Predicting Soil Sorption Coefficients of an Environmental Pollutant Herbicide (Diuron) Using a Neural Network Model

Abstract

Background: A wide range of groundwater and soil pollutions - due to diuron herbicide - have resulted in intensive studies on their effects and transport in the environment. Modeling of sorption coefficients is an effective technique to investigate the effects and behavior of environmental pollutants such as diuron. Objectives: The purpose of the current study was to present an exact model with minimum required inputs, to predict the soil sorption coefficients (K d ) and the soil organic carbon sorption coefficients (K oc ) of diuron, in order to eliminate the need for time-consuming and costly laboratory experiments. Intelligent models based on artificial neural networks (ANNs) were used to achieve this objective. Materials and Methods: Data of this study were driven from the sorption studies, carried out on soils from a paddock under pasture at Flaxley Agriculture Centre, Mount Lofty Ranges, South Australia. Results: The multilayer perceptron (MLP) artificial neural networks (ANN) model with 1-6-1 layout, predicted nearly 98% of the variance of K d as well as 94% of the K oc variations with soil organic carbon content. Conclusions: Results showed that ANN is a powerful tool for predicting sorption coefficients using soil organic carbon content variations.