In Press (Summer)
Back to the articles list |
Back to browse issues page
Department of Reclamation of Arid and Mountainous regions Engineering, Faculty of Natural Resources, University of Tehran, Karaj, Iran. , ansari.ghojghar@ut.ac.ir
Abstract: (17 Views)
Accurate understanding of the temporal and spatial variations of destructive climatic phenomena is a fundamental step toward integrated management of these hazards. In recent years, the use of dual hybrid models for simulating dust storms has attracted significant attention. This study models the frequency index of dust storm days in Khuzestan province over a 50-year period (1971–2020). The performance of the triple hybrid model GRNN–AF–ARMA in predicting this index was compared with single models GRNN and ARMA, as well as dual hybrid models GRNN–AF, ARMA–AF, and GRNN–ARMA. The results demonstrated the superior performance of the triple hybrid model. Subsequently, to enhance prediction accuracy and stability, the Ensemble Kalman Filter (EnKF) was employed to correct and optimize the outputs of the hybrid model. The findings indicated that integrating this filter with the triple hybrid model significantly improved the correlation coefficient and the Nash–Sutcliffe efficiency. Therefore, the proposed model exhibits high capability in predicting the temporal and spatial patterns of dust storms in Khuzestan province. The outcomes of this research can serve as an effective tool for management planning and the implementation of preventive measures to mitigate damages caused by dust storms.
Keywords: Artificial Neural Network, Box-Jenkins Method, FDSD Index, Artificial Plants Algorithm, Modeling.
Type of Study: Research |
Received: 2026/04/21
Received: 2026/04/21
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
