Modeling of Atrazine Adsorption onto Surficial Sediments in the System of Cadmium and Malathion Co-Existed

Based on the data of single atrazine (AT) adsorption, co-sorption of AT and cadmium (Cd), and co-sorption of AT and malathion (Ma) onto surficial sediments (SSs), a BP artificial neural network (ANN) model was established, in which the inputs were selected as concentrations of AT, Cd, Ma and the output was the amount of AT adsorption onto SSs, in order to simulate the interaction of co-existed Ma and Cd on AT adsorption. The results verified that the model could simulate the effect of Ma, Cd on the adsorption of AT onto SSs, there was only less than 8.19% of average relative error existed between the values predicated through the BP model and experimental ones. The determination coefficient between the fitting curve and the Nash-Sutcliffe simulation efficiency coefficient (NSC) was 0.9973 (> 0.80), indicating the model could describe AT adsorption onto SSs in the system of Cd and Ma co-existed well. The influence of Cd and Ma on the adsorption of AT in the AT-Cd-Ma system could be also predicted via the established BP ANN model. The results show that the presence of Cd in the AT-Cd-Ma system will enhance the adsorption of AT and the higher Cd concentration, the less AT will be adsorbed onto SSs; moreover, the co-existed Ma will restrain the adsorption of AT due to competitive sorption behavior.