Application of Fourier-transform infrared spectroscopy for quantification of chemical parameters in peat samples

The demand for the characterization of soil properties on a wide geographical scale with a high spatial resolution is constantly growing to implement various competitive studies, including climate and ecology studies. However, conventional soil analysis methods are time-consuming and expensive. Fourier-transform infrared spectroscopy (FTIR) has the potential to provide an alternative solution for the rapid and cost-effective determination of soil chemical and physical parameters. In this study, we calibrated a mid-infrared diffuse reflectance Fourier transform spectrometer (MIR-DRIFTS) for the determination of commonly analysed soil parameters such as pH, carbon (C), nitrogen (N), phosphorous (P), calcium (Ca), magnesium (Mg), potassium (K) and humic acid content and evaluated the performance of the elaborated soil analysis MIR-DRIFTS methods. The spectrometer calibration was performed by partial least squares regression analysis using soil spectra data acquired by the spectrometer and soil parameter reference values for the same samples acquired by traditional soil analysis methods. For calibration and validation of the MIR-DRIFTS methods, soil spectra were acquired in 3 replicates for a total of 1891 peat and peaty soils. To evaluate the performance of the elaborated MIR-DRIFTS methods, their uncertainty was evaluated by estimating the precision and accuracy of the results obtained. The highest potential of MIR-DRIFTS performance was observed for the determination of pH, C, N, and humic acid content, with an estimated combined uncertainty ranging around 6%, 6–28%, 16–20%, and 26–50%, respectively, depending on the analyte concentration. The estimated combined uncertainty of Ca, P, Mg, and K concentrations determined by MIR-DRIFTS ranged from 28–49%, 34–37%, 33–57%, and 57–100%, respectively. It was observed that the MIR-DRIFTS method’s performance is highly sensitive to the availability of samples used for method calibration. An increased number and diversity of samples used for calibration can improve the performance of MIR-DRIFTS methods of soil parameter determination.