Monte Carlo simulation in the theory of wave propagation in random media: I. Modified and Feynman path integrals

Abstract

Abstract The purpose of this article (comprising parts I and II) is to develop and test the approach of combining a path-integral technique and a complex-valued Monte Carlo method to calculate the highest moments of the Green function of the stochastic wave equation for a random medium against the background of large-scale inhomogeneities. In part I, the new modified path-integral representations of the Green function moments of the stochastic wave equation have been developed. The limiting transition of these representations to the Feynman path integrals corresponding to the parabolic approximation is discussed. Path-integral representations for Green function moments are given for three models: a model of the stochastic wave equation and models of parabolic and Markov approximations. The Metropolis algorithm underlying the Monte Carlo method for calculating real and complex-valued path integrals is discussed in brief. Numerical results are presented in part II of the article.