chaotic was demonstrated numerically in 1989-1992 (Laskar 1989, Nature 338, 237; Sussman & Wisdom 1992, Science 257, 56): the maximum Lyapunov exponent (the asymptotic rate of exponential divergence of nearby orbits) corresponds to a Lyapunov time of only about 5 million years, yet the orbits remain statistically stable over billions of years. The main destabilization channel is the resonance g1−g5g_1-g_5 between Mercury's perihelion precession frequency g1g_1 and Jupiter's nearly constant frequency g5g_5: chaotic diffusion of g1g_1 down to g5g_5 pumps Mercury's eccentricity until its orbit crosses that of Venus; direct numerical ensembles give roughly a 1% probability of such an event within 5 Gyr (Laskar & Gastineau 2009, Nature 459, 817).

Which secular resonances actually generate the chaos was disputed for three decades: Laskar's analytical studies proposed the resonances 2(g3−g4)−(s3−s4)2(g_3-g_4)-(s_3-s_4) and (g3−g4)−(s3−s4)(g_3-g_4)-(s_3-s_4) between Earth- and Mars-dominated modes, but no analytical corroboration existed. Mogavero & Laskar (2021, A&A 655, A1) built a forced secular model of the inner planets (the outer planets' regular secular motion imposed as a quasi-periodic forcing, first-order averaging in the inner masses, computer-algebra expansion in eccentricities and inclinations) that reproduces the maximum Lyapunov exponent and the statistics of Mercury's high eccentricities. An ensemble of one thousand 100-Gyr integrations showed that the probability density function (PDF) of the time of the first secular collision (geometric intersection of two planetary orbits) is reasonably reproduced by a one-dimensional Wiener process — idealized Brownian motion — for g1g_1, with a reflecting barrier at its observed upper edge (a boundary the process cannot cross) and an absorbing barrier at g1=g5g_1=g_5 (where the process stops and destabilization is deemed to occur); the hitting time (first-passage time) PDF of this process matches the empirical shape, with fitted parameters T1=27.6T_1=27.6 Gyr and α=0.9\alpha=0.9. The same paper showed that diffusion estimates based only on the leading resonances give a characteristic time about fifty times too short, implying unknown dynamical constraints slow the diffusion. A companion statistical study (Hoang et al. 2021, A&A 654, A156) established the marginal PDFs of the fundamental frequencies, including the asymmetric, bounded-above distribution of g1g_1 that motivates the reflecting barrier.

Three later developments define the present frontier. First, Mogavero & Laskar (2022, A&A 662, L3) implemented canonical perturbation theory - the systematic averaging of Hamiltonian dynamics through near-identity changes of coordinates - in the computer-algebra system TRIP, computed the half-widths of all secular resonances at work along the forced dynamics, and resolved the old attribution question: (g3−g4)−(s3−s4)(g_3-g_4)-(s_3-s_4) is the widest resonance, 2(g3−g4)−(s3−s4)2(g_3-g_4)-(s_3-s_4) also leads, the chaotic zone they span is confirmed by resonance overlap (the regime in which adjacent resonances are wide enough to overlap and generate widespread chaos), and a second chaotic source lies in the g1−g5g_1-g_5, g2−g5g_2-g_5, s1−s2s_1-s_2 subspace. Second, Hoang, Mogavero & Laskar (2022, MNRAS) showed that the Hamiltonian truncated at degree 4 in eccentricities and inclinations is practically stable over 5 Gyr and that destabilization is driven by degree-6 terms — an analogy to the Fermi-Pasta-Ulam-Tsingou problem of unexpectedly slow thermalization. Third, Mogavero, Hoang & Laskar (2023, Phys. Rev. X 13, 021018) computed the full Lyapunov spectrum (the complete set of exponential divergence rates of the dynamics), uncovered a hierarchy of timescales, and identified three symmetries of the leading resonances whose breaking is weak, implying quasi-integrals of motion — dynamical quantities approximately conserved by the leading resonances — with strong evidence that they constrain the chaotic diffusion and explain the metastability. On the stochastic side, Woillez & Bouchet (2020, Phys. Rev. Lett. 125, 021101) had derived the hitting-time statistics, including instanton (rare-event) asymptotics, only for a simplified Mercury model with other inner-planet frequencies frozen; that model destabilizes about ten times faster than realistic ones. Abbot and collaborators improved the empirical statistics with rare-event sampling (2021) and large validated N-body ensembles (2023), and then showed (2024, ApJ 967, 121) that the diffusive model underpredicts Mercury's instability probability by a factor of 3-10,000 on timescales below 5 Gyr, proposing instead a phenomenological subdiffusive model (a process whose typical displacement grows more slowly than the square root of time) with parameters tuned to the N-body trajectories of g1g_1, and posing the open questions of why subdiffusion fits better than ordinary diffusion and why g1g_1 has an upper bound but no protective lower bound. What no work has done is derive the effective stochastic process — its identity, coefficients, and anomalous character — from the Hamiltonian itself.

Problem Statement

Starting from the planetary secular Hamiltonian (or the full N-body dynamics) of the inner Solar System, derive an effective first-principles stochastic description of the slow chaotic diffusion that destabilizes it. Concretely: (i) identify the dynamical variable or combination of variables (for example the perihelion precession frequency g1g_1, a coordinate involving the quasi-integrals of motion, or another combination) that actually performs the effective stochastic evolution towards the destabilizing resonance g1−g5g_1-g_5; (ii) derive the transport coefficients of the corresponding effective process — the diffusion coefficient, or the parameters of the subdiffusive/anomalous process that later work shows fits the short-time statistics better — from the resonant structure of the Hamiltonian (resonance half-widths, harmonic amplitudes, symmetry-breaking terms at degree six and beyond) rather than by fitting to numerical trajectories; (iii) account, within the same derivation, for the dynamical constraints (quasi-integrals associated with the resonance symmetries, the bounded-above distribution of g1g_1) that make the destabilization orders of magnitude slower than naive resonance-overlap estimates; and (iv) show that the resulting theory quantitatively reproduces the observed statistics of Mercury's destabilization times — the peak of the hitting-time PDF, the rare-event behavior at times below 5 Gyr, and the long-time decay — thereby also explaining why simplified models with frozen frequencies predict timescales about ten times shorter. The effective stochastic process, its coefficients, and its domain of validity must follow from the Hamiltonian rather than be fitted.

References

  1. J. Laskar (1989). "A numerical experiment on the chaotic behaviour of the Solar System." Nature 338, 237-238. DOI: 10.1038/338237a0. https://doi.org/10.1038/338237a0
  2. G. J. Sussman & J. Wisdom (1992). "Chaotic Evolution of the Solar System." Science 257 (5066), 56-62. DOI: 10.1126/science.257.5066.56. https://doi.org/10.1126/science.257.5066.56
  3. J. Laskar & M. Gastineau (2009). "Existence of collisional trajectories of Mercury, Mars and Venus with the Earth." Nature 459, 817-819. DOI: 10.1038/nature08096. https://doi.org/10.1038/nature08096
  4. E. Woillez & F. Bouchet (2020). "Instantons for the destabilization of the inner Solar System." Physical Review Letters 125, 021101. DOI: 10.1103/PhysRevLett.125.021101. arXiv:1910.04005. https://arxiv.org/abs/1910.04005
  5. F. Mogavero & J. Laskar (2021). "Long-term dynamics of the inner planets in the Solar System." Astronomy & Astrophysics 655, A1. DOI: 10.1051/0004-6361/202141007. arXiv:2105.14976. https://arxiv.org/abs/2105.14976
  6. N. H. Hoang, F. Mogavero & J. Laskar (2021). "Chaotic diffusion of the fundamental frequencies in the Solar System." Astronomy & Astrophysics 654, A156. DOI: 10.1051/0004-6361/202140989. https://doi.org/10.1051/0004-6361/202140989
  7. F. Mogavero & J. Laskar (2022). "The origin of chaos in the Solar System through computer algebra." Astronomy & Astrophysics 662, L3. DOI: 10.1051/0004-6361/202243327. arXiv:2205.03298. https://arxiv.org/abs/2205.03298
  8. N. H. Hoang, F. Mogavero & J. Laskar (2022). "Long-term instability of the inner Solar system: numerical experiments." Monthly Notices of the Royal Astronomical Society 514, 1342-1350. DOI: 10.1093/mnras/stac1299. https://doi.org/10.1093/mnras/stac1299
  9. F. Mogavero, N. H. Hoang & J. Laskar (2023). "Timescales of Chaos in the Inner Solar System: Lyapunov Spectrum and Quasi-integrals of Motion." Physical Review X 13, 021018. DOI: 10.1103/PhysRevX.13.021018. arXiv:2305.01683. https://arxiv.org/abs/2305.01683
  10. D. S. Abbot, R. J. Webber, S. Hadden, D. Seligman & J. Weare (2021). "Rare Event Sampling Improves Mercury Instability Statistics." The Astrophysical Journal 923, 236. DOI: 10.3847/1538-4357/ac2fa8. https://doi.org/10.3847/1538-4357/ac2fa8
  11. D. S. Abbot, D. M. Hernandez, S. Hadden, R. J. Webber, G. P. Afentakis & J. Weare (2023). "Simple Physics and Integrators Accurately Reproduce Mercury Instability Statistics." The Astrophysical Journal 944, 190. DOI: 10.3847/1538-4357/acb6ff. https://doi.org/10.3847/1538-4357/acb6ff
  12. D. S. Abbot, R. J. Webber, D. M. Hernandez, S. Hadden & J. Weare (2024). "Mercury's Chaotic Secular Evolution as a Subdiffusive Process." The Astrophysical Journal 967, 121. DOI: 10.3847/1538-4357/ad3e5f. arXiv:2306.11870. https://arxiv.org/abs/2306.11870