Uncertainties in Estimation of a Possible Upper Limit for Environmental Extremes

Sammanfattning

In statistical risk analysis, estimation of return levels is often of importance. In particular, the uncertainties of the estimates are crucial. The longer the return period, the more uncertain is the estimate of the return level, or phrased in terms of statistical modeling, the associated confidence intervals become wider. In this note, we examine closer the relationship between the return level and the upper limit occurring for the case of a negative shape parameter in a fitted generalized extreme value distribution. Examples based on sea level data close to Swedish nuclear power plants are given, as well as results from simulation studies. The findings may be of interest for planners and engineers that analyze extreme-value data for risk and safety analysis.

Nyckelord

estimation; international ocean; method; illustration; proceedings; extreme; distribution; artificial intelligence; thirty-fourth; ringhal

Publicerad i

Titel: Proceedings of the 34th International Ocean and Polar Engineering Conference
Utgivare: ISOPE

Konferens

The 34th International Ocean and Polar Engineering Conference (ISOPE), 2024, June 16-21, Rhodes, Greece

SLU författare

• Rydén, Jesper

  • Institutionen för energi och teknik, Sveriges lantbruksuniversitet
    

UKÄ forskningsämne

Sannolikhetsteori och statistik


Permanent länk till denna sida (URI)

https://res.slu.se/id/publ/131764