Study on spectral feature characterizing and soil nitrogen content prediction

Near infrared reflectance spectroscopy offers the potential for rapid and cost-effective soil analysis. The spectra features of the soil samples at different soil particles, manually crushed; <0.25; .5-1; 1-2; 2-5 and >5 mm were collected and analyzed. The results indicated that particle size affected strongly the absorbance of the spectra. Accurate N prediction was obtained when the particle sizes vary from 0.5-1.0; 1.0-2.0 and 2.0-5 mm with r of 0.82; 0.81 and 0.82, respectively. Dried and humid soil samples reflectance spectra were also obtained to analyze the influence of the moisture content on soil spectra features. Poor N prediction was obtained when the soil kept its natural moisture with r of 0.57. This research will be useful for future applicability of NIR spectroscopy as a technique to make prediction on site.