Article Contents
Article Contents

# The dynamics of pulse solutions for reaction diffusion systems on a star shaped metric graph with the Kirchhoff's boundary condition

• *Corresponding author: Shin-Ichiro Ei

Dedicated to the memory of Professor Masayasu Mimura

The first author was supported in part by JST CREST (No. JPMJCR14D3) and JSPS KAKENHI Grant Numbers JP22H01129

• In this paper, we consider motions of localized patterns for reaction-diffusion systems of general types on a metric star graph which consists of several half-lines with a common end point called "the junction point", where the Kirchhoff boundary condition is imposed. Assuming the existence and the stability of pulse and front like patterns for corresponding 1dimensional problems of reaction-diffusion systems, we rigorously derive ordinary differential equations describing the motions of them on a metric star graph. As the application, the attractive motion of a single pulse solution for the Gray-Scott model toward the junction point is shown. It is also shown that a single front solution of Allen-Cahn equation is repulsive against the junction point. The motion of multi pulse solutions and front solutions are also treated.

Mathematics Subject Classification: Primary: 35B32, 35B35, 35K55, 35K57; Secondary: 35Q92.

 Citation:

• Figure 1.  Schematic graph of a metric star graph in the case that $R = 3$, so called "Y-shaped metric graph"

Figure 2.  The flows for the ODE (19). (A): Flows in the case 1-a) are drawn for $R = 5$. Each flow corresponds to (a): $l_1(0) = l_2(0)$, (b): $0 < l_1(0) - l_2(0) < |\frac{1}{\alpha}\log\frac{2}{R-2}|$, (c): $0 < l_2(0) - l_1(0) < |\frac{1}{\alpha}\log\frac{2}{R-2}|$. (B): Flows in the case 1-c) ($R = 3$) are drawn. Each flow corresponds to (a): $l_1(0) = l_2(0)$, (b): $0 < l_1(0) - l_2(0) < \frac{\log 2}{\alpha}$, (c): $0 < l_2(0) - l_1(0) < \frac{\log 2}{\alpha}$

Figure 3.  Movement of a single front solution for (23) by a numerical simulation. (A) denotes the location $l_1(t)$ of the zero level point of the front solution. (B) draws the time evolution of $l_1(t)$. The movement apart from $O$ is observed. In the numerical simulations, each length of $\Omega_j$ is taken sufficiently large more than 10 with the Neumann boundary condition in the opposite side of $O$. The same treatments are done for other figures of numerical simulations as in Figures 4, 5, and 6

Figure 4.  Movement of two front solutions for (23) in the case when $R = 5$ and $r = 2$ corresponding to 2-a) when $l_2(0) - l_1(0) > \frac{1}{\alpha}\log\frac{2}{3}$ is satisfied. (A) denotes the locations $l_1(t)$, $l_2(t)$ of the zero level points of the front solutions on $\Omega_1$ and $\Omega_2$, respectively. (B) draws the time evolutions of $l_1(t)$ and $l_2(t)$. It is observed that $l_2(t)$ is monotone increasing in time while $l_1(t)$ goes toward $O$ for a while and then turns apart from $O$. (C) is the enlargement of the part surrounded by the dot circle in (B). The movement of ${^t}(l_1(t), l_2(t))$ approaching the set $\tilde {m}_2$ is also observed

Figure 5.  Movement of two front solutions for (23) in the case when $R = 4$ and $r = 2$ which corresponds to the case of 2-b). (A) denotes the locations $l_1(t)$, $l_2(t)$ of the zero level sets of the front solutions on $\Omega_1$ and $\Omega_2$, respectively. (B) draws the time evolutions of $l_1(t)$ and $l_2(t)$. It is observed that ${^t}(l_1(t), l_2(t))$ approaches a stationary state in the set $\tilde {m}_2$

Figure 6.  Movement of two front solutions for (23) corresponding to 2-c) when $l_2(0) - l_1(0) > \frac{\log 2}{\alpha}$ is satisfied. (A) denotes the locations $l_1(t)$, $l_2(t)$ of the zero level points of the front solutions on $\Omega_1$ and $\Omega_2$, respectively. (B) draws the time evolutions of $l_1(t)$ and $l_2(t)$. It is observed that $l_2(t)$ is monotone decreasing in time while $l_1(t)$ goes apart from $O$ for a while and then turns toward $O$. The movement of ${^t}(l_1(t), l_2(t))$ approaching the set $\tilde {m}_2$ is also observed

Figure 7.  The profile of a stationary pulse solution of (24)

•  [1] D. Beniaguev, I. Segev and M. London, Single cortical neurons as deep artificial neural networks, Neuron, 109 (2021), 2727-2739, e3. doi: 10.1016/j.neuron.2021.07.002. [2] J. Carr and R. L. Pego, Metastable patterns in solutions of $u_t = \varepsilon ^2u_xx+f(u)$, Comm, Pure Appl. Math., 42 (1989), 523-576.  doi: 10.1002/cpa.3160420502. [3] A. Doleman, R. A. Gardner and T. J. Kaper, Stability analysis of singular patterns in the 1-D Gray-Scott model, Physica D, 122 (1998), 1-36.  doi: 10.1016/S0167-2789(98)00180-8. [4] Y. Du, B. Lou, R. Peng and M. Zhou, The Fisher-KPP equation over simple graphs: Varied persistence states in river networks, J. Math. Biol., 80 (2020), 1559-1616.  doi: 10.1007/s00285-020-01474-1. [5] S.-I. Ei, The motion of weakly interacting pulses in reaction diffusion systems, J. Dynam. Differential Equations, 14 (2002), 85-137.  doi: 10.1023/A:1012980128575. [6] S.-I. Ei and T. Ishimoto, Effect of boundary conditions on the dynamics of a pulse solution for reaction-diffusion systems, Networks and Heterogeneous Media, 8 (2013), 191-209.  doi: 10.3934/nhm.2013.8.191. [7] S.-I. Ei and J. Wei, Dynamics of metastable localized patterns and its application to the interaction of spike solutions for the Gierer-Meinhardt system in 2 dimensional space, Japan J. Ind. Appl. Math., 19 (2002), 181-226.  doi: 10.1007/BF03167453. [8] P. C. Fife, Mathematical Aspects of Reacting and Diffusing Systems, Springer, 1979. [9] G. Fusco and J. Hale, The Slow motion manifold, dormant instability and singular perturbations, J. Dynamics and Different Equations, 1 (1989), 75-94.  doi: 10.1007/BF01048791. [10] D. Henry, Geometric Theory of Semilinear Parabolic Equations, Lecture Notes in Mathematics, 840, Springer-Verlag, 1981. [11] L. Hodgkin and A. F. Huxley, A quantitative description of membrane current and its application to conduction and excitation in nerve, J. Physiol, 1 (1952), 500-544. [12] S. Iwasaki, Asymptotic convergence of solutions of Keller-Segel equations in network shaped domains,, Nonlinear Analysis, 197 (2020), 111839, 28 pp. doi: 10.1016/j.na.2020.111839. [13] S. Jimbo and Y. Morita, Entire solutions to reaction-diffusion equations in multiple half-lines with a junction, J. Differential Equations, 267 (2019), 1247-1276.  doi: 10.1016/j.jde.2019.02.008. [14] S. Jimbo and Y. Morita, Asymptotic behavior of entire solutions to reaction-diffusion equations in an infinite star graph, Discrete Contin. Dyn. Syst., 41 (2021), 4013-4039.  doi: 10.3934/dcds.2021026. [15] S. Jimbo and Y. Takazawa, Y-shaped graph and time entire solutions of a semilinear parabolic equation, preprint. [16] K. Kawasaki and T. Ohta, Kink dynamics in one-dimensional nonlinear systems, Physica A, 116 (1982), 573-593.  doi: 10.1016/0378-4371(82)90178-9. [17] J. Keener and J. Sneyd, Mathematical Physiology I: Cellular Physiology, Second Edit. Springer, 2009. doi: 10.1007/978-0-387-79388-7. [18] J. D. Murray, Mathematical Biology I, Springer-Verlag, Berlin, 3rd edition, 2002. [19] Y. Nishiura, Coexistence of Infinity Many Stable Solutions to Reaction Diffusion Systems in Singular Limit Dynamics Reported, vol.3(New Series), Springer, 1994, 25-103. [20] Y. Nishiura, T. Teramoto and X. Yuan, Heterogeneity-induced spot dynamics for a three-component reaction-diffusion system, Communications on Pure and Applied Analysis, 11 (2012), 307-338.  doi: 10.3934/cpaa.2012.11.307. [21] W. Rall, G. M. Shepherd, T. S. Reese and M. W. Brightman, Dendro-dendritic synaptic pathway for inhibition in the olfactory bulb, Exp. Neurol., 14 (1966), 44-56.  doi: 10.1016/0014-4886(66)90023-9. [22] E. Yanagida, Stability of nonconstant steady states in reaction-diffusion systems on graphs, Japan J. Indust. Appl. Math., 18 (2001), 25-42.  doi: 10.1007/BF03167353.
Open Access Under a Creative Commons license

Figures(7)