2411.00175
Thick Arnold tongues
Mark Levi, Alexey Okunev
correctmedium confidence
- Category
- math.DS
- Journal tier
- Strong Field
- Processed
- Sep 28, 2025, 12:56 AM
- arXiv Links
- Abstract ↗PDF ↗
Audit review
The paper constructs, for the inertial particle system x¨=-(1/ε)(x˙-v(x)) with v=J∇H and H(x,y)=cos x cos y - a x + b y, a global attracting 2D invariant manifold and a Poincaré section on the torus, reducing the dynamics to a monotone one-parameter family of circle maps with two flat spots. It proves: (i) existence of a common drift slope for almost all initial data; (ii) Cantor-like dependence of the drift slope on α=a/b with plateaus at rationals; (iii) zero Hausdorff dimension of the irrational set; and (iv) exponential “steepness to all orders” at plateau endpoints. Key ingredients include an explicit first-order reduction ẋ=v+εf with f=−(Dv)v+O(ε), a contraction estimate for the forward Poincaré map via negative time-averaged divergence, and a monotonicity/expansivity framework establishing Theorem 1.3 for circle maps with flat spots. These address precisely the obstacles the model deemed unresolved. See the paper’s Proposition 3.1 (Fenichel reduction) and explicit f (and div f) formulae, Proposition 3.2 (phase portrait, section, and monotone family Qs), the contraction estimate for P, the rotation-number-to-drift link, and Theorem 1.1 enumerating items (1)–(4) with their proofs via Theorem 1.3.
Referee report (LaTeX)
\textbf{Recommendation:} minor revisions \textbf{Journal Tier:} strong field \textbf{Justification:} This work rigorously establishes a devil’s staircase of drift directions for inertial particles in a cellular flow under weak forcing, introducing thick Arnold tongues with zero-dimensional irrational complement and exponential endpoint steepness. The reduction from the inertial ODE to a monotone family of circle maps with flat spots is executed cleanly, with careful control of the Poincaré map via a negative time-averaged divergence estimate and a precise parameter monotonicity argument. The paper fills a gap explicitly noted as challenging by others and provides proofs that appear correct and complete. Minor clarifications regarding the exact description of exceptional initial data and parameter dependencies would strengthen presentation.