Mid‐ and Near‐Infrared Spectroscopic Determination of Carbon in a Diverse Set of Soils from the Brazilian National Soil Collection

Abstract

Calibrations for soil carbon content measured by combustion (total carbon, TC) and chromate oxidation by a modified Walkley‐Black method (Walkley‐Black carbon, WBC) from the Brazilian National Soil Collection were made using Fourier‐transform near (1100 to 2500 nm; NIRS) and mid‐infrared diffuse reflectance (2,500 to 25,000 nm; DRIFTS) spectroscopy combined with partial least squares (PLS). Calibration sets of sample populations of different carbon ranges, soil taxonomic classes, and soil textural groups were established. These are for TC ranges between 0.4 to 555.0, 0.4 to 99.1, and 0.4 to 39.9 g kg^?1: for WBC 0.2 to 401.0, 0.2 to 66.0, and 0.2 to 66.0, and 0.2 to 30.0 g kg^?1: for soil taxonomic classes Ferralsols and Acrisols; and for soil textural groups very clayey, clayey, and medium textures were examined. Calibrations obtained for the largest TC and WBC ranges were better compared to the lower ones, but lower root mean squared deviation (RMSD) and relative difference (RD=RMSD/mean value) were found for the lower carbon ranges. Taxonomic soil class was not an adequate criterium for calibration set formation. Soil texture had effect on calibrations, especially using NIR, because of the particle size effect to which NIR was more sensitive than mid‐IR. In general, DRIFTS showed better performance than NIRS. NIRS only outperformed DRIFTS when used with calibration set fairly homogeneous in its particle size distribution. Results demonstrated that while calibrations can be developed using either DRIFTS or NIRS for even a very diverse set of soil samples, which will determine C over a wide range of concentrations inherent in such a diverse set, it is desirable to seperate sample populations by soil textural properties and choose the adequate spectral range (NIR or mid‐IR) based on the textural group, for calibration development to achieve more accurate results.