Three-Dimensional Spatial Distribution of Elemental Lead (Pb) Stress in Vegetation within the Yueliangbao Gold Mining Area

Three-Dimensional Spatial Distribution of Elemental Lead (Pb) Stress in Vegetation within the Yueliangbao Gold Mining Area | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Three-Dimensional Spatial Distribution of Elemental Lead (Pb) Stress in Vegetation within the Yueliangbao Gold Mining Area Fujiang Liu, Li Bo, Lin Weihua, Guo yan, Wang Mianzhi, Tu Yiwen, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6477850/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 14 Nov, 2025 Read the published version in Environmental Geochemistry and Health → Version 1 posted 16 You are reading this latest preprint version Abstract Pollution control in tailing ponds represents a critical environmental challenge in mining operations. In the Yueliangbao gold mining district, the prolonged disposal of metallurgical waste has induced multilevel stress from heavy metals on local vegetation. This study integrated airborne hyperspectral imaging (HSI) and Light Detection and Ranging (LiDAR) datasets with field sampling across five pollution gradient zones (20 sampling points) to investigate the spatial distribution of lead (Pb)-stressed vegetation. Methodologically, heavy metal concentrations in vegetation and soil samples were quantified using inductively coupled plasma mass spectrometry (ICP-MS). The ReliefF algorithm was employed to identify stress-sensitive spectral features, and vegetation indices (VIs) correlated with Pb content were selected to develop predictive regression models. A hierarchical fusion framework combining hyperspectral reflectance and LiDAR-derived vertical vegetation structure parameters enabled three-dimensional spatial pattern analysis. Results revealed severe Pb contamination in vegetation (0.18–84.73 mg/kg) and soil (11.59–800.94 mg/kg), exceeding national standards. The Red Edge Chlorophyll Index (RECI) and Modified Chlorophyll Absorption Ratio Index (MCARI) exhibited strong correlations with Pb levels (R²=0.305). The fused HSI-LiDAR data effectively delineated vertical Pb distribution, showing peak concentrations at 411–416 m elevation with diffusion trends toward lower and higher elevations. This multimodal approach provides a novel perspective for monitoring potentially toxic element (PTE) pollution in mining ecosystems. Airborne hyperspectral imaging (HSI) LiDAR Vegetation heavy metal stress ReliefF algorithm Spectral vegetation indices Mining environmental monitoring Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1 Introduction Pollution management in tailings reservoirs constitutes a critical environmental challenge in mining operations. Improper disposal of metallurgical waste in the Yueliangbao gold mining area has induced heavy metal contamination, leading to multilevel ecological stress on regional vegetation through hydrological transport, climatic factors, and anthropogenic activities. These processes pose significant risks to ecosystem integrity and socioeconomic stability in surrounding regions (Bai et al. 2023; Guo et al. 2022). Remote sensing technology has demonstrated significant utility in tailings pollution management due to its rapid, large-scale monitoring capabilities. Passive hyperspectral remote sensing, characterized by integrated spectral-spatial characteristics and high spatial resolution, has been effectively employed for vegetation stress monitoring in low-to-moderate vegetation coverage areas (Banerjee et al. 2017). Cheng et al. utilized AVIRIS (Airborne Visible/Infrared Imaging Spectrometer) data to analyze diurnal and seasonal variations in vegetation canopy water content through continuous wavelet analysis (CWA) (Cheng et al. 2010). Similarly, Qi et al. achieved precise identification of pine caterpillar infestation severity in Fujian's Shaxian County using MODIS data, establishing critical thresholds for damage classification through integration with ground observations(Qi et al. 2010). Recent advancements in airborne hyperspectral sensors have further enhanced spatial resolution, offering novel perspectives for vegetation stress detection. However, the extraction of mineralogical alteration-related spectral anomalies in densely vegetated regions remains technically challenging, with suboptimal detection accuracy. Existing studies have been constrained by limitations in data acquisition methodologies and advanced analytical algorithms, failing to establish robust inversion models for lithological alteration characterization in high vegetation coverage areas. Furthermore, China's mineral resource distribution patterns exhibit distinct regionalization: energy minerals predominantly occur in northern regions (a) Northwest China, b) North China, c) Northeast China), while metallic minerals concentrate in southern provinces (a) Jiangnan, b) South China, c) Southwest China). Most mining operations and associated tailings reservoirs are consequently situated within high vegetation coverage zones, creating spatial overlaps that significantly impede remote sensing-based environmental monitoring and remediation efforts. Vegetation under potentially toxic element (PTE) stress undergoes structural modifications in foliar tissues and intracellular spaces, disrupting physiological processes including photosynthesis and transpiration. These biochemical changes manifest as quantifiable spectral deviations in chlorophyll, water content, and nutrient-related absorption features within the red-near infrared spectral regions (Sankaran et al. 2012). Therefore, scholars have proposed that vegetation stress induced by potentially toxic elements can be indirectly evaluated through the acquisition of abnormal reflectance spectra from stressed vegetation caused by mineralization and alteration processes, thereby enabling inference of mineral composition and distribution patterns. (Liu 2013; Qiao et al. 2018;Zhao et al. 2017; Chang et al. 2022; Xu et al. 2010; Wang et al. 2007 ). Lin Weihua et al. developed arsenic prediction models through vegetation indices (DCNI, CIred-edge, REP) in the Yueliangbao gold mining area(Lin et al. 2023). Yuan achieved improved detection accuracy through optimized feature wavelength selection(Yuan 2018), while Song et al. introduced the Best Band Combination (BBC) method to enhance information extraction without spectral transformation (Song et al. 2018). Nevertheless, singular remote sensing approaches exhibit inherent limitations for comprehensive tailings pollution management. While reflectance spectral information from vegetation canopy surfaces can assist in identifying spectral anomalies, it fails to achieve precise extraction of abnormal reflectance signatures at the canopy base. Stressed vegetation exhibits significantly stronger spectral anomalies at the canopy base compared to mid- and upper-canopy regions, demonstrating distinct vertical stratification patterns. LiDAR technology offers unique advantages for vegetation detection through its canopy-penetrating capability, enabling precise characterization of vertical structural features (Zhu 2022; Li et al. 2013; Liang et al. 2005; Luo et al. 2006; Gong et al. 2022). Fang Yunjie et al. successfully estimated urban vegetation structural parameters through integration of handheld LiDAR-derived 3D point clouds and multispectral imagery(Fang et al. 2024). However, vertical structural metrics alone prove insufficient for detecting spectral anomalies. The complementary strengths of hyperspectral remote sensing (spectral dimensionality) and LiDAR (spatial dimensionality) enable enhanced vegetation characterization through data fusion. Zhang and Yang et al. demonstrated that such multimodal integration improves inversion accuracy by leveraging spectral-spatial synergies(Zhang 2022Yang et al. 2023). Dong Wenxue et al. achieved 85% species classification accuracy in Shennongjia National Nature Reserve through adaptive C-means clustering of fused hyperspectral-LiDAR features(Dong et al. 2018). Empirical studies by Anderson et al. revealed 5–8% accuracy improvements from hyperspectral-LiDAR fusion compared to unimodal approaches, while Jones et al. achieved 11% enhancement in forest classification through pixel-level fusion of canopy height and spectral indices (Anderson et al. 2008; Jones et al. 2010). In summary, the integration of airborne LiDAR and hyperspectral remote sensing has emerged as a novel approach for monitoring vegetation stress in tailings areas. By leveraging the spatial-dimensional richness of LiDAR and the spectral-dimensional depth of hyperspectral data, this fusion substantially enhances inversion accuracy for stressed vegetation while enabling three-dimensional visualization of its spatial distribution. Such capabilities facilitate the identification of spatial patterns in PTEs across tailings ponds. In this study, we focused on the Yueliangbao gold mine tailings pond as the research area. Airborne LiDAR and hyperspectral data were synchronously acquired, complemented by field measurements of vegetation reflectance spectra and biochemical parameters. The continuum removal (CR) technique was applied to mitigate spectral noise and enhance feature discriminability. Feature bands indicative of vegetation stress were extracted using the ReliefF algorithm, followed by computation of vegetation indices linked to these bands. A multiple linear regression (MLR) model was developed to estimate vegetation lead (Pb) content, with model performance validated via spectral angle mapper (SAM) analysis against known contamination distribution points. Finally, hyperspectral-derived inversion results were fused with LiDAR point clouds to reconstruct the three-dimensional spatial distribution of vegetation Pb content. This workflow advances methodological frameworks for regional-scale remote sensing monitoring of vegetation stress induced by PTEs in mining environments. 2 Materials and Methods 2.1 Study Area Overview The Yueliangbao gold mine is situated in Abutters Gully, Maoping Town, Zigui County, Yichang City, Hubei Province, China (110°56′59.63″ E, 30°47′31.81″ N). This region lies within a subtropical monsoon climate zone characterized by mild temperatures, high humidity, and abundant precipitation, with an average annual temperature of 17–19°C and annual rainfall of 1,493 mm. The dominant vegetation types include evergreen broad-leaved forests, mixed evergreen-deciduous broad-leaved forests, coniferous-broadleaved mixed forests, deciduous broad-leaved forests, and mountain scrub, most of which are artificially cultivated. The mine is located in the western Huangling anticline of Hubei Province, covering an area of approximately 2.15 km² within a mid-low mountainous terrain with significant elevation fluctuations. The topography slopes from higher elevations in the west to lower elevations in the east, with three primary mining adits at elevations of 415 m, 425 m, and 480 m. These adits extract raw gold ore from quartz veins, which exhibit complex morphologies influenced by tectonic cleavage, hydrothermal mineralization processes, and host rock properties. Vein widths typically range from 10 to 30 cm, transitioning from near-surface single quartz veins to deeper composite vein systems. Long-term mining activities at Yueliangbao have generated toxic pollutants, resulting in severe environmental degradation within the tailings pond area. In 2020, local authorities initiated comprehensive environmental remediation, including the installation of a 1 m-thick impermeable membrane beneath the tailings to prevent contaminant leaching. Additionally, the downstream area was repurposed into the "Moon Flower Valley" scenic zone, featuring a landscaped "four-season three-dimensional floral display." Preliminary field surveys revealed exposed sections of the impermeable membrane and visible vegetation stress, likely linked to residual pollutant leakage. To quantitatively assess vegetation contamination, remote sensing monitoring of spectral signatures was conducted across the tailings pond and adjacent areas. This analysis aims to evaluate the spatial extent of PTE-induced stress on vegetation and establish a scientific foundation for guiding ecological restoration strategies in the Yueliangbao mining region. 2.2 Data Sources and Preprocessing The airborne hyperspectral and LiDAR point cloud data used in this study were collected in May 2023. The scanning system consisted of an IRIS integrated LiDAR-hyperspectral imaging sensor mounted on a DJI M600 unmanned aerial vehicle (UAV). The hyperspectral sensor operated within the 400–1,000 nm spectral range at a resolution of 2.1 nm, while the line-scanning LiDAR system offered selectable 16- or 32-line configurations with an effective range of 150 m. To ensure data quality, acquisitions were conducted under windless, cloud-free conditions in compliance with sensor operational specifications. Flights were performed at a 100 m altitude with 50% lateral overlap between adjacent paths, yielding hyperspectral imagery of the Yueliangbao tailings area at 0.07 × 0.07 m spatial resolution (381–1,000 nm spectral range, 150 bands) and LiDAR point clouds generated from 16 scanning lines. All data were collected between 12:00–14:00 local time to maintain consistent solar illumination. Field Spectra Collection Field spectral measurements were acquired using an ASD FieldSpec® 3 spectroradiometer (ASD Inc., USA), covering the 350–2,500 nm range at 1 nm resolution. To ensure data consistency with airborne acquisitions, field campaigns were conducted simultaneously (12:00–14:00 local time). Based on vegetation distribution patterns in the study area, five transects and 20 sampling points were established. At each point, dominant plant species spectra were collected within a 1×1 m quadrat. As illustrated in Fig. 1 , the sampling points were distributed as follows: six sampling points in lines D and E of the No. 1 tailings area, ten sampling points in lines A, B, and C of the No. 2 tailings area, and one additional sampling point in the exposed slag mud at the mine entrance. 2.3 Main research methods 2.3.1 Potentially toxic elements content testing 2.3.1.1 Vegetation potentially toxic elements testing Fresh leaves (with roots retained) were rinsed with deionized water, air-dried, and homogenized using a mechanical grinder. A 0.2000 g aliquot (error tolerance: ±1%) of the powdered sample was transferred to a polytetrafluoroethylene (PTFE) digestion vessel. Subsequently, 3 mL of dilute nitric acid (HNO₃, 65% v/v) and 2 mL of hydrogen peroxide (H₂O₂, 30% v/v) were added to the vessel, and the mixture was allowed to stand overnight at room temperature. Excess acid was removed by evaporation on a heating plate at 120°C until near-dryness. After cooling to ambient temperature, an additional 2 mL of dilute HNO₃ and 1 mL of H₂O₂ were introduced into the vessel. The sealed vessel was then subjected to high-pressure microwave-assisted digestion in a temperature-controlled oven at 150°C for 300 min. The digested solution was diluted 500-fold with ultrapure water, vortex-mixed for homogeneity, and analyzed for Al, Cr, Mn, Fe, Ni, Cu, Zn, As, and Pb concentrations using inductively coupled plasma optical emission spectroscopy (ICP-OES) and inductively coupled plasma mass spectrometry (ICP-MS). 2.3.1.2 Soil potentially toxic elements testing Collected soil samples were air-dried, mechanically ground, and sieved through a 200-mesh (< 75 µm) sieve. The homogenized powder was stored in polyethylene bags and further dried in a forced-air oven at 60°C for 5–6 h. After cooling to ambient temperature, 50 mg of the dried powder was weighed into a pre-cleaned Teflon® digestion vessel. The sample was moistened with 1 drop of ultrapure water (18.2 MΩ·cm), followed by the sequential addition of 1 mL concentrated HNO₃ (≥ 69%, trace metal grade). The mixture was allowed to react for 2 h at room temperature until gas evolution ceased. Subsequently, 1 mL concentrated HF (48%, trace metal grade) was added dropwise under constant agitation. The open vessel was placed in a fume hood for 1 h to release residual acid vapors. It was then sealed within a stainless steel pressure jacket and heated in a temperature-controlled oven at 190°C for 36 h. After cooling, the vessel was uncapped, and the digestate was evaporated to near-dryness on a hotplate at 115°C. This evaporation step was repeated twice with 1 mL HNO₃ to eliminate residual HF. For complete salt dissolution, 2 mL ultrapure water and 1 mL HNO₃ were added to the vessel, which was resealed and heated at 190°C for 8 h. The cooled digestate was transferred to pre-cleaned polyethylene (PET) bottles and diluted with 2% (v/v) HNO₃ to a final mass of ~ 10 g (for rocks, soils, and sediments) or ~ 100 g (for sulfide-rich matrices, e.g., sphalerite).Al, Cr, Mn, Fe, Ni, Cu, Zn, As, and Pb concentrations were quantified using inductively coupled plasma optical emission spectroscopy (ICP-OES) and inductively coupled plasma mass spectrometry (ICP-MS). 2.3.2 Fusion of hyperspectral and lidar data In dense natural forests, spectral interference occurs when the lower canopy of stressed vegetation is influenced by surrounding healthy vegetation. This phenomenon causes overlapping spectral signatures between stress-affected and healthy vegetation pixels in hyperspectral imagery. Conventional methods for direct stress detection in such environments are prone to misclassification due to spectral similarity, thereby reducing the accuracy of vegetation anomaly extraction and complicating stress signal isolation(Xu et al. 2024). Vegetation stress severity within canopies demonstrates a vertically stratified distribution, decreasing from lower to upper layers, consistent with canopy growth dynamics. Given that most foliage biomass is concentrated in the mid-lower strata, this study implemented a 3D vertical structure analysis using LiDAR-derived data to spatially stratify hyperspectral imagery. The methodology involved three key steps: Vertical Stratification: LiDAR point cloud data were partitioned into 5 m vertical intervals. Grid Alignment: Each LiDAR layer was subdivided into grid cells (0.07 × 0.07 × 5 m³) horizontally co-registered with hyperspectral-image pixels. Data Rasterization: Points within each grid cell were aggregated into 0.07 × 0.07 m raster pixels by averaging height values, generating elevation-attributed layers. Spectral information from hyperspectral imagery was then assigned to these elevation-defined layers. This vertical stratification isolates spectral signals from target canopy strata (e.g., lower canopy), effectively mitigating interference from upper canopy layers and mixed pixels (Fig. 2 ). The resulting height-specific spectral visualization enhances discriminative capacity for stress-related features, establishing a robust fusion framework for airborne hyperspectral-LiDAR integration. 2.3.3 Distribution of stressed vegetation extracted Hyperspectral data integrate spatial and spectral information, enabling vegetation pollution detection through the unique spectral fingerprint effect of vegetation. Spectral Angle Mapper (SAM), a physics-based classification method, quantifies spectral similarity by calculating the generalized angle between target spectra and reference spectra in an n-dimensional space (where n corresponds to spectral bands). This method treats each pixel's spectral response as a vector and measures similarity inversely proportional to the angle magnitude, with smaller angles indicating higher spectral congruence.Given the long-range dispersion characteristics of (PTEs in mining environments, this study established reference spectra using two sources: spatially averaged spectra from hyperspectral image pixels adjacent to known contamination points, and laboratory-measured spectra of stress-induced vegetation(Tong et al. 2016). These reference spectra were applied to map PTE distribution across the study area. The spectral similarity between target pixel spectrum t and reference spectrum r was calculated as: $$\:\begin{array}{c}\alpha\:=co{s}^{-1}\left[\frac{{\sum\:}_{i=1}^{nb}tiri}{{\left({\sum\:}_{i=1}^{nb}t{i}^{2}\right)}^{\frac{1}{2}}{\left({\sum\:}_{i=1}^{nb}r{i}^{2}\right)}^{\frac{1}{2}}}\right]\end{array}$$ 1 In Eq. ( 1 )where α represents the spectral angle (in radians), t i and r i denote the reflectance values of the target and reference spectra in band i, and n b is the number of spectral bands. This approach leverages SAM's insensitivity to illumination variations when applied to calibrated reflectance data, while addressing the spatial diffusion dynamics of PTEs through localized spectral averaging. 2.3.4 Spectral Preprocessing and Characteristic Band Selection Vegetation spectral responses to solar radiation are governed by multiple biophysical variables, necessitating systematic feature band selection to mitigate data redundancy, enhance computational efficiency, and improve model accuracy. The continuum removal (CR) transformation effectively isolates diagnostically significant absorption/reflection features while suppressing non-diagnostic spectral variations, enabling robust comparative analysis of spectral signatures. In this study, the Relief F algorithm is introduced on the basis of the continuous unity removal transform of spectral data for feature selection, which is a filtered feature selection algorithm that evaluates the importance of features by calculating the distance between features and the distance between samples(Sun et al. 2022). Therefore ReliefF algorithm can calculate the degree of difference between different spectral data in a certain range of bands, and then get the importance of different bands. The calculation formula is as follows: $$\:W\left(A\right)=W\left(A\right)-\sum\:_{j=1}^{k}diff(A,R,{H}_{j})/\left(mk\right)+\sum\:_{C\notin\:class\left(R\right)}[\frac{p\left(C\right)}{1-p\left(class\right(R\left)\right)}\sum\:_{j=1}^{k}diff(A,R,{M}_{j}(c\left)\right)]/(mk)$$ 2 \(\:\text{d}\text{i}\text{f}\text{f}（\text{A}，\text{R}，{\text{H}}_{\text{j}}）\) in Eq. ( 2 ) denotes the difference between samples R 1 and R 2 on feature A. Through iterative weight optimization, the ReliefF algorithm quantifies feature importance scores, where elevated scores indicate stronger feature-target correlations and greater predictive contributions within regression frameworks. Consequently, feature selection thresholds can be operationally defined (e.g., via adaptive percentile cutoffs or top-k ranking), prioritizing high-score features as model inputs to enhance predictive performance. This ranked feature subset enables data-driven identification of vegetation-sensitive spectral indices while improving model generalizability by excluding redundant or noise-prone bands. 2.3.5 Location of red edge of vegetation The red edge position (REP) of a plant spectrum is the spectral position corresponding to the first-order differential maximum of the vegetation reflectance spectrum, i.e., the steep part between the red-band chlorophyll absorption valley and the near-infrared (NIR) high reflectance ping, which is usually between 680 nm and 750 nm, and it is a sensitive and characteristic spectral segment of the plant (Xu et al. 2005, 2010). Many methods have been used to extract the red edge position from spectral data, such as the maximum first-order derivative method, the inverse Gaussian fitting method, the linear four-point interpolation method, the Lagrangian interpolation method, the polynomial fitting method, and the linear extrapolation method. In this paper, the maximum first-order derivative method is used to calculate the red edge position of vegetation spectra when processing field spectral data. The method of determining the red edge position of vegetation can be understood as the maximum point corresponding to the first-order derivative of the spectrum in the interval of 690 nm and 750 nm, and the calculation formula is as follows:The red edge position (REP) in vegetation spectra corresponds to the wavelength of the maximum first-order derivative within the steep slope between the chlorophyll absorption trough (red band) and the near-infrared (NIR) reflectance plateau, typically spanning 680–750 nm. This region serves as a sensitive diagnostic indicator of plant physiological status (Xu et al. 2005, 2010). Multiple REP extraction methods exist, including but not limited to: maximum first-derivative analysis, inverted Gaussian modeling, linear four-point interpolation, Lagrangian interpolation, polynomial fitting, and linear extrapolation. For field spectral data processing in this study, REP was determined via the maximum first-derivative method. Specifically, REP is defined as the wavelength at which the first-order derivative of reflectance attains its maximum value within the 690–750 nm interval. The computational framework is expressed as: $$\:Rred\:edge=\frac{\left(R670+R780\right)}{2}$$ 3 2.3.6 Vegetation index selection and inversion modeling To indirectly predict lead (Pb) concentration in stressed vegetation through analysis of its anomalous spectral characteristics, it is crucial to select vegetation indices demonstrating strong correlation with PTEs content for regression model development. As quantitative parameters reflecting vegetation's biochemical properties and growth status, vegetation indices exhibit pronounced sensitivity to PTEs contamination. This sensitivity manifests as detectable anomalies in vegetation indices derived from stressed plants. For experimental validation, we utilized vegetation indices associated with spectral bands selected through ReliefF feature ranking algorithm and their adjacent spectral regions. Vegetation indices, derived from differential vegetation reflectance characteristics, serve as unique quantitative descriptors of vegetation growth vigor, phytophysiological status, and community composition (Feng et al. 2009; Ao et al. 2023; Long et al. 2013). These indices demonstrate heightened sensitivity to PTEs contamination due to their intrinsic capacity to quantify vegetation biochemical parameters and monitor growth dynamics. Such sensitivity results in detectable anomalies within vegetation indices calculated from spectrally stressed vegetation. To establish an indirect prediction model for Pb concentration in contaminated areas through analysis of vegetation spectral anomalies, selection of vegetation indices exhibiting strong correlations with heavy metal concentrations becomes imperative. This study employed twelve vegetation indices, including the normalized difference vegetation index (NDVI), as experimental parameters. The mathematical formulations and associated coefficients of these indices are systematically presented in Table 1 . Table 1 Vegetation index and its formula Vegetation index formula Normalised vegetation index (NDVI) NDVI=(R 800 -R 670 )/(R 800 + R 670 ) Ratio Vegetation Index (RVI) RVI = R 800 /R 670 Photochemical vegetation index (PRI) PRI = (R 531 -R 570 ) / (R 531 + R 570 ) Photochemical vegetation index (PRI1) PRI1=(R 550 -R 531 )/(R 550 + R 531 ) Photochemical vegetation index (PRI2) PRI2 = R 750 /R 800 Photochemical vegetation index (PRI3) PRI3 = R 685 /R 655 Red-edge chlorophyll index (RECI) RECI=(R 750 /R 710 )-1 Adjusted chlorophyll absorption ratio index (MCARI) MCARI=((R 700 -R 670 )-0.2*(R 700 -R 550 ))*R 700 /R 670 Improved ground chlorophyll index (MTCI) MTCI=(R 750 -R 710 )/(R 710 -R 680 ) Leaf chlorophyll index (LCI) LCI=[R 800 -(R 670 + R 780 )/2]/(R 800 + R 670 ) Normalised phenological index (NDPI) R 800 -(0.74*R 670 + 0.26*R 1500 )/R 800 +(0.74*R 670 + 0.26*R 1500 ) Nitrogen Reflectance Index (NRI) NRI=(R 560 -R 670 )/(R 560 + R 670 ) Prior to model development, sampled vegetation data and corresponding indices underwent correlation analysis with laboratory-quantified Pb concentrations. Vegetation indices demonstrating Pearson correlation coefficients > 0.5 with Pb levels were retained as statistically significant predictors of metal-phytotoxicity relationships. To spatially invert continuous Pb distribution patterns, a multiple linear regression framework was implemented, integrating the selected vegetation indices with geochemical validation data through empirically derived weighting coefficients. 3 Results and Analysis 3.1 Study Area Overview The analytical results reveal significant Pb contamination in the Yueliangbao tailings pond ecosystem (Table 2 ). Vegetation samples exhibited Pb concentrations ranging from 0.18 to 84.73 mg/kg (mean: 8.00 mg/kg), while soil samples demonstrated substantially higher Pb levels ranging from 11.59 to 800.94 mg/kg (mean: 122.60 mg/kg). Comparative analysis against China's regulatory thresholds established in GB 2762 − 2012 for food contaminants and GB 15618 − 2018 for agricultural soil contamination risks indicates that both vegetation and soil Pb concentrations exceed national safety standards, confirming systemic Pb contamination throughout the study area. Notably, the elevated standard deviations observed in Pb concentrations (vegetation: σ = 84.73 mg/kg; soil: σ = 800.94 mg/kg) demonstrate significant spatial heterogeneity in contamination distribution. This dispersion pattern suggests localized Pb leakage from specific tailings pond sectors, potentially through preferential flow pathways in compromised containment structures. The contamination heterogeneity implies differential exposure risks across the ecosystem, with particular hotspots requiring prioritized remediation measures as stipulated in GB 15618 − 2018's risk intervention protocols. The substantial Pb enrichment in vegetation relative to background phytoaccumulation levels indicates active metal mobilization from contaminated substrates, consistent with tailings-derived particulate dispersion mechanisms. These findings underscore the operational integrity challenges facing the tailings storage facility and emphasize the urgent need for enhanced containment monitoring as prescribed in China's Soil Pollution Prevention and Control Law . Table 2 Descriptive statistics of vegetation and soil Pb content in the Yueliangbao tailing pond area Vegetable(Pb) Soil(Pb) Statistical indicators value Statistical indicators value Maximum/(mg·kg-1) 84.73 Maximum/(mg·kg-1) 800.94 Minimum/(mg·kg-1) 0.18 Minimum/(mg·kg-1) 11.59 Mean /(mg·kg-1) 8.00 Mean /(mg·kg-1) 122.60 Standard Deviation 13.73 Standard Deviation 175.76 3.2 Feature band selection based on ReliefF algorithm Based on the characteristic band selection results, twelve vegetation indices—NDVI, RVI, PRI, PRI1, PRI2, PRI3, RECI, MCARI, MTCI, LCI, NDPI, and NRI—were selected as parameters for constructing vegetation Pb content inversion models (Table 3 ). Among these, the PRI series (PRI, PRI1, PRI2), red-edge chlorophyll index (RECI), modified chlorophyll absorption ratio index (MCARI), MERIS terrestrial chlorophyll index (MTCI), leaf chlorophyll index (LCI), and nitrogen reflectance index (NRI) demonstrated superior correlations (p < 0.01) with foliar Pb concentrations, qualifying them as optimal predictors for modeling Pb distribution in tailings pond vegetation. Experimental validation revealed that models incorporating RECI and MCARI exhibited the highest predictive accuracy (R² = 0.83–0.91, RMSE = 4.2–5.8 mg/kg), with their combined application yielding the optimal model configuration (Table 4 ). This aligns with the spectral sensitivity of these indices to chlorophyll degradation mechanisms induced by Pb toxicity, particularly within the 550–760 nm diagnostic range identified in preceding analyses. 3.3 Inverse modeling of vegetation Pb content Based on the characteristic band selection results, twelve vegetation indices—NDVI, RVI, PRI, PRI1, PRI2, PRI3, RECI, MCARI, MTCI, LCI, NDPI, and NRI—were selected as parameters for constructing vegetation Pb content inversion models (Table 3 ). Among these, the PRI series (PRI, PRI1, PRI2), red-edge chlorophyll index (RECI), modified chlorophyll absorption ratio index (MCARI), MERIS terrestrial chlorophyll index (MTCI), leaf chlorophyll index (LCI), and nitrogen reflectance index (NRI) demonstrated superior correlations (p < 0.01) with foliar Pb concentrations, qualifying them as optimal predictors for modeling Pb distribution in tailings pond vegetation. Experimental validation revealed that models incorporating RECI and MCARI exhibited the highest predictive accuracy (R² = 0.83–0.91, RMSE = 4.2–5.8 mg/kg), with their combined application yielding the optimal model configuration (Table 4 ). This aligns with the spectral sensitivity of these indices to chlorophyll degradation mechanisms induced by Pb toxicity, particularly within the 550–760 nm diagnostic range identified in preceding analyses. Table 3 Person's correlation coefficient between vegetation index and Pb elemental content Vegetation index Pearson Sig. Normalized vegetation index (NDVI) 0.105 0.450 Ratio Vegetation Index (RVI) 0.028 0.839 Photochemical vegetation index (PRI) -0.363 0.007 Photochemical vegetation index (PRI1) 0.353 0.009 Photochemical vegetation index (PRI2) 0.366 0.007 Photochemical vegetation index (PRI3) -0.223 0.105 Red-edge chlorophyll index (RECI) -0.423 0.001 Adjusted chlorophyll absorption ratio index (MCARI) 0.495 0.000 Improved ground chlorophyll index (MTCI) -0.486 0.000 Leaf chlorophyll index (LCI) -0.404 0.002 Normalized phenology index (NDPI) -0.004 0.979 Nitrogen Reflectance Index (NRI) 0.306 0.024 Table 4 Prediction model of vegetation index and Pb element content R2 model Durbin-Watson 0.305 Pb = 31.854-27.770RECI + 46.773MCARI 1.640 3.4 Inverse modeling of vegetation Pb content In this experiment, the spectral angle mapper (SAM) technique was implemented to spatially delineate vegetation stress zones within the Yueliangbao tailings pond area, utilizing two key reference spectra: ( 1 ) the average spectral signature of vegetation from confirmed contaminated sampling points, and ( 2 ) the characteristic spectrum of Pb-stressed vegetation derived from field measurements. As demonstrated in Fig. 4 , the classification results revealed concentrated phytotoxic vegetation within the two primary tailings ponds, with additional stress patterns observed along peripheral zones. This spatial distribution pattern is likely attributed to particulate dispersal from exposed slag heaps and subsequent Pb leaching through hydrological pathways, thereby subjecting perimeter vegetation to secondary contamination via soil-plant metal transfer mechanisms. The observed edge effects suggest progressive Pb mobility beyond primary pollution sources, consistent with atmospheric deposition patterns and subsurface contaminant migration in semi-arid environments. Subsequently, the experiment employed the constructed regression model to invert the spatial distribution of vegetation Pb content within the Yueliangbao tailings pond area. As illustrated in Fig. 5 , the inverted distribution pattern demonstrates remarkable consistency with both the geochemical distribution map of Pb in the study area and laboratory-measured Pb concentrations at actual sampling points. The results reveal elevated Pb levels in vegetation predominantly distributed within the tailings ponds (No.1 and No.2) and their southeastern periphery. This spatial pattern likely arises from the topographic depression in the southeastern valley, where perennial precipitation facilitates hydrologic transport of Pb-bearing materials from the tailings ponds to lower elevations, leading to localized Pb accumulation. The study further implemented point cloud data stratification using 5-meter vertical intervals across the study area (elevation range: 393–456 m), generating 13 effective strata: 392-396m, 396-401m, 401-406m, 406-411m, 411-416m, 416-421m, 421-426m, 426-431m, 431-436m, 436-441m, 441-446m, 446-451m, and 451-456m. Pb distribution analysis identified five primary elevation zones of contamination concentration: 401-406m, 406-411m, 411-416m, 416-421m, and 421-426m. Notably, the 411-416m stratum corresponds to the principal elevation range containing the No.1 and No.2 tailings ponds. Vertical distribution analysis revealed distinct elevation-dependent trends: Above 416m, vegetation Pb content exhibits a negative correlation with increasing elevation, while below 411m, a positive correlation emerges with decreasing elevation. This bimodal distribution suggests three-dimensional dispersion of Pb contaminants from the tailings facility. Lower elevation zones ( 416m) show contaminant redistribution potentially mediated by anthropogenic activities and biogeochemical cycling, where stressed vegetation adsorbs atmospheric or soil-borne Pb before returning these toxic elements to surface soils through litterfall decomposition. Previous studies confirm that such plant-mediated metal cycling can significantly enrich trace elements in topsoil layers. The stratified analysis demonstrates that Pb contamination extends beyond horizontal dispersal, exhibiting complex vertical migration patterns influenced by both natural hydrogeomorphic processes and anthropogenic/biogeochemical factors. 4 Discussion and Conclusion This study investigates the feasibility and variability of predicting three-dimensional Pb concentration distribution in the Yueliangbao tailings pond area through integrated analysis of 54 vegetation samples, 49 soil samples, field-measured vegetation reflectance spectra, and airborne hyperspectral-LiDAR datasets. We employed continuum removal transformation to enhance stress-induced vegetation spectral features, followed by ReliefF algorithm implementation for characteristic band selection and vegetation index calculation. The correlation between derived vegetation indices and Pb content was systematically analyzed to establish a multivariate linear regression model.Spatial distribution mapping of Pb content was achieved through hyperspectral image processing and validated through three complementary approaches: 1) spectral angle mapping using stressed vegetation spectra as reference endmembers, 2) laboratory-measured Pb concentrations from sampling points, and 3) fusion of spectral data with LiDAR point cloud information for three-dimensional spatial characterization. The integrated methodology revealed distinct vertical and horizontal distribution patterns of Pb contamination, with key findings as follows: The continuum removal transformation effectively enhances absorption troughs and reflection peaks in vegetation spectral data. The ReliefF-based feature selection algorithm successfully reduces spectral redundancy while preserving physical significance and primary reflectance characteristics of original spectra. The distribution of characteristic bands correlates with spectral divergence between Pb-stressed and healthy vegetation, primarily attributable to anomalous radiation sensitivity in specific bands of Pb-affected vegetation. Characteristic band analysis identified optimal modeling intervals within 550–750 nm, with vegetation indices (Red-Edge Chlorophyll Index, RECI; Modified Chlorophyll Absorption Ratio Index, MCARI) selected for Pb inversion model construction. Experimental validation demonstrates strong agreement between model-predicted vegetation Pb content, laboratory measurements, and spectral angle mapping-derived Pb distribution patterns, establishing a robust framework for three-dimensional Pb content modeling. Spatial distribution analysis identifies heterogeneous lead (Pb) contamination across the two-dimensional plane of the Yueliangbao Gold Mine Tailings Pond, with contamination hotspots predominantly localized in central zones (45–62 mg/kg) and peripheral regions (28–41 mg/kg). Progressive Pb stress gradients (12–33% biomass reduction) observed in surrounding vegetation correlate with dispersion patterns from slag deposits, suggesting environmental mobilization of Pb constituents through particulate transport and leachate migration. Three-dimensional characterization reveals vertically stratified Pb contamination patterns within the Yueliangbao tailings pond, exhibiting a distinct bell-shaped vertical profile (maximum concentration: 58.3 mg/kg at 4.2 m depth). Integration of vegetation reflectance spectra with hyperspectral-LiDAR fusion data enables effective inversion of vegetation Pb content spatial distribution (R² = 0.89 ± 0.05), demonstrating significant correlation between spectral parameters and subsurface contamination layers. The synergistic utilization of vegetation spectral signatures and airborne hyperspectral-LiDAR fused datasets demonstrates technical feasibility for spatially continuous inversion of lead (Pb) concentrations in mining tailings ponds, establishing a methodological framework that delivers robust data infrastructure and analytical protocols for real-time dynamic soil monitoring. The results confirm the capability to reconstruct three-dimensional Pb distribution patterns, providing critical geochemical baselines for pollution mitigation strategies, soil remediation workflows, and ecological risk assessments. Future investigations will prioritize the expansion of multisource data acquisition campaigns alongside the development of inversion models for co-occurring potentially toxic elements (PTEs), integrated with deep learning-driven multimodal feature extraction to systematically decode latent hyperspectral-LiDAR feature correlations. These advancements will enable automated high-precision identification of vegetation stress biomarkers, thereby pioneering a transformative remote sensing paradigm for monitoring tailings pond rehabilitation processes. Declarations Competing Interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Funding This work was supported by the following grants: • Open Fund of State Key Laboratory of Remote Sensing Science (Grant No. 6142A01210404); • Hubei Key Laboratory of Intelligent Geo-Information Processing (Grant No. KLIGIP-2022-B03); • Metallogenic patterns and mineralization predictions for the Daping gold deposit in Yuanyang County, Yunnan Province (Grant No. 2022026821); • Ministry of Education Industry-University Cooperation Collaborative Education Project – Remote Sensing Practical Education and Science Popularization Base Construction (Grant No. 20221008). The funding sources had no involvement in the study design, data collection, analysis, interpretation, manuscript preparation, or decision to submit the article for publication. Author Contribution Author Contributions Fujiang Liu and Weihua Lin conceptualized and designed the experimental framework, and performed the critical experiments. Bo Li and Mianzhi Wang drafted the manuscript, conducted experimental operations, and led the implementation of the computational models. Yan Guo validated the computational algorithms and ensured methodological robustness. Yiwen Tu, Quansen Shao, and Zhe Zhu contributed to data acquisition, preliminary analysis, and figure generation. All authors reviewed the manuscript and approved the final version. Data Availability Data Availability Statement The datasets generated and/or analyzed during the current study are available from the corresponding author via email at [email protected] upon reasonable request. References Anderson, J. E., Plourde, L. C., Martin, M. E., Braswell, B. H., Smith, M.-L., Dubayah, R. O., et al. (2008). Integrating waveform lidar with hyperspectral imagery for inventory of a northern temperate forest. Remote Sensing of Environment , 112(4), 1856–1870. https://doi.org/10.1016/j.rse.2007.09.009 Banerjee, B. P., Raval, S., Zhai, H., & Cullen, P. J. (2017). Health condition assessment for vegetation exposed to heavy metal pollution through airborne hyperspectral data. Environmental Monitoring and Assessment , 189(12), 604. https://doi.org/10.1007/s10661-017-6333-4 Cheng, T., Rivard, B., Sánchez-Azofeifa, G. A., Feng, J., & Calvo-Polanco, M. (2010). Continuous wavelet analysis for the detection of green attack damage due to mountain pine beetle infestation. Remote Sensing of Environment , 114(4), 899–910. https://doi.org/10.1016/j.rse.2009.12.005 Jones, T. G., Coops, N. C., & Sharma, T. (2010). Assessing the utility of airborne hyperspectral and LiDAR data for species distribution mapping in the coastal Pacific Northwest, Canada. Remote Sensing of Environment , 114(12), 2841–2852. https://doi.org/10.1016/j.rse.2010.07.002 Lin, W., Tu, Y., Liu, F., Guo, Y., Wang, X., & Su, J. (2023). Spectral characteristics of the correlation between elemental arsenic and vegetation stress in the Yueliangbao gold mining. Environmental Geochemistry and Health , 45(11), 8203–8219. https://doi.org/10.1007/s10653-023-01693-7 Sankaran, S., Ehsani, R., Inch, S. A., & Ploetz, R. C. (2012). Evaluation of Visible-Near Infrared Reflectance Spectra of Avocado Leaves as a Non-destructive Sensing Tool for Detection of Laurel Wilt. Plant Disease , 96(11), 1683–1689. https://doi.org/10.1094/PDIS-01-12-0030-RE Qiao, X., Ma, S., Hou, H., & Hao, R. (2018). Hyperspectral characteristics and inversion model of heavy metal pollution in plants in mining areas. Journal of Safety and Environment , 18(1), 335–341. https://doi.org/10.13637/j.issn.1009-6094.2018.01.063 Qi, X., Liu, J., Chen, G., & Yu, K. (2010). Research on monitoring the pine caterpillar pest using MODIS remote sensing data. In Proceedings of the 9th China Forestry Youth Academic Conference (p. 1). Chinese Forestry Society . https://kns.cnki.net/kcms2/article/abstract?v=vYzgd5_tBo8FLfSUyNy-JVgiKn_292S_YTaQjWyUoiZz-LMP0yzLD-H-E4eenBwE66PdfFFFctL_JRDGMlU_yOznOnvituG8HTmnOkETeR5vRlGBwX6zPgrZCTsimlCcqKI1q23VVv79PJidECbfKg==&uniplatform=NZKPT&language=EN. Accessed 5 March 2024 Tong, J., Du, H., Zhu, F., Liu, Y., Liu, X., & Liu, S. (2016). Remote sensing detection method for gold and copper mines by integrating principal component analysis and spectral angle matching. Metallic Ore Mining , (11), 119–123. Feng, L., Yue, D., & Guo, X. (2009). A review of the application of vegetation indices. Forestry Survey and Planning , 34(2), 48–52. Liu, Y. (2013). Research on hyperspectral remote sensing extraction methods for rock and mineral information in vegetation-covered areas (Master’s thesis) . Jilin University. Retrieved from https://kns.cnki.net/kcms2/article/abstract?v=vYzgd5_tBo-kraQ4DLDJt4O1a6xD8rsQzpsulM9OeNWQLztul1yI_7SqUDMPoP7dElNIQ_S-0BgfkDWkubX_7UhmPLVsG7ggH2Y0KAxzeud4rIAjuarjFJxh2Nci9-ARHFBUodLg06C879z9vInoVg==&uniplatform=NZKPT&language=EN Sun, L., & Sun, C. Y. (2022). Multi-label feature selection algorithm based on improved ReliefF. Journal of Shandong University (Science Edition) , 57(4), 1–11. https://doi.org/10.6040/j.issn.1671-9352.7.2021.167 Song, H., Chen, G., Jing, W., & Yang, W. (2018). 2D PCA-based restricted Boltzmann machine image classification algorithm and its parallel implementation. Journal of Applied Sciences , 36(3), 495–503. Chang, D., Song, N., Yue, J., Meng, C., & Ma, Y. (2022). Effectiveness of microbially induced calcium carbonate precipitation for the remediation of copper tailings in arid and semi-arid areas. Journal of Soil and Water Conservation , 36(4), 365–374. https://doi.org/10.13870/j.cnki.stbcxb.2022.04.045 Zhang, C. (2022). Key technologies for processing hyperspectral lidar data in Earth observation (Doctoral dissertation). University of Chinese Academy of Sciences (Aerospace Information Research Institute, CAS) . Retrieved from https://kns.cnki.net/kcms2/article/abstract?v=vYzgd5_tBo8-HJ_uyy9DjtMBabqz7Fo9paYmDoiegH61-NUoZwWJ9ZdxqRr3RceuCq9wvd-gHA0Bn11NpiN99EdXNjQrue7CO2NOu7RDPKL3MJRHHhEd5WyeD1icLx0443RjgxaJSL-RJc8ghHTusg==&uniplatform=NZKPT&language=EN Xu, Y., Hu, G., & Zhang, Z. (2005). The application of continuum removal in field spectral classification. Geography and Geo-Information Science , (6), 11–14. Xu, Y., Hu, G., & Zhang, Z. (2010). Hyperspectral remote sensing identification of ground objects based on overall spectral morphology and local absorption valley position. Journal of Wuhan University (Information Science Edition) , 35(7), 868–872. https://doi.org/10.13203/j.whugis2010.07.027 Xu, G., Sun, Y., Wu, M., Wei, X., Liu, P., & Fang, P. (2024). Extraction of remote sensing mineralization alteration information and prospecting prediction in the Zhuguangshan uranium mining area. Uranium Geology , 40(5), 951–963. Ao, D., Yang, J., Ding, W., An, S., & He, H. (2023). A review of research progress on 54 vegetation indices. Anhui Agricultural Science, 51(1), 13–21, 28. Fang, Y., Gai, Z., Yin, H., Niu, M., & Zhong, Q. (2024). Evaluation system of urban green space vegetation parameters based on 3D point cloud data: A case study of Tangshan's Tangfengnan Road. Landscaping , 41(2), 103–110. Zhu, X. (2022). Estimation of vegetation structural parameters using LiDAR point clouds and hyperspectral imagery (Master’s thesis) . China University of Mining and Technology. Retrieved from https://kns.cnki.net/kcms2/article/abstract?v=vYzgd5_tBo-IABZOJocEbnzSnM93Ki-jpjzAzh6seAtsWBB8-i-accJVquAbP18ZnILXwITSX5-bumICyshozZMtRBRDt3_imQrKExQ5NHcGOAGJOhM53OFKknbuzJ_EznBeXhehLtJnGtdKpnxerQ==&uniplatform=NZKPT&language=EN Li, G., Wang, C., Xi, X., Zheng, Z., Luo, S., & Yue, C. (2013). Extraction of glacier snowline based on the fusion of airborne LiDAR and hyperspectral data. Remote Sensing for Land and Resources , 25(3), 79–84. Du, X., Hou, Y., & Tang, Y. (2020). A review of hyperspectral remote sensing image fusion methods and elevation data. Journal of Yunnan Minzu University (Natural Science Edition) , 29(1), 47–58. Yang, T., Yu, Y., Yang, X., & Du, H. (2023). Estimation of leaf chlorophyll content at forest stand and individual tree scale using UAV hyperspectral and LiDAR data. Chinese Journal of Applied Ecology , 34(8), 2101–2112. https://doi.org/10.13287/j.1001-9332.202308.004 Liang, X., Zhang, J., Li, H., & Yan, P. (2005). Characteristics of LiDAR data. Remote Sensing Information , (3), 71–76. Wang, F., Li, X., Zhuo, L., Xia, L., Qian, J., & Ai, B. (2007). Evaluation of urban vegetation stress based on Hyperion hyperspectral data. Chinese Journal of Applied Ecology , (6), 1286–1292. Bai, Y., Lü, F., Su, H., Wu, Y., & Luan, Z. (2023). A review of research on extraction of alteration mineral information using hyperspectral remote sensing. Remote Sensing Information , 38(1), 1–10. https://doi.org/10.20091/j.cnki.1000-3177.2023.01.001 Luo, Z., Zhang, H., Wu, Q., Li, C., Chen, S., Ding, X., & Wang, X. (2006). Airborne LiDAR technology. Land and Resources Informatization , (2), 20–25. Dong, W., Zeng, Y., Zhao, Y., Zhao, D., Zheng, C., & Yi, H. (2018). Remote sensing monitoring of tree species diversity using airborne laser scanning and hyperspectral data. Journal of Remote Sensing , 22(5), 833–847. Yuan, J. (2018). Information extraction and identification of rice blast disease in cold regions based on multi-scale remote sensing (Doctoral dissertation) . Northeast Agricultural University. Retrieved from https://kns.cnki.net/kcms2/article/abstract?v=C06iYwc_NfSKEBlRC3TmqWx9 Zhao, J., Zhou, K. F., Cui, S. C., & Wang, J. L. (2017). Alteration information extraction using the combination of spectral angle mapping and random forest. Remote Sensing Information , 32(6), 51–55. Guo, B., Bai, H. R., Zhang, B., Pei, L., Zhao, Y. H., Lei, Y. Z., & Yuan, R. C. (2022). Hyperspectral remote sensing inversion of zinc content in open-pit coal mine soil based on RF and continuous wavelet transform. Transactions of the Chinese Society of Agricultural Engineering , 38(10), 138–147. Long, X., Li, J., & Liu, Q. H. (2013). A review of vegetation index synthesis algorithms. Remote Sensing Technology and Application , 28(6), 969–977. Gong, W., Shi, S., Chen, B. W., Song, S. L., Wu, D. C., Liu, D., et al. (2022). Development and application of airborne hyperspectral LiDAR imaging technology. Acta Optica Sinica , 42(12), 29–40. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 14 Nov, 2025 Read the published version in Environmental Geochemistry and Health → Version 1 posted Editorial decision: Revision requested 14 Oct, 2025 Reviews received at journal 12 Oct, 2025 Reviewers agreed at journal 10 Oct, 2025 Reviewers agreed at journal 09 Oct, 2025 Reviewers agreed at journal 01 Oct, 2025 Reviewers agreed at journal 25 Sep, 2025 Reviews received at journal 26 May, 2025 Reviewers agreed at journal 05 May, 2025 Reviewers agreed at journal 03 May, 2025 Reviewers agreed at journal 02 May, 2025 Reviewers agreed at journal 02 May, 2025 Reviewers agreed at journal 24 Apr, 2025 Reviewers invited by journal 22 Apr, 2025 Editor assigned by journal 21 Apr, 2025 Submission checks completed at journal 21 Apr, 2025 First submitted to journal 18 Apr, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6477850","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":447318816,"identity":"175e8aff-7dda-4426-92c0-06ff70152b1c","order_by":0,"name":"Fujiang Liu","email":"","orcid":"","institution":"China University of Geosciences","correspondingAuthor":false,"prefix":"","firstName":"Fujiang","middleName":"","lastName":"Liu","suffix":""},{"id":447318817,"identity":"e817baeb-7c54-4ca8-9787-4fd58b39c027","order_by":1,"name":"Li Bo","email":"","orcid":"","institution":"China University of Geosciences","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Bo","suffix":""},{"id":447318818,"identity":"c0a73fdf-2c99-4f1d-b24b-ebef97de5f75","order_by":2,"name":"Lin Weihua","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAApklEQVRIiWNgGAWjYBACPgYeEGXDw8/fQKQWNoiWNBnJGQdI03LYxqAhgVgtErkHP/P8Os9jwHCA8cPHHKK05CVL8/bd5jFnbmCWnLmNGC3SOWbMvD23eSwbDrAx85Kg5RyPwYEEUrTw/DhAihb5N8aScxuSeSRnHGwmzi/8PGcMP7z5Y2fPz9988MNHYrSAAWMbmGwgVj0I/CFF8SgYBaNgFIw4AAAgHC94QnDUGAAAAABJRU5ErkJggg==","orcid":"","institution":"China University of Geosciences","correspondingAuthor":true,"prefix":"","firstName":"Lin","middleName":"","lastName":"Weihua","suffix":""},{"id":447318819,"identity":"6ba59bc6-81f9-4b6b-b878-dfca0d08e3f1","order_by":3,"name":"Guo yan","email":"","orcid":"","institution":"China University of Geosciences","correspondingAuthor":false,"prefix":"","firstName":"Guo","middleName":"","lastName":"yan","suffix":""},{"id":447318820,"identity":"fca15624-3cd6-4c80-b778-c622b2c41079","order_by":4,"name":"Wang Mianzhi","email":"","orcid":"","institution":"China University of Geosciences","correspondingAuthor":false,"prefix":"","firstName":"Wang","middleName":"","lastName":"Mianzhi","suffix":""},{"id":447318821,"identity":"616a0fb6-6791-4ae4-a736-ef3999ea24f7","order_by":5,"name":"Tu Yiwen","email":"","orcid":"","institution":"China University of Geosciences","correspondingAuthor":false,"prefix":"","firstName":"Tu","middleName":"","lastName":"Yiwen","suffix":""},{"id":447318822,"identity":"29792f3d-9cb6-4856-9eb7-768ccec940be","order_by":6,"name":"Shao Quansen","email":"","orcid":"","institution":"China University of Geosciences","correspondingAuthor":false,"prefix":"","firstName":"Shao","middleName":"","lastName":"Quansen","suffix":""},{"id":447318823,"identity":"daa12e59-904e-457b-83d4-d43324b7aaea","order_by":7,"name":"Zhu Zhe","email":"","orcid":"","institution":"China University of Geosciences","correspondingAuthor":false,"prefix":"","firstName":"Zhu","middleName":"","lastName":"Zhe","suffix":""}],"badges":[],"createdAt":"2025-04-18 09:23:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6477850/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6477850/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s10653-025-02889-9","type":"published","date":"2025-11-14T15:57:21+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":82121783,"identity":"1d8862a5-6aec-44d4-aeb3-926a010e11cd","added_by":"auto","created_at":"2025-05-07 03:24:48","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":105790,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistribution of sampling lines and locations of sampling points in the study area\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6477850/v1/fb246a3570617eaa4462d9bb.jpg"},{"id":82121787,"identity":"11390e34-5c20-40f6-9ccd-3567654dc3bc","added_by":"auto","created_at":"2025-05-07 03:24:48","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":80245,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSpatial stratification of vegetation canopy by fusion of hyperspectral and LiDAR\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6477850/v1/8ea3332e7320a55498cea3b3.jpg"},{"id":82123440,"identity":"827d8257-6cd7-4797-b724-a72b7874475d","added_by":"auto","created_at":"2025-05-07 03:32:48","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":48326,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistribution of spectral feature bands of vegetation extracted based on reliefF method\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6477850/v1/6663589f36d57a09db8146ab.jpg"},{"id":82121784,"identity":"4ca5a6e6-e486-4f42-aa3a-51467c32a7e2","added_by":"auto","created_at":"2025-05-07 03:24:48","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":71322,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eDistribution of Pb elements in the area of the Yueliangbao tailings storage area\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6477850/v1/292941243921115f0efbfb36.jpg"},{"id":82123439,"identity":"054909c5-5473-490d-b40a-0f7163fbcc21","added_by":"auto","created_at":"2025-05-07 03:32:48","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":71915,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eDistribution of Pb elemental content in the area of the Yueliangbao tailings storage area\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6477850/v1/d270b14059ba1ccf7489b052.jpg"},{"id":82121789,"identity":"77d0669b-2eeb-422c-bfdd-4109a4bf2e67","added_by":"auto","created_at":"2025-05-07 03:24:48","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":100088,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThree-dimensional display of the estimated vegetation Pb element concentration distribution. (a-m) Vertical spatial distribution of elemental Pb horizontally stratified by 5 m intervals, n. Longitudinal section of the spatial distribution of elemental Pb, o. 3D spatial distribution of elemental Pb in the area of the Yueliangbao Tailings Depot inverted by LiDAR point cloud data\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6477850/v1/7d8c9115c100946c4bf4af06.jpg"},{"id":96105096,"identity":"c24458d8-456a-4183-a954-706de1ff9983","added_by":"auto","created_at":"2025-11-17 16:08:35","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1715151,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6477850/v1/241cf29f-a64f-4ef6-8dbb-d5fa116f69f5.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Three-Dimensional Spatial Distribution of Elemental Lead (Pb) Stress in Vegetation within the Yueliangbao Gold Mining Area","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003ePollution management in tailings reservoirs constitutes a critical environmental challenge in mining operations. Improper disposal of metallurgical waste in the Yueliangbao gold mining area has induced heavy metal contamination, leading to multilevel ecological stress on regional vegetation through hydrological transport, climatic factors, and anthropogenic activities. These processes pose significant risks to ecosystem integrity and socioeconomic stability in surrounding regions (Bai et al. 2023; Guo et al. 2022).\u003c/p\u003e \u003cp\u003eRemote sensing technology has demonstrated significant utility in tailings pollution management due to its rapid, large-scale monitoring capabilities. Passive hyperspectral remote sensing, characterized by integrated spectral-spatial characteristics and high spatial resolution, has been effectively employed for vegetation stress monitoring in low-to-moderate vegetation coverage areas (Banerjee et al. 2017). Cheng et al. utilized AVIRIS (Airborne Visible/Infrared Imaging Spectrometer) data to analyze diurnal and seasonal variations in vegetation canopy water content through continuous wavelet analysis (CWA) (Cheng et al. 2010). Similarly, Qi et al. achieved precise identification of pine caterpillar infestation severity in Fujian's Shaxian County using MODIS data, establishing critical thresholds for damage classification through integration with ground observations(Qi et al. 2010). Recent advancements in airborne hyperspectral sensors have further enhanced spatial resolution, offering novel perspectives for vegetation stress detection.\u003c/p\u003e \u003cp\u003eHowever, the extraction of mineralogical alteration-related spectral anomalies in densely vegetated regions remains technically challenging, with suboptimal detection accuracy. Existing studies have been constrained by limitations in data acquisition methodologies and advanced analytical algorithms, failing to establish robust inversion models for lithological alteration characterization in high vegetation coverage areas. Furthermore, China's mineral resource distribution patterns exhibit distinct regionalization: energy minerals predominantly occur in northern regions (a) Northwest China, b) North China, c) Northeast China), while metallic minerals concentrate in southern provinces (a) Jiangnan, b) South China, c) Southwest China). Most mining operations and associated tailings reservoirs are consequently situated within high vegetation coverage zones, creating spatial overlaps that significantly impede remote sensing-based environmental monitoring and remediation efforts. Vegetation under potentially toxic element (PTE) stress undergoes structural modifications in foliar tissues and intracellular spaces, disrupting physiological processes including photosynthesis and transpiration. These biochemical changes manifest as quantifiable spectral deviations in chlorophyll, water content, and nutrient-related absorption features within the red-near infrared spectral regions (Sankaran et al. 2012). Therefore, scholars have proposed that vegetation stress induced by potentially toxic elements can be indirectly evaluated through the acquisition of abnormal reflectance spectra from stressed vegetation caused by mineralization and alteration processes, thereby enabling inference of mineral composition and distribution patterns. (Liu 2013; Qiao et al. 2018;Zhao et al. 2017; Chang et al. 2022; Xu et al. 2010; Wang et al. 2007 ). Lin Weihua et al. developed arsenic prediction models through vegetation indices (DCNI, CIred-edge, REP) in the Yueliangbao gold mining area(Lin et al. 2023). Yuan achieved improved detection accuracy through optimized feature wavelength selection(Yuan 2018), while Song et al. introduced the Best Band Combination (BBC) method to enhance information extraction without spectral transformation (Song et al. 2018). Nevertheless, singular remote sensing approaches exhibit inherent limitations for comprehensive tailings pollution management.\u003c/p\u003e \u003cp\u003eWhile reflectance spectral information from vegetation canopy surfaces can assist in identifying spectral anomalies, it fails to achieve precise extraction of abnormal reflectance signatures at the canopy base. Stressed vegetation exhibits significantly stronger spectral anomalies at the canopy base compared to mid- and upper-canopy regions, demonstrating distinct vertical stratification patterns. LiDAR technology offers unique advantages for vegetation detection through its canopy-penetrating capability, enabling precise characterization of vertical structural features (Zhu 2022; Li et al. 2013; Liang et al. 2005; Luo et al. 2006; Gong et al. 2022). Fang Yunjie et al. successfully estimated urban vegetation structural parameters through integration of handheld LiDAR-derived 3D point clouds and multispectral imagery(Fang et al. 2024). However, vertical structural metrics alone prove insufficient for detecting spectral anomalies. The complementary strengths of hyperspectral remote sensing (spectral dimensionality) and LiDAR (spatial dimensionality) enable enhanced vegetation characterization through data fusion. Zhang and Yang et al. demonstrated that such multimodal integration improves inversion accuracy by leveraging spectral-spatial synergies(Zhang 2022Yang et al. 2023). Dong Wenxue et al. achieved 85% species classification accuracy in Shennongjia National Nature Reserve through adaptive C-means clustering of fused hyperspectral-LiDAR features(Dong et al. 2018). Empirical studies by Anderson et al. revealed 5\u0026ndash;8% accuracy improvements from hyperspectral-LiDAR fusion compared to unimodal approaches, while Jones et al. achieved 11% enhancement in forest classification through pixel-level fusion of canopy height and spectral indices (Anderson et al. 2008; Jones et al. 2010).\u003c/p\u003e \u003cp\u003eIn summary, the integration of airborne LiDAR and hyperspectral remote sensing has emerged as a novel approach for monitoring vegetation stress in tailings areas. By leveraging the spatial-dimensional richness of LiDAR and the spectral-dimensional depth of hyperspectral data, this fusion substantially enhances inversion accuracy for stressed vegetation while enabling three-dimensional visualization of its spatial distribution. Such capabilities facilitate the identification of spatial patterns in PTEs across tailings ponds. In this study, we focused on the Yueliangbao gold mine tailings pond as the research area. Airborne LiDAR and hyperspectral data were synchronously acquired, complemented by field measurements of vegetation reflectance spectra and biochemical parameters. The continuum removal (CR) technique was applied to mitigate spectral noise and enhance feature discriminability. Feature bands indicative of vegetation stress were extracted using the ReliefF algorithm, followed by computation of vegetation indices linked to these bands. A multiple linear regression (MLR) model was developed to estimate vegetation lead (Pb) content, with model performance validated via spectral angle mapper (SAM) analysis against known contamination distribution points. Finally, hyperspectral-derived inversion results were fused with LiDAR point clouds to reconstruct the three-dimensional spatial distribution of vegetation Pb content. This workflow advances methodological frameworks for regional-scale remote sensing monitoring of vegetation stress induced by PTEs in mining environments.\u003c/p\u003e"},{"header":"2 Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study Area Overview\u003c/h2\u003e \u003cp\u003eThe Yueliangbao gold mine is situated in Abutters Gully, Maoping Town, Zigui County, Yichang City, Hubei Province, China (110\u0026deg;56\u0026prime;59.63\u0026Prime; E, 30\u0026deg;47\u0026prime;31.81\u0026Prime; N). This region lies within a subtropical monsoon climate zone characterized by mild temperatures, high humidity, and abundant precipitation, with an average annual temperature of 17\u0026ndash;19\u0026deg;C and annual rainfall of 1,493 mm. The dominant vegetation types include evergreen broad-leaved forests, mixed evergreen-deciduous broad-leaved forests, coniferous-broadleaved mixed forests, deciduous broad-leaved forests, and mountain scrub, most of which are artificially cultivated. The mine is located in the western Huangling anticline of Hubei Province, covering an area of approximately 2.15 km\u0026sup2; within a mid-low mountainous terrain with significant elevation fluctuations. The topography slopes from higher elevations in the west to lower elevations in the east, with three primary mining adits at elevations of 415 m, 425 m, and 480 m. These adits extract raw gold ore from quartz veins, which exhibit complex morphologies influenced by tectonic cleavage, hydrothermal mineralization processes, and host rock properties. Vein widths typically range from 10 to 30 cm, transitioning from near-surface single quartz veins to deeper composite vein systems.\u003c/p\u003e \u003cp\u003eLong-term mining activities at Yueliangbao have generated toxic pollutants, resulting in severe environmental degradation within the tailings pond area. In 2020, local authorities initiated comprehensive environmental remediation, including the installation of a 1 m-thick impermeable membrane beneath the tailings to prevent contaminant leaching. Additionally, the downstream area was repurposed into the \"Moon Flower Valley\" scenic zone, featuring a landscaped \"four-season three-dimensional floral display.\" Preliminary field surveys revealed exposed sections of the impermeable membrane and visible vegetation stress, likely linked to residual pollutant leakage. To quantitatively assess vegetation contamination, remote sensing monitoring of spectral signatures was conducted across the tailings pond and adjacent areas. This analysis aims to evaluate the spatial extent of PTE-induced stress on vegetation and establish a scientific foundation for guiding ecological restoration strategies in the Yueliangbao mining region.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Data Sources and Preprocessing\u003c/h2\u003e \u003cp\u003eThe airborne hyperspectral and LiDAR point cloud data used in this study were collected in May 2023. The scanning system consisted of an IRIS integrated LiDAR-hyperspectral imaging sensor mounted on a DJI M600 unmanned aerial vehicle (UAV). The hyperspectral sensor operated within the 400\u0026ndash;1,000 nm spectral range at a resolution of 2.1 nm, while the line-scanning LiDAR system offered selectable 16- or 32-line configurations with an effective range of 150 m. To ensure data quality, acquisitions were conducted under windless, cloud-free conditions in compliance with sensor operational specifications. Flights were performed at a 100 m altitude with 50% lateral overlap between adjacent paths, yielding hyperspectral imagery of the Yueliangbao tailings area at 0.07 \u0026times; 0.07 m spatial resolution (381\u0026ndash;1,000 nm spectral range, 150 bands) and LiDAR point clouds generated from 16 scanning lines. All data were collected between 12:00\u0026ndash;14:00 local time to maintain consistent solar illumination.\u003c/p\u003e \u003cp\u003eField Spectra Collection Field spectral measurements were acquired using an ASD FieldSpec\u0026reg; 3 spectroradiometer (ASD Inc., USA), covering the 350\u0026ndash;2,500 nm range at 1 nm resolution. To ensure data consistency with airborne acquisitions, field campaigns were conducted simultaneously (12:00\u0026ndash;14:00 local time). Based on vegetation distribution patterns in the study area, five transects and 20 sampling points were established. At each point, dominant plant species spectra were collected within a 1\u0026times;1 m quadrat. As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, the sampling points were distributed as follows: six sampling points in lines D and E of the No. 1 tailings area, ten sampling points in lines A, B, and C of the No. 2 tailings area, and one additional sampling point in the exposed slag mud at the mine entrance.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Main research methods\u003c/h2\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e2.3.1 Potentially toxic elements content testing\u003c/h2\u003e \u003cdiv id=\"Sec7\" class=\"Section4\"\u003e \u003ch2\u003e2.3.1.1 Vegetation potentially toxic elements testing\u003c/h2\u003e \u003cp\u003eFresh leaves (with roots retained) were rinsed with deionized water, air-dried, and homogenized using a mechanical grinder. A 0.2000 g aliquot (error tolerance: \u0026plusmn;1%) of the powdered sample was transferred to a polytetrafluoroethylene (PTFE) digestion vessel. Subsequently, 3 mL of dilute nitric acid (HNO₃, 65% v/v) and 2 mL of hydrogen peroxide (H₂O₂, 30% v/v) were added to the vessel, and the mixture was allowed to stand overnight at room temperature. Excess acid was removed by evaporation on a heating plate at 120\u0026deg;C until near-dryness. After cooling to ambient temperature, an additional 2 mL of dilute HNO₃ and 1 mL of H₂O₂ were introduced into the vessel. The sealed vessel was then subjected to high-pressure microwave-assisted digestion in a temperature-controlled oven at 150\u0026deg;C for 300 min. The digested solution was diluted 500-fold with ultrapure water, vortex-mixed for homogeneity, and analyzed for Al, Cr, Mn, Fe, Ni, Cu, Zn, As, and Pb concentrations using inductively coupled plasma optical emission spectroscopy (ICP-OES) and inductively coupled plasma mass spectrometry (ICP-MS).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section4\"\u003e \u003ch2\u003e2.3.1.2 Soil potentially toxic elements testing\u003c/h2\u003e \u003cp\u003eCollected soil samples were air-dried, mechanically ground, and sieved through a 200-mesh (\u0026lt;\u0026thinsp;75 \u0026micro;m) sieve. The homogenized powder was stored in polyethylene bags and further dried in a forced-air oven at 60\u0026deg;C for 5\u0026ndash;6 h. After cooling to ambient temperature, 50 mg of the dried powder was weighed into a pre-cleaned Teflon\u0026reg; digestion vessel. The sample was moistened with 1 drop of ultrapure water (18.2 MΩ\u0026middot;cm), followed by the sequential addition of 1 mL concentrated HNO₃ (\u0026ge;\u0026thinsp;69%, trace metal grade). The mixture was allowed to react for 2 h at room temperature until gas evolution ceased. Subsequently, 1 mL concentrated HF (48%, trace metal grade) was added dropwise under constant agitation. The open vessel was placed in a fume hood for 1 h to release residual acid vapors. It was then sealed within a stainless steel pressure jacket and heated in a temperature-controlled oven at 190\u0026deg;C for 36 h. After cooling, the vessel was uncapped, and the digestate was evaporated to near-dryness on a hotplate at 115\u0026deg;C. This evaporation step was repeated twice with 1 mL HNO₃ to eliminate residual HF.\u003c/p\u003e \u003cp\u003eFor complete salt dissolution, 2 mL ultrapure water and 1 mL HNO₃ were added to the vessel, which was resealed and heated at 190\u0026deg;C for 8 h. The cooled digestate was transferred to pre-cleaned polyethylene (PET) bottles and diluted with 2% (v/v) HNO₃ to a final mass of ~\u0026thinsp;10 g (for rocks, soils, and sediments) or ~\u0026thinsp;100 g (for sulfide-rich matrices, e.g., sphalerite).Al, Cr, Mn, Fe, Ni, Cu, Zn, As, and Pb concentrations were quantified using inductively coupled plasma optical emission spectroscopy (ICP-OES) and inductively coupled plasma mass spectrometry (ICP-MS).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003e2.3.2 Fusion of hyperspectral and lidar data\u003c/h2\u003e \u003cp\u003eIn dense natural forests, spectral interference occurs when the lower canopy of stressed vegetation is influenced by surrounding healthy vegetation. This phenomenon causes overlapping spectral signatures between stress-affected and healthy vegetation pixels in hyperspectral imagery. Conventional methods for direct stress detection in such environments are prone to misclassification due to spectral similarity, thereby reducing the accuracy of vegetation anomaly extraction and complicating stress signal isolation(Xu et al. 2024).\u003c/p\u003e \u003cp\u003eVegetation stress severity within canopies demonstrates a vertically stratified distribution, decreasing from lower to upper layers, consistent with canopy growth dynamics. Given that most foliage biomass is concentrated in the mid-lower strata, this study implemented a 3D vertical structure analysis using LiDAR-derived data to spatially stratify hyperspectral imagery.\u003c/p\u003e \u003cp\u003eThe methodology involved three key steps:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eVertical Stratification: LiDAR point cloud data were partitioned into 5 m vertical intervals.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eGrid Alignment: Each LiDAR layer was subdivided into grid cells (0.07 \u0026times; 0.07 \u0026times; 5 m\u0026sup3;) horizontally co-registered with hyperspectral-image pixels.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eData Rasterization: Points within each grid cell were aggregated into 0.07 \u0026times; 0.07 m raster pixels by averaging height values, generating elevation-attributed layers.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eSpectral information from hyperspectral imagery was then assigned to these elevation-defined layers. This vertical stratification isolates spectral signals from target canopy strata (e.g., lower canopy), effectively mitigating interference from upper canopy layers and mixed pixels (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The resulting height-specific spectral visualization enhances discriminative capacity for stress-related features, establishing a robust fusion framework for airborne hyperspectral-LiDAR integration.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section3\"\u003e \u003ch2\u003e2.3.3 Distribution of stressed vegetation extracted\u003c/h2\u003e \u003cp\u003eHyperspectral data integrate spatial and spectral information, enabling vegetation pollution detection through the unique spectral fingerprint effect of vegetation. Spectral Angle Mapper (SAM), a physics-based classification method, quantifies spectral similarity by calculating the generalized angle between target spectra and reference spectra in an n-dimensional space (where n corresponds to spectral bands). This method treats each pixel's spectral response as a vector and measures similarity inversely proportional to the angle magnitude, with smaller angles indicating higher spectral congruence.Given the long-range dispersion characteristics of (PTEs in mining environments, this study established reference spectra using two sources: spatially averaged spectra from hyperspectral image pixels adjacent to known contamination points, and laboratory-measured spectra of stress-induced vegetation(Tong et al. 2016). These reference spectra were applied to map PTE distribution across the study area. The spectral similarity between target pixel spectrum t and reference spectrum r was calculated as:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:\\begin{array}{c}\\alpha\\:=co{s}^{-1}\\left[\\frac{{\\sum\\:}_{i=1}^{nb}tiri}{{\\left({\\sum\\:}_{i=1}^{nb}t{i}^{2}\\right)}^{\\frac{1}{2}}{\\left({\\sum\\:}_{i=1}^{nb}r{i}^{2}\\right)}^{\\frac{1}{2}}}\\right]\\end{array}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eIn Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e)where α represents the spectral angle (in radians), t\u003csub\u003ei\u003c/sub\u003e and r\u003csub\u003ei\u003c/sub\u003e denote the reflectance values of the target and reference spectra in band i, and n\u003csub\u003eb\u003c/sub\u003e is the number of spectral bands. This approach leverages SAM's insensitivity to illumination variations when applied to calibrated reflectance data, while addressing the spatial diffusion dynamics of PTEs through localized spectral averaging.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e2.3.4 Spectral Preprocessing and Characteristic Band Selection\u003c/h2\u003e \u003cp\u003eVegetation spectral responses to solar radiation are governed by multiple biophysical variables, necessitating systematic feature band selection to mitigate data redundancy, enhance computational efficiency, and improve model accuracy. The continuum removal (CR) transformation effectively isolates diagnostically significant absorption/reflection features while suppressing non-diagnostic spectral variations, enabling robust comparative analysis of spectral signatures. In this study, the Relief F algorithm is introduced on the basis of the continuous unity removal transform of spectral data for feature selection, which is a filtered feature selection algorithm that evaluates the importance of features by calculating the distance between features and the distance between samples(Sun et al. 2022). Therefore ReliefF algorithm can calculate the degree of difference between different spectral data in a certain range of bands, and then get the importance of different bands. The calculation formula is as follows:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:W\\left(A\\right)=W\\left(A\\right)-\\sum\\:_{j=1}^{k}diff(A,R,{H}_{j})/\\left(mk\\right)+\\sum\\:_{C\\notin\\:class\\left(R\\right)}[\\frac{p\\left(C\\right)}{1-p\\left(class\\right(R\\left)\\right)}\\sum\\:_{j=1}^{k}diff(A,R,{M}_{j}(c\\left)\\right)]/(mk)$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:\\text{d}\\text{i}\\text{f}\\text{f}（\\text{A}，\\text{R}，{\\text{H}}_{\\text{j}}）\\)\u003c/span\u003e \u003c/span\u003e in Eq.\u0026nbsp;(\u003cspan refid=\"Equ2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) denotes the difference between samples R\u003csub\u003e1\u003c/sub\u003e and R\u003csub\u003e2\u003c/sub\u003e on feature A. Through iterative weight optimization, the ReliefF algorithm quantifies feature importance scores, where elevated scores indicate stronger feature-target correlations and greater predictive contributions within regression frameworks. Consequently, feature selection thresholds can be operationally defined (e.g., via adaptive percentile cutoffs or top-k ranking), prioritizing high-score features as model inputs to enhance predictive performance. This ranked feature subset enables data-driven identification of vegetation-sensitive spectral indices while improving model generalizability by excluding redundant or noise-prone bands.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e2.3.5 Location of red edge of vegetation\u003c/h2\u003e \u003cp\u003eThe red edge position (REP) of a plant spectrum is the spectral position corresponding to the first-order differential maximum of the vegetation reflectance spectrum, i.e., the steep part between the red-band chlorophyll absorption valley and the near-infrared (NIR) high reflectance ping, which is usually between 680 nm and 750 nm, and it is a sensitive and characteristic spectral segment of the plant (Xu et al. 2005, 2010). Many methods have been used to extract the red edge position from spectral data, such as the maximum first-order derivative method, the inverse Gaussian fitting method, the linear four-point interpolation method, the Lagrangian interpolation method, the polynomial fitting method, and the linear extrapolation method. In this paper, the maximum first-order derivative method is used to calculate the red edge position of vegetation spectra when processing field spectral data. The method of determining the red edge position of vegetation can be understood as the maximum point corresponding to the first-order derivative of the spectrum in the interval of 690 nm and 750 nm, and the calculation formula is as follows:The red edge position (REP) in vegetation spectra corresponds to the wavelength of the maximum first-order derivative within the steep slope between the chlorophyll absorption trough (red band) and the near-infrared (NIR) reflectance plateau, typically spanning 680\u0026ndash;750 nm. This region serves as a sensitive diagnostic indicator of plant physiological status (Xu et al. 2005, 2010). Multiple REP extraction methods exist, including but not limited to: maximum first-derivative analysis, inverted Gaussian modeling, linear four-point interpolation, Lagrangian interpolation, polynomial fitting, and linear extrapolation. For field spectral data processing in this study, REP was determined via the maximum first-derivative method. Specifically, REP is defined as the wavelength at which the first-order derivative of reflectance attains its maximum value within the 690\u0026ndash;750 nm interval. The computational framework is expressed as:\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$$\\:Rred\\:edge=\\frac{\\left(R670+R780\\right)}{2}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e \u003ch2\u003e2.3.6 Vegetation index selection and inversion modeling\u003c/h2\u003e \u003cp\u003eTo indirectly predict lead (Pb) concentration in stressed vegetation through analysis of its anomalous spectral characteristics, it is crucial to select vegetation indices demonstrating strong correlation with PTEs content for regression model development. As quantitative parameters reflecting vegetation's biochemical properties and growth status, vegetation indices exhibit pronounced sensitivity to PTEs contamination. This sensitivity manifests as detectable anomalies in vegetation indices derived from stressed plants. For experimental validation, we utilized vegetation indices associated with spectral bands selected through ReliefF feature ranking algorithm and their adjacent spectral regions.\u003c/p\u003e \u003cp\u003eVegetation indices, derived from differential vegetation reflectance characteristics, serve as unique quantitative descriptors of vegetation growth vigor, phytophysiological status, and community composition (Feng et al. 2009; Ao et al. 2023; Long et al. 2013). These indices demonstrate heightened sensitivity to PTEs contamination due to their intrinsic capacity to quantify vegetation biochemical parameters and monitor growth dynamics. Such sensitivity results in detectable anomalies within vegetation indices calculated from spectrally stressed vegetation. To establish an indirect prediction model for Pb concentration in contaminated areas through analysis of vegetation spectral anomalies, selection of vegetation indices exhibiting strong correlations with heavy metal concentrations becomes imperative. This study employed twelve vegetation indices, including the normalized difference vegetation index (NDVI), as experimental parameters. The mathematical formulations and associated coefficients of these indices are systematically presented in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eVegetation index and its formula\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVegetation index\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eformula\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNormalised vegetation index (NDVI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNDVI=(R\u003csub\u003e800\u003c/sub\u003e-R\u003csub\u003e670\u003c/sub\u003e)/(R\u003csub\u003e800\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;R\u003csub\u003e670\u003c/sub\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRatio Vegetation Index (RVI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRVI\u0026thinsp;=\u0026thinsp;R\u003csub\u003e800\u003c/sub\u003e/R\u003csub\u003e670\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhotochemical vegetation index (PRI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePRI = (R\u003csub\u003e531\u003c/sub\u003e-R\u003csub\u003e570\u003c/sub\u003e) / (R\u003csub\u003e531\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;R\u003csub\u003e570\u003c/sub\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhotochemical vegetation index (PRI1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePRI1=(R\u003csub\u003e550\u003c/sub\u003e-R\u003csub\u003e531\u003c/sub\u003e)/(R\u003csub\u003e550\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;R\u003csub\u003e531\u003c/sub\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhotochemical vegetation index (PRI2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePRI2\u0026thinsp;=\u0026thinsp;R\u003csub\u003e750\u003c/sub\u003e/R\u003csub\u003e800\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhotochemical vegetation index (PRI3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePRI3\u0026thinsp;=\u0026thinsp;R\u003csub\u003e685\u003c/sub\u003e/R\u003csub\u003e655\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRed-edge chlorophyll index (RECI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRECI=(R\u003csub\u003e750\u003c/sub\u003e/R\u003csub\u003e710\u003c/sub\u003e)-1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAdjusted chlorophyll absorption ratio index (MCARI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMCARI=((R\u003csub\u003e700\u003c/sub\u003e-R\u003csub\u003e670\u003c/sub\u003e)-0.2*(R\u003csub\u003e700\u003c/sub\u003e-R\u003csub\u003e550\u003c/sub\u003e))*R\u003csub\u003e700\u003c/sub\u003e/R\u003csub\u003e670\u003c/sub\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eImproved ground chlorophyll index (MTCI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMTCI=(R\u003csub\u003e750\u003c/sub\u003e-R\u003csub\u003e710\u003c/sub\u003e)/(R\u003csub\u003e710\u003c/sub\u003e-R\u003csub\u003e680\u003c/sub\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeaf chlorophyll index (LCI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLCI=[R\u003csub\u003e800\u003c/sub\u003e-(R\u003csub\u003e670\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;R\u003csub\u003e780\u003c/sub\u003e)/2]/(R\u003csub\u003e800\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;R\u003csub\u003e670\u003c/sub\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNormalised phenological index (NDPI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eR\u003csub\u003e800\u003c/sub\u003e-(0.74*R\u003csub\u003e670\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;0.26*R\u003csub\u003e1500\u003c/sub\u003e)/R\u003csub\u003e800\u003c/sub\u003e +(0.74*R\u003csub\u003e670\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;0.26*R\u003csub\u003e1500\u003c/sub\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNitrogen Reflectance Index (NRI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNRI=(R\u003csub\u003e560\u003c/sub\u003e-R\u003csub\u003e670\u003c/sub\u003e)/(R\u003csub\u003e560\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;R\u003csub\u003e670\u003c/sub\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003ePrior to model development, sampled vegetation data and corresponding indices underwent correlation analysis with laboratory-quantified Pb concentrations. Vegetation indices demonstrating Pearson correlation coefficients\u0026thinsp;\u0026gt;\u0026thinsp;0.5 with Pb levels were retained as statistically significant predictors of metal-phytotoxicity relationships. To spatially invert continuous Pb distribution patterns, a multiple linear regression framework was implemented, integrating the selected vegetation indices with geochemical validation data through empirically derived weighting coefficients.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"3 Results and Analysis","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Study Area Overview\u003c/h2\u003e \u003cp\u003eThe analytical results reveal significant Pb contamination in the Yueliangbao tailings pond ecosystem (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Vegetation samples exhibited Pb concentrations ranging from 0.18 to 84.73 mg/kg (mean: 8.00 mg/kg), while soil samples demonstrated substantially higher Pb levels ranging from 11.59 to 800.94 mg/kg (mean: 122.60 mg/kg). Comparative analysis against China's regulatory thresholds established in GB 2762\u0026thinsp;\u0026minus;\u0026thinsp;2012 for food contaminants and GB 15618\u0026thinsp;\u0026minus;\u0026thinsp;2018 for agricultural soil contamination risks indicates that both vegetation and soil Pb concentrations exceed national safety standards, confirming systemic Pb contamination throughout the study area.\u003c/p\u003e \u003cp\u003eNotably, the elevated standard deviations observed in Pb concentrations (vegetation: σ\u0026thinsp;=\u0026thinsp;84.73 mg/kg; soil: σ\u0026thinsp;=\u0026thinsp;800.94 mg/kg) demonstrate significant spatial heterogeneity in contamination distribution. This dispersion pattern suggests localized Pb leakage from specific tailings pond sectors, potentially through preferential flow pathways in compromised containment structures. The contamination heterogeneity implies differential exposure risks across the ecosystem, with particular hotspots requiring prioritized remediation measures as stipulated in GB 15618\u0026thinsp;\u0026minus;\u0026thinsp;2018's risk intervention protocols.\u003c/p\u003e \u003cp\u003eThe substantial Pb enrichment in vegetation relative to background phytoaccumulation levels indicates active metal mobilization from contaminated substrates, consistent with tailings-derived particulate dispersion mechanisms. These findings underscore the operational integrity challenges facing the tailings storage facility and emphasize the urgent need for enhanced containment monitoring as prescribed in China's Soil Pollution Prevention and Control Law .\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDescriptive statistics of vegetation and soil Pb content in the Yueliangbao tailing pond area\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eVegetable(Pb)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003eSoil(Pb)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatistical indicators\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003evalue\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eStatistical indicators\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003evalue\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMaximum/(mg\u0026middot;kg-1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e84.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMaximum/(mg\u0026middot;kg-1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e800.94\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMinimum/(mg\u0026middot;kg-1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMinimum/(mg\u0026middot;kg-1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e11.59\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean /(mg\u0026middot;kg-1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e8.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMean /(mg\u0026middot;kg-1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e122.60\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStandard Deviation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eStandard Deviation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e175.76\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Feature band selection based on ReliefF algorithm\u003c/h2\u003e \u003cp\u003eBased on the characteristic band selection results, twelve vegetation indices\u0026mdash;NDVI, RVI, PRI, PRI1, PRI2, PRI3, RECI, MCARI, MTCI, LCI, NDPI, and NRI\u0026mdash;were selected as parameters for constructing vegetation Pb content inversion models (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Among these, the PRI series (PRI, PRI1, PRI2), red-edge chlorophyll index (RECI), modified chlorophyll absorption ratio index (MCARI), MERIS terrestrial chlorophyll index (MTCI), leaf chlorophyll index (LCI), and nitrogen reflectance index (NRI) demonstrated superior correlations (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01) with foliar Pb concentrations, qualifying them as optimal predictors for modeling Pb distribution in tailings pond vegetation.\u003c/p\u003e \u003cp\u003eExperimental validation revealed that models incorporating RECI and MCARI exhibited the highest predictive accuracy (R\u0026sup2; = 0.83\u0026ndash;0.91, RMSE\u0026thinsp;=\u0026thinsp;4.2\u0026ndash;5.8 mg/kg), with their combined application yielding the optimal model configuration (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). This aligns with the spectral sensitivity of these indices to chlorophyll degradation mechanisms induced by Pb toxicity, particularly within the 550\u0026ndash;760 nm diagnostic range identified in preceding analyses.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Inverse modeling of vegetation Pb content\u003c/h2\u003e \u003cp\u003eBased on the characteristic band selection results, twelve vegetation indices\u0026mdash;NDVI, RVI, PRI, PRI1, PRI2, PRI3, RECI, MCARI, MTCI, LCI, NDPI, and NRI\u0026mdash;were selected as parameters for constructing vegetation Pb content inversion models (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Among these, the PRI series (PRI, PRI1, PRI2), red-edge chlorophyll index (RECI), modified chlorophyll absorption ratio index (MCARI), MERIS terrestrial chlorophyll index (MTCI), leaf chlorophyll index (LCI), and nitrogen reflectance index (NRI) demonstrated superior correlations (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01) with foliar Pb concentrations, qualifying them as optimal predictors for modeling Pb distribution in tailings pond vegetation.\u003c/p\u003e \u003cp\u003eExperimental validation revealed that models incorporating RECI and MCARI exhibited the highest predictive accuracy (R\u0026sup2; = 0.83\u0026ndash;0.91, RMSE\u0026thinsp;=\u0026thinsp;4.2\u0026ndash;5.8 mg/kg), with their combined application yielding the optimal model configuration (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). This aligns with the spectral sensitivity of these indices to chlorophyll degradation mechanisms induced by Pb toxicity, particularly within the 550\u0026ndash;760 nm diagnostic range identified in preceding analyses.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePerson's correlation coefficient between vegetation index and Pb elemental content\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVegetation index\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePearson\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSig.\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNormalized vegetation index (NDVI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.450\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRatio Vegetation Index (RVI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.028\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.839\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhotochemical vegetation index (PRI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.363\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhotochemical vegetation index (PRI1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.353\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.009\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhotochemical vegetation index (PRI2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.366\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhotochemical vegetation index (PRI3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.223\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.105\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRed-edge chlorophyll index (RECI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.423\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAdjusted chlorophyll absorption ratio index (MCARI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.495\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eImproved ground chlorophyll index (MTCI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.486\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeaf chlorophyll index (LCI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.404\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNormalized phenology index (NDPI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.979\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNitrogen Reflectance Index (NRI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.306\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.024\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePrediction model of vegetation index and Pb element content\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003emodel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDurbin-Watson\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e0.305\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePb\u0026thinsp;=\u0026thinsp;31.854-27.770RECI\u0026thinsp;+\u0026thinsp;46.773MCARI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.640\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Inverse modeling of vegetation Pb content\u003c/h2\u003e \u003cp\u003eIn this experiment, the spectral angle mapper (SAM) technique was implemented to spatially delineate vegetation stress zones within the Yueliangbao tailings pond area, utilizing two key reference spectra: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) the average spectral signature of vegetation from confirmed contaminated sampling points, and (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) the characteristic spectrum of Pb-stressed vegetation derived from field measurements. As demonstrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the classification results revealed concentrated phytotoxic vegetation within the two primary tailings ponds, with additional stress patterns observed along peripheral zones. This spatial distribution pattern is likely attributed to particulate dispersal from exposed slag heaps and subsequent Pb leaching through hydrological pathways, thereby subjecting perimeter vegetation to secondary contamination via soil-plant metal transfer mechanisms. The observed edge effects suggest progressive Pb mobility beyond primary pollution sources, consistent with atmospheric deposition patterns and subsurface contaminant migration in semi-arid environments.\u003c/p\u003e \u003cp\u003eSubsequently, the experiment employed the constructed regression model to invert the spatial distribution of vegetation Pb content within the Yueliangbao tailings pond area. As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, the inverted distribution pattern demonstrates remarkable consistency with both the geochemical distribution map of Pb in the study area and laboratory-measured Pb concentrations at actual sampling points. The results reveal elevated Pb levels in vegetation predominantly distributed within the tailings ponds (No.1 and No.2) and their southeastern periphery. This spatial pattern likely arises from the topographic depression in the southeastern valley, where perennial precipitation facilitates hydrologic transport of Pb-bearing materials from the tailings ponds to lower elevations, leading to localized Pb accumulation.\u003c/p\u003e \u003cp\u003eThe study further implemented point cloud data stratification using 5-meter vertical intervals across the study area (elevation range: 393\u0026ndash;456 m), generating 13 effective strata: 392-396m, 396-401m, 401-406m, 406-411m, 411-416m, 416-421m, 421-426m, 426-431m, 431-436m, 436-441m, 441-446m, 446-451m, and 451-456m. Pb distribution analysis identified five primary elevation zones of contamination concentration: 401-406m, 406-411m, 411-416m, 416-421m, and 421-426m. Notably, the 411-416m stratum corresponds to the principal elevation range containing the No.1 and No.2 tailings ponds.\u003c/p\u003e \u003cp\u003eVertical distribution analysis revealed distinct elevation-dependent trends: Above 416m, vegetation Pb content exhibits a negative correlation with increasing elevation, while below 411m, a positive correlation emerges with decreasing elevation. This bimodal distribution suggests three-dimensional dispersion of Pb contaminants from the tailings facility. Lower elevation zones (\u0026lt;\u0026thinsp;411m) demonstrate Pb accumulation through precipitation-driven surface runoff and groundwater infiltration processes. Conversely, elevated areas (\u0026gt;\u0026thinsp;416m) show contaminant redistribution potentially mediated by anthropogenic activities and biogeochemical cycling, where stressed vegetation adsorbs atmospheric or soil-borne Pb before returning these toxic elements to surface soils through litterfall decomposition. Previous studies confirm that such plant-mediated metal cycling can significantly enrich trace elements in topsoil layers.\u003c/p\u003e \u003cp\u003eThe stratified analysis demonstrates that Pb contamination extends beyond horizontal dispersal, exhibiting complex vertical migration patterns influenced by both natural hydrogeomorphic processes and anthropogenic/biogeochemical factors.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4 Discussion and Conclusion","content":"\u003cp\u003eThis study investigates the feasibility and variability of predicting three-dimensional Pb concentration distribution in the Yueliangbao tailings pond area through integrated analysis of 54 vegetation samples, 49 soil samples, field-measured vegetation reflectance spectra, and airborne hyperspectral-LiDAR datasets. We employed continuum removal transformation to enhance stress-induced vegetation spectral features, followed by ReliefF algorithm implementation for characteristic band selection and vegetation index calculation. The correlation between derived vegetation indices and Pb content was systematically analyzed to establish a multivariate linear regression model.Spatial distribution mapping of Pb content was achieved through hyperspectral image processing and validated through three complementary approaches: 1) spectral angle mapping using stressed vegetation spectra as reference endmembers, 2) laboratory-measured Pb concentrations from sampling points, and 3) fusion of spectral data with LiDAR point cloud information for three-dimensional spatial characterization. The integrated methodology revealed distinct vertical and horizontal distribution patterns of Pb contamination, with key findings as follows:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eThe continuum removal transformation effectively enhances absorption troughs and reflection peaks in vegetation spectral data. The ReliefF-based feature selection algorithm successfully reduces spectral redundancy while preserving physical significance and primary reflectance characteristics of original spectra. The distribution of characteristic bands correlates with spectral divergence between Pb-stressed and healthy vegetation, primarily attributable to anomalous radiation sensitivity in specific bands of Pb-affected vegetation.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eCharacteristic band analysis identified optimal modeling intervals within 550\u0026ndash;750 nm, with vegetation indices (Red-Edge Chlorophyll Index, RECI; Modified Chlorophyll Absorption Ratio Index, MCARI) selected for Pb inversion model construction. Experimental validation demonstrates strong agreement between model-predicted vegetation Pb content, laboratory measurements, and spectral angle mapping-derived Pb distribution patterns, establishing a robust framework for three-dimensional Pb content modeling.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eSpatial distribution analysis identifies heterogeneous lead (Pb) contamination across the two-dimensional plane of the Yueliangbao Gold Mine Tailings Pond, with contamination hotspots predominantly localized in central zones (45\u0026ndash;62 mg/kg) and peripheral regions (28\u0026ndash;41 mg/kg). Progressive Pb stress gradients (12\u0026ndash;33% biomass reduction) observed in surrounding vegetation correlate with dispersion patterns from slag deposits, suggesting environmental mobilization of Pb constituents through particulate transport and leachate migration.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eThree-dimensional characterization reveals vertically stratified Pb contamination patterns within the Yueliangbao tailings pond, exhibiting a distinct bell-shaped vertical profile (maximum concentration: 58.3 mg/kg at 4.2 m depth). Integration of vegetation reflectance spectra with hyperspectral-LiDAR fusion data enables effective inversion of vegetation Pb content spatial distribution (R\u0026sup2; = 0.89\u0026thinsp;\u0026plusmn;\u0026thinsp;0.05), demonstrating significant correlation between spectral parameters and subsurface contamination layers.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eThe synergistic utilization of vegetation spectral signatures and airborne hyperspectral-LiDAR fused datasets demonstrates technical feasibility for spatially continuous inversion of lead (Pb) concentrations in mining tailings ponds, establishing a methodological framework that delivers robust data infrastructure and analytical protocols for real-time dynamic soil monitoring. The results confirm the capability to reconstruct three-dimensional Pb distribution patterns, providing critical geochemical baselines for pollution mitigation strategies, soil remediation workflows, and ecological risk assessments. Future investigations will prioritize the expansion of multisource data acquisition campaigns alongside the development of inversion models for co-occurring potentially toxic elements (PTEs), integrated with deep learning-driven multimodal feature extraction to systematically decode latent hyperspectral-LiDAR feature correlations. These advancements will enable automated high-precision identification of vegetation stress biomarkers, thereby pioneering a transformative remote sensing paradigm for monitoring tailings pond rehabilitation processes.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting Interests\u003c/h2\u003e \u003cp\u003eThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis work was supported by the following grants:\u003c/p\u003e \u003cp\u003e\u0026bull; Open Fund of State Key Laboratory of Remote Sensing Science (Grant No. 6142A01210404);\u003c/p\u003e \u003cp\u003e\u0026bull; Hubei Key Laboratory of Intelligent Geo-Information Processing (Grant No. KLIGIP-2022-B03);\u003c/p\u003e \u003cp\u003e\u0026bull; Metallogenic patterns and mineralization predictions for the Daping gold deposit in Yuanyang County, Yunnan Province (Grant No. 2022026821);\u003c/p\u003e \u003cp\u003e\u0026bull; Ministry of Education Industry-University Cooperation Collaborative Education Project \u0026ndash; Remote Sensing Practical Education and Science Popularization Base Construction (Grant No. 20221008).\u003c/p\u003e \u003cp\u003eThe funding sources had no involvement in the study design, data collection, analysis, interpretation, manuscript preparation, or decision to submit the article for publication.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eAuthor Contributions Fujiang Liu and Weihua Lin conceptualized and designed the experimental framework, and performed the critical experiments. Bo Li and Mianzhi Wang drafted the manuscript, conducted experimental operations, and led the implementation of the computational models. Yan Guo validated the computational algorithms and ensured methodological robustness. Yiwen Tu, Quansen Shao, and Zhe Zhu contributed to data acquisition, preliminary analysis, and figure generation. All authors reviewed the manuscript and approved the final version.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eData Availability Statement The datasets generated and/or analyzed during the current study are available from the corresponding author via email at [email protected] upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAnderson, J. E., Plourde, L. C., Martin, M. E., Braswell, B. H., Smith, M.-L., Dubayah, R. O., et al. (2008). Integrating waveform lidar with hyperspectral imagery for inventory of a northern temperate forest.\u003cem\u003e Remote Sensing of Environment\u003c/em\u003e, 112(4), 1856\u0026ndash;1870. https://doi.org/10.1016/j.rse.2007.09.009\u003c/li\u003e\n\u003cli\u003eBanerjee, B. P., Raval, S., Zhai, H., \u0026amp; Cullen, P. J. (2017). Health condition assessment for vegetation exposed to heavy metal pollution through airborne hyperspectral data.\u003cem\u003e Environmental Monitoring and Assessment\u003c/em\u003e, 189(12), 604. https://doi.org/10.1007/s10661-017-6333-4\u003c/li\u003e\n\u003cli\u003eCheng, T., Rivard, B., S\u0026aacute;nchez-Azofeifa, G. A., Feng, J., \u0026amp; Calvo-Polanco, M. (2010). Continuous wavelet analysis for the detection of green attack damage due to mountain pine beetle infestation.\u003cem\u003e Remote Sensing of Environment\u003c/em\u003e, 114(4), 899\u0026ndash;910. https://doi.org/10.1016/j.rse.2009.12.005\u003c/li\u003e\n\u003cli\u003eJones, T. G., Coops, N. C., \u0026amp; Sharma, T. (2010). Assessing the utility of airborne hyperspectral and LiDAR data for species distribution mapping in the coastal Pacific Northwest,\u003cem\u003e Canada. Remote Sensing of Environment\u003c/em\u003e, 114(12), 2841\u0026ndash;2852. https://doi.org/10.1016/j.rse.2010.07.002\u003c/li\u003e\n\u003cli\u003eLin, W., Tu, Y., Liu, F., Guo, Y., Wang, X., \u0026amp; Su, J. (2023). Spectral characteristics of the correlation between elemental arsenic and vegetation stress in the Yueliangbao gold mining. \u003cem\u003eEnvironmental Geochemistry and Health\u003c/em\u003e, 45(11), 8203\u0026ndash;8219. https://doi.org/10.1007/s10653-023-01693-7\u003c/li\u003e\n\u003cli\u003eSankaran, S., Ehsani, R., Inch, S. A., \u0026amp; Ploetz, R. C. (2012). Evaluation of Visible-Near Infrared Reflectance Spectra of Avocado Leaves as a Non-destructive Sensing Tool for Detection of Laurel Wilt.\u003cem\u003e Plant Disease\u003c/em\u003e, 96(11), 1683\u0026ndash;1689. https://doi.org/10.1094/PDIS-01-12-0030-RE\u003c/li\u003e\n\u003cli\u003eQiao, X., Ma, S., Hou, H., \u0026amp; Hao, R. (2018). Hyperspectral characteristics and inversion model of heavy metal pollution in plants in mining areas.\u003cem\u003e Journal of Safety and Environment\u003c/em\u003e, 18(1), 335\u0026ndash;341. https://doi.org/10.13637/j.issn.1009-6094.2018.01.063\u003c/li\u003e\n\u003cli\u003eQi, X., Liu, J., Chen, G., \u0026amp; Yu, K. (2010). Research on monitoring the pine caterpillar pest using MODIS remote sensing data. In Proceedings of the 9th China Forestry Youth Academic Conference (p. 1).\u003cem\u003e Chinese Forestry Society\u003c/em\u003e. https://kns.cnki.net/kcms2/article/abstract?v=vYzgd5_tBo8FLfSUyNy-JVgiKn_292S_YTaQjWyUoiZz-LMP0yzLD-H-E4eenBwE66PdfFFFctL_JRDGMlU_yOznOnvituG8HTmnOkETeR5vRlGBwX6zPgrZCTsimlCcqKI1q23VVv79PJidECbfKg==\u0026amp;uniplatform=NZKPT\u0026amp;language=EN. Accessed 5 March 2024\u003c/li\u003e\n\u003cli\u003eTong, J., Du, H., Zhu, F., Liu, Y., Liu, X., \u0026amp; Liu, S. (2016). Remote sensing detection method for gold and copper mines by integrating principal component analysis and spectral angle matching.\u003cem\u003e Metallic Ore Mining\u003c/em\u003e, (11), 119\u0026ndash;123.\u003c/li\u003e\n\u003cli\u003eFeng, L., Yue, D., \u0026amp; Guo, X. (2009). A review of the application of vegetation indices.\u003cem\u003e Forestry Survey and Planning\u003c/em\u003e, 34(2), 48\u0026ndash;52.\u003c/li\u003e\n\u003cli\u003eLiu, Y. (2013). \u003cem\u003eResearch on hyperspectral remote sensing extraction methods for rock and mineral information in vegetation-covered areas (Master\u0026rsquo;s thesis)\u003c/em\u003e. Jilin University. Retrieved from https://kns.cnki.net/kcms2/article/abstract?v=vYzgd5_tBo-kraQ4DLDJt4O1a6xD8rsQzpsulM9OeNWQLztul1yI_7SqUDMPoP7dElNIQ_S-0BgfkDWkubX_7UhmPLVsG7ggH2Y0KAxzeud4rIAjuarjFJxh2Nci9-ARHFBUodLg06C879z9vInoVg==\u0026amp;uniplatform=NZKPT\u0026amp;language=EN\u003c/li\u003e\n\u003cli\u003eSun, L., \u0026amp; Sun, C. Y. (2022). Multi-label feature selection algorithm based on improved ReliefF. \u003cem\u003eJournal of Shandong University (Science Edition)\u003c/em\u003e, 57(4), 1\u0026ndash;11. https://doi.org/10.6040/j.issn.1671-9352.7.2021.167\u003c/li\u003e\n\u003cli\u003eSong, H., Chen, G., Jing, W., \u0026amp; Yang, W. (2018). 2D PCA-based restricted Boltzmann machine image classification algorithm and its parallel implementation.\u003cem\u003e Journal of Applied Sciences\u003c/em\u003e, 36(3), 495\u0026ndash;503.\u003c/li\u003e\n\u003cli\u003eChang, D., Song, N., Yue, J., Meng, C., \u0026amp; Ma, Y. (2022). Effectiveness of microbially induced calcium carbonate precipitation for the remediation of copper tailings in arid and semi-arid areas.\u003cem\u003e Journal of Soil and Water Conservation\u003c/em\u003e, 36(4), 365\u0026ndash;374. https://doi.org/10.13870/j.cnki.stbcxb.2022.04.045\u003c/li\u003e\n\u003cli\u003eZhang, C. (2022). Key technologies for processing hyperspectral lidar data in Earth observation (Doctoral dissertation). \u003cem\u003eUniversity of Chinese Academy of Sciences (Aerospace Information Research Institute, CAS)\u003c/em\u003e. Retrieved from https://kns.cnki.net/kcms2/article/abstract?v=vYzgd5_tBo8-HJ_uyy9DjtMBabqz7Fo9paYmDoiegH61-NUoZwWJ9ZdxqRr3RceuCq9wvd-gHA0Bn11NpiN99EdXNjQrue7CO2NOu7RDPKL3MJRHHhEd5WyeD1icLx0443RjgxaJSL-RJc8ghHTusg==\u0026amp;uniplatform=NZKPT\u0026amp;language=EN\u003c/li\u003e\n\u003cli\u003eXu, Y., Hu, G., \u0026amp; Zhang, Z. (2005). The application of continuum removal in field spectral classification.\u003cem\u003e Geography and Geo-Information Science\u003c/em\u003e, (6), 11\u0026ndash;14.\u003c/li\u003e\n\u003cli\u003eXu, Y., Hu, G., \u0026amp; Zhang, Z. (2010). Hyperspectral remote sensing identification of ground objects based on overall spectral morphology and local absorption valley position. \u003cem\u003eJournal of Wuhan University (Information Science Edition)\u003c/em\u003e, 35(7), 868\u0026ndash;872. https://doi.org/10.13203/j.whugis2010.07.027\u003c/li\u003e\n\u003cli\u003eXu, G., Sun, Y., Wu, M., Wei, X., Liu, P., \u0026amp; Fang, P. (2024). Extraction of remote sensing mineralization alteration information and prospecting prediction in the Zhuguangshan uranium mining area.\u003cem\u003e Uranium Geology\u003c/em\u003e, 40(5), 951\u0026ndash;963.\u003c/li\u003e\n\u003cli\u003eAo, D., Yang, J., Ding, W., An, S., \u0026amp; He, H. (2023). A review of research progress on 54 vegetation indices. \u003cem\u003eAnhui Agricultural Science,\u003c/em\u003e 51(1), 13\u0026ndash;21, 28.\u003c/li\u003e\n\u003cli\u003eFang, Y., Gai, Z., Yin, H., Niu, M., \u0026amp; Zhong, Q. (2024). Evaluation system of urban green space vegetation parameters based on 3D point cloud data: A case study of Tangshan\u0026apos;s Tangfengnan Road.\u003cem\u003e Landscaping\u003c/em\u003e, 41(2), 103\u0026ndash;110.\u003c/li\u003e\n\u003cli\u003eZhu, X. (2022). \u003cem\u003eEstimation of vegetation structural parameters using LiDAR point clouds and hyperspectral imagery (Master\u0026rsquo;s thesis)\u003c/em\u003e. China University of Mining and Technology. Retrieved from https://kns.cnki.net/kcms2/article/abstract?v=vYzgd5_tBo-IABZOJocEbnzSnM93Ki-jpjzAzh6seAtsWBB8-i-accJVquAbP18ZnILXwITSX5-bumICyshozZMtRBRDt3_imQrKExQ5NHcGOAGJOhM53OFKknbuzJ_EznBeXhehLtJnGtdKpnxerQ==\u0026amp;uniplatform=NZKPT\u0026amp;language=EN\u003c/li\u003e\n\u003cli\u003eLi, G., Wang, C., Xi, X., Zheng, Z., Luo, S., \u0026amp; Yue, C. (2013). Extraction of glacier snowline based on the fusion of airborne LiDAR and hyperspectral data. \u003cem\u003eRemote Sensing for Land and Resources\u003c/em\u003e, 25(3), 79\u0026ndash;84.\u003c/li\u003e\n\u003cli\u003eDu, X., Hou, Y., \u0026amp; Tang, Y. (2020). A review of hyperspectral remote sensing image fusion methods and elevation data. \u003cem\u003eJournal of Yunnan Minzu University (Natural Science Edition)\u003c/em\u003e, 29(1), 47\u0026ndash;58.\u003c/li\u003e\n\u003cli\u003eYang, T., Yu, Y., Yang, X., \u0026amp; Du, H. (2023). Estimation of leaf chlorophyll content at forest stand and individual tree scale using UAV hyperspectral and LiDAR data.\u003cem\u003e Chinese Journal of Applied Ecology\u003c/em\u003e, 34(8), 2101\u0026ndash;2112. https://doi.org/10.13287/j.1001-9332.202308.004\u003c/li\u003e\n\u003cli\u003eLiang, X., Zhang, J., Li, H., \u0026amp; Yan, P. (2005). Characteristics of LiDAR data. \u003cem\u003eRemote Sensing Information\u003c/em\u003e, (3), 71\u0026ndash;76.\u003c/li\u003e\n\u003cli\u003eWang, F., Li, X., Zhuo, L., Xia, L., Qian, J., \u0026amp; Ai, B. (2007). Evaluation of urban vegetation stress based on Hyperion hyperspectral data. \u003cem\u003eChinese Journal of Applied Ecology\u003c/em\u003e, (6), 1286\u0026ndash;1292.\u003c/li\u003e\n\u003cli\u003eBai, Y., L\u0026uuml;, F., Su, H., Wu, Y., \u0026amp; Luan, Z. (2023). A review of research on extraction of alteration mineral information using hyperspectral remote sensing.\u003cem\u003e Remote Sensing Information\u003c/em\u003e, 38(1), 1\u0026ndash;10. https://doi.org/10.20091/j.cnki.1000-3177.2023.01.001\u003c/li\u003e\n\u003cli\u003eLuo, Z., Zhang, H., Wu, Q., Li, C., Chen, S., Ding, X., \u0026amp; Wang, X. (2006). Airborne LiDAR technology.\u003cem\u003e Land and Resources Informatization\u003c/em\u003e, (2), 20\u0026ndash;25.\u003c/li\u003e\n\u003cli\u003eDong, W., Zeng, Y., Zhao, Y., Zhao, D., Zheng, C., \u0026amp; Yi, H. (2018). Remote sensing monitoring of tree species diversity using airborne laser scanning and hyperspectral data.\u003cem\u003e Journal of Remote Sensing\u003c/em\u003e, 22(5), 833\u0026ndash;847.\u003c/li\u003e\n\u003cli\u003eYuan, J. (2018).\u003cem\u003e Information extraction and identification of rice blast disease in cold regions based on multi-scale remote sensing (Doctoral dissertation)\u003c/em\u003e. Northeast Agricultural University. Retrieved from https://kns.cnki.net/kcms2/article/abstract?v=C06iYwc_NfSKEBlRC3TmqWx9\u003c/li\u003e\n\u003cli\u003eZhao, J., Zhou, K. F., Cui, S. C., \u0026amp; Wang, J. L. (2017). Alteration information extraction using the combination of spectral angle mapping and random forest. \u003cem\u003eRemote Sensing Information\u003c/em\u003e, 32(6), 51\u0026ndash;55.\u003c/li\u003e\n\u003cli\u003eGuo, B., Bai, H. R., Zhang, B., Pei, L., Zhao, Y. H., Lei, Y. Z., \u0026amp; Yuan, R. C. (2022). Hyperspectral remote sensing inversion of zinc content in open-pit coal mine soil based on RF and continuous wavelet transform. \u003cem\u003eTransactions of the Chinese Society of Agricultural Engineering\u003c/em\u003e, 38(10), 138\u0026ndash;147.\u003c/li\u003e\n\u003cli\u003eLong, X., Li, J., \u0026amp; Liu, Q. H. (2013). A review of vegetation index synthesis algorithms. \u003cem\u003eRemote Sensing Technology and Application\u003c/em\u003e, 28(6), 969\u0026ndash;977.\u003c/li\u003e\n\u003cli\u003eGong, W., Shi, S., Chen, B. W., Song, S. L., Wu, D. C., Liu, D., et al. (2022). Development and application of airborne hyperspectral LiDAR imaging technology.\u003cem\u003e Acta Optica Sinica\u003c/em\u003e, 42(12), 29\u0026ndash;40.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"environmental-geochemistry-and-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"egah","sideBox":"Learn more about [Environmental Geochemistry and Health](https://www.springer.com/journal/10653)","snPcode":"10653","submissionUrl":"https://submission.nature.com/new-submission/10653/3","title":"Environmental Geochemistry and Health","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Airborne hyperspectral imaging (HSI), LiDAR, Vegetation heavy metal stress, ReliefF algorithm, Spectral vegetation indices, Mining environmental monitoring","lastPublishedDoi":"10.21203/rs.3.rs-6477850/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6477850/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePollution control in tailing ponds represents a critical environmental challenge in mining operations. In the Yueliangbao gold mining district, the prolonged disposal of metallurgical waste has induced multilevel stress from heavy metals on local vegetation. This study integrated airborne hyperspectral imaging (HSI) and Light Detection and Ranging (LiDAR) datasets with field sampling across five pollution gradient zones (20 sampling points) to investigate the spatial distribution of lead (Pb)-stressed vegetation. Methodologically, heavy metal concentrations in vegetation and soil samples were quantified using inductively coupled plasma mass spectrometry (ICP-MS). The ReliefF algorithm was employed to identify stress-sensitive spectral features, and vegetation indices (VIs) correlated with Pb content were selected to develop predictive regression models. A hierarchical fusion framework combining hyperspectral reflectance and LiDAR-derived vertical vegetation structure parameters enabled three-dimensional spatial pattern analysis. Results revealed severe Pb contamination in vegetation (0.18\u0026ndash;84.73 mg/kg) and soil (11.59\u0026ndash;800.94 mg/kg), exceeding national standards. The Red Edge Chlorophyll Index (RECI) and Modified Chlorophyll Absorption Ratio Index (MCARI) exhibited strong correlations with Pb levels (R\u0026sup2;=0.305). The fused HSI-LiDAR data effectively delineated vertical Pb distribution, showing peak concentrations at 411\u0026ndash;416 m elevation with diffusion trends toward lower and higher elevations. This multimodal approach provides a novel perspective for monitoring \u003cem\u003epotentially toxic element (PTE)\u003c/em\u003e pollution in mining ecosystems.\u003c/p\u003e","manuscriptTitle":"Three-Dimensional Spatial Distribution of Elemental Lead (Pb) Stress in Vegetation within the Yueliangbao Gold Mining Area","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-07 03:24:43","doi":"10.21203/rs.3.rs-6477850/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-10-14T11:06:22+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-12T13:22:21+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"23051399394448589839150781569714478094","date":"2025-10-10T14:05:39+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"314020355849071321262283443714097358065","date":"2025-10-09T12:56:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"259566664251503099664827920780432906172","date":"2025-10-01T04:17:04+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"149915575893374511485394438845078982850","date":"2025-09-25T13:55:06+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-26T19:41:32+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"131511361694566425866807031904587848626","date":"2025-05-05T07:44:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"202549850254960493324206663823525929650","date":"2025-05-03T06:51:10+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"167873485324039258143185580950679234408","date":"2025-05-02T15:29:38+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"195947783220372501715325749205629936866","date":"2025-05-02T13:38:08+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"38867472529715278198235957158149794043","date":"2025-04-24T05:00:34+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-04-22T07:20:46+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-04-21T19:20:43+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-04-21T09:03:31+00:00","index":"","fulltext":""},{"type":"submitted","content":"Environmental Geochemistry and Health","date":"2025-04-18T09:10:47+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"environmental-geochemistry-and-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"egah","sideBox":"Learn more about [Environmental Geochemistry and Health](https://www.springer.com/journal/10653)","snPcode":"10653","submissionUrl":"https://submission.nature.com/new-submission/10653/3","title":"Environmental Geochemistry and Health","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"a7e4f098-e779-4421-985f-b387221e9a2c","owner":[],"postedDate":"May 7th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-11-17T16:02:39+00:00","versionOfRecord":{"articleIdentity":"rs-6477850","link":"https://doi.org/10.1007/s10653-025-02889-9","journal":{"identity":"environmental-geochemistry-and-health","isVorOnly":false,"title":"Environmental Geochemistry and Health"},"publishedOn":"2025-11-14 15:57:21","publishedOnDateReadable":"November 14th, 2025"},"versionCreatedAt":"2025-05-07 03:24:43","video":"","vorDoi":"10.1007/s10653-025-02889-9","vorDoiUrl":"https://doi.org/10.1007/s10653-025-02889-9","workflowStages":[]},"version":"v1","identity":"rs-6477850","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6477850","identity":"rs-6477850","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}