Soil analysis to estimate soil fertility parameters is of great importance for precision agriculture but nowadays it still relies mainly on complex and time-consuming laboratory methods. Optical measurement techniques can provide a suitable alternative. Raman spectroscopy is of particular interest due to its ability to provide a molecular fingerprint of individual soil components. To overcome the major issue of strong fluorescence interference inherent to soil, we applied shifted excitation Raman difference spectroscopy (SERDS) using an in-house-developed dual-wavelength diode laser emitting at 785.2 and 784.6 nm. To account for the intrinsic heterogeneity of soil components at the millimeter scale, a raster scan with 100 individual measurement positions has been applied. Characteristic Raman signals of inorganic (quartz, feldspar, anatase, and calcite) and organic (amorphous carbon) constituents within the soil could be recovered from intense background interference. For the first time, the molecule-specific information derived by SERDS combined with partial least squares regression was demonstrated for the prediction of the soil organic matter content (coefficient of determination R2 = 0.82 and root mean square error of cross validation RMSECV = 0.41%) as important soil fertility parameter within a set of 33 soil specimens collected from an agricultural field in northeast Germany.
Keywords: SERDS; Shifted excitation Raman difference spectroscopy; dual-wavelength diode laser; fluorescence removal; soil constituents; soil organic matter.