Use of visual-diagnostic color parameters of soils and optical reflectometry for determination of organic carbon content

Abstract


 
 
 To get additional tools for the assessment of carbon sequestration, along with the visual assessment of soil coloration with the applying of A. H. Munsell’s atlas, the analysis of color and spectral characteristics of soil using portable colorimeter NixPro and reflectometer Our Sci Reflectometer was carried out in this study. Elemental analysis of soil samples using X-ray fluorescence analysis was performed and the content of organic carbon was estimated. The spectral range of reflected light, which correlates most with the content of organic soil substance, was singled out. Based on the data, received by methods of reflectometry and colorimetry, prognostic regression models were constructed. A multiple linear regression equation with a statistically authentic luminosity predictor (L*) (R2=0.61) was obtained. It allows describing the link between the content of the organic substance in the studied soils and the parameters of the color setting system CIELab, as well as the equation describing 69 % of the data link dispersion between the integrated reflection coefficient and the organic carbon content of the soil. The link between the integral reflection coefficient and the total organic substance content was found. The most correlated spectral range with the content of organic substance – 500–632 nm was singled out. Regression models, which were based exclusively on the spectral data of pre-treated H2O2 soils, increased their predictability by 8–10 %. Approaches that can complement the tools for rapid determination of the organic carbon content in the soil were presented in the work. Researchers are expanding their arsenal of technical support for estimation of color or spectral coefficients of light reflection, based on which it is possible to conduct geospatial analysis and determine the content of the organic substance in low-humus soils with a probability of 69 %.