Near Infrared Spectroscopy for Rapid Prediction of Soil Organic Carbon Content in Agricultural Soils

Soil organic carbon (SOC) or C-organic is a key component of soil quality that affects the properties of organic materials and soil mixtures.It also holds practical value and importance in agriculture.Traditionally, determining SOC has involved expensive and time-consuming procedures that require the use of chemicals and may cause pollution.Therefore, there is a need for an alternative method that is fast, environmentally friendly, and cost-effective, as they are key important factors in precision agriculture practices the near infrared reflectance spectroscopy (NIRS) technique can be considered as a suitable option since it is non-destructive, requires simple preparation, and does not cause pollution.The main objective of this study is to apply the NIRS technique to predict SOC levels and classify soils based on geographical and soil land-use characteristics.Soil samples were collected from four different locations, and their spectra data were acquired in the wavenumbers range of 4000-10,000 cm -1 .A prediction model for SOC was developed using the NIR spectra data and the partial least squares regression (PLSR) method, followed by k-fold cross-validation.The results demonstrated that the NIRS technique successfully predicted SOC levels, with a maximum correlation coefficient (r) of 0.96 and a residual predictive deviation (RPD) index of 4.05, indicating excellent model performance.In conclusion, the NIRS technique can be applied as a rapid and nondestructive method for predicting C-organic levels and classifying soil characteristics.