Paperlane

no. 17 · Wednesday 18 November 2026

Nonlinear Stability of Spiky Patterns for a Three-Dimensional Chemotaxis–Consumption Model with Signal-Dependent Motilities

Lin Chang, Song Xu and Yuling Xu

math.AP

Abstract

We prove the nonlinear asymptotic stability of one-dimensional spiky steady states for a three-dimensional chemotaxis–consumption system with signal-dependent motility $w^p$ ($p>0$, $α\ge2$) in the half-cylinder $Ω=\mathbb{R}_+\times\mathbb{T}^2$. The cell density satisfies a no-flux boundary condition at $x_1=0$, while the chemical concentration satisfies $w(0,x',t)=b>0$ and vanishes as $x_1\to+\infty$. For any prescribed steady-state cell mass $λ>0$, we show that any sufficiently small compatible initial perturbation with zero mass generates a unique global strong solution that converges to the spiky steady state in $L^\infty(Ω)$, with the chemical concentration converging uniformly relative to the steady profile, namely $\|w(\cdot,t)/W-1\|_{L^\infty(Ω)}\to0$ as $t\to\infty$. The stability analysis for an arbitrarily large steady mass $λ>0$ in three dimensions involves four main difficulties: the algebraic degeneracy of the diffusion coefficient $W^p(x_1)\sim (1+x_1)^{-2/(α-1)}$ as $x_1\to\infty$, the $O(λ)$ linear coupling between the density perturbation and the chemical logarithmic gradient on the transversal nonzero modes, the nonlinear boundary conditions at $x_1=0$ produced by the Hopf–Cole transformation, and the different spatial weights required across derivative orders. We resolve these difficulties by introducing the algebraic weights $(ω_1,ω_2,ω_3)=(W^{-p}U^{-1},U^{-1},W^pU^{-1})$, coupling a weighted Neumann inverse on the transversal nonzero modes with the scalar logarithmic potential $Ψ+q$ to cancel the $O(λ)$ linear coupling terms exactly, performing tangential and time-differentiated weighted energy estimates whose boundary traces vanish identically at $x_1=0$, and closing the time-derivative energy before recovering the full weighted Hessian dissipation from the parabolic system.

fields
Analysis of PDEs
pdf ↗ page arXiv:2611.22803 ↗