From benchtop to handheld MIR for soil analysis: Predicting lime requirement and organic matter in agricultural soils
2021
Recent development of handheld mid-infrared (MIR) spectrometers holds promising applications for soil analysis. However, problems arise due to water content and other in-situ characteristics, and compatibility with existing spectral libraries and calibrations developed on benchtop instruments. This study examined the potential of handheld MIR for sampling and analysis of field fresh soils. A handheld MIR-DRIFTS instrument (Agilent 4300 handheld FTIR) was used to determine primary soil fertility metrics, specifically, lime requirement (LR) and organic matter (%OM). LR was predicted from handheld collected spectra using calibration models developed from (1) an existing benchtop spectral library and (2) a newly developed model using handheld spectra. Predictions of handheld spectra from the benchtop model exhibited a bias of 2.14 t ha⁻¹, however this was improved (−0.13 t ha⁻¹) when calibrating the model from spectra collected from the handheld instrument. To overcome the adverse effects of moisture on spectral predictions of LR and %OM, moisture correction spectra for seven different moisture classes were applied to soil spectra under field conditions. The moisture correction for the prediction of LR under field moist condition was partially successful and could correct for soils with moisture contents below the permanent wilting point. Prediction of %OM using moisture correction was effective across all moisture categories, and predicted values came within 1% of laboratory values. This research highlights the challenges of handheld MIR in the field, such as finding moisture corrections, practices for field spectra acquisition and the use of spectral libraries and serves as a starting point for future research.
Afficher plus [+] Moins [-]Mots clés AGROVOC
Informations bibliographiques
Cette notice bibliographique a été fournie par National Agricultural Library
Découvrez la collection de ce fournisseur de données dans AGRIS