Assessing Flood Risk through Frequency Analysis at Sayong River Station

Abstract

This research investigates extreme flood events at Sayong River Station using two statistical distributions: the Generalized Logistic (GLO) and Generalized Extreme Value (GEV) distributions. The study employs the L-moment method for parameter estimation and evaluates quantile estimates for return periods of 10, 50, and 100 years. The primary objective is to determine which distribution provides a more accurate representation of extreme flood behavior. Quantile estimates derived from the GLO distribution are 197.3153 m³/s for the 10-year return period, 363.8308 m³/s for the 50-year return period, and 469.9711 m³/s for the 100-year return period show better alignment with observed data compared to the GEV distribution. The GLO distribution's superior performance indicates its enhanced ability to capture the tail behavior of extreme floods, providing more reliable estimates for flood risk management. The findings emphasize the importance of selecting an appropriate distribution model for accurate flood risk assessment. The efficacy of the GLO distribution in representing extreme values underscores its appropriateness for forecasting extreme flood magnitudes and providing guidance for infrastructure design and flood mitigation measures.