Questions: N-Body Planetary Dynamics and Orbital Integration

Question 1

Two N-body simulations of the same planetary system are run with all parameters identical except that one planet's initial position differs by one millimeter. After 100 million simulated years, the two simulations predict completely different orbital states for several planets. This outcome most accurately indicates:

• A. A software bug in the numerical integrator that accumulates errors over time
• B. Insufficient time resolution — the integrator's time step was too large
• C. The fundamental chaotic nature of gravitational N-body dynamics, where tiny initial differences amplify exponentially
• D. An unrealistically large perturbation — one millimeter exceeds measurement uncertainty for real planets

Answer: (C)

Chaotic dynamics — not numerical error — is the correct interpretation. Chaos means that the system's sensitivity to initial conditions is intrinsic to the differential equations governing it, not a property of the numerical method. Even a perfect integrator (if one existed) would show divergence, because the underlying gravitational N-body system has a positive Lyapunov exponent: small differences in initial conditions grow exponentially with time. Options 0 and 1 attribute the divergence to numerical artifacts, which misunderstands the physics.

Question 2

Why do planetary dynamicists run hundreds of N-body simulations with slightly varied initial conditions rather than one very long, high-precision simulation?

• A. To average out random numerical errors that accumulate differently in each run
• B. Because individual long-term trajectories are unreliable due to chaos — ensemble statistics give meaningful probabilistic answers where single trajectories cannot
• C. To test whether different integrators (symplectic vs. Runge-Kutta) agree over long timescales
• D. Because computational resources are insufficient for a single simulation long enough to cover billion-year timescales

Answer: (B)

The motivation is epistemic, not computational. Because the N-body system is chaotic, any individual simulation's exact trajectory becomes physically meaningless after the chaos timescale (tens of millions of years for the inner solar system). What remains meaningful is the statistics of outcomes across an ensemble: the probability that Mercury's eccentricity exceeds a threshold, the fraction of simulations showing a planet ejection within 5 Gyr. Ensemble simulations shift the question from 'where will this planet be?' to 'what is the probability distribution of outcomes?' — the only type of question the dynamics can actually answer reliably.

Question 3

Symplectic integrators solve the N-body problem exactly and eliminate accumulated numerical error over long integrations.

• T. True
• F. False

Answer: False

Symplectic integrators do not eliminate error — they preserve the geometric structure of Hamiltonian mechanics (specifically, the symplectic structure of phase space). This causes them to conserve energy and angular momentum far better than generic methods like Runge-Kutta over very long integrations, making them the preferred tool for billion-year planetary simulations. But they do not produce exact solutions. Moreover, the underlying chaotic divergence persists regardless of integrator quality — symplectic or not, two trajectories with slightly different initial conditions will eventually diverge exponentially.

Question 4

The solar system is substantially stable over its remaining lifetime — no planet is at risk of orbital instability before the Sun becomes a red giant.

• T. True
• F. False

Answer: False

N-body simulations of the solar system show that it is not perfectly stable over gigayear timescales. There is approximately a 1% probability that Mercury's orbit becomes chaotically unstable before the Sun exhausts its hydrogen fuel — potentially colliding with Venus or the Sun, or being ejected from the solar system. Jupiter and Saturn's near 5:2 resonance (the Great Inequality) drives slow oscillations in inner planet orbits that can, in rare simulation runs, push Mercury into crossing orbits. This is a genuine dynamical result that cannot be obtained from two-body or perturbation theory.

Question 5

Why does adding a third body to a two-body gravitational system fundamentally change what kinds of predictions are possible, and what technique do planetary scientists use to compensate?

Model answer: The two-body problem has an exact closed-form solution (Kepler's ellipses), enabling precise, arbitrarily long-range predictions. Adding a third body eliminates this — the three-body problem has no general closed-form solution, and the dynamics become chaotic: initially nearby trajectories diverge exponentially, making precise long-term prediction of individual trajectories impossible. Planetary scientists compensate with ensemble simulations: running many integrations with slightly varied initial conditions and interpreting results statistically. Instead of predicting where a planet will be, they predict the probability distribution of outcomes — e.g., the probability Mercury's eccentricity exceeds 0.6 within 5 billion years.

The shift from exact to statistical prediction is not a temporary limitation awaiting a better algorithm — it is a fundamental consequence of the chaos that emerges in gravitational systems with three or more bodies. The Lyapunov timescale (after which trajectories diverge significantly) for the inner solar system is roughly 5 million years, far shorter than the timescales of interest. Ensemble statistics survive where individual trajectories do not.