Models for the diffusion process of particle were proposed.
The study contains two main parts as follows. The one is an analysis of an equation of motion of a particle and the other is the concentration of the models of atmospheric turbulence to estimate velocity which are substituted into velocity terms of the equation of motion of a particle. A movement of particle is governed by forces of gravity, viscosity of air, pressure, accerelation and Basset. The effect of each force was evaluated. The following results were recognized.(1) Forces of gravity and viscosity are important, if the particle is small.(2) The movement of particle is almost similar to that of air when the diameter of particle is smaller than several microns, in the conditions that the density is 3.0 g/cm³ and the integral time scale of turbulent eddy is 9.0 sec.
Three models to simulate atmospheric turbulence were evaluated.
These are a simple Monte Carlo method (Model I), a first order Markov process model taken into account past wind speed (Model II) and a stochastic differential equation taken into account past noise for wind speed (Model III). As a result, it was recognized that Model III is best to express turbulence characteristics as variance, integral time scale and energy spectrum of wind velocity, and atmospheric stability by combining with a higher order self-regression model to self-correlation function of wind velocity.