Search Results (899)

The downstream irrigation district of the Yarkant River basin has experienced increasing soil salinization driven by shallow groundwater levels, constraining the sustainable development of regional agriculture. However, the dynamic relationship between soil salinity and groundwater depth in this region remains unclear, limiting the effectiveness of saline–alkali land remediation strategies based on groundwater level regulation. In this study, field data were collected in 2025 on total soil salinity, concentrations of eight major ions, groundwater depth, and groundwater salinity in the irrigation district. The spatiotemporal distribution patterns of soil salinity, groundwater depth, and groundwater salinity were analyzed, along with their interrelationships. The soils in the irrigation district are predominantly mildly to moderately saline. Overall, soil salinity exhibits clear seasonal patterns, characterized by accumulation due to evaporation in spring and autumn and dilution through irrigation in summer. The dominant anions in the soil were SO₄²⁻ and Cl⁻, while Ca²⁺ and Na⁺ were the dominant cations, indicating a chloride–sulfate salinity type. Soil salinity shows a significant positive correlation with groundwater mineralization. A clear Boltzmann function relationship was identified between soil salinity and groundwater depth, revealing a critical groundwater depth of 2.10–2.18 m for salt accumulation in the irrigation district. The critical groundwater depths corresponding to soil salinity and major salt ions, from lowest to highest, are Cl⁻ < Na⁺ < total salts < SO₄²⁻ < Ca²⁺. Random forest regression analysis identified the main factors influencing soil salinity and their relative importance, ranked from highest to lowest as follows: groundwater depth > Na⁺ > Cl⁻ > groundwater salinity > Ca²⁺ > SO₄²⁻ > Mg²⁺ > HCO₃⁻ > K⁺ > CO₃²⁻. Maintaining groundwater depth below the critical threshold and focusing on groundwater ions that strongly influence soil salinity can effectively alleviate soil salinization in the lower Yarkant River irrigation district caused by shallow, highly mineralized groundwater. Full article

Origin–destination (OD) flow prediction is fundamental to intelligent transportation systems, yet existing diffusion-based models face two critical limitations. First, they inadequately exploit spatial semantics, focusing primarily on temporal dependencies or topological correlations while neglecting urban functional heterogeneity encoded in Points of Interest (POIs). Second, static embedding fusion cannot dynamically capture semantic importance variations during denoising—particularly during traffic surges in POI-dense areas. To address these gaps, we propose the Cross-Attention Diffusion Model (CADM), a semantically conditioned framework for short-term OD flow forecasting. CADM integrates POI embeddings as spatial semantic priors and employs cross-attention to enable semantic-guided denoising, facilitating dynamic spatiotemporal feature fusion. This design adaptively reweights regional representations throughout reverse diffusion, enhancing the model’s capacity to capture complex mobility patterns. Experiments on real-world datasets demonstrate that CADM achieves balanced performance across multiple metrics. At the 30 min horizon, CADM attains the lowest RMSE of 5.77, outperforming iTransformer by 1.9%, while maintaining competitive performance at the 15 min horizon. Ablation studies confirm that removing POI features increases prediction errors by 15–20%, validating the critical role of semantic conditioning. These findings advance semantic-aware generative modeling for spatiotemporal prediction and provide practical insights for intelligent transportation systems, particularly for newly established transportation hubs or functional zone reconfigurations where semantic understanding is essential. Full article

The Hermies-Ruyaulcourt site (Pas-de-Calais), investigated within the “Canal Seine-Nord Europe” project, provides an exceptional record of pedosedimentary dynamics throughout the last interglacial-glacial cycle (Eemian–Weichselian). Eight stratigraphic trenches, correlated along 350 m, reveal several pedosedimentary units strongly influenced by local topography. This study combines sedimentological and micromorphological analyses with optically stimulated luminescence (OSL) dating. For OSL ages, a correction of the water content calculation protocol was developed, accounting for long-term moisture variations during burial. Nine OSL ages from humic horizons of the Early Glacial (MIS 5d-5a) and colluvial deposits of the Lower Pleniglacial (MIS 4) represent the first robust chronological dataset for these periods in northern France. Their internal consistency and agreement with existing thermoluminescence ages on burnt flints support their reliability. Moreover, geomorphological analysis highlights intense erosional phases which are interpreted as rapid permafrost destabilisation events linked to the melting of large ice-wedge networks around 60–55 ka and 30 ka (thermokarst erosion gullies). These investigations thus enable the chronology of the loess–palaeosols and the link with associated climatic events to be refined. This leads to a spatio-temporal model describing the evolution of Last Glacial environments in Western Europe, providing a robust reference for studying the Neanderthal occupation of the area. Full article

Aquaculture is the fastest-growing food production sector globally. As its largest producer, China plays a pivotal role in ensuring aquatic food supply and supporting the blue economy. Despite its massive scale, a systematic understanding of the geographic distribution, structural composition, and drivers of China’s aquaculture value chain remains limited. We comprehensively characterized the sector’s composition, spatiotemporal evolution, and structural dynamics. We compiled and analyzed over 2.85 million enterprise registration records from the TianYanCha database, applying rigorous industry classification, spatial mapping, correlation analysis, and bottleneck assessment with natural and socioeconomic variables. Results show that policy reforms, notably the 2013 Company Law amendment and 2016 aquaculture certification measures, drove sharp increases in enterprise registrations, particularly in retail and farming. Enterprises are highly clustered in the Yangtze River Basin, Pearl River Delta, and southeastern coast, with inland expansion along major river systems. Strong interdependencies exist among sectors, while wholesale remains numerically scarce, forming a structural bottleneck. Standardization levels are low. Foreign investment, though under 5%, concentrated in processing and distribution, contributed to advanced technologies in the 1990s–2000s. These findings highlight rapid formalization, regional clustering, and structural imbalances, suggesting that enhancing formalization and addressing intermediary bottlenecks could improve sector resilience and efficiency. Full article

Leaf area index (LAI) is a pivotal biophysical parameter linking vegetation physiological processes and macro-ecological functions. Accurate large-scale LAI estimation is indispensable for agricultural management, climate change research, and ecosystem modeling. However, existing methods fail to efficiently extract integrated spatial-spectral-temporal features and lack targeted modeling of spatio-temporal dependencies, compromising the accuracy of LAI products. To address this gap, we propose STC-DeepLAINet, a Transformer-GCN hybrid deep learning architecture integrating spatio-temporal correlations via the following three synergistic modules: (1) a 3D convolutional neural networks (CNNs)-based spectral-spatial embedding module capturing intrinsic correlations between multi-spectral bands and local spatial features; (2) a spatio-temporal correlation-aware module that models temporal dynamics (by “time periods”) and spatial heterogeneity (by “spatial slices”) simultaneously; (3) a spatio-temporal pattern memory attention module that retrieves historically similar spatio-temporal patterns via an attention-based mechanism to improve inversion accuracy. Experimental results demonstrate that STC-DeepLAINet outperforms eight state-of-the-art methods (including traditional machine learning and deep learning networks) in a 500 m resolution LAI inversion task over China. Validated against ground-based measurements, it achieves a coefficient of determination (R²) of 0.827 and a root mean square error (RMSE) of 0.718, outperforming the GLASS LAI product. Furthermore, STC-DeepLAINet effectively captures LAI variability across typical vegetation types (e.g., forests and croplands). This work establishes an operational solution for generating large-scale high-precision LAI products, which can provide reliable data support for agricultural yield estimation and ecosystem carbon cycle simulation, while offering a new methodological reference for spatio-temporal correlation modeling in remote sensing inversion. Full article

Financial fraud detection is critical to modern economic security, yet remains challenging due to collusive group behavior, temporal drift, and severe class imbalance. Most existing graph neural network (GNN) detectors rely on pairwise edges and correlation-driven learning, which limits their ability to represent high-order group interactions and makes them vulnerable to spurious environmental cues (e.g., hubs or temporal bursts) that correlate with labels but are not necessarily causal. We propose Causal-DHG, a dynamic hypergraph framework that integrates hypergraph modeling, causal intervention, and multi-view contrastive learning. First, we construct label-agnostic hyperedges from publicly available metadata to capture high-order group structures. Second, a Multi-Head Spatio-Temporal Hypergraph Attention encoder models group-wise dependencies and their temporal evolution. Third, a Causal Disentanglement Module decomposes representations into causal and environment-related factors using HSIC regularization, and a dictionary-based backdoor adjustment approximates the interventional prediction $P(Y\mid do(C\left)\right)$ to suppress spurious correlations. Finally, we employ self-supervised multi-view contrastive learning with mild hypergraph augmentations to leverage unlabeled data and stabilize training. Experiments on YelpChi, Amazon, and DGraph-Fin show consistent gains in AUC/F1 over strong baselines such as CARE-GNN and PC-GNN, together with improved robustness under feature and structural perturbations. Full article

Point-of-Interest (POI) recommendation predicts users’ future check-ins based on their historical trajectories and plays a key role in location-based services (LBS). Traditional approaches such as collaborative filtering and matrix factorization model user–POI interaction matrices fail to fully leverage spatio-temporal information and semantic attributes, leading to weak performance on sparse and long-tail POIs. Recently, Graph Neural Networks (GNNs) have been applied by constructing heterogeneous user–POI graphs to capture high-order relations. However, they still struggle to effectively integrate spatio-temporal and semantic information and enhance the discriminative power of learned representations. To overcome these issues, we propose Spatio-Temporal and Semantic Dual-Channel Contrastive Alignment for POI Recommendation (S²DCRec), a novel framework integrating spatio-temporal and semantic information. It employs hierarchical relational encoding to capture fine-grained behavioral patterns and high-level semantic dependencies. The model jointly captures user–POI interactions, temporal dynamics, and semantic correlations in a unified framework. Furthermore, our alignment strategy ensures micro-level collaborative and spatio-temporal consistency and macro-level semantic coherence, enabling fine-grained embedding fusion and interpretable contrastive learning. Experiments on real-world datasets, Foursquare NYC, and Yelp, show that S²DCRec outperforms all baselines, improving F1 scores by 4.04% and 3.01%, respectively. These results demonstrate the effectiveness of the dual-channel design in capturing both sequential and semantic dependencies for accurate POI recommendation. Full article

To achieve a dynamic balance between economic development and ecological protection, it is necessary to analyze the enabling mechanism of new quality productive (NQP) on the green and low-carbon transformation of the construction industry. Based on the panel data of 107 cities in the Yangtze River Economic Belt (YERB) from 2011 to 2023, this study analyzes the spatiotemporal characteristics of carbon emissions in the construction industry (CECI), and explores the impact, mechanism, regional heterogeneity and spatial spillover effect of NQP on carbon emission intensity in the construction industry (CEICI). From 2011 to 2023, CECI increased in low amplitude but weakened the spatial concentration. The overall level of NQP in the YREB region shows a trend of first a slight decline and then a steady increase, and the development disparities among different regions have continued to widen. The NQP is significantly negatively correlated with the CEICI. The impact effect shows a gradient distribution pattern of “upstream > downstream > midstream”, and there is also a spatial spillover effect. Moreover, the mechanism analysis shows that the NQP can influence CEICI by the green technological innovation (GTI) and industrial structure upgrading (ISU). Moreover, the mediating effect of GTI (−0.4019) is greater than ISU (−0.1049). These results can help to formulate policies on NQP for reducing building emissions. Full article

Under the ongoing trend of climate warming and increasing humidity on the Qinghai–Tibet Plateau, the Three River Source Region (TRSR) has shown strong sensitivity to global climate change. Its vegetation change is particularly worthy of attention and research. The Normalized Difference Vegetation Index (NDVI) is a key indicator for assessing the growth status of vegetation. However, the insufficiency of existing NDVI datasets in terms of spatiotemporal continuity has limited the accuracy of long-term vegetation change studies. This study proposed a machine learning-based downscaling framework that integrates the Moderate-resolution Imaging Spectroradiometer (MODIS) NDVI and the Global Inventory Monitoring and Modeling System (GIMMS) NDVI data to reconstruct a long-term, high-resolution NDVI dataset. Unlike conventional statistical fusion approaches, the proposed framework employs machine learning-based nonlinear relationships to generate long-term, high-resolution NDVI data. Three machine learning algorithms—Random Forest (RF), LightGBM, and CatBoost—were evaluated. Their performance was validated using the MODIS NDVI as reference, with the coefficient of determination (R²), root mean square error (RMSE), mean absolute error (MAE), and Pearson’s correlation coefficient (R) as evaluation metrics. Based on model comparison, the CatBoost model was identified as the optimal algorithm for spatiotemporal data fusion (R² = 0.9014, RMSE = 0.0674, MAE = 0.0445), significantly outperforming RF and LightGBM models and demonstrating stronger capability for NDVI spatiotemporal reconstruction. Using this model, a long-term, 1 km monthly GIMMS-MODIS NDVI dataset from 1982 to 2014 was successfully reconstructed. On the basis of this dataset, the spatiotemporal variation characteristics of vegetation in the TRSR from 1982 to 2014 were systematically analyzed. The research results show that: (1) The constructed long-series high-resolution NDVI dataset has a high consistency with MODIS NDVI data; (2) From 1982 to 2014, the NDVI in the TRSR showed an increasing trend, with an average growth rate of 0.0020/10a (p < 0.05). NDVI showed obvious spatial heterogeneity, characterized by a decreasing gradient from southeast to northwest. (3) The Yellow River source exhibited the most evident vegetation recovery, the Yangtze River Source area showed a moderate improvement, whereas the Lancang River Source area displayed little noticeable change. (4) Broad-leaved forests experienced the most significant growth, while cultivated vegetation displayed a marked tendency toward degradation. This study provides both a high-accuracy long-term NDVI product for the TRSR and a methodological foundation for advancing vegetation dynamics research in other high-altitude regions. Full article

Sabina vulgaris is a keystone shrub species endemic to arid northwestern China, renowned for its exceptional drought tolerance, sand fixation capabilities, and critical role in desert ecosystem stability. This study investigates the impact of coal mining activities on the spatiotemporal dynamics of S. vulgaris shrublands in the ecologically fragile Mu Us Sandy Land, focusing on the Longde Coal Mine adjacent to the Shenmu S. vulgaris Nature Reserve. Utilizing seven periods (2013–2025) of 2 m resolution Gaofen-1 (GF-1) satellite imagery spanning 12 years of mining operations, we implemented a deep learning approach combining UAV-derived hyperspectral ground truth data and the SegU-Net semantic segmentation model to map shrub distribution via GF-1 data with high precision. Classification accuracy was rigorously validated through confusion matrix analysis (incorporating the Kappa coefficient and overall accuracy metrics). Results reveal contrasting trends: while the S. vulgaris Protection Area exhibited substantial expansion (e.g., Southern Section coverage grew from 2.6 km² in 2013 to 7.88 km² in 2025), mining panels experienced significant degradation. Within Panel 202, coverage declined by 15.4% (58.4 km² to 49.5 km²), and Panel 203 showed a 18.5% decrease (3.16 km² to 2.57 km²) over the study period. These losses correlate spatially and temporally with mining-induced groundwater depletion and land subsidence, disrupting the shrub’s shallow-root water access strategy. The study demonstrates that coal mining drives fragmentation and coverage reduction in S. vulgaris communities through mechanisms including (1) direct vegetation destruction, (2) aquifer disruption impairing drought adaptation, and (3) habitat fragmentation. These findings underscore the necessity for targeted ecological restoration strategies integrating groundwater management and progressive reclamation in mining-affected arid regions. Full article

Reforestation efforts, notably the massive Grain for Green Project (GFGP), have significantly greened China’s Loess Plateau (LP) but intensified regional water limitations. This study aims to systematically characterize the spatio-temporal dynamics and the critical legacy effects of moisture stress on eWUE to evaluate ecosystem sustainability under accelerated climate change. Using 2001–2020 MODIS GPP and ET data and the comprehensive Temperature–Vegetation–Precipitation Drought Index (TVPDI), we analyzed the trends, spatial patterns, and lagged correlations on the LP. We find the LP’s mean eWUE was 1.302 g C kg⁻¹ H₂O, exhibiting a robust increasing trend of 0.001 g C kg⁻¹ H₂O a⁻¹ (p < 0.05), primarily driven by a faster increase in gross primary productivity (GPP) than evapotranspiration (ET). Spatially, areas with significant increases in eWUE concentrated in the afforested south and central LP. Concurrently, the region experienced a mild drought state (mean TVPDI: 0.557) with a concerning drying trend of 0.003 yeyr⁻¹, highlighting persistent water stress. Crucially, eWUE exhibited high and spatially divergent sensitivity to drought. A striking 69.64% of the region showed a positive correlation between eWUE and the TVPDI, suggesting that vegetation may adjust its physiological functions to adapt to drought. However, this correlation varied across vegetation types, with grasslands showing the highest positive correlation (0.415) while woody vegetation types largely showed a negative correlation. Most importantly, our analysis reveals a pronounced drought legacy effect: the correlation between eWUE and drought in the previous two years was stronger than in the current year, indicating multi-year cumulative moisture deficit rather than immediate climatic forcing (precipitation and temperature). These findings offer a critical scientific foundation for optimizing water resource management and developing resilient “right tree, right place” ecological restoration strategies on the LP, mitigating the ecological risks posed by prolonged drought legacy. Full article

Extreme climates pose increasing threats to ecosystems, particularly in ecologically fragile regions such as the Yellow River Basin (YRB). Leaf area index (LAI) reflects vegetation response to climatic stressors, yet spatiotemporal dynamics of such responses under future climate scenarios remain poorly understood. This study examined LAI responses to extreme climatic factors across the YRB from 2025 to 2065, utilizing Coupled Model Intercomparison Project Phase 6 (CMIP6) outputs under three Shared Socioeconomic Pathways (SSP) scenarios. Partial least squares regression was performed using historical consistency-validated and future scenario LAI data alongside 26 extreme climate indices to identify extreme climate impacts on vegetation dynamics. Time-lag and cumulative effect analyses using Pearson correlation further quantified the potential impacts of extreme climate on future vegetation dynamics. Results indicate that the regionally averaged LAI in the YRB exhibits a consistent increasing trend under all three SSP scenarios, with linear rates of 0.0016–0.0020 yr⁻¹ and the highest values under SSP5-8.5, accompanied by clear scenario-dependent spatial differences in LAI distribution and vegetation response to extreme climates, particularly in the lag and cumulative effects that depend on local hydro-climatic conditions. Partial least squares regression results identified annual total wet-day precipitation, frost days, growing season length, summer days, and ice days as the dominant extreme climate indices regulating LAI variability. In the arid and semiarid Loess Plateau regions, relatively long lag and cumulative effects imply vegetation vulnerability to delayed or prolonged climatic stress, necessitating enhanced soil and water conservation practices. These findings support region-specific ecological conservation and climate mitigation strategies for the YRB and other ecologically vulnerable watersheds. Full article

Accurate streamflow forecasting is crucial for water resources management and flood mitigation, yet it remains challenging due to the complex dynamics of hydrological systems. Conventional data-driven approaches often struggle to effectively capture spatio-temporal evolution characteristics, particularly the dynamic interdependencies among streamflow gauges. This study proposes a novel deep learning architecture, termed DynaSTG-Former. It employs a multi-channel dynamic graph constructor to adaptively integrate three spatial dependency patterns: physical topology, statistical correlation, and trend similarity. A dual-stream temporal predictor is designed to collaboratively model long-range dependencies and local transient features. In an empirical study within the Delaware River Basin, the model demonstrated exceptional performance in multi-step-ahead forecasting (12-, 36-, and 72 h). It achieved basin-scale Kling–Gupta Efficiency (KGE) values of 0.961, 0.956, and 0.855, significantly outperforming baseline models such as LSTM, GRU, and Transformer. Ablation studies confirmed the core contribution of the dynamic graph module, with the Pearson correlation graph playing a dominant role in error reduction. The results indicate that DynaSTG-Former effectively enhances the accuracy and stability of streamflow forecasts and demonstrates its strong robustness at the basin scale. It thus provides a reliable tool for precision water management. Full article

Resource-based regions play an indispensable role as strategic bases for national energy and raw material supply in the global industrialization and urbanization process. However, intensive and large-scale natural resource exploitation—particularly mineral extraction—often triggers dramatic land use/cover changes, leading to a series of problems including cultivated land degradation, ecological function deterioration, and human settlement environment degradation. However, a systematic understanding of the functional transitions within the land use system and their drivers in such regions remains limited. This study takes Shenmu City, a typical resource-based city in the ecologically vulnerable Loess Plateau, as a case study to systematically analyze the transition characteristics and driving mechanisms of land use functions from 2000 to 2020. By constructing an integrated “element–structure–function” analytical framework and employing a suite of methods, including land use transfer matrix, Spearman correlation analysis, and random forest with SHAP interpretation, we reveal the complex spatiotemporal evolution patterns of production–living–ecological functions and their interactions. The results demonstrate that Shenmu City has undergone rapid land use transformation, with the total transition area increasing from 27,394.11 ha during 2000–2010 to 43,890.21 ha during 2010–2020. Grassland served as the primary transition source, accounting for 66.5% of the total transition area, while artificial surfaces became the main transition destination, receiving 38.6% of the transferred area. The human footprint index (SHAP importance: 4.011) and precipitation (2.025) emerged as the dominant factors driving land use functional transitions. Functional interactions exhibited dynamic changes, with synergistic relationships predominating but showing signs of weakening in later periods. The findings provide scientific evidence and a transferable analytical framework for territorial space optimization and ecological restoration management not only in Shenmu but also in analogous resource-based regions facing similar development–environment conflicts. Full article

Spatiotemporally continuous monitoring of soil nitrogen is essential for rational farmland nitrogen management and non-point source pollution control. This study focused on a typical wheat-maize rotation system in the North China Plain under four nitrogen fertilizer application levels (N0: 0 kg/ha; N200: 200 kg/ha; N400: 400 kg/ha; N600: 600 kg/ha). By integrating soil profile sampling with UAV-based hyperspectral remote sensing, we identified soil nitrogen distribution characteristics and established a retrieval relationship between hyperspectral data and seasonal soil nitrogen dynamics. Results showed that higher nitrogen fertilizer levels significantly increased soil nitrogen content, with N400 and N600 causing nitrate nitrogen (NO₃⁻-N) peaks in both surface and deep layers indicating leaching risk. Hyperspectral imagery at the jointing stage, combined with PLSR and XGBoost-SHAP models, effectively retrieved NO₃⁻-N at 0–50 cm depths. Canopy spectral traits correlated with nitrogen leaching and deep accumulation, suggesting they can serve as early indicators of leaching risk. The “sky-ground” collaborative approach provides conceptual and technical support for precise nitrogen management and pollution control. Full article