Estimation of Suspended Sediment in Dam-impacted Streams Using Regression-based Methods

Öz

The estimation of suspended sediment (SS) is a crucial factor for dam reservoirs, river ecosystems, structural changes in water resources, various operational activities, environmental safety, and water management. In this study, classical regression analysis (CRA) and multivariate adaptive regression splines (MARS) methods were used to estimate the mean daily SS (mg/L) of the Patapsco River in the United States. Mean daily discharge (Q), turbidity (T) and SS data from the Catonsville and Elkridge observation stations were used to build the models. For each station, 1199 (65.74%) of the 1824 data recorded between October 2016 and September 2021 were used for training, 300 (16.45%) for validation and 325 (17.81%) for testing datasets. The performance of the developed models was evaluated using the root mean square error, the mean absolute error, and the Nash-Sutcliffe efficiency coefficient statistics. Three different models were developed for estimating SS using Q and T parameters. The model including both Q and T parameters together showed higher estimation performance than the other models at both stations. In addition, the CRA method showed high performance in the training dataset, while the MARS method showed high performance in the validation and testing datasets. It was concluded that the MARS method demonstrated superior estimation performance compared to the CRA method.

Anahtar Kelimeler

• Suspended Sediment
• MARS
• Classical Regression Analysis
• Modeling
• Patapsco River Basin

Regresyon Tabanlı Yöntemler Kullanılarak Baraj Etkisindeki Akarsularda Askıda Katı Madde Tahmini

Öz

Askıda katı madde (AKM) tahmini, baraj gölleri, nehir ekosistemleri, su kaynaklarının yapısal değişiklikleri, çeşitli operasyonel faaliyetler, çevre güvenliği ve su yönetimi bakımından oldukça önemlidir. Bu çalışmada Amerika Birleşik Devletleri’ndeki Patapsco Nehri’nin günlük ortalama AKM (mg/L) değerlerini tahmin etmek için klasik regresyon analizi (KRA) ve çok değişkenli uyarlanabilir regresyon eğrileri (ÇDURE) yöntemleri kullanılmıştır. Modelleri oluşturmak için Catonsville ve Elkridge gözlem istasyonlarından alınan günlük ortalama debi (Q), bulanıklık (T) ve AKM verileri kullanılmıştır. Ekim 2016 ile Eylül 2021 dönemini kapsayan 1824 veriden 1199’u eğitim (%65.74), 300’ü doğrulama (%16.45) ve 325’i test (%17.81) seti için kullanılmıştır. Geliştirilen modellerin performansları ortalama karesel hatanın karekökü, ortalama mutlak hata ve Nash-Sutcliffe verimlilik katsayısı istatistikleri kullanılarak değerlendirilmiştir. AKM tahmini için Q ve T parametreleri ile üç farklı model kurulmuştur. Q ve T parametrelerinin birlikte kullanıldığı model diğer modellere kıyasla her iki istasyonda da daha yüksek tahmin performansı göstermiştir. Ayrıca eğitim veri setinde KRA yöntemi, doğrulama ve test veri setlerinde ise ÇDURE yöntemi yüksek performans göstermiştir. ÇDURE yönteminin KRA yöntemine kıyasla daha yüksek tahmin performansı sergilediği sonucuna ulaşılmıştır.

Anahtar Kelimeler

• Askıda Katı Madde
• ÇDURE
• Klasik Regresyon Analizi
• Modelleme
• Patapsco Nehri Havzası

Kaynakça

1. Abed, M. S., Kadhim, F. J., Almusawi, J. K., Imran, H., Bernardo, L. F. A., & Henedy, S. N. (2023). Utilizing multivariate adaptive regression splines (MARS) for precise estimation of soil compaction parameters. Applied Sciences, 13(21), Article 11634. https://doi.org/10.3390/app132111634
2. Achite, M., Yaseen, Z. M., Heddam, S., Malik, A., & Kisi, O. (2022). Advanced machine learning models development for suspended sediment prediction: comparative analysis study. Geocarto International, 37(21), 6116–6140. https://doi.org/10.1080/10106049.2021.1933210
3. Adnan, R. M., Liang, Z., El-Shafie, A., Zounemat-Kermani, M., & Kisi, O. (2019). Prediction of suspended sediment load using data-driven models. Water, 11(10), Article 2060. https://doi.org/10.3390/w11102060
4. Adnan, R., Wang, M., Masood, A., Kisi, O., Shahid, S., & Zounemat-Kermani, M. (2025). Applications of advanced optimized neuro fuzzy models for enhancing daily suspended sediment load prediction. Computer Modeling in Engineering & Sciences, 143(1), Article 1249. https://doi.org/10.32604/cmes.2025.062339
5. Asadi, M. A., Mokhtari, L. G., Zandi, R., & Naemitabar, M. (2025). Modeling, evaluation and forecasting of suspended sediment load in Kal-e Shur River, Sabzevar Basin, in northeast of Iran. Applied Water Science, 15(3), Article 44. https://doi.org/10.1007/s13201-025-02361-0
6. Bayazıt, M. (1981). Hidrolojide istatistik yöntemler. İstanbul Teknik Üniversitesi Matbaası.
7. Bayram, A., Kankal, M., & Onsoy, H. (2012). Estimation of suspended sediment concentration from turbidity measurements using artificial neural networks. Environmental Monitoring and Assessment, 184(7), 4355–4365. https://doi.org/10.1007/s10661-011-2269-2
8. Cashman, M. J., Gellis, A. C., Boyd, E., Collins, M. J., Anderson, S. W., McFarland, B. D., & Ryan, A. M. (2021). Channel response to a dam‐removal sediment pulse captured at high‐temporal resolution using routine gage data. Earth Surface Processes and Landforms, 46(6), 1145–1159. https://doi.org/10.1002/esp.5083

9. Collins, M. J., Snyder, N. P., Boardman, G., Banks, W. S., Andrews, M., Baker, M. E., & Wilcock, P. (2017). Channel response to sediment release: Insights from a paired analysis of dam removal. Earth Surface Processes and Landforms, 42(11), 1636–1651. https://doi.org/10.1002/esp.4122
10. Collins, M. J., Baker, M. E., Cashman, M. J., Miller, A., & Van Ryswick, S. (2024). Impounded sediment and dam removal: Erosion rates and proximal downstream fate. Earth Surface Processes and Landforms, 49(9), 2690–2703. https://doi.org/10.1002/esp.5850
11. Campos, J. A., & Pedrollo, O. C. (2021). A regional ANN-based model to estimate suspended sediment concentrations in ungauged heterogeneous basins. Hydrological Sciences Journal, 66(7), 1222–1232. https://doi.org/10.1080/02626667.2021.1918695
12. Conoscenti, C., Rotigliano, E., Cama, M., Caraballo-Arias, N. A., Lombardo, L., & Agnesi, V. (2016). Exploring the effect of absence selection on landslide susceptibility models: A case study in Sicily, Italy. Geomorphology, 261, 222–235. https://doi.org/10.1016/j.geomorph.2016.03.006
13. Cui, Y., Collins, M. J., Andrews, M., Boardman, G. C., Wooster, J. K., Melchior, M., & McClain, S. (2019). Comparing 1-D sediment transport modeling with field observations: Simkins Dam removal case study. International Journal of River Basin Management, 17(2), 185–197. https://doi.org/10.1080/15715124.2018.1508024
14. Demirci, M., & Baltaci, A. (2013). Prediction of suspended sediment in river using fuzzy logic and multilinear regression approaches. Neural Computing and Applications, 23(1), 145–151. https://doi.org/10.1007/s00521-012-1280-z
15. Dukalang, H. H., & Otok, B. W. (2025). Modified partial least square structural equation model with multivariate adaptive regression spline: Parameter estimation technique and applications. MethodsX, 14, Article 103381. https://doi.org/10.1016/j.mex.2025.103381
16. Essam, Y., Huang, Y. F., Birima, A. H., Al-Hashimi, A., & Zain, M. (2022). Predicting suspended sediment load in Peninsular Malaysia using support vector machine and deep learning algorithms. Scientific Reports, 12, Article 302. https://doi.org/10.1038/s41598-021-04419-w
17. Friedman, J. H. (1991). Multivariate adaptive regression splines. The Annals of Statistics, 19(1), 1–67. https://doi.org/10.1214/aos/1176347963
18. Friedman, J. H., & Roosen, C. B. (1995). An introduction to multivariate adaptive regression splines. Statistical Methods in Medical Research, 4, 197–217. https://doi.org/10.1177/0962280295004003
19. Halverson, J. B. (2019). Flood City, USA: Ellicott City faces latest historical flooding. Weatherwise, 72(2), 12–18. https://doi.org/10.1080/00431672.2019.1559267
20. Hameed, M., Sharqi, S. S., Yaseen, Z. M., & Al-Ansari, N. (2017). Application of artificial intelligence (AI) techniques in water quality index prediction: A case study in tropical region, Malaysia. Neural Computing and Applications, 28, 893–905. https://doi.org/10.1007/s00521-016-2404-7
21. Huang, C. S., Legett, H. D., Plough, L. V., Aguilar, R., Fitzgerald, C., Gregory, B., Heggie, K., Lee, B., Richie, K. D., Harbold, W., & Ogburn, M. B. (2023). Early detection and recovery of river herring spawning habitat use in response to a mainstem dam removal. PLOS One, 18(5), Article e0284561. https://doi.org/10.1371/journal.pone.0284561
22. Jung, B. M., Fernandes, E. H., Moller Jr, O. O., & Garcia-Rodriguez, F. (2020). Estimating suspended sediment concentrations from river discharge data for reconstructing gaps of information of long-term variability studies. Water, 12(9), Article 2382. https://doi.org/10.3390/w12092382
23. Katatchambo, A. Y., & Bingöl, Ş. (2025). Comparison of regression based functions and ANN models for predicting the compressive strength of geopolymer mortars. Scientific Reports, 15(1), Article 11652. https://doi.org/10.1038/s41598-025-96772-3
24. Kemper, J. T. (2021). Patapsco River, Maryland. https://storymaps.arcgis.com/stories/99fe45668d5843ebb5e784a2da2e7dbd
25. Keshtegar, B., Piri, J., Hussan, W. U., Ikram, K., Yaseen, M., Kisi, O., Adnan, R. M., Adnan, M., & Waseem, M. (2023). Prediction of sediment yields using a data-driven radial M5 tree model. Water, 15(7), Article 1437. https://doi.org/10.3390/w15071437
26. Khan, M. Y. A., Tian, F., Hasan, F., & Chakrapani, G. J. (2019). Artificial neural network simulation for prediction of suspended sediment concentration in the River Ramganga, Ganges Basin, India. International Journal of Sediment Research, 34(2), 95–107. https://doi.org/10.1016/j.ijsrc.2018.09.001
27. Kisi, O., & Zounemat-Kermani, M. (2016). Suspended sediment modeling using neuro-fuzzy embedded fuzzy c-means clustering technique. Water Resources Management, 30, 3979–3994. https://doi.org/10.1007/s11269-016-1405-8
28. Kumantaş, M., Mete, B., Nacar, S., & Bayram, A. (2025). Estimation of suspended sediment concentration using regression analysis: A case study from the Patapsco River Basin, USA. In Y. Kaya, E. K. Oztekin, I. H. Hasan, N. Hasim, S. Oztekin, S. Kosunalp & U. Yildirim (Eds.), Proceedings of the 2nd international conference on engineering, natural sciences, and technological developments (pp. 405–411). Bayburt, Türkiye.
29. Malik, A., Kumar, A., Kisi, O., & Shiri, J. (2019). Evaluating the performance of four different heuristic approaches with Gamma test for daily suspended sediment concentration modeling. Environmental Science and Pollution Research, 26(22), 22670–22687. https://doi.org/10.1007/s11356-019-05553-9
30. Meshram, S. G., Safari, M. J. S., Khosravi, K., & Meshram, C. (2021). Iterative classifier optimizer-based pace regression and random forest hybrid models for suspended sediment load prediction. Environmental Science and Pollution Research, 28, 11637–11649. https://doi.org/10.1007/s11356-020-11335-5
31. Mete, B., Nacar, S., Bayram, A., & Baki, O. T. (2023). Estimation of total suspended solids concentration in streams using regression and artificial neural networks methods. Journal of Natural Hazards and Environment, 9(1), 125–135. https://doi.org/10.21324/dacd.1133981
32. Mohammadi, B., Guan, Y., Moazenzadeh, R., & Safari, M. J. S. (2021). Implementation of hybrid particle swarm optimization-differential evolution algorithms coupled with multi-layer perceptron for suspended sediment load estimation. Catena, 198, Article 105024. https://doi.org/10.1016/j.catena.2020.105024
33. Moriasi, D. N., Arnold, J. G., Van Liew, M. W., Bingner, R. L., Harmel, R. D., & Veith, T. L. (2007). Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Transactions of the ASABE, 50(3), 885–900. https://doi.org/10.13031/2013.23153
34. Nacar, S., Kankal, M., & Hınıs, M. A. (2018). Çok değişkenli uyarlanabilir regresyon eğrileri (ÇDURE) ile günlük akarsu akımlarının tahmini-Haldizen Deresi örneği. Gümüşhane Üniversitesi Fen Bilimleri Dergisi, 8(1), 38–47.
35. Nguyen, S. Q., Nguyen, L. C., Ngo-Duc, T., & Ouillon, S. (2024). Applying a machine learning-based method for the prediction of suspended sediment concentration in the Red River Basin. Modeling Earth Systems and Environment, 10(2), 2675–2692. https://doi.org/10.1007/s40808-023-01915-y
36. Parsaie, A., Haghiabi, A. H., Saneie, M., & Torabi, H. (2016). Prediction of energy dissipation on the stepped spillway using the multivariate adaptive regression splines. ISH Journal of Hydraulic Engineering, 22(3), 281–292. https://doi.org/10.1080/09715010.2016.1201782
37. Raja, M. N. A., & Shukla, S. K. (2021). Multivariate adaptive regression splines model for reinforced soil foundations. Geosynthetics International, 28(4), 368–390. https://doi.org/10.1680/jgein.20.00049
38. Sahraei, M. A., Duman, H., Çodur, M. Y., & Eyduran, E. (2021). Prediction of transportation energy demand: Multivariate adaptive regression splines. Energy, 224, Article 120090. https://doi.org/10.1016/j.energy.2021.120090
39. Samantaray, S., & Sahoo, A. (2021). Prediction of suspended sediment concentration using hybrid SVM-WOA approaches. Geocarto International, 37(19), 5609–5635. https://doi.org/10.1080/10106049.2021.1920638
40. Samui, P., & Kothari, D. P. (2012). A multivariate adaptive regression spline approach for prediction of maximum shear modulus and minimum damping ratio. Engineering Journal, 16(5), 69–78. https://doi.org/10.4186/ej.2012.16.5.69
41. Shiri, N., Shiri, J., Nourani, V., & Karimi, S. (2020). Coupling wavelet transform with multivariate adaptive regression spline for simulating suspended sediment load: Independent testing approach. ISH Journal of Hydraulic Engineering, 28(sup1), 356–365. https://doi.org/10.1080/09715010.2020.1801528
42. Steffy, L. Y., & Shank, M. K. (2018). Considerations for using turbidity as a surrogate for suspended sediment in small, ungaged streams: Time‐series selection, streamflow estimation, and regional transferability. River Research and Applications, 34(10), 1304–1314. https://doi.org/10.1002/rra.3373
43. Teixeira, L. C., Mariani, P. P., Pedrollo, O. C., dos Reis Castro, N. M., & Sari, V. (2020). Artificial neural network and fuzzy inference system models for forecasting suspended sediment and turbidity in basins at different scales. Water Resources Management, 34(11), 3709–3723. https://doi.org/10.1007/s11269-020-02647-9
44. Terrain360. (2025). Patapsco River Terrain360 Map Collection, Baltimore’s crown jewel. 19 Ocak 2026’da https://www.terrain360.com/map/patapsco-river/visual#10.7/39.1856/-76.5401 adresinden alındı. Travers, P. (2016). The Patapsco: Baltimore’s river of history. Shiffer Publishing.
45. Van, C. P., Le, H., & Chin, L. V. (2023). Estimation of daily suspended sediment concentration in the Ca River Basin using a sediment rating curve, multiple regression, and long short-term memory model. Journal of Water and Climate Change, 14(12), 4356–4375. https://doi.org/10.2166/wcc.2023.229
46. Wang, Y. M., T. Kerh, & S. (2009). Traore. Neural networks approaches for modelling river suspended sediment concentration due to tropical storms. Global Nest Journal, 11(4), 457–466.
47. Yilmaz, B., Aras, E., Nacar, S., & Kankal, M. (2018). Estimating suspended sediment load with multivariate adaptive regression spline, teaching-learning based optimization, and artificial bee colony models. Science of the Total Environment, 639, 826–840. https://doi.org/10.1016/j.scitotenv.2018.05.153
48. Zabihi, M., Pourghasemi, H. R., Pourtaghi, Z. S., & Behzadfar, M. (2016). GIS-based multivariate adaptive regression spline and random forest models for groundwater potential mapping in Iran. Environmental Earth Sciences, 75(8), Article 665. https://doi.org/10.1007/s12665-016-5424-9