Prediksi Kekeringan Ekstrem di Jawa Timur Menggunakan Generalized Pareto Distribution (GPD)

Abstract

Extreme droughtin East Java is an event that has a significan timpact on the agricultural sector, water availability, and socio-economic conditions of the community. This study aims to analyze the characteristics of extreme drought using the Standardized Precipitation Index (SPI) and predict extreme rainfall events using the Generali zed Pareto Distribution (GPD) based on the return level. The data used is monthly rainfall data for the period 1990-2024 from BMKG observation stations in Malang Regency, East Java. SPI analysis was conducted to identify drought patterns based on rainfall in the dry season period (April-September). The results of the analysis show that extreme drought occurs more frequently at the end of the dry season (July-September) than at the beginning of the dry season (April-June), with some years experiencing SPI below-1.5, indicating extreme drought conditions. Extreme data were identified using the Peaks Over Threshold (POT) method with a threshold of 2.00 mm. The parameters of the GPD distribution were estima ted using Maximum Likelihood Estimation (MLE), with the results of shape ξ = 1.297 and scale σ = 2.983. The Kolmogorov-Smirnov test showed that the extreme data fit the GPD distribution (p-value = 0.301). The prediction results show that in the next 1-10 years, extreme rainfall tends to increase as the time period increases, so the risk of extreme drought decreases. Prediction accuracy may decrease in longer periods due to uncertainty in the estimation of the return level value. The results of this study provide insight into the pattern of extreme drought in East Java and the potential for future occurrence. The findings can serve as a reference in drought mitigation planning and water resources managementtoreduceits negative impacts.

Keywords: Extreme drought, Standardized Precipitation Index (SPI), Generalized Pareto Distribution (GPD)