No pattern formation in a quasilinear chemotaxis model with local sensing

Philippe Laurençot,Ariane Trescases
2024-04-18
Abstract:Convergence to spatially homogeneous steady states is shown for a chemotaxis model with local sensing and possibly nonlinear diffusion when the intrinsic diffusion rate $\phi$ dominates the inverse of the chemotactic motility function $\gamma$, in the sense that $(\phi\gamma)'\ge 0$. This result encompasses and complies with the analysis and numerical simulations performed in Choi \& Kim (2023). The proof involves two steps: first, a Liapunov functional is constructed when $\phi\gamma$ is non-decreasing. The convergence proof relies on a detailed study of the dissipation of the Liapunov functional and requires additional technical assumptions on $\phi$ and $\gamma$.
Analysis of PDEs
What problem does this paper attempt to address?
The paper primarily investigates the pattern formation problem in a class of chemotaxis models with local sensing capabilities, particularly focusing on whether the system converges to a spatially uniform steady state when the intrinsic diffusion rate \( \phi \) is sufficiently large relative to the inverse of the chemotactic sensitivity function \( \gamma \). Specifically, the authors consider a coupled system of partial differential equations describing the dynamics of cell density \( u \) and the concentration of a chemical substance \( v \) secreted by the cells. This system includes a term reflecting cell diffusion and chemotaxis, as well as an equation describing the diffusion of the chemical substance. By introducing a Lyapunov functional and conducting a detailed study of its dissipation, the authors prove that under certain conditions, namely when the derivative of \( \phi/\gamma \) is non-negative, the solution of the system converges to a spatially uniform steady-state solution. The specific results include: 1. When the derivative of \( \phi/\gamma \) is non-negative, it is proven that the solution \( v \) converges to the mean value \( M \) in the \( L^q \) space (\( q \in [1,2) \)), and \( u \) also converges to \( M \) in the time-averaged sense. 2. Furthermore, under certain technical assumptions on the initial conditions, stronger convergence properties for \( u \) can be obtained. 3. For specific forms of the \( \phi \) and \( \gamma \) functions, such as power-law or exponential functions, the authors provide more detailed convergence results. In summary, this paper mathematically demonstrates that under appropriate conditions, chemotaxis models with local sensing capabilities do not form complex spatial patterns but instead converge to a uniform state. This result is significant for understanding the mechanisms of pattern formation in biological systems.