Abstract
The current soil nutrient estimation method is laborious, repetitive, time-consuming and costly, making it less efficient for large-scale soil fertility assessment in precision agriculture practice. This paper discussed the feasibility of visible and near-infrared (Vis-NIR) spectroscopy as an alternative method for rapid measurement of total nitrogen in the soil, which is more efficient for a huge paddy field area. For this purpose, Vis-NIR reflectance spectra (350 – 1750 nm) were acquired on 200 soil samples using spectrometers. Partial Least Squares Regression (PLSR) with full (leave-one-out) cross-validation was used to develop the calibration model between the Vis-NIR soil spectra and the total nitrogen obtained by chemical analysis in laboratory. The coefficient of determination (R2val) and residual prediction deviation (RPD) of the developed calibration model for total nitrogen (Ntot) was 0.78 and 1.86, respectively. The predicted total nitrogen map generated based on the Vis-NIR spectroscopy was comparable with the laboratory analysis’s measured map. This result indicates that the Vis-NIR infrared spectroscopy is the potential to be used for total nitrogen estimation in soil.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
