Oscillating Turing patterns, chaos and strange attractors in a reaction-diffusion system augmented with self- and cross-diffusion terms

Benjamin Aymard · 2024

Read Oscillating Turing patterns, chaos and strange attractors in a reaction-diffusion system augmented with self- and cross-diffusion terms by Benjamin Aymard (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-17076.txt.

Introduction

Benjamin Aymard studies how order and chaos emerge in a spreading, reacting system in this 2024 mathematics paper. The model tracks two species that diffuse and react, with extra terms for self-diffusion and cross-diffusion. Such systems are known for Turing patterns, the regular stripes and spots that appear when diffusion destabilizes a uniform state.

Aymard builds a Fourier spectral method to follow the system's steady states and orbits with high accuracy. That accuracy matters when tracing periodic orbits and testing their stability. Varying one control parameter across four diffusion regimes, he finds new kinds of Turing patterns and oscillating patterns born through Hopf bifurcations. Further changes trigger period doubling that ends in chaos. The paper closes by reporting new strange attractors, the tangled sets that chaotic orbits trace in phase space.

Catalog record

CREATORBenjamin Aymard
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-17076

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Oscillating Turing patterns, chaos and strange attractors in a reaction-diffusion system augmented with self- and cross-diffusion terms Benjamin Aymard Abstract In this article we introduce an original model in order to study the emergence of chaos in a reaction diffusion system in the presence of self- and cross-diffusion terms. A Fourier Spectral Method is derived to approximate equilibria and orbits of the latter. Special attention is paid to accuracy, a necessary condition when one wants to catch periodic orbits and to perform their linear stability analysis via Floquet multipliers. Bifurcations with respect to a single control parameter are studied in four different regimes of diffusion: linear diffusion, self-diffusion for each of the two species, and cross-diffusion. Key observations are made: development of original Turing patterns, Hopf bifurcations leading to oscillating patterns and period doubling cascades leading to chaos. Eventually, original strange attractors are reported in phase space. Keywords: Self-diffusion; Cross-diffusion; Turing instability; Hopf bifurcation; Floquet theory; Period doubling cascade; Chaos; Strange attractors. 1 Introduction When does a deterministic dynamical system become unpredictible? In his 1814 memoir on probabilities [ 1 ] , Pierre-Simon de Laplace hypothetised that one could predict the future of a dynamical system given the exact knowledge of positions and impulsion of its components at an initial time. Later, James Clerk Maxwell observed in 1876 [ 2 ] that for a class of systems, an error on the initial condition could dramatically modify its outcome. Eventually, Henri Poincaré’s work in celestial mechanics from 1892 [ 4 ] led him to discover the unpredictability of trajectories in the three body problem. In doing so, he provided fundamental tools, of what would later become the chaos theory. Beyond its philosophical aspect, and its origin in celestial mechanics, chaos theory finds practical applications in everyday life, in problems ranging from meteorology [ 10 ] to electrical engineering, passing by chemical reactions [ 12 ] , plasma physics [ 9 ] , laser physics [ 23 ] and mechanics [ 22 ] . Understanding chaotic systems is a challenging task, with numerous potential applications. Using the tools of bifurcation theory [ 28 , 30 ] , several mechanisms explaining the developement of chaos have been identified. The key idea of this aproach is to evaluate the sequence of bifurcations (loss of stability of equilibria or periodic orbits) while a control parameter is varied. Among the identified scenarios is the period doubling cascade. Discovered in 1976 in numerical experiments on the logistic map (see [ 11 ] ), then observed in numerous experiments, such as in electronic circuits in 1981 by Linsay [ 18 ] , or in mercury convection rolls in 1982 by Libchaber [ 19 ] , it consists of a sequence of period doubling of periodic orbits. After each period doubling, the distance between consecutive bifurcation diminishes, and eventually, the period tends to infinity for a finite value of the control param€ter: the orbit becomes aperiodic. Careful measurements of distances between consecutive bifurcation values led to the discovery of the Feigenbaum constant [ 15 , 16 ] , revealing a concept of universality of the period doubling cascade. It was then observed in many other models, in particular in the Rössler model for chemical reaction [ 12 ] , the Mackey-Glass system in blood cells population dynamics [ 13 ] , neuron models of Fitzhugh-Nagumo type [ 20 ] , the Kuramoto-Sivashinsky system modeling flamme front propagation [ 21 ] , the Chua circuit [ 24 ] , the forced Dufing oscillator [ 5 ] , among others. Interested reader may find a historical note about period doubling in [ 34 ] . However, even if many road to chaos have been identified, the question of understanding how does the presence of self- and/or cross-diffusion affect the development of chaos in reaction diffusion systems, remains unclear. This work, which aims precisely at answering this question, stands at the crossroads of three domains that appeared during the 20th century. First, the domain of morphogenesis, initiated by the seminal paper of Alan Turing [ 8 ] , studying the appearance of periodically spatial patterns, as a result of an instability created by the interaction between linear diffusion, and nonlinear reaction. Second, the domain of spatio-temporal chaos, notably following the works of Kuramoto on diffusion induced chaos, where, for instance, the interaction between a Hopf bifurcation and a Turing instability can create oscillating patterns, possibly becoming chaotic [ 7 , 14 ] . Third, the domain of self- and cross-diffusion, motivated by questions in population dynamics and in chemistry, notably the works of Shigesada Kawasaki Terramoto [ 17 ] and Vanag and Epstein [ 27 ] , and more recently in [ 35 , 39 ] . There is therefore a need for a simple model, with bifurcations driven by a single parameter, gathering those three elements, for which we could study in a continuous way the development of chaos. Due to the presence of nonlinearities on both reaction and diffusion terms, this kind of system is, in general, not analytically solvable, except in special cases. Therefore, reliable numerical methods [ 29 ] have to be designed in order to study them numerically. To this end, high accuracy is a necessary condition when one wants to catch periodic orbits, and to perform their linear stability analysis via Floquet multipliers [ 3 ] . The outline of the paper is the following. First, an original model is introduced, containing self- and cross-diffusion, with bifurcations controled by a single parameter. Together with the model, a Fourier spectral method is derived, and theoretical convergence guarantees are provided. Second, bifurcations, of both equilibria and periodic orbits, are studied in four regimes: linear diffusion, activator self-diffusion, inhibitor self-diffusion, and cross-diffusion. Finally, the road to chaos is identified in each case, and the strange attractors that appear in the chaotic regime, are described. 2 Model and method 2.1 Model In [ 26 , 33 ] the authors introduce a model containing a richness of behavior, while being relatively simple, as the bifurcations are driven by one control parameter C , making it, as they mention, a good laboratory for studying oscillation and chaos in two species reaction diffusion models. We will refer to this model as the classical BVAM model. Along those lines, our aim is to introduce a generalization of their model, by introducing non linear diffusion terms. Let us consider the dynamics of two species, u_{1} and u_{2} , on a domain \Omega=[-L_{x},L_{x}] , described by the reaction diffusion system: \displaystyle\frac{\partial u_{1}}{\partial t} \displaystyle=\Delta(d_{1}u_{1}+d_{11}u_{1}^{3}+d_{12}u_{2}^{2}u_{1})+\eta(u_{1}+au_{2}-Cu_{1}u_{2}-u_{1}u_{2}^{2}), \displaystyle\frac{\partial u_{2}}{\partial t} \displaystyle=\Delta(d_{2}u_{2}+d_{22}u_{2}^{3}+d_{12}u_{1}^{2}u_{2})+\eta(bu_{2}+Hu_{1}+Cu_{1}u_{2}+u_{1}u_{2}^{2}), (1) with \eta , a , b , C and H reaction parameters, and d_{i},d_{ii},d_{ij} for i,j=1,2,i\not=j diffusion parameters. Let us remark that, when d_{11}=d_{22}=d_{12}=0 , the model ( 1 ) reduces to the original BVAM model. The problem is mathematically closed by adding initial conditions and periodic boundary conditions on the domain \Omega . In their original work [ 33 ] , the authors considered Dirichlet boundary conditions. Our choice of periodic boundary conditions here is motivated by two reasons: this reduces the importance of the central point of the domain, later used to define a phase space, and it allows to easily use Fourier analysis [

Read the complete text