Noise reduction of monthly discharge time series using wavelet and chaos theory

Document Type : Original Article

Authors

1 Assistant Professor, Department of Civil Engineering, Faculty of Technical and Engineering of Marand, University of Tabriz, Tabriz. Iran

2 Department of water and Environmental Engineering, Faculty of Civil Engineering, University of Tabriz, Tabriz, Iran

Abstract

River streamflow predicting is necessary to optimal protection and management of water resources. Improving forecasting of river flow is many researcher's goals. The noise on hydrological data is present, which reduces the efficiency of Data-Driven Models, Therefore, in recent research, in addition to paying attention to model development, importance has also been given to identification and reducing of noise in the data used. For this purpose, monthly streamflow prediction of the two rivers (Ajichay and Sufichay) has been carried out using an ANN model in three different cases of input data to model, including: 1- raw data 2- noise-reduced data utilizing the chaos theory method, 3- noise-reduced data utilizing the wavelet.

To examine and compare the results, the evaluation criteria are calculated in three different cases of input data to ANN and the results of three cases are listed respectively. In Sufichay river, RMSE is 0.078, 0.045 and 0.034, R^2 is 0.846, 0.974 and 0.979, and the Nash-Sutcliffe value is 0.715, 0.948 and 0.958. In Ajichay river, RMSE values in these three cases were 0.126, 0.06, and 0.078, respectively, and R^2 were obtained 0.74, 0.94, and 0.91, and NSE were 0.545, 0.882, and 0.815. Therefore, it can be concluded that the using noise-reduced data cause to improve the accuracy of model prediction results. We observed that in Ajichay, the chaos method and in Sufichay, the wavelet method had better results, and one of these methods cannot be definitively introduced as the better method than the other for reducing data noise.

Keywords

Main Subjects

References

Attar, N.F., Sattari, M.T. & Apaydin, H. A. (2024). Novel stochastic tree model for daily streamflow prediction based on a noise suppression hybridization algorithm and efficient uncertainty quantification. Water Resources Management, 38, 1943–1964. https://doi.org/10.1007/s11269-023-03688-6.
Bahrami A.R. & Asil Gharebaghi, S. (2023). Improved noise reduction method for chaotic time series using neural network and singular spectrum analysis, Modares Mechanical Engineering, 24 (1), 53-63. https://doi.org/10.22034/MME.24.1.53.
Boustani, M., Farzin, S., Mousavi, S.F. & Karami, H. (2019). Effect of denoise reduction of time series on its analysis using Chaos theory (case study: Zayandehrud river), Eco Hydrology, 6(1), 15-27. https://doi.org/10.22059/ije.2018.260455.906.
Daneshvar Vousoughi, F. and Samadzadeh, R. (2021). Predicting runoff with pre-processing approaches in Ardabil plain. Journal of Hydrogeomorphology, 8(26), 116-99. doi: 10.22034/hyd.2021.44060.1570.
Davaie Markazi, A.H. & Nazarahari, M. (2015). Application of DWT for acoustic signal identification of ship using feature extraction methods and ensemble learning, Modares Mechanical Engineering, 15(8), 75-84. https://doi.org/20.1001.1.10275940.1394.15.8.10.0.
Guo, S., Wen, Y., Zhang, X. & Chen, H. (2023). Runoff prediction of lower Yellow river based on CEEMDAN–LSSVM–GM (1, 1) model. Scientific Reports, 13(1), 1511. https://doi.org/10.1038/s41598-023-28662-5.
Karunasingha, D.S.K. & Liong, S.Y. (2018). Enhancement of chaotic hydrological time series prediction with real-time noise reduction using Extended Kalman Filter, Journal of Hydrology, 565, 737-746. https://doi.org/10.1016/j.jhydrol.2018.08.044.
Kazemzadeh, M., Malekian, A., Moghaddamnia, A. R. & Khalighi Sigaroudi, Sh. (2017). Shift changes and heterogeneity analysis of hydro-climate variables (a case study: Aji Chai watershed), Eco Hydrology, 4(1), 163-175. https://doi.org/10.22059/ije.2017.60899.
Kocak K. Saylan L. & Sen O. 2000. Nonlinear Time Series Prediction of O3 Concentration in Istanbul. Atmosphere Environment, 34: 1267-1271. https://doi.org/10.1016/S1352-2310(99)00323-4.
Malekani, L. (2020). Reduced chaotic noise to improve the accuracy of estimates of monthly flow (case study: Nahandchai, Aharchai and Lighvanchai Rivers), Journal of Irrigation and Water Engineering, 11(1), 89-103. https://doi.org/10.22125/iwe.2020.114955.
Meng, J., Wang, Y., Guo, H., & Ding, Y. (2023). Application of Wavelet Denoising Algorithm in Monthly Runoff Series of Fuchun River Hydropower Station. International Seminar on Computer Science and Engineering Technology (SCSET) 719-723. https://doi.org/10.1109/SCSET58950.2023.00162.
Nourani, V., Andalib, G., & Sadikoglu, F. (2017). Multi-station streamflow forecasting using wavelet denoising and artificial intelligence models. Procedia Computer Science, 120, 617-624. ‏
Partovyan, A., Nourani, V. & Alami, M.T. (2018). Noise injection-denoising techniques to improve artificial intelligence-based rainfall runoff modelling, Water Resource Engineering, 11(36), 81-94. https://doi.org/20.1001.1.20086377.1397.11.36.8.9.
Rezaei, H. & Jabbari Gharabagh, S. (2017). Noise reduction effect on Chaotic analysis of Naziuchay river flow, Water and Soil Science, 27(3), 239-250. https://doi.org/10.22098/mmws.2021.9431.1043.
Schreiber, T. & Kantz. H. (1998). Nonlinear projective filtering II: Application to real time series, arXiv preprint chao-dyn/9805025.  https://doi.org/10.48550/arXiv.chao-dyn/9805025.
Wang, Y. Y., Wang, W. C., Xu, D. M., Zhao, Y. W., & Zang, H. F. (2024). A compound approach for ten-day runoff prediction by coupling wavelet denoising, attention mechanism, and LSTM based on GPU parallel acceleration technology. Earth Science Informatics, 17(2), 1281-1299. https://doi.org/10.1007/s12145-023-01212-3.
Xiao, H., Hu, D. & Wang, J. (2022). Threshold selection of wavelet denoising based on optimization algorithms, International Conference on Innovations and Development of Information Technologies and Robotics (IDITR), Chengdu, China, 2022, pp. 88-92. https://doi.org/10.1109/IDITR54676.2022.9796485.
Yahyavi Rahimi, A. (2013). Using the threshod method to obtain error-free input in runoff sediment modelling using the artificial neural network model method, MSc in Water Engineering, Faculty of civil Engineering, University of Tabriz.
Yang, Y., Li, W., & Liu, D. (2024). Monthly Runoff Prediction for Xijiang River via Gated Recurrent Unit, Discrete Wavelet Transform, and Variational Modal Decomposition. Water, 16(11), 1552. https://doi.org/10.3390/w16111552.
Zerouali, B., Al-Ansari, N., Chettih, M., Mohamed, M., Abda, Z., Santos, C. A. G., & Elbeltagi, A. (2021). An enhanced innovative triangular trend analysis of rainfall based on a spectral approach. Water, 13(5), 727.  https://doi.org/10.3390/w13050727.

Files

History

  • Receive Date: 19 February 2025
  • Revise Date: 10 August 2025
  • Accept Date: 16 August 2025

Share

How to cite

Statistics