The Berry–Balazs transition of the semiclassical Wigner function: rigorous transition time and post-transition morphology for whorl-forming (stable) classical motion

Berry and Balazs (J. Phys. A 12, 625 (1979)) showed that the Wigner function $W(q,p,t)$ of an evolving semiclassical state hugs the classically transported phase-space curve $C_t$, which develops 'whorls' under stable motion and 'tendrils' under unstable motion, and conjectured a transition time $t_c$ beyond which $W$ can no longer follow $C_t$, with $t_c = O(\hbar^{-2/3})$ for whorls and $t_c = O(\ln \hbar^{-1})$ for tendrils. The logarithmic scaling for unstable motion and the fates of the quantum-maps program and of the quantum-pendulum numerics that motivated the paper have since been settled. What remains open is the stable (whorl) branch and the detailed post-transition structure: no rigorous derivation of $t_c$ for whorl-forming motion exists, the specific $\hbar^{-2/3}$ conjecture has been neither verified nor refuted (competing analyses of localized packets give $\hbar^{-1/2}$-type Ehrenfest times), and the paper's quantitative predictions for the post-transition Wigner function — divergences near invariant curves tangent to the initial curve, step discontinuities in the locally averaged probability density, the oscillation spectrum, and the proliferating cusp catastrophes that herald the transition — have never been systematically confirmed. The problem is to derive, and verify numerically or by exactly solvable models, the transition time and the full post-transition morphology of the Wigner function for stable classical motion, including the uniform description of $W$ on the catastrophe set where many stationary points coalesce.

Background

The Wigner function $W(q,p,t)$ is a phase-space representation of a quantum state: it is the Weyl transform of the density matrix (the phase-space image of the state obtained by a standard linear change of variables), is real (but generally not positive), and reproduces the correct quantum marginal distributions in $q$ and $p$. Under classical dynamics a phase-space point moves along a Hamiltonian trajectory; Berry and Balazs considered states whose initial form corresponds to a curve $C_0$ in the two-dimensional phase space $(q,p)$ of a system with one degree of freedom — for example a Gaussian wave packet, whose Wigner function concentrates on a small blob straddling $C_0$. Under the classical flow $C_0$ is transported into a curve $C_t$, and their central result is a semiclassical formula, built from the time-dependent WKB (Van Vleck) propagator — the leading-order semiclassical approximation to the quantum evolution kernel — showing that the region where $W$ is large hugs $C_t$, with quantum oscillations whose local wavelength is set by the geometry of $C_t$ and its neighbours, and by the Airy scale $\hbar^{2/3}$ near caustics (turning points, where $C_t$ is perpendicular to the $q$ axis; the $\hbar^{2/3}$ width is the fringe scale of the Airy-function diffraction pattern at such a fold). As $t$ grows, $C_t$ becomes convoluted. Berry and Balazs classified the convolutions into 'whorls' — ever tighter windings around an elliptic (stable) fixed point, whose length grows linearly in time — and 'tendrils' — infinitely stretched and folded filaments produced by unstable (chaotic) motion, whose length grows exponentially. Because a quantum phase-space distribution cannot resolve structures on areas smaller than $O(\hbar)$, $W$ cannot follow the arbitrarily fine classical detail, and at a 'transition time' $t_c$ the semiclassical picture must fail: this is the physical origin of what the later literature calls the Ehrenfest time, the time up to which quantum and classical evolution track each other. Later work settled large parts of the original picture. The logarithmic Ehrenfest time $t_E \sim \lambda^{-1}\ln(1/\hbar)$ for chaotic systems ($\lambda$ the Lyapunov exponent, the rate of exponential separation of nearby classical trajectories) is now standard, and Silvestrov and Beenakker even used the Berry–Balazs WKB itself to show that the approximation survives until $(7/6)\lambda^{-1}\ln(A/\hbar)$ in one dimension; rigorously, Schubert proved that expectation values in Lagrangian states follow their classical limits up to the Ehrenfest time for Anosov (uniformly hyperbolic) systems. The quantum-maps program announced in the paper was carried out by Berry, Balazs, Tabor and Voros. The quantum-pendulum (kicked-rotor) computations of Casati, Chirikov, Izraelev and Ford that motivated the speculation that $W$ spreads more slowly than $C_t$ were later explained by dynamical localization — the Anderson localization (localization of waves in a disordered medium) of the rotor's quasi-energy eigenstates in momentum space (Fishman, Grempel and Prange) — a mechanism different from, but sufficient for, that puzzle. For stable motion, however, the situation is not settled: Zhao and Wu find an integrable Ehrenfest time proportional to $\hbar^{-1/2}$ for localized packets (with quantum revivals at $\sim\hbar^{-1}$), Michel and Zell find stable many-body models whose quantum break time scales like a power of $1/\hbar$ or never 'breaks', and Schubert's rigorous integrable result only states a time range 'much larger' than the chaotic one. None of these addresses the Berry–Balazs curve-based estimate $t_c = O(\hbar^{-2/3})$, which follows from the Airy scale $\hbar^{2/3}$ and the shrinking inter-loop distance $D(t) \sim 1/t$ of a whorl. In parallel, the fine-scale phase-space oscillations predicted for the post-transition state — structure on scales below $\hbar$ — became the subject of sub-Planck structure and decoherence (Zurek), and the general late-time fate of isolated quantum states became the equilibration/eigenstate-thermalization program (Rigol, Dunjko and Olshanii), which conjectures that single eigenstates of generic isolated systems already encode thermal averages; but the specific morphological predictions of Berry and Balazs — equations (36)–(38) of their paper — were not tested. A 'cusp catastrophe' here is the elementary-catastrophe event at which two turning points of $\psi$ are born as an inflection of $C_t$ rotates (locally described by the Pearcey function, a canonical diffraction integral for the cusp), and the 'catastrophe set' $\mathscr{L}$ is the locus of phase-space points on which the stationary-phase points governing $W$ coalesce.

Problem Statement

For a system with one degree of freedom whose classical phase-space motion transports an initial curve $C_0$ into the convoluted curve $C_t$ (whorls around stable fixed points, tendrils under unstable motion), determine rigorously — and verify, by exactly solvable models or controlled numerics — the transition beyond which the Wigner function $W(q,p,t)$ of the evolving semiclassical state can no longer follow the details of $C_t$, and characterize the state after that transition. Specifically:

1. Derive the transition time $t_c$ for whorl-forming (stable/integrable) motion and settle the Berry–Balazs conjecture $t_c = O(\hbar^{-2/3})$, which followed from the Airy scale $\hbar^{2/3}$ and the inter-loop distance $D(t) \sim 1/t$ of a developing whorl; in particular decide how this exponent relates to the $\hbar^{-1/2}$-type Ehrenfest times found for localized packets and to the 'much larger' rigorous integrable time ranges, clarifying for which initial states (extended curves vs localized blobs) and observables each scaling holds. Because the transition admits more than one quantitative criterion — the $O(\hbar)$ resolution-area scale and the Airy scale $\hbar^{2/3}$ — $t_c$ for whorl-forming motion is criterion-dependent.
2. Determine quantitatively the post-transition Wigner function and wavefunction for a fully developed whorl: the local average of $|\psi|^2$ (including the predicted step discontinuities near actions where invariant curves are tangent to $C_0$), the spectrum of quantum oscillations, the classical-limit form of $W$ with its $|I-I_c|^{-1/2}$ singular structure ($I$ the classical action, $I_c$ the action of the invariant curve tangent to $C_0$), and how these are smoothed by finite $\hbar$.
3. Verify the conjectured precursors of the transition — the proliferation of cusp catastrophes of the wavefunction $\psi$ (pairs of turning points born at inflections of $C_t$, locally described by the Pearcey function, the canonical diffraction integral for the cusp) — and construct the uniform asymptotic description of $W$ on the catastrophe set for highly convoluted curves where many stationary points coalesce, going beyond the two-stationary-point (simple stationary-state) case treated in the source paper. The chaotic (tendril) logarithmic scaling and the special cases resolved elsewhere (quantum maps; the kicked-rotor energy saturation via dynamical localization) are context, not targets; the requested results above jointly support the single objective of completing the Berry–Balazs theory of the quantum–classical transition for the Wigner function.

References

1. M. V. Berry and N. L. Balazs, 'Evolution of semiclassical quantum states in phase space', J. Phys. A: Math. Gen. 12, 625-642 (1979), DOI 10.1088/0305-4470/12/5/012, https://doi.org/10.1088/0305-4470/12/5/012
2. M. V. Berry, N. L. Balazs, M. Tabor and A. Voros, 'Quantum maps', Annals of Physics 122, 26-63 (1979), DOI 10.1016/0003-4916(79)90296-3, https://doi.org/10.1016/0003-4916(79)90296-3
3. G. Casati, B. V. Chirikov, F. M. Izraelev and J. Ford, 'Stochastic behavior of a quantum pendulum under a periodic perturbation', in Stochastic Behavior in Classical and Quantum Hamiltonian Systems, Lecture Notes in Physics 93, 334-352 (1979), DOI 10.1007/BFb0021757, https://doi.org/10.1007/BFb0021757
4. S. Fishman, D. R. Grempel and R. E. Prange, 'Chaos, Quantum Recurrences, and Anderson Localization', Physical Review Letters 49, 509-512 (1982), DOI 10.1103/PhysRevLett.49.509, https://doi.org/10.1103/PhysRevLett.49.509
5. P. W. O'Connor, S. Tomsovic and E. J. Heller, 'Semiclassical dynamics in the strongly chaotic regime: breaking the log time barrier', Physica D 55, 340-357 (1992), DOI 10.1016/0167-2789(92)90064-T, https://doi.org/10.1016/0167-2789(92)90064-T
6. S. Tomsovic and E. J. Heller, 'Long-time semiclassical dynamics of chaos: The stadium billiard', Physical Review E 47, 282-299 (1993), DOI 10.1103/PhysRevE.47.282, https://doi.org/10.1103/PhysRevE.47.282
7. W. H. Zurek, 'Sub-Planck structure in phase space and its relevance for quantum decoherence', Nature 412, 712-717 (2001), DOI 10.1038/35089017, https://doi.org/10.1038/35089017 (arXiv version, titled 'Sub-Planck spots of Schroedinger cats and quantum decoherence', at arXiv:quant-ph/0201118, https://arxiv.org/abs/quant-ph/0201118)
8. P. G. Silvestrov and C. W. J. Beenakker, 'Ehrenfest times for classically chaotic systems', Physical Review E 65, 035208 (2002), DOI 10.1103/PhysRevE.65.035208, https://doi.org/10.1103/PhysRevE.65.035208 (arXiv:nlin/0111042, https://arxiv.org/abs/nlin/0111042)
9. P. P. de M. Rios and A. M. Ozorio de Almeida, 'On the propagation of semiclassical Wigner functions', Journal of Physics A: Mathematical and General 35, 2609-2617 (2002), DOI 10.1088/0305-4470/35/11/307, https://doi.org/10.1088/0305-4470/35/11/307 (arXiv:math-ph/0111012, https://arxiv.org/abs/math-ph/0111012)
10. R. Schubert, 'Semiclassical behaviour of expectation values in time evolved Lagrangian states for large times', Communications in Mathematical Physics 256, 239-254 (2005), DOI 10.1007/s00220-005-1319-4, https://doi.org/10.1007/s00220-005-1319-4 (arXiv:math/0402038, https://arxiv.org/abs/math/0402038)
11. M. Rigol, V. Dunjko and M. Olshanii, 'Thermalization and its mechanism for generic isolated quantum systems', Nature 452, 854-858 (2008), DOI 10.1038/nature06838, https://doi.org/10.1038/nature06838
12. Y. Zhao and B. Wu, 'Quantum-classical correspondence in integrable systems', Science China Physics, Mechanics & Astronomy 62, 997011 (2019), DOI 10.1007/s11433-018-9348-2, https://doi.org/10.1007/s11433-018-9348-2 (arXiv:1801.06389, https://arxiv.org/abs/1801.06389)
13. C. Zhao and B. Wu, 'Ehrenfest Time at the Transition from Integrable Motion to Chaotic Motion', Chinese Physics Letters 38, 030502 (2021), DOI 10.1088/0256-307X/38/3/030502, https://doi.org/10.1088/0256-307X/38/3/030502
14. X. Han and B. Wu, 'Ehrenfest breakdown of the mean-field dynamics of Bose gases', Physical Review A 93, 023621 (2016), DOI 10.1103/PhysRevA.93.023621, https://doi.org/10.1103/PhysRevA.93.023621 (arXiv:1506.04020, https://arxiv.org/abs/1506.04020)
15. M. Michel and S. Zell, 'The Timescales of Quantum Breaking', Fortschritte der Physik 71, 2300163 (2023), DOI 10.1002/prop.202300163, https://doi.org/10.1002/prop.202300163 (arXiv:2306.09410, https://arxiv.org/abs/2306.09410)
16. M. Ćosić, S. Petrović and S. Bellucci, 'On the phase-space catastrophes in dynamics of the quantum particle in an optical lattice potential', Chaos 30, 103107 (2020), DOI 10.1063/1.5140528, https://doi.org/10.1063/1.5140528
17. A. M. Ozorio de Almeida, 'Vestiges of quantum oscillations in the open evolution of semiclassical states', Journal of Physics A: Mathematical and Theoretical 50, 245302 (2017), DOI 10.1088/1751-8121/aa6b8f, https://doi.org/10.1088/1751-8121/aa6b8f (arXiv:1703.08533, https://arxiv.org/abs/1703.08533)
18. G. M. Lando, O. Giraud and D. Ullmo, 'Computing Quantum Mean Values in the Deep Chaotic Regime', Physical Review Letters 132, 260401 (2024), DOI 10.1103/PhysRevLett.132.260401, https://doi.org/10.1103/PhysRevLett.132.260401