DDA 2015 – The Evidence for Slow Migration of Neptune from the Inclination Distribution of Kuiper Belt Objects

This is one of a series of notes taken during the 2015 meeting of the AAS Division on Dynamical Astronomy, 3-7 May, at CalTech. An index to this series (all the papers presented at the meeting) is here.

David Nesvorny (SWRI)


Much of the dynamical structure of the Kuiper Belt can be explained if Neptune migrated over several AU, and/or if Neptune was scattered to an eccentric orbit during planetary instability. An outstanding problem with the existing formation models is that the distribution of orbital inclinations predicted by them is narrower than the one inferred from observations. Here we perform numerical simulations of the Kuiper belt formation starting from an initial state with Neptune at $20\lt a^{N,0} \lt 30$ AU and a dynamically cold outer disk extending from beyond $a^{N,0}$ to 30 AU. Neptune’s orbit is migrated into the disk on an e-folding timescale $1 \le \tau \le 100$ Myr. A small fraction ($\sim10^{-3}$) of disk planetesimals become implanted into the Kuiper belt in the simulations. By analyzing the orbital distribution of the implanted bodies in different cases we find that the inclination constraint implies that $\tau \ge 10$ Myr and $a^{N,0} \le 26$ AU.The models with $\tau \lt 10$ Myr do not satisfy the inclination constraint, because there is not enough time for various dynamical processes to raise inclinations. The slow migration of Neptune is consistent with other Kuiper belt constraints, and with the recently developed models of planetary instability/migration. Neptune’s eccentricity and inclination are never large in these models ($e^N \lt 0.1$, $i^N \lt 2$ deg), as required to avoid excessive orbital excitation in the $\gt 40$ AU region, where the Cold Classicals presumably formed.


  • Early SS evolution
    • giant planets emerged from dispersing protopl disk on compact orbits (inside massive belt)
    • planetesimal driven migration?
    • dynamical instability?
    • giant planets now spread from 5 to 30 AU
  • Kuiper Belt is the best clue to evolution of Neptune’s orbit
    • KB structure is complex (plot: $e$ vs $a$)
    • between 3:2 and 2:1 MMRs: a mess, but hot and cold populations
    • where did hot population come from (including high-$i$ 3:2 objects)?
      • model: too many Plutinos compared to observations
  • New model
    • 4 outer planets
    • ICs:
      • Neptune starting points: 22, 24, 26, 28 AU
      • Neptune migration e-folding timescales 1, 3, 10, 30, 100 Myr
    • 1e6 particles, Rayleigh initial distribution
    • swift_rmvs3 integrator
      • 500 cores of Pleiades supercomputer
    • 20 jobs total, most stopped 1 Gyr, interesting ones to 4 Gyr
    • $\rightarrow$ result matches observed distribution
      • 24 AU, 30 Myr
    • but too manyPlutinos(?)
      • observational bias?
        • cf Petit et al. 2012
      • CFEPS detection simulator
        • agreement (of hot population) is actually pretty good
    • Gomes capture mechanism:Gomes 2003
      • 2:1 MMR secular structure is complex
  • Conclusions:
    • Neptune migrated into a massive cometary disk at $\lt 30$ AU
    • Neptune’s migration hadto be slow
      • need time to increase inclinations
    • Model also explains other KB properties
    • Initial disk had to be $\sim 20 M_\oplus$

Leave a Reply