Complex attractors and patterns in reaction-diffusion systems (Q1743981)

This paper is devoted to the analysis of long time behavior and pattern formation in the following class of reaction-diffusion systems. Let \(u=u(x,y,t)\) be a scalar function and \({\mathbf v}={\mathbf v} (x,y,t)\) be a vector function, both posed on an infinite domain \(\Omega\times \{t\geq 0\}\), where \(\Omega= (-\infty,\infty)\times(0,1)\), such that \((u, {\mathbf v})\) solves the system \[ \begin{aligned} u_t=d\Delta u +f(u,{\mathbf v})+ \zeta(x,y),\\ {\mathbf v}_t={\mathbf D}\Delta {\mathbf v} +{\mathbf g}(u,{\mathbf v})+ {\boldsymbol \eta}(x,y). \end{aligned} \] Here \(\zeta\) and \(\eta_i\) are smooth functions representing external sources, \(g_i\) and \(f \) are sufficiently smooth. The diffusion is described by a positive \(d \ll 1\) and a diagonal matrix \({\mathbf D}\) with entries \(D_i \gg d\). More precisely, it is assumed that \(d=\gamma^{2s}\) with \(s \in (0,0.01)\) and \(D_i=\gamma^{-p_0} d_i\), where \(d_i\) is independent of \(\gamma\) for any \(i\), and \(p_0=1-\frac{3s}{2}+\frac{s_1}{2}\) with \(0\leq s_1\leq \frac{s}{10}.\) The assumption of the smallness of \(\gamma\) imposes a slow-fast structure on the dynamics of this system. The author formulates sufficient conditions on the reaction terms \(f\) and \(\mathbf g\) which in case of scalar \(v\) and \(g\) can be simply formulated as the existence of a root of \(f_v\) which is not a root of \( g_{u} \) or the existence of a root of \(g_u\) which is not a root of \( f_{v} \) and requires a longer, more detailed description in case of multicomponent vectors \(\mathbf v\) and \(\mathbf g\). The author then combines Turing's idea of importance of the significant difference in the rates of diffusion of the components of the system with the method of realization of vector fields developed by Poláčik and, as a result, demonstrates the complex nature of the long-time behavior of the system as a function of \(\gamma\), \(\zeta\), and \(\boldsymbol \eta\), for fixed reaction terms \(f\) and \(\mathbf g\).