Optical sensing for the rapid prediction of soil geochemistry

Proposed legislation to secure and maintain soil quality in Europe has generated interest surrounding how best to characterize soil geochemistry, and how to assess and monitor soil contamination. Many soil laboratories are now... [ view full abstract ]

Proposed legislation to secure and maintain soil quality in Europe has generated interest surrounding how best to characterize soil geochemistry, and how to assess and monitor soil contamination. Many soil laboratories are now equipped with technology platforms in portable visible near-infrared (vis-NIR), mid-infrared (MIR) and X-ray fluorescence (XRF) spectrometers. These technologies have complementary capabilities, where XRF is known to accurately measure the soil’s inorganic element concentration, and vis-NIR and MIR have the ability to estimate the soil’s organic component and mineralogy suites. The objective of this study was to investigate if there was a benefit to using these techniques in a synergistic way for the estimation of soil properties. This hypothesis was tested using agricultural topsoils (n = 322) sampled from the National Soil Database. Data mining was used to estimate soil geochemistry from the vis-NIR and MIR, and in a novel way from the XRF spectral data. The predictions were combined into a single prediction outcome, using formal methods called model-averaging procedures. Prediction accuracy for a suite of soil geochemistry (40 total elements) was much improved using this approach, and the total number of well predicted elements increased from 15 to 25. Most notable is the large number of trace elements (As, Cd, Co, Cu, Hg, Mn, Ni and Zn) predicted to good or reasonable accuracy in non-polluted soils. It is concluded that the synergistic use of vis-NIR/MIR and XRF spectral methods are well placed as a tool to permit large scale routine soil geochemical monitoring (i.e. soil contamination).