1: \begin{abstract}

2: In this article, the solution of a statistical inverse problem $M = AU+\Ec$ by the Bayesian approach is studied where

3: $U$ is a function on the unit circle $\T$, i.e., a periodic signal. The mapping $A$ is a smoothing linear operator

4: and $\Ec$ a Gaussian noise. The connection to the solution of  

5: a finite-dimensional computational model $M_{kn} = A_k U_n + \Ec_k$ is discussed.

6: Furthermore, a novel hierarchical prior model for obtaining edge-preserving conditional mean estimates is introduced.

7: The convergence of the method with respect to finer discretization is studied and

8: the posterior distribution is shown to converge weakly.

9: Finally, theoretical findings are illustrated by a numerical example with simulated data.

10: \end{abstract}

11: