Revue d'économie et de statistique appliquée
Volume 18, Numéro 1, Pages 45-55
2021-06-29

Modeling Average Maximum Monthly Temperature Using The Generalized Pareto Distribution At The Jijel Weather Station From 1988 To 2018

Authors : Cheraitia Hassen .

Abstract

The objective of this article is to estimate the two parameters of the generalized Pareto distribution (GPD) that allows us to calculate return levels of temperature for different return periods. we worked on a monthly average temperatures series of the Jijel weather station (Northern Algeria), after the determination of the threshold (u_0 = 18 °C), we estimated the parameters of the GPD by the maximum likelihood method, it was found that the generalized Pareto distribution of type II (with bounded tails) is more suitable for the maximum monthly temperatures. Once the estimated model has been validated we calculated returns levels for different periods, according to this model. We must wait about 100 years to record an average monthly temperature of 28.56 °C.

Keywords

generalized Pareto distribution ; maximum likelihood estimate ; average monthly temperature ; return level