Regional Frequency Analysis of Drought Indexes by L-Moments Method

Abstract

Drought has been experienced frequently in Turkey in the last two decades as the effects of withdrawn water resources become more evident. Drought; causes problems for planners and managers. For this reason, in order to provide an accurate framework for sustainable water resources management, it is necessary to investigate the characteristics of drought events and to estimate the return periods of the drought with the help the regional frequency analysis. In this study, firstly, the characteristics of meteorological droughts in the Kızılırmak Basin which has semi-arid climate characteristics, were determined by using the Standardized Precipitation Index (SPI), Z-Score Index (ZSI), China-Z Index (CZI) and Modified China-Z Index (MCZI) as a measure of drought severity and also the applicability and performance of the selected indices to the basin were investigated. Secondly, regional frequency analysis was performed by using L-moment methods for the maximum drought severity values obtained for each year as a result of the application of the four drought indices on a 12-month time scale. According to the results of the meteorological drought analysis of the four drought indices, it was noted that the most severe and long-lasting droughts occurred mainly in the 2000s, the drought severity values increased as the return period increased and also the drought severity values obtained by MCZI method give the extraordinary results than other indices. Thus, it has been seen that the MCZI method is generally not suitable for use in the basin. According to the results of regional frequency analysis, Kızılırmak Basin, which is H1<1, was determined as acceptable homogeneous for all four indices according to the Hosking and Wallis homogeneity test. For each index, the optimum regional distribution function was investigated and Pearson type‒III distribution for SPI and ZSI; distribution of general extreme values for CZI; for MCZI, the generalized logistic distribution was determined as the most appropriate distribution. As a result of the index-flood frequency analysis calculated by using the most appropriate distribution, regional drought severity maps were created for the study area with the Inverse Distance Weighting (IDW) Method for the return periods between 5 years and 1000 years. Using the maps obtained, it is feasible to predict the drought probability of any point in the basin that does not have adequate data for hydrological investigations.