Visible, Near-infrared And Short Wave Infrared Research Articles

Visible and infrared lab spectroscopy for soil texture classification: Analysis of entire spectra v/s reduced spectra

Soil, directly or indirectly, impacts seven of the Sustainable Development Goals set by the United Nations. Soil texture is a fundamental physical property that controls other properties ranging from soil stability, erodibility, compaction to water holding capacity, nutrient availability and fertility. While Visible Near-Infrared and Shortwave-Infrared (VNIR/SWIR) laboratory spectroscopy has been already used for soil texture classification, Mid Infrared (MIR) region and unison of the VNIR/SWIR and MIR region warrant quantification. Thus, the objectives of this work were: (1) to investigate the critical spectral regions for soil texture classification, (2) to compare the use of entire spectral bands v/s subset of spectral bands obtained via the feature selection method, and (3) to quantify the degree of misclassification in the neighbouring classes for a given textural class. This study utilized an ICRAF–ISRIC global soil spectral library consisting of 3643 VNIR/SWIR, MIR soil spectra and the USDA soil textural classes. Two classifiers were used for soil texture classification using the three databases (VNIR/SWIR, MIR, VNIR + MIR) with entire spectral bands and three databases with only the Partial Information Correlation (PIC) selected bands (VNIR/SWIR_PIC, MIR_PIC, VNIR + MIR_PIC). The mean confusion matrix, overall accuracy and kappa were calculated to evaluate classifiers' performances over 100 iterations. Two additional measures were proposed for partitioning inaccuracies between neighbouring and far classes: Correct-Neighbouring-Far classes distribution matrix and added neighbourhood accuracy. This work highlighted that: (1) the combined VNIR/SWIR + MIR region provided the best texture classification, followed by MIR and then VNIR/SWIR region, (2) the use of PIC selected bands provided lower classification performances when compared to using all bands but a massive reduction in the number of bands allowing to reduce model complexity, (3) the misclassifications were predominantly in the neighbouring classes rather than far classes, for a given texture class. Finally, the textural classes with poor classification performance had low areal and sample representation. These insights may launch a number of soil texture classification studies worldwide based on global soil spectral databases, further helping map soils and the associated ecosystem services it provides.

Prediction of lead in agricultural soils: An integrated approach using machine learning, terrain attributes and reflectance spectra

The study aimed to predict lead (Pb) concentrations in agricultural soils by analyzing visible, near-infrared, and shortwave infrared (VNIR-SWIR) spectra and terrain attributes under three distinct approaches. Approach 1 relied solely on the VNIR-SWIR data, with the cubist_FD (cubist_first derivative) modeling technique as the most effective, achieving a coefficient of determination (R2) of 0.63, a concordance correlation coefficient (CCC) of 0.51, a mean absolute error (MAE) of 6.87 mg kg-1, and a root mean square error (RMSE) of 8.66 mg kg-1. Approach 2 combined VNIR-SWIR spectra and terrain attributes, where SNV (standard normal variate) preprocessing technique and extreme gradient boosting (EGB) yielded superior results with R2 of 0.75, CCC of 0.65, MAE of 5.48 mg kg-1, and minimal RMSE of 7.34 mg kg-1. Approach 3 solely employed terrain attributes, and the combination with cubist modeling demonstrated the best prediction results, with R2 of 0.75, CCC of 0.66, MAE of 6.18 mg kg-1, and RMSE of 7.71 mg kg-1. The cumulative assessment identified that the fusion of terrain properties, SNV-transformed VNIR-SWIR spectra, and EGB as the optimal method for estimating Pb concentrations in agricultural soil yielded the highest R2 value and minimal error. Comparatively, Gaussian process regression (GPR), artificial neural network (ANN), and support vector machine (SVM) techniques achieved high R2 values in approaches 2 and 3 but also exhibited higher estimation errors. In conclusion, the study underscores the significance of using relevant auxiliary datasets and appropriate modeling techniques for accurate Pb concentration predictions with minimal error in agricultural soils. The findings contribute valuable insights for developing successful soil management strategies based on predictive modeling.

Removal of environmental influences for estimating soil texture fractions based on ZY1 satellite hyperspectral images

Soil texture is one of the important factors affecting the physical and chemical properties of soil, and is generally divided into sand, silt, and clay. Understanding its spatial distribution is crucial for agricultural and environmental decision-making. Estimation of soil texture fractions (STFs) can be achieved through visible, near-infrared, and shortwave infrared (VNIR-SWIR) spectra, making hyperspectral imaging technology particularly promising for low-cost and large-scale monitoring of STFs. However, the process of image acquisition is inevitably affected by soil environmental factors. In this study, a soil texture fractions estimation method for hyperspectral images was proposed based on Orthogonal Signal Correction (OSC). Firstly, the OSC algorithm was applied to eliminate the impact of soil environmental factors on the image spectra. Then, the spectral feature was enhanced by Fractional Order Derivative (FOD). Subsequently, a Partial Least Squares Regression (PLSR) model was established based on the optimal bands selected by Competitive Adaptive Reweighted Sampling (CARS). A total of 107 soil samples were obtained from the agricultural area in Lishu County, Jilin Province, China. The hyperspectral images of the ZY1-02D and ZY1-02E satellites were acquired simultaneously in this area. The results showed that the OSC algorithm effectively reduced the influence of soil environmental factors on image spectra, and significantly improved the accuracy of STFs estimation. After applying OSC to the original image spectra, the estimation accuracies (R2) for sand, silt, and clay content were improved from 0.52, 0.51, 0.50 to 0.81, 0.65, and 0.70 respectively. Compared to External Parameter Orthogonalization (EPO) correction, the model accuracy and stability were higher with OSC correction. The selected results from CARS indicated that the characteristic spectral bands of soil texture are concentrated around 500 nm, 750 nm, 900 nm, and 2200 nm. Further analysis indicates that the distribution trends of the soil texture fractions in this study were consistent with previous studies, demonstrating the good applicability of the estimation method.

Predicting the Surface Soil Texture of Cultivated Land via Hyperspectral Remote Sensing and Machine Learning: A Case Study in Jianghuai Hilly Area

Soil reflectance spectra and hyperspectral images have great potential to monitor and evaluate soil texture in large-scale scenarios. In hilly areas, sand, clay, and silt have similar spectral characteristics in visible, near-infrared, and short-wave infrared (VNIR-SWIR) reflection spectra. Soil texture spectra belong to mixed spectra despite some differences in particle size, mineral composition, and water content, making their distinction difficult. The accurate identification of the content within different particle sizes is difficult as it involves capturing spectral reflection features. Therefore, this study aimed to predict soil texture content through machine learning and unmixing the soil texture’s spectra while also comparing their respective modelling performances. Taking typical cultivated land in the Jianghuai hills as an example, the GaoFen-5 Advanced Hyperspectral Imaging (GF-5 AHSI) laboratory spectra of soil samples were used to predict sand, silt, and clay particle contents using partial least squares regression (PLSR) and convolutional neural networks (CNNs). The entire spectra of VNIR-SWIR regions were smoothed, and the dimensions were reduced via principal component analysis (PCA). The prediction models of sand, silt, and clay particle content were constructed, and inversion maps were generated using AHSI. The results showed that the PCA-CNN model achieved a higher prediction precision than the PCA-PLSR in both ASD and GF-5 data. Clay content exhibited the highest predictive performance with a coefficient of determination (R2) of 0.948 and 0.908 and a root mean square error (RMSE) of 26.51 g/kg and 31.24 g/kg, respectively, which represented a 39.0% and 79.8% increase in R2 and a 57% and 57.1% decrease in RMSE compared to that of the PCA-PLSR. This method indicates that the PCA-CNN model can effectively achieve nonlinear interactions between multiple spectral components and better model and fit spectral mixing processes; moreover, it provides an alternative method for investigating the spatial distribution of soil texture.

Spectral mapping of zeolite bearing paleolake deposits at Lake Tecopa, California and its implications for mapping zeolites on Mars

• Zeolitic tuff at paleolake Tecopa investigated using in-situ and orbital methods. • Reasons for limited orbital detections of zeolites in paleolake. • Importance of ground investigation to interpreting orbital data. While non-analcime zeolites are plausible mineral constituents of the Martian surface, their lack of prominent spectral characteristics in the visible to shortwave infrared wavelength region makes them difficult to identify using orbital data. This lack of detection has led to the assumption that they are limited or absent. Here we assess the capability of mapping non-analcime zeolites using different mapping techniques and present the possible reasons that could lead to limited detection of zeolites using an analog study from Paleolake Tecopa, California. We used spaceborne hyperspectral Hyperion and multispectral ASTER data to identify zeolites following X-ray Diffraction (XRD) and visible-near infrared and shortwave infrared (VNIR-SWIR) spectral analysis of collected samples containing zeolites and other associated minerals. We used four of the most common spectral mapping methods: (1) band ratios, (2) minimum noise fraction (MNF), (3) spectral angle mapper (SAM), and (4) linear spectral unmixing (LSU). We were able to map zeolite-rich tuff beds using carefully selected band indices, MNF band combinations, and the pixel spectra corresponding to a zeolite-rich abandoned quarry area. Hyperion results show less overall accuracy than the ASTER image products mostly due to the low signal-to-noise ratio of the Hyperion data. The spectral characteristics of zeolites in zeolite-rich pixels are masked by phyllosilicates (e.g., smectite) present in the ground resolution cell (GRC), making it difficult to identify zeolites in phyllosilicate-bearing deposits solely based on zeolite spectra. Therefore, the SAM product derived using the pixel spectrum of the well-exposed zeolite-rich quarry shows the highest accuracy in Hyperion image products. ASTER band ratio Band 2/Band 1, representing the ferric ion, was able to identify the zeolite-rich beds with the highest accuracies. However, most of the zeolite-rich paleolake beds were difficult to identify since they are buried by other beds, mixed with other beds due to physical weathering, or the areas of exposed beds are smaller than the GRC of the satellite. The results imply that the paucity of detected zeolites on Mars, as we see at Lake Tecopa, does not preclude their wider presence, either beneath other materials, obscured by surface dust, or mixed with more spectrally dominant phases. :

An integrated approach for rapid exploration of carbonatites and related mineral resources

AbstractCarbonatites are proven significant repositories of several critical and strategic elements such as rare earth elements, niobium, thorium, and uranium. Owing to their economic significance, mapping of carbonatites and associated mineral deposits has occupied prominent place in mineral resource exploration programs. In this study an integrated approach was developed to map carbonatite and related mineral deposits in the Loe‐Shilman, Northwest Himalaya of Pakistan, using remotely sensed advance space‐borne thermal emission and reflection radiometer (ASTER) multispectral data and visible near infrared and short‐wave infrared (VNIR‐SWIR) spectral characteristics of minerals in these deposits. Several image enhancement techniques, including band ratio (i.e., B4/B3), principal component and minimum noise fraction transformation (PC6 and MNF5, respectively) helped in highlighting the targeted rocks. The results demonstrate the suitability of ASTER data for discriminating carbonatite related mineral deposits from other surrounding lithologies. Results obtained from these methods were validated through field observations in the area and further confirmed through petrographic and chemical analyses of collected specimens. Field data have also served as training data to perform a supervised classification, allowing further improvement of the mapping results. Moreover, the obtained results from the techniques used for exploring carbonatites and related mineral deposits were stacked together for comparison with each other, to check their sensitivities, and assess their efficiency and accuracy. Generally, all these methods successfully highlighted carbonatites and related mineral deposits; however, when used integratively they exhibit higher degree of accuracy, and has proven to be relatively rapid and cost‐effective.

The use of Vis-NIR-SWIR spectroscopy in the prediction of soil available ions after application of rock powder

Some of the problems attributed to traditional laboratory analyses that limit the correct assessment of the nutrient contents in the soil are time requirements and high cost of the soil nutrient determinations. To solve these problems, a study was carried out to evaluate the use of visible, near-infrared, and short-wave infrared (Vis-NIR-SWIR) spectroscopy in the prediction of soil available ions submitted to the application of rock powders. The study was carried out on an Arenosol in Paranavaí City/Brazil. Treatments (rock powders) were arranged within a split-plot system designed in randomized blocks with four repetitions. Sugarcane was cultivated for 14 months after the application of rock powders. Later, 96 soil samples were collected for measuring the pH and available ions P, K+, Ca2+, Mg2+, S-SO42-, Si, Cu2+, Fe2+, Mn2+, and Zn2+ as well as spectral reading through a Vis-NIR-SWIR spectroradiometer to predict the soil chemical attributes through the partial least square regression (PLS) technique. The results showed that the elements K+, Ca2+, Mg2+, Cu2+, and Fe2+ could be predicted with a reasonable rightness degree (R2p > 0.50, RPDp > 1.40) from spectral models. However, for the attributes pH, P, S-SO42-, Si, Mn2+, and Zn2+, there were no satisfactory models (R2p < 0.50, RPDp < 1.40). Thus, the application of rock powder changed the spectral curves and, because of that, allows the building of PLS models to predict the elements K+, Ca2+, Mg2+, Cu2+, and Fe2+. Therefore, Vis-NIR-SWIR spectroscopy is a promising alternative to the routine analyses of soil fertility since it has advantages such as fast analytical speed, low cost, easy to operate, non-destructive, and environmentally friendly, because it does not use harmful chemicals.

Performance of hyperspectral data in predicting and mapping zinc concentration in soil

Reflectance spectroscopy in visible, near-infrared, and short-wave infrared (VNIR-SWIR) region has been recognized as a promising alternative for prediction of heavy metal concentration in soil. Compared with VNIR-SWIR reflectance spectroscopy, VNIR reflectance spectroscopy is less affected by atmospheric water vapor and has relatively high signal to noise ratio. The performances of VNIR and VNIR-SWIR hyperspectral data in predicting and mapping heavy metal concentration in soil were explored. In this study, laboratory spectra of soil samples collected from an agricultural area and Advanced Hyperspectral Imaging (AHSI) remote sensing imagery were used to predict and map zinc (Zn) concentration with genetic algorithm and partial least squares regression (GA-PLSR). The entire spectral regions of VNIR-SWIR and VNIR and spectral subsets extracted from the entire spectral regions were used in the prediction. For the laboratory spectra, the combination of the spectral bands extracted from the absorption features at 500 nm and in 600–800 nm obtained the highest prediction accuracy with the root mean square error (RMSE) and coefficient of determination (R2) values of 8.90 mg kg−1 and 0.72. For soil spectra from AHSI remote sensing imagery, the highest prediction accuracy was achieved by using the spectral bands extracted from the absorption feature in 600–800 nm with the RMSE and R2 values of 9.02 mg kg−1 and 0.75. Soil Zn concentration maps were generated with the established prediction models using AHSI remote sensing imagery. Analysis on the Zn concentration maps shows that the prediction model established using the spectral bands extracted from the absorption feature in 600–800 nm has a better performance in mapping Zn concentration. The results indicate that VNIR hyperspectral data outperforms VNIR-SWIR hyperspectral data in predicting and mapping Zn concentration in soil, which provides an alternative to the application of hyperspectral data in soil science.

Soil texture prediction through stratification of a regional soil spectral library

Knowing the spatial distribution of soil texture, which is a physical property, is essential to support agricultural and environmental decision making. Soil texture can be estimated using visible, near infrared, and shortwave infrared (Vis-NIR-SWIR) spectroscopy. However, the performance of spectroscopic models is variable because of soil heterogeneity. Currently, few studies address the effects of soil sample variability on the performance of the models, especially for larger spectral libraries that include soils that are more heterogeneous. Therefore, the objectives of this study were to: i) apply Vis-based color parameters on the stratification of a regional soil spectral library; ii) evaluate the performance of the predictive models generated from the spectral library stratification; iii) compare the performance of stratified models (SMs) and the model without stratification (WSM), and iv) explain possible changes in prediction accuracy based on the SMs. Thus, a regional soil spectral library with 1 535 samples from the State of Santa Catarina, Brazil was used. Soil reflectance data were obtained by Vis-NIR-SWIR spectroscopy in the laboratory using a spectroradiometer covering the 350–2 500 nm spectral range. Sand, silt, and clay fractions were determined using the pipette method. Twenty-two components of color parameters were derived from the Vis spectrum using the colorimetric models. A cubist regression algorithm was used to assess the accuracy of the applicability of the initial models (SMs and WSM) and of the validation between the clusters. Fractional order derivatives (FODs) at 0.5, 1.5, and 2 intervals were used to explain possible changes in the performance of the SMs. The SMs with higher contents of clay and iron oxides obtained the highest accuracy, and the most important spectral bands were identified, mainly in the 480–550 and 850–900 nm ranges and the 1 400, 1 900, and 2 200 nm bands. Therefore, stratification of soil spectral libraries is a good strategy to improve regional assessments of soil resources, reducing prediction errors in the qualitative determination of soil properties.

Study on the Pretreatment of Soil Hyperspectral and Na+ Ion Data under Different Degrees of Human Activity Stress by Fractional-Order Derivatives

Soluble salts in saline soil often exist in the form of salt base ions, and excessive water-soluble base ions can harm plant growth. As one of the water-soluble base ions, Na+ ion, is the main indicator of the degree of soil salinization. The pretreatment of visible, near infrared and short-wave infrared (VNIR-SWIR) spectroscopy data is the key to establishing a high-precision inversion model, and a proper pretreatment method can fully extract the effective information hidden in the hyperspectral data. Meanwhile, different degrees of human activity stress will have an impact on the ecological environment of oases. However, there are few comparative analyses of the data pretreatment effects for soil water-soluble base ions on the environment under different human interference conditions. Therefore, in this study, the difference in the degree of soil disturbance caused by human activities was used as the basis for dividing the experimental area into lightly disturbed area (Area A), moderately disturbed area (Area B) and severely disturbed zone (Area C). The Grünwald-Letnikov fractional-order derivative (FOD) was used to preprocess the VNIR-SWIR spectroscopic data measured by a FieldSpec®3Hi-Res spectrometer, which could fully extract the useful information hidden in the FOD of the VNIR-SWIR spectroscopy results and avoid the loss of information caused by the traditional integer-order derivative (1.0-order, 2.0-order) pretreatment. The spectrum pretreatment was composed of five transform spectra (R, R, 1/R, lgR, 1/lgR) and 21 FOD methods (step size is 0.1, derivative range is from 0.0- to 2.0-order). In addition, this manuscript compares and analyzes the pretreatment advantages between fractional-order and integer-order. The main results were as follows: (1) Grünwald-Letnikov FOD can reveal the nonlinear characteristics and variation laws of the field hyperspectral of saline soil, namely, due to the continuous performance of the order selection, the FOD accurately depicts the details of spectral changes during the derivation process, and improves the resolution between the peaks of the hyperspectral spectrum. (2) There is a big difference in the shape of the correlation coefficient curve between the original hyperspectral and Na+ at different FOD. The correlation coefficient curve has a clear outline in rang of the 0.0- to 0.6-order, and the change trend is gentle, which presents a certain gradual form. With the continuous increase of the order of the FOD, the change range of the correlation coefficient curve is gradually increased, and the fluctuation is greater between the 1.0-order and the 2.0-order. (3) Regardless of the transformation spectrum and different interference regions, the improvement effect of the FOD on the correlation between hyperspectral and Na+ is significantly better than that of the integer-order derivative. Comparative analysis shows that he percentage of increase of the former is more than 3%, and the highest is more than 17%.

Estimation of leaf area index and leaf chlorophyll content in Sporobolus densiflorus using hyperspectral measurements and PROSAIL model simulations

ABSTRACT Previous studies have shown the potential of remote-sensing tools to monitor coastal wetlands at a landscape scale. Several biophysical parameters are typically used to evaluate ecosystem conditions such as the chlorophyll content, an indicator of photosynthesis activity. In natural environments characterized by a large fraction of standing litter – such as the marshes on the Atlantic coast of the Buenos Aires Province – spectral indices designed for green vegetation change their typical response and, hence, their biophysical interpretation requires more attention. In this work, a theoretical study was performed to determine if it is possible to detect and eventually quantify the abrupt reductions of leaf chlorophyll content () in Sporobolus densiflorus – the dominant vegetation in these marshes – using hyperspectral data. To achieve this, in situ radiometric measurements in the VIS-NIR-SWIR (Visible, Near-Infrared and Short Wave Infrared) spectral region and biological data, acquired over S. densiflorus specimens in several campaigns, were used to set up an inversion procedure based on the radiative transfer model PROSAIL. By applying this model, simulated reflectances that fit the measured reflectances were obtained and by means of this inversion a theoretical canopy reflectance data set for S. densiflorus was modelled using the PROSAIL parameters. The performance of several vegetation indices typically used to estimate chlorophyll content was studied using the simulated and modelled reflectances, among which MTCI (MERIS Terrestrial Chlorophyll Index), Macc and MCARI/OSAVI (Modified Chlorophyll Absorption in Reflectance Index/Optimized Soil-Adjusted Index) indices showed significant correlation with . By means of addressing the performance of these indices together with BLRs (Baseline Residuals), a two-step VI-LUT (Vegetation Index – Look-Up Table) inversion model was proposed to retrieve , which first corrects the effect of variable leaf area index (LAI) using a BLR index (using bands at 800, 1100 and 1300 nm), and then retrieves by either using MCARI/OSAVI or another BLR (using bands at 657, 672 and 700 nm). Even though both procedures perform similarly well in the estimation of (coefficient of determination about 0.6) the two steps BLR-based approach is preferred, given that these indices are a priori less sensitive to undesired atmospheric effects.

Field characterisation of rock art paintings using non-invasive reflectance spectroscopy in the search for organic paint binders at Genealogy and Stickman Rockshelters in the Weld Range (Western Australia)

This paper reports on the search for organic binder materials in rock art paintings at Genealogy and Stickman Rockshelters in Wajarri Yamaji country, Weld Range (Western Australia). Portable visible, near infrared and shortwave infrared (VNIR-SWIR) reflectance spectroscopy requires no physical sampling or surface preparation. Eleven motifs were analysed in-situ. Only one of the motifs had absorption features characteristic of organic molecular bonds and resembled hematite and blood experimental paint reference materials. Constituent hematite and clay pigment minerals defined the spectra of the remaining ten motifs. This pilot study demonstrates that non-invasive VNIR-SWIR spectroscopy is an effective field characterisation tool for assessing painted rock art motifs for constituent organic materials and mineralogy including clay minerals and iron oxides. Without the need to take samples, the technique will be invaluable for rock art conservation, and as a screening tool in the search for organic materials suited for C14 direct dating in mineral pigmented rock art paint.

Simultaneous prediction of soil properties from VNIR-SWIR spectra using a localized multi-channel 1-D convolutional neural network

The use of visible near-infrared and shortwave-infrared (VNIR-SWIR) diffuse reflectance spectroscopy for the estimation of soil properties is increasingly maturing with large-scale soil spectral libraries (SSLs) of laboratory spectra developed across the globe. Such an SSL is the publicly available LUCAS topsoil database with approximately 20,000 soil samples encompassing 23 countries of the European Union. A wide variety of machine learning tools have been applied to the LUCAS SSL to predict some of the soil samples’ physicochemical properties with different degrees of accuracy. In this paper, we developed and examined the use of a novel one-dimensional convolutional neural network (CNN) to simultaneously predict ten physicochemical properties of the LUCAS SSL. Leveraging on the use of multiple-input channels it uses as model inputs the absorbance spectra along with some pre-processed spectra developed using standard techniques. Moreover, it exploits the use of local spectral neighborhoods to perform an adaptive error-correction mechanism. This novel localized multi-channel 1-D CNN was applied to all the available physicochemical properties of the LUCAS SSL and was statistically compared with the current state-of-the-art where it was shown to statistically outperform its counterparts, as well as with other CNNs where it exhibited the best performance. In particular, for the mineral soil samples, the RMSE for the Clay content was 4.80% (R2 0.86), for soil organic carbon the RMSE was 10.96 g kg−1 (R2 0.86), while for total nitrogen the RMSE was 0.66 g kg−1 (R2 0.83).

Removing Moisture Effect on Soil Reflectance Properties: A Case Study of Clay Content Prediction

Visible, near-infrared and shortwave-infrared (VNIR-SWIR) spectroscopy is an efficient approach for predicting soil properties because it reduces the time and cost of analyses. However, its advantages are hampered by the presence of soil moisture, which masks the major spectral absorptions of the soil and distorts the overall spectral shape. Hence, developing a procedure that skips the drying process for soil properties assessment directly from wet soil samples could save invaluable time. The goal of this study was twofold: proposing two approaches, partial least squares (PLS) and nearest neighbor spectral correction (NNSC), for dry spectral prediction and utilizing those spectra to demonstrate the ability to predict soil clay content. For these purposes, we measured 830 samples taken from eight common soil types in Israel that were sampled at 66 different locations. The dry spectrum accuracy was measured using the spectral angle mapper (SAM) and the average sum of deviations squared (ASDS), which resulted in low prediction errors of less than 8% and 14%, respectively. Later, our hypothesis was tested using the predicted dry soil spectra to predict the clay content, which resulted in R2 of 0.69 and 0.58 in the PLS and NNSC methods, respectively. Finally, our results were compared to those obtained by external parameter orthogonalization (EPO) and direct standardization (DS). This study demonstrates the ability to evaluate the dry spectral fingerprint of a wet soil sample, which can be utilized in various pedological aspects such as soil monitoring, soil classification, and soil properties assessment.

Using interpretable fuzzy rule-based models for the estimation of soil organic carbon from VNIR/SWIR spectra and soil texture

In this paper, the use of a novel evolutionary fuzzy rule-based system (FRBS) for the prediction of Soil Organic Carbon from visible, near-infrared, and short-wave infrared (VNIR/SWIR) spectra and the textural information as additional predictor is examined. Compared to other techniques, the proposed model generates a compact set of rules with a high interpretation degree, mapping local input to local output regions. This is achieved through an evolutionary learning procedure which is applied to establish linguistic rules and assist in the interpretation of the association between spectra and the target property. The rule base may be also decomposed into texture-specific sets of rules, allowing a more detailed analysis on a per textural class basis. These intrinsic properties enable the development of spectral prototype signatures and sparse feature utilization histograms at different levels of aggregation, i.e. per textural class and/or output region. The proposed model is applied to the LUCAS topsoil database comprised of roughly 18,000 mineral samples across 23 European Union member-states. We first demonstrate the enhanced interpretation capabilities of our fuzzy approach, which can assist in the extraction of fruitful knowledge governing the association between soil properties and VNIR/SWIR spectra. The model is then compared with other contemporary approaches, namely PLS, SVM, and Cubist. The results indicate that our approach produced compact and interpretable results with fair prediction accuracies (equivalent with the best approach).

Time takes its toll: Detection of organic binder media in ochre paints with visible near-infrared and short-wave infrared reflectance spectroscopy

Organic binder media in ochre rock art have so far resisted in-situ spectroscopic analysis. In this study visible near-infrared and short-wave infrared (VNIR-SWIR) spectroscopy was investigated for its usefulness in characterising experimental paints made with ochre and local (Weld Range, Western Australia) organic materials. VNIR-SWIR spectroscopy is a proven technique for the detection of mineral molecular compounds. Historic art conservators have successfully detected organic binder materials in 500-year old Renaissance art with near-infrared (NIR) spectroscopy. In this study, the chemical morphological spectral features of organic materials in newly made ochre paints, and UV accelerated 100-year aged ochre paints, were investigated. The analysis revealed complexities that would affect the detection of organic materials as anthropogenic in rock art. The first of these concerned the nature of the organic materials. There were significant deteriorative spectral changes caused by reactions initially with hematite (red ochre), and additional changes caused by accelerated UV light, resulting in remnant features different from the raw materials. These would render current references and databases useless for identification. The spectroscopic identification of organic binders in old rock art may depend on such remnants. Secondly, the composition of naturally occurring red ochre included organic materials as natural constituents, confounding the potential to differentiate those added by humans. Finally, this study has challenged the assumption that all organic materials, in particular animal fats, might function as binders to chemically adhere pigment. This paper presents a novel approach to the characterisation and identification of the constituents of ochre and organic paints, using VNIR-SWIR spectroscopy.

A hyperspectral study (V-NIR-SWIR) of the Nechalacho REE-Nb-Zr deposit, Canada

This paper reports the results of a visible, near-infrared and short wave infrared (V-NIR-SWIR) spectroscopic study of the Nechalacho rare earth element (REE)-Nb-Zr deposit (Canada), associated rare metal pegmatites and select samples from other REE deposits. The zoning of minerals produced by hydrothermal alteration is traced using V-NIR-SWIR spectroscopic measurements in combination with geological, geochemical and mineralogical data. The minerals detected using V-NIR-SWIR spectroscopy (analcime, ankerite, biotite, chlorite, illite, REE-fluorocarbonates) display complex spatial distributions that were strongly controlled by the pathways and composition of multiple stages of hydrothermal fluids as well as the precursor lithology. Magmatic-hydrothermal analcime ± amphibole alteration was overprinted by regional biotite + magnetite + illite + ankerite + chlorite + REE-fluorocarbonate replacement. A qualitative mass change analysis indicates that the latter alteration (main-stage) was associated with significant gains in Mg and losses of Na at temperatures of ~300 °C, as indicated by the presence of phengitic to muscovitic illite. Low-temperature ankerite ± chlorite alteration was mainly fluid-controlled and associated with the formation of hydrothermal REE-fluorocarbonate minerals that crystallized as a result of the interaction of aqueous-carbonic hydrothermal fluids with eudialyte syenite protoliths, locally leading to leaching and re-deposition of light REE. The ankerite ± chlorite alteration also affected pegmatites peripheral to the Nechalacho intrusion and surrounding hydrothermally altered zones. Analysis of a variety of REE-mineral specimens and geochemically characterized bulk rock samples showed that the relative depths of the 741 and 864 nm NIR absorption features of Nd can be used as qualitative and quantitative proxies for the detection of the REE. In combination with fast measurement and processing capabilities, this allows field-based V-NIR-SWIR spectroscopy to be deployed as a powerful tool for regional and deposit-scale REE exploration, deposit characterization and mine production grade control.

Evaluating the effect of spatial subsetting on subpixel unmixing methodology applied to ASTER over a hydrothermally altered terrain

This study investigates the effect of spatial subsets of Advanced Spaceborne Thermal Emission and Reflection radiometer (ASTER) L1B visible-near infrared and short wave-infrared (VNIR-SWIR) data on matched filtering results at the central part of Kerman magmatic arc, where abundant porphyry copper deposits exist. The matched filtering (MF) procedure was run separately at sites containing hydrothermal minerals such as sericite, kaolinite, chlorite, and jarosite to map the abundances of these minerals on spatial subsets containing 100, 75, 50, and 25 percent of the original scene. Results were evaluated by comparing the matched filtering scores with the mineral abundances obtained by semi-quantitative XRD analysis of corresponding field samples. It was concluded that MF method should be applied to the whole scene prior to any data subsetting.

Inference model to predict heavy metals of Bahr El Baqar soils, Egypt using spectroscopy and chemometrics technique

With 35 soil samples gathered from the Bahr El Baqar region, Egypt, the objective of this study was to propose an inference model for prediction of heavy metals (Cd, Cu, Pb, Zn, Ni, Mn, Cr, Co, Fe) with visible near-infrared and short wave infrared (VNIR-SWIR) region (350–2500 nm). For a better understanding of the mechanism that allows the estimation of heavy metals with reflectance spectroscopy, statistical analysis was first made. The fingerprint region of 538, 578, 630, 870, 1900, 2240, and 2376 nm is very useful in recognizing small differences in the heavy metals of Bahr El Baqar region. Partial least squares regression (PLSR) is able to some extent (moderate accuracy) to model heavy metals with laboratory spectra parameters. The best coefficients of determination (R 2) between predicted and chemically analyzed concentrations were for Mn, 0.62; Pb, 0.66; Zn, 0.66, Ni, 0.69, and Cu, 0.60. Iron influences the soil reflectance in the VNIR-SWIR region. This is due to the electronic transition of iron cations (2+, 3+). Mn, Pb, Ni, Cu, and Zn were successfully predicted using PLSR. Only Cr (0.59) and Cd (0.52) content was predicted fairly. The analysis of correlation between heavy metal and soil constitutes of Fe2O3, Al2O3 and OM, which represented the clay minerals, iron oxides and organic matter, respectively, could support the above predicated binding forms of heavy metals. The order of the correlation coefficients from high to low between metal and wavelength is Cu > Zn > Mn > Co > Pb > Cr > Ni > Cd. This is almost the same as the order of their correlation coefficients with Fe2O3, Al2O3 and OM. These findings also validate a mechanism to predict heavy metals that have no absorption features in reflectance spectra. The results concerning the relationships between spectral parameters and heavy metals concentration indicate that iron oxides, clay, and OM play an important role in the prediction of soil heavy metals using reflectance spectra. To identify the specific wavelengths for prediction of the studied heavy metals, the regression coefficient with OM, Al2O3, and Fe2O3 were used. The most significant peaks for Zn prediction are found in the wavelengths of 430, 570, and 1700 nm; for Cu prediction at 538, 1259, 1500, and 2184 nm; for Pb significant peaks are at 440, 578, 915, and 1894 nm; for Cr at 860, 1001, and 2376 nm; for Ni at 870, 909, and 2240 nm; for Cd at 630, and 1270 nm; for Co at 1900, and 2300 nm. As the current results suggested, large concentrations of heavy metals can be predicted using reflectance spectroscopy. However, it is much more difficult to determine small concentrations of heavy metals in soils of Bahr El Baqar region. Organic matter masks spectral signatures, making it difficult to identify metal quantities at these wavelengths. So, a new concept “inference approach” in light of key soil constituents that absorb VNIR-SWIR radiation would became a powerful tool to enhance the accuracy of heavy metal prediction. Due to the interference effects of organic matter and low heavy metal concentration, the selection of the most sensitive original spectral bands still did not result in highly accurate prediction. Therefore, it is important to predict heavy metals through OM, iron oxides, or clays first. These results indicate that it is possible to predict heavy metals in soils using the rapid and economic reflectance spectroscopy.

Semi-blind source separation for the estimation of the clay content over semi-vegetated areas using VNIR/SWIR hyperspectral airborne data

Abstract Visible, near-infrared and short wave infrared (VNIR/SWIR) hyperspectral imagery has proven to be a useful technique for mapping the soil surface properties over bare soils pixels. Multivariate regression models are usually built linking a set of soil surface properties (response Y-variables) to a set of imaging reflectance spectra over bare soil pixels (predictor X-variables), and then, they are applied to all bare soil pixels to map the soil surface properties. The applicability of VNIR/SWIR hyperspectral imagery for soil properties mapping decreases when surfaces are partially covered by vegetation. The objective of this research was to develop a “Double-Extraction” approach for clay content estimation over semi-vegetated surfaces and to evaluate its performance using VNIR/SWIR HyMap airborne data acquired in a Mediterranean region over an area of 24 km2. The “Double-Extraction” approach consists of 1) an extraction of a soil reflectance spectrum, s ^ soil , using a semi-blind source separation (SBSS) technique applied to couples of semi-vegetated spectra and 2) an extraction of clay content from the soil reflectance spectrum s ^ soil using a multivariate regression method. The source separation approach is semi-blind due to the use of available knowledge about expected soil and vegetation spectra. The multiplicative algorithm of Lee & Seung, belonging to the family of non-negative matrix factorization (NMF) methods, is used to solve the blind source separation (BSS) problem. The multivariate regression method used in this study is the partial least squares regression (PLSR) method. The “Double-Extraction” approach was compared to a “Direct” approach consisting of the application of the multivariate regression model built from bare soil spectra over the semi-vegetated area. Our results showed poor prediction performances for both approaches when applied to all pixels; however, a slight improvement was observed when correcting the bias prediction that occurs when using the PLSR model. Conversely, satisfactory prediction performances were obtained by restricting the prediction to the weakly vegetated area (NDVI