Quantitative analysis of heavy metals in soil via hierarchical deep neural networks with X-ray fluorescence spectroscopy

Wanqi Yang,Fusheng Li,Shubin Lyu,Qinglun Zhang,Yanchun Zhao

DOI: https://doi.org/10.1039/d3ja00120b

2023-01-01

Journal of Analytical Atomic Spectrometry

Abstract:Potentially toxic metals in soils are accurately predicted using hierarchical depth neural networks by X-ray fluorescence spectroscopy.

spectroscopy,chemistry, analytical

What problem does this paper attempt to address?

A deep spectral prediction network to quantitatively determine heavy metal elements in soil by x-ray fluorescence

Qinglun Zhang,Fusheng Li,Wanqi Yang

DOI: https://doi.org/10.1039/d3ja00392b

2024-01-05

Journal of Analytical Atomic Spectrometry

Abstract:An accurate and effective analytical method to measure the concentration of heavy metal elements (HMEs) in soil is of great significance for combating soil pollution, repairing ecosystems, and guiding agricultural cropland. Non-destructive, rapid, and in situ measurements make x-ray fluorescence spectroscopy (XRF) a popular tool for analyzing HMEs in soil. However, due to complex matrix effects and spectral line interference, the analytical accuracy is limited. In this study, an effective deep learning method is proposed to accurately determine the concentration of HMEs in soil by combining with the XRF. Firstly, soil spectra are acquired based on a handheld energy dispersive x-ray fluorescence spectrometer (ED-XRF). Secondly, according to the spectral continuity, cross-space correlation, and local correlation of XRF, a feature mining coordination module (FMC) is proposed. FMC is composed of a global spectral attention module (GSA) and a local multiscale feature extraction module (LMSFE), it can simultaneously achieve overall attention to the spectrogram and local feature modeling. Finally, a deep spectral prediction network (DSPFormer) is proposed based on the FMC module to achieve an accurate estimation of the concentration of five HMEs (Ti, Mn, Cu, Zn, Pb). The effectiveness of the method is demonstrated by comparing it with other advanced soil analysis algorithms. The coefficients of determination of DSPFormer for five HMEs (Ti, Mn, Cu, Zn, Pb) are 0.9559, 0.9627, 0.9658, 0.9584, and 0.9664, respectively. The results indicate that DSPFormer effectively mitigates the matrix effect and spectral line interference present in XRF. In summary, the deep learning method based on self-attention and convolutional neural network provides new theoretical guidance for soil HME analysis.

spectroscopy,chemistry, analytical
Hyperspectral Inversion of Soil Heavy Metals in Three-River Source Region Based on Random Forest Model

Wei Zhou,Han Yang,Lijuan Xie,Haoran Li,Lu Huang,Yapeng Zhao,Tianxiang Yue

DOI: https://doi.org/10.1016/j.catena.2021.105222

IF: 6.367

2021-01-01

CATENA

Abstract:Hyperspectral remote sensing technology has considerable research value in monitoring and evaluating soil heavy metal pollution. In this study, the Three-River Source Region was taken as the study area. The occurrence relationship of six heavy metals in soil, such as Mn, Cu, Zn, Pb, Cr, Ni, with soil organic matter, clay minerals, and iron-manganese oxides, was studied through the determination and analysis of soil samples and the collection of soil reflectance spectrum. Spectral transformation was carried out by first derivative, second derivative, inverse-log, continuum removal and multiple scattering correction of the spectrum. The correlation between soil heavy metal content and soil spectrum was analyzed to select the characteristic band, and partial least squares (PLS) method, support vector machine (SVM) method and random forest (RF) model were used to build inversion model based on characteristic band. Then the best combination of spectral transformation and inversion model were explored. The results showed that Pb contents were the twice of the background in Qinghai province. The combination spectrum processing method can improve the correlation between spectrum and heavy metals. The location and quantity of characteristic bands of six heavy metals are different. The accuracy of RF was significantly better than that of SVM and PLS for all six heavy metal (i.e. pb: R-RF(2) = 0.83, R-SVM(2) = 0.62, R-PLS(2) = 0.18), and the model effective of soil properties in non-polluted sites were reliable (i.e. clay: R-RF(2) = 0.93, R-SVM(2) = 0.87, R-PLS(2) = 0.74). This study can provide technical support for the larger-scale monitoring of soil heavy metal content and heavy metal pollution assessment.
Quantitative analysis of heavy metals in soil by X-ray fluorescence with improved variable selection strategy and bayesian optimized support vector regression

Xin Lu,Fusheng Li,Wanqi Yang,Pengfei Zhu,Shubin Lv

DOI: https://doi.org/10.1016/j.chemolab.2023.104842

IF: 4.175

2023-04-29

Chemometrics and Intelligent Laboratory Systems

Abstract:As a sensitive analytical technique, x-ray fluorescence (XRF) has received increasing attention in elemental analysis applications. In this paper, XRF technique coupled with improved variable selection strategy and bayesian optimized support vector regression was proposed for predicting heavy metals in the soil. First of all, an iterative variational mode decomposition (VMD) framework was provided to tackle continuous background interference in the XRF spectra of soil. The corrected spectra were then utilized for quantitative analysis of heavy metals. Aiming to implement accurate heavy metal prediction, we resorted to support vector regression (SVR) to establish the non-linear model. Combined with Bayesian optimization algorithm, the bayesian optimized SVR (BO-SVR) with optimal hyperparameters was obtained. However, severe collinearity between input variables can affect the accuracy and stability of the model. Towards this, the research utilized sensitivity analysis as the variable selection approach prior to BO-SVR prediction of XRF datasets. The experimental results revealed that BO-SVR assisted with VMD and sensitivity analysis yielded the best model outcome. Moreover, considerable contributions were observed by employing sensitivity analysis. It showed Rp2 of 0.9472 (for Cu), 0.7534 (for As), 0.9374 (for Zn), and 0.9283 (for Pb), and RMSEP of 58.0472 (for Cu), 16.5271 (for As), 63.2069 (for Zn) and 64.0662 (for Pb) without using variable selection technique. Resorting to sensitivity analysis, the Rp2 of Cu, As, Zn and Pb achieved 0.9918, 0.9526, 0.9611 and 0.9541, respectively and the RMSEP of Cu, As, Zn and Pb achieved 22.2204, 11.6868, 38.8753 and 32.9387, respectively. The results of this study clearly illustrated that the proposed model is an effective method for the analysis of heavy metals in soil.

automation & control systems,computer science, artificial intelligence,instruments & instrumentation,statistics & probability,mathematics, interdisciplinary applications,chemistry, analytical
Quantitative analysis of potentially toxic elements in soil by XRF based on efficient reinforcement learning and sparse partial least squares

Shubin Lyu,Fusheng Li,Wanqi Yang,Qinglun Zhang,Jin Su,Ruqing Zhao,Xin Lu

DOI: https://doi.org/10.1039/d3ja00406f

2024-01-25

Journal of Analytical Atomic Spectrometry

Abstract:Measuring accurate concentration levels of potentially toxic elements in agricultural soils is necessary to prevent hazards to human health. X-ray fluorescence spectroscopy (XRF) allows rapid, non-destructive detection of potentially toxic elements in agricultural soils. However, the high-dimensional spectral information associated with low concentrations of potentially toxic elements is difficult to be mined effectively, affecting the quantitative performance. Reinforcement learning is a potent paradigm for variable spatial search that allows effective modeling of high-dimensional data. This paper proposes a novel XRF spectral variable selection algorithm based on efficient Q-learning and sparse partial least squares (SPLS) to establish the relationship between XRF spectra and a single potentially toxic element of low concentrations (i.e., QSP). Firstly, an energy-dispersive X-ray fluorescence (EDXRF) spectrometer is used to obtain XRF spectral information of soil samples. Then, a novel efficient Q-learning XRF spectral variable optimization algorithm is proposed to select spectral variables related to a single potentially toxic element of low concentrations. Finally, the efficient Q-learning algorithm combines with the proposed SPLS to predict the concentrations of four potentially toxic elements, respectively. Compared with other advanced algorithms, the QSP achieves the lowest root mean squared error of prediction (RMSEP) at 8.540, 2.217, and 40.667 for Cr, Cd, and Ba, respectively, with the highest coefficients of determination ( ) of 0.988, 0.854, and 0.958. The concentration prediction results of QSP for Pb are similarly competitive. These results show that combining the QSP with XRF provides a new solution to improve quantitative performance for potentially toxic elements of low concentrations.

spectroscopy,chemistry, analytical
Estimation of metal elements content in soil using x-ray fluorescence based on multilayer perceptron

Fang Li,Xiaofeng Zhang,Anxiang Lu,Li Xu,Dong Ren,Tianyan You

DOI: https://doi.org/10.1007/s10661-022-09750-x

IF: 3.307

2022-01-14

Environmental Monitoring and Assessment

Abstract:X-ray fluorescence (XRF) is widely used to rapidly detect heavy metals in soil. Spectra processing has been an important research topic to improve accuracy. In this study, 80 soil samples were analyzed by XRF under indoor conditions, where different preprocessing and quantitative analysis methods were compared in terms of prediction accuracy. Denoising algorithms were used to preprocess the soil spectra before establishing prediction models for As, Pb, Cu, Cr, and Cd in soil. The influence of denoising methods on the prediction effects of different models was compared and discussed. The results on five heavy metal spectra show that the proper spectral preprocessing method can effectively improve the prediction performance of the model. The multilayer perceptron model provides promising analysis and modeling for the five metal elements. The determination coefficients (R2) of the models were 0.857, 0.976, 0.977, 0.995, and 0.886, respectively. The proposed method provides the potential to support accurate quantitation by XRF analysis.

environmental sciences
An integrated CBLA-Net with fractional discrete wavelet transform and frequency-based CARS to predict heavy metal elements by XRF

Wanqi Yang,Fusheng Li,Qinglun Zhang,Shubin Lyu

DOI: https://doi.org/10.1016/j.aca.2024.343073

2024-09-22

Abstract:Background: X-ray fluorescence (XRF) emerges as a promising technique for estimating heavy metal elements. However, XRF spectra typically contain a significant amount of environmental information and signal noise, and the relationship between spectral intensity and element concentration is difficult to quantify using a single model, thereby reducing the predictive performance for low concentration elements. Results: This paper proposed a comprehensive framework for predicting elemental concentrations, encompassing preprocessing, variable selection, decision-making, to enable fast, non-destructive, and accurate estimation of element concentrations in soil. Firstly, an optimal denoising method based on fractional discrete wavelet transform (FDWT) was introduced to enhance signal quality. Furthermore, the frequency-based competitive adaptive reweighted sampling (FCARS) algorithm was employed for feature selection of XRF spectral variables, allowing extraction of the most informative features from the complex spectral data. Finally, a novel deep learning network, called ConvBiLSTM-Attention (CBLA-Net), was designed to achieve precise estimation of heavy metal elements concentration. Compared with other advanced algorithms, The CBLA-Net demonstrated the highest accuracy for V, Cr, Mn, Zn, Cd, and Pb, achieving the coefficient of determination (R2) of 0.9730, 0.9874, 0.9952, 0.9921, 0.9518, and 0.9741, respectively. The CBLA-Net not only effectively extract local features and capture global information, but also combines attention mechanism to focus on key information. Significance: The proposed novel deep learning quantitative framework, including preprocessing, feature selection, and CBLA-Net decision-making, significantly enhances the accuracy of elemental content prediction. It provides a new approach for accurately assessing the concentration of heavy metal elements in soil.
Application of the Deep Learning Algorithm to Identify the Spatial Distribution of Heavy Metals at Contaminated Sites

Jun Man,Lingzao Zeng,Jian Luo,Weiliang Gao,Yijun Yao

DOI: https://doi.org/10.1021/acsestengg.1c00224

IF: 7.1

2022-01-01

ACS ES&T Engineering

Abstract:Accurate estimation of the spatial distribution of pollutants, such as heavy metals, is indispensable in investigation of contaminated sites. In this paper, we develop a novel deep learning (DL) algorithm, in which the data value of the target point is predicted by a nearest-neighbor neural network. Using 20 hypothetical sites and a real-world one, the developed DL is compared with several commonly used kriging algorithms (e.g., ordinary kriging). By taking the results of ordinary kriging as the benchmark, the interpolation accuracy can be improved by an average of 38.2% and 11.2-36.6%, respectively. The comparisons also indicate that the new method performs much better in cases with significant spatial variability by alleviating the smoothing effect and the edge effect in the kriging. Upon further examination of the developed method, the prediction accuracy is found to first increase and then decrease with the number of neighbor points. Moreover, the influence of the sampling density is limited if the number exceeds a certain threshold (e.g., n = 64 in our case). As a preliminary attempt of applying the DL algorithm at individual contaminated sites, this work provides a general alternative method to identify the spatial distribution of heavy metals.
Simulation and prediction research of heavy metal pollution in soil of typical sites based on multi-source heterogeneous data

XU Yang,CHEN Jiansong,WANG Zhidong,JIANG Fangming,ZHANG Qingyu,TANG Kuo,JIANG Hongqiang,DENG Jinsong

DOI: https://doi.org/10.13671/j.hjkxxb.2023.0090

2023-01-01

Abstract:Soil is considered the fundamental factor involved in the survival and development of human beings. The risk of heavy metal pollution in soil has been also increasing yearly along with the continuous advancement of industrialization and urbanization in China. In order to study the scientific perspective of heavy metal pollution prediction of site soil based on non-sampling multi-source heterogeneous data as well as exploring the importance and influence rules of different variables in different heavy metal pollution, 104 typical polluted sites in Zhejiang Province have been selected as the research objects. The site soil heavy metal pollution risk identification index system has been constructed based on the principle of mass balance, including four major variables of enterprise production, atmospheric deposition, plant enrichment, and soil leaching along with twenty Secondary variables. Random Forest and Feature Importance analyzing methods have been used to establish prediction models and explore the importance of variables using detailed site soil pollution survey results. The results have shown that:（1） The mean value of the determination coefficient(R mean ~2=0.75) of the prediction model based on the site mean value of Random Forest has been found higher than that of the models based on the site maximum value(R mean ~2=0.62). The predicting ability of the model method towards the mean values of heavy metal pollution in sites has been found better than that of the maximum values in sites. The preciseness of the prediction models based on the mean values has been ranked as nickel, mercury, cadmium, lead, copper and arsenic from high to low, while the rank based on the maximum values is cadmium, nickel, arsenic, lead, copper and mercury from high to low. The determination coefficient of the prediction model based on the Nemerow comprehensive pollution index（R~2=0.84） has been found higher than the average level of the single heavy metal prediction models and lower than the mean-based prediction models of nickel（R~2=0.92） and mercury（R~2=0.91）；（2） The model has shown poor predicting ability for extremely high values, and the multi-source heterogeneous database has shown weak representing ability for such samples, which could be the main reason why the overall accuracy of the model based on the maximum value has been found lower than that of the model based on the mean value；（3）Based on the average importance of the characteristics, the industrial-environmental impact grade（15.57%）, the proportion of soil clay particles（9.55%）, the highway density（8.44%）, the enterprise operation time（7.31%）, the solar radiation intensity（7.06%）, and the enterprise establishment time（6.56%） have been addressed as the variable indicators closely related to soil heavy metal pollution. The purpose of this study has been set to provide a new approach to the analysis of heavy metal pollution in soil, which includes using multi-source heterogeneous data to identify potential areas where heavy metal pollution risks may exist, analyzing the important relationship between different heavy metal elements and feature variables, and providing more accurate information for the investigation of heavy metal pollution in soil and environmental protection.
Estimation of Heavy Metals in Agricultural Soils Using Vis-NIR Spectroscopy with Fractional-Order Derivative and Generalized Regression Neural Network

Xitong Xu,Shengbo Chen,Liguo Ren,Cheng Han,Donglin Lv,Yufeng Zhang,Fukai Ai

DOI: https://doi.org/10.3390/rs13142718

IF: 5

2021-07-10

Remote Sensing

Abstract:With the development of industrialization and urbanization, heavy metal contamination in agricultural soils tends to accumulate rapidly and harm human health. Visible and near-infrared (Vis-NIR) spectroscopy provides the feasibility of fast monitoring of the variation of heavy metals. This study explored the potential of fractional-order derivative (FOD), the optimal band combination algorithm and different mathematical models in estimating soil heavy metals with Vis-NIR spectroscopy. A total of 80 soil samples were collected from an agriculture area in Suzi river basin, Liaoning Province, China. The spectra for mercury (Hg), chromium (Cr), and copper (Cu) of the samples were obtained in the laboratory. For spectral preprocessing, FODs were allowed to vary from 0 to 2 with an increment of 0.2 at each step, and the optimal band combination algorithm was applied to the spectra after FOD. Then, four mathematical models, namely, partial least squares regression (PLSR), adaptive neural fuzzy inference system (ANFIS), random forest (RF) and generalized regression neural network (GRNN), were used to estimate the concentration of Hg, Cr and Cu. Results showed that high-order FOD had an excellent effect in highlighting hidden information and separating minor absorbing peaks, and the optimal band combination algorithm could remove the influence of spectral noise caused by high-order FOD. The incorporation of the optimal band combination algorithm and FOD is able to further mine spectral information. Furthermore, GRNN made an obvious improvement to the estimation accuracy of all studied heavy metals compared to ANFIS, PLSR, and RF. In summary, our results provided more feasibility for the rapid estimation of Hg, Cr, Cu and other heavy metal pollution areas in agricultural soils.

environmental sciences,imaging science & photographic technology,remote sensing,geosciences, multidisciplinary
Estimation of Heavy Metal Content in Soil Based on Machine Learning Models

Shuaiwei Shi,Meiyi Hou,Zifan Gu,Ce Jiang,Weiqiang Zhang,Mengyang Hou,Chenxi Li,Zenglei Xi

DOI: https://doi.org/10.3390/land11071037

IF: 3.905

2022-07-09

Land

Abstract:Heavy metal pollution in soil is threatening the ecological environment and human health. However, field measurement of heavy metal content in soil entails significant costs. Therefore, this study explores the estimation method of soil heavy metals based on remote sensing images and machine learning. To accurately estimate the heavy metal content, we propose a hybrid artificial intelligence model integrating least absolute shrinkage and selection operator (LASSO), genetic algorithm (GA) and error back propagation neural network (BPNN), namely the LASSO-GA-BPNN model. Meanwhile, this study compares the accuracy of the LASSO-GA-BPNN model, SVR (Support Vector Regression), RF (Random Forest) and spatial interpolation methods with Huanghua city as an example. Furthermore, the study uses the LASSO-GA-BPNN model to estimate the content of eight heavy metals (including Ni, Pb, Cr, Hg, Cd, As, Cu, and Zn) in Huanghua and visualize the results in high resolution. In addition, we calculate the Nemerow index based on the estimation results. The results denote that, the simultaneous optimization of BPNN by LASSO and GA can greatly improve the estimation accuracy and generalization ability. The LASSO-GA-BPNN model is a more accurate model for the estimate heavy metal content in soil compared to SVR, RF and spatial interpolation. Moreover, the comprehensive pollution level in Huanghua is mainly low pollution. The overall spatial distribution law of each heavy metal content is very similar, and the local spatial distribution of each heavy metal is different. The results are of great significance for soil pollution estimation.

environmental studies
Hyperspectral sensing of heavy metals in soil by integrating AI and UAV technology

DOI: https://doi.org/10.1007/s10661-022-10125-5

IF: 3.307

2022-06-23

Environmental Monitoring and Assessment

Abstract:Given the limitation of conventional soil pollution monitoring through mapping which is a costly, time-consuming work, the study aims to establish an image recognition model to identify the source of pollution automatically. The study choses a contaminated land and then use a non-destructive instrument that can quickly and effectively measure the content of heavy metals. A two concentration prediction models of Ni, Cu, Zn, Cr, Pb, As, Cd, and Hg using hyperspectral imaging were developed, Decision Tree and Back Propagation Neural Network, in combination of particle swarm optimization employed for optimization algorithm. As a result, random forest is more accurate than the forecast result of back propagation neural network. This study has established an excellent Cu and Cr model, which can accurately capture the pollution source. In addition, through aerial photographs, we also found that there were also high pollution reactions on the banks of the river. The developed model is beneficial for high pollution areas which can be quickly found, thereby following investigation and remediation work can be carried out with less time and cost consuming comparing with the conventional soil monitoring.

environmental sciences
Hyperspectral monitor of soil chromium contaminant based on deep learning network model in the Eastern Junggar coalfield

Yuan Wang,Hongbing Ma,Jingzhe Wang,Li Liu,Matti Pietikäinen,Zipeng Zhang,Xiangyue Chen

DOI: https://doi.org/10.1016/j.saa.2021.119739

2021-08-05

Abstract:In China, over 10% of cultivated land is polluted by heavy metals, which can affect crop growth, food safety and human health. Therefore, how to effectively and quickly detect soil heavy metal pollution has become a critical issue. This study provides a novel data preprocessing method that can extract vital information from soil hyperspectra and uses different classification algorithms to detect levels of heavy metal contamination in soil. In this experiment, 160 soil samples from the Eastern Junggar Coalfield in Xinjiang were employed for verification, including 143 noncontaminated samples and 17 contaminated soil samples. Because the concentration of chromium in the soil exists in trace amounts, combined with the fact that spectral characteristics are easily influenced by other types of impurity in the soil, the evaluation of chromium concentrations in the soil through hyperspectral analysis is not satisfactory. To avoid this phenomenon, the pretreatment method of this experiment includes a combination of second derivative and data enhancement (DA) approaches. Then, support vector machine (SVM), k-nearest neighbour (KNN) and deep neural network (DNN) algorithms are used to create the discriminant models. The accuracies of the DA-SVM, DA-KNN and DA-DNN models were 95.61%, 95.62% and 96.25%, respectively. The results of this experiment demonstrate that soil hyperspectral technology combined with deep learning can be used to instantly monitor soil chromium pollution levels on a large scale. This research can be used for the management of polluted areas and agricultural insurance applications.
Digital soil mapping of heavy metals using multiple geospatial data: Feature identification and deep neural network

Qian Liu,Bin Du,Li He,Yun Zeng,Yu Tian,Zihong Zhang,Ran Wang,Tiezhu Shi

DOI: https://doi.org/10.1016/j.ecolind.2023.110863

IF: 6.9

2023-09-04

Ecological Indicators

Abstract:Monitoring the spatial distribution and sources of heavy metals (HM) in soil is essential for avoiding health risks and achieving sustainable soil utilization. Multiple geospatial data, including remote sensing, climate, soil and topography data, were used to extract environmental covariates. Additionally, the spatial scene was employed as the alternative data of land use/land cover to describe the urban functions and human activity intensity in more detail. After converting to a uniform resolution of 30 m, these environmental covariates were adopted to characterize several common HM in soil, including copper (Cu), chromium (Cr), lead (Pb), nickel (Ni), and zinc (Zn). The RReliefF algorithm was used to identify several important variables. The quantification models of HM were established using back-propagation neural network (BPNN) and deep neural network (DNN). Besides, the impact of distance from the spatial scenes on HM were analyzed. The result demonstrated that the spatial scene is a key environmental covariate in estimating HM in soil. Compared with BPNN, the DNN model provided better accuracy (R 2 = 0.67–0.75) for estimation of five HM elements. Therefore, the DNN model was used to map HM concentrations at a grid scale of 30 m. The spatial scenes with the highest risk of HM pollution are industrial areas, residential areas, road, and commercial areas, and the concentration of HM is negatively correlated with the distance from these spatial scenes. The effective impact distances of industrial and residential areas are about 2000 m, and the effective impact distances of road and commercial areas are 500 m.

environmental sciences
Improved mapping of heavy metals in agricultural soils using machine learning augmented with spatial regionalization indices

Xudong Ma,Dong-Xing Guan,Chaosheng Zhang,Tao Yu,Cheng Li,Zhiliang Wu,Bo Li,Wenda Geng,Tiansheng Wu,Zhongfang Yang

DOI: https://doi.org/10.1016/j.jhazmat.2024.135407

2024-08-03

Abstract:The accurate spatial mapping of heavy metal levels in agricultural soils is crucial for environmental management and food security. However, the inherent limitations of traditional interpolation methods and emerging machine-learning techniques restrict their spatial prediction accuracy. This study aimed to refine the spatial prediction of heavy metal distributions in Guangxi, China, by integrating machine learning models and spatial regionalization indices (SRIs). The results demonstrated that random forest (RF) models incorporating SRIs outperformed artificial neural network and support vector regression models, achieving R2 values exceeding 0.96 for eight heavy metals on the test data. Hierarchical clustering for feature selection further improved the model performance. The optimized RF models accurately predicted the heavy metal distributions in agricultural soils across the province, revealing higher levels in the central-western regions and lower levels in the north and south. Notably, the models identified that 25.78 % of agricultural soils constitute hotspots with multiple co-occurring heavy metals, and over 6.41 million people are exposed to excessive soil heavy metal levels. Our findings provide valuable insights for the development of targeted strategies for soil pollution control and agricultural soil management to safeguard food security and public health.
Using Multispectral Variables to Estimate Heavy Metals Content in Agricultural Soils: A Case of Suburban Area in Tianjin, China

Na Sun,Li Zhang,Lu Wang,Huiyue Su,Zhikui Chen,Huan Yu,Bo Li

DOI: https://doi.org/10.1016/j.geodrs.2022.e00540

IF: 4.201

2022-01-01

Geoderma Regional

Abstract:Efficient estimation and mapping of soil heavy metal concentrations based on rapid and environmentally friendly multispectral data can provide a reference for the prevention and control of non-point source pollution and safety in ecosystems. In this study, the spatial variation of soil heavy metals (As, Cd, Cr, Cu, Ni, and Zn) in Wuqing was analyzed with 394 and 57 soil samples gathered in 2005 and 2019. Based on these samples, as well as multispectral variables (raw bands and their respective vegetation index and derivative transformation), random forest-ordinary kriging (RFOK) and artificial neural network-ordinary kriging (ANNOK) were combined to predict spatial variations of heavy metal content. Results indicated that the mean values of soil As, Cd, Cr, Cu, and Zn in 2019 exceeded their respective concentrations in 2005 by 1.40, 1.36, 1.12, 1.12, and 1.35 times, and exceeded the standard rate of background values (BGV) in 2019 for all soil heavy metals (except for Ni). Compared to RF, ANN, and RFOK, the ANNOK method exhibited the best estimation performance for As (R20052= 69.47% andR20192= 65.03%), Cd (R20052= 64.48% andR20192= 64.20%), Cr (R20052= 72.02% andR20192= 70.65%), Cu (R20052= 72.05% andR20192= 70.53%), Ni (R20052= 69.37% andR20192= 66.03%), and Zn (R20052= 71.16% andR20192= 67.11%) in the validation dataset. Mapping of heavy metals derived from ANNOK revealed a similar distribution in 2005 and 2019, where the high concentrations of all soil heavy metals were in the southern part of the study area. The effects of multispectral bands as indicated by vegetation index and derivative transformation exhibited sensitivity for predicting soil heavy metals based on hybrid geostatistical methods. Overall, greater efficiency of spatial estimation for the content of soil heavy metals is important for the prevention and control of non-point source pollution.
Accurate prediction of spatial distribution of soil potentially toxic elements using machine learning and associated key influencing factors identification: A case study in mining and smelting area in southwestern China

Kai Li,Guanghui Guo,Degang Zhang,Mei Lei,Yingying Wang

DOI: https://doi.org/10.1016/j.jhazmat.2024.135454

2024-08-07

Abstract:Accurate prediction of spatial distribution of potentially toxic elements (PTEs) is crucial for soil pollution prevention and risk control. Achieving accurate prediction of spatial distribution of soil PTEs at a large scale using conventional methods presents significant challenges. In this study, machine learning (ML) models, specially artificial neural network (ANN), random forest (RF), and extreme gradient boosting (XGB), were used to predict spatial distribution of soil PTEs and identify associated key factors in mining and smelting area located in Yunnan Province, China, under the three scenarios: (1) natural + socioeconomic + spatial datasets (NS), (2) NS + irrigation pollution index (IPI) datasets, (3) NS + IPI + deposition (DEPO) datasets. The results highlighted the combination of NS+IPI+DEPO yielded the highest predictive accuracy across ML models. Particularly, XGB exhibited the highest performance for As (R2 =0.7939), Cd (R2 =0.6679), Cu (R2 =0.8519), Pb (R2 =0.8317), and Zn (R2 =0.7669), whereas RF performed the best for Ni (R2 =0.7146). The feature importance and Shapley additive explanation (SHAP) analysis revealed that DEPO and IPI were the pivotal factors influencing the distribution of soil PTEs. Our findings highlighted the important role of DEPO in spatial distribution prediction of soil PTEs, which has often been ignored in previous studies.
Prediction of heavy metals in soils of an arid area based on multi-spectral data.

Qingyu Guan,Rui Zhao,Feifei Wang,Ninghui Pan,Liqin Yang,Na Song,Chuanqi Xu,Jinkuo Lin

DOI: https://doi.org/10.1016/j.jenvman.2019.04.109

IF: 8.7

2019-01-01

Journal of Environmental Management

Abstract:With the rapid and extensive development of industry and agriculture, the soil environment inevitably becomes contaminated with heavy metals, thus creating adverse environmental conditions for flora and fauna. The traditional methods for combining field sampling with laboratory analysis of soil heavy metals are limited not only because they are time-consuming and expensive, but also because they are unable to obtain adequate information about the spatial distribution characteristics of heavy metals in soil over a large area. Three hundred and ninety-four soil samples (Gobi and farmland) were collected in an arid area in Jiuquan in Northwest China and analyzed for elements concentrations. Based on these measured concentrations, as well as rapid and environmentally friendly remote sensing (multi-spectral data), stepwise multiple linear regression (SMLR) and partial least-squares regression (PLS) were combined to predict concentrations and distributions of heavy metals in the soils of the study area. Furthermore, laboratory data were used to assess the accuracy of the prediction results. Obtained results suggest that the SMLR and PLS models were able to predict the metals contents in the study area. The concentrations of Cr, Ni, V and Zn could be predicted by two regression models, while those of Cu and Mn were predicted more accurately when they were attached to the SMLR model. The spatial distribution of heavy metals derived from the two models is consistent with measured values, indicating that it is reasonable to predict the concentrations of heavy metals in the soil of the study area using the multi-spectral data.
Prediction of Heavy Metal Concentrations in Contaminated Sites from Portable X-ray Fluorescence Spectrometer Data Using Machine Learning

Feiyang Xia,Tingting Fan,Yun Chen,Da Ding,Jing Wei,Dengdeng Jiang,Shaopo Deng

DOI: https://doi.org/10.3390/pr10030536

IF: 3.5

2022-03-09

Processes

Abstract:Portable X-ray fluorescence (pXRF) spectrometers provide simple, rapid, nondestructive, and cost-effective analysis of the metal contents in soils. The current method for improving pXRF measurement accuracy is soil sample preparation, which inevitably consumes significant amounts of time. To eliminate the influence of sample preparation on PXRF measurements, this study evaluates the performance of pXRF measurements in the prediction of eight heavy metals’ contents through machine learning algorithm linear regression (LR) and multivariate adaptive regression spline (MARS) models. Soil samples were collected from five industrial sites and separated into high-value and low-value datasets with pXRF measurements above or below the background values. The results showed that for Cu and Cr, the MARS models were better than the LR models at prediction (the MARS-R2 values were 0.88 and 0.78; the MARS-RPD values were 2.89 and 2.11). For the pXRF low-value dataset, the multivariate MARS models improved the pXRF measurement accuracy, with the R2 values improved from 0.032 to 0.39 and the RPD values increased by 0.02 to 0.37. For the pXRF high-value dataset, the univariate MARS models predicted the content of Cu and Cr with less calculation. Our study reveals that machine learning methods can better predict the Cu and Cr of large samples from multiple contaminated sites.

engineering, chemical
Prediction of heavy metal content in soil of cultivated land: Hyperspectral technology at provincial scale

Xia Fang,Peng Jie,Wang Qian-Long,Zhou Lian-Qing,Shi Zhou

DOI: https://doi.org/10.11972/j.issn.1001-9014.2015.05.014

2015-01-01

JOURNAL OF INFRARED AND MILLIMETER WAVES

Abstract:A total of 643 farmland topsoil samples distributed in 36 counties and cities of Zhejiang Province were collected. The correlation between contents of Ni, Cu, As, Hg, Zn, Cr, Cd and Pb and that of organic matter was probed by measuring the reflectance of soil samples in visible-near infrared light band. The characteristic wave bands of heavy metal elements and organic matter were compared. The partial least squares regression (PLSR) model for the content of each heavy metal element was established. The results indicated that Ni and Cr have the best correlation with organic matter, while As has the worst, with the correlation coefficients 0.54, 0.59 and 0.20, respectively. The distance between heavy metal elements and organic matter in the first three principal components loading diagram was inversely proportional to their correlation coefficient. Degree of overlap between different heavy metal elements and organic matter at hyperspectral sensitive band and the positive and negative consistency of regression coefficients varied greatly, the greater the correlation with organic matter is, the higher degree of overlap is, and the better the positive and negative consistency. In PLSR models of heavy metals, models for Ni and Cr performed well in modeling and predicting with a good ability of quantificational prediction, with RPD values of 1.94 and 1.80 separately. The remaining models for other 6 heavy metals could only conduct distinguishing for high and low values with RPD values ranged from 1.00 to 1.4. The results of this study provide certain theoretical assistance and reference for hyperspectral remote sensing monitoring of soil contamination by heavy metals in large-scale areas.
Research on XRF-visNIR Soil Heavy Metal Exceedance Analysis Method based on GAS Transformation and PCANet

Qingya Wang,Liangliang Tao,Fusheng Li,Zhichun Wu,Yaoyi Cai,Shubin Lyu

DOI: https://doi.org/10.1039/d4ja00161c

2024-06-27

Journal of Analytical Atomic Spectrometry

Abstract:Analyzing and rapidly screening the phenomenon of soil heavy metal exceedance remains a challenge for the fusion technology of X-ray fluorescence (XRF) and visible near infrared spectroscopy (visNIR). To address this, a new XRF-visNIR fusion method based on Gramian Angular Summation (GAS) transformation and Principal Component Analysis Network (PCANet) feature extraction is proposed. This method transforms XRF and visNIR data into two-dimensional images through GAS conversion, followed by feature extraction using PCANet. This reduces data dimensions and extracts important information about soil heavy metals. In the experimental phase, a large number of soil samples were collected from Hongfeng Lake area and tested for spectral information using XRF-visNIR. By constructing and training a deep learning network, soil heavy metal pollution was classified and assessed. The results show that this method has achieved significant results in the analysis of soil heavy metal exceedance. The optimized GASF_PCANet_CNN can rapidly and accurately identify seven kinds of heavy metal pollution exceedance (Pb, Cd, As, Cr, Cu, Zn, Ni). Deployed on an embedded platform, it can achieve quick feedback of screening results, with average accuracy, average recall rate, average precision, and average F1 score being 95.07%, 95.90%, 95.17%, and 95.53% respectively. The XRF-visNIR soil heavy metal analysis method based on GAS transformation and PCANet proposed in this study provides an efficient and reliable analytical means for monitoring soil heavy metal exceedance, actively promoting soil pollution management and environmental protection work.

spectroscopy,chemistry, analytical

Quantitative analysis of heavy metals in soil via hierarchical deep neural networks with X-ray fluorescence spectroscopy

A deep spectral prediction network to quantitatively determine heavy metal elements in soil by x-ray fluorescence

Hyperspectral Inversion of Soil Heavy Metals in Three-River Source Region Based on Random Forest Model

Quantitative analysis of heavy metals in soil by X-ray fluorescence with improved variable selection strategy and bayesian optimized support vector regression

Quantitative analysis of potentially toxic elements in soil by XRF based on efficient reinforcement learning and sparse partial least squares

Estimation of metal elements content in soil using x-ray fluorescence based on multilayer perceptron

An integrated CBLA-Net with fractional discrete wavelet transform and frequency-based CARS to predict heavy metal elements by XRF

Application of the Deep Learning Algorithm to Identify the Spatial Distribution of Heavy Metals at Contaminated Sites

Simulation and prediction research of heavy metal pollution in soil of typical sites based on multi-source heterogeneous data

Estimation of Heavy Metals in Agricultural Soils Using Vis-NIR Spectroscopy with Fractional-Order Derivative and Generalized Regression Neural Network

Estimation of Heavy Metal Content in Soil Based on Machine Learning Models

Hyperspectral sensing of heavy metals in soil by integrating AI and UAV technology

Hyperspectral monitor of soil chromium contaminant based on deep learning network model in the Eastern Junggar coalfield

Digital soil mapping of heavy metals using multiple geospatial data: Feature identification and deep neural network

Improved mapping of heavy metals in agricultural soils using machine learning augmented with spatial regionalization indices

Using Multispectral Variables to Estimate Heavy Metals Content in Agricultural Soils: A Case of Suburban Area in Tianjin, China

Accurate prediction of spatial distribution of soil potentially toxic elements using machine learning and associated key influencing factors identification: A case study in mining and smelting area in southwestern China

Prediction of heavy metals in soils of an arid area based on multi-spectral data.

Prediction of Heavy Metal Concentrations in Contaminated Sites from Portable X-ray Fluorescence Spectrometer Data Using Machine Learning

Prediction of heavy metal content in soil of cultivated land: Hyperspectral technology at provincial scale

Research on XRF-visNIR Soil Heavy Metal Exceedance Analysis Method based on GAS Transformation and PCANet