Effect of Moisture Content on Prediction of Organic Carbon and pH Using Visible and Near‐Infrared Spectroscopy

Abstract

This study was undertaken to investigate the effect of moisture content (MC) on the prediction accuracy of soil organic C (SOC) and pH of soils collected from Turkey and the United Kingdom using a fiber-type visible and near infrared (Vis-NIR) spectrophotometer. The diffuse reflectance spectra of 270 soil samples were measured under six gravimetric MC levels of 0, 5,10,15, 20, and 25%. Partial least squares (PLS) regression analyses with full cross-validation were performed to establish models for SOC and pH. Before PLS analysis, the entire spectra were randomly split three times into calibration (80%) and validation (207n) sets. Results showed that the prediction performance of SOC in the validation set was successful, with root mean square errors of prediction (RMSEPs) of 1.26 to 1.55% and residual prediction deviations (RPDs) of 2.29 to 2.8.3, and rather poor for pH, with RMSEPs of 0.65 to 0.85 and RPDs of 1.29 to 1.65. The best accuracy achieved for SOC was for dry soil samples (RMSEP = 1.26%, RPD = 2.83), whereas the worst accuracy was for wet soil samples with 5% MC (RMSEP -1.55%, RPD = 2.29). The best result for pH was obtained for dry samples (RMSEP -0.70%, RPD = 1.65), although this accuracy was comparable to that of the 107o MC soil samples (RMSEP = 0.65%, RPD -1.60). The ANOVA supported the conclusion that there was a significant effect of MC on prediction accuracy, although this effect was larger for SOC (P < 0.0000) than pH (P < 0.05).Abbreviations: MC, moisture content; SOC, soil organic carbon; PLS, partial least squares; RMSEP, root mean square error of prediction; RPD, residual prediction deviation; Vis-NIR, visible and near infrared.

Comparing the artificial neural network with parcial least squares for prediction of soil organic carbon and pH at different moisture content levels using visible and near-infrared spectroscopy

“…Mouazen et al (2006) found that variable soil MC decreased the predictive accuracies of several soil properties, including total C and pH. Morgan et al (2009) arrived at the same conclusion for SOC and inorganic C. Likewise, Tekin et al (2012), using the same data set as that of the current study, reported significantly improved results for prediction of SOC using dry soil samples, as compared to wet ones. These authors concluded that MC significantly affects the predictive performance of both SOC and pH, although this effect was found to be greater for the former than for the latter.…”

“…Many researchers have successfully measured soil organic carbon (SOC) using vis-NIR spectroscopy (Mouazen et al, 2007;Gomez et al, 2008;Vasquez et al, 2008;Leone et al, 2012;Tekin et al, 2012). A comprehensive analysis of the literature was carried out by Stenberg et al (2010), confirming the possibility of successful measurement of SOC with vis-NIR, which was attributed to the direct spectral response of C in the NIR range.…”

“…All samples were dried in a laboratory oven at 65 o C for 24 h, ground, and sieved in a 2 mm sieve. Before spectral measurement, dry samples were weighed to determine the amount of water by weight to be added to achieve the specified gravimetric MC levels of 5, 10, 15, 20, and 25 % (Tekin et al, 2012). To achieve homogeneous MC distribution, samples in cups after wetting were closed with plastic covers overnight until optical measurements.…”

Comparison between artificial neural network and partial least squares for on-line visible and near infrared spectroscopy measurement of soil organic carbon, pH and clay content

“…It was obvious that the performance of OC and CC models was better than that of pH model, which could be attributed to the direct spectral response in the NIR spectral range mainly Stenberg et al, 2010). The better performance of ANN-OC model, as compared to the corresponding PLSR-models might be attributed to a non-linear behavior documented for OC Stenberg, 2010;Tekin et al, 2012), which seemed to be overcome by the nature of ANN, in solving non-linear problems. However, the reason why ANN provided a better performance for OC and pH (showing no non-linear behaviour), as compared to PLSR was not clear.…”

Estimating Soil Organic Carbon of Cropland Soil at Different Levels of Soil Moisture Using VIS-NIR Spectroscopy

“…As MC increases, the reflectance generally decreases nonlinearly and independent of soil type [9,10]. Therefore, most studies have demonstrated that dried soil samples are preferable for modeling of soil properties [11,12]. Unfortunately, soils may significantly vary in moisture in field conditions.…”