1: \begin{abstract}

2: %We investigate the weak rate of convergence of a space-time discrete approximation scheme for semilinear parabolic stochastic evolution equations driven by additive square-integrable Lévy noise.

3: %We investigate the weak approximation error for space-time discretizations of semilinear parabolic stochastic evolution equations driven by additive square-integrable Lévy noise.

4: We investigate the weak order of convergence for space-time discrete approximations of semilinear parabolic stochastic evolution equations driven by additive square-integrable Lévy noise.

5: To this end, the Malliavin regularity of the solution is analyzed and recent results on refined Malliavin-Sobolev spaces from the Gaussian setting are extended to a Poissonian setting.

6: For a class of path-dependent test functions, we obtain that the weak rate of convergence is twice the strong rate.

7: %It is shown that the solution to stochastic evolution equations in Hilbert space, driven by a square integrable Levy process without Gaussian part, is regular in the sense of Malliavin. Moreover, recent results on refined Sobolev-Malliavin spaces from the Gaussian setting are extended to the setting of the present paper. They are applied to prove essentially sharp weak rate of convergence of a full discretization of semilinear stochastic reaction-diffusion equations with additive noise. Discretization is done by means of the finite element method and a backward Euler time-stepping.

8: \end{abstract}

9: