Investigating the ability of Visible -NIR spectrometry to estimate some soil properties

Knowledge of soil chemical properties is important for understanding its engineering characteristics in road construction. Reflectance spectrometry is a non-destructive, fast and environment friendly method, but unfortunately no research has been conducted to use this tool to estimate the forest soil characteristics in Iran. The present study was conducted to investigate the ability of VNIR spectrometry to estimate some chemical properties including organic matter, CaCO3, pH and EC in the forest soil. For this purpose, 45 soil samples were collected from Namkhaneh series in Kheyroud Forest and those properties were measured by standard methods. The reflectance of sample was measured in dark room using Fieldspec ASD spectroradiometer. PLSR (Partial Least Squares Regression) method was used to predict soil properties in reflectance and first derivative mode. The results of RPD shows that OM has a very good estimate in two modes of reflectance spectra and derivative reflectance spectra (PRD>2) and pH, EC have acceptable estimate (1.42), but CaCO3 has poor prediction in two modes of reflectance spectra and derivative reflectance spectra (PRD<1.4). In general, based on the results of this study, spectrometry can be used to predict the above-mentioned properties except CaCO3 as a fast and an acceptable method and this method can simultaneously estimate several soil properties.