A Mixed FEM for a Time-Fractional Fokker–Planck Model

We propose and analyze a mixed finite element method for the spatial approximation of a time-fractional Fokker–Planck equation in a convex polyhedral domain, where the given driving force is a function of space. Taking into account the limited smoothing properties of the model, and considering an appropriate splitting of the errors, we employed a sequence of clever energy arguments to show optimal convergence rates with respect to both approximation properties and regularity results. In particular, error bounds for both primary and secondary variables are derived in \(L^2\) $L^2$ -norm for cases with smooth and nonsmooth initial data. We further investigate a fully implicit time-stepping scheme based on a convolution quadrature in time generated by the backward Euler method. Our main result provides pointwise-in-time optimal \(L^2\) $L^2$ -error estimates for the primary variable. Numerical examples are then presented to illustrate the theoretical contributions.