Research Information System
Environmental Earth Sciences, vol.84, no.1, 2025 (SCI-Expanded)
Drought is a significant natural disaster with adverse effects on both social and ecological systems. Unlike other natural disasters, drought develops slowly and gradually, complicating its early detection and often resulting in severe impacts on affected regions. Consequently, accurate and dependable drought monitoring is essential for devising effective mitigation strategies. Standardized drought indices are vital tools in drought monitoring, providing a means to quantify and characterize drought events. Most standardized drought indices utilize the Standardized Precipitation Index (SPI) method, which is valued for its simplicity and flexibility. However, this study contends that the SPI method lacks several critical elements, particularly in practice, such as determining the most suitable probability distribution for hydrometeorological variables. Therefore, this study proposes a novel methodology for calculating standardized drought indices and assesses its performance against conventional and nonparametric standardized indices, employing various methods capable of capturing complex dependencies. The novel methodology involves identifying the best-fit probability distributions for each data group through various goodness-of-fit tests. This approach ensures that each group is modeled optimally, considering the seasonal variations inherent to each group. The Seyhan River Basin has been chosen as a case study for the proposed methodology. The drought characteristics of the basin are analyzed using indices derived from the new methodology, the conventional SPI method, and the nonparametric method. Additionally, trend analyses were performed on the calculated indices to identify any directional changes in drought patterns within the Seyhan River Basin. The performance of the proposed methodology was evaluated by analyzing its relationship with nonparametric standardized indices and comparing it to the relationship between conventional standardized indices and nonparametric standardized indices. The results show that the newly proposed methodology outperforms the conventional SPI method across various dependence measures, suggesting it captures the underlying data structure more effectively than the SPI method.