Polluted soil–plant interaction analysis and soil classification based on laser-induced breakdown spectroscopy and machine learning

Abstract

A new method is introduced for the swift and precise detection of soil pollution and its effects on crops. Soil quality is essential for human well-being, with heavy metal pollution presenting considerable risks to both the ecological environment and human health. In crops, heavy metal contamination primarily occurs through mediums such as soil and water sources. This study introduces a system combining Laser-Induced Breakdown Spectroscopy (LIBS) with machine learning (ML) to analyze garlic contaminated by soil and the soil used for its cultivation. The simulation conducted in this study focuses on the impact of heavy metal-contaminated soil on garlic. Detection results indicate a significant influence of soil on garlic, resulting in heavy metal accumulation. Further analysis shows that metals from contaminated soil accumulate differently in various garlic plant parts, as per spectral data, underscoring the need for targeted detection methods to assess crop contamination. Conducting LIBS analysis on various soil samples enables the classification of different soil types. This indicates that tracing the origin of contaminated garlic through its residual soil is feasible. These findings imply the feasibility of tracing contaminated garlic's origin through its residual soil.