Special Session 4: Control and Optimization

On strong convergence of time and full numerical schemes for the stochastic 2D Navier-Stokes equations

Annie MILLET
University Paris 1 Panth\`eon-Sorbonne
France
Co-Author(s):    Hakima Bessaih
Abstract:
We prove the strong (i.e. in the set of square integrable random variables) convergence of a fully or semi-implicit time discretization, as well as of a full space-time discretization based on finite elements, of the 2D Navier-Stokes equations on the torus subject to a random perturbation. The initial condition belongs to in $H^1$ ; it is deterministic or has exponential moments. The speed of convergence is polynomial in case of an additive noise; it depends on the viscosity and the strength of the noise. The argument is based on convergence of a localized scheme, and on exponential moments of the solution to the stochastic 2D Navier-Stokes equations, and of its fully implicit time scheme. This work improves previous results which only described the speed of convergence in probability of these numerical schemes.