1: \begin{abstract} In this paper we develop   

2: adaptive numerical schemes for certain nonlinear variational problems.

3: The discretization of the variational problems is done by

4: a suitable frame decomposition of the solution, i.e.,

5: a complete, stable, and redundant expansion. The discretization

6: yields an equivalent nonlinear problem   on

7: $\ell_2(\mathcal{N})$, the space of   frame coefficients. The

8: discrete problem is then adaptively solved using approximated

9: nested fixed point and Richardson type iterations. We investigate the 

10: convergence, 

11: stability, and optimal complexity of the scheme. 

12: %%This shows a theoretical advantage of these adaptive  

13: %methods in comparison to  adaptive finite element schemes, for example, 

14: %whose convergence and complexity are usually difficult to study theoretically. 

15: This constitutes a theoretical advantage, for example, with

16: respect to adaptive finite element schemes for which convergence

17: and complexity results are usually hard to prove.

18: The use of frames is further motivated by their redundancy, which, at 

19: least numerically, 

20: has been shown to improve the conditioning of the corresponding discretization

21: matrices.

22: Also frames are usually easier to construct than Riesz bases. 

23: Finally, we show how to apply the adaptive scheme we

24: propose  for finding an approximation to the solution 

25: of the PDE governing magnetohydrodynamic (MHD) flows, once 

26: suitable frames are constructed.

27: \end{abstract}

28: