No abstract

No abstract

No abstract

No abstract

No abstract

We discuss the problem of Poincaré recurrences in area-preserving maps and the universality of their decay at long times. The work is related to to the results presented in Refs. [1,2] The first one states that the asymptotic decay P (τ ) ∝ 1/τ 3 , which is expected from the scaling theory of universal phase-space structure in the vicinity of critical golden curve (see Refs. 1-2, 12-18 in [2]), is not valid. Such a conclusion is based on where N tot = 10 12 . New data for 8 < log τ ≤ 9 indeed demonstrate noticeable deviations from the theoretical estimates for P (τ ) obtained from the exit times τ n from golden resonance scales r n (Fig. 1 in [2]). These deviations are rather intriguing. Indeed, the local properties of critical golden curve are known to be self-similar and universal (e.g. phase-space structure, size of resonances and local diffusion rate, see e.g. Refs. 2, 15 in [2]). Moreover, the exit times τ n , found in [2] for a first time, give an example of a non-local characteristic which follows the universal scaling law. The arguments based on this scaling lead to the asymptotic decay P (τ ) ∝ 1/τ 3 which however can start after extremely long times τ > τ a . In [2] we have shown that at least τ a > τ g ≈ 2×10 5 and it is not excluded that this time scale is still much longer [3], and is not yet reached even in the simulations presented in [1].If to assume this then there is an interesting possibility to see if P (τ ) would have some universal properties on the presently available (intermediate) time scales (τ ≤ 10 9 ). This is the second aspect of the quarry on which the authors of the Comment tend to give a negative answer. To this end we show in Fig.

No abstract

No abstract