Prediction of volumetric soil organic carbon from field‐moist intact soil cores

This study investigated the potential for visible–near‐infrared (vis–NIR) spectroscopy to predict locally volumetric soil organic carbon (SOC) from spectra recorded from field‐moist soil cores. One hundred cores were collected from a 71‐ha arable field. The vis–NIR spectra were collected every centimetre along the side of the cores to a depth of 0.3 m. Cores were then divided into 0.1‐m increments for laboratory analysis. Reference SOC measurements were used to calibrate three partial least‐squares regression (PLSR) models for bulk density (ρ_b), gravimetric SOC (SOC_g) and volumetric SOC (SOC_v). Accurate predictions were obtained from averages of spectra from those 0.1‐m increments for SOC_g (ratio of performance to inter‐quartile (RPIQ) = 5.15; root mean square error (RMSE) = 0.38%) and SOC_v (RPIQ = 5.25; RMSE = 4.33 kg m^?3). The PLSR model for ρ_b performed least well, but still produced accurate results (RPIQ = 3.76; RMSE = 0.11 Mg m^?3). Predictions for ρ_b and SOC_g were combined to compare indirect and direct predictions of SOC_v. No statistical difference in accuracy between these approaches was detected, suggesting that the direct prediction of SOC_v is possible. The PLSR models calibrated on the 10‐cm depth intervals were also applied to the spectra originally recorded on a 1‐cm depth increment. While a bigger bias was observed for 1‐cm than for 10‐cm predictions (1.13 and 0.19 kg m^?3, respectively), the two populations of estimates were not distinguishable statistically. The study showed the potential for using vis–NIR spectroscopy on field‐moist soil cores to predict SOC at high depth resolutions (1 cm) with locally derived calibrations.     

基于可见近红外和中红外近地面光谱数据融合的土壤有机碳含量反演

  目的  以传统的实验室分析方法进行大规模土壤有机碳（SOC）含量调查耗时、费力、成本高昂,以土壤可见近红外（VNIR）、中红外（MIR）光谱或两光谱数据融合手段能够快速预测SOC含量,但预测精度不一、特别是光谱数据融合技术应用于土柱样本的效果尚待考察。  方法  从全球土壤光谱库筛选出同时具有VNIR光谱、MIR光谱和SOC含量的677个土柱共计3755个土样。光谱数据经Savitzky–Golay平滑和一阶微分预处理后,用Kennard–Stone算法进行建模和验证的集合划分,使用偏最小二乘回归与随机森林方法分别建立以VNIR、MIR以及两者融合的VNMIR光谱为自变量的SOC含量预测模型,并对模型精度进行评估。  结果  MIR光谱模型的SOC预测精度优于VNIR光谱模型,VNMIR光谱模型预测精度低于MIR光谱模型但优于VNIR光谱模型。  结论  使用光谱数据融合技术预测SOC含量并非一定比使用单一光谱数据的精度高,就本例而言使用MIR光谱数据构建预测模型的方法是快速、准确预测大尺度时空范围SOC含量的最 佳手段。     

Can Near or Mid‐Infrared Diffuse Reflectance Spectroscopy Be Used to Determine Soil Carbon Pools?

Abstract

The objective of this study was to compare mid‐infrared (MIR) an near‐infrared (NIR) spectroscopy (MIRS and NIRS, respectively) not only to measure soil carbon content, but also to measure key soil organic C (SOC) fractions and the δ¹³C in a highly diverse set of soils while also assessing the feasibility of establishing regional diffuse reflectance calibrations for these fractions. Two hundred and thirty‐seven soil samples were collected from 14 sites in 10 western states (CO, IA, MN, MO, MT, ND, NE, NM, OK, TX). Two subsets of these were examined for a variety of C measures by conventional assays and NIRS and MIRS. Biomass C and N, soil inorganic C (SIC), SOC, total C, identifiable plant material (IPM) (20× magnifying glass), the ratio of SOC to the silt+clay content, and total N were available for 185 samples. Mineral‐associated C fraction, δ¹³C of the mineral associated C, δ¹³C of SOC, percentage C in the mineral‐associated C fraction, particulate organic matter, and percentage C in the particulate organic matter were available for 114 samples. NIR spectra (64 co‐added scans) from 400 to 2498 nm (10‐nm resolution with data collected every 2 nm) were obtained using a rotating sample cup and an NIRSystems model 6500 scanning monochromator. MIR diffuse reflectance spectra from 4000 to 400 cm^?1 (2500 to 25,000 nm) were obtained on non‐KBr diluted samples using a custom‐made sample transport and a Digilab FTS‐60 Fourier transform spectrometer (4‐cm^?1 resolution with 64 co‐added scans). Partial least squares regression was used with a one‐out cross validation to develop calibrations for the various analytes using NIR and MIR spectra. Results demonstrated that accurate calibrations for a wide variety of soil C measures, including measures of δ¹³C, are feasible using MIR spectra. Similar efforts using NIR spectra indicated that although NIR spectrometers may be capable of scanning larger amounts of samples, the results are generally not as good as achieved using MIR spectra.     

Accounting for the effects of water and the environment on proximally sensed vis–NIR soil spectra and their calibrations

Visible–near infrared (vis–NIR) spectroscopy can be used to estimate soil properties effectively using spectroscopic calibrations derived from data contained in spectroscopic databases. However, these calibrations cannot be used with proximally sensed (field) spectra because the spectra in these databases are recorded in the laboratory and are different to field spectra. Environmental factors, such as the amount of water in the soil, ambient light, temperature and the condition of the soil surface, cause the differences. Here, we investigated the use of direct standardization (DS) to remove those environmental factors from field spectra. We selected 104 sensing (sampling) sites from nine paddy fields in Zhejiang province, China. At each site, vis–NIR spectra were recorded with a portable spectrometer. The soils were also sampled to record their spectra under laboratory conditions and to measure their soil organic matter (SOM) content. The resulting data were divided into training and validation sets. A subset of the corresponding field and laboratory spectra in the training set (the transfer set) was used to derive the DS transfer matrix, which characterizes the differences between the field and laboratory spectra. Using DS, we transferred the field spectra of the validation samples so that they acquired the characteristics of spectra that were measured in the laboratory. A partial least squares regression (PLSR) of SOM on the laboratory spectra of the training set was then used to predict both the original field spectra and the DS‐transferred field spectra. The assessment statistics of the predictions were improved from R² = 0.25 and RPD = 0.35 to R² = 0.69 and RPD = 1.61. We also performed independent predictions of SOM on the DS‐transferred field spectra with a PLSR derived using the Chinese soil spectroscopic database (CSSD), which was developed in the laboratory. The R² and RPD values of these predictions were 0.70 and 1.79, respectively. Predictions of SOM with the DS‐transferred field spectra were more accurate than those treated with external parameter orthogonalisation (EPO), and more accurate than predictions made by spiking. Our results show that DS can effectively account for the effects of water and environmental factors on field spectra and improve predictions of SOM. DS is conceptually straightforward and allows the use of calibrations made with laboratory‐measured spectra to predict soil properties from proximally sensed (field) spectra, without needing to recalibrate the models.     

Improved analysis and modelling of soil diffuse reflectance spectra using wavelets

Diffuse reflectance spectroscopy using visible (vis), near‐infrared (NIR) and mid‐infrared (mid‐IR) energy can be a powerful tool to assess and monitor soil quality and function. Mathematical pre‐processing techniques and multivariate calibrations are commonly used to develop spectroscopic models to predict soil properties. These models contain many predictor variables that are collinear and redundant by nature. Partial least squares regression (PLSR) is often used for their analysis. Wavelets can be used to smooth signals and to reduce large data sets to parsimonious representations for more efficient data storage, computation and transmission. Our aim was to investigate their potential for the analyses of soil diffuse reflectance spectra. Specifically we wished to: (i) show how wavelets can be used to represent the multi‐scale nature of soil diffuse reflectance spectra, (ii) produce parsimonious representations of the spectra using selected wavelet coefficients and (iii) improve the regression analysis for prediction of soil organic carbon (SOC) and clay content. We decomposed soil vis‐NIR and mid‐IR spectra using the discrete wavelet transform (DWT) using a Daubechies’s wavelet with two vanishing moments. A multiresolution analysis (MRA) revealed their multi‐scale nature. The MRA identified local features in the spectra that contain information on soil composition. We illustrated a technique for the selection of wavelet coefficients, which were used to produce parsimonious multivariate calibrations for SOC and clay content. Both vis‐NIR and mid‐IR data were reduced to less than 7% of their original size. The selected coefficients were also back‐transformed. Multivariate calibrations were performed by PLSR, multiple linear regression (MLR) and MLR with quadratic polynomials (MLR‐QP) using the spectra, all wavelet coefficients, the selected coefficients and their back transformations. Calibrations by MLR‐QP using the selected wavelet coefficients produced the best predictions of SOC and clay content. MLR‐QP accounted for any nonlinearity in the data. Transforming soil spectra into the wavelet domain and producing a smaller representation of the data improved the efficiency of the calibrations. The models were computed with reduced, parsimonious data sets using simpler regressions.     

Comparison of Near Infrared Reflectance Spectroscopy with Combustion and Chemical Methods for Soil Carbon Measurements in Agricultural Soils

As interest in soil organic carbon (SOC) dynamics increases, so do needs for rapid, accurate, and inexpensive methods for quantifying SOC. Objectives were to i) evaluate near infrared reflectance (NIR) spectroscopy potential to determine SOC and soil organic matter (SOM) in soils from across Tennessee, USA; and ii) evaluate potential upper limits of SOC from forest, pasture, no-tillage, and conventional tilled sites. Samples were analyzed via dry-combustion (SOC), Walkley–Black chemical SOM, and NIR. In addition, the sample particle size was classified to give five surface roughness levels to determine effects of particle size on NIR. Partial least squares regression was used to develop a model for predicting SOC as measured by NIR by comparing against SOM and SOC. Both NIR and SOM correlated well (R² > 0.9) with SOC (combustion). NIR is therefore considered a sufficiently accurate method for quantifying SOC in soils of Tennessee, with pasture and forested systems having the greatest accumulations.Abbreviations SOC, soil organic carbon; NIR, Near Infrared Reflectance Spectroscopy; MTREC, Middle Tennessee Research and Education Center; RECM, Research and Education Center at Milan; PREC, Plateau Research and Education Center; PLS, Partial least squares.     

Quantifying moisture and roughness with Support Vector Machines improves spectroscopic soil organic carbon prediction

The challenges of Vis‐NIR spectroscopy are permanent soil surface variations of moisture and roughness. Both disturbance factors reduce the prediction accuracy of soil organic carbon (SOC) significantly. For improved SOC prediction, both disturbance effects have to be determined from Vis‐NIR spectra, which is especially challenging for roughness. Thus, an approach for roughness quantification under varying moisture and its impact on SOC assessment using Support Vector Machines is presented here.     

苹果栽培区土壤参数的近红外及中红外测定

Soil quality monitoring is important in precision agriculture.This study aimed to examine the possibility of assessing the soil parameters in apple-growing regions using spectroscopic methods.A total of 111 soil samples were collected from 11 typical sites of apple orchards,and the croplands surrounding them.Near-infrared(NIR) and mid-infrared(MIR) spectra,combined with partial least square regression,were used to predict the soil parameters,including organic matter(OM) content,pH,and the contents of As,Cu,Zn,Pb,and Cr.Organic matter and pH were closely correlated with As and the heavy metals.The NIR model showed a high prediction accuracy for the determination of OM,pH,and As,with correlation coefficients(r) of 0.89,0.89,and 0.90,respectively.The predictions of these three parameters by MIR showed reduced accuracy,with r values of 0.77,0.84,and 0.92,respectively.The heavy metals could also be measured by spectroscopy due to their correlation with organic matter.Both NIR and MIR had high correlation coefficients for the determination of Cu,Zn,and Cr,with standard errors of prediction of 2.95,10.48,and 9.49 mg kg-1 for NIR and 3.69,5.84,and 6.94 mg kg-1 for MIR,respectively.Pb content behaved differently from the other parameters.Both NIR and MIR underestimated Pb content,with r values of 0.67 and 0.56 and standard errors of prediction of 3.46 and 2.99,respectively.Cu and Zn had a higher correlation with OM and pH and were better predicted than Pb and Cr.Thus,NIR spectra could accurately predict several soil parameters,metallic and nonmetallic,simultaneously,and were more feasible than MIR in analyzing soil parameters in the study area.     

Improved estimates of organic carbon using proximally sensed vis–NIR spectra corrected by piecewise direct standardization

We investigated the use of piecewise direct standardization (PDS) to remove the effects of water and other environmental factors from proximally sensed (field) visible–near infrared (vis–NIR) spectra. Our hypothesis was that the PDS‐standardized field spectra can be used to predict soil carbon effectively with calibrations derived from existing spectroscopic databases of spectra recorded in the laboratory on dried, ground and sieved samples. In our experiments we used field spectra recorded in situ with a portable spectrometer at 124 sites in 11 paddy fields in Zhejiang Province, China. We sampled the soil at these same sites, recorded their spectra in the laboratory and measured their soil organic carbon (SOC) contents with a conventional laboratory technique. Two‐thirds of the samples were used to relate the laboratory spectra to SOC by partial least squares regression (PLSR), and the remaining one‐third was used as an independent validation dataset. We selected a representative set of samples from corresponding field and laboratory spectra that we could use as the PDS transfer set. Piecewise direct standardization was used to relate each wavelength in the laboratory spectra to the corresponding wavelength and its neighbours in the field spectra. The field spectra of the validation samples were then corrected with PDS so that they acquired the characteristics of the spectra measured under laboratory conditions. The approach was evaluated by (i) quantifying the similarity between the PDS‐standardized spectra and their corresponding laboratory spectra, (ii) measuring the accuracy of their SOC predictions on the independent validation dataset and (iii) comparing these results with those of direct standardization (DS). Both PDS and DS led to considerable improvements in the predictions of SOC (R² = 0.71, R² = 0.60, respectively), compared with those with original field spectra (R² = 0.03). However, fewer transfer samples were needed with PDS to obtain similar results.     

Sample preparation and selection for qualitative and quantitative analyses of soil organic carbon with mid‐infrared reflectance spectroscopy

Mid‐infrared spectroscopy (MIRS) is a well‐established analytical tool for qualitative and quantitative analysis of soil samples. However, effects of soil sample grinding procedures on the prediction accuracy of MIR models and on qualitative spectral information have not been well investigated and, in consequence, not standardized up to now. Further, the effects of soil sample selection on the accuracy of MIR prediction models has not been quantified yet. This study investigated these effects by using 180 well‐characterized soil samples that were ground for different times (0, 2 or 4 minutes) and then used for MIR measurements. To study the impact of sample preparation, soil spectra were subjected to principal component analyses (PCA), multiple regression and partial least square (PLS) analysis. The results indicate that the prediction accuracy of MIR models for soil organic carbon (SOC) and pH and the qualitative spectral information were better overall for lightly ground (2 minutes) soil samples compared with intensively (4 minutes) or unground soil samples. Whereas the grinding procedure did not show any effect on spectra of clay minerals, spectral information for quartz and for SOC was modified. Even though it is difficult to recommend a global standardized soil sample grinding procedure for MIR measurements because of different mill types available within laboratories, we highly recommend using an internally standardized grinding procedure. Moreover, we show that neither land use nor soil sampling depth influences the prediction of the SOC content. However, sand and clay content substantially affect the score vectors used by the PLS algorithm to predict the SOC content. Thus, we recommend using soil samples similar in texture for more precise SOC calibration models for MIR spectroscopy.     

Near-infrared spectroscopy for determination of soil organic C,microbial biomass C and C and N fractions in a heterogeneous sample of German arable surface soils

Visual and near-infrared spectroscopy (vis-NIRS) is an established method to estimate soil properties. However, only limited information is available to estimate C and N fractions in a heterogeneous sample. The objectives of our study were to determine estimation accuracies of vis-NIRS using two software for soil organic carbon (SOC), total nitrogen (N_t), pH, texture, and C and N fractions (light (LF), mineral (MF), labile, intermediate and passive fractions) in a heterogeneous sample (consisting of 51 units with different mineralogy) and to compare these results with those obtained by mid-infrared spectroscopy (MIRS). Analyzing vis-NIRS spectra and the mentioned properties showed a possibility to distinguish between high and low values for SOC (residual prediction deviation (RPD) = 1.90) and N_t (RPD = 1.93). Sand and clay could be estimated, whereas pH and silt could not. No useful estimation was possible for N-LF, passive C, intermediate C or intermediate N. C-LF, C-MF and N-MF could be differentiated between high and low values, whereas for passive N the estimation was approximate quantitative. MIRS reached one or two times higher estimation categories than vis-NIRS for SOC, N_t, pH and texture, suggesting that MIRS has a higher potential to estimate soil properties in a heterogeneous sample.     

Prediction of atrazine sorption coefficients in soils using mid-infrared spectroscopy and partial least-squares analysis

This study explored the potential of mid-infrared spectroscopy (MIR) with partial least-squares (PLS) analysis to predict sorption coefficients (Kd) of pesticides in soil. The MIR technique has the advantage of being sensitive to both the content and the chemistry of soil organic matter and mineralogy, the important factors in the sorption of nonionic pesticides. MIR spectra and batch Kd values of atrazine were determined on a set of 31 soil samples as reference data for PLS calibration. The samples, with high variability in soil organic carbon content (SOC), were chosen from 10 southern Australian soil profiles (A1, A2, B, and C in one case). PLS calibrations, developed for the prediction of Kd from the MIR spectra and reference Kd data, were compared with predictions from Koc-based indirect estimation using SOC content. The reference Kd data for the 31 samples ranged from 0.31 to 5.48 L/kg, whereas Koc ranged from 30 to 680 L/kg. Both coefficients generally increased with total SOC content but showed a relatively poor coefficient of determination (R2 = 0.53; P > 0.0001) and a high standard error of prediction (SEP =1.22) for the prediction of Kd from Koc. This poor prediction suggested that total SOC content alone could explain only half of the variation in Kd. In contrast, the regression plot of PLS predicted versus measured Kd resulted in an improved correlation, with R2 = 0.72 ( P > 0.0001) and standard error of cross-validation (SECV) = 0.63 for three PLS factors. With the advantages of MIR-PLS in mind, (i) more accurate prediction of Kd, (ii) an ability to reflect the nature and content of SOC as well as mineralogy, and (iii) high repeatability and throughput, it is proposed that MIR-PLS has the potential for an improved and rapid assessment of pesticide sorption in soils.     

Exploring the predictability of soil texture and organic matter content with a commercial integrated soil profiling tool

In soil mapping, combining information from conceptually different proximal soil sensors can increase the accuracy of prediction and robustness of the model when compared with using individual sensors. In this study the predictability of soil texture (clay, silt and sand fractions) and soil organic matter (SOM) content was tested with a commercial integrated soil profiling tool that included sensors for measuring apparent electrical conductivity (EC_a), reflectance in the visible and near‐infrared (vis‐NIR) parts of the electromagnetic spectrum and insertion force (IF). The measurements were made at 20 locations on each of two Swedish farms. At every location, sensor measurements were made at 1.5‐cm intervals from the soil surface to a depth of 0.8 m. Soil samples were collected close to the sensor measurement points and analysed for texture and SOM content. Farm‐specific calibrations were developed for texture and SOM with each sensor separately and with combinations of all three sensors. The calibrations were made using both partial least squares regression (PLSR) and simple linear regression. The results for the two farms were quite consistent in terms of rank in prediction performance between the individual sensors and the sensor combinations. The vis‐NIR spectrometer was the best individual sensor for predicting the soil properties tested on both farms, with root mean square error of cross‐validation (RMSECV) of 0.3–0.5% for SOM, about 6% for clay and silt and 10–11% for sand. The inclusion of IF reduced the RMSECV for predictions of SOM content by about 10%. For soil texture, including EC_a reduced the RMSECV on average for all particle size fractions by 5–10%. However, the small improvements obtained by combining sensors do not provide strong support for combining vis‐NIR sensor measurements with measurements of EC_a and or IF.     

Improving the prediction performance of a large tropical vis‐NIR spectroscopic soil library from Brazil by clustering into smaller subsets or use of data mining calibration techniques

Effective agricultural planning requires basic soil information. In recent decades visible near‐infrared diffuse reflectance spectroscopy (vis‐NIR) has been shown to be a viable alternative for rapidly analysing soil properties. We studied 7172 samples of seven different soil types collected from several regions of Brazil and varying in organic matter (OM) (0.2–10.3%) and clay content (0.2–99.0%). The aim was to explore the possibility of enhancing the performance of vis‐NIR data in predicting organic matter and clay content in this library by dividing it into smaller sub‐libraries on the basis of their vis‐NIR spectra. We used partial least square regression (PLSR) models on the sub‐libraries and compared the results with PLSR and two non‐linear calibration techniques, boosted regression trees (BT) and support vector machines (SVM) applied to the whole library. The whole library calibrations for clay performed well (ME (modelling efficiency) > 0.82; RMSE (root mean squared error) < 10.9%), reflecting the influence of the direct spectral responses of this property in the vis‐NIR range. Calibrations for OM were reasonably good, especially in view of the very small variation in this property (ME > 0.60; RMSE < 0.55%). The best results were, however, found when dividing the large library into smaller subsets by using variation in the mean‐normalized or first derivative spectra. This divided the global data set into clusters that were more uniform in mineralogy, regardless of geographical origin, and improved predictive performance. The best clustering method improved the RMSE in the validation to 8.6% clay and 0.47% OM, which corresponds to a 21% and 15% reduction, respectively, as compared with whole library PLSR. For the whole library, SVM performed almost equally well, reducing RMSE to 8.9% clay and 0.48% OM.     

Prediction of soil organic matter using a spatially constrained local partial least squares regression and the Chinese vis–NIR spectral library

We need to determine the best use of soil vis–NIR spectral libraries that are being developed at regional, national and global scales to predict soil properties from new spectral readings. To reduce the complexity of a calibration dataset derived from the Chinese vis–NIR soil spectral library (CSSL), we tested a local regression method that combined geographical sub‐setting with a local partial least squares regression (local‐PLSR) that uses a limited number of similar vis–NIR spectra (k‐nearest neighbours). The central idea of the local regression, and of other local statistical approaches, is to derive a local prediction model by identifying samples in the calibration dataset that are similar, in spectral variable space, to the samples used for prediction. Here, to derive our local regressions we used Euclidean distance in spectral space between the calibration dataset and prediction samples, and we also used soil geographical zoning to account for similarities in soil‐forming conditions. We tested this approach with the CSSL, which comprised 2732 soil samples collected from 20 provinces in the People's Republic of China to predict soil organic matter (SOM). Results showed that the prediction accuracy of our spatially constrained local‐PLSR method (R² = 0.74, RPIQ = 2.6) was better than that from local‐PLSR (R² = 0.69, RPIQ = 2.3) and PLSR alone (R² = 0.50, RPIQ = 1.5). The coupling of a local‐PLSR regression with soil geographical zoning can improve the accuracy of local SOM predictions using large, complex soil spectral libraries. The approach might be embedded into vis–NIR sensors for laboratory analysis or field estimation.     

Usefulness of near‐infrared spectroscopy for the prediction of chemical and biological soil properties in different long‐term experiments

The prediction accuracy of visible and near‐infrared (Vis‐NIR) spectroscopy for soil chemical and biological parameters has been variable and the reasons for this are not completely understood. Objectives were (1) to explore the predictability of a series of chemical and biological properties for three different soil populations and—based on these heterogeneous data sets—(2) to analyze possible predictive mechanisms statistically. A number of 422 samples from three arable soils in Germany (a sandy Haplic Cambisol and two silty Haplic Luvisols) of different long‐term experiments were sampled, their chemical and biological properties determined and their reflectance spectra in the Vis‐NIR region recorded after shock‐freezing followed by freeze‐drying. Cross‐validation was carried out for the entire population as well as for each population from the respective sites. For the entire population, excellent prediction accuracies were found for the contents of soil organic C (SOC) and total P. The contents of total N and microbial biomass C and pH were predicted with good accuracy. However, prediction accuracy for the other properties was less: content of total S was predicted approximately quantitatively, whereas Vis‐NIR spectroscopy could only differentiate between high and low values for the contents of microbial N, ergosterol, and the ratio of ergosterol to microbial biomass C. Contents of microbial biomass P and S, basal respiration, and qCO₂ could not be predicted. Prediction accuracies were greatest for the entire population and the Luvisol at Garte, followed by the Luvisol at Hohes Feld, whereas the accuracy for the sandy Cambisol was poor. The poor accuracy for the sandy Cambisol may have been due to only smaller correlations between the measured properties and the SOC content compared to the Luvisols or due to a general poor prediction performance for sandy soils. Another reason for the poor accuracy may have been the smaller range of contents in the sandy soil. Overall, the data indicated that the accuracy of predictions of soil properties depends largely on the population investigated. For the entire population, the usefulness of Vis‐NIR for the number of chemical and biological soil properties was evident by markedly greater correlation coefficients (measured against Vis‐NIR predicted) compared to the Pearson correlation coefficients of the measured properties against the SOC content. However, the cross‐validation results are valid only for the closed population used in this study.     

Using Interlaboratory Proficiency Data to Guide NIR/MIR Calibrations

There is growing interest in the use of near-range and/or midrange infrared (IR) diffuse reflectance spectroscopy (NIR and MIR) as nondestructive alternatives to chemical testing of soils. This trend is supported by research on how best to correlate IR spectral data with results obtained by conventional laboratory measurements. While for soils there is growing interest in developing local and national calibrations using “legacy” data, the proven analytical performance of provider laboratories now and earlier, the moisture status of reported results, and the method of soil preparation warrant greater attention. Examples for soil carbon (C) and total soil nitrogen (N) from Australasian interlaboratory proficiency testing across multiple years from 1993 are provided to demonstrate the magnitude of past and present measurement uncertainties, including the effects of method and different concentrations. The evidence is sufficient to require those commissioned to develop NIR and MIR calibrations to subject their prototype calibrations to external peer review by participating in credible, independent interlaboratory proficiency testing programs for ≥12 months, including checks on soil moisture status and possible effects of sample preparation. To rate as credible for most uses, the prototype results should be within the interquartile range for each sample and ideally there should be no outliers and few stragglers. Across the period of assessment (1993–2008), users of Walkley and Black organic C and Kjeldahl digestion for total soil N (Kjeldahl method does not measure total N, but most of the organic N plus an undetermined proportion of nitrate and nitrate present in the sample; quantitative inclusion of both requires a modification of the Kjeldahl procedure) declined as use of furnace technologies for soil C and N increased linearly. There is a strong case to commission two or three well-performing and experienced laboratories to reanalyze samples in “legacy” soil collections prior to finalizing predictive relationships with NIR/MIR spectra for the same samples.     

Near‐infrared spectroscopy for rapid estimation of microbial properties in reclaimed mine soils

Several chemical and microbial properties of mine soils need to be measured for comprehensive assessment of the reclamation success. The objective of this study was to evaluate the ability of NIR spectroscopy to predict organic C (C_org), total N (N_t), and several microbial properties of mine soils reclaimed for forestry. Soils samples (n = 154) were collected at two reclaimed areas in central and S Poland, and their spectra in the NIR region (including the visible range, 400–2500 nm) were recorded. A half of the samples was used to develop calibration equations, and another half was used for validation. The modified partial least squares regression was applied to build calibration equations using the whole spectrum (0 to 2nd derivative). The best predictions were obtained for C_org and N_t (ratio of standard deviation to standard error of prediction in the validation stage [RPD] = 3.4 and 4.1; the regressions coefficients [a] of linear regression [measured against predicted values] = 0.94 and 0.96; correlation coefficients [r] = 0.96 and 0.97, respectively). Very well predictive models applicable for quantitative measurements were obtained also for microbial biomass, basal respiration, and the activities of dehydrogenase and acid phosphatase (RPD = 2.3–2.5, a = 0.90–0.99, r = 0.90–0.92). Prediction of urease activity was slightly worse (RPD = 2.1, a = 0.88, r = 0.87) but sufficient for rough estimation. The obtained results indicated the ability of NIR spectroscopy to predict complex soil microbial properties. Therefore, application of this analytical method may improve the assessment of recovery of microbial functions in reclaimed post‐mining barrens.     

Determination of organic carbon and nitrogen in particulate organic matter and particle size fractions of Brookston clay loam soil using infrared spectroscopy

The objective of this study was to determine whether models developed from infrared spectroscopy could be used to estimate organic carbon (C) content, total nitrogen (N) content and the C:N ratio in the particulate organic matter (POM) and particle size fraction samples of Brookston clay loam. The POM model was developed with 165 samples, and the particle size fraction models were developed using 221 samples. Soil organic C and total N contents in the POM and particle size fractions (sand, 2000–53 µm; silt, 53–2 µm; clay, <2 µm) were determined by using dry combustion techniques. The bulk soil samples were scanned from 4000 to 400 cm^?1 for mid‐infrared (MIR) spectra and from 8000 to 4000 cm^?1 for near‐infrared (NIR) spectra. Partial least squares regression (PLSR) analysis and the ‘leave‐one‐out' cross‐validation procedure were used for the model calibration and validation. Organic C and N content and C:N ratio in the POM were well predicted with both MIR‐ and NIR‐PLSR models ( = 0.84–0.92; = 0.78–0.87). The predictions of organic C content in soil particle size fractions were also very good for the model calibration ( = 0.84–0.94 for MIR and = 0.86–0.92 for NIR) and model validation ( = 0.79–0.94 for MIR and = 0.84–0.91 for NIR). The prediction of MIR‐ and NIR‐PLSR models for the N content and the C:N ratio in the sand and clay fractions was also satisfactory ( = 0.73–0.88; = 0.67–0.85). However, the predictions for the N content and C:N ratio in the silt fraction were poor ( = 0.23–0.55; = 0.20–0.40). The results indicate that both MIR and NIR methods can be used as alternative methods for estimating organic C and total N in the POM and particle size fractions of soil samples. However, the NIR model is better for estimating organic C and N in POM and sand fractions than the MIR model, whereas the MIR model is superior to the NIR model for estimating organic C in silt and clay fractions and N in clay fractions.     

Discrimination of Australian soil horizons and classes from their visible–near infrared spectra

A soil's reflectance spectrum in the visible and near infrared is rich in information. It is an integrative property of the soil that measures its colour, the abundance of iron oxides, clay minerals and carbonates, the amount of water and organic matter and its particle size. We explored the merit of discriminating between soil horizons, of which we had 13 654 samples, and soil orders from the Australian Soil Classification, of which we had samples from 1697 profiles with designated horizons, by analysing quantitatively their diffuse reflectance spectra in the visible‐near infrared (vis–NIR) range (350–2500 nm). We re‐sampled the spectra to 10‐nm intervals and converted them to logarithms of deviations from their convex hulls. We then transformed them to canonical variates, which we display as scatter diagrams in low‐order canonical planes. The minerals, colour and organic constituents thought to be responsible for their discrimination are identified. Each spectrum was re‐examined and allocated to the group whose centroid was nearest in the canonical space. Topsoil horizons (A, A2 and transitional AB and AC horizons) were distinguishable from subsoils (B, C and transitional BC horizons). Vertosols, Ferrosols, Podosols (Podzols), Organosols, Calcarosols, Rudosols, Sodosols, Hydrosols, Kandosols and Kurosols were in general well separated from other soil orders and were assigned to their own orders such that 80% or more were correctly allocated. These orders possess characteristics that are easily distinguished by vis–NIR spectroscopy. Re‐allocations to other orders were interpretable and could be related to identifying features of the ASC classification. Our results show that spectra distinguish soil horizons and soil type. They suggest that vis–NIR spectroscopy could make an important contribution to the definition and identification of classes in an effective system of soil classification.