Toxicity Prediction of Landfill Leachate-Contaminated Crops Using Machine Learning Models Based on PAH and Heavy Metal Concentrations

Abstract

The unregulated disposal of municipal solid waste in landfills leads to the generation of leachate, a polluted liquid that carries various harmful substances capable of endangering environmental quality and human health. This study investigates the extent of crop contamination in agricultural fields situated near landfill sites, with particular emphasis on the infiltration of polycyclic aromatic hydrocarbons (PAHs) and heavy metals into the surrounding soil. A total of 600 samples, comprising both soil and crop specimens, were systematically collected from five prominent landfill locations in South India. These samples were analyzed for sixteen PAHs and eight heavy metals using Gas Chromatography-Mass Spectrometry (GC-MS) and Atomic Absorption Spectrophotometry (AAS). The measured pollutant concentrations were classified as either “safe” or “unsafe” in accordance with internationally recognized safety standards. Several supervised machine learning models, including Artificial Neural Network (ANN), Support Vector Machine (SVM), Random Forest (RF), and K-Nearest Neighbour (KNN), were applied to predict the toxicity status of the samples, with Principal Component Analysis (PCA) employed to simplify the data and identify key contributing factors. Among the models tested, the ANN delivered the highest predictive accuracy at 97.8%, outperforming the other algorithms. The developed AI-driven framework offers a reliable and scalable approach for environmental risk assessment and promotes safer agricultural practices in landfill-affected regions.