Abstract
The photoacoustic spectroscopy (PAS) technique has emerged as a valuable tool for the study of soil materials. The present work developed the experimental set-up for near infrared (800–2400 nm) photoacoustic spectroscopy (NIR-PAS), which consists of photoacoustic (PA) accessory, tungsten bromine lamp, monochromator, chopper and lock-in-amplifier, and this spectrometer was first applied to record the spectra of soils (n = 50) collected from the Loess Plateau of China; partial least-squares regression (PLSR) models and leave-one-out cross-validation were used to predict soil organic matter (SOM), clay, carbonate and available phosphorus content. The spectra varied among different soil samples; SOM played an important role in the spectral appearance, and the intensity and position of five typical absorption bands significantly shifted owing to the variances in the components and structure of SOM. The PLSR model demonstrated a good performance in the prediction of SOM with a root mean square error of 2.5 g kg−1 and a ratio of standard deviation to prediction error of 2.3. This pilot study demonstrates that NIR-PAS exhibits typical infrared absorptions and may be suitable for analysing soil samples.
Keywords
Get full access to this article
View all access options for this article.
