| Abstract: |
| Under certain assumptions, the solution of parabolic semilinear SPDEs driven by additive space-time white noise (in dimension 1), converges as time gets large, to a unique invariant probability distribution. I will discuss how the error due to spatial and temporal discretization can be analyzed. The (weak) orders of convergence are equal to $1$ and $1/2$ respectively, due to low regularity properties of the process. I will present some new techniques to design higher-order integrators for temporal discretization. This work is supported by theoretical analysis and numerical experiments. |
|