Abstract

A new likelihood method to estimate the extremal index, together with an application to temperature data, is presented. Conditions for the validity of the model are discussed and diagnostics are proposed.

Details

Actions