Search Results (3,153)

Multi-view clustering (MvC) aims to discover cluster structures by exploiting complementary information across views. Most existing MvC methods assume that same-index observations across views describe the same semantic instance. In practice, however, index-aligned observations can be semantically unrelated. This inconsistency between observed index-level correspondence and underlying semantic correspondence is known as noisy correspondence (NC). Learning from such mismatched pairs imposes erroneous cross-view constraints and distorts clustering. Many existing methods only suppress unreliable pairs. This discriminative strategy avoids incorrect alignment but also excludes suspicious pairs from cross-view learning. To reuse these pairs without enforcing incorrect correspondence, we propose Reliability-Aware Gaussian Residual Counterpart Generation. Using reliability estimates derived from cross-view losses, the framework retains observed counterparts for reliable pairs and routes unreliable pairs to counterpart generation. For each unreliable pair, prototype-level semantic transport locates a matched target-view prototype. A Gaussian residual model estimated from reliable target-view samples captures variations around this prototype. The framework samples a residual from this model and adds it to the prototype center, yielding a semantically matched yet diverse counterpart. Random walk-based intra-view contrastive learning further preserves neighborhood structures. Experiments on Scene15, LandUse21, Reuters, and CCV20 achieve the best average ACC, NMI, and ARI across the evaluated NC ratios. Ablation and transfer studies further support the effectiveness of the proposed design. Full article

To address the limited synergy between integrated water–fertilizer technologies and intelligent application equipment, as well as the low water–nitrogen use efficiency in rapeseed production systems in southwestern China—which together constrain the large-scale adoption of smart fertigation equipment—this study conducted a two-season field experiment using an integrated water–fertilizer application system. The experiment included two irrigation regimes (W1: 60% ET_c; W2: 100% ET_c) and three nitrogen rates (F1: 220, F2: 300, F3: 360 kg N ha⁻¹), plus a rainfed control (CK), to quantify rapeseed responses to water–nitrogen interactions under an equipment-based fertigation framework. Results showed that water–nitrogen interactions significantly regulated rapeseed physiological processes, growth, yield formation, and resource-use efficiency. Compared with CK, appropriate water and nitrogen supply markedly enhanced PSII efficiency and overall energy conversion. In particular, W2F2 increased leaf photosynthetic rate by 51.2% and 50.8% across the two seasons. Water–nitrogen coupling also improved yield components such as branch number and thousand-seed weight, thereby increasing final yield. Although nitrogen partial factor productivity declined with increasing N rates, smart fertigation improved economic returns to varying degrees. EWM-TOPSIS results indicated that W2F2 consistently achieved the highest comprehensive performance across both seasons. By balancing yield, seed quality, resource efficiency, and economic benefit, this treatment represents the optimal strategy under intelligent fertigation systems. Overall, this study provides a decision-oriented optimization framework based on crop physiological responses under smart water–fertilizer equipment, offering practical guidance for intelligent fertigation deployment in rapeseed systems in southwestern China. Full article

The variation features of metallic element grades can reflect the enrichment level of mineral deposits. Thus, quantitative characterization of the spatial patterns of metallogenic elements is fundamental to studying ore-forming processes and implementing mineral exploration. This study applies fractal theory and self-developed MATLAB computational scripts to process Cu grade datasets from 28 drillholes within the Pulang porphyry copper deposit, Yunnan Province. Both rescaled range (R/S) analysis and correlation integral methods were applied to clarify the spatial patterns of Cu grades in drill-cores. The calculated Hurst exponents ranged from 0.510 to 0.636, which demonstrated the persistent variation of Cu grades along the vertical direction of drillholes. This work further explored the correlation between Cu mineralization and fluctuations in correlation dimension (D_C), with D_C values spanning 0.011–2.873. Results indicate steep fractal gradient zones host high-grade copper ore bodies, and fractal dimension is a robust indicator to trace the migration of hydrothermal fluids. The Hurst exponents of Cu grade sequences correlate strongly with mineralization intensity, and ore-bearing veins extend continuously throughout all sampled drillholes. Accordingly, fractal gradients can be utilized to depict prospective zones for favorable mineralization in uncharted regions. This methodology may be applicable to other structurally controlled mineral deposits where similar fracture-controlled mineralization occurs, though further testing on different deposit types is needed. Full article

The determination of the angle of arrival (AoA) of Ku-band microwave signals is of great significance in satellite communications, spectrum monitoring, and national defense security, which has created an urgent demand for high-precision and interference-resistant passive detection techniques. In this study, we propose a Rydberg-atom-based passive measurement method for estimating the AoA of Ku-band incident signals. Based on well-established quantum optical techniques, the proposed method converts the EIT-AT splitting intervals induced by waves or signals incident from different directions in Rydberg atoms within a single vapor cell into the electric-field intensity at the atomic sensor. An electric-field-intensity–angle response model is then established for AoA estimation of a 13.6 GHz wave or signal. Simulation results demonstrate that the proposed method can determine the incident direction within a certain angular range, with an estimation error of less than 3.8°. Without the need for self-calibration and free from interference caused by actively emitted probing electromagnetic waves or reference electric fields, the proposed approach enables AoA determination of incident waves or signals using a single vapor cell with a length of 5 cm. Full article

This study quantifies same-grid-cell, same-date co-occurrence of low precipitation and low operational wind-power density (WPD) in Southwest China during 1979–2018. Daily precipitation and WRF-derived 10 m wind fields were analyzed on a 271 × 271 grid with 10-, 15-, 20-, and 30-day rolling windows. At each grid cell, rolling precipitation totals and rolling means of the daily operational-WPD index were ranked against a fixed seasonal distribution. Low conditions were defined by $u\le 0.1587$, with sensitivity analyses at $u\le 0.10$ and $u\le 0.20$. Low precipitation represented a sub-monthly meteorological-drought condition. Domain compound frequency declined from 2.207% at 10 days to 1.704% at 30 days. Warm-season frequency was highest at every window, ranging from 2.624% to 2.028%. R3 had the highest compound frequency at the central threshold, whereas R5 had the lowest. Under the 1% regional active-area rule, median compound runs lasted three to five days. Spatial-block intervals changed with block size, so the 40-year record does not support a robust domain-wide trend conclusion. The results provide a 1979–2018 climatology of atmospheric precipitation–WPD co-occurrence. Full article

Balancing food production, ecosystem service value, and agricultural irrigation water consumption is a major challenge for sustainable agricultural development in China. However, quantitative evidence at the national scale remains limited on whether crop planting structure optimization derived from models can effectively achieve coordination among these three objectives. Existing studies mainly focus on individual crops or localized regions and rarely integrate the spatiotemporal evolution, influencing factors, and multi-objective optimization of major staple crops within a unified framework. This study developed a progressive framework integrating spatiotemporal evolution analysis, influencing factor identification, and planting structure optimization for wheat, rice, and maize. Spatial autocorrelation analysis, center-of-gravity shift analysis, and pixel-based image differencing were applied to reveal crop evolution patterns across China from 2000 to 2025. A five-dimensional indicator system comprising 17 quantitative indicators was developed through multiple experiments using four large language models, and the Random Forest algorithm was employed to identify key influencing factors and their relative importance. Based on these factors, optimization constraints were constructed, and a multi-objective fuzzy linear programming model combined with the NSGA-II algorithm was used to determine optimal crop area allocation across 28 provincial-level regions. The three staple crops exhibited a significant pattern of northward shift, eastward expansion, and southern contraction. Precipitation and market accessibility were common core influencing factors, ranking among the top five factors in all nine Random Forest models. Crop-specific factors, including soil available phosphorus for rice, accumulated active temperature for wheat, and soil pH for maize, explained differences in spatial responses among crops and provided a scientific basis for optimization modeling and coordinated improvement of food production, ecosystem service value, and irrigation water consumption. The optimized scheme increased total grain output by 3.6%, improved ecosystem service value by 11.0%, and reduced irrigation water consumption by 45.7% compared with the actual planting structure, all 28 provinces achieved improvement or stability in the three indicators simultaneously. Based on optimized crop allocation patterns, national planting structures were summarized into regional models, including a rice–maize dual-core system in Northeast China, wheat–maize rotation in the Huang-Huai-Hai Plain, rice-dominated systems in the middle and lower Yangtze River Basin and South China, water-efficient dryland farming in Northwest China, and a diversified balanced system in Southwest China. These findings provide a quantitative reference for optimizing China’s staple crop planting structure. Full article

Improving fruit quality is a core challenge in apple production in Southwest China, where long-term clean tillage has driven red soil acidification and fertility decline. However, how different cover crop species influence fruit quality through soil fungal pathways in this region remains unclear. Here, we compared five cover crop treatments—natural grass, alfalfa, perennial ryegrass, white clover, and orchardgrass—against clean tillage in a field experiment, integrating ITS sequencing with multivariate statistical analyses to evaluate their effects on soil properties, fungal communities, and apple fruit quality. Orchardgrass (DG) showed the most favorable performance, increasing sugar–acid ratio (+28.74%) and vitamin C content (+42.16%), followed by alfalfa and natural grass, while ryegrass had the weakest effects. Soil fertility improvement followed the same ranking, and soil pH, organic matter, and nitrogen were identified as key factors associated with fungal community shifts. Fungal community composition and diversity were closely linked to soil properties and fruit quality traits, with sugar–acid ratio co-regulated by nitrogen, pH, and fungal diversity, while vitamin C content was directly associated with cover crop treatment. These findings suggest that orchardgrass may be a promising cover crop for enhancing fruit flavor and nutritional quality under the subtropical red-soil conditions tested in this study, offering practical implications for cover crop selection, though the long-term stability of these effects requires further validation through multi-year trials. Full article

GDSL esterases/lipases constitute a large and functionally versatile gene family in plants, yet their systematic characterization in bamboo—perennial woody grasses with exceptionally rapid shoot elongation—remains scarce. Here, we performed a genome-wide identification of the GDSL family in allohexaploid Dendrocalamus farinosus and characterized the function of a candidate gene in secondary wall synthesis. A total of 265 DfGDSL genes were identified and classified into nine clades; rice xylan deacetylases BS1 and DARX1 fell within Clade VIII, which harbors 75 members. Chromosomal distribution, exon–intron organization, conserved motifs, and collinearity analyses indicated that whole-genome and tandem duplications drove family expansion, with purifying selection as the predominant evolutionary force. Expression profiling across tissues and shoot developmental stages pinpointed DfGDSL58 as the sole highly expressed OsDARX1 homolog in D. farinosus, showing specific upregulation during rapid elongation (50–400 cm). Its promoter contains auxin-responsive elements, and exogenous NAA treatment significantly induced its expression, with a peak at 6 h. Heterologous overexpression of DfGDSL58 (designated DfDACX1) in tobacco increased plant height and basal diameter while reducing lignin and hemicellulose deposition, xylem width, and transcript levels of xylan synthase (NtIRX9) and lignin biosynthetic genes (NtC4H, NtCOMT, NtCAD), whereas cellulose content and cellulose synthase genes (NtCESA4, NtCESA7) remained largely unchanged. Collectively, these findings demonstrate that DfDACX1 negatively regulates secondary wall thickening and promotes longitudinal growth, likely via modulating xylan deacetylation and downstream lignin biosynthesis. This study presents the first systematic characterization of the GDSL family in D. farinosus and identifies DfDACX1 as a key modulator of the trade-off between cell wall deposition and rapid shoot elongation, offering mechanistic insights into bamboo’s extraordinary growth and a potential target for engineering plant architecture. Full article

In this study, bacterial α-diversity, represented by Chao1 richness and Shannon diversity, was quantified and mapped across the acidic red and yellow soils of southern China by integrating a comprehensive meta-database of 920 georeferenced observations with multi-source environmental data and machine-learning models. Chao1 and Shannon were modeled using Random Forest, eXtreme Gradient Boosting, and Support Vector Machine algorithms. Among these, Random Forest consistently achieved the highest predictive accuracy and robustness (R² > 0.83). SHapley Additive exPlanations (SHAP) analysis revealed that soil pH acted as the fundamental filter for both diversity dimensions, while they were further shaped by distinct environmental controls. The Normalized Difference Vegetation Index (NDVI) negatively constrained species richness, reflecting the prevalence of intensively managed monoculture systems in this region, whereas temperature exerted a strong negative control on community evenness. Spatial predictions showed pronounced regional contrasts among the Southwest, Middle–Lower Yangtze, and Southern China agricultural zones, with uncertainty patterns closely linked to terrain complexity and data density. The results demonstrate that bacterial richness and evenness respond to distinct yet interacting environmental controls, highlighting the necessity of multidimensional diversity assessment. This study provides a scalable and interpretable framework for digital soil mapping of microbial diversity, offering spatially explicit insights for managing acid-sensitive agroecosystems under environmental change. Full article

The Taiyuan Formation in the Daniudi Gas Field of the Ordos Basin was formed in a tidal flat mixed sedimentary environment. It represents a typical tight sandstone reservoir with significant exploration and development potential. However, current research on the reservoir evaluation of the Taiyuan Formation is relatively limited. This study focuses on the Taiyuan Formation in the Da 17 well area of the Daniudi Gas Field, employing methods such as petrographic analysis, X-ray diffraction, porosity/permeability measurements, mercury injection capillary pressure, and scanning electron microscopy to conduct a detailed investigation of sand dams in the tidal flat environment. By combining reservoir characteristics with detailed sedimentary microfacies classification, and on the basis of discussing the relationship between reservoir heterogeneity and natural gas production capacity, this study performs reservoir classification and evaluation. The characteristics and distribution patterns of relatively high-quality reservoirs are clarified. Through the analysis of lithological characteristics and sedimentary markers, it is determined that the Taiyuan Formation in the Da 17 well area represents a typical tidal flat sedimentary environment, with sand dams serving as the primary development sites for sand bodies. The Taiyuan Formation sandstone is mainly composed of medium- to fine-grained lithic quartz sandstone and lithic sandstone. The predominant pore types include primary pores, secondary pores, and fractures. The porosity of the reservoir ranges from 0.3% to 14.2%, while the permeability varies from 0.006 to 29.7 × 10⁻³ μm², classifying it as a typical tight sandstone reservoir. Using the gas testing method, the lower limit of the physical properties of the Taiyuan Formation reservoir has been determined to be 3.9%. Considering the heterogeneity characteristics and productivity information of the sand layers, the reservoirs of the Taiyuan Formation are classified into three types: Type I reservoirs are predominantly composed of pebbly coarse sandstone and coarse sandstone, with porosity > 10%, permeability > 0.8 × 10⁻³ μm², mainly developed in the center of sand dams, exhibiting medium heterogeneity; Type II reservoirs consist of coarse sandstone with porosity ranging from 4% to 10% and permeability from 0.1 to 0.8 × 10⁻³ μm², with the main body of the sand dams belonging to this reservoir type, also showing medium heterogeneity; Type III reservoirs are composed of medium to fine sandstone, with porosity < 4% and permeability < 0.1 × 10⁻³ μm², mainly developed at the edges of sand dams, exhibiting strong heterogeneity. By integrating the distribution characteristics of sedimentary facies sand bodies, the planar distribution characteristics of the various reservoir types are clarified. Full article

Three-dimensional (3D) S-wave velocity field modeling is a critical task in seismic exploration, but balancing modeling accuracy and geological rationality remains challenging due to sparse observation data and inherent limitations of existing methods. To address this issue, we propose an Ordinary Kriging-Thin Plate Spline-Graph Convolutional Network (OTG) fusion model. It dynamically integrates the global trend capture capability of Ordinary Kriging, the local smooth processing ability of Thin Plate Spline, and the nonlinear feature fitting performance of Graph Convolutional Network (GCN) via an adaptive gating fusion mechanism. A multi-dimensional physical constraint loss function is further designed to ensure the geophysical plausibility of interpolation results. Validated on a dataset from 105 seismic stations in Southwest China using spatial cross-validation and random repeated experiments, the full OTG model achieves a root mean square error (RMSE) of 0.1148 km/s, a mean absolute percentage error (MAPE) of 1.9614%, and a Pearson correlation coefficient (PCC) of 0.9801. Compared with the optimal traditional method (OK) and the state-of-the-art hybrid method (DeepKriging), the proposed model reduces the root mean square error (RMSE) by 42.8% and 8.2%, respectively. This study demonstrates that the OTG model realizes the complementary advantages of traditional geoscientific methods and deep learning, providing a reliable engineering solution for high-precision 3D S-wave velocity structure interpolation in seismic exploration. Full article

Understanding the spatiotemporal dynamics of carbon emissions and developing differentiated governance strategies for urban agglomerations are essential for achieving regional low-carbon transformation. This study aims to identify the spatiotemporal patterns, driving mechanisms, and development-stage differences of carbon emissions across China’s urban agglomerations and to establish a type-specific governance framework. Based on multi-source geospatial and socioeconomic data from 19 urban agglomerations for the period 2006–2023, this study integrates spatial autocorrelation analysis, standard deviation ellipse analysis, hotspot analysis, random forest regression with SHAP interpretation, K-medoid clustering, and the Environmental Kuznets Curve (EKC) model to systematically examine emission evolution, influencing factors, and governance pathways. The results indicate the following: (1) carbon emissions in China’s urban agglomerations increased continuously during the study period and exhibited significant spatial heterogeneity, characterized by a “high east–low west” pattern, expanding eastern emission hotspots, and a gradual southwest shift in the emission centroid; (2) industrial structure and economic development level were identified as the dominant factors associated with carbon-emission differences, while energy efficiency, urbanization, and population density showed heterogeneous relationships across regions; (3) five carbon-emission development types were identified, including high-carbon high-development, transition-pressure, resource-dependent, stable-development, and low-carbon potential agglomerations, each exhibiting distinct development characteristics and governance requirements; and (4) EKC analysis revealed differentiated development stages among these types, suggesting that carbon governance should be tailored according to regional development conditions, dominant drivers, and emission-transition stages. This study provides an integrated geospatial modeling framework for understanding carbon-emission heterogeneity and offers scientific support for differentiated low-carbon planning and collaborative governance of urban agglomerations. Full article

This exploratory case study assesses ecotourism adaptability across three zones within the proposed Southwest Karst National Park (Guangxi), where UNESCO-recognised geological heritage intertwines with diverse ethnic cultures. To fill the gap in zone-specific integrated assessment frameworks, we employed a two-stage entropy-weighted TOPSIS method—calibrated via Delphi-AHP adjustments within a Social-Ecological Systems framework—using 2023–2024 data, complemented by ethnographic observations. Preliminary results indicate divergent trajectories: the Tiankeng zone exhibits high geological uniqueness and market appeal (composite score 0.74); the Mulun zone shows relatively balanced social–ecological coupling (0.58); and the Underground River zone displays pronounced asymmetry between vulnerability and development potential (0.16). Accordingly, tentative daily visitor caps of 3000, 8000, and 14 persons, respectively, and a pilot adaptability alert threshold of 0.20 for this specific case are proposed. These findings reveal a “vulnerability–capacity inverse matching” dynamic, informing differentiated governance: high-adaptability zones may allow moderate tourism with real-time monitoring, whereas low-adaptability zones require strict conservation-first measures and phased institutional capacity-building prior to any development. This preliminary framework offers actionable insights for sustainable management, conceptually aligned with SDG 11.4 and SDG 15.9. Full article

Clostridium perfringens type A causes neonatal clostridial enteritis, hemorrhagic enteritis, and sudden death syndrome in cattle. This study examined a fatal outbreak at a fattening farm housing 58 adult yaks in Southwest China, with an incidence of 22.41% (13/58) and a 100% case fatality rate. Histopathology of deceased yaks showed widespread multi-organ hemorrhage and severe hemorrhagic enteritis. A specific PCR assay amplified the alpha-toxin gene of Clostridium perfringens, while no other pathogens were detected. Clostridium perfringens was isolated from small intestinal contents using TSC agar, designated strain CQ1, and identified as type A through toxinotyping. Whole-genome sequencing revealed that strain CQ1 carries the core alpha-toxin gene along with key accessory virulence determinants (pfoA, colA, and cloSI). Phylogenomic analysis based on core single-nucleotide polymorphisms revealed that strain CQ1 is clustered into a distinct, independent evolutionary lineage with an existing historical yak isolate and a canine-origin isolate. Intraperitoneal challenge in mice confirmed its extreme lethality, causing 100% mortality within 36 h with severe intestinal necrosis. These findings provide fundamental genetic resources for understanding the virulence mechanisms and genetic background of this pathogen and offer critical insights for preventing and controlling clostridial diseases in high-altitude yak farming. Full article

During underground mining, stope stability is affected by excavation-induced stress redistribution and nonlinear degradation of rock mass strength. The conventional Mathews stability graph method employs an empirical stress factor A, which does not explicitly consider the nonlinear relationship between stress conditions and rock mass strength. In this study, the generalized Hoek–Brown criterion was introduced to define the maximum stress factor (MSF), and a modified stress factor A′ was developed by considering tensile and shear failure mechanisms. The proposed method incorporates the nonlinear stress–stability relationship of underground stopes and improves the reliability of stability assessment. A copper mine in southwest China was selected as a case study. The rock mass quality indices and stress parameters of ten representative stopes were obtained through field investigations, stope roof stress measurements, discontinuity surveys, and laboratory tests. The modified stress factor A′ was incorporated into the Mathews stability graph to account for excavation-induced stress redistribution and rock mass strength degradation. Compared with the conventional method, the modified approach generally reduced the stability numbers of the investigated stopes, with an average reduction of approximately 25.6% (excluding D1780-1, where the confinement strengthening effect resulted in a slight increase in stability number). The revised stability classifications show good consistency with the FLAC^3D simulation results and field observations, providing supporting evidence for the application of the proposed method to underground stope stability assessment in the studied mine. Full article