Laser-Induced Breakdown Spectroscopy and Chemometric Tools for Soil Fertility Classification and Carbon Calibration

In this study, laser-induced breakdown spectroscopy (LIBS) proved to be a promising alternative for direct and rapid multi-element analysis of 54 samples from three different areas, requiring minimal sample preparation. A calibration model was developed and validated for the carbon (C) line at$247,86 \text{nm}$. Furthermore, machine learning techniques were applied to distinguish soils fertilized with solid and liquid fertilizers, as well as non-fertilized soils, achieving an accuracy above$\mathbf{9 8. 3 \%}$. Principal Component Analysis (PCA) was used to cluster these soils, confirming the separation of the groups.