Paper Title
Multivariate Extreme Value Theory and Application to Environment

Abstract
In this paper, we are interested in modeling the dependence structure for a multivariate extreme value distribution. A Monte Carlo simulation is conducted to compare the performance of different nonparametric estimators of the dependence function. Experimental results indicate that the Zhang estimator has better performance than the other dependence function estimators. We finally carry out an empirical application to the air pollution data in Tunisia, in order to illustrate the usefulness of the dependence function estimators to describe the link between extreme ozone concentrations observed in some polluted stations. Index Terms� multivariate extreme value distribution, dependence function, air pollution, ozone modeling