Performance evaluation of signal decomposition methods in monthly precipitation estimation (case study: Telezang station)

Articles in Press

Document Type : Research Paper

Authors

1 Faculty of Water and Environmental Engineering, Shahid Chamran University of Ahvaz, Ahvaz, Iran.

2 Professor, Faculty of Water Engineering, Shahid Chamran University of Ahvaz

10.22059/jwim.2025.396662.1239

Abstract

Precipitation forecasting is of great importance due to its impact on agriculture, natural disaster management, and water supply. Therefore, in this study, the monthly precipitation of the Telezang station from 1966 to 2020 was modeled using the Variable Mode Decomposition (VMD) and Complete Ensemble Empirical Mode Decomposition (CEEMD) methods. Input data were defined for the Support Vector Machine (SVM) model based on four scenarios. In the first scenario, monthly precipitation values with up to four lags were considered as model inputs. In the second scenario, in addition to the lagged precipitation data, a periodic term was added to the input patterns of the model. In the third and fourth scenarios, the monthly precipitation data were decomposed using CEEMD and VMD, respectively, and provided to the model. The findings of this research indicated that adding the periodic term slightly improved the model’s performance. Additionally, a comparison of the results from the data preprocessing methods using VMD and CEEMD showed that the VMD-SVM model outperformed the CEEMD-SVM model significantly, reducing the MAE index by an average of approximately 35.25 mm compared to the standalone model and 13.77 mm compared to the CEEMD-SVM model, while also achieving greater accuracy.

Keywords

Main Subjects

Water and Irrigation Management

Articles in Press, Accepted Manuscript
Available Online from 04 September 2025

Files

Share

How to cite

Statistics