Artificial neural network for modelling the sediments accumulation in Es-Saada reservoir (North-Western Algeria)
by Mustapha Sidi Adda; Djilali Yebdri; Djilali Baghdadi; Sarita Gajbhiye Meshram
International Journal of Hydrology Science and Technology (IJHST), Vol. 17, No. 1, 2024

Abstract: Sediment deposition represents an important aspect of dam reservoir exploitation and management, as it relates to several operational and environmental problems. This study aimed to model the spatiotemporal evolution of the sediment accumulation in the Es-Saada reservoir (North-Western Algeria) using an artificial neural network (ANN) under low data conditions. The ANN model calibration was applied to the chronological period between the bathymetric surveys in 1986 and 2000, and the model verification was performed using data from a third survey conducted in 2003. The simulation of the reservoir bed presented acceptable results compared to the measured data (mean error of 7.76%). The model can provide predictive capacity curve for an average gap of 0.068 to the real curve, with a signification of 93.2%. It would be concluded that using determinist models for predicting sediment accumulation in reservoirs is complicated and needs all system details, while the application of ANN presents an adequate and uncomplicated method for predicting sediment distribution in dam reservoirs and also reservoir volume reduction in an approximate way.

Online publication date: Fri, 01-Dec-2023

  Free full text Access Free full text access

If you still need assistance, please email subs@inderscience.com