Automated Selection of r for the r Largest Order Statistics Approach with Adjustment for Sequential Testing

TL;DR

This paper introduces new sequential testing methods, including bootstrap and entropy-based tests, for selecting the optimal number of order statistics in extreme value analysis, improving accuracy and computational efficiency.

Contribution

It develops novel sequential tests for choosing r in the r largest order statistics approach, with methods that control error rates and enhance computational speed.

Findings

Both tests effectively detect model misspecification in simulations.

The procedures control false discovery and familywise error rates.

Applications demonstrate improved extreme event analysis.

Abstract

The r largest order statistics approach is widely used in extreme value analysis because it may use more information from the data than just the block maxima. In practice, the choice of r is critical. If r is too large, bias can occur; if too small, the variance of the estimator can be high. The limiting distribution of the r largest order statistics, denoted by GEVr, extends that of the block maxima. Two specification tests are proposed to select r sequentially. The first is a score test for the GEVr distribution. Due to the special characteristics of the GEVr distribution, the classical chi-square asymptotics cannot be used. The simplest approach is to use the parametric bootstrap, which is straightforward to implement but computationally expensive. An alternative fast weighted bootstrap or multiplier procedure is developed for computational efficiency. The second test uses the difference in estimated entropy between the GEVr and GEV(r-1) models, applied to the r largest order statistics and the r-1 largest order statistics, respectively. The asymptotic distribution of the difference statistic is derived. In a large scale simulation study, both tests held their size and had substantial power to detect various misspecification schemes. A new approach to address the issue of multiple, sequential hypotheses testing is adapted to this setting to control the false discovery rate or familywise error rate. The utility of the procedures is demonstrated with extreme sea level and precipitation data.