Uranium contamination from mining and natural sources poses a major environmental and health risk, as soluble uranium U(VI) readily migrates through groundwater systems. Microbial reduction to insoluble U(IV) via dissimilatory metal-reducing bacteria offers a sustainable remediation method, relying on extracellular electron transfer (EET) to shuttle electrons to extracellular acceptors. MR-1 (.MR-1) serves as a model organism for this process, but its EET efficiency is hindered by limited endogenous redox mediators and biofilm conductivity. Despite advances in genetic engineering, the potential of synthetic microbial communities to enhance EET through interspecies interactions remains underexplored. Here we show a synthetic consortium comprising .MR-1 and a non-U-reducing isolate, LXZ1 (.LXZ1), that fully reduces U(VI) within 48 h, compared to only 60 % reduction by .MR-1 alone. This enhancement stems from .LXZ1-secreted pyocyanin, which binds selectively to .MR-1's outer-membrane cytochrome OmcA, shifting its redox potential to facilitate directional electron flow along a thermodynamic gradient. Concurrently, conductive extracellular DNA released by . LXZ1 promotes electron transport and aggregate formation, as evidenced by electrochemical assays, transcriptomics, and molecular dynamics simulations. These synergistic mechanisms alleviate proton-transfer limitations and upregulate metabolic pathways, boosting overall EET rates. By harnessing natural microbial cooperation, this approach provides insights into community-driven metal reduction and paves the way for efficient, scalable bioremediation strategies in contaminated sites.

Buildings are among the largest contributors to global energy consumption and carbon emissions, making their transformation essential for advancing environmental sustainability goals. Innovative technologies such as artificial intelligence (AI) and digital twins (DTs) offer powerful tools for optimizing performance in smart, green, and zero-energy buildings. However, existing research remains fragmented-AI and AI-driven DT applications are often confined to isolated functions or specific building types-resulting in a limited, non-cohesive understanding of their collective potential in the built environment. This fragmentation, in turn, has hindered the development of integrated strategies that link building-level efficiencies with the broader environmental objectives of smart cities. To address these interrelated gaps, this study conducts a comprehensive systematic review of leading-edge AI and AI-powered DT solutions applied across smart, green, and zero-energy buildings. It aims to provide a holistic understanding of how these solutions enhance environmental performance through the analysis of key building-related indicators. By synthesizing, comparing, and evaluating recent research, it examines how AI and AI-powered DT technologies facilitate integrated, system-level strategies that promote environmentally sustainable smart practices across the built environment. The study reveals that AI enhances smart buildings by enabling dynamic energy optimization, occupant-centered environmental control, improved thermal comfort, renewable energy integration, and predictive system management. In green buildings, AI contributes to greater resource efficiency, minimizes construction and operational waste, promotes the use of sustainable materials, strengthens cost estimation and risk assessment processes, and supports adaptive design strategies. For zero-energy buildings, AI facilitates multi-objective optimization, advances explainable and transparent AI-driven control systems, supports performance benchmarking against net and nearly zero-energy standards, and enables renewable energy integration tailored to diverse climatic and regulatory contexts. Furthermore, AI-powered DTs enable real-time environmental monitoring, predictive analytics, anomaly detection, and adaptive operational strategies, thereby enhancing building performance, energy optimization, and resilience. At broader spatial scales, these technologies foster interconnected urban ecosystems, advancing environmental sustainability, sustainable development, and smart city initiatives. Building on these insights, this study introduces a novel integrated framework that positions AI and AI-driven DTs as systemic enablers of environmentally sustainable smart built and urban environments, emphasizing their cross-scale convergence in promoting carbon neutrality, circular economy principles, climate resilience, and regenerative urban strategies. The findings offer actionable pathways for advancing research agendas, inform practical strategies for building and urban system design, and provide evidence-based recommendations for policymakers committed to fostering more intelligent, sustainable, and resilient urban futures. This work establishes AI and AI-driven DTs as transformative catalysts for realizing the next generation of resource-efficient, carbon-neutral, and ecologically integrated urban ecosystems.

The Arctic, though remote, is exceptionally vulnerable to chemical contaminants that threaten its fragile ecosystems. Bisphenols (BPs), a class of endocrine-disrupting chemicals used in plastics and resins, are now detected across the Arctic, but the risks posed by their many analogues are poorly understood. Most studies have focused on documenting their presence, leaving a critical gap in our understanding of whether these compounds bioaccumulate in Arctic food webs and to what extent local, within-Arctic pollution contributes to the overall burden. Here we show, through a comprehensive analysis of 32 BPs in 134 samples from a Norwegian Arctic food web, that multiple BP analogues not only bioaccumulate but also biomagnify from plankton up to polar bears. We found that 5,5'-(1-methylethylidene)bis [(1,1'-biphenyl)-2-ol] (BPPH) exhibited the highest trophic magnification factor (TMF = 2.3), and we documented total BP concentrations in polar bear tissues up to 1396 ng g wet weight, orders of magnitude higher than in lower-trophic-level species. Furthermore, our analysis identified distinct local pollution sources, such as a firefighting training site releasing 2,4,6-trichlorophenol (2,4,6-TBP) and landfill leachate contributing other BPs to the local environment. These findings provide the first evidence of trophic magnification for multiple BPs in a polar food chain and underscore the urgent need to incorporate food-web dynamics and local source management into ecological risk assessments for the Arctic.

Wastewater-based epidemiology (WBE) offers a unique window into the health and habits of communities through the analysis of pollutants and biomarkers in sewage. Traditionally focused on small molecules, such as pharmaceuticals and illegal drugs, recent advances in environmental proteomics have expanded WBE to include large biomolecules such as proteins. Notably, novel sampling methods using polymeric probes and high-resolution mass spectrometry have facilitated the detection of human and animal proteins, both soluble and in particulate material, linking them to specific populations and industrial activities. An immunological dimension to this approach is fundamental to include the recognition of host immunoglobulins, immune-response proteins, and pathogen antigens in wastewater, potentially serving as indicators of community immune status, infection prevalence, and vaccination coverage. This review consolidates the latest advancements in environmental proteomics as applied to WBE, emphasizing an immunological perspective as a comprehensive tool for assessing population health and environmental conditions to bridge environmental monitoring, public health, and clinical diagnostics.

Non-point source pollution from agricultural activities poses a significant threat to water quality by introducing excess nutrients like nitrogen into aquatic ecosystems, leading to issues such as eutrophication and groundwater contamination. In agricultural watersheds, nitrate transport involves intricate physical, chemical, and biological processes influenced by meteorological conditions, hydrological features, and spatial topologies, making accurate short-term predictions challenging. Traditional data-driven deep learning models often fail to incorporate physical constraints and complex spatiotemporal dynamics, limiting their interpretability and predictive accuracy. Here we show a hierarchical transformer and graph neural network model that accurately predicts watershed nitrate concentrations by integrating multi-source data and simulating pollutant migration. The model captures nonlinear multivariate temporal patterns through hierarchical transformers, fuses global meteorological and local hydrological features via neural networks, and models runoff topologies with physically constrained graph neural networks. For predicting the concentration changes of pollutants discharged from watersheds, it outperforms baselines like multi-layer perceptrons, recurrent neural networks, and long short-term memory networks, with state-of-the-art performance in root mean square error, mean absolute error, and . Ablation studies confirm the essential roles of multi-source data integration and watershed topological modeling in enhancing performance. This method of directly modeling physical processes by leveraging the characteristics of different neural network architectures opens up a new path for addressing the interpretability problem in neural earth system modeling, apart from the process-guided deep learning and differentiable modelling methods.

Hyperthermophilic composting (HC) represents a promising approach for converting organic solid waste into valuable resources by leveraging extreme temperatures to enhance microbial degradation and detoxification processes. In this high-temperature environment, microbial communities undergo dynamic succession, where thermophilic bacteria dominate and drive efficient organic matter transformation through adapted metabolic pathways and stress responses. These adaptations include the stabilization of cellular structures and enzymes, often mediated by heat shock proteins (HSPs) that prevent protein misfolding under thermal stress. However, the integrated mechanisms linking community-level functional shifts to molecular-level protein remodeling in thermophiles during HC remain poorly understood. Here we show a coordinated interaction of functional succession and molecular adaptations within thermophilic bacteria in HC, which collectively achieve heat resistance. This interaction encompasses enhanced metabolic and genetic modules, accounting for 97 % of the variance observed in thermophile abundance. Metagenomic analyses revealed upregulation of translation, transcription, amino acid metabolism, and cell wall biosynthesis, coupled with suppression of mobilome functions to maintain genomic stability, as confirmed by partial least squares path modeling and Boruta analyses. Molecular dynamics simulations of key enzymes from the thermophile further demonstrated intrinsic structural rigidity, reduced hydrophobic exposure, and hierarchical chaperone activity involving DNAJ, DNAK, and GroEL for protein repair. These findings enhance our comprehension of microbial thermotolerance and establish a foundation for optimizing composting efficiency and advancing heat-resistant microbial applications in biotechnology and waste management. Additionally, they offer insights into the evolution of thermophiles, protein engineering, and stress adaptation across various biological and industrial systems, thereby promoting the integration of environmental engineering and systems biology.

Microbial electrorespiration harnesses bacteria to drive reductive dechlorination, offering a sustainable method to remediate environments contaminated with persistent chlorinated organic pollutants (COPs). However, aquifers' complex hydrogeological and hydrochemical conditions, combined with uneven COP distribution, create substantial spatial and temporal variability in biochemical reactions, environmental factors, and microbial communities. Traditional trial-and-error experiments are labor-intensive and slow, impeding the quick identification of conditions that accelerate dechlorination rates. Here we show that a machine learning framework, integrating experimental design with cathodic biofilm data, uncovers key interrelationships among environmental variables, dechlorination kinetics, electrochemical properties, and functional microbes, enabling rapid optimization of bioelectrodechlorination. Trained on literature-derived datasets using models such as extreme gradient boosting, random forest, and multilayer perceptron, this framework identifies temperature and cathode potential as primary drivers in experimental design while highlighting key biofilm genera, including , , , , , , , and . It supports inverse design to determine optimal parameters-such as cathode potential, temperature, and additives-for dechlorinating representative COPs, including tetrachloroethene, trichloroethene, and 1,2-dichloroethane, achieving reaction rate predictions with errors below 6 %. This approach surpasses conventional methods by increasing efficiency, cutting costs, and accelerating bioremediation without extensive laboratory testing. By incorporating microbial community insights into predictive models, our data-driven strategy advances the scalable application of microbial electrorespiration for COP-contaminated water remediation and paves the way for broader bioelectrochemical uses in environmental engineering.

Microbial contamination in building plumbing systems poses significant risks to public health at the point of use. Stagnation and warm temperatures are well-known drivers of microbial regrowth, but the effects of common short-term stagnation in touchless sensor faucets-widely used for hygiene and comfort-remain poorly understood. Here we show that microbial water quality in touchless sensor faucets changes during short-term stagnation (0.25-10 h) at varying temperatures (10, 30, and 40 °C). We identify two pivotal time points-2 and 4 h-where microbial diversity decreases and concentrations increase significantly, driven by accelerated chlorine decay and biofilm contributions. Heating to 30 °C maximizes microbial biomass (measured as ATP) but minimizes proliferation, whereas 40 °C reduces biomass while promoting growth. These findings reveal a temperature-dependent microbial water quality guarantee period of 2-4 h, beyond which flushing is necessary to mitigate health risks. Optimizing faucet temperatures between 30 and 40 °C could balance microbial safety, user comfort, and energy efficiency, offering practical guidance for managing water quality in modern plumbing systems.

Urban greenspaces enhance human well-being and promote sustainable development in rapidly urbanizing regions by delivering vital ecosystem services, including cooling, air purification, and recreation. In China, where cities accommodate a large share of the population amid persistent environmental pressures, disparities in greenspace exposure pose a major obstacle to equitable access; these disparities arise from geographic, climatic, socioeconomic, and landscape factors. Although awareness of such inequalities is growing, their long-term trajectories, demographic and city-scale patterns, and viable spatial optimization approaches remain largely unexplored. Here we show that greenspace exposure inequality across 246 Chinese cities increased by 25 % from 2000 to 2020 and is projected to rise further by 12.2-15.7 % by 2050 under middle-of-the-road and fossil-fueled development scenarios, disproportionately affecting older, less-educated women and megacity residents. Geodetector and random forest analyses reveal that this rise results from interactions among greenspace coverage, population density, and patch connectivity, which explain 83.9 % of the inequality. A network-based optimization approach that improves patch connectivity-without expanding total greenspace-can reduce disparities by 10.3-20.8 %, with greater efficacy in high-inequality cities and among vulnerable populations. Our results highlight how precise landscape interventions can advance social equity in greenspace access, supporting Sustainable Development Goal 11 for inclusive, resilient urban environments.

Marine fuel oil (MFO) spills in tropical coastal environments are under-characterized despite increasing risk from maritime activities. Microbial and geochemical responses to the June 2024 Marine Honour MFO spill on Singapore's intertidal sediments were analyzed in real time over 185 days. Using metagenomics and hydrocarbon profiling, microbial community shifts and hydrocarbon degradation were quantified across visibly oiled (high-impact) and clean (low-impact) sites. Microbiomes at all sites adapted rapidly to the spill through increased diversity and abundance of genes encoding alkane and aromatic compound degradation, detoxification, and biosurfactant production. The dominant hydrocarbon-degrading bacteria differed markedly from those reported in other crude oil spills and in regions with different climates. Oil deposition intensity strongly influenced microbial succession and hydrocarbon-degrading gene profiles, and this reflected early toxicity constraints in heavily oiled areas. The persistence of hydrocarbon degradation genes beyond hydrocarbon detection in sediments suggested long-term functional priming may occur. The study provides novel genome-resolved insight into the microbial response to MFO pollution, advances understanding of marine environmental biodegradation, and provides urgently needed baseline data for oil spill response strategies in Southeast Asia and beyond.

Freshwater reservoirs are critical for water management but face increasing impacts from climate change, which alters their thermal regimes and affects ecosystem functions globally. In temperate regions, surface water temperatures have risen at rates often surpassing those of air temperature, driven by atmospheric warming, hydrological processes, and reservoir morphometry. However, long-term studies on reservoir-specific thermal responses, particularly short-term variability, remain scarce. An important question is how environmental drivers influence both long-term warming trends and daily thermal fluctuations in managed water bodies. Here we show that over 31 years (1991-2021), surface water temperatures in 35 Czech reservoirs increased by an average of 0.59 °C per decade, with air temperature, altitude, and retention time as primary predictors of mean temperatures. A novel corrected metric for day-to-day variability () revealed that inflow rate, depth, and retention time strongly influence short-term fluctuations, and trends positively correlated with warming rates, indicating linked drivers of thermal reorganization. Seasonal patterns showed strongest warming in April, with an anomaly of minimal change in May, likely tied to regional climatic shifts. These findings elucidate climate-driven thermal dynamics in reservoirs, highlighting the interaction of climatic and local factors. By combining statistical modeling with process-based indicators, this study informs adaptive strategies to mitigate impacts on water quality, stratification, and biodiversity under changing climates.

Terrestrial pathogens are increasingly being detected in marine organisms, raising concerns about ecosystem sustainability, biodiversity loss, and threats to human health. Over the past two decades, reports of microbial contaminants crossing from land to sea have increased, suggesting shifts in pathogen ecology driven by environmental changes and human activities. Pathogens originating on land can spread, adapt, and persist in marine environments, infecting a wide range of hosts and potentially re-entering terrestrial environments. Despite growing recognition of this issue, a comprehensive understanding of the distribution, diversity, and transmission pathways of these pathogens in marine ecosystems remains limited. In this Review, we provide a global analysis of terrestrial pathogen contamination in marine animal populations. Drawing from pathogen detection data across 66 countries, we used phylogenetic methods to infer land-to-sea transmission routes. We identified 179 terrestrial pathogen species, including 38 bacterial, 39 viral, 80 parasitic, and 22 fungal species, in 20 marine host species. Terrestrial pathogens are not only widespread but also highly diverse in marine ecosystems, highlighting the frequency and ecological significance of cross-system microbial exchange. By revealing the scale and complexity of land-to-sea pathogen flow, we show that climate change, pollution, and other anthropogenic pressures may intensify pathogen spillover events, with potential feedback effects on terrestrial systems. This highlights the urgent need for integrated surveillance and policy frameworks acknowledging the interconnectedness of terrestrial and marine health. Our work advocates a One Health approach to microbial ecology, stressing the need to safeguard marine and human populations from emerging cross-system threats.

Humans shape the world through policies, practices, and behavior that create environmental heterogeneity. Political and critical ecology offer frameworks for understanding how societies have historically and currently used power, policies, and practices to shape environmental landscapes and conditions, ultimately influencing the ecology and evolution of biodiversity. We suggest that integrating political and critical ecology can enhance our understanding of anthropogenic influences, such as luxury effects and legacy effects, including redlining-a form of structural racism implemented in the United States. Here, we review the consequences of legacy and luxury effects on urban ecosystems, with a focus on their impact on the fauna and flora. We propose that legacy and luxury effects can have independent and interdependent influences on ecological diversity, abundance, biological invasions, and pollution exposure. Although these effects can persist, environmental remediation may provide a pathway to restorative justice. We also discuss , herbaceous plants with the potential to mitigate the impacts of cadmium, a notorious environmental contaminant whose disposition parallels redlining patterns. Phytoremediation can contribute to biofuels, biofoundries, and the green economy, offering solutions to restore affected communities. By applying political and critical ecology lenses, we can identify socio-ecological mechanisms that affect humans and the environment. These insights can inform the development of green infrastructure to help remediate adverse effects. Ideally, these approaches provide pathways to address historical injustices, enhance equity, and restore ecological landscapes.

Large language models (LLMs) are revolutionizing specialized fields by enabling advanced reasoning and data synthesis. Environmental science, however, poses unique hurdles due to its interdisciplinary scope, specialized jargon, and heterogeneous data from climate dynamics to ecosystem management. Despite progress in subdomains like hydrology and climate modeling, no integrated framework exists to generate high-quality, domain-specific training data or evaluate LLM performance across the discipline. Here we introduce a unified pipeline to address this gap. It comprises EnvInstruct, a multi-agent system for prompt generation; ChatEnv, a balanced 100-million-token instruction dataset spanning five core themes (climate change, ecosystems, water resources, soil management, and renewable energy); and EnvBench, a 4998-item benchmark assessing analysis, reasoning, calculation, and description tasks. Applying this pipeline, we fine-tune an 8-billion-parameter model, EnvGPT, which achieves 92.06 ± 1.85 % accuracy on the independent EnviroExam benchmark-surpassing the parameter-matched LLaMA-3.1-8B baseline by ∼8 percentage points and rivaling the closed-source GPT-4o-mini and the 9-fold larger Qwen2.5-72B. On EnvBench, EnvGPT earns top LLM-assigned scores for relevance (4.87 ± 0.11), factuality (4.70 ± 0.15), completeness (4.38 ± 0.19), and style (4.85 ± 0.10), outperforming baselines in every category. This study reveals how targeted supervised fine-tuning on curated domain data can propel compact LLMs to state-of-the-art levels, bridging gaps in environmental applications. By openly releasing EnvGPT, ChatEnv, and EnvBench, our work establishes a reproducible foundation for accelerating LLM adoption in environmental research, policy, and practice, with potential extensions to multimodal and real-time tools.

Industrial polycyclic aromatic hydrocarbons (PAHs) pollution threatens soil ecosystems worldwide, posing persistent risks due to their toxicity and intricate transport dynamics. In steelworks, a major PAH emitter, contaminant distribution arises from multifaceted interactions between production activities and geological features, complicating the elucidation of underlying mechanisms. Previous studies have largely overlooked the inherent heterogeneity in these influences, focusing instead on global relationships that may bias assessments of pollution drivers and PAH migration. Here we show heterogeneity, nonlinearity, and multifactor interactions in PAH contamination at a steelworks site using a multidimensional framework that integrates machine learning and spatial analysis. Applied to 3339 soil samples and nine influencing factors, the framework reveals distance to production facilities as the dominant driver, with a 60-m impact radius; production factors exert stronger effects on 2-3-ring PAHs than on 4-6-ring PAHs, particularly in deeper soil layers at depths of 9-20 m. Soil moisture and clay content synergistically control PAH mobility across strata, elevating the framework's explanatory power from 0.5 to 0.9 and enabling precise delineation of dynamics. This modular approach not only advances mechanistic insights into industrial PAH pollution but also provides scalable guidance for targeted prevention and remediation strategies across diverse contaminated sites.

Dissolved organic matter (DOM) is a key determinant of heavy metal fate in aquatic environments, influencing their mobility, toxicity, and bioavailability. Derived from natural sources such as soil and vegetation decomposition, natural DOM (N-DOM) typically features humic-like substances with abundant oxygen-containing functional groups that stabilize heavy metals through complexation. However, microplastic-derived DOM (MP-DOM), increasingly prevalent due to plastic degradation, may interact differently with heavy metals, potentially exacerbating environmental risks amid rising plastic pollution. Yet, how heavy metals drive molecular transformations in MP-DOM versus N-DOM remains unclear, hindering accurate pollution assessments. Here, we compare interactions between N-DOM and MP-DOM with cadmium, chromium (Cr), copper, and lead from both fluorescence and molecular perspectives. Our results show that N-DOM, dominated by humic-like substances (46.0-57.3 %), lignin-like (55.0-64.9 %), and tannin-like (10.1-17.6 %) compounds, forms more stable heavy metal complexes via carboxyl, phenolic hydroxyl, and ether groups than MP-DOM. By contrast, MP-DOM-enriched in protein/phenolic-like substances (13.8-24.0 %), condensed aromatic (12.1-28.5 %), and protein/aliphatic-like (8.6-12.4 %) compounds-yields less stable complexes and is highly susceptible to Cr-induced oxidation. Mass-difference network analysis and density functional theory calculations further reveal that both DOM types undergo heavy-metal-triggered decarboxylation and dealkylation, but N-DOM retains complex structures, whereas MP-DOM degrades into smaller, hazardous molecules such as phenol and benzene. This study underscores the potential for heavy metals to exacerbate the ecological risks associated with the transformation of MP-DOM, providing crucial insights to inform global risk assessment and management strategies in contaminated waters where plastic and metal pollution co-occur.

Understanding and mitigating PM and ozone (O) pollution remains challenging due to the nonlinear atmospheric chemistry and spatially heterogeneous nature of pollutant emissions. Traditional forward modeling approaches suffer from high computational cost and limited diagnostic resolution to precisely attribute emissions sources at fine spatial, temporal, and chemical scales. Adjoint modeling has emerged as an efficient alternative, enabling high-resolution, multi-pollutant source attribution in a single integrated framework; however, its application to simultaneous PM-O pollution episodes is limited, particularly in densely populated regions experiencing complex co-pollutant interactions. Here we apply a newly developed multiphase adjoint of the Community Multiscale Air Quality (CMAQ) model to quantify the emission sensitivities of PM and O concentrations during pollution episodes in major urban agglomerations. Our results indicate that local emissions predominantly drive PM concentrations, contributing up to 79 μg m. In contrast, O episodes are largely initiated by regional transport (3.8-7.3 ppbv), surpassing local emission contributions during episode onset. The sensitivity analyses reveal distinct spatial emission signatures and pollutant-specific influences from critical precursors, including volatile organic compounds (VOCs; up to 15.9 ppbv O, 11.4 μg m PM), nitrogen oxides (NO ; 16.6 ppbv O, 13.8 μg m PM), and ammonia (NH; up to 8.7 μg m PM). This study demonstrates the diagnostic strength and predictive capabilities of adjoint modeling in unraveling complex source-receptor relationships. By offering detailed, pollutant-specific emission sensitivity information, our approach provides a robust foundation for precision-driven emission control strategies and improved cross-regional policy coordination, substantially advancing air quality management frameworks.

The marine biosphere profoundly influences atmospheric chemistry and climate through its carbon cycle. Viruses, the most abundant and diverse entities in marine ecosystems, significantly shape global carbon dynamics by infecting microbes and altering their metabolism. Both DNA and RNA viruses drive these processes in surface oceans, yet their roles in the deep sea-a sunlight-independent ecosystem that stores vast carbon reserves-remain largely unexplored. Here we show that viruses regulate the microbial carbon cycle in the deep-sea biosphere, based on viromic analysis of 66 global sediment samples spanning 1900 to 24,000 years. We identified 324,772 DNA viruses and 61,066 RNA viruses, revealing high diversity and long-term persistence. These viruses co-participate in host carbon metabolism via synergistic genes that encode carbohydrate-active enzymes, with DNA viruses primarily aiding synthesis and RNA viruses supporting decomposition. Integrated virome and microbiome data indicate that viral genes form novel metabolic branches, compensating for host deficiencies and enhancing pathway efficiency in processes like fructose-mannose and pyruvate metabolism. Our findings position deep-sea viruses as key regulators of marine microbial carbon cycling, with implications for global biogeochemical models and climate resilience. This work offers the first holistic perspective on DNA and RNA viruses in deep-sea carbon dynamics, illuminating their ecological significance across geological timescales.

Transboundary hydrological basins span international borders and are essential to global water systems, human development, and environmental sustainability. Nearly 40 % of the world's population lives within these basins, which supply critical resources such as freshwater, food, energy, and biodiversity. Yet their sustainability remains poorly understood, as existing assessments often overlook the unique social, environmental, and political complexities of transboundary basins. This study addresses that gap by developing and applying a systematic framework to assess Sustainable Development Goals (SDGs) progress across 310 transboundary basins worldwide. Here we show that transboundary basins score significantly lower on average SDGs achievement (an SDG Index score of 42 on a scale of 0-100) compared to national averages (a score of 67), with considerable variation between regions. We identify four distinct types of transboundary basins in terms of SDGs achievement and associated challenges. We also show that progress on a specific set of goals can drive broader sustainability within each basin type. Notably, achieving clean water (SDG 6), sustainable economic growth (SDG 8), and healthy livelihoods (SDG 3) is linked to overall SDGs success in 38 % of transboundary basins worldwide. Our results highlight the importance of basin-level analysis for revealing sustainability patterns overlooked by national assessments. This framework can inform future basin research and support policy development in transboundary regions.

Waste activated sludge from wastewater treatment plants poses a major environmental challenge, with its high moisture content complicating disposal and resource recovery processes across global industries. Efficient sludge management requires precise moisture monitoring to optimize treatment methods, reduce costs, and enhance outcomes such as anaerobic digestion and composting. Traditional approaches for moisture measurement are time-intensive and batch-based, while emerging techniques, such as infrared or nuclear magnetic resonance methods, suffer from inaccuracies, high costs, or limitations in real-time applications. Here we show that sludge jet characteristics, reflecting its non-Newtonian fluid properties, can be captured via high-speed imaging and analyzed with deep learning to accurately predict moisture content within 20 s. By developing a laboratory-scale system of instantaneous capturing of activated sludge jet expansion images (iCASJEI), we acquired over 11,000 jet images across 79-94 % moisture ranges and trained convolutional neural networks, with VGG-16 outperforming AlexNet and LeNet under optimized conditions (0.2 MPa pressure, 4 mm aperture), achieving 93.5 % validation accuracy at 2 % precision and 87.6 % at 1 % precision. These findings show that incorporating iCASJEI to extract non-Newtonian fluid characteristics from sludge jets with deep learning algorithms can substantially reduce testing time for sludge moisture content. This approach could also be applicable to other sectors where non-Newtonian fluid characteristics enable real-time moisture detection in viscous liquids.