Evaluation of the support vector regression (SVR) and the random forest (RF) models accuracy for streamflow prediction under a data-scarce basin in Morocco

• Published in: Discover applied sciences Vol. 6; no. 6; p. 306
• Main Authors: Bargam, Bouchra, Boudhar, Abdelghani, Kinnard, Christophe, Bouamri, Hafsa, Nifa, Karima, Chehbouni, Abdelghani
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 03.06.2024; Springer Nature B.V
• Subjects: Algorithms; Applied and Technical Physics; Arid zones; Artificial intelligence; Basins; Chemistry/Food Science; Drought; Earth Sciences; Engineering; Environment; Geographic information systems; Hydrology; Learning algorithms; Machine learning; Materials Science; Mountains; Neural networks; Precipitation; Predictions; Regression; Resource management; River basins; Root-mean-square errors; Runoff; Semi arid areas; Semiarid zones; Stream discharge; Stream flow; Streamflow forecasting; Support vector machines; Water resources; Water resources management; Watersheds; Wavelet transforms; Morocco; Atlas Mountains; Streamflow prediction; Machine learning techniques; Support vector regression (SVR); Temporal variability of streamflow; Semi-arid regions; Water resources
• ISSN: 3004-9261; 2523-3963; 3004-9261; 2523-3971
• DOI: 10.1007/s42452-024-05994-z

Streamflow prediction is a key variable for water resources management. It becomes more important in semi-arid regions such as the Tensift river basin in Morocco, where water resources are facing a severe drought and the demand is continuously increasing. The present analysis focuses on evaluating Machine Learning techniques, namely support vector regression (SVR) and Random Forest (RF) against the multiple linear regression (MLR) for daily streamflow forecasting in the mountainous sub-basin of Rheraya between 2003 and 2016. The results show that SVR performed best, followed by RF and MLR. In measurable terms and regarding mean performance, SVR exhibited the higher Nash–Sutcliffe efficiency score (NSE = 0.59) and a lower root mean squared error (RMSE = 1.18  m 3 s - 1 ) compared to RF (NSE = 0.53, RMSE = 1.18  m 3 s - 1 ) and MLR (NSE = 0.54, RMSE = 1.01  m 3 s - 1 ). Furthermore,the available time series was too short to properly capture the full range of streamflow variability, which reduced the prediction performance outside of the calibration conditions. These findings suggest that ML algorithms, particularly SVR, can provide accurate streamflow estimation useful for water resources management when trained on a representative period. The results highlight the capacity of Machine Learning algorithms, specifically SVR, to augment streamflow prediction for enhanced water resource management in arid regions.