Heavy metal pollution was surveyed in 2019 and 2021 in cultivated soils from suburban areas, industrial surroundings, and main agricultural lands on the south bank of Taihu Lake. Overall, based on the Chinese national soil environmental quality standard grade II criteria, 21.71, 6.98, 0.78, and 0.78% of suburban and industrial surrounding soils exceeded the standards for mercury (Hg), cadmium (Cd), nickel (Ni), and copper (Cu), while those of main agricultural soils were 4.40, 4.98, 13.60, 5.41%, respectively. The comprehensive pollution indices of soils in the suburban and industrial surrounding area were 0.696 and 0.660, respectively. In contrast, the comprehensive pollution index was 0.648 in soils that were far from the point pollution sites (CK). For locally produced rice grains, the proportions of samples exceeding the Chinese National Food Safety and Health Criteria standards were 12.06% for Hg, 10.64% for Cd, and 17.73% for Zn. Overall, heavy metal pollution was more severe in soils within the suburban areas than the other areas. The survey in 2021 showed that the comprehensive pollution indices of soils in vegetable base, grain function area, modern agricultural park, long-term application of organic manure were 0.748, 0.685, 0.618 and 0.662, respectively. Therefore, suburban area was a priority for preventing and remediating heavy metal pollution, and vegetable bases were the key areas of heavy metal pollution in farmland.

ABSTRACT Mangrove ecosystems play a critical role in supporting coastal aquaculture by providing essential ecosystem services and reducing HM exposure to adjacent aquatic environments. Despite their importance, mangroves have declined sharply worldwide due to conversion for aquaculture and urban development, leading to substantial HM release and transport from sediments into surrounding waters. Quantitative assessment of mangroves’ capacity to regulate heavy metals remains limited, as monitoring trace-level HM concentrations over large spatial scales is technically challenging. This study presents a novel and automated ensemble learning framework for large-scale prediction of arsenic (As) and lead (Pb) concentrations in mangrove soils using multisource Earth observation data. Optical Sentinel-2 (MSI) combined with C-band SAR (Sentinel-1), and L-band SAR (ALOS-2 PALSAR-2) data were integrated with field measurements from 101 soil cores to derive complementary spectral, vegetation and soil indices, textural, and backscatter features. Multiple machine learning models were trained and systematically optimized using stacking and five-fold cross-validation within the AutoGluon framework. The weighted Level-2 ensemble consistently outperformed seven Level-1 base learners, achieving high predictive accuracy for both metals (R 2 > 0.75). The hybrid genetic algorithm-particle swarm optimization (GA-PSO) approach for optimal feature selection further improved performance, increasing R 2 to 0.816 (As) and 0.886 (Pb), while reducing RMSE to 4.266 and 6.293 mg kg−1, respectively. The proposed workflow demonstrates the added value of multisensor data fusion, feature selection, and automated ensemble learning for mapping trace-level soil contaminants in complex coastal environments. This is the first systematic, large-scale assessment of heavy metal accumulation in Vietnamese mangroves that integrates optical and multi-frequency SAR data with advanced ensemble modelling. The framework is computationally efficient, scalable, and transferable, offering a practical solution for regional to national-scale environmental monitoring and ecological risk assessment from space.

This study investigates the impact of environmental factors on land use change in Cross River Basin catchment area of Nigeria using geospatial techniques. The findings provide useful insights into land use planning, management regimes, and agricultural practices in the Cross River Basin as a critical component of the Nigerian ecosystem. The elevation distribution suggests characteristic basin morphology with water flowing from elevated peripheries toward central lowlands, which influences drainage patterns, microclimate variations, and land use suitability. Majority of the study area exhibits gentle slopes suitable for various land uses including agriculture and low-density development. Flow accumulation in the vast majority of the terrain exhibits minimal accumulation in the basin area. Total wetness values range from 2.04 to 27.83, with the majority of the terrain exhibiting values between 2.04 and 8.5, indicating low to moderate moisture retention capacity across most of the basin. Normalized Difference Vegetation Index presents values indicating continued vegetation dynamics over the decade, while Bare Soil Index reveals concerning trends, particularly in the northern and eastern sections where its values increased substantially from predominantly negative values in 2015 to widespread positive values by 2020-2025, suggesting sustained vegetation loss, soil disturbance, or land use conversion amongst other findings thus availing stakeholders with relevant data for key policy decisions.

From biowastes to risks? Impact of biosolids treatment and dose on antibiotic resistance in agricultural soils - A mesocosm study.

Phytophthora blight of Panax notoginseng caused by the fungal pathogen Phytophthora cactorum is a typical soilborne disease, that severely devastates the P. notoginseng industry, and soil bacterial diversity is closely related to the development of Phytophthora blight. A systematic investigation into the abundance of Phytophthora spp. as well as the structural and functional characteristics of the soil bacterial community was carried out using quantitative PCR, high-throughput sequencing of the bacterial 16S rRNA gene, and KEGG pathway analysis. Results showed Phytophthora spp. was most abundant in the rhizosphere soil of diseased plants (RSD) (9.3017 copies ng⁻¹ DNA) and least in the rhizosphere soil of healthy plants (RSH) (0.0169 copies ng⁻¹ DNA). For bacterial α-diversity, ACE and Chao1 indices of diseased soils (RSD and root-zone soil of diseased plants, ZSD) were significantly higher than healthy soils (RSH and root-zone soil of healthy plants, ZSH) and furrow soil (FS) (p < 0.05), with no difference in Shannon index (p > 0.05). β-diversity analyses revealed distinct clustering of the same soil type and inter-group compositional differences. Dominant taxa analysis showed higher Actinomycetota and Bacteroidota in healthy soils, while Verrucomicrobiota, Planctomycetota, and Ramlibacter were more abundant in diseased soils. KEGG functional analysis indicated that the abundances of the "Signal transduction" and "Membrane transport" pathways in the RSD group were higher than those in other groups, while the abundances of the "Metabolism of terpenoids and polyketides" and "Biosynthesis of other secondary metabolism" pathways in the RSD group were lower than those in healthy soils. This suggests that the lack of microorganisms synthesizing anti-oomycete secondary metabolites may contribute to the enrichment of pathogens in the RSD group. This study reveals the association mechanisms between the abundance of soil pathogens, bacterial community structure, and functions in P. notoginseng soils, providing a theoretical basis for the prevention and control of soil-borne diseases and the regulation of soil microecology in P. notoginseng cultivation.

Heavy metal (HM) accumulation is a significant environmental concern that endangers human health and the ecosystem due to increasing natural and anthropogenic activities. This study assessed the toxicity of fourteen different HMs in the soil samples taken from the agricultural sites situated along the national highways in Bilaspur, Chhattisgarh, India. Using a range of soil indices, such as contamination factor (CF), pollution load index (PLI), geo-accumulation index (Igeo), ecological risk assessment (ERA), and risk index (RI), the study examined the risk associated with the HMs such as boron, aluminum, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zinc, arsenic, molybdenum, cadmium, and lead. Their potential sources of origin were assessed using multivariate statistical techniques, coupled with positive matrix factorization (PMF) and self-organizing map (SOM). The CF and Igeo results showed that 4.80% of the soil sites were extremely polluted and 54.0% were moderately polluted. Comprehensive multivariate analysis combining PMF and SOM identified both geogenic and mixed anthropogenic sources, with vehicular pollution and agricultural activities emerging as the major contributors to health risk. This study advances the knowledge of HM contamination in agroecosystems and helps in developing the future strategies to reduce HM exposure in the environment.

This study develops an integrated framework to evaluate soil by combining climatic records, soil analyses, geomorphology, and remote sensing. Forty-three soil profiles were sampled (0-100 cm) and analysed for texture, calcium carbonate (CaCO3), organic matter, pH, electrical conductivity (EC), exchangeable sodium percentage (ESP), cation exchange capacity (CEC), gypsum, and macronutrients nitrogen (N), phosphorus (P), and potassium (K). Long-term meteorological data (1975-2021) were partitioned into three periods (1975-2001, 2002-2011, 2012-2021) to construct time-series of temperature, rainfall, and Normalized Difference Vegetation Index (NDVI), highlighting vegetation dynamics under hyper-arid conditions. Landsat imagery (2001, 2011, 2021) was processed using the Landsat Ecosystem Disturbance Adaptive Processing System (LEDAPS) and the Landsat Surface Reflectance Code (LaSRC), and spectral indices quantified vegetation vigor, soil brightness, moisture, and salinity. Composite indices soil indices were normalized (0-1) and weighted using the Analytic Hierarchy Process (AHP), then integrated into a Decision Support System (DSS) to generate spatial maps of degradation and sustainability. Results revealed severe salinity in sabkhas, nutrient deficits in sandy plains, and resilient loamy plains as sustainable hotspots. Declining vegetation in degraded units and stability in reclaimed lands were confirmed through NDVI time series analysis. More than 80% agreement with field observations was achieved by the DSS framework, providing reproducible, policy-relevant tools for prioritizing reclamation, conservation, and agricultural expansion.

Abstract The aim of this work is a multi‐scale, multi‐sensor geomorphological characterization of the Becca d'Aver deep‐seated gravitational slope deformation (DsGSD). Particular attention was given to the pseudo‐badlands morphologies that crop out in the area, which are producing sediments that are subsequently reactivated by debris flows. A geomorphological map at a scale of 1:10000 was generated in the Cretaz–Comba Basset basin area, left side of the Aosta Valley in northern Italy, ending at the Champagne fan. The study area is characterized by gravitational and runoff associated processes that interact with anthropic structures. However, pseudo‐badlands landforms represent the main source of sediments providing loose materials highly connected with the main drainage system. The landforms in the area were characterized and mapped using Google Earth and remote sensing interpretation as well as with field campaigns. Especially radar (Interferometric Synthetic Aperture Radar, InSAR) and optical (multispectral) data were employed to map active deformation areas and detect bare soil using the Bare Soil Index (BSI). Furthermore, we conducted a detailed terrain analysis (TA) and derived the Geological Strength Index (GSI). Moreover, we applied the connectivity index (IC) model and used remotely sensed data to assess the contribution of pseudo‐badlands to the general sediment transport. The results highlight how altered bedrock materials and anthropic deposits of mined serpentinite (waste deposits) contribute to the provision of sediments that are related to hydrogeological hazard in the area.

Transformation of sediment from mercury reservoir to potential mercury source in drought-stricken wetlands influenced by climate change.

Topsoil sand content (TSC) is a critical indicator of soil degradation in black soil regions, yet its long-term dynamics remain poorly quantified. To address this, we developed an automated Landsat-based framework on Google Earth Engine (GEE) for mapping cropland TSC across the Northeast China Black Soil Region (NCBSR) from 1984 to 2023. The methodology integrates a hierarchical bare-soil extraction strategy using the Normalized Difference Bare Soil Index (NDBSI), Normalized Difference Vegetation Index (NDVI), and Normalized Difference Tillage Index (NDTI) with a Random Forest (RF) model optimized by three-band spectral indices and a “prediction-first” compositing workflow. Results demonstrate that the bare-soil extraction achieved an overall accuracy of 96%, while the TSC retrieval model maintained robust performance with a coefficient of determination (R²) of 0.80 and a root mean square error (RMSE) of 9.68%, together with satisfactory temporal transferability. Long-term mapping revealed a significant biphasic evolutionary trajectory: 23.4% of croplands experienced soil coarsening predominantly before 2000, followed by a partial reversal and stabilization in later decades. This framework provides a high-resolution, multi-decadal baseline for monitoring soil physical degradation and supports sustainable agricultural management in global black soil regions.

The results of the study of the total amount and mobile forms of Cd, Pb, Zn, Cu, Ni, Co and Cr are presented. The studies were conducted in the Bryansk Region, Russia, in the landscapes of the Desna River floodplain. Soil samples were collected using the soil key method (92 soil pits). High variability of gross amount and mobile forms of cadmium, lead, zinc, copper, nickel, cobalt and chromium in alluvial soils of the Desna River is shown. The presence of a correlation between the gross content of trace elements in the soil and its granulometric composition is established. According to the concentration clarke value, the elements are grouped in a descending series: Cu &gt; Pb &gt; Zn &gt; Co &gt; Ni &gt; Cr &gt; Cd, similar to that in alluvial soils of other rivers in the region. A geochemical index of floodplain soils of the Desna River has been compiled, which allows one to judge the potential reserve of trace elements. The mobility series of trace elements Cd &gt; Ni &gt; Pb &gt; Zn &gt; Cu &gt; Co &gt; Cr has been established, similar to that in soils of a different genesis. A correlation has been established between the content of the mobile form of cadmium, copper, nickel and cobalt and the content of physical clay in the soil. The degree of zinc, copper and cobalt supply of floodplain soils of the Desna River is shown. The need for their additional amount when growing agricultural products is noted.

Glauconite is an iron- and potassium-rich mineral of the mica family. It evolves at the soil-water interface through chemical exchange, its maturity linked to exposure duration at the seafloor and source element availability. Glauconite sands have been discovered at offshore wind lease areas along the U.S. Atlantic Outer Continental Shelf, leading to uncertainties in pile foundation installation and long-term performance. This paper describes site characterization activities from the Piling in Glauconitic Sand (PIGS) Joint Industry Project test site, located along the New Jersey Coastal Plain. In situ testing, laboratory-based geological, microstructure, and soil index testing, and advanced soil behavior measurements are presented in detail. Comparisons to selected silica sand-based cone penetration testing (CPT) correlations are made, highlighting the cautions needed for deriving soil parameters in this unique material. The measured properties and observed behavior exhibit a transition from sand-like to clay-like behavior during particle crushing due to compression and shear stresses from impact pile driving, which deviate from conventional sedimentary clays and sands. This transition can lead to changes in soil microstructure, plasticity, strength, and permeability, among others. The extensive dataset provides a reference for geotechnical practioners and researchers encountering other glauconite sand deposits or similar transitional sediments.

This study evaluates the influence of routine soil index properties on the prediction of optimum moisture content (wopt) and maximum dry unit weight (γdmax), which are the primary outcomes of the Proctor compaction test, using machine learning (ML) methods. A curated database of fine-grained soils (n = 465, drawn from 15 sources) included gravel content (GC), sand content (SC), fines content (FC), liquid limit (LL), plastic limit (PL), plasticity index (PI), specific gravity (Gs), wopt, and γdmax. After correlation-based feature filtering, three models were developed: Generalized Additive Model (GAM), Random Forest (RF), and Extreme Gradient Boosting (XGBoost). The training used nested cross-validation with Bayesian optimization, corresponding to an overall 80–20 train-test split. The model performance was evaluated using R2, RMSE, MAE, MAPE, r, and the overfitting ratio calculated for the test set. For wopt, the best GAM model achieved R2 = 0.84 and RMSE = 2.16%, outperforming RF and XGBoost. For γdmax, the best GAM and XGBoost models reached R2 = 0.79 and RMSE = 0.76 kN/m3, respectively. SHapley Additive exPlanations (SHAP), model-based importance scores, and single ablation analyses consistently identified LL and PL as the most influential predictors, and FC provided secondary contributions, while GC and Gs added little once LL and PL had been included. Moreover, paired-feature ablation confirmed the joint influence of LL and PL on the prediction. Overall, all three models predicted compaction parameters with good accuracy; however, GAM models achieved comparable or better predictive metric values than the ensembles (RF and XGBoost) while offering interpretability through plots linking soil indices with the predicted outcomes. This balance of accuracy and interpretability supports GAM as the preferred model for prediction modeling.

Prediction of compressive index of clayey soils: A novel physics-guided neural network with consistency-gated fusion and ensemble uncertainty

In this study, soil analyses based on the Unified Soil Classification System (USCS), Atterberg limits, water content values, and groundwater level parameters were carried out for a 133.65 m² study area using a geographic information system (GIS). For this purpose, selected borehole data from a total of 210 boreholes located in the city center of Elazığ were used, and GIS-based model prediction maps were generated using the Inverse Distance Weighting (IDW) method. As a result, it was modeled that, within the study area, ML, GM, and CL soil classes are dominant at shallow depths (1.5–4.5 m), GM occurs together with MH at intermediate depths (7.5–9 m), and ML, SM, and GM types become dominant again at deeper levels (15 m). In addition, it was modeled that water content values of 11–20% are dominant across the area, while other ranges (0–10%, 21–30%, and 31–44%) occur at varying proportions at different depths. Overall, the distribution indicates predominantly low to medium water content, and the groundwater potential of the study area is not considered rich. According to the liquid limit, plastic limit, and plasticity index values, two distinct soil zones were identified, as demonstrated by the IDW-based model prediction maps.

Predictive models of the initial shear modulus of unsaturated sandy and silty soils from soil index properties

Monitoring urban expansion in arid regions is complicated by the spectral similarity between impervious surfaces and bare soil. Although machine learning classifiers on platforms like Google Earth Engine (GEE) offer effective solutions, their performance in these environments has not been systematically benchmarked. This study addresses this gap by comparing five supervised ML algorithms—Random Forest (RF), Support Vector Machine (SVM), Gradient Boosted Trees (GBT), Classification and Regression Tree (CART), and k-Nearest Neighbor (KNN)—for binary urban and non-urban mapping. We applied this analysis to Riyadh, Saudi Arabia, using a 35-year Landsat time series from 1990 to 2025. Annual, radiometrically consistent median composites were generated in GEE, with the 2025 composite based on imagery from January to September 2025. A custom ten-band feature stack, including indices such as the Bare Soil Index (BSI), was used for classification. The Random Forest model (RF-100) achieved the highest accuracy (Overall Accuracy = 0.977, Kappa = 0.954) and was selected for final land-use and land-cover mapping. Validation with 600 independent samples per epoch and comparison to the ESA WorldCover 2020 dataset confirmed the robustness of the results. The analysis found a 293% increase in Riyadh’s built-up area, from 416 km 2 in 1990 to 1,219 km 2 in 2025, with a notable slowdown in growth after 2010. Variable importance analysis showed that the Bare Soil Index (BSI) and the Normalized Difference Built-up Index (NDBI) were the most significant features for class separation, offering key methodological insights for arid urban remote sensing. This work provides a transferable methodological framework for classifier and feature selection in arid environments and a high-accuracy spatiotemporal dataset establishing a baseline for assessing sustainable urban development under Saudi Vision 2030.

Abstract Aims This study assessed the comparative impact of irrigation with sewage wastewater (SWW) and mixed sewage-industrial wastewater (SIWW) on soil fertility indicators and heavy metal accumulation in clayey soils and rice grains, using Nile water (NW) irrigation as a control. Materials and Methods Samples were collected from 21 sites in Dakahlia Governorate, Egypt, representing three irrigation sources: seven sites each irrigated by SWW, SIWW, and NW. Results Soils and rice grains from wastewater-irrigated sites showed significantly elevated levels of salinity, nutrients, organic matter, and bioavailable heavy metals compared to the Nile water control. Soil pH was the sole exception, showing no significant difference. SWW sources exhibited the highest levels of salinity, biochemical oxygen demand, nutrients, and heavy metals. This directly corresponded with elevated levels of these constituents in the soil and, subsequently, in rice grains, demonstrating a clear source-pathway-receptor linkage. Evaluation of soil fertility using the Applied System for Land Evaluation software classified into two classes: good-C2 under, and fair-C3 under. Heavy metal levels in soil, water, and plants were within safe limits. The contamination factor values followed the order: Ni ˃ Cd ˃ Cu ˃ Pb ˃ Zn. The pollution load index of soils under SWW was higher than that of soils under SIWW. Conclusions Wastewater is a valuable source of nutrients that significantly improves soil fertility. However, the associated accumulation of soluble salts and bioavailable cadmium, which entered the food chain, necessitates immediate, targeted monitoring and source control to ensure long-term agricultural sustainability.

ABSTRACT Mangrove ecosystems along Pakistan’s coastline provide critical ecological services, including coastal protection, carbon sequestration, and biodiversity support, but face increasing threats from sea-level rise, salinity intrusion, and reduced freshwater flows. This study integrates multi-index remote sensing (RS), machine learning (ML), and climate projection modeling to assess mangrove health, spatial dynamics, and future vulnerability from 2015 to 2024. Sentinel-2 imagery was analyzed using the Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Modified Soil-Adjusted Vegetation Index (MSAVI), Normalized Difference Water Index (NDWI), Modified Normalized Difference Water Index (MNDWI), Normalized Difference Moisture Index (NDMI), Moisture Vegetation Index (MVI), Salinity Index (SI), and Bare Soil Index (BSI). A Random Forest (RF) classifier implemented in Google Earth Engine (GEE) mapped mangrove extent and quantified temporal change. Results show a net loss of ~10 km² over the decade, despite 104 km² of gains and 114 km² of losses. Approximately 458 km² remained stable, while 166 km² exhibited transitional dynamics, indicating localized resilience. Climate projections from the Coupled Model Intercomparison Project Phase 6 (CMIP6) under Shared Socioeconomic Pathways (SSP2-4.5 and SSP5-8.5) suggest relative stability under moderate emissions but heightened vulnerability under high-emission scenarios. This integrated framework supports climate-adaptive mangrove conservation and long-term coastal resilience planning.

The accumulation of potentially toxic elements (PTEs) in agricultural soils poses significant risks to food safety and ecosystem health, necessitating rapid and cost-effective monitoring approaches. While inductively coupled plasma optical emission spectroscopy (ICP-OES) provides accurate PTE quantification, its high cost, time requirements, and chemical reagent necessitate the search for green, fast, robust, and cost-effective alternatives. This study aims to evaluate the suitability of mid-infrared Fourier transform infrared (MIR-FTIR) spectroscopy combined with partial least squares regression (PLSR) as an alternative rapid method for predicting PTE concentrations and calculating pollution indices in semi-arid agricultural soils of central Morocco. A total of 67 surface soil samples (0–20 cm) were collected from three distinct soil types: Lithic Calciustolls (n=23), Typic Haplusterts (n=23), and Typic Calciustolls (n=21). Ten PTEs (As, Ba, Cd, Cu, Mn, Pb, Se, Sr, Ti, and Zn) were measured by ICP-OES and predicted using MIR-FTIR (4000–400 cm −1 ) coupled with PLSR. Mean PTE concentrations varied substantially across soil types, with Cd ranging from 0.95 to 3.91 mg·kg −1 , Sr from 56.25 to 535.14 mg·kg −1 , and Zn from 38.23 to 59.63 mg·kg −1 . PTE Pollution Index (PI) was calculated using both datasets for comparative pollution assessment. Results demonstrated strong to excellent predictive performance (R² = 0.82-0.95) with the highest correlations for Ba, Zn, and Sr. PI calculations showed exceptional concordance between methods (mean PI: 1.54 for both), with all samples classified as low pollution. FTIR spectroscopy maintains the same geochemical relationships as ICP-OES (correlation differences &amp;lt;0.083), confirming method equivalence for soil pollution indexation. This approach offers significant advantages for large-scale monitoring programs while maintaining classification accuracy for environmental risk assessment.