Modeling Magnetotelluric Surveys Based on Stochastic Path Integral
Hongyu Zhou, Maokun Li, Fan Yang, Shenheng Xu, Aria Abubakar · 2024
We study a probabilistic modeling of magnetotelluric (MT) surveys based on stochastic path integral. Based on the Feynman-Kac formula, the probabilistic solution of Helmholtz equation is formulated by the expectation of path integrals of a series of random walk paths. The studied method based on stochastic path integral transforms the computational bottleneck from solving matrix equations to large-scale matrix multiplication, which is more suitable for massively parallel computing platforms. Numerical experiments show that as the number of paths increases, the numerical error of the studied method can effectively converge, and it also exhibits the considerable potential to efficiently accelerate the computation.