Detecting distributional changes in samples of independent block maxima using probability weighted moments

DOI10.1007/S10687-016-0273-1MaRDI QIDQ1675709zbMATH OpenOpenAlexWikidataFDO

Authors Ivan Kojadinovic, Philippe Naveau

Publication date 2 November 2017

Published in Extremes (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1507.06121

zbMATH Keywords

block maximacumulative sum statistics (CUSUM)generalized extreme value distributionprobability weighted moment estimatorweighted empirical process

Mathematics Subject Classification ID

Nonparametric hypothesis testing (62G10) Asymptotic properties of nonparametric inference (62G20) Statistics of extreme values; tail inference (62G32) Applications of statistics to environmental and related topics (62P12)

Abstract: The analysis of seasonal or annual block maxima is of interest in fields such as hydrology, climatology or meteorology. In connection with the celebrated method of block maxima, we study several tests that can be used to assess whether the available series of maxima is identically distributed. It is assumed that block maxima are independent but not necessarily generalized extreme value distributed. The asymptotic null distributions of the test statistics are investigated and the practical computation of approximate p-values is addressed. Extensive Monte-Carlo simulations show the adequate finite-sample behavior of the studied tests for a large number of realistic data generating scenarios. Illustrations on several environmental datasets conclude the work.

Review text

The block maxima (BM) method in extreme value theory is a useful tool in the analysis of data in climatology, hydrology and meteorology. Statistical properties of maximum likelihood estimators and probability weighted moments (PWM) are considered by several authors. In the present paper, the question of distributional changes is investigated: The block maxima approach consists of observation period into nonoverlapping periods; the new observations are the maxima in each period. The distribution of these block maxima is modeled by the generalized extreme value (GEV) distribution. The question arises, whether the maxima have all the same distribution or not. To answer this question the authors propose test statistics which are related to the CUSUM statistics for the change-point detection. To formulate the test procedures the PWM method is recalled and generalized. The authors derive the limit distribution of the test statistics under the null hypothesis -- a result on the weak convergence of the sequential weighted uniform empirical process and of certain sequential empirical processes constructed from the PWM estimators. Based on this limit statement the computation of the \(p\)-values is discussed. Monte Carlo experiments demonstrate the finite-sample properties of the proposed test procedures. Further, the test is applied to real climate data.

Recommendations

• On the block maxima method in extreme value theory: PWM estimators
• Improving probability-weighted moment methods for the generalized extreme value distribu\-tion
• Maximum likelihood estimators based on the block maxima method
• Skew generalized extreme value distribution: probability-weighted moments estimation and application to block maxima procedure
• Maximum likelihood estimation for the Fréchet distribution based on block maxima extracted from a time series

Cites work

• Dependent multiplier bootstraps for non-degenerate U-statistics under mixing conditions with applications
• Extreme value theory. An introduction.
• Extremes and related properties of random sequences and processes
• scientific article; zbMATH DE number 3141621 (Why is no real title available?)
• scientific article; zbMATH DE number 1048663 (Why is no real title available?)
• Improving probability-weighted moment methods for the generalized extreme value distribu\-tion
• Introduction to empirical processes and semiparametric inference
• Nonparametric tests for change-point detection à la Gombay and Horváth
• On the block maxima method in extreme value theory: PWM estimators
• Parameter and Quantile Estimation for the Generalized Pareto Distribution
• Rank-based inference for bivariate extreme-value copulas
• Return level bounds for discrete and continuous random variables
• Time Series Regression with a Unit Root
• Weak convergence and empirical processes. With applications to statistics

Cited in

(5)

• On the block maxima method in extreme value theory: PWM estimators
• Sources of uncertainty in the extreme value statistics of climate data
• Detecting change in UK extreme precipitation using results from the climateprediction.net BBC climate change experiment
• On the disjoint and sliding block maxima method for piecewise stationary time series
• Statistics for heteroscedastic time series extremes

Describes a project that uses

Uses Software

• ismev
• evd
• npcp
• R

This page was built for publication: Detecting distributional changes in samples of independent block maxima using probability weighted moments

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1675709)