STOCHASTIC-EQUATIONS WITH MULTIFRACTAL RANDOM INCREMENTS FOR MODELING TURBULENT DISPERSION

Previously studied stochastic models of one-particle dispersion in stationary, isotropic, and homogeneous turbulence are reconsidered and intermittency corrections sought. Known Lagrangian intermittency effects, in the form of multifractal scaling, independently derived from Eulerian measurements [M. S. Borgas, Philos. Trans. R. Soc. London Ser. A 342, 379 (1993)], are used to develop a new model. The previous models and approaches are shown to be inadequate. The new model incorporating Lagrangian intermittency satisfies Thomson's well-mixed criterion [J. Fluid Mech. 180, 529 (1987)] and gives almost-Gaussian mean-concentration distributions for Gaussian sources. The trajectories generated by the model are not fractal, in agreement with the results of Borgas. The practical impact of intermittency upon dispersion is found to be small.