Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Turquie → Critical drought severity-duration-frequency curves based on precipitation deficit

İstanbul Teknik Üniversitesi (2019)

Critical drought severity-duration-frequency curves based on precipitation deficit


Titre : Critical drought severity-duration-frequency curves based on precipitation deficit

Yağış açığı cinsinden kritik kuraklık şiddet-süre-frekans eğrileri

Auteur : ÇAVUŞ Yonca

Université de soutenance : İstanbul Teknik Üniversitesi

Grade : Master of Science (MS) 2019

Drought is one of the most disastrous natural phenomena that causes scarcity and lack of water in hydrological basins such as the Seyhan River Basin in Turkey. The precipitation is the merely source of the natural water resources in the country which has a semi-arid climate with a dry and hot summer. It changes also depending on seasonality within the year. Because of the decrease in precipitation and increase in evapotranspiration, water resources have dropped remarkably and it is certain that the precipitation deficit is a more difficult problem to overcome. Accordingly, it is of significance to determine how much the precipitation deficit is in terms of water supply in this region. The main purpose of this study is to define as the drought severity/intensity-duration-frequency curves based on precipitation deficit. In the study, drought analysis was performed by using the Standardized Precipitation Index (SPI) for 19 meteorological stations in the Seyhan River basin. In order to find SPI values of each station, the monthly precipitation data were used. An executable file developed by the National Drought Mitigation Center, University of Nebraska–Lincoln, USA was used to calculate the SPIk (k = 1, 3, 6, 9, 12, 24 months) values. After the SPIk values were obtained, dry and wet periods were determined from the SPIk time series for each time scale. Drought characteristics are determined from the SPI time series. Droughts are characterized by severity, duration and frequency. Also, time scale is important when SPI is used for the drought. Severity can be replaced by intensity, the average severity over the length of the dry period. Drought duration from D = 1 month (at minimum) up to the longest duration (48 months at maximum when exists) were considered. Severity values of SPIk series were determined in dry periods. For each year, one or higher number of droughts are likely to be observed. Drought with the highest severity in each year is defined as the critical drought of the year. When more than one drought is observed within a year, no matter how long the drought period in a year is, it is assigned as the critical drought of this particular year. When no drought is observed in a year, the critical drought severity is not calculated and a zero value is assigned to the critical drought severity of this particular year. Frequency analysis was applied on the critical severity time series to determine the best-fit probability distribution function among followings commonly used in the literature : General Extreme Value (GEV), Log-normal 2 (LN2), Log-normal 3 (LN3), Gamma 2 (G2), Gamma 3 (G3), Log-Pearson type III (LP3), Weibull 2 (W2), Weibull 3 (W3). Since the critical severity values include zeros, the total probability theorem was applied after the frequency analysis. The best-fit probability distribution functions were determined and their parameters were calculated for each D-month drought of the SPIk series. For the frequency analysis, critical severity series with at least 10 years at length were considered ; i.e., no frequency analysis was applied on a critical severity time series shorter than 10 years. The General Extreme Value (GEV) distribution was found as the best-fit distribution to the critical severities almost for all drought durations in the meteorological stations in Seyhan River basin considering SPIk for time scales k = 1, 3, 6, 9, 12, 24 and 48 months. The goodness-of-fit of the probability distribution function was checked with the Anderson Darling statistical test. The critical drought severities corresponding to different return periods were calculated with the frequency analysis. As the drought duration increases, the critical drought severity increases for each of the k-time scales. Accordingly, the severity-duration-frequency curve was obtained. Also, the intensity values were calculated as the ratio of the drought severity to its duration. On the other hand, the relationship between the precipitation and corresponding SPI values was examined with regression analysis. A logistic function was found suitable to use in the regression.


Version intégrale

Page publiée le 9 janvier 2023