Statistical analysis for stationary time series at extreme levels: New estimators for the limiting cluster size distribution

The serial dependence of a stationary time series at extreme levels may be captured by the limiting cluster size distribution. New estimators based on blocks declustering scheme are proposed and analyzed both theoretically means large-scale simulation study. A sliding version is shown to outperform disjoint version. In contrast some competitors from literature, only depend one tuning parameter chosen statistician.