Analysis of a spatio-temporal advection-diffusion model for human behaviors during a catastrophic event (Q6572308)

The paper is concerned with the spatio-temporal macroscopic first-order non-conservative pedestrians model, namely \N\[\N\left\{\begin{array}{l} \partial_t\rho_i+\nabla\cdot q_i(\rho)=f_i(\rho_1,\ldots,\rho_5),\text{ in }[0,T)\times \Omega,\\\Nq_i\cdot n=q_i^0\cdot n,\text{ in }[0,T)\times \partial \Omega,\\\N\rho_i(0)=\rho_i^0,\text{ in }\Omega. \end{array}\right.\tag{1}\N\]\NHere \(\Omega\subset \mathbb{R}^2\) is a bounded domain with smooth boundary \(\partial \Omega\), \(\rho_i\) is the density of population representing a specific human behavior, \(q_i:=-d_i\nabla \rho_i+\rho_i\vec v_i(\rho)\) is the corresponding flux, \(f_i\) is a given nonlinear coupling reaction term, and \(\rho_i^0\) is the initial population density, for \(i=1,\ldots,5\). This model describes the evolution of a population in a sudden, unexpected and without warning signs disaster. By using the theory of semigroups of bounded linear operators, the authors prove the local existence, uniqueness and regularity of a solution to problem \((1)\). They also give sufficient conditions on the parameters which ensure the positivity and the \(L^1\)-boundedness of this solution. Numerical simulations for several scenarios of evacuation of a population in an emergency situation are also presented.