Optical sensing and chemometric analysis of soil organic carbon – a cost effective alternative to conventional laboratory methods?

Abstract

Soil organic carbon (SOC) is frequently determined by the Walkley‐Black (WB) method. A limitation of the test is incomplete oxidation of the carbon fraction and underestimation of SOC. Automated dry combustion methods are expensive and slow. Optical sensing and chemometric analysis offer the potential of an economical method capable of quantifying SOC fractions. The aim of this study was to identify the best SOC analysis method to facilitate maximum sampling resolution based on the cost per sample, analytical accuracy and time. A comparative evaluation was made of five techniques; (1) the WB method, (2) total combustion by total organic C analyser, (3) infrared (IR) diffuse reflectance spectroscopy, (4) a portable spectroradiometer and (5) laboratory hyperspectral imaging. This involved assessing equipment costs, consumables and time to derive total analytical cost. The benefits were sample throughput and analytical accuracy. Instrumentation represented the largest input to analytical cost and for optical methods was governed by the spectral range. In contrast to dry combustion, this cost is offset by high sample throughput and minimal consumable requirements for IR spectroscopy and hyperspectral imaging. Hyperspectral imaging is identified as the most rapid technique with potential to scan about 720 samples per day at 90% less cost than the WB method. The opportunity cost of hyperspectral imaging is to forfeit some analytical accuracy associated with the dry combustion method. Dry combustion, despite its high cost per sample, incurs no further costs associated with updating prediction models or developing site or soil specific correction factors.

A comparison of point and imaging visible-near infrared spectroscopy for determining soil organic carbon

“…The decision on which instrument to employ for SOC will depend on the operator's decision to spend more time on sample preparation for point spectroscopy or more time on chemometric analysis for hyperspectral imaging. An added advantage for hyperspectral imaging is that multiple samples can be presented to the hyperspectral imaging system at once permitting greater economy of scale, 37 and with further method development in place soil samples may be run as intact samples eliminating the need for sample preparation. Figure 7.…”

Building a near infrared spectral library for soil organic carbon estimation in the Limpopo National Park, Mozambique

“…Laboratory spectroscopy is widely-applied in chemometrics (Geladi and Kowalski, 1986) and recently also to soil characterization (Brown et al, 2006;Shepherd and Walsh, 2002). It offers rapid and about 50% cheaper soil analysis (Cécillon et al, 2009a;O'Rourke and Holden, 2011), and, as an added benefit it is non-destructive, so samples can be analyzed repeatedly.…”

Prediction of Soil Organic Carbon at the European Scale by Visible and Near InfraRed Reflectance Spectroscopy

“…Visible and Near InfraRed (Vis-NIR) diffuse reflectance spectroscopy has been applied in soil analysis over the last 20 years [6] and has been demonstrated to accurately measure several soil attributes at minimal costs [7] and with satisfactory analytical errors [8]. Vis-NIR spectroscopy is currently used in laboratory conditions, but its application in-situ and even on air- or space-borne platforms is growing [9].…”