A Stochastic Process for the Dynamics of the Turbulent Cascade

Abstract

Velocity increments over a distance r and turbulent energy dissipation on a box of size r are well described by the multifractal models of fully developed turbulence. These quantities and models however, do not involve time-correlations and therefore are not a detailed test of the dynamics of the turbulent cascade. If the time development of the turbulent cascade, in the inertial range, is related to the lifetime of the eddies at different length scales, the time correlations may be described by a stochastic process on a tree with jumping kernels which are a function of the ultrametric (tree) distance. We obtain the solutions of the Chapman-Kolmogorov equation for such a stochastic process, with jumping kernels depending on the ultrametric distance, but with an arbitrarily specified invariant probability measure. We then show how to use these solutions to compute the time correlations in the turbulent cascade.