1: \begin{abstract}

2: We introduce some mathematical framework for functional extreme

3: value theory and provide basic definitions and tools. In

4: particular we introduce a functional domain of attraction approach

5: for stochastic processes, which is more general than the usual one

6: based on weak convergence.

7: 

8: The distribution function $G$ of a continuous max-stable process

9: on $[0,1]$ is introduced and it is shown that $G$ can be

10: represented via a norm on functional space, called $D$-norm. This

11: is in complete accordance with the multivariate case and leads to

12: the definition of functional generalized Pareto distributions (GPD) $W$.

13: These satisfy $W=1+\log(G)$ in their upper tails, again in

14: complete accordance with the uni- or multivariate case.

15: 

16: Applying  this framework to copula processes we derive

17: characterizations of the domain of attraction condition for copula

18: processes in terms of tail equivalence with a functional GPD.

19: 

20: $\delta$-neighborhoods of a functional GPD are introduced and it

21: is shown that these are characterized by a polynomial rate of

22: convergence of functional extremes, which is well-known in the

23: multivariate case.

24: \end{abstract}