Search

Our study investigates the capability of the environmental mapping and analysis program (EnMAP) scenes simulated using the EnMAP end-to-end simulator software (EeteS) based on the AISA Eagle airborne data to predict chlorophyll-a (Chl-a) and total suspended solids (TSS) as two of the most crucial water quality indicators. Three machine learning (ML) approaches (principal component regression(PCR), partial least square regression (PLSR) and random forest (RF)) were employed to establish links between the simulated image spectra and the above-mentioned water attributes of the samples collected from several inland water reservoirs within the southern part of the Czech Republic. Airborne hyperspectral images were also used to develop a model to compare its performance with models developed based on the simulated EnMAP data. Adequate prediction accuracy was obtained for both Chl-a (R2 = 0.89, RMSE = 43.06 g/L, and Lin’s concordance correlation coefficient (LCCC) = 0.91) and TSS (R2 = 0.91, RMSE = 17.53 mg/L, and LCCC = 0.94), which were close enough to those obtained from the airborne hyperspectral images. Chl-a and TSS correlated with the wavelengths around 550 nm and 700 to 750 nm of the red and near-infrared (NIR) regions. In addition, the spatial distribution maps derived from the simulated EnMAP were comparable to those obtained from the AISA Eagle airborne data. Overall, it can be concluded that the simulated EnMAP image successfully and reliably predicted and spatially mapped the selected biophysical properties of the small inland water bodies.

The Environmental Mapping and analysis program (EnMAP) is a new Earth observation satellite which will use imaging spectroscopy to obtain a diagnostic characterisation of the Earth's surface and record changes. Since we hypothesis that imaging spectroscopy can significantly improve the accuracy of predicting and assessing water quality traits of small in-land waters, our study investigates the capability of the simulated EnMAP data to predict chlorophyll-a (Chl-a) and total suspended solids (TSS) as two of the most crucial water quality indicators. Three machine learning (ML) approaches(i.e., The methods used were Principal Component Regression(PCR), Partial Least Square Regression (PLSR) and Random Forest (RF)) were employed to establish links between the simulated image spectra and the above-mentioned water attributes of the samples collected from several in-land water reservoirs within the southern part of the Czech Republic. Additionally, an airborne hyperspectral image likewise was used for developing a model to compare its performance with models developed based on simulated EnMAP data. According to the results, adequate prediction accuracy was obtained for Chl-a (RF: R\(^2\) = 0.89, RMSE = 43.06, and LCCC = 0.91 ) and TSS (RF: R\(^2\) = 0.91, RMSE = 17.53, and LCCC = 0.94), which was close enough to those obtained for the airborne hyperspectral image. Chl-a and TSS showed a correlation with around 550 and 650 nm and from 700 nm to 800 nm of the red and near-infrared (NIR) regions.The spatial distribution maps derived from the simulated EnMAP were comparable to those obtained by the source image, particularly in water bodies with relatively high and low contents of the water attributes. Overall, it can be concluded that the simulated EnMAP image was successful and reliable in the prediction and spatial mapping of the selected biophysical properties of the small in-land water bodies.

Finding an appropriate satellite image as simultaneous as possible with the sampling time campaigns is challenging. Fusion can be considered as a method of integrating images and obtaining more pixels with higher spatial, spectral and temporal resolutions. This paper investigated the impact of Landsat 8-OLI and Sentinel-2A data fusion on prediction of several toxic elements at a mine waste dump. The 30 m spatial resolution Landsat 8-OLI bands were fused with the 10 m Sentinel-2A bands using various fusion techniques namely hue-saturation-value (HSV), Brovey, principal component analysis (PCA), Gram-Schmidt (GS), wavelet, and area-to-point regression kriging (ATPRK). ATPRK was the best method preserving both spectral and spatial features of Landsat 8-OLI and Sentinel-2A after fusion. Furthermore, the partial least squares regression (PLSR) model developed on genetic algorithm (GA)-selected laboratory visible-near infrared-shortwave infrared (VNIR-SWIR) spectra yielded more accurate prediction results compared to the PLSR model calibrated on the entire spectra. It was hence, applied to both individual sensors and their ATPRK-fused image. In case of the individual sensors, except for As, Sentinel-2A provided more robust prediction models than Landsat 8-OLI. However, the best performances were obtained using the fused images, highlighting the potential of data fusion to enhance the toxic elements' prediction models.

There is a need to update soil maps and monitor soil organic carbon (SOC) in the upper horizons or plough layer for enabling decision support and land management, while complying with several policies, especially those favoring soil carbon storage. This review paper is dedicated to the satellite-based spectral approaches for SOC assessment that have been achieved from several satellite sensors, study scales and geographical contexts in the past decade. Most approaches relying on pure spectral models have been carried out since 2019 and have dealt with temperate croplands in Europe, China and North America at the scale of small regions, of some hundreds of km2: dry combustion and wet oxidation were the analytical determination methods used for 50% and 35% of the satellite-derived SOC studies, for which measured topsoil SOC contents mainly referred to mineral soils, typically cambisols and luvisols and to a lesser extent, regosols, leptosols, stagnosols and chernozems, with annual cropping systems with a SOC value of ~15 g·kg−1 and a range of 30 g·kg−1 in median. Most satellite-derived SOC spectral prediction models used limited preprocessing and were based on bare soil pixel retrieval after Normalized Difference Vegetation Index (NDVI) thresholding. About one third of these models used partial least squares regression (PLSR), while another third used random forest (RF), and the remaining included machine learning methods such as support vector machine (SVM). We did not find any studies either on deep learning methods or on all-performance evaluations and uncertainty analysis of spatial model predictions. Nevertheless, the literature examined here identifies satellite-based spectral information, especially derived under bare soil conditions, as an interesting approach that deserves further investigations. Future research includes considering the simultaneous analysis of imagery acquired at several dates i.e., temporal mosaicking, testing the influence of possible disturbing factors and mitigating their effects fusing mixed models incorporating non-spectral ancillary information.

For decades, the Příbram district in the Czech Republic has been affected by industrial and mining activities. These activities are important sources of heavy metal pollutants that are detrimental to soil quality. A recent study examined visible–near-infrared (VNIR), shortwave-infrared (SWIR) and X-ray fluorescence (XRF) spectroscopy to model soil lead (Pb) content in a selected area located in the Příbram district. Following that study, and using the same chemical and geographical data, we examined the spatial distribution of Pb content in the soil, with a combination of different traditional spatial analyses (Moran’s I, hotspot analysis, and Kriging) that were significantly validated. One of the novel points of this work is the use of the Getis–Ord hotspot analysis before the execution of a Kriging interpolation model to better emphasize clustering patterns. The results indicated that Pb was a spatially dependent soil property and through extensive in situ sampling, it was possible to generate a very accurate Kriging interpolation model. The high-Pb hotspots coincided with topographic obstacles that were modeled using topographic profiles extracted from the open-source Google Earth platform, indicating that Pb content does not always exhibit a direct relationship with topographic height as a result of runoff, due to the contribution of topographic steps. This observation provides a new perspective on the relationship between Pb content and topographic patterns.

We conducted a systematic review and inventory of recent research achievements related to spaceborne and aerial Earth Observation (EO) data-driven monitoring in support of soil-related strategic goals for a three-year period (2019–2021). Scaling, resolution, data characteristics, and modelling approaches were summarized, after reviewing 46 peer-reviewed articles in international journals. Inherent limitations associated with an EO-based soil mapping approach that hinder its wider adoption were recognized and divided into four categories: (i) area covered and data to be shared; (ii) thresholds for bare soil detection; (iii) soil surface conditions; and (iv) infrastructure capabilities. Accordingly, we tried to redefine the meaning of what is expected in the next years for EO data-driven topsoil monitoring by performing a thorough analysis driven by the upcoming technological waves. The review concludes that the best practices for the advancement of an EO data-driven soil mapping include: (i) a further leverage of recent artificial intelligence techniques to achieve the desired representativeness and reliability; (ii) a continued effort to share harmonized labelled datasets; (iii) data fusion with in situ sensing systems; (iv) a continued effort to overcome the current limitations in terms of sensor resolution and processing limitations of this wealth of EO data; and (v) political and administrative issues (e.g., funding, sustainability). This paper may help to pave the way for further interdisciplinary research and multi-actor coordination activities and to generate EO-based benefits for policy and economy.

Soil contamination is a worldwide crisis, which diminishes food and agricultural production. Alterations in the soil environment due to soil contamination cause biophysical and biochemical changes in vegetation. Due to dynamic nature of these changes, early monitoring can permit for preventive interferences before intense and sometimes inevitable vegetation and soil problems occur. As plants are rooted in soil substrate, vegetation changes can be used as bio-indicators of soil conditions. Traditionally, vegetation changes have been usually determined by visual analysis or detected after major destructive sampling during the growth period. As the characteristics of vegetation influence its spectral properties, effective remote and non-contact detection methods offer an alternative and near real-time way for detecting plant changes, even prior to visual symptoms and negative effects appearance. The aim of the current study is to review the potential of optical proximal and remote sensing techniques at different platforms for indirect assessment of plant–soil interactions via monitoring vegetation anomalies related to soil contamination. It is strongly felt that this rapidly progressing technological direction will permit extending the use of the techniques to geology, soil science and precision agriculture and an overall broad range of applications.

Soil contamination by potentially toxic elements (PTEs) is intensifying under increasing industrialization. Thus, the ability to efficiently delineate contaminated sites is crucial. Visible–near infrared (vis–NIR: 350–2500 nm) and X-ray fluorescence (XRF: 0.02–41.08 keV) spectroscopic techniques have attracted tremendous attention for the assessment of PTEs. Recently, the application of fused vis–NIR and XRF spectroscopy, which is based on the complementary effect of data fusion, is also increasing. Moreover, different data manipulation methods, including feature selection approaches, affect the prediction performance. This study investigated the feasibility of using single and fused vis–NIR and XRF spectra while exploring feature selection algorithms for the assessment of key soil PTEs. The soil samples were collected from one of the most heavily polluted areas of the Czech Republic and scanned using laboratory vis–NIR and XRF spectrometers. Univariate filter (UF) and genetic algorithm (GA) were used to select the bands of greater importance for the PTE prediction. Support vector machine (SVM) was then used to train the models using the full-range and feature-selected spectra of single sensors and their fusion. It was found that XRF spectra alone (primarily GA-selected) performed better than single vis–NIR and fused spectral data for predictions of PTEs. Moreover, the prediction models that were derived from the fused data set (particularly the GA-selected) enhanced the models’ accuracies as compared with the single vis–NIR spectra. In general, the results suggest that the GA-selected spectra obtained from the single XRF spectrometer (for As and Pb) and from the fusion of vis–NIR and XRF (for Pb) are promising for accurate quantitative estimation detection of the mentioned PTEs.

Visible–near infrared–shortwave infrared (VNIR–SWIR) spectroscopy is being increasingly used for soil organic carbon (SOC) assessment. Common practice consists of scanning soil samples using a single spectrometer. Considerations have rarely been documented of the effects of using multiple instruments and scanning conditions on SOC model calibration that occur when merging soil spectral libraries (SSLs), particularly in soils with low SOC concentration and using both field spectroradiometers and laboratory fixed spectrometers. To address this gap, we scanned 143 low-SOC-content soil samples using three spectrometers (ASD FieldSpec 3, ASD FieldSpec 4, and FOSS XDS) and four setup features - FOSS, contact probe (CP), dark-box (DB), and open laboratory (LAB) - at three laboratories. The application of an internal soil standard (ISS) to align one laboratory spectrum with another for spectral correction and spectral merging of various SSLs was examined. SOC models were developed using i) data from each single spectrometer – single laboratory separately and ii) merged data from multiple spectrometers – different laboratories, applying the 1st derivatives of spectra and random forest (RF) regression. The results indicate that the spectral shape and wavelength position of key features obtained from all spectrometers and setups did not show any noticeable differences, though spectra based on FOSS setup, particularly on low-SOC samples, demonstrated greater range in absolute derivative values regardless of ISS application. The derivative ISS-corrected spectra showed less variation among different spectrometers compared to their uncorrected raw reflectance spectra. All single spectrometer models predicted SOC reasonably well. However, the spectra acquired by the FOSS setup predicted SOC more accurately (R2 = 0.77, RPIQ = 3.30, RMSE = 0.22 %, and SD = 0.04) than the spectra acquired by the other setups. The models derived from merged uncorrected raw reflectance spectra yielded poor results (R2 = 0.48, RPIQ = 2.33, RMSE = 0.33 %, and SD = 0.10); nevertheless, assessment of SOC using the 1st derivative ISS-corrected merged SSLs considerably improved the prediction accuracy (R2 = 0.70, RPIQ = 3.10, RMSE = 0.25 %, and SD = 0.06). Hence, the derivative spectra coupled with the ISS correction improved the accuracy of SOC prediction models obtained from the merged soil spectra collected in different environments using different instruments. We therefore recommend application of the ISS spectral alignment method linked to the 1st derivative approach to enhance the compilation of SSLs at the regional and global scales for SOC assessment.

Weathering and oxidation of sulphide minerals in mine wastes release toxic elements in surrounding environments. As an alternative to traditional sampling and chemical analysis methods, the capability of proximal and remote sensing techniques was investigated in this study to predict Chromium (Cr) concentration in 120 soil samples collected from a dumpsite in Sarcheshmeh copper mine, Iran. The samples’ mineralogy and Cr concentration were determined and were then subjected to laboratory reflectance spectroscopy in the range of Visible–Near Infrared–Shortwave Infrared (VNIR–SWIR: 350–2500 nm). The raw spectra were pre-processed using Savitzky-Golay First-Derivative (SG-FD) and Savitzky-Golay Second-Derivative (SG-SD) algorithms. The important wavelengths were determined using Partial Least Squares Regression (PLSR) coefficients and Genetic Algorithm (GA). Artificial Neural Networks (ANN), Stepwise Multiple Linear Regression (SMLR) and PLSR data mining methods were applied to the selected spectral variables to assess Cr concentration. The developed models were then applied to the selected bands of Aster, Hyperion, Sentinel-2A, and Landsat 8-OLI satellite images of the area. Afterwards, rasters obtained from the best prediction model were segmented using a binary fitness function. According to the outputs of the laboratory reflectance spectroscopy, the highest prediction accuracy was obtained using ANN applied to the SD pre-processed spectra with R2 = 0.91, RMSE = 8.73 mg/kg and RPD = 2.76. SD-ANN also showed an acceptable performance on mapping the spatial distribution of Cr using the ordinary kriging technique. Using satellite images, SD-SMLR provided the best prediction models with R2 values of 0.61 and 0.53 for Hyperion and Sentinel-2A, respectively. This led to the higher visual similarity of the segmented Hyperion and Sentinel-2A images with the Cr distribution map. This study’s findings indicated that applying the best prediction models obtained by spectroscopy to the selected wavebands of Hyperion and Sentinel-2A satellite imagery could be considered a promising technique for rapid, cost-effective and eco-friendly assessment of Cr concentration in highly heterogeneous mining areas.

Soil Carbon (C) is central to the functioning of ecosystems and climate change mitigation. It represents the largest terrestrial pool and much of it, is stored in forest soils. Soil Organic Carbon (SOC) in a forest varies not only laterally, but also vertically (i.e., with depth). However, the SOC content of forest soil horizons has not been investigated over large scales, despite its importance for framing our understanding of soil function. Visible–Near Infrared (vis–NIR) reflectance spectroscopy enables rapid and cost-effective examination of forest SOC distribution, both laterally and vertically. This study aims to evaluate the potential of vis–NIR spectroscopy for classifying and predicting the SOC concentration of organic and mineral horizons in forests of the Czech Republic. We investigated 1080 forest sites across the country, each with five soil horizons, representing the Litter (L), Fragmented (F), and Humus (H) organic horizons, as well as the A1 (depth of 2–10 cm) and A2 (depth of 10–40 cm) mineral horizons. We, then, used Support Vector Machines (SVMs) to classify the soil horizons based on their spectra and also to model the SOC concentration of (i) the profile (organic and mineral horizons together), (ii) only the organic horizons, (iii) only the mineral horizons, and (iv) each individual horizon separately. The models were validated using 10-repeated 10-fold cross-validation. Results show that the SVM with radial basis kernel could accurately classify the soil horizons (Correct Classification Rate (CCR) of 70% and Kappa coefficient of 0.63). The SOC model developed for the soil profile performed well (R2 = 0.76 and RMSE = 1.63%). The model of the combined organic horizons was considerably more accurate than that of the combined mineral horizons (R2 = 0.78 and R2 = 0.53, respectively). Estimates of SOC in the individual soil horizons had R2 values greater than 0.63 but those of the F and A1 models were better with R2 > 0.70. The study indicates that vis–NIR spectroscopy can effectively characterize the SOC concentration of the highly variable forest soil horizons in the Czech Republic.

Visible–near infrared–shortwave infrared (VIS–NIR–SWIR) spectroscopy is being increasingly used for soil organic carbon (SOC) assessment. Common practice consists of scanning soil samples using a single spectrometer. Considerations have rarely been documented of the effects of using multiple instruments and scanning conditions on SOC model calibration that occur when merging soil spectral libraries (SSLs), particularly in soils with low SOC concentration and using both field spectroradiometers and laboratory fixed spectrometers. To address this gap, we scanned 143 low-SOC-content soil samples using three spectrometers (ASD FieldSpec 3, ASD FieldSpec 4, and FOSS XDS) and four setup features - FOSS, contact probe (CP), dark-box (DB), and open laboratory (LAB) - at three laboratories. The application of an internal soil standard (ISS) to align one laboratory spectrum with another for spectral correction and spectral merging of various SSLs was examined. SOC models were developed using i) data from each single spectrometer – single laboratory separately and ii) merged data from multiple spectrometers – different laboratories, applying the 1st derivatives of spectra and random forest (RF) regression. The results indicate that the spectral shape and position obtained from all spectrometers and setups did not show any noticeable differences, though spectra based on FOSS setup, particularly on low-SOC samples, demonstrated the highest absolute derivative values regardless of ISS application. The derivative ISS-corrected spectra showed less variation among different spectrometers compared to their uncorrected reflectance spectra. All single spectrometer models predicted SOC reasonably well. However, the spectra acquired by the FOSS setup predicted SOC more accurately (R2 = 0.77 and SD = 0.04) than the spectra acquired by the other setups. The models derived from merged uncorrected reflectance spectra yielded poor results (R2 = 0.48 and SD = 0.10); nevertheless, assessment of SOC using the 1st derivative ISS-corrected merged SSLs considerably improved the prediction accuracy (R2 = 0.70 and SD = 0.06). Hence, the derivative spectra coupled with the ISS correction improved the accuracy of SOC prediction models obtained from merged soil spectra collected in different environments using different instruments. We therefore recommend application of the ISS spectral alignment method linked to the 1st derivative approach to enhance the compilation of SSLs at the regional and global scales for SOC assessment.

Forest ecosystems are among the main parts of the biosphere; however, they have been endangered from the significant elevation and harmful effects of air and soil pollutants, including potentially toxic elements (PTEs). The concentration of PTEs in forest soils varies not only laterally but also vertically with depth. Forest surface organic horizons are of particular interest in forest ecosystem monitoring due to their role as stable adsorbents of the deposited atmospheric substances. Therefore, the main purpose of this study was to conduct rapid examinations of forest soils PTEs (Cr, Cu, Pb, Zn, and Al), testing the capability of VIS--NIR spectroscopy coupled with machine learning (ML) techniques (partial least square regression (PLSR), support vector machine regression (SVMR), and random forest (RF)) and fully connected neural network (FNN), a deep learning (DL) approach, in forest organic horizons. One-thousand-and-eighty forested sites across the Czech Republic at two soil layers, defining the fragmented (F) and humus (H) organic horizons, were investigated (total 2160 samples). PTEs as well as total Fe and SOC, as auxiliary data, were conventionally and spectrally determined and modelled in the combined organic horizons (F + H) and in each individual horizon using the ML and DL algorithms. Results indicated that the concentration of all PTEs was higher in the horizon H compared to the F horizon. Although the spectral reflectance of samples tended to decrease with increased PTEs concentration. Strongly significant positive correlations between all PTEs and total Fe in all horizons were obtained, which were higher in the H and F + H horizons than the F horizon. The highest correlations of PTEs with the spectra were at 460--590~nm, which is mostly linked to the presence of Fe-oxide. These results show the importance of Fe for spectral prediction of PTEs. Cr and Al were the most accurately predicted elements, regardless of the applied learning technique. SVMR provided the best results in assessing the H horizon (e.g., R\(^2\) = 0.88 and root mean square error (RMSE) = 3.01~mg/kg, and R\(^2\) = 0.82 and RMSE = 1682.25~mg/kg for Cr and Al, respectively); however, FNN predicted the combined F + H horizons the best (R\(^2\) = 0.89 and RMSE = 2.95~mg/kg, and R\(^2\) = 0.86 and RMSE = 1593.64~mg/kg for Cr and Al, respectively) due to the larger number of samples. In the F horizon, almost no parameters were predicted adequately. This study shows that given the availability of larger sample sizes, FNN can be a more promising technique compared to ML methods for assessment of Cr and Al concentration based on national spectral data in the forests of the Czech Republic.

Any strategy to change the Carbon (C) pool has a substantial effect on the functionality of numerous ecosystem functions, the detachment of Soil Organic Carbon (SOC), the atmospheric carbon dioxide (CO2) concentration, and climate change mitigation. As the largest amount of the world's C is stored in forests soils, the importance of forest SOC management is highlighted. The total SOC in a forest varies not only laterally, but also vertically (i.e., with depth). However, the SOC storage of different forest soil horizons has not been investigated in a national scale thoroughly, despite their potential to frame our understanding of soil function. Visible--Near Infrared (vis--NIR) reflectance spectroscopy enables rapid examination of the horizontal distribution of forest SOC, overcoming the limitations of traditional soil assessment methods. This study aims to evaluate the potential of vis--NIR spectroscopy in characterizing and predicting the SOC content of organic and mineral horizons in forests. We investigate 1080 forested sites across the Czech Republic at five individual soil layers, representing the Litter (L), Fragmented (F), and Humus (H) organic horizons, as well as the A1 (depth of 2--10 cm) and A2 (depth of 10--40 cm) mineral horizons (for a total of 5400 samples). We, then, use Support Vector Machines (SVMs) to classify the soil horizons based on their spectra and also to predict the SOC content of (i) the profile (all organic and mineral horizons together), (ii) the combined organic horizons, (iii) the combined mineral horizons, and (iv) each individual horizon separately. The models are validated using 10-repeated 10-fold cross validation. The results show that there is at least more than seven times as much SOC in the combined organic horizons, compared to the combined mineral horizons, with more variation in the deeper layers. The SVM with radial based kernel is a reliable classifier for classification of soil horizons, with Correct Classification Rate (CCR) of 70% and Kappa coefficient of 0.63. All individual horizon SOCs are successfully predicted with low error and with R2 values higher than 0.63. However, the prediction accuracies of the F and A1 models are greater, compared to others (R2~0.70 and very low-biased spatial estimates). We conclude that the modelling of SOC with vis--NIR spectra in different soil horizons of highly heterogeneous forests in the Czech Republic is practical. This study provides an example of how general pedological knowledge can be used to define depth functions of SOC for forested sites.

Any strategy to change Carbon (C) pool would have a substantial effect on functionality of numerous ecosystem functions, detachment of Soil Organic Carbon (SOC), atmospheric carbon dioxide (CO2) concentration, and climate change mitigation. As the largest amount of the world’s C is stored in forests soils, the importance of forest SOC management is highlighted. Total SOC in forest varies not only laterally but also vertically with depth; however, the SOC storage of lower soil horizons have not been investigated enough despite their potential to frame our understanding of soil functioning. Visible–Near Infrared (vis–NIR) reflectance spectroscopy enables rapid examinations of the horizontal distribution of forest SOC, overcoming limitations of traditional soil assessment. This study aims to evaluate the potential of vis–NIR spectroscopy for characterizing the SOC contents of organic and mineral horizons in forests. We investigated 1080 forested sites across the Czech Republic at five individual soil layers, representing the Litter (L), Fragmented (F), and Humus (H) organic horizons, and the A1 (depth of 2–10 cm) and A2 (depth of 10–40 cm) mineral horizons (total 5400 samples). We then used Support Vector Machine (SVM) to model the SOC contents of (i) the profile (all organic and mineral horizons together), (ii) the combined organic horizons, (iii) the combined mineral horizons, and (iv) each individual horizon separately. The models were validated using 10-repeated 10-fold cross validation. Results showed that there was at least more than seven times as much SOC in the combined organic horizons compared to the combined mineral horizons with more variation in deeper layers. All individual horizons’ SOC was successfully predicted with low error and R2 values higher than 0.63; however, the prediction accuracy of F and A1 was greater compared to others (R2 > 0.70 and very low-biased spatial estimates). We have shown that modelling of SOC with vis–NIR spectra in different soil horizons of highly heterogeneous forests of the Czech Republic is practical.

Effective measurement and management of soil organic carbon (SOC) are essential for ecosystem function and food production. SOC has an important influence on soil properties and soil quality. Conventional SOC analysis is expensive and time-consuming. The development of spectral imaging sensors enables the acquisition of larger amounts of data using cheaper and faster methods. In addition, satellite remote sensing offers the potential to perform surveys more frequently and over larger areas. This research aimed to measure SOC content with colour as an indirect proxy. The measurements of soil colour were made at an agricultural site of the Czech Republic with an inexpensive digital camera and the Sentinel-2 remote sensor. Various soil colour spaces and colour indices derived from the (i) reflectance spectroscopy in the selected wavelengths of the visible (VIS) range (400–700 nm), (ii) RGB digital camera, and (iii) Sentinel-2 visible bands were used to train models for prediction of SOC. For modelling, we used the machine learning method, random forest (RF), and the models were validated with repeated 5-fold cross-validation. For prediction of SOC, the digital camera produced R2 = 0.85 and RMSEp = 0.11%, which had higher R2 and similar RMSEp compared to those obtained from the spectroscopy (R2 = 0.78 and RMSEp = 0.09%). Sentinel-2 predicted SOC with lower accuracy than other techniques; however, the results were still fair (R2 = 0.67 and RMSEp = 0.12%) and comparable with other methods. Using a digital camera with simple colour features was efficient. It enabled cheaper and accurate predictions of SOC compared to spectroscopic measurement and Sentinel-2 data.

The optimum rate and application timing of Nitrogen (N) fertilizer are crucial in achieving a high yield in rice cultivation; however, conventional laboratory testing of plant nutrients is time-consuming and expensive. To develop a site-specific spatial variable rate application method to overcome the limitations of traditional techniques, especially in fields under a double-cropping system, this study focused on the relationship between Soil Plant Analysis Development (SPAD) chlorophyll meter readings and N content in leaves during different growth stages to introduce the most suitable stage for assessment of crop N and prediction of rice yield. The SPAD readings and leaf N content were measured on the uppermost fully expanded leaf at panicle formation and booting stages. Grain yield was also measured at the end of the season. The analysis of variance, variogram, and kriging were calculated to determine the variability of attributes and their relationship, and finally, variability maps were created. Significant linear relationships were observed between attributes, with the same trends in different sampling dates; however, accuracy of semivariance estimation reduces with the growth stage. Results of the study also implied that there was a better relationship between rice leaf N content (R2Â =Â 0.93), as well as yield (R2Â =Â 0.81), with SPAD readings at the panicle formation stage. Therefore, the SPAD-based evaluation of N status and prediction of rice yield is more reliable on this stage rather than at the booting stage. This study proved that the application of SPAD chlorophyll meter paves the way for real-time paddy N management and grain yield estimation. It can be reliably exploited in precision agriculture of paddy fields under double-cropping cultivation to understand and control spatial variations. Keywords: Spatial variability, Non-invasive measurement, Precision farming, Decision support

The physical and chemical attributes of soils are strongly influenced by the nature of the minerals they contain and their concentration. Thus, soil texture is directly dependent on the content in clayey minerals, which influences a number of characteristics such as water dynamics. Although the mineralogical composition of soil is usually determined by X-ray diffraction spectroscopy, this technique is expensive and time-consuming, and uses toxic materials, all of which makes it impractical for obtaining large data sets. Also, available methods for acquiring, interpreting and examining visible–near infrared–shortwave infrared (VIS–NIR–SWIR) spectra are largely ineffective with tropical soils. The aim of this work was to ascertain whether VIS–NIR–SWIR reflectance spectroscopy (350–2500 nm) is useful for identifying minerals in subtropical soils as classified by textural class. For this purpose, soil samples were collected at 66 points at three different soil depths (0–20, 20–40 and 40–60 cm) over a study area located in the State of Rio Grande do Sul (Brazil). The soil texture were determined with the pipette method, and soil spectra were recorded on a FieldSpec Pro VIS-NIR-SWIR laboratory spectrophotometer. Soil minerals were identified, and their proportions determined, from the second-derivative of the Kubelka–Munk (KM) function for the spectra. Five main minerals were thus identified from their spectral signatures, namely: hematite, goethite, kaolinite, chlorite and illite. Identification of the minerals was facilitated by classifying the samples according to texture. The higher the clay content was, the higher was the spectral amplitude of the minerals identified. Those textural classes with the highest clay contents exhibited the greatest proportions of iron oxides and of clay minerals such as kaolinite. These relationships allowed more comprehensive analysis of the soils and expeditious characterization of the study area in terms of texture and mineralogy with a view to facilitating decision-making agricultural support policies.

This study aims to identify the influence of soil organic matter (OM) content and calcium carbonates (CaCO3) on soil reflectance and select the optimum spectral bands for discriminating between topsoils of different soil types situated in the irrigated perimeter of the Triffa plain (Morocco) using VIS-NIR reflectance spectroscopy. Soil samples were collected from the plow layer in 26 sampling sites. The spectral measurements were conducted in the field using an ASD Fieldspec portable spectroradiometer (350–2500 nm), while the soil samples were analyzed in the laboratory. The spectral data were pre-processed to remove the noise effects and then analyzed with the CovSel (selected covariance) method, validated by linear discriminant analysis in order to select the most optimal spectral variables to discriminate between topsoils of different soil types in the plain. The results of the soils reflectance curves showed that low reflectance intensity marked the soils with high OM contents throughout the VIS spectrum. The influence of the soil OM content was very apparent in the VIS range (between 580–750 nm). Regarding the CaCO3 content, it was noted that the soil samples with a high percentage of CaCO3 increase the reflectance in all spectral domains situated between 350 and 2500 nm. The spectral bands of 1999, 686, 1280, 2340 and 1951 nm were the most optimal for the soil discrimination in the Triffa plain. This study concluded that the VIS-NIR spectroscopy demonstrates an excellent ability to characterize and discriminate between topsoils in the Triffa plain.

Extensive efforts are currently being devoted to the establishment of soil spectral libraries on regional, national, continent-wide and global domains. In particular, a new development goes in the direction of global harmonized soil spectral databases that shall be acquired following common standards and procedures so that they can be merged with other soil spectral libraries. The choice of standards is important, especially with respect to their behaviors when laboratory conditions, such as humidity, change. In this study, we test the application and robustness of the Internal Soil Standard (ISS) spectral re-alignment procedure on an extended number of soil samples acquired at different laboratories. In particular, a focus is placed on the comparison between two standards Lucky Bay (LB) and Wylie Bay (WB), for their performance with different humidity laboratory conditions. LB and WB are almost pure quartz sands from Australia. For this, 71 soil samples from Israel with different mineralogical background and variable soil organic matter contents are scanned at two laboratories. The scanning took place with different spectral measurement protocols and extreme diverse conditions in terms of laboratory humidity and moist. All scans are completed by the repeated scans of the two Australian white sands through all scan batches to harmonize the spectral measurements. Our results show that the ISS minimizes the visual spectral variation, aligns the minor and extreme systematic changes of the protocols, and makes them more stable. Furthermore, the LB and WB standards do not exhibit equal performance regarding to relatively dry and humid conditions. The WB standard provides more stable and satisfactory results in humid condition. However, the high performance of the LB sample in spectral correction is still observed, particularly in laboratories with lower moist. Accordingly, these analyses suggest that in the ISS re-alignment procedure, the WB sample is more reliable to be used in humid laboratory condition.

The need for soil information is higher now than ever before. Agriculture and the way in which we use and manage our soils are being changed with the concerns over food security and global climate change. Mainly, soil data is necessary to be used in soil and natural resource management, e.g. for environmental modeling for a better understanding of soil processes and reducing risks in decision-making. Conventional soil survey cannot efficiently offer these data because the techniques are time-consuming and expensive. Proximal Soil Sensing (PSS), which has become a multidisciplinary area, aims to develop field-based techniques for acquiring information on the soil from close by, or within the soil. It can be used to monitor soil both surface and subsurface spatial and temporal information rapidly, cheaply and with less labor. This chapter reports on developments in PSS and its application in soil science and environmental assessment with specific attention to Iraq soils. It will review some of the technologies that may be used for PSS and proposes a framework for their use in Iraq soil and environmental monitoring. The chapter brings together ideas and examples from developing and using proximal sensors for applications such as precision agriculture and soil contamination monitoring, where there is a particular need for high spatial resolution information.

The application of multivariate geostatistical and statistical methods remain valuable tools for environmental pollution assessment. In particular, stochastic simulation techniques like sequential Gaussian simulation (SGS) and the self-organizing map artificial neural networks (SeOM-ANNs) have facilitated the understanding of the spatial distribution of potentially toxic elements (PTEs) in polluted soils. However, there is a dearth of literature on the application of SGS and SeOM-ANN in mapping potentially toxic elements (PTE) in heavily polluted mining and smelter affected floodplain soils. This study shows the applicability SGS and SeOM–ANN which is a powerful visualization tool for the categorization of PTEs [Cadmium (Cd), Arsenic (As), Antimony (Sb), Lead (Pb) and Zinc (Zn)] levels together with selected soil properties [oxidizable carbon (Cox) and soil reaction (pH_H2O)] in one of the most polluted mining floodplain soils in Europe. A k-means algorithm was used to classify distinct clusters which were visually unclear based on the SeOM–ANN Neighbor distance plot (U-Matrix). The k-means resulted in 5 distinct clusters. Cluster 1 to 5 based on SeOM–ANN for all PTEs revealed an increase in concentration levels in the same order (1–5) while for soil properties the trend was not clear. The soils were successfully assessed based on different intensity level combinations and k-means clustering results efficiently mapped into a spatial distribution map. High concentration levels of the PTEs were noticed in the northern parts of the study area based on the conditional Gaussian simulations (CGSs) generated through SGS, while low levels were prominent in the southwestern parts. The hotspot areas were comparable with the k-means spatial distribution maps. It is recommended that special attention be paid to the identified hotspots for possible remediation. This study further demonstrates the usefulness of geostatistics and advanced statistical methods in site-specific planning and implementation of remediation measures for polluted mining floodplain soils.

The surface organic horizons in forest soils have been affected by air and soil pollutants, including potentially toxic elements (PTEs). Monitoring of PTEs requires a large number of samples and adequate analysis. Visible–near infrared (vis–NIR: 350–2500 nm) spectroscopy provides an alternative method to conventional laboratory measurements, which are time-consuming and expensive. However, vis–NIR spectroscopy relies on an empirical calibration of the target attribute to the spectra. This study examined the capability of vis–NIR spectra coupled with machine learning (ML) techniques (partial least squares regression (PLSR), support vector machine regression (SVMR), and random forest (RF)) and a deep learning (DL) approach called fully connected neural network (FNN) to assess selected PTEs (Cr, Cu, Pb, Zn, and Al) in forest organic horizons. The dataset consists of 2160 samples from 1080 sites in the forests over all the Czech Republic. At each site, we collected two samples from the fragmented (F) and humus (H) organic layers. The content of all PTEs was higher in horizon H compared to F horizon. Our results indicate that the reflectance of samples tended to decrease with increased PTEs concentration. Cr was the most accurately predicted element, regardless of the algorithm used. SVMR provided the best results for assessing the H horizon (R2 = 0.88 and RMSE = 3.01 mg/kg for Cr). FNN produced the best predictions of Cr in the combined F + H layers (R2 = 0.89 and RMSE = 2.95 mg/kg) possibly due to the larger number of samples. In the F horizon, the PTEs were not predicted adequately. The study shows that PTEs in forest soils of the Czech Republic can be accurately estimated with vis–NIR spectra and ML approaches. Results hint in availability of a large sample size, FNN provides better results.

The monitoring and quantification of soil carbon provide a better understanding of soil and atmosphere dynamics. Visible-near-infrared-short-wave infrared (VIS-NIR-SWIR) reflectance spectroscopy can quantitatively estimate soil carbon content more rapidly and cost-effectively compared to traditional laboratory analysis. However, effective estimation of soil carbon using reflectance spectroscopy to a great extent depends on the selection of a suitable preprocessing sequence and data-mining algorithm. Many efforts have been dedicated to the comparison of conventional chemometric techniques and their optimization for soil properties prediction. Instead, the current study focuses on the potential of the new data-mining engine PARACUDA-II®, recently developed at Tel-Aviv University (TAU), by comparing its performance in predicting soil oxidizable carbon (Cox) against common data-mining algorithms including partial least squares regression (PLSR), random forests (RF), boosted regression trees (BRT), support vector machine regression (SVMR), and memory based learning (MBL). To this end, 103 soil samples from the Pokrok dumpsite in the Czech Republic were scanned with an ASD FieldSpec III Pro FR spectroradiometer in the laboratory under a strict protocol. Spectra preprocessing for conventional data-mining techniques was conducted using Savitzky-Golay smoothing and the first derivative method. PARACUDA-II®, on the other hand, operates based on the all possibilities approach (APA) concept, a conditional Latin hypercube sampling (cLHs) algorithm and parallel programming, to evaluate all of the potential combinations of eight different spectral preprocessing techniques against the original reflectance and chemical data prior to the model development. The comparison of results was made in terms of the coefficient of determination (R2) and root-mean-square error of prediction (RMSEp). Results showed that the PARACUDA-II® engine performed better than the other selected regular schemes with R2 value of 0.80 and RMSEp of 0.12; the PLSR was less predictive compared to other techniques with R2 = 0.63 and RMSEp = 0.29. This can be attributed to its capability to assess all the available options in an automatic way, which enables the hidden models to rise up and yield the best available model.

Soil Organic Carbon (SOC) is a useful representative of soil fertility and an essential parameter in controlling the dynamics of various agrochemicals in soil. Soil texture is also used to calculate soil's ability to retain water for plant growth. SOC and soil texture are thus important parameters of agricultural soils and need to be regularly monitored. Optical satellite remote sensing offers the potential for frequent surveys over large areas. In addition, the recently-operated Sentinel-2 missions provide free imagery. This study compared the capabilities of Sentinel-2 for monitoring and mapping of SOC and soil texture (clay, silt and sand content) with those obtained from airborne hyperspectral (CASI/SASI sensors) and lab ASD FieldSpec spectroradiometer measurements at four agricultural sites in the Czech Republic. Combination of 10 extracted bands of the Sentinel-2 and 18 spectral indices, as independent variables, were used to train prediction models and then produce spatial distribution maps of the selected attributes. Results showed that the prediction accuracy based on lab spectroscopy, airborne and Sentinel-2 in the majority of the sites was adequate for SOC and fair for clay; however, Sentinel-2 imagery could not be used to detect and map variations in silt and sand. The SOC and clay maps derived from the airborne and spaceborne datasets showed similar trend, with both performing better where SOC levels were relatively high, though at the highest levels Sentinel-2 was able to create the SOC map more precisely than the airborne sensors. Taken across all SOC levels measured in the reference data, Sentinel-2 results were marginally lower than lab spectroscopy and airborne imagery, but this reduction in precision may be offset by the extensive geographical coverage and more frequent revisit characteristic of satellite observation. The increased temporal revisit and area are expected to be positive enhancements to the acquisition of high-quality information on variations in SOC and clay content of bare soils.

In this study, the performances of hyperspectral airborne and superspectral spaceborne spectral imaging to derive selected Soil Organic Carbon (SOC) were analyzed and compared in agricultural sites of the Czech Republic. The main aim was to assess the potential of superspectral Sentinel-2 satellite for the prediction and mapping of the attribute. The prediction accuracy based on airborne and spaceborne techniques in majority of the sites was adequate for SOC. Comparing the spatial distribution maps of SOC derived from the airborne and spaceborne data showed a similar trend at both platforms. The SOC maps also confirmed that in areas with a high level of SOC, Sentinel-2 was able to detect SOC even more precisely than the airborne sensors. Although a decrease in the model and map performances was obvious in the case of parameters with low contents. The findings of the current research showed that superspectral Sentinel-2 allows for the estimation and mapping of SOC. The study also emphasized the importance of the superspectral Sentinel-2 data in soil characteristics assessment with a frequent revisit-time over larger areas than it currently is with laboratory and airborne instruments. Certainly, the repeatability of the Sentinel-2 products is still a work in progress and with the Sentinel-2B, a revisit-time of five-day and the temporal frequency of cloud-free acquisitions will be further increased. Accordingly, much more data will be freely available in the near future, which will have a significant influence on the obtaining of high-quality soil data.

Soil degradation includes a number of processes, ranging from soil erosion to soil contamination, which reduce the capability of soil to work as a base for vegetation roots. Methods to quantify soi...