Generalized Gumbel model for r-largest order statistics, with an application to peak streamflow

Abstract The effective use of available information in extreme value analysis is critical because extreme values are scarce. Thus, applying the r-largest order statistics (rLOS) instead of the block maxima is encouraged in general. The Gumbel distribution for rLOS (rGD) has been employed for modelin...

Full description

Saved in:
Bibliographic Details
Main Authors: Yire Shin, Jeong-Soo Park
Format: Article
Language:English
Published: Nature Portfolio 2025-03-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-024-83273-y
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850029776312467456
author Yire Shin
Jeong-Soo Park
author_facet Yire Shin
Jeong-Soo Park
author_sort Yire Shin
collection DOAJ
description Abstract The effective use of available information in extreme value analysis is critical because extreme values are scarce. Thus, applying the r-largest order statistics (rLOS) instead of the block maxima is encouraged in general. The Gumbel distribution for rLOS (rGD) has been employed for modeling the r-largest data. However, the rGD is not flexible enough to capture the variability of the r-largest data because only two parameters are used. This study extends the rGD to the generalized Gumbel distribution for rLOS (rGGD) to address this problem, which incorporates three parameters including a shape parameter. We derive some probability functions of the rGGD. The maximum likelihood estimation, delta method, entropy difference test for r selection, and cross-validated likelihood are considered for inference. The usefulness and practical effectiveness of the rGGD are illustrated by Monte Carlo simulation and an application to the peak streamflow data at Oykel Bridge in the UK. This new r-largest model should be helpful for the design of engineering structures to prevent severe damage by extreme events.
format Article
id doaj-art-947f459e3fa54d6a87ada835974a4591
institution DOAJ
issn 2045-2322
language English
publishDate 2025-03-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-947f459e3fa54d6a87ada835974a45912025-08-20T02:59:24ZengNature PortfolioScientific Reports2045-23222025-03-0115111610.1038/s41598-024-83273-yGeneralized Gumbel model for r-largest order statistics, with an application to peak streamflowYire Shin0Jeong-Soo Park1Department of Statistics, Chonnam National UniversityDepartment of Statistics, Chonnam National UniversityAbstract The effective use of available information in extreme value analysis is critical because extreme values are scarce. Thus, applying the r-largest order statistics (rLOS) instead of the block maxima is encouraged in general. The Gumbel distribution for rLOS (rGD) has been employed for modeling the r-largest data. However, the rGD is not flexible enough to capture the variability of the r-largest data because only two parameters are used. This study extends the rGD to the generalized Gumbel distribution for rLOS (rGGD) to address this problem, which incorporates three parameters including a shape parameter. We derive some probability functions of the rGGD. The maximum likelihood estimation, delta method, entropy difference test for r selection, and cross-validated likelihood are considered for inference. The usefulness and practical effectiveness of the rGGD are illustrated by Monte Carlo simulation and an application to the peak streamflow data at Oykel Bridge in the UK. This new r-largest model should be helpful for the design of engineering structures to prevent severe damage by extreme events.https://doi.org/10.1038/s41598-024-83273-yExperiments using unknown populationFlood frequency analysisFour-parameter kappa distributionReturn levelStructural design
spellingShingle Yire Shin
Jeong-Soo Park
Generalized Gumbel model for r-largest order statistics, with an application to peak streamflow
Scientific Reports
Experiments using unknown population
Flood frequency analysis
Four-parameter kappa distribution
Return level
Structural design
title Generalized Gumbel model for r-largest order statistics, with an application to peak streamflow
title_full Generalized Gumbel model for r-largest order statistics, with an application to peak streamflow
title_fullStr Generalized Gumbel model for r-largest order statistics, with an application to peak streamflow
title_full_unstemmed Generalized Gumbel model for r-largest order statistics, with an application to peak streamflow
title_short Generalized Gumbel model for r-largest order statistics, with an application to peak streamflow
title_sort generalized gumbel model for r largest order statistics with an application to peak streamflow
topic Experiments using unknown population
Flood frequency analysis
Four-parameter kappa distribution
Return level
Structural design
url https://doi.org/10.1038/s41598-024-83273-y
work_keys_str_mv AT yireshin generalizedgumbelmodelforrlargestorderstatisticswithanapplicationtopeakstreamflow
AT jeongsoopark generalizedgumbelmodelforrlargestorderstatisticswithanapplicationtopeakstreamflow