İpsala su ölçüm istasyonunda gelecekteki su debisi tahminleri için makine öğrenmesinin uygulanması

Öz

Son zamanlarda, akış parametrelerinin tahmin edilmesi önemli ölçüde artmıştır; bunlardan birisi de su akış hızının tahminidir. Bu fiziksel parametrenin bilinmeyen gelecekteki değerinin tahmini, bazı konularda, insanlara yol göstermesi açısından çok önemlidir. Örneğin, bir nehrin su akış tahmini söz konusu olduğunda, ani sellere karşı müdahale açısından, su akış hızının gelecekteki değerlerinin tahmini çok hayatidir. Bu çalışmada, Türkiye'de akan Ergene Nehri'ne ait ölçülen deneysel verilerin bir sonraki gözlemlenen su akış hızı değerinin tahmininde makine öğrenmesi yöntemlerini kullanan birkaç algoritma önerilmiştir. Çalışma alanı olarak bu nehir üzerinde kurulu Ipsala su ölçüm istasyonu seçilmiştir. Çalışmada kullanılan algoritmalar; toplamda 93 farklı model oluşturmuştur. Ergene Nehri'nin su akış hızı tahminleri, bulanık c-ortalamalar ve uzun kısa süreli bellek algoritmalarının istatistiksel doğruluk ölçütleri açısından üstün sonuçlar ürettiğini göstermiştir. Özellikle; uyarlanabilir nöro bulanık çıkarım sistemine ait bulanık c-ortalamaları yaklaşımının, su akış hızının gelecekteki değer tahmininde çok doğru ve etkili bir yöntem olduğu sonucuna varılmıştır; bu yöntem, istatistiksel eşleştirilmiş t-testi ve farklı tahliye seviyesi senaryoları ile de yenilik olarak doğrulanmıştır.

Anahtar Kelimeler

• Uyarlanabilir nöro-bulanık çıkarım sistemi
• Veri tahmini
• Uzun kısa süreli bellek
• Nehir su akış debisi

Application of machine learning for future water flow rate estimations on Ipsala water metering station

Öz

Latterly, forecasting of flow parameters has significantly increased; one of those is estimation of water flow rate. Prediction of unknown future value of this physical parameter is very important in terms of guiding people on some issues. For example, when water flow estimation of a river is considered, estimation of future values of water flow rate is momentous in terms of a response to flash floods. A couple of algorithms that use machine learning methods have been proposed in this study, in forecast of one-ahead observed value of water flow rate of measured experimental data belonging to Ergene river, flowing in Türkiye. Ipsala water metering station established on this river was chosen as study area. Algorithms used in the study construct a total of 93 different models. Water flow rate anticipations of Ergene river have indicated that fuzzy c-means and long short-term memory had generated superior results in terms of statistical accuracy measures. It has been concluded that especially fuzzy c-means of adaptive neuro fuzzy inference system approach is very accurate and effective method in future value prediction of water flow rate, which is also validated as novelty by statistical paired t-test and different discharge level scenarios.

Anahtar Kelimeler

• Adaptive neuro-fuzzy inference system
• Data prediction
• Long short-term memory
• River water flow rate

Kaynakça

1. N-Sci Technologies 2024, The importance of hydropower. https://nsci.ca/2019/08/14/the-importance-of-hydro-power/, Accessed 24 December 2025.
2. Renewable Energy Policy Network for the 21st Century 2023, Global status report. https://www.ren21.net/, Accessed 24 December 2025.
3. A. Ilhan, Forecasting of river water flow rate with machine learning. Neural Computing and Applications, 34, 20341–20363, 2022. https://doi.org/10.10 07/s00521-022-07576-9.
4. K. Benmouiza and A. Cheknane, Clustered ANFIS network using fuzzy c-means, subtractive clustering, and grid partitioning for hourly solar radiation forecasting. Theoretical and Applied Climatology, 137, 31–43, 2019. https://doi.org/10.1007/s00704-018-2576-4.
5. A. Sahraei, A. Chamorro, P. Kraft and L. Breuer, Application of machine learning models to predict maximum event water fractions in streamflow. Frontiers in Water, 3, 652100, 2021. https://doi.org/10.3389/frwa.2021.652100.
6. L. Xiao, M. Zhong and D. Zha, Runoff forecasting using machine-learning methods: Case study in the middle reaches of Xijiang river. Frontiers in Big Data, 4, 752406, 2022. https://doi.org/10.3389/fdata.20 21.752406.
7. Y. Zhou, Y. Wang, Y. Zhang and W. Wan, Interpretable machine learning for predicting and optimizing pressure extremes in pipeline water hammer effects based on the method of characteristics. Water Resources Management, 39, 4679–4706, 2025. https://doi.org/10.1007/s11269-025-04175-w.
8. K. Roushangar and F. Homayounfar, Prediction characteristics of free and submerged hydraulic jumps on horizontal and sloping beds using SVM method. KSCE Journal of Civil Engineering, 23(11), 4696–4709, 2019. https://doi.org/10.1007/s12205-019-1070-6.

9. M. Heidarnejad, J. Feili, M. Fuladipanah and Upaka Rathnayake, Machine learning-based discharge coefficent estimation in trapezoidal-arched labyrinth weirs. Asian Journal of Water, Environment and Pollution, 22(6), 73–88, 2025. https://doi.org/10.36922/AJWEP025120081.
10. Z. Li, T. Mu, X. Li, P. Li, J. Feng, H. Xu, C. Liu and S. Qian, Physics-informed neural network for hydraulic prediction in open-channel water transfer projects with sparse monitoring data. Water Research, 287, 124507, 2025. https://doi.org/10.1016/j.watres.2025.124507.
11. R. Dong, J. Du and C. Liu, Deep recurrent neural networks for water hammer transient prediction and dynamic protection optimization in long distance pipelines. Scientific Reports, 16, 8687, 2026. https://doi.org/10.1038/s41598-026-41915-3.
12. M. A. Moustafa, A. Amin, Z. A. Bello, K. A. M. Ali, H. A. A. Sayed, Y. K. Osman, S. Hassan, M. Aurangzaib and M. H. Fayed, Enhanced deep learning model for predicting hydraulic performance in recycled porous pipe irrigation systems. Scientific Reports, 15, 34286, 2025. https://doi.org/10.1038/s41598-025-20354-6.
13. A. J. G. Orduna, O. E. C. Hernandez, A. A. Pastrana, H. M. Ramos and M. P. Sanchez, Physics-informed machine learning for hybrid modelling of water leakages induced by rapid valve manoeuvres in water distribution networks. Engineering Applications of Artificial Intelligence, 167, 113887, 2026. https://doi.org/10.1016/j.engappai.2026.113887.
14. T. Tugrul, Forecasting future droughts with machine learning. Niğde Ömer Halisdemir University Journal of Engineering Sciences, 15, 1819926, 2026. https://doi.org/10.28948/ngumuh.1819926.
15. M. S. Dogan, Simulating hydropower reservoir operations of the Yamula Dam with machine learning. Niğde Ömer Halisdemir University Journal of Engineering Sciences, 12(4), 1426-1435, 2023. https://doi.org/10.28948/ngumuh.1314793.
16. M. N. Ciner, M. Guler, E. Namlı, M. Samastı, M. Ulu, I. B. Peker and S. Gulbaz, the forecast of streamflow through Göksu stream using machine learning and statistical methods. Water, 16, 1125, 2024. https://doi.org/10.3390/w16081125.
17. H. Aktas and E. Polat, Automatic detection and classification systems based on YOLOv11 model for walnut processing lines. Niğde Ömer Halisdemir University Journal of Engineering Sciences, 14(4), 1638–1646, 2025. https://doi.org/10.28948/ngumuh.1723155.
18. F. Uysal, Comparative analysis of various machine learning and deep learning approaches for car resale price prediction in the Turkish market. Niğde Ömer Halisdemir University Journal of Engineering Sciences, 13(1), 342–349, 2024. https://doi.org/10.28948/ngumuh.1353526.
19. T. Ozkul and A. Topalli, Comparison of machine learning methods for limited predictive maintenance. Niğde Ömer Halisdemir University Journal of Engineering Sciences, 14(1), 183–191, 2025. https://doi.org/10.28948/ngumuh.1465282.
20. S. Zeybek, B. Alkın and Y. Kaya, Crime detection on social media data using deep learning and machine learning methods. Niğde Ömer Halisdemir University Journal of Engineering Sciences, 14(1), 175–182, 2025. https://doi.org/10.28948/ngumuh.1551734.
21. F. Koylu and Y. Atılkan, Overtaking with deep reinforcement learning methods in autonomous vehicles. Niğde Ömer Halisdemir University Journal of Engineering Sciences, 13(2), 429–439, 2024. https://doi.org/10.28948/ngumuh.1331354.
22. B. H. Yilmaz and C. Kose, The effects of sensor and feature level fusion methods in multimodal emotion analysis. Niğde Ömer Halisdemir University Journal of Engineering Sciences, 13(4), 1093–1099, 2024. https://doi.org/10.28948/ngmuh.1469289.
23. H. Z. Abyaneh, A. M. Nia, M. B. Varkeshi, S. Marofi and O. Kisi, Performance evaluation of ANN and ANFIS models for estimating garlic crop evapotranspiration. Journal of Irrigation and Drainage Engineering, 137(5), 280–286, 2011. https://doi.org/10.1061/(ASCE)IR.1943-4774.0000298.
24. J. R. Jang, ANFIS: Adaptive-network-based fuzzy inference system. IEEE Transactions on Systems, Man, and Cybernetics, 23(3), 665–685, 1993. https://doi.org/10.1109/21.256541.
25. O. Karakus, E. E. Kuruoglu and M. A. Altinkaya, One-day ahead wind speed/power prediction based on polynomial autoregressive model. IET Renewable Power Generation, 11(11), 1430–1439, 2017. https://doi.org/10.1049/iet-rpg.2016.0972.
26. H. Tabari, O. Kisi, A. Ezani and P. H. Talaee, SVM, ANFIS, regression and climate based models for reference evapotranspiration modeling using limited climatic data in a semi-arid highland environment. Journal of Hydrology, 444-445, 78–89, 2012. https://doi.org/10.1016/j.jhydrol.2012.04.007.
27. Mathworks 2025, Long short-term memory networks. https://www.mathworks.com/help/deeplearning/ug/long-short-term-memory-networks.html, Accessed 24 December 2025.
28. K. M. Chandy, and S. Taylor, An Introduction to Parallel Programming. MA: Jones and Bartlett, Boston, 1992.
29. S. Hochreiter and J. Schmidhuber, Long short-term memory. Neural Computation, 9(8), 1735–1780, 1997. https://doi.org/10.1162/neco.1997.9.8.1735.
30. S. Zahroh, Y. Hidayat, R. S. Pontoh, A. Santoso, A. T. Bon, and S. Firman, Modeling and forecasting daily temperature in Bandung. Proceedings of the International Conference on Industrial Engineering and Operations Management, pp. 406-412, Riyadh, Saudi Arabia, 2019.
31. A. G. Salman, Y. Heryadi, E. Abdurahman and W. Suparta, Single layer & multi-layer long short-term memory (LSTM) model with intermediate variables for weather forecasting. Procedia Computer Science, 135, 89–98, 2018. https://doi.org/10.1016/j.procs.2018.08.153.
32. Liu and L. Liu, Predicting housing price in China based on long short-term memory incorporating modified genetic algorithm. Soft Computing, 23, 11829–11838, 2019. https://doi.org/10.1007/s00500-018-03739-w.
33. International Renewable Energy Agency 2023, Renewable energy statistics. https://www.irena.org/, Accessed 24 December 2025.
34. International Renewable Energy Agency 2024, Renewable energy statistics. https://www.irena.org/, Accessed 24 December 2025.
35. Devlet Su Isleri (DSI-2025), 11th regional directorate of DSI of Türkiye. http://edirnenehir.dsi.gov.tr/, Accessed 24 December 2025.