Reflectance and imaging spectroscopy applied to detection of petroleum hydrocarbon pollution in bare soils

Development of a method for detecting and classifying hydrocarbon-contaminated soils via laser-induced breakdown spectroscopy and machine learning algorithms

In recent years, there has been a significant increase in oil exploration and exploitation activities, resulting in spills that pose a severe threat to the environment and public health. The present work aims to develop a method to detect and classify hydrocarbon-contaminated soils that is useful for analyzing contaminated sites. The method combines machine learning algorithms with data obtained via the laser-induced breakdown spectroscopy (LIBS) technique. The first stage involved optimizing the experimental parameters of the LIBS technique from eleven soil samples contaminated with different hydrocarbons and one sample used for control purposes. To classify the samples effectively, a robust and interpretable method was required. Linear discriminant analysis (LDA) was chosen for its ability to identify the linear combination of features that best separates classes while maintaining simplicity and interpretability. To address overfitting risks and reduce dimensionality, principal component analysis (PCA) was applied before LDA. This preprocessing step optimized the classification of samples contaminated with eleven different hydrocarbon sources and distinguished them from the control class. The results revealed accuracies greater than 90%. The model was also used to discriminate subsets of classes that shared similarities, which could be revealed from the analysis of the entire class set. The approach also successfully classified related classes, such as gasoline and different oils, achieving 100% accuracy in all cases. This enhanced capacity to identify and differentiate hydrocarbons with LIBS and machine learning marks a significant advancement in environmental monitoring.

Methane, Ethane, and Propane Detection Using a Quartz-Enhanced Photoacoustic Sensor for Natural Gas Composition Analysis

A compact and portable gas sensor based on quartz-enhanced photoacoustic spectroscopy (QEPAS) for the detection of methane (C1), ethane (C2), and propane (C3) in natural gas (NG)-like mixtures is reported. An interband cascade laser (ICL) emitting at 3367 nm is employed to target absorption features of the three alkanes, and partial least-squares regression analysis is employed to filter out spectral interferences and matrix effects characterizing the examined gas mixtures. Spectra of methane, ethane, and propane mixtures diluted in nitrogen are employed to train and test the regression algorithm, achieving a prediction accuracy of ∼98%, ∼96%, and ∼93% on C1, C2, and C3, respectively. With respect to previously reported QEPAS sensors for natural gas analysis, the high prediction accuracy as well as the capability to discriminate and detect C3 within natural gas-like complex mixtures provided by the employment of partial least-squares regression mark significant improvements. Furthermore, these results enable an improved performance of the sensor for in situ, real-time, and online natural gas composition analysis.

Spatial distribution of polycyclic aromatic hydrocarbons in sediment deposits in a Seine estuary tributary by hyperspectral imaging

Water bodies allow the storage of sediments from their catchment areas, including sediments containing persistent contaminants. This study used visible and near-infrared hyperspectral imaging to characterize the composition of sediment deposits collected in Martot Pond (France) and to reconstruct the volume of polycyclic aromatic hydrocarbon (PAH) contaminated sediments in the pond. Additionally, combining this method with polychlorinated biphenyl (PCB) analysis enhanced the age model associated with these sediments. To achieve this, indicators of oxides and chlorophyll a (and its derivatives) were employed to correlate various sediment cores, and to propose a sedimentary filling mode for the pond. Furthermore, one sedimentary unit, which appears homogeneous but of variable size within the pond, exhibited repetitive alternations associated with tidal cycles due to a defect in the Martot dam, corresponding to 34 +/- 3 days. A chemometric approach was used to model PAHs with near-infrared hyperspectral imaging data (validation determination coefficient of 0.85, Root Mean Squared Error of Prediction of 1.64 mg/kg). This model was then applied to other cores, coupled with the sedimentary filling mode in the pond, allowing the reconstruction of the volume of PAH contamination. Thus, this study demonstrates that hyperspectral imaging is a powerful tool for estimating various contaminants in sediments: not only is it much faster than conventional chromatographic methods, it also provides a more detailed understanding of a sample, and even of a site through the correlation of multiple core samples.

Real-time in situ detection of petroleum hydrocarbon pollution in soils via a novel optical methodology

Petroleum hydrocarbon (PHC) contamination in soils is considered one of the most serious problems currently, of which the detection and identification is a fairly significant but challenging work. Conventional methods to do such work usually need complex sample pretreatment, consume much time and fail to do the in-situ detection. This paper set out to create a novel systematic methodology to realize the goals accurately and efficiently. Based on laser-induced breakdown spectroscopy (LIBS) and self-improved machine learning methods, the innovative methodology only uses extremely simple devices to do the real-time in situ detection and identification work and even realize the quantitative analysis of pollution level accurately. In the study, clean soils mixed with petroleum were served as polluted samples, clean soils to be the blank group for comparison. Based on the elemental information from the spectra obtained by LIBS, machine learning methods were improved and helped optimized the algorithm to identify the PHC polluted soil samples for the first time. Furthermore, a novel model was designed to perform the quantitative analysis of the concentration of PHC pollution in soils, which can be applied to detect the degree of PHC contamination in soils accurately. Finally, the harmful volatile component of the PHC polluted soils was also successfully and identified despite its extremely minimal content in the air. The newly-designed methodology is novel and efficient, which has extensive application prospect in the real-time in situ detection of petroleum hydrocarbon pollution.

Diagnostic features emerging in near-infrared reflectance spectroscopy for low petroleum hydrocarbon pollution after spectral subtraction

Diagnostic features in near-infrared reflectance spectroscopy (NIRS) are the foundation of knowledge-based approach of petroleum hydrocarbon determination. However, a significant challenge arises when analyzing samples with low levels of petroleum hydrocarbon pollution, as they often lack distinctive diagnostic features in their sample NIRS spectra, limiting the effectiveness of this approach. To address this issue, we have developed a technical workflow for diagnostic spectrum construction and parameterization based on spectral subtraction. This method was applied on a set of NIRS spectra from soil samples that were contaminated with petroleum hydrocarbons (ranged between 178 and 1716 mg/kg of total petroleum hydrocarbon). Then two diagnostic features for low-level petroleum hydrocarbon pollution were found: (1) An overall downward concave emerged on diagnostic spectrum within both 2290-2370 nm and 1700-1780 nm for all low pollution levels even below 200 mg/kg; (2) An indicative pattern of asymmetric "W-shaped" double absorption valley occurred for those exceeding 1000 mg/kg, and its valleys located near 2310 nm, 2348 nm or 1727 nm, 1762 nm stably. These two features on diagnostic spectrum could be parameterized to detect, and the detection limit was at least about 10-50 times lower than that based on sample spectrum. These findings update our understanding on the detectability of spectral response from low petroleum hydrocarbon pollution, and widely extend the application of knowledge-based NIRS approach in either field detection or remote sensing identification for environmental management.

Influence of intermolecular interactions on the infrared complex indices of refraction for binary liquid mixtures

This paper investigates the accuracy of deriving the composite optical constants of binary mixtures from only the complex indices of refraction of the neat materials. These optical constants enable the reflectance spectra of the binary mixtures to be modeled for multiple scenarios (e.g., different substrates, thicknesses, volume ratios), which is important for contact and standoff chemical detection. Using volume fractions, each mixture's complex index of refraction was approximated via three different mixing rules. To explore the impact of intermolecular interactions, these predictions are tested by experimental measurements for two representative sets of binary mixtures: (1) tributyl phosphate combined with n-dodecane, a non-polar medium, to represent mixtures which primarily interact via dispersion forces and (2) tributyl phosphate and 1-butanol to represent mixtures with polar functional groups that can also interact via dipole-dipole interactions, including hydrogen bonding. The residuals and the root-mean-square error between the experimental and calculated index values are computed and demonstrate that for miscible liquids in which the average geometry of the cross-interactions can be considered isotropic (e.g., dispersion), the refractive indices of the mixtures can be modeled using composite n and k values derived from volume fractions of the neat liquids. Conversely, in spectral regions where the geometry of the cross-interactions is more restricted and anisotropic (e.g., hydrogen bonding), the calculated n and k values vary from the measured values. The impact of these interactions on the reflectance spectra are then compared by modeling a thin film of the binary mixtures on an aluminum substrate using both the measured and the mathematically computed indices of refraction.

Monitoring of Iron Ore Quality through Ultra-Spectral Data and Machine Learning Methods

Currently, most mining companies conduct chemical analyses by X-ray fluorescence performed in the laboratory to evaluate the quality of Fe ore, where the focus is mainly on the Fe content and the presence of impurities. However, this type of analysis requires the investment of time and money, and the results are often available only after the ore has already been sent by the processing plant. Reflectance spectroscopy is an alternative method that can significantly contribute to this type of application as it consists of a nondestructive analysis technique that does not require sample preparation, in addition to making the analyses available in more active ways. Among the challenges of working with reflectance spectroscopy is the large volume of data produced. However, one way to optimize this type of approach is to use machine learning techniques. Thus, the main objective of this study was the calibration and evaluation of models to analyze the quality of Fe from Sinter Feed collected from deposits in the Carajás Mineral Province, Brazil. To achieve this goal, machine learning models were tested using spectral libraries and X-ray fluorescence data from Sinter Feed samples. The most efficient models for estimating Fe were the Adaboost and support vector machine and our results highlight the possibility of application in the samples without the need for preparation and optimization of the analysis time, providing results in a timely manner to contribute to decision-making in the production chain.

Thermal desorption treatment of petroleum hydrocarbon-contaminated soils of tundra, taiga, and forest steppe landscapes

The results of field, analytical, and experimental research at a number of production facilities reflect the properties of oil-contaminated soils in 3 landscapes: the permafrost treeless Arctic ecosystem, boreal forest, and temperate-climate grassland-woodland ecotone. Laboratory studies have revealed the concentrations of petroleum hydrocarbons in soils, ranging from medium levels of 2000-3000 mg/kg to critical figures over 5000 mg/kg, being 2-25 times higher than the permissible content of oil products in soils. The experimentally applied thermal effects for the oil products desorption from the soil allowed finding an optimal regime: the treatment temperature from 25 to 250 °C reduces the concentrations to an acceptable value. The conditions are environmentally sound, given that the complete combustion point of humates is ca. 450 °C. The outcomes suggest the eco-friendly solution for soil remediation, preserving the soil fertility in fragile cold environments and in more resilient temperate climates, where revitalized brownfields are essential for food production.

Computed tomography imaging spectrometry based on superiorization and guided image filtering

Computed tomography imaging spectrometry (CTIS) is a snapshot hyperspectral imaging technique that can obtain a three-dimensional (${2D +}\lambda$) data cube of the target scene within a single exposure. Previous studies of CTIS suggest that reconstructions usually suffer from severe artifacts due to the limited number of projections available. To overcome this limitation, an iterative algorithm combining superiorization and guided image filtering is proposed to explore the intrinsic properties of the hyperspectral data cube as well as the characteristics of zero-order diffraction for the first time, to the best of our knowledge. Results from both simulative studies and proof-of-concept experiments demonstrate its superiority in suppressing artifacts and improving precision over the frequently used expectation maximization algorithm.

Statistical Evaluation of Quantities Measured in the Detection of Soil Air Pollution of the Environmental Burden

The article highlights the investigation of the relationships between measured quantities during the atmospheric geochemical survey of contaminated soil and the environmental burden of the industrial establishment in eastern Slovakia. Statistical data processing was undertaken from the measured values of pollutants. The basic statistical characteristics of the monitored indicators were defined here. With the help of regressive and correlative analysis, dependency was confirmed between examined values, further expressed by a mathematical relationship. We analysed variability of the measured variables due to the influence of changed input quantities by the non-parametric Wilcox test. The statistical data processing helps us to identify the dependency between the measured values and improves valorization of the pollution of a given environmental burden. This was due to the handling of organic pollutants and the production of basic organic and inorganic chemicals stated for other industries. Chemical analysis of soil air helps us to determine the extent and amount of soil contamination by pollutants. Individual pollutants have their own characteristic properties and their negative effects on biota, the environment and humans are different.

A two-step approach for detecting Striga in a complex agroecological system using Sentinel-2 data

Information on weed occurrence within croplands is vital but is often unavailable to support weeding practices and improve cropland productivity assessments. To date, few studies have been conducted to estimate and map weed abundances within agroecological systems from spaceborne images over wide-area landscapes, particularly for the genus Striga. Therefore, this study attempts to increase the detection capacity of Striga at subpixel size using spaceborne high-resolution imagery. In this study, a two-step classification approach was used to detect Striga (Striga hermonthica) weed occurrence within croplands in Rongo, Kenya. Firstly, multidate and multiyear Sentinel-2 (S2) data (2017 to 2018) were utilized to map cropland and non-cropland areas using the random forest algorithm within the Google Earth Engine. The non-cropland class was thereafter masked out from a single date S2 image of the 13^th of December 2017. The remaining cropland area was then used in a subpixel multiple endmember spectral mixture analysis (MESMA) to detect Striga occurrence and infestation using endmembers (EMs) obtained from the in-situ hyperspectral data. The gathered in-situ hyperspectral data were resampled to the spectral waveband configurations of S2 and three representative EMs were inferred, namely: (1) Striga, (2) crop and other weeds, and (3) soil. Overall classification accuracies of 88% and 78% for the pixel-based cropland mapping and subpixel Striga detection were achieved, respectively. Furthermore, an F-score (0.84) and a root mean square error (0.0075) showed that the MESMA subpixel algorithm provides plausible results for predicting the relative abundance of Striga within each S2 pixel at a landscape scale. The capability of MESMA together with a cropland classification hierarchical approach was thus proven to be suited for Striga detection in a heterogenous agroecological system. These results can be used to guide in the adaptation, mitigation, and remediation of already infested areas, thereby avoiding further Striga infestation of new croplands.

WorldView-3 mapping of Tarmat deposits of the Ras Rakan Island, Northern Coast of Qatar: Environmental perspective

This study characterizes the spectral behavior of tarmats and maps the tarmat deposits found along the coast of Ras Rakan Island off Qatar using WorldView-3 (WV-3) sensor data. The laboratory spectra of tar materials showed diagnostic absorptions features at 0.6 and 1.1 μm in the visible and near-infrared (VNIR) and 1.52, 1.73, 2.04, and 2.31 μm in the short wave infrared (SWIR) region. The panchromatic grayscale image and FCC showed the tarmat deposit as a linear warp feature between beach and water. The mapping of deposits using WV-3 data by decorrelation stretch and Linear Spectral Unmixing (LSU) methods discriminated the tarmats from the sandy soil, vegetation and sabkha features in a different tone. The capability of WV-3 sensor and the potential of image processing methods were verified by mapping the tar distribution of the Ras Ushayriq and NE of Al Ruwais.

Monitoring oil contamination in vegetated areas with optical remote sensing: A comprehensive review

The monitoring of soil contamination deriving from oil and gas industry remains difficult in vegetated areas. Over the last decade, optical remote sensing has proved helpful for this purpose. By tracking alterations in vegetation biochemistry through its optical properties, multi- and hyperspectral remote sensing allow detecting and quantifying crude oil and petroleum products leaked following accidental leakages or bad cessation practices. Recent advances in this field have led to the development of various methods that can be applied either in the field using portable spectroradiometers or at large scale on airborne and satellite images. Experiments carried out under controlled conditions have largely contributed to identifying the most important factors influencing the detection of oil (plant species, mixture composition, etc.). In a perspective of operational use, an important effort is still required to make optical remote sensing a reliable tool for oil and gas companies. The current methods used on imagery should extend their scope to a wide range of contexts and their application to upcoming satellite-embedded hyperspectral sensors should be considered in future studies.

Estimating persistent oil contamination in tropical region using vegetation indices and random forest regression

The persistence of soil contamination after cessation of oil activities remains a major environmental issue in tropical regions. The assessment of the contamination is particularly difficult on vegetated sites, but promising advances in reflectance spectroscopy have recently emerged for this purpose. This study aimed to exploit vegetation reflectance for estimating low concentrations of Total Petroleum Hydrocarbons (TPH) in soils. A greenhouse experiment was carried out for 42 days on Cenchrus alopecuroides (L.) under realistic tropical conditions. The species was grown on oil-contaminated mud pit soils from industrial sites, with various concentrations of TPH. After 42 days, a significant decrease in plant growth and leaf chlorophyll and carotenoid contents was observed for plants exposed to 5-19 g kg^-1 TPH in comparison to the controls (p < 0.05). Conversely, pigment contents were higher for plants exposed to 1 g kg^-1 TPH (hormesis phenomenon). These modifications proportionally affected the reflectance of C. alopecuroides at leaf and plant scales, especially in the visible region around 550 and 700 nm. 33 vegetation indices were used for linking the biochemical and spectral responses of the species to oil using elastic net regressions. The established models indicated that chlorophylls a and b and β-carotene were the main pigments involved in the modifications of reflectance (R² > 0.7). The same indices also succeeded in estimating the concentrations of TPH using random forest regression, at leaf and plant scales (RMSE = 1.46 and 1.63 g kg^-1 and RPD = 5.09 and 4.44, respectively). Four out of the 33 indices contributed the most to the models (>75%). This study opens up encouraging perspectives for monitoring the cessation of oil activities in tropical regions. Further researches will focus on the application of our approach at larger scale, on airborne and satellite imagery.

Hyperspectral Imaging and Hierarchical PLS-DA Applied to Asbestos Recognition in Construction and Demolition Waste

Asbestos-Containing Materials (ACMs) are hazardous and prohibited to be sold or used as recycled materials. In the past, asbestos was widely used, together with cement, to produce “asbestos cement-based” products. During the recycling process of Construction and Demolition waste (C&DW), ACM must be collected and deposited separately from other wastes. One of the main aims of the recycling strategies applied to C&DW was thus to identify and separate ACM from C&DW (e.g., concrete and brick). However, to obtain a correct recovery of C&DW materials, control methodologies are necessary to evaluate the quality and the presence of harmful materials, such as ACM. HyperSpectral Imaging (HSI)-based sensing devices allow performing the full detection of materials constituting demolition waste. ACMs are, in fact, characterized by a spectral response that nakes them is different from the “simple” matrix of the material/s not embedding asbestos. The described HSI quality control approach is based on the utilization of a platform working in the short-wave infrared range (1000–2500 nm). The acquired hyperspectral images were analyzed by applying different chemometric methods: Principal Component Analysis for data exploration and hierarchical Partial Least-Square-Discriminant Analysis (PLS-DA) to build classification models. Following this approach, it was possible to set up a repeatable, reliable and efficient technique able to detect ACM presence inside a C&DW flow stream. Results showed that it is possible to discriminate and identify ACM inside C&DW. The recognition is potentially automatic, non-destructive and does not need any contact with the investigated products.

Toward Quantifying Oil Contamination in Vegetated Areas Using Very High Spatial and Spectral Resolution Imagery

Recent remote sensing studies have suggested exploiting vegetation optical properties for assessing oil contamination, especially total petroleum hydrocarbons (TPH) in vegetated areas. Methods based on the tracking of alterations in leaf biochemistry have been proposed for detecting and quantifying TPH under controlled and field conditions. In this study, we expand their use to airborne imagery, in order to monitor oil contamination at a larger scale. Airborne hyperspectral images with very high spatial and spectral resolutions were acquired over an industrial site with oil-contamination (mud pits) and control sites both colonized by Rubus fruticosus L. The method of oil detection exploiting 14 vegetation indices succeeded in classifying the sites in the case of high TPH contamination (overall accuracy ≥ 91.8%). Two methods, based on either the PROSAIL (PROSPECT + SAIL) radiative transfer model or elastic net multiple regression, were also developed for quantifying TPH. Both methods were tested on reflectance measurements in the field, at leaf and canopy scales, and on the image, and achieved accurate predictions of TPH concentrations (RMSE ≤ 3.28 g/kg−1 and RPD ≥ 1.90). The methods were validated on additional sites and open up promising perspectives of operational application for oil and gas companies, with the emergence of new hyperspectral satellite sensors.

Application of PROSPECT for estimating total petroleum hydrocarbons in contaminated soils from leaf optical properties

Recent advances in hyperspectral spectroscopy suggest making use of leaf optical properties for monitoring soil contamination in oil production regions by detecting pigment alterations induced by Total Petroleum Hydrocarbons (TPH). However, this provides no quantitative information about the level of contamination. To achieve this, we propose an approach based on the inversion of the PROSPECT model. 1620 leaves from five species were collected on a site contaminated by 16 to 77 g.kg^-1 of TPH over a 14-month period. Their spectral signature was measured and used in PROSPECT model inversions to retrieve leaf biochemistry. The model performed well for simulating the spectral signatures (RMSE < 2%) and for estimating leaf pigment contents (RMSE ≤ 2.95 μg.cm^-2 for chlorophylls). Four out of the five species exhibited alterations in pigment contents when exposed to TPH. A strong correlation was established between leaf chlorophyll content and soil TPH concentrations (R² ≥ 0.74) for three of them, allowing accurate predictions of TPH (RMSE =3.20 g.kg^-1 and RPD = 5.17). The accuracy of predictions varied by season and improved after the growing period. This study demonstrates the capacity of PROSPECT to estimate oil contamination and opens up promising perspectives for larger-scale applications.

Hyperspectral Imaging in Environmental Monitoring: A Review of Recent Developments and Technological Advances in Compact Field Deployable Systems

The development and uptake of field deployable hyperspectral imaging systems within environmental monitoring represents an exciting and innovative development that could revolutionize a number of sensing applications in the coming decades. In this article we focus on the successful miniaturization and improved portability of hyperspectral sensors, covering their application both from aerial and ground-based platforms in a number of environmental application areas, highlighting in particular the recent implementation of low-cost consumer technology in this context. At present, these devices largely complement existing monitoring approaches, however, as technology continues to improve, these units are moving towards reaching a standard suitable for stand-alone monitoring in the not too distant future. As these low-cost and light-weight devices are already producing scientific grade results, they now have the potential to significantly improve accessibility to hyperspectral monitoring technology, as well as vastly proliferating acquisition of such datasets.