Akademik Veri Yönetim Sistemi
9TH INTERNATIONAL CONFERENCE ONGLOBAL PRACTICE OF MULTIDISCIPLINARY SCIENTIFIC STUDIES, La Habana, Küba, 21 - 28 Ocak 2025, ss.197, (Özet Bildiri)
In recent years, many studies have been conducted in the literature using various drought indices to monitor and evaluate drought and drought characteristics. The most wellknown of these is the Standardized Precipitation Index (SPI). In SPI calculations, it is seen that the gamma probability distribution function (PDF) is often used as a reference in modeling data sets determined according to selected time scales without considering the suitability tests or only for the raw data set scale where measurements are made. Gamma cannot be the most appropriate PDF for the original data sets frequently created in arid and semiarid areas and occasionally in wetlands. Despite this, it is seen that in scientific articles, projects, and applications produced by researchers, and in free software developed using Python and R software, the gamma PDF is generally used without any suitability check, and rarely the most appropriate probability distribution models are considered. The climate of the study area, precipitation characteristics, and, in short, physical and regional conditions are effective in selecting the appropriate model. Therefore, in studies conducted on drought indices, it is critical to test different distributions for all data sets and select the most appropriate models. Within the scope of this research, the reliability of the results obtained by comparing the drought analysis results of Gamma and other PDFs like Weibull and normal functions, whose suitability was checked with the goodness of fit tests, such as the Kolmogorov-Smirnov statistical test according to 1, 3 and 12-month timescales for the observation data within the Kocaeli province and presented directly in the literature, will be evaluated.