Authors: Nur Amalina Mat Jan; Ani Shabri; Jean Hounkpè; Basri Badyalina
Addresses: Department of Mathematics, Faculty of Science, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia ' Department of Mathematics, Faculty of Science, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia ' Laboratory of Applied Hydrology, National Water Institute, University of Abomey-Calavi, Abomey-Calavi BP 2008, Benin ' Department of Mathematics, Faculty of Science, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia
Abstract: Global change has raised concerns among hydrologists about the use of the stationary assumption (independent and identically distributed flood series) in infrastructure-designed methods. This confirms the necessity of evaluating the stationary or non-stationary behaviour of hydrological variables before deriving flood plans for infrastructure projects and flood mitigation. Trends were evaluated in the annual maximal streamflow of 49 stations in Peninsular Malaysia, using the Mann Kendall and Spearman Rho trend tests. Three models, a stationary model (GEV0) and two non-stationary models, with: 1) location parameter; 2) location and log-transformed scale parameters as a linear function of time (GEV2), were considered for stations with significant trend. It was found that a quarter of the analysed stations show statistically significant trends in their annual maximal streamflow. These results indicate the importance of taking into consideration the non-stationary behaviour of the flood series in order to improve the quality of flood estimation.
Keywords: stationary; non-stationary; streamflow; flood frequency analysis; FFA; generalised extreme value distribution; Mann-Kendall test; Spearman's rho trend test; Malaysia.
International Journal of Water, 2018 Vol.12 No.2, pp.116 - 140
Received: 03 Oct 2016
Accepted: 28 Jan 2017
Published online: 16 Apr 2018 *