Prediction of soil macronutrients content using near-infrared spectroscopy

Near-infrared reflectance (NIR) spectroscopy is a technique that shows many possibilities in the field of testing chemical and physical properties of materials. Soil and agricultural products are not excluded of this practical application. The objective of this study was to investigate the potential of NIR spectroscopy to estimate nitrogen (N), phosphorus (P), potassium (K), organic matter (OM) and pH content in a loamy mixed soil. A 165 soil samples were taken from the field, 135 samples spectra were used during the calibration and validation stage, and 30 samples spectra were used for the prediction of N, P, K, OM and pH content. Principal component analysis/partial least square (PCA/PLS) was used as multivariate analysis technique to correlate wavelength information with each constituent concentration. The correlation coefficient (r) between measured and predicted values of N, OM and pH were 0.93, 0.93 and 0.91, and standard error of prediction (SEP) were 3.28, 0.06 and 0.07, respectively, which showed that NIR had potential to accurately predict these constituents in this soil. Unfortunately, it showed that NIR was not a good tool for P and K prediction with r, 0.47 and 0.68, and SEP, 33.70 and 26.54, respectively. The results showed that NIR could be a good tool for precision farming application.