Search Results (224)

Conventional clustering techniques for urban water consumption profiling treat each household as an independent entity, thereby disregarding the spatial, socioeconomic, and infrastructural contexts that jointly govern demand behavior. This structural limitation prevents the extraction of contextually coherent consumption profiles—a critical shortcoming for utility managers who must design spatially targeted conservation interventions. To overcome this, we propose Simple GLAC, a novel graph clustering framework that leverages graph neural networks with an adaptive attention mechanism to dynamically model these complex interdependencies. The model’s end-to-end training jointly optimizes a latent representation for cluster cohesion, separation, and spatial homogeneity, where each household’s multi-month consumption record serves as the node feature vector encoding temporal consumption patterns. Evaluated on a large-scale real-world dataset of 4590 residential households across four distinct graph topologies, Simple GLAC consistently achieves superior multi-metric performance over both traditional and graph-based benchmarks, yielding interpretable and operationally actionable consumption profiles aligned with the spatial, administrative, socioeconomic, and infrastructural dimensions of urban water governance in the studied context. This work provides a data-driven tool for utility managers to deploy targeted water conservation strategies, with findings grounded in a Colombian mid-sized city and generalization to broader urban settings identified as a priority direction for future work. Full article

This paper proposes an experimental, intelligent optimization approach to improve the thermal cooling performance of an overclocked graphics processing unit (GPU). A closed-loop liquid-cooling system was built and tested utilizing deionized water and a silver (Ag) nanofluid coolant (0.015% vol.) across a variety of microchannel heat sink topologies with varying fin spacing. Key thermal performance indicators, including GPU temperature, coolant outlet temperature, and thermal resistance, were measured at different coolant flow rates. Experiments revealed that raising the flow velocity and decreasing the fin gap considerably enhanced cooling performance, while the Ag nanofluid consistently lowered GPU temperature by 1–3 °C compared to water. An Artificial Neural Network (ANN) surrogate model was constructed and trained using experimental data to support predictive analysis and system optimization, achieving excellent predictive accuracy with low RMSE. The trained ANN model was combined with the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to perform multi-objective optimization, aiming to minimize GPU temperature and thermal resistance while improving heat removal. The Pareto-optimal solutions revealed that nanofluid-based cooling offered the best trade-off circumstances, with optimal designs occurring at moderate flow rates and small fin spacing. The ANN-NSGA-II multi-objective optimization results indicated that the best thermal performance of the GPU cooling system was achieved when using Ag nanofluid (0.015 vol.%) as the coolant, with an optimal coolant flow rate in the range of 1.30–1.84 LPM and an optimal fin/channel spacing of 0.57–0.71 mm, producing GPU temperatures of 29.18–29.66 °C, coolant outlet temperatures of 29.06–29.41 °C, and a minimized thermal resistance of 0.0106–0.0152 °C/W; thus, overall, the suggested ANN-NSGA-II framework works well as a practical design tool for improving GPU cooling systems and may be used to other high-heat-flux electronic thermal management applications. Full article

Soil nematodes are integral to soil micro-food webs and serve as sensitive bioindicators of soil ecological condition. However, how forest vegetation and soil properties interact to shape nematode community assembly, network structure, and functional stability remains inadequately understood. Using 18S rRNA gene amplicon sequencing coupled with phylogenetic null modeling, we examined soil nematode communities across four forest types along a succession gradient. Although taxonomic diversity (e.g., Shannon and Pielou indices) differed significantly among forest types, network topology and stochastic assembly processes were more closely associated with community restructuring and co-occurrence patterns. This suggests that, while diversity is not irrelevant, network- and assembly-based metrics provide complementary and often more sensitive indicators of nematode community responses to forest type. Co-occurrence network analysis revealed that mixed forests fostered more complex and potentially stable networks, whereas plantations formed dense but potentially vulnerable networks. Assembly processes were not dominated by strong deterministic selection (|βNTI| ≤ 2 for most comparisons), a pattern consistent with undominated processes (e.g., ecological drift, weak environmental filtering). Dispersal limitation played a negligible role in this system. Partial Least Square Path Modeling identified spatial factors and key soil properties (e.g., pH, electrical conductivity, soil water content, and organic carbon) as primary drivers of community structure. Our findings indicate that assessing soil food web health should integrate network analysis and stochasticity metrics rather than rely solely on taxonomic diversity. For sustainable forest management, mixed-species stands are preferable to monoculture plantations, and maintaining soil physicochemical heterogeneity is critical for community stability. Full article

Water stress is a major abiotic constraint limiting the growth and productivity of alfalfa (Medicago sativa L.). To elucidate the adaptive mechanisms and identify key drought-tolerance genes, physiological measurements were integrated with multi-omics analyses of cultivar ‘Tamu 1’ under three water treatments: waterlogging (100% field water capacity), normal irrigation (80% FWC), and drought (light: 60% FWC, moderate: 40% FWC, severe: 20% FWC). Water stress markedly inhibited plant growth, induced oxidative stress, and reduced the photosynthetic capacity. Compared with waterlogging stress (DAMs: n = 71; DEGs: n = 313), drought stress resulted in a substantially greater number of differentially accumulated metabolites (DAMs, n = 1504) and differentially expressed genes (DEGs, n = 8006). Weighted gene co-expression network analysis (WGCNA) identified six key modules and ten hub genes associated with stress responses. Integrated transcriptomic and metabolomic analyses further revealed four major responsive pathways: starch and sucrose metabolism, phenylpropanoid and flavonoid metabolism, glutathione metabolism, and zeatin biosynthesis. Based on integrative criteria, including differential expression (|log₂FC| ≥ 1, adjusted p < 0.05), WGCNA modules significantly associated with drought-related traits (R² > 0.6), as well as functional annotation and protein–protein interaction (PPI) network topology, 28 candidate genes associated with drought tolerance were identified, of which six were further validated by quantitative real-time PCR (qRT-PCR). These findings highlight key metabolic pathways and regulatory modules underlying alfalfa responses to water stress and provide valuable candidate gene resources for improving drought tolerance. Full article

Although Graph Neural Networks (GNNs) are widely regarded as an ideal tool for capturing spatial dependencies in river basins, their effectiveness in hydrological forecasting is severely challenged by a topology paradox: under a purely data-driven paradigm, GNNs fail to spontaneously learn physical laws, generating predictions that lack physical interpretability and frequently violate mass conservation. To address this fundamental problem, this paper proposes a physics-informed graph learning framework integrated with an explicit, differentiable water balance constraint (WB-GNN). By reconstructing the continuity equation into a differentiable loss function, we directly embed physical conservation as a strong inductive bias into the neural network’s training objective. We comprehensively evaluated the model on two large-sample datasets (LamaH-CE and CAMELS) against state-of-the-art baselines, including EA-LSTM and unconstrained Pure-GNN. Quantitative results demonstrate that the proposed physical constraint successfully awakens the potential of river network topology. On the LamaH-CE dataset, WB-GNN achieved a Nash-Sutcliffe Efficiency (NSE) of 0.86 and a Root Mean Square Error (RMSE) of 9.2 ${\mathrm{m}}^3/\mathrm{s}$, outperforming both the domain-specific EA-LSTM (NSE: 0.83) and the unconstrained Pure-GNN (NSE: 0.74). Crucially, the introduction of the differentiable constraint reduced the Physical Inconsistency Ratio (PIR) by an order of magnitude-from 39.8% in the unconstrained model to just 4.3%. Similar robust improvements were validated across the highly heterogeneous CAMELS dataset. These quantifiable results confirm that the proposed method not only achieves superior forecasting accuracy but also fundamentally guarantees physical trustworthiness, making it highly robust for critical decision-making in extreme flood events. Full article

Proton exchange membrane water electrolysis is a promising technology for large-scale green hydrogen production due to its high efficiency, compact design, and rapid dynamic response. However, its commercialization is strictly limited by high material costs, durability issues, and complex multiphysics coupling within the membrane electrode assembly. This work provides a comprehensive and critical review of key physicochemical processes and advanced predictive modeling approaches for PEMWEs. To capture recent paradigm shifts, we introduce an innovative multi-dimensional classification framework—incorporating spatial resolution, temporal dynamics, and methodological paradigms—to critically evaluate lumped-parameter, continuum, microscale, and multiscale models, explicitly defining their applicability bounds and inherent limitations. The fundamental mechanisms governing electrode kinetics, membrane water transport, and gas–liquid two-phase flow are analyzed, establishing state-of-the-art quantitative benchmarks for microstructural parameters and advanced 3D flow field topologies under high-current-density and high-pressure regimes. Furthermore, we systematically examine model validation rigor, typical prediction errors, and the critical failure of static models in capturing dynamic property shifts during extreme bubble breakthrough. Recent breakthroughs integrating in situ diagnostics, pore-scale simulations, density functional theory, and Physics-Informed Neural Networks are extensively discussed. Future efforts must prioritize mechanical–electrochemical–thermal coupling, transient degradation prognostics, and machine learning-driven predictive digital twin technologies to overcome current empirical limitations and accelerate the gigawatt-scale deployment of PEMWE systems. Full article

Water ecosystem service flows (WESFs) help address spatial mismatches in water resources and support basin resilience. However, their dynamic evolution and nonlinear drivers under climate change and intensive human activities remain poorly understood. This study evaluates the spatiotemporal dynamics of WESFs in the Yangtze River Basin (YRB) from 2005 to 2022 by integrating dynamic flow analysis with mechanism interpretation. We developed an integrated framework coupling SWAT hydrological simulations with a proxy-based spatial allocation approach for social water demand. Using the Water Stress Index (WSI) and river topology, dynamic inter-regional WESFs were simulated. Furthermore, an interpretable machine learning approach was employed to identify the nonlinear effects of multiple driving factors. Results reveal a persistent supply–demand mismatch: supply exhibited a northwest–southeast gradient (averaging 567.21 mm annually), while demand concentrated in mid-lower plains and urban corridors. The flow network, which accounts for accumulated upstream inflow, demonstrated a stable “upstream supply, mid-reach transmission, and downstream benefit” pattern, highlighting downstream reliance on upstream inputs. Driving analysis identified land surface and vegetation as the largest associated driver category, while climate–hydrology and human activity were not cleanly separable. Climate provided the hydro-climatic conditions for redistribution. Nonlinear responses and blue–green interactions were also identified, informing transboundary ecological compensation and regional water-resilience management. Full article

Leak localization and maintenance in water distribution networks (WDNs) are essential for reducing water losses and operating costs; however, they usually require extensive monitoring and large datasets. This work proposes a methodology that combines topological sectorization of a hydraulic node network and deep learning techniques to improve leak location by selecting representative nodes to reduce the spatial dimensionality of the WDNs. The network is partitioned using a Spectral Clustering algorithm to identify key nodes based on a weighted criterion that considers pressure variability, flow rate, and proximity to the centroid. Subsequently, a Bidirectional Long Short-Term Memory (Bi-LSTM) neural network classifies the cluster and sub-cluster where a leak occurs, using pressure and flow time series simulated in EPANET. This methodology was validated on the L-Town network, achieving an accuracy of 99.94% for cluster classification and 99.82% for sub-clusters, with a validation loss of 0.024%. During validation with 117 unseen leakage scenarios, the model reached an overall effectiveness of 85%. Moreover, Spectral Clustering outperformed K-Means in preserving physical connectivity. These results confirm the efficiency of the proposed methodology and highlight its potential for application in other hydraulic networks. Full article

Against the backdrop of rapid global urbanization, peri-urban villages universally face the dual dilemmas of landscape homogenization and the imbalance between heritage preservation and functional renewal. As a typical representative, the “Shanghai-style Jiangnan” villages feature an open water–land chessboard pattern and linear water-house parallel organization, which are distinctly different from the closed and introverted texture of traditional Suzhou-Hangzhou water towns. Such villages urgently need to balance the continuation of the original spatial fabric and the adaptation of modern functions. Existing studies on rural landscapes mostly focus on the static vertical identification of single elements, lacking a systematic quantitative analysis of the horizontal topological relationships among multiple elements, making it difficult to accurately define the spatial boundaries between preservation and renewal. This study takes Xinyuan Village in Jinshan District, Shanghai, as an empirical subject to construct a model for the vertical gene decoding of the “Point-Line-Network” and horizontal topology coupling of “Surface Gene.” By introducing a landscape sensitivity assessment combined with the Entropy Weight Method (EWM) and GIS (Geographic Information System) spatial Kernel Density Estimation (KDE), a quantifiable landscape control heat map is generated. The study identifies the nested original fabric structure of the “house-water-field-forest-road” and the spatial landscape differentiation characteristics in Xinyuan Village and delineates three-tier differentiated zoning controls through dual-verified heat maps. Validated based on Xinyuan Village, this method effectively resolves the conflict between rural preservation and renewal and realizes the transformation from static museum-style preservation to refined adaptive zoning. It provides specific practical strategies for the renewal of “Shanghai-style Jiangnan” villages and offers a quantitative morphological reference for enhancing the spatial resilience and living heritage of peri-urban villages, while its cross-regional transferability needs further verification. Full article

Global climate change has altered temperature regimes, hydrological stability, and redox dynamics in inland waters, yet the continental-scale impact of these alterations on sediment microbiomes remains poorly understood. Here, we compiled 562 publicly available 16S rRNA gene datasets from lake sediments across five major climatic zones in China to examine how climatic gradients influence microbial diversity, community assembly, and interaction networks, as well as their associated taxonomic composition and environmental responses. Sediment microbiomes showed clear spatial differentiation in both α- and β-diversity, accompanied by climatic zone-specific taxonomic signatures and biomarker taxa. Community assembly also varied markedly across climatic zones, with stochasticity and dispersal limitation dominating in colder regions, transitional assembly in the south temperate zone, and stronger selective or high-turnover dynamics in the warm subtropics. Importantly, random forest models revealed a clear transition from climate-dominated to anthropogenic-dominated control in sediment microbiome organization: microbial variation in the plateau and temperate regions was primarily associated with climatic and geographic constraints, whereas anthropogenic factors played a more important role in shaping community differentiation in the central subtropical zone. By integrating diversity patterns, taxonomic composition, assembly processes, and network topology, we further propose a three-stage conceptual pattern of sediment microbial community organization along climatic gradients, shifting from a persistence-dominated regime in the cold plateau regions, to an efficiency-dominated regime in the temperate zones, and finally to a plasticity-dominated regime in the warm subtropical regions. These findings would provide a continental-scale framework for understanding sediment microbiome responses to coupled climatic and anthropogenic forcing in inland waters, with implications for future water quality management and ecosystem conservation. Full article

Inter-basin water transfer projects (IBWTPs) play a pivotal role in alleviating the spatiotemporal imbalances of water resources. However, their operation is exposed to multiple, highly interdependent safety risks that can significantly undermine system stability and water supply reliability. Existing studies predominantly focus on isolated risk factors or rely heavily on subjective data, which limits their ability to capture the complex interrelationships among risks and reveal their underlying propagation mechanisms. To address these limitations, this study proposes a novel risk correlation analysis framework that integrates Interpretive Structural Modeling (ISM) with copula functions. ISM is first employed as a preprocessing tool to structure expert knowledge and develop an initial risk correlation framework. It is then used to hierarchically organize the complex interrelationships among risks. Subsequently, copula functions are utilized to model nonlinear dependencies and tail behaviors among risk variables. This enables a quantitative assessment of correlation strengths and facilitates the construction of a risk topological network. An empirical case study is conducted based on the Middle Route of the South-to-North Water Diversion Project. The results reveal 13 significant correlations among six second-level risk categories. Natural risks (e.g., floods and geological hazards) are identified as the primary driving factors. They exhibit a strong positive correlation (0.6155) with engineering risks and serve as the most critical nodes for proactive risk prevention and control. Engineering risks function as central intermediary hubs in the risk transmission process, whereas water quality and economic risks are characterized as terminal endpoints. Furthermore, three principal risk propagation pathways are identified: (1) natural risks → engineering risks → economic risks; (2) natural risks → operational scheduling risks → social risks; and (3) engineering risks → water quality risks → economic risks. The resulting risk topological network demonstrates significant small-world properties, indicating highly efficient risk transmission within the system. Ultimately, this study provides a robust quantitative approach for analyzing risk interactions in complex engineering systems and enriches the theoretical framework of engineering risk management. It also identifies critical nodes and key transmission pathways for risk prevention and control in IBWTPs, thereby offering significant practical implications for operational safety. Full article

Aquatic–terrestrial ecotones are highly dynamic biogeochemical hotspots where hydrological fluctuations profoundly influence microbial community structure and ecosystem functioning. However, the mechanisms underlying microbial community responses across hydrological gradients remain insufficiently understood. In this study, 16S rRNA gene sequencing was used to comparatively analyze bacterial communities in the waterward and landward zones of the drawdown area of the Danjiangkou Reservoir. The results showed that bacterial community composition differed significantly between the two zones, and waterlogging markedly increased bacterial α-diversity. Community variation was primarily associated with key environmental factors, including total phosphorus (TP), soil moisture content (SMC), and nitrate nitrogen (NO₃⁻-N). Compared with the landward zone, stochastic processes contributed more to community assembly in the waterward zone, which also exhibited higher network complexity and topological stability. In addition, several keystone taxa were identified, suggesting their potential roles in maintaining network structure and ecological stability. Functional prediction further revealed distinct metabolic potentials between zones, with enhanced anaerobic and redox-related functions in the waterward zone and predominantly aerobic metabolism in the landward zone. These findings suggest that hydrological fluctuations reshape bacterial community structure and potential ecological functions by jointly regulating water availability and nutrient dynamics. This study provides new insights into microbial ecological processes in reservoir riparian zones and offers a scientific basis for the management of aquatic–terrestrial ecotone ecosystems. Full article

Urban flood resilience is an important indicator for measuring a city’s capacity to respond to and recover from flood disasters. However, existing assessments often lack a long-term hydrological baseline. This study establishes the historical water system of Yuan Dadu (1267–1368 CE) as a control scenario to benchmark the flood resilience of modern Beijing. By integrating a historical geographic reconstruction with a hydrological–hydrodynamic simulation and the fuzzy analytic hierarchy process (FAHP), the research quantifies structural differences in resilience profiles between the nature-adapted historical system and the modern engineering-dominated system. The results indicate that Yuan Dadu’s urban flood resilience index (UFRI) is 3.44 and modern Beijing’s is 3.28. Despite modern Beijing’s significant advantage in drainage facility density (0.61 km/km²) and emergency management, the system exhibits a functional substitution failure, where gray infrastructure has failed to fully compensate for a 26% reduction in the unit area storage capacity (from 6.4 to 4.7 × 10⁴ m³/km²) and a 48.4% decline in the water system structural complexity. The findings indicate that, in rapidly urbanized cities on alluvial plains with high impervious coverage, expanding drainage networks alone may be insufficient to offset losses in a natural hydraulic buffering capacity. Accordingly, planning strategies are proposed that integrate distributed micro-storage and restore topological connectivity to recreate system-level hydraulic buffering functions. Full article

This study examines the carbonation behavior and CO₂ storage potential of a Ca-rich alkali-activated binder produced entirely from industrial residues-ladle furnace slag (LFS), coal ash (CA), and cement kiln dust (CKD). The system was designed as a one-part alkali-activated material (AAM), with CKD acting as an internal activator, and subjected to ambient curing, water curing, and accelerated CO₂ curing at ambient pressure. Phase evolution, microstructural development, and pore-structure characteristics were investigated using X-ray diffraction, FTIR spectroscopy, DSC–TG analysis, scanning electron microscopy, and X-ray micro-computed tomography, together with measurements of density, water absorption, and compressive strength. Loss-on-ignition measurements combined with chemical analysis were further used to quantify CO₂ uptake and evaluate the degree of carbonation of the binder system. CO₂ curing fundamentally altered the reaction pathway of the binder, shifting it from hydration-dominated to carbonation-controlled phase evolution, leading to the decomposition of calcium-bearing hydrates and complete carbonation of non-hydraulic γ-belite with the formation of vaterite, aragonite, and calcite. These transformations induced pronounced microstructural densification, reflected in a near-doubling of compressive strength (>48 MPa), increased apparent density, reduced water absorption, and simplified pore-network topology. A preliminary carbon footprint assessment indicates that the production of 1 m³ of the developed LFS–CA–CKD concrete generates about 14.36 kg CO₂-eq, while the carbonation process enables significant CO₂ sequestration, resulting in a net negative carbon balance. The results demonstrate that controlled carbonation is an effective post-treatment strategy for waste-derived alkali-activated binders, enabling simultaneous performance enhancement and permanent CO₂ sequestration. Full article

Underwater crack detection is critical for the structural health monitoring of concrete dams; however, complex turbid environments and limited computational resources on underwater robots pose significant challenges. This study proposes HAIS-SegFormer, a lightweight segmentation network utilizing a Mix Transformer backbone. We introduce a tandem Hybrid Attention mechanism—cascading Coordinate Attention (CoordAtt) and Convolutional Block Attention Modules (CBAM)—to preserve long-range topological connectivity and refine local edge details. Furthermore, a Feature Inhibition Module (FIM), modeled after biological lateral inhibition, is designed to actively suppress high-frequency background noise such as water plants. Experimental results on an underwater crack dataset demonstrate that HAIS-SegFormer achieves a favorable trade-off between segmentation accuracy (71.66% mIoU) and computational efficiency (73 FPS, 3.80 M parameters). The proposed framework provides a robust and resource-efficient solution for automated underwater inspections. Full article