Visible And Near-infrared Reflectance Spectroscopy Research Articles

Analysis of four heavy metal concentrations in sediments from the Jiaozhou Bay, China by visible and near infrared spectroscopy (225–975 nm)

As an important component ofbiogeochemical cyclein coastal ecosystems, sediments are the sink of heavy metals. Therefore, distribution and dynamics of heavy metals in sediments could assess ecological quality and predict ecological risks. In the new era, rapid and green technology are highly needed, especially that could determine multi-parameters simultaneously. Here, we explored a new method to rapidly determine concentrations of heavy metals in sediments by visible and near infrared reflectance spectroscopy (VIRS).We sampled sediments in the Jiaozhou Bay, China, collected their reflectance spectra, and measured concentrations of four heavy metals (As, Cr, Cu, and Zn). Heavy metal models were established and evaluated using substances highly correlated with heavy metals. This study provides an effective reference for rapid analysis of As, Cr, Cu, and Zn simultaneously in sediments, at least in the Jiaozhou Bay, and for ecological environment protection and resource development of the Jiaozhou Bay.

Partial least square regression based machine learning models for soil organic carbon prediction using visible–near infrared spectroscopy

Monitoring and assessment of soil organic carbon (SOC) are critical for maintaining and enhancing the productivity of agricultural soils. The SOC is commonly determined through soil sampling and subsequent laboratory analysis using chemical methods. This method though very precise is time-consuming, labour-intensive and expensive. Contrarily, visible and near-infrared reflectance spectroscopy (VNIRS) may be utilised to estimate SOC in a quick, labour-saving, and cost-effective manner. In this study, 72 soil samples were collected for SOC estimation and spectra collection. This current work proposes to investigate the use of PLSR scores in place of raw spectral reflectance to increase both the computation and model efficiency by reducing the number of input variables while retaining the maximum information present in the original data. With the existing indices, ratio and normalized difference indices were calculated in all possible combinations and were regressed to SOC content to identify the best-performing indices. Ten different multivariate models were evaluated for SOC estimation using full-spectrum and partial least squares regression (PLSR) scores. The results revealed that reflectance gradually increased with increasing soil depth and decreasing SOC. The prediction models developed using existing indices were observed to be poor in predicting the SOC with the R2 values ranging from 0.009 to 0.34. The best spectral indices for estimating SOC were RI (R1888, R2015) and NDI (R1888, R2015) with R2 of 0.60, 0.61 and 0.39, 0.43 for calibration and validation datasets, respectively. The PLSR score-based multivariate models outperformed solo multivariate and optimized index-based models. Our study suggested that VNIRS with PLSR combined multivariate models can reliably be used for fast and non-invasive estimation of SOC.

Mapping soil arsenic pollution at a brownfield site using satellite hyperspectral imagery and machine learning.

Heavy metal contamination is ubiquitous in brownfields. Traditional site investigation employs geostatistical interpolation methods (GIMs) to predict the distribution of soil pollutants after soil sampling and chemical analysis. However, the heterogeneity of soil pollution in brownfields makes the assumptions of GIMs no longer valid and further undermines the accuracy of soil investigation. In the present study, a satellite hyperspectral image processing and machine learning method was developed to map arsenic pollution at a brownfield site. To eliminate the noise caused by atmospheric factors and increase the efficiency of spectral data, 1.3 million spectral indexes (SIs) were constructed and 1171 of them were selected due to their high correlations with soil arsenic. Five machine learning methods, i.e., Random forest (RF), ExtraTrees, Adaptive Boosting, Extreme Gradient Trees, and Gradient Descent Boosting Trees (GDB) were built to predict soil arsenic. The RF method was found to render the best performance (r = 0.78), reducing 30 % of prediction errors compared with traditional GIMs. RF also maintained a relatively higher level of accuracy (r = 0.56) when the sampling grids increased to 100 m, which was higher than that of GIMs under a 50 m sampling grid (r = 0.42), revealing that the proposed method can provide more accurate results with fewer sampling points, namely less investigation cost. It was indicated that the second derivate was the most efficient preprocessing method to remove spectral noise and normalized difference (ND) was the most reliable spectral index construction strategy. Based on uncertainty analysis, the heterogeneity of soil arsenic distribution was considered the most influential factor causing prediction errors. This study demonstrates that machine learning based on satellite visible and near-infrared reflectance spectroscopy (VNIR) is a promising approach to map soil arsenic contamination at brownfield sites with high accuracy and low cost.

Spectra transfer based learning for predicting and classifying soil texture with short-ranged Vis-NIRS sensor

Classifying soil texture is critical to investigate soil processes and functions influencing agronomic decisions and environmental stewardship. Despite the multiple benefits together with accurate estimation potential of visible and near-infrared reflectance spectroscopy (Vis-NIRS) over the traditional method of texture analysis, the prediction accuracy of the Vis-NIRS decreases due to the negative effect of soil moisture content (MC) on soil spectra. This study evaluated the performance of piecewise direct standardization (PDS) and direct standardization (DS) for eliminating the soil MC influence from fresh (non-processed wet) soil spectra to improve the accuracy of soil texture classification using a short-ranged (305–1700 nm) Vis-NIRS (Tec5 Technology, Germany). A set of 96 composite soil samples was collected from 18 fields, and the second set of 294 non-composite samples was collected from 10 of these 18 fields. All soil samples were scanned using a Vis-NIRS sensor after drying, grinding, and sieving, while the non-composite samples were additionally scanned in wet conditions. After analysing the composite samples for percent sand, silt, and clay, partial least squares regression (PLSR) models were calibrated (72 samples in calibration set) and validated (23 samples in validation set). Then, the PLSR model was used to predict the textural fractions for the wet, air-dried, DS and PDS transferred spectra of non-composite soil samples followed by their texture classification. Validation results indicated that the Vis-NIRS sensor predicted sand, silt, and clay fractions with excellent accuracies [coefficient of determination (R2) = 0.88–0.93; residual prediction deviation (RPD) = 2.91–3.74; ratio of the performance to interquartile range (RPIQ) = 2.12–5.02; root means square error (RMSE) = 3.62–8.06 %). Spectra transferred by the PDS and DS improved the texture classification accuracy up to 70 % and 90 %, respectively, while the wet soil spectra misclassified most fields (accuracy = 20 %). Unlike PDS, the DS predicted particles’ distribution seemed approximately similar to laboratory observations. Therefore, it was concluded that transferring the wet Vis-NIRS spectra by the DS algorithm can provide a rapid and accurate prediction of particle fractions and classification of soil texture classes with minimized effort, time, and cost.

Reflectance Spectroscopy with Multivariate Methods for Non-Destructive Discrimination of Edible Oil Adulteration.

The visible and near-infrared (Vis-NIR) reflectance spectroscopy was utilized for the rapid and nondestructive discrimination of edible oil adulteration. In total, 110 samples of sesame oil and rapeseed oil adulterated with soybean oil in different levels were produced to obtain the reflectance spectra of 350–2500 nm. A set of multivariant methods was applied to identify adulteration types and adulteration rates. In the qualitative analysis of adulteration type, the support vector machine (SVM) method yielded high overall accuracy with multiple spectra pretreatments. In the quantitative analysis of adulteration rate, the random forest (RF) combined with multivariate scattering correction (MSC) achieved the highest identification accuracy of adulteration rate with the full wavelengths of Vis-NIR spectra. The effective wavelengths of the Vis-NIR spectra were screened to improve the robustness of the multivariant methods. The analysis results suggested that the competitive adaptive reweighted sampling (CARS) was helpful for removing the redundant information from the spectral data and improving the prediction accuracy. The PLSR + MSC + CARS model achieved the best prediction performance in the two adulteration cases of sesame oil and rapeseed oil. The coefficient of determination () and the root mean square error () of the prediction set were 0.99656 and 0.01832 in sesame oil adulterated with soybean oil, and the and were 0.99675 and 0.01685 in rapeseed oil adulterated with soybean oil, respectively. The Vis-NIR reflectance spectroscopy with the assistance of multivariant analysis can effectively discriminate the different adulteration rates of edible oils.

Comparison of soil organic carbon stocks predicted using visible and near infrared reflectance (VNIR) spectra acquired in situ vs. on sieved dried samples: Synthesis of different studies

There is increasing demand for data on soil organic carbon (SOC) stock (SSOC; kgC m−2), but the acquisition of such data, which relies on the determination of volumetric SOC content (SOCv; gC dm−3), is often tedious or complex. Visible and near infrared reflectance spectroscopy (VNIRS) has proven useful for soil characterization, but has rarely been used for direct prediction of SOCv. The objectives of this work were: (i) to compare SOCv predictions using VNIR spectra collected in situ vs. on 2-mm sieved air-dried soil (laboratory conditions), on three sample sets separately (with in situ spectra collected differently for each set); and (ii) to assess SOCv prediction in independent validation using laboratory spectra from all sets.Predictions of SOCv were more accurate using laboratory than in situ spectra for two sets, but not for the third set, where coarse particles content was rather high and variable. Considering the total set of laboratory spectra, predictions in independent validation (leave-one-site-out) yielded accurate SOCv and SSOC predictions (standard errors of prediction were 1.9 gC dm−3 and 0.36 kgC m−2 at 0–30 cm depth, respectively). This result was achieved using local partial least squares regression (PLSR), based on spectral neighbors, which noticeably outperformed global PLSR (which uses all calibration samples equally), as often reported when using large soil spectral libraries for independent validation.Finally, this work demonstrated that SSOC could be quantified accurately using a VNIRS library built on archive soil samples, which offers important perspectives for SSOC accounting.

Concurrent starch accumulation in stump and high fruit production in coffee (Coffea arabica).

In coffee, fruit production on a given shoot drops after some years of high yield, triggering pruning to induce resprouting. The timing of pruning is a crucial farmer's decision affecting yield and labour. One reason for fruit production drop could be the exhaustion of resources, particularly the non-structural carbohydrates (NSC). To test this hypothesis in a Coffea L. arabica agroforestry system, we measured the concentrations of NSC, carbon (C) and nitrogen (N) in leaves, stems and stumps of the coffee plants, 2 and 5years after pruning. We also compared shaded vs full sun plants. For that purpose, both analytical reference and visible and near infrared reflectance spectroscopy (VNIRS) methods were used. As expected, concentrations of biochemical variables linked to photosynthesis activity (N, glucose, fructose, sucrose) decreased from leaves to stems, and then to stumps. In contrast, variables linked more closely to plant structure and reserves (total C, C:N ratio, starch concentration) were higher in long lifespan organs like stumps. Shading had little effect on most measured parameters, contrary to expectations. Concentrations of N, glucose and fructose were higher in 2-year-old organs. Conversely, starch concentration in perennial stumps was three times higher 5years after pruning than 2years after pruning, despite high fruit production. Therefore, the drop in fruit production occurring after 5-6years was not due to a lack of NSC on plant scale. Starch accumulation in perennial organs concurrently to other sinks, such as fruit growth, could be considered as a 'survival' strategy, which may be a relic of the behaviour of wild coffee (a tropical shade-tolerant plant). This study confirmed that VNIRS is a promisingly rapid and cost-effective option for starch monitoring (coefficient of determination for validation, R2val=0.91), whereas predictions were less accurate for soluble sugars, probably due to their too similar spectral signature.

Estimation of Organic Carbon in Anthropogenic Soil by VIS-NIR Spectroscopy: Effect of Variable Selection

Visible and near-infrared reflectance (VIS-NIR) spectroscopy is widely applied to estimate soil organic carbon (SOC). Intense and diverse human activities increase the heterogeneity in the relationships between SOC and VIS-NIR spectra in anthropogenic soil. This fact results in poor performance of SOC estimation models. To improve model accuracy and parsimony, we investigated the performance of two variable selection algorithms, namely competitive adaptive reweighted sampling (CARS) and random frog (RF), coupled with five spectral pretreatments. A total of 108 samples were collected from Jianghan Plain, China, with the SOC content and VIS-NIR spectra measured in the laboratory. Results showed that both CARS and RF coupled with partial least squares regression (PLSR) outperformed PLSR alone in terms of higher model accuracy and less spectral variables. It revealed that spectral variable selection could identify important spectral variables that account for the relationships between SOC and VIS-NIR spectra, thereby improving the accuracy and parsimony of PLSR models in anthropogenic soil. Our findings are of significant practical value to the SOC estimation in anthropogenic soil by VIS-NIR spectroscopy.

Predicting total petroleum hydrocarbons in field soils with Vis-NIR models developed on laboratory-constructed samples.

Accurate quantification of petroleum hydrocarbons (PHCs) is required for optimizing remedial efforts at oil spill sites. While evaluating total petroleum hydrocarbons (TPH) in soils is often conducted using costly and time-consuming laboratory methods, visible and near-infrared reflectance spectroscopy (Vis-NIR) has been proven to be a rapid and cost-effective field-based method for soil TPH quantification. This study investigated whether Vis-NIR models calibrated from laboratory-constructed PHC soil samples could be used to accurately estimate TPH concentration of field samples. To evaluate this, a laboratory sample set was constructed by mixing crude oil with uncontaminated soil samples, and two field sample sets (F1 and F2) were collected from three PHC-impacted sites. The Vis-NIR TPH models were calibrated with four different techniques (partial least squares regression, random forest, artificial neural network, and support vector regression), and two model improvement methods (spiking and spiking with extra weight) were compared. Results showed that laboratory-based Vis-NIR models could predict TPH in field sample set F1 with moderate accuracy (R2 >.53) but failed to predict TPH in field sample set F2 (R2 <.13). Both spiking and spiking with extra weight improved the prediction of TPH in both field sample sets (R2 ranged from .63 to .88, respectively); the improvement was most pronounced for F2. This study suggests that Vis-NIR models developed from laboratory-constructed PHC soil samples, spiked by a small number of field sample analyses, can be used to estimate TPH concentrations more efficiently and cost effectively compared with generating site-specific calibrations.

Prediction of Soil Oxalate Phosphorus using Visible and Near-Infrared Spectroscopy in Natural and Cultivated System Soils of Madagascar

Phosphorus is among the main limiting nutrients for plant growth and productivity in both agricultural and natural ecosystems in the tropics, which are characterized by weathered soil. Soil bioavailable P measurement is necessary to predict the potential growth of plant biomass in these ecosystems. Visible and near-infrared reflectance spectroscopy (Vis-NIRS) is widely used to predict soil chemical and biological parameters as an alternative to time-consuming conventional laboratory analyses. However, quantitative spectroscopic prediction of soil P remains a challenge owing to the difficulty of direct detection of orthophosphate. This study tested the performance of Vis-NIRS with partial least square regression to predict oxalate-extractable P (Pox) content, representing available P for plants in natural (forest and non-forest including fallows and degraded land) and cultivated (upland and flooded rice fields) soils in Madagascar. Model predictive accuracy was assessed based on the coefficient of determination (R2), the root mean squared error of cross-validation (RMSECV), and the residual predictive deviation (RPD). The results demonstrated successful Pox prediction accuracy in natural (n = 74, R² = 0.90, RMSECV = 2.39, and RPD = 3.22), and cultivated systems (n = 142, R² = 0.90, RMSECV = 48.57, and RPD = 3.15) and moderate usefulness at the regional scale incorporating both system types (R² = 0.70, RMSECV = 71.87 and RPD = 1.81). These results were also confirmed with modified bootstrap procedures (N = 10,000 times) using selected wavebands on iterative stepwise elimination–partial least square (ISE–PLS) models. The wavebands relevant to soil organic matter content and Fe content were identified as important components for the prediction of soil Pox. This predictive accuracy for the cultivated system was related to the variability of some samples with high Pox values. However, the use of “pseudo-independent” validation can overestimate the prediction accuracy when applied at site scale suggesting the use of larger and dispersed geographical cover sample sets to build a robust model. Our study offers new opportunities for P quantification in a wide range of ecosystems in the tropics.

Machine Learning Based On-Line Prediction of Soil Organic Carbon after Removal of Soil Moisture Effect

It is well-documented in the visible and near-infrared reflectance spectroscopy (VNIRS) studies that soil moisture content (SMC) negatively affects the prediction accuracy of soil attributes. This work was undertaken to remove the negative effect of SMC on the on-line prediction of soil organic carbon (SOC). A mobile VNIR spectrophotometer with a spectral range of 305–1700 nm and spectral resolution of 1 nm (CompactSpec, Tec5 Technology, Germany) was used for the spectral measurements at four farms in Flanders, Belgium. A total of 381 fresh soil samples were collected and divided into a calibration set (264) and a validation set (117). The validation samples were processed (air-dried and grind) and scanned with the same spectrophotometer in the laboratory. Three SMC correction methods, namely, external parameter orthogonalization (EPO), piecewise direct standardization (PDS), and orthogonal signal correction (OSC) were used to correct the on-line fresh spectra based-on its corresponding laboratory spectra. Then, the Cubist machine learning method was used to develop calibration models of SOC using the on-line spectra (after correction) of the calibration set. Results indicated that the EPO-Cubist outperformed the PDS-Cubist and the OSC-Cubist, with considerable improvements in the prediction results of SOC (coefficient of determination (R2) = 0.76, ratio of performance to deviation (RPD) = 2.08, and root mean square error of prediction (RMSEP) = 0.12%), compared with the corresponding uncorrected on-line spectra (R2 = 0.55, RPD = 1.24, and RMSEP = 0.20%). It can be concluded that SOC can be accurately predicted on-line using the Cubist machine learning method, after removing the negative effect of SMC with the EPO method.

Rapid Identification and Prediction of Cadmium-Lead Cross-Stress of Different Stress Levels in Rice Canopy Based on Visible and Near-Infrared Spectroscopy

Accurate detection of cadmium (Cd) and lead (Pb)-induced cross-stress on crops is essential for agricultural, ecological environment, and food security. The feasibility to diagnose and predict Cd–Pb cross-stress in agricultural soil was explored by measuring the visible and near-infrared reflectance of rice leaves. In this study, two models were developed—namely a diagnostic model and a prediction model. The diagnostic model was established based on visible and near-infrared reflectance spectroscopy (VNIRS) datasets with Support Vector Machine (SVM), followed by leave-one-out cross-validation (LOOCV). A partial least-squares (PLS) regression, as the prediction model was employed to predict the foliar concentration of Cd and Pb contents. To accurately calibrate the two models, a rigorous greenhouse experiment was designed and implemented, with 4 levels of treatments on each of the Cd and Pb stress on rice. Results show that with the appropriate pre-processing, the diagnostic model can identify 79% of Cd and 85% of Pb stress of any levels. The significant bands that have been used mainly distributed between 681–776 nm and 1224–1349 nm for Cd stress and 712–784 nm for Pb stress. The prediction model can estimate Cd with coefficient of determination of 0.7, but failed to predict Pb accurately. The results illustrated the feasibility to diagnose Cd stress accurately by measuring the visible and near-infrared reflectance of rice canopy in a cross-contamination soil environment. This study serves as one step forward to heavy metal pollutant detection in a farmland environment.

Improved Mapping of Soil Heavy Metals Using a Vis-NIR Spectroscopy Index in an Agricultural Area of Eastern China

Visible and near-infrared reflectance (Vis-NIR) spectroscopy can provide low-cost and high-density data for mapping various soil properties. However, a weak correlation between the spectra and measurements of soil heavy metals makes spectroscopy difficult to use in predicting incipient risk areas. In this study, we introduce a new spectral index (SI) based on Vis-NIR spectra and use it as a covariate in ordinary cokriging (OCK) to improve the mapping of soil heavy metals. The SI was defined from the highest covariance between spectra and heavy metal content in the partial least squares regression (PLSR) model. The proposed mapping approach was compared with an ordinary kriging (OK) predictor that uses only soil heavy metal data and an OCK predictor that uses soil organic matter (SOM) and Fe as covariates. To this end, a total of 100 topsoil (0-20 cm) samples were collected in an agricultural area near Longkou City, and the contents of As, Pb and Zn in the soil were determined. The results showed that OCK with the SI provided better results in terms of unbiasedness and accuracy compared to other comparative methods. Additionally, we explored the SI through simple strategies based on spectral analysis and correlation statistics and found that the SI synthesized most of the soil properties affected by heavy metals and was not limited to Fe and SOM. In summary, the SI method is cost-effective for improving soil heavy metal mapping and can be applied to other areas.

Potential of VNIR spectroscopy for prediction of clay mineralogy and magnetic properties, and its paleoclimatic application to two contrasting Quaternary soil deposits

The properties of phyllosilicates, iron oxides (oxyhydroxides), and magnetic minerals in soils can respond to changes in pedogenic weathering intensity and past climate. Numerous paleoclimate reconstructions using soil mineralogy have made use of only one or two of these mineral types, generally in order to limit analytical effort or expense, thereby foregoing comparisons between different mineralogic proxies. Quaternary soil deposits in China afford a unique opportunity for assessing the potential of visible and near-infrared reflectance (VNIR) spectroscopy for analysis of soil minerals, chemical weathering, and paleoclimate reconstruction. Here, we investigated two typical Quaternary soil types exhibiting different weathering degrees, i.e., a loess-paleosol sequence in northern China and a red earth deposit in southern China. Clay and iron mineralogy from the soils was compared with VNIR data using partial least squares regression (PLSR) and continuum-removal-based multiple linear regression (MLRCR) approaches. VNIR spectra show potential for analysis of pedogenic weathering and paleoclimate reconstruction in weakly to highly weathered soils in that spectral behavior is strongly related to both clay-mineral composition and magnetic properties. Both higher concentrations of minerals and their wider variability can improve model predictive ability. Phyllosilicates and non-magnetic Fe-oxides are predicted more accurately for highly weathered red earth deposits, whereas soil magnetic properties are predicted more accurately for loess-paleosols. The predictors identified by the MLRCR method (D2200, D680, and D900) and the ranges constrained by the PLSR method (650–700 and 2150–2250 nm) deserve more attention in future studies of spectral pedology and chemical weathering. We suggest that VNIR spectroscopy has the potential to be an alternative and broadly applicable tool for reconstructing paleoclimate using soil-mineral properties, substituting for a series of common mineralogic analyses. This would significantly reduce the workload and expense of experiments and improve the resolution of paleoclimate studies.

Quantification of soil organic carbon stock in urban soils using visible and near infrared reflectance spectroscopy (VNIRS) in situ or in laboratory conditions

Urban soils, like other soils, can be sink or source for atmospheric carbon dioxide, and due to urban expansion, are receiving increasing attention. Studying their highly variable attributes requires high-density sampling, which can hardly be achieved using conventional approaches. The objective of this work was to determine the ability of visible and near infrared reflectance spectroscopy (VNIRS) to quantify soil organic carbon (SOC) concentration (gC kg−1) and stock (gC dm−3, or MgC ha−1 for a given depth layer) in parks and sealed soils of two French cities, Marseille and Nantes, using spectra collected on pit walls or in laboratory conditions (air dried, 2 mm sieved samples).Better VNIRS predictions were achieved using laboratory than in situ spectra (R2 ≈ 0.8–0.9 vs. 0.7–0.8 in validation), and for sample SOC concentration than stock (R2val up to 0.83 in situ and 0.95 in the laboratory vs. 0.78 and 0.89, respectively). Stock was conventionally calculated according to four methods that variably account for coarse particles (>2 mm); and it was better predicted when coarse particles were not taken into account. This was logical using laboratory spectra, collected on 2 mm sieved samples; but concerning in situ spectra, this suggested the operator tended to put the spectrometer beside the coarsest particles during spectrum acquisition. This point is worth considering for urban soils, often rich in coarse particles.Stocks were then aggregated at the profile level: SOC stock prediction was more accurate at profile than sample level when using laboratory spectra (R2val = 0.94 vs. 0.89, respectively), probably due to uncertainty compensation; but this was not the case when using in situ spectra, possibly because samples collected for SOC analysis and corresponding VNIRS scans were not at the exact same location.This work demonstrates VNIRS usefulness for quantifying SOC stock time- and cost-effectively, in urban soils especially.

Estimating heavy metal concentrations in suburban soils with reflectance spectroscopy

Soil contamination by heavy metals has become a serious environmental issue worldwide. Rapidly and reliably obtaining heavy metal concentrations in soil is vital for soil monitoring and remediation. Visible and near-infrared reflectance (VNIR) spectroscopy provides a promising method for the estimation of heavy metal concentrations over large areas. Ninety-three soil samples were collected from a suburb of Wuhan City, Hubei Province, China, and their reflectance spectra were measured in the laboratory. This study aimed to (i) examine the feasibility of using soil reflectance spectra to estimate the concentrations of Cd, Pb, As, Cr, Cu and Zn in suburban soils; (ii) compare the performances of different spectral pretreatments and (iii) explore the mechanism underlying the estimation of heavy metal concentration from VNIR spectra. In particular, we proposed a strategy for the mechanism investigation that combined PCA biplot analysis, correlation and partial correlation analyses. Partial least-square regression was adopted to calibrate the VNIR model. Results showed that the VNIR model provided acceptable estimation accuracies for Cr, As and Cd concentrations with the ratio of the performance to deviation (RPD) values of 2.70, 1.81 and 1.63, respectively, but unsatisfactory estimation accuracies for Pb, Cu and Zn concentrations with the RPD values of 0.70–1.03. Savitzky–Golay smoothing outperformed other spectral pretreatments. The mechanisms underlying the estimation of the six studied heavy metals varied on a case-to-case basis. Specifically, the spectral estimation of Cd (Group I) concentration was attributed to its close correlations with soil organic matter (SOM). Cr and As (Group II) concentrations could be estimated by the VNIR model on the basis of their close correlations with Fe. Pb, Cu and Zn (Group III) concentrations, however, had weak correlations with neither SOM nor Fe, resulting in poor estimations. The proposed strategy on mechanism investigation for heavy metals could be transferred to other study areas. In summary, VNIR spectroscopy combined with the PLSR model is an alternative method for the rapid monitoring of some heavy metal pollution in suburban soils.

Visible and near infrared reflectance spectroscopy for field-scale assessment of Stagnosols properties

Spectral data contain information on soil organic and mineral composition, which can be useful for soil quality monitoring. The objective of research was to evaluate hyperspectral visible and near infrared reflectance (VNIR) spectroscopy for field-scale prediction of soil properties and assessment of factors affecting soil spectra. Two hundred soil samples taken from the experiment field (soil depth: 30 cm; sampling grid: 15 × 15 m) were scanned using portable spectroradiometer (350–1050 nm) to identify spectral differences of soil treated with ten different rates of mineral nitrogen (N) fertilizer (0–300 kg N/ha). Principal component analysis revealed distinction between higher- and lower-N level treatments conditioned by differences in soil pH, texture and soil organic matter (SOM) composition. Partial least square regression resulted in very strong correlation and low root mean square error (RMSE) between predicted and measured values for the calibration (C) and validation (V) dataset, respectively (SOM, %: R&lt;sub&gt;C&lt;/sub&gt;&lt;sup&gt;2&lt;/sup&gt; = 0.75 and R&lt;sub&gt;V&lt;/sub&gt;&lt;sup&gt;2&lt;/sup&gt; = 0.74; RMSE&lt;sub&gt;C&lt;/sub&gt; = 0.334 and RMSE&lt;sub&gt;V&lt;/sub&gt; = 0.346; soil pH: R&lt;sub&gt;C&lt;/sub&gt;&lt;sup&gt;2&lt;/sup&gt; = 0.78 and R&lt;sub&gt;V&lt;/sub&gt;&lt;sup&gt;2&lt;/sup&gt; = 0.62; RMSE&lt;sub&gt;C&lt;/sub&gt; = 0.448 and RMSE&lt;sub&gt;V&lt;/sub&gt; = 0.591). Results indicated that hyperspectral VNIR spectroscopy is an efficient method for measurement of soil functional attributes within precision farming framework.

Predicting nickel concentration in soil using reflectance spectroscopy associated with organic matter and clay minerals

Heavy metal contamination in soil has become a serious environmental problem. Visible and near-infrared reflectance spectroscopy (VNIRS) is recognized as an alternative to rapidly predict heavy metal concentration in soil. The correlation between soil spectrally active constituents and heavy metals provides a mechanism for the prediction. Generally, the entire VNIR spectral region (VNIR-SR) without discrimination is used. Considering the adsorption and retention of heavy metals on soil spectrally active constituents, a method was proposed to predict heavy metal concentration in soil using VNIRS. Organic matter and clay minerals have strong sorption and retention for Ni in soil. Therefore, the spectral bands associated with organic matter and clay minerals were used to predict nickel (Ni) concentration to validate the proposed method. In this study, two sets of reflectance spectra of soil samples collected in Chenzhou and Hengyang, Hunan Province, China were used. The prediction model was calibrated with a combination of genetic algorithm and partial least squares regression (GA-PLSR). In Chenzhou, the ratio of prediction to deviation (RPD) and the coefficient of determination (R2) were improved from 1.566 and 0.577 to 2.139 and 0.773 by the prediction using the spectral bands associated with organic matter and clay minerals compared with the prediction using the entire VNIR-SR. The RPD and R2 were improved from 1.805 and 0.672 to 3.144 and 0.892 by the proposed method in Hengyang. To further validate the proposed method, the soil samples from Chenzhou were reallocated to calibration and validation sets according to Ni concentration. The RPD and R2 were improved from 1.193 and 0.267 to 2.396 and 0.818 by using the spectral bands associated with organic matter and clay minerals. The results indicate that the proposed method is effective in predicting Ni concentration and has the potential to predict other heavy metals in soil.

Determination of Lignin Monomer Contents in Rice Straw Using Visible and Near-infrared Reflectance Spectroscopy

Genetic modification of plant lignin composition is an important strategy for improving biomass enzymatic digestibility without sacrificing the normal growth of the plant. However, the absence of fast and convenient methods for rapid determination of lignin composition has impeded corresponding research. Near-infrared reflectance spectroscopy (NIRS) analysis has potential as a solution for this dilemma, while the NIRS measurement for expediently assaying lignin composition in rice straws is still lacking. In this study, visible and near-infrared reflectance spectroscopy (VIS/NIRS) and modified partial least squares (MPLS) method were combined to develop calibration models for predicting the lignin monomer contents in a diverse rice population. Four optimal equations for predicting the content of p-hydroxyphenyl (H), guaiacyl (G), and syringyl (S) lignin units and their total amount (H + S + G) were generated with acceptable determination coefficients for calibration (0.85 to 0.93), cross-validation (0.75 to 0.88), and external validation (0.82 to 0.88), and the ratio performance deviation (RPD, 2 to 3.01). This study was the first to demonstrate that VIS/NIRS could give a sufficiently accurate prediction of lignin monomer contents in rice and could be applied for rapid assessments of large-scale rice straw samples.

Monitoring soil lead and zinc contents via combination of spectroscopy with extreme learning machine and other data mining methods

In order to limit pollution risk and develop proper remediation strategies, soil quality has to be controlled by rapid and sustainable monitoring measures. Visible and near-infrared reflectance spectroscopy (VisNIR) is an attractive surrogate to time-consuming and costly classical soil assessment protocols. It highly depends on selecting appropriate data mining methods for regression analysis. In this study, performance of a state of the art learning algorithm called extreme learning machine (ELM), was evaluated through comparing with the other calibration methods proposed in the literature for predicting lead (Pb) and Zinc (Zn) concentrations. Solid samples collected from a mine waste dump (n=120) were scanned using a Fieldspec3 portable spectroradiometer with a measurement range of (350–2500nm) in a laboratory. Transformation of the reflectance spectra to absorbance was followed by three pre-processing scenarios including Savitzky-Golay smoothing (SG), first derivative (FD) and second derivative (SD). Partial Least Square Regression (PLSR), Support Vector Machine (SVM) and neural networks with two learning algorithms models (back propagation and extreme learning machine), were calibrated on spectral features selected by genetic algorithm, and then applied to predict soil metal concentrations. The best prediction accuracy was obtained by FD-ELM method with R2p, RMSEp, concordance correlation coefficient and RPD values of 0.93, 63.01, 0.98 and 5.92 for Pb and 0.87, 167.90, 0.91 and 5.62 for Zn, respectively. Study of the prediction mechanism proved that element sorption by spectrally active Fe-oxide and clay contents of the soil was the major mechanism by which the spectrally featureless Pb and Zn ions can be predicted. The spatial patterns of predicted toxic elements showed that FD-ELM had the most similarity with those maps obtained by interpolating measured values. Over all, it is concluded that reflectance spectroscopy combined with the ELM algorithm is a rapid, inexpensive and accurate tool for indirect evaluation of Pb and Zn and mapping their spatial distribution in dumpsite soils of Sarcheshmeh copper mine.