Environmental Science & Technology Letters Environ.最新文献

China’s East–Data–West–Computingstrategyplaces data centers in cooler climates, but climate alone does notdetermine water savings. This study combines public cooling evidence,hourly weather data for 2015–2024, a physical heat-rejectionmodel, and provincial electricity-water factors in a project-specificbreak-even analysis. An audit of 93 mapped records identifies 12 uniquewestern hub projects and four named or cluster-level eastern alternatives,yielding 48 comparisons. Cooling operation is reconstructed from publicevidence where possible and otherwise represented by explicit climate-basedscenarios. Under the central assumptions, western combined operationalwater intensity is lower in 27 of 48 pairs, with changes ranging from−63.5% to +110.4%. Against Guangdong, matched wet-tower designsand facility energy efficiency yield higher mean western consumptionbecause of electricity-related water use. Low-water configurationscan reverse this penalty, but their advantage depends on the easternalternative, facility energy overhead, and electricity supply. Localwater stress adds a separate constraint on residual campus demand.Westward siting therefore saves operational water only when the destinationproject crosses the threshold set by the eastern alternative it replaces.

Levoglucosan (LEV) has long beenused as a tracer for biomass burningemissions due to its presumed atmospheric stability. However, recentstudies suggest that its degradation rate varies significantly underdifferent environmental conditions. This work investigates the heterogeneousoxidation of LEV by hydroxyl radicals (•OH) and the influenceof initial particle acidity on its degradation. Experiments were conductedin a smog chamber using internally mixed LEV particles with eitherammonium sulfate (AS) or ammonium bisulfate (ABS). The chemical compositionof the particle was monitored using a high-resolution aerosol massspectrometer (HR-AMS) during the experiment. The results show thatlow-initial pH particles accelerate LEV degradation by 7–8times relative to high-initial pH particles, with mean second-orderreaction rate constants of 4.46 × 10^–11 and6.03 × 10^–12 cm³ molecule^–1 s^–1, respectively. Acidic conditions also favorthe formation of highly oxygenated products, some of which likelyfragment and evaporate into the gas phase. These findings imply thatacidic particles significantly reduce LEV’s atmospheric lifetime(from 40 to 3 h) and its reliability as a biomass burning tracer,highlighting the need to incorporate acidity effects into atmosphericmodels.

Biodiversity is central to the functioning, stability,and climaterelevance of marine vegetated ecosystems, yet its role in macroalgalcarbon sequestration remains underappreciated in current climate andbiodiversity frameworks. Macroalgae contribute substantially to coastalprimary production, habitat formation, and nutrient cycling, whilesupporting diverse ecological communities across temperate, tropical,and polar seas. However, their environmental value extends beyondthat of carbon capture alone. In biologically diverse macroalgal systems,modulated biodiversity can also help control toxic materials arisingfrom anthropogenic activities, including CO₂, SOx, andNOx, by integrating these pollutants into existing elemental biogeochemicalcycles and enabling their transformation into more benign and potentiallyuseful chemical forms. In this review, we examine macroalgal biodiversityas a determinant of marine carbon sequestration potential across naturalecosystems, restoration efforts, and cultivated systems. We synthesizecurrent understanding of how taxonomic and functional diversity influencecarbon fixation, biomass stability, trophic transfer, detrital export,nutrient retention, pollutant buffering, and resistance to environmentalstress. We further assess key uncertainties surrounding carbon permanence,ecological trade-offs, measurement, reporting, and verification, andthe risks associated with simplified or monoculture-based deploymentstrategies. We argue that biodiversity should be treated not merelyas a cobenefit of macroalgal systems but as a central design principlefor evaluating their ecological performance, climate relevance, andbroader capacity to restore biogeochemical balance in anthropogenicallystressed environments.

Per- and polyfluoroalkyl substances (PFAS) have beenimplicatedin neurodevelopmental toxicity, yet direct evidence of their presencewithin the central nervous system (CNS) of children remains limited.We quantified 12 PFAS in cerebrospinal fluid (CSF) samples from 47pediatric patients undergoing clinically indicated lumbar puncturein Nanjing, China. PFAS were detected in 100% of samples, with PFOA(detection frequency: 98%), PFHxS (96%), PFOS (89%), and 6:2 Cl-PFESA(68%) most prevalent; their CSF distribution patterns were generallyconsistent with those observed in pediatric serum samples from nearbyregions. Median concentrations were numerically higher in childrenunder 2 and under 6 years of age, but these differences were not statisticallysignificant (p = 0.450 and p = 0.322;rank-biserial r = 0.110 and 0.145) and should beinterpreted as nonsignificant in this small, clinically heterogeneoussample. The presence of PFAS in children’s CSF adds to thegrowing concerns about early-life neurotoxicant exposure. Importantly,this study involved a small cohort (n = 47) of criticallyill children. Given the limited sample size, the underpowered agesubgroups, and the potential influence of disease-related changesin barrier permeability, the measured concentrations should not begeneralized to healthy children, and all age-related comparisons areexploratory.

Interdisciplinary, international research collaborationsare essentialto expanding resilient global mineral supply chains sustainably andresponsibly. We identify key research needs to achieve these objectivesthat address mining water consumption and contamination through tailingmanagement and upstream process innovations. We also explore meansof reducing solid waste from mines by repurposing it as a constructionmaterial and recovering minerals from aqueous waste streams. Introducingnew technologies that improve water efficiency and reduce waste generationwill bring trade-offs, including potentially increasing energy consumption.Accordingly, systems analysis is an essential tool for exploring thesetrade-offs and identifying innovations that can minimize them. Thesetechnologies will potentially influence how communities adjacent tomines experience social and environmental effects. Developing andimplementing sensor networks to support community-based monitoringof the changes that adoption of these technologies will bring cansupport community-mine dialogue. Finally, energy system models canexplore the potential regional or global effects of these changeson overall mining sector environmental effects. Innovation in themining sector therefore requires collaboration among engineers, scientists,social scientists, policy experts, and others across academia, industry,governments, and nongovernmental organizations. Developing a secure,resilient, mineral supply chain that maintains or restores environmentalquality while meeting global mineral demand requires supporting andsustaining interdisciplinary collaborations among nations.

Effective real-time control of wastewater treatment plants(WWTPs)remains challenging due to highly variable influent conditions, whileregulation of external carbon sources (ECS) is critical for sustainingnitrogen removal and meeting increasingly stringent effluent standards.This study developed a systematic framework integrating online learning(OL) and reinforcement learning (RL) to optimize effluent qualityand ECS dosing in a full-scale WWTP. Based on one year of high-frequencysensor data, Transformer (TF) was identified as the best-performingmodel among six machine learning methods, with a mean R² of 0.91 across multiple effluent indicators. SHAP was employed toidentify key features, which informed the design of a key-indicator-basedtime-decayed online learning (KTOL) strategy, enabling the Transformermodel to maintain stable predictive performance under dynamic operatingconditions. Building on this, RL enabled real-time ECS regulation.During a 10-day practical test, KTOL-TF-RL outperformed conventionalfeedback control by improving effluent compliance and reducing ECSusage by 21.8%. Monte Carlo simulations demonstrated robustness againstsensor distortion and malfunction. Overall, the KTOL-TF-RL frameworkdemonstrates the potential of intelligent control as a promising pathwayfor achieving cost-effective and resilient wastewater treatment.

Hourlyair-pollution fields are needed for exposure assessmentand air-quality management; however, ground monitors are sparse, andsatellite observations are intermittent. Data-driven models may reproducestation observations while leaving unmonitored regions poorly constrained.This mismatch raises a central concern for observation-free hourlymapping: station-level skill may not imply a physically reliable pollutionfield. To test this, we trained a neural operator to predict near-surfaceCO over Mainland China from emissions, meteorology, and terrain, usingground observations only for training. Transport regularization wasimposed through a full-grid advection-diffusion-reaction residualthat balances emissions, transport, diffusion, and decay. On the unseenyear 2022, this constraint increased hourly station agreement fromR = 0.57 to 0.64 and structural agreement with an assimilation-informedreanalysis from R = 0.55 to 0.74. In a blind-region test, where localmonitors were removed from training, R increased from 0.51 to 0.58.Transport regularization therefore reduces nonphysical extrapolationand enables more stable hourly pollution-field reconstruction withoutobservations at prediction time.

Ion partitioning is a key step influencing ion selectivityin nanoporousmembranes, yet its thermodynamic origin remains insufficiently understoodbecause Donnan and non-Donnan effects are often coupled. Here, wedevelop a thermodynamic framework to operationally decompose and quantifynon-Donnan and Donnan contributions to ion partitioning for both co-ionsand counterions. Using quartz crystal microbalance measurements combinedwith temperature-dependent analysis, we estimate non-Donnan partitioncoefficients and associated enthalpic and entropic changes. Non-Donnanpartitioning is closely associated with ion dehydration and ion–membraneinteractions and exhibits enthalpy–entropy compensation withadditional ion–membrane interactions for divalent ions. Wefurther evaluate the thermodynamic contribution of Donnan partitioningand show that electrostatic interactions affect co-ions and counterionsdifferently. We infer that those electrostatic interactions may beassociated with cation dehydration or hydration-shell reorganization.This work provides a thermodynamic basis for understanding ion-specificselectivity in polyamide membranes.

Reliable quantification of neutralper- and polyfluoroalkylsubstances(nPFAS) via headspace analysis remains hindered by unrecognized analyticalbiases. We reveal a severe, previously unquantified artifact arisingfrom sorptive interaction between nPFAS and commonly used polytetrafluoroethylene(PTFE)/silicone septa, resulting in time-dependent signal loss exceeding90% within 12 h. Incorporating septum sorption into the headspaceequilibrium model shows that this artifact systematically distortsair–water partition coefficients (K_aw), artificially masking true nPFAS volatility and compromising environmentalfate modeling. A simple aluminum-foil barrier successfully eliminatesseptum artifacts, restoring analyte stability and enabling reliable K_aw determination. These findings reveal a criticalbut underappreciated source of systematic error in headspace workflows,highlighting the need to account for sorptive artifacts to understandnPFAS environmental behavior.