Abstract/Summary

Unstable periodic orbits (UPOs), exact periodic solutions of the evolution equation, offer a very powerful framework for studying chaotic dynamical systems, as they allow one to dissect their dynamical structure. UPOs can be considered the skeleton of chaotic dynamics, its essential building blocks. In fact, it is possible to prove that in a chaotic system, UPOs are dense in the attractor, meaning that it is always possible to find a UPO arbitrarily near any chaotic trajectory. We can thus think of the chaotic trajectory as being approximated by different UPOs as it evolves in time, jumping from one UPO to another as a result of their instability. In this thesis we provide a contribution towards the use of UPOs as a tool to understand and distill the dynamical structure of chaotic dynamical systems. We will focus on two models, characterised by different properties, the Lorenz-63 and Lorenz-96 model. The process of approximation of a chaotic trajectory in terms of UPOs will play a central role in our investigation. In fact, we will use this tool to explore the properties of the attractor of the system under the lens of its UPOs. In the first part of the thesis we consider the Lorenz-63 model with the classic parameters’ value. We investigate how a chaotic trajectory can be approximated using a complete set of UPOs up to symbolic dynamics’ period 14. At each instant in time, we rank the UPOs according to their proximity to the position of the orbit in the phase space. We study this process from two different perspectives. First, we find that longer period UPOs overwhelmingly provide the best local approximation to the trajectory. Second, we construct a finite-state Markov chain by studying the scattering of the trajectory between the neighbourhood of the various UPOs. Each UPO and its neighbourhood are taken as a possible state of the system. Through the analysis of the subdominant eigenvectors of the corresponding stochastic matrix we provide a different interpretation of the mixing processes occurring in the system by taking advantage of the concept of quasi-invariant sets. In the second part of the thesis we provide an extensive numerical investigation of the variability of the dynamical properties across the attractor of the much studied Lorenz ’96 dynamical system. By combining the Lyapunov analysis of the tangent space with the study of the shadowing of the chaotic trajectory performed by a very large set of unstable periodic orbits, we show that the observed variability in the number of unstable dimensions, which shows a serious breakdown of hyperbolicity, is associated with the presence of a substantial number of finite-time Lyapunov exponents that fluctuate about zero also when very long averaging times are considered.