Validity of formal expansions for singularly perturbed competition-diffusion systems

TL;DR

This paper rigorously validates formal asymptotic expansions for two-species competition-diffusion systems near the sharp interface limit, confirming the convergence of transition layer profiles to traveling waves and analyzing interface motion via mean curvature flow.

Contribution

It provides a rigorous proof of the convergence of transition layers to traveling waves and extends Liouville type theorems to multi-species systems with periodic coefficients.

Findings

Transition layers converge to traveling wave solutions as ε→0

Interface motion follows mean curvature flow with a driving force

Liouville type theorems hold for multi-species systems with periodic coefficients

Abstract

We consider a two-species competition-diffusion system involving a small parameter $\varepsilon>0$ and discuss the validity of formal asymptotic expansions of solutions near the sharp interface limit $\varepsilon\approx0$. We assume that the corresponding ODE system has two stable equilibria. As in the scalar Allen--Cahn equation, it is known that the motion of the sharp interfaces of such systems is governed by the mean curvature flow with a driving force. The formal expansion also suggests that the profile of the transition layers converges to that of a traveling wave solution as $\varepsilon\rightarrow0$. In this paper, we rigorously verify this latter ansatz for a large class of initial data. The proof relies on a rescaling argument, the super--subsolution method and a Liouville type theorem for eternal solutions of parabolic systems. Roughly speaking, the Liouville type theorem states that any eternal solution that lies between two traveling waves is itself a traveling wave. The same Liouville type theorem was established for the scalar Allen--Cahn equation by Berestycki and Hamel. In view of their importance, we prove the Liouville type theorems in a rather general framework, not only for two-species competition-diffusion systems but also for $m$-species cooperation-diffusion systems possibly with time periodic or spatially periodic coefficients.