Load sediments quantification in Algerian North-West basins by ANN (Artificial Neurons Network) method

Abstract

Due to the complexity of basins morphometric parameters and the hydroclimatic irregularity of the semi-arid regions of North Africa, solid transport has been still far from being clearly assessed. This study attempts to shed light on this problem; in order to conceive a global model for the suspended sediment load quantification, taking into account all stream waters of the North-West area of Algerian. The calculation is based on the use of the ANN artificial neurons network method, which has proven its success and its reliability in several fields of research. The collected data are measured in hydrometric stations of several basins, such as Cheliff, Tafna, Macta, and Oran’s basins. The results obtained using the ANN method are sufficiently reliable, the best correlations were obtained for each studied stream water exceeding 97% (specific model to each station), and 90% in the case of a global model characterizing for all studied stations, which allows the extracted model to give better estimation of the suspended solid flow rates for any measured liquid flow rate of the north-west Algerian basins.