Title: Rainfall frequency analysis using extreme values distributions in the steppe region of Western Algeria

Authors: Abdelkader Boucefiane; Mohamed Meddi; Jean Pièrre Laborde; Saeid Eslamian

Addresses: University of Sciences and Technology USTO, MB – Bp 1505, 31000, Oran, Algeria; LGEE – Higher National School of Hydraulic Engineering, BP 31, 09000 Blida, Algeria ' LGEE – Higher National School of Hydraulic Engineering, BP 31, 09000 Blida, Algeria ' Hydroinformatics and Water Engineering Department, Polytech'Nice-Sophia School, Nice Sophia Antipolis University, 930 Road of Colles – BP 145 06903, Sophia Antipolis Cedex, France ' Water Engineering Department, Isfahan University of Technology, Isfahan, 8415683111, Iran

Abstract: The aim of this paper is to study the possibility to use the generalised extreme value (GEV) distribution by adjusting the three parameters for 65 rainfall stations located in steppe region of western Algeria. Analysis of maximum daily rainfall with Gumbel model, which traditionally used shows that it is not perfectly suited to steppe region of western Algeria. In fact, the small hydrological sample sizes generally available affect remarkably the estimation of parameters and leads to underestimate the frequency of that value. Therefore, it opts for the hypothesis that maximum daily rainfalls in the study region are distributed according to GEV distributions but k is a regional constant to be determined. A value of k = -0.053 is an acceptable hypothesis in relation to the uncertainties on the local estimates. This value of k allows also in finding an observations number in agreement with the estimated return periods.

Keywords: maximum daily rainfall; rainfall frequency analysis; generalised extreme values; GEVs; Gumbel; Algeria; steppe regions.

DOI: 10.1504/IJHST.2014.068451

International Journal of Hydrology Science and Technology, 2014 Vol.4 No.4, pp.348 - 367

Received: 26 Jul 2014
Accepted: 03 Nov 2014
Published online: 15 Apr 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article