Hurst exponent: an indicator of weak chaos
Carregando...
Data
Autores
Orientador
Coorientador
Pós-graduação
Física - IGCE
Curso de graduação
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (UNESP)
Tipo
Tese de doutorado
Direito de acesso
Acesso aberto

Resumo
The coexistence of regular and chaotic regions in Hamiltonian phase space gives rise to weak chaos, a dynamical regime characterized by slow decay of correlations and anomalous transport. A common mechanism underlying weak chaos in such systems is stickiness, whereby chaotic orbits become intermittently trapped near stability islands, behaving almost regularly for long but finite times before escaping. Traditional chaos indicators, such as the Lyapunov exponent, converge slowly in the presence of stickiness and require knowledge of the equations of motion. In this thesis, we propose the Hurst exponent as an alternative indicator to characterize weak chaos. Originally developed in hydrology to quantify long-term dependence in time series, the Hurst exponent is computed solely from the orbit’s data, requiring no underlying dynamical equations. We apply this method to three paradigmatic area-preserving maps: the standard map, the bouncer model, and the kicked Harper map. We show that the Hurst exponent clearly distinguishes regular, weakly chaotic, and strongly chaotic motion. Moreover, a finite-time analysis reveals multimodal distributions of the finite-time Hurst exponent, where each peak corresponds to a distinct hierarchical level of the islands-around-islands structure, a feature not captured by finite-time Lyapunov exponents. Weextend the analysis to fractional dynamical systems, specifically the Riemann-Liouville fractional standard map, which exhibits cascade of bifurcations type trajectories (CBTT), a distinctive behavior in which a single trajectory undergoes a sequence of bifurcations as it evolves in time, alternating between different dynamical regimes. The Hurst exponent successfully identifies transitions between strongly chaotic, weakly chaotic, and periodic-like regimes along a single trajectory, again producing multimodal distributions. The results establish the Hurst exponent as a practical, robust, and computationally efficient tool for detecting weak chaos and dynamical traps.
Descrição
Idioma
Inglês



