We consider a bistable reaction-diffusion-advection system as a population model describing the growth of biological individuals which move by diffusion and chemotaxis. Assuming that the diffusion rate and the chemotactic rate are both very small with the growth rate, we introduce the limiting system by using the singular limit procedure and study the dynamics of growth patterns arising in this system. It is shown that in two dimensional domain planar traveling front solutions are transversally stable when the chemotaxis effect is weak and, when it becomes stronger, they are destabilized. But, these results depend on the sign of the second derivative of the sensitive function. Numerical simulations reveal that the destabilized solution evolves into complex patterns with dynamic spot structures and so on. To explain the appearance of these complex patterns, we obtain the estimate of the attractor dimension from below and it tends to infinity as the chemotaxis effect is increasing.

抄録全体を表示

The purpose of this article is to describe some properties of solutions to the Keller-Segel model and a simplified Keller-Segel model. The Keller-Segel model was introduced to describe chemotactic aggregation of cellular slime molds, and the simplified Keller-Segel model was also introduced to describe the same phenomenon. Moreover, the simplified Keller-Segel model describes the gravitational interaction of particles. The solutions to those models represent the density of cells or particles. Then, the solutions are functions of spatial and time variables. Those systems have blowup solutions. We say that a solution blows up at finite time, if maximum value of the solution in the spatial domain tends to infinity as time variable tends to the finite time. We think that the blowup corresponds to aggregation of the cells or the particles. In this article, we describe some results about blowup solutions and the relation between those results and the biological phenomenon.

抄録全体を表示

We consider the following reaction-diffusion equation (KS)_m. Throughout this article, we assume that N≥1, and that m, q, and γ are the constants satisfying m>1, q≥2, γ≥0, respectively. This equation is often called the Keller-Segel model describing the motion of the chemotaxis molds. Here u(x,t) and v(x,t) denote the density of amoebae and the concentration of the chemo-attractant, respectively. In the semi-linear case i.e., m=1, up to the present there have been many results for q=2. As for the quasi-linear case m>1, the balance between diffusion strength m and non-linearity effect q plays an important role for existence of global solutions to (KS)_m. In this article, we introduce recent results on global existence and blow-up in a finite time of solution of (KS)_m with the critical exponet q=m+2/N. Furthermore, we introduce a significant feature of degenerate one such as the property of finite speed of propagation for solutions, which is different phenomena from that of non-degenerate one. In addition, we characterize the interface curve as the solution of a certain ordinary differential equation associated with (KS)_m. We also state the so-called ε-regularity theorem of (KS)_m. Finally, we give several open problems.

抄録全体を表示

近年急速に研究が進展している生命現象の一つである,腫瘍による血管新生について,数理的背景が明確な二つの数理モデルを用いて考えることにする.まず,オスマーとスティーヴンスによる放物型方程式-常微分方程式系の時間大域解の存在と漸近挙動について考える.このモデルは本来,強化ランダム歩行の理論から導かれたものであるが,レヴィンとスリーマンによって腫瘍による血管新生のメカニズムの理解に適用された.次に,基礎医学の研究結果に基づいたアンダーソンとチャプレンによる腫瘍血管新生モデルについて同様な観点から考えることにする.しかしこれら2モデルについて,これまで関連付けられた扱いはされなかった.そこで統一的な数学的手法を用いて,解の存在と漸近挙動を調べ,数学的妥当性を保証し,さらに両モデルの数学的構造を比較検討する.これにより両モデルの同等性について数学的議論をし,お互いのモデルの妥当性の相互補完,この生命現象を支配する数理原理の特徴づけ,解の可解性と性質の観点からの数学的枠組みを与える.ここで用いられた数学的方法を,関連した生体現象の数学解析に適用する.

抄録全体を表示

We are concerned with numerical methods for the Keller-Segel system that describes the aggregation of slime molds resulting from their chemotactic features. Whereas the system has positivity and mass conservation properties, it is not certain that numerical schemes for the system preserve these conservation properties. In the present article, we review two conservative numerical schemes proposed by the author and discuss how to choose the time increments and the space meshes in order to realize those conservation properties. The first one is the finite-difference method that makes use of the semi-implicit time discretization with the time-increment control and the upwind difference approximation. The second is the finite-element method that is an application of Baba-Tabata's conservative upwind finite element. Conservative properties are proved via the theory of M-matrices. Error analysis of the finite-element scheme is also summarized. In particular, we have error estimates with explicit convergence rates by virtue of the analytical semigroup theory.

抄録全体を表示