Seminar| Institute of Mathematical Sciences
Time: Friday, November 20th, 2026,16:00-17:00
Location: IMS S408
Speaker: Wenju Zhao, Shandong University
Abstract: This talk focuses on a discrete finite element method and its error analysis for the backward stochastic Navier–Stokes equations. Backward stochastic differential equations and their partial differential equation counterparts arise naturally in stochastic control, dynamic risk measures, financial derivatives pricing, data assimilation, and inverse problems under uncertainty. They describe how a system evolves backward in time from a prescribed terminal state or a future target. Stable and efficient numerical methods for backward stochastic fluid equations are thus of theoretical interest and provide useful computational tools for prediction, control, and inversion in complex stochastic dynamical systems. We propose a discrete finite element scheme that addresses the challenges posed by stochastic perturbations, the incompressibility constraint, velocity–pressure coupling, and nonlinear convection. The scheme has a natural parallel structure in probability space and is well suited to simulations with large numbers of stochastic samples. Our analysis combines techniques from stochastic analysis and finite element approximation theory with mathematical induction. Particular attention is paid to error propagation due to the nonlinear convection term $u\cdot\nabla u$. We establish optimal-order convergence estimates in time and space for the velocity u, the pressure p, and the auxiliary stochastic process z. Numerical experiments confirm the theoretical results and demonstrate the effectiveness and stability of the proposed method.