International Journal of Analysis and Applications | 2026
Authors: Deetae N.; Khamrot P.; Gaketem T.
DOI: 10.28924/2291-8639-24-2026-4
Journal: International Journal of Analysis and Applications
Year: 2026
Publisher: Etamaths Publishing
Document Type: Article
Open Access: All Open Access; Gold Open Access
Cited by: 0
This study focuses on the statistical modeling of extreme water volumes at the Kwae Noi Bamrung Daen Dam in Phitsanulok Province using extreme value theory (EVT). The objective is to predict high water levels that may pose risks to dam safety and operations. Historical monthly water volume data from 2011 to 2023 were analyzed using the generalized extreme value (GEV) distribution. The Jarque-Bera test confirmed a non-normal distribution (p = 0.02906), justifying the use of EVT. Maximum likelihood estimation yielded parameter estimates of µ = 446.58, σ = 228.68, and ξ =-0.13. Goodness-of-fit tests (K-S and A-D) confirmed the adequacy of the GEV model, with p-values of 0.3559 and 0.1124, respectively. The model estimated that extreme water volumes exceeding 950 million cubic meters are expected approximately once every 25 years. These findings contribute to more accurate hydrological forecasting, improved early warning systems, and enhanced water resource management policies. The study supports risk-informed decision-making in flood-prone regions and advances the application of EVT in dam safety and environmental planning. © 2026 the author(s).
extreme value theory; generalized extreme value distribution; KM transformation; maximum likelihood estimation; return level