1: \begin{abstract}

2: We propose and analyze novel adaptive algorithms for the numerical solution

3: of elliptic partial differential equations with parametric uncertainty.

4: Four different marking strategies are employed for refinement of stochastic

5: Galerkin finite element approximations.

6: The algorithms are driven by the energy error reduction estimates derived from

7: two-level \textsl{a~posteriori} error indicators for spatial approximations and

8: hierarchical \textsl{a~posteriori} error indicators for parametric approximations.

9: The focus of this work is on the mathematical foundation of the

10: adaptive algorithms in the sense of rigorous convergence analysis.

11: In particular, we prove that the proposed algorithms

12: drive the underlying energy error estimates to zero.

13: \end{abstract}

14: