Uncertainty Quantification in Extreme Flood Prediction: Geomorphic Analysis of Statistical Parameters in Arid Basins (Case Study: Qareh Aghaj River Basin

Zeinli, Hamid; Elmizadeh, Heeva

doi:10.22111/jneh.2025.52184.2117

Uncertainty Quantification in Extreme Flood Prediction: Geomorphic Analysis of Statistical Parameters in Arid Basins (Case Study: Qareh Aghaj River Basin

Articles in Press

Document Type : Research Article

Authors

Hamid Zeinli ¹
Heeva Elmizadeh ²

¹ MS. in Geomorphology, Khorramshahr University of Marine Science and Technology, Khorramshahr, Iran

² Associate Prof., Department of Marine Geology, Faculty of Marine Natural Resources, Khorramshahr University of Marine Science and Technology, Khorramshahr, Iran

10.22111/jneh.2025.52184.2117

Abstract

The Qarah-Aghaj basin, as one of the most significant watersheds in Fars Province, has experienced numerous destructive flood events in recent decades, causing extensive damage to regional infrastructure. This study presents a comprehensive analysis of flood behavior in the Qarah-Aghaj basin through an integrated approach that evaluates four statistical distributions: Gumbel Type I, Generalized Extreme Value (GEV), Generalized Pareto (GPD), and log-normal. The distribution parameters were fitted using three distinct methods: Method of Moments (MOM), L-moments, and Maximum Likelihood Estimation (MLE). The goodness-of-fit was rigorously assessed through multiple criteria, including the Kolmogorov-Smirnov test, Akaike Information Criterion (AIC), and Root Mean Square Error (RMSE). Key findings demonstrate the superior performance of the GEV distribution in flood modeling for this basin. The GEV distribution showed optimal agreement with observational data, as evidenced by the lowest Kolmogorov-Smirnov statistic (0.058) and AIC value (234.5). Furthermore, it provided the most accurate predictions with an RMSE of 112.3 m³/s and a positive bias of 2.1%. These results indicate heavy-tailed behavior in the flood frequency distribution, primarily attributable to unique geomorphological characteristics, including steep slopes (mean 12.5%), high drainage density (1.8 km/km²), and extensive karst formations (covering approximately 60% of the basin area). The analysis underscores the necessity of employing advanced distributions like GEV for complex hydrological systems in arid basins such as Qarah-Aghaj, potentially reducing computational errors by up to 20%. The study conclusively establishes that a single distribution cannot be universally applied across all arid basins, emphasizing the critical importance of considering basin-specific characteristics in flood frequency analysis. These findings provide a scientific basis for updating flood management standards in similar regions, with significant implications.

Keywords

Main Subjects

Environmental Hazards

References

References [in Persian]

Jabbari, I., Ghobadian, R., & Jadidi, A. (2023). Impact of April 2019 Flood on Morphology of the Confluence Zone of Dinavar and Gamasiab Rivers Using SRH-2D. Numerical Model. Geography and Development, 21(70), 1-26. doi: 10.22111/gdij.2023.7401. [In Persian]

Zeinli, H. (2023). Geomorphological Analysis of Flood Effects in Tang Karzin Basin (Master's thesis in Hydrogeomorphology). Khorramshahr University of Marine Science and Technology, 1-107. [In Persian]

References [in English]

Bashirgonbad, M., Moghaddam Nia, A., Khalighi-Sigaroodi, S. (2024). A hydro-climatic approach for extreme flood estimation in mountainous catchments. Applied Water Science, 14, 98. https://doi.org/10.1007/s13201-024-02149-8

Cui, L., Hu, G., Zhu, Y. (2025). Multi-strategy improved snow ablation optimizer: a case study of optimization of kernel extreme learning machine for flood prediction. Artificial Intelligence Review, 58, 181. https://doi.org/10.1007/s10462-025-11192-z

Das, P., Posch, A., Barber, N. (2024). Hybrid physics-AI outperforms numerical weather prediction for extreme precipitation nowcasting. npj Climate and Atmospheric Science, 7, 282. https://doi.org/10.1038/s41612-024-00834-8

Ghobadi, M., Ahmadipari, M. (2024). Enhancing Flood Susceptibility Modeling: a Hybrid Deep Neural Network with Statistical Learning Algorithms for Predicting Flood Prone Areas. Water Resources Management, 38, 2687-2710. https://doi.org/10.1007/s11269-024-03770-7

Huang, J., Hong, Y., Sun, D. (2025). Urban flood depth prediction using an improved LSTM model incorporating precipitation forecasting. Natural Hazards, 121, 8305-8326. https://doi.org/10.1007/s11069-024-07065-3.