Instability Thresholds and Dynamics of Mesa Patterns in Reaction-Diffusion Systems

Abstract

We consider reaction-diffusion systems of two variables with Neumann boundary conditions on a finite interval with diffusion rates of different orders. Solutions of these systems can exhibit a variety of patterns and behaviours; one common type is called a mesa pattern; these are solutions that in the spatial domain exhibit highly localized interfaces connected by almost constant regions. The main focus of this thesis is to examine three different mechanisms by which the mesa patterns become unstable. These patterns can become unstable due to the effect of the heterogeneity of the domain, through an oscillatory instability, or through a coarsening effect from the exponentially small interaction with the boundary. We compute instability thresholds such that, as the larger diffusion coefficient is increased past this threshold, the mesa pattern transitions from stable to unstable. As well, the dynamics of the interfaces making up these mesa patterns are determined. This allows us to describe the mechanism leading up to the instabilities as well as what occurs past the instability threshold. For the oscillatory solutions, we determine the amplitude of the oscillations. For the coarsening behaviour, we determine the motion of the interfaces away from the steady state. These calculations are accomplished by using the methods of formal asymptotics and are verified by comparison with numerical computations. Excellent agreement between the asymptotic and the numerical results is found.