Earth and Environmental Sciences

Abstract. Carbon dioxide removal (CDR) strategies such as ocean alkalinity enhancement (OAE) are likely required in addition to rapid emissions reductions to limit global warming to well below 2 °C. However, the long-term efficiency of OAE and its potential to mitigate climate change and ocean acidification remain uncertain. Here, we investigate efficiencies, climate and ocean acidification responses of idealized OAE using a fully coupled, emission-driven Earth system model across three global warming stabilization scenarios (1.5, 2, and 3 °C) spanning 1861–2500. OAE is implemented as a continuous global surface alkalinity addition of 0.14 Pmol yr−1 following the CDRMIP protocol from 2026 onward. Our results show that OAE reduces atmospheric CO2 by 73–130 ppm by 2500, with larger reductions under higher warming scenarios and during the first 100 to 200 years of alkalinity addition. In contrast, global surface air temperature decreases nearly linearly by 0.14–0.17 °C per century across all scenarios, indicating that the cooling rate due to OAE is largely insensitive to the emission pathway and background warming level. The interpretation of OAE efficiency depends strongly on the chosen metric. The global gross ocean carbon capture efficiency of about 0.79 remains close to the theoretical maximum, reflecting the negative emissions through OAE, whereas the net ocean capture and atmospheric CO2 reduction efficiencies are substantially lower and decline over time due to carbon cycle feedbacks in response to lowered atmospheric CO2. OAE mitigates ocean acidification, at the surface as well as in the interior ocean, with most centennial-scale mitigation arising from atmospheric CO2 drawdown, an effect shared with other CDR approaches. Direct chemical effects of added alkalinity contribute transiently and diminish over time as the ocean–atmosphere system equilibrates. Overall, our results underscore that rapid emission reductions remain the most effective strategy for achieving the Paris Agreement goals and mitigating ocean acidification.

The Fundão dam collapse in Mariana (Southeast Brazil) released metal-rich tailings into the Doce River basin and its tributaries, including the Gualaxo do Norte River. Metal concentrations in aquatic environments vary seasonally, generally increasing during the rainy season. Exposure to metals may lead to behavioural alterations and changes in acetylcholinesterase (AChE) activity in fish. This study investigated the effects of water from the Gualaxo do Norte River (collected during the dry and rainy seasons) on the behaviour and cholinergic activity of Astyanax lacustris . One hundred fish were assigned to four treatments: dechlorinated water during the dry (CDS) and rainy (CRS) seasons, and river water from the dry (DS) and rainy (RS) seasons. Fish were exposed to the treatments for 15 days, and behaviours were recorded for 10 days. After that, brain and muscle AChE activity was analysed in the fish. Metals such as Al, Cu, Fe, Hg, Mn, Pb, and Zn showed higher concentrations in the water collected during the rainy season. Fish from the DS group showed increased activity, whereas those from the RS group showed reduced activity, reduced sociability, and lower aggressiveness. Muscular AChE activity increased in DS and RS groups relative to CDS and CRS. In contrast, cerebral AChE activity was higher in DS (compared to CDS and RS) and lower in RS (compared to CRS). Our results indicate that water collected during the rainy season may pose a greater threat to fish health because of increased metal runoff. Even a decade after the dam rupture, residual contamination continues to affect fish physiology and behaviour, emphasising the persistent ecological risks and the urgent need for long-term monitoring and management of impacted aquatic ecosystems. • Water collected in the rainy season (RS) was richer in metal concentrations. • Fish from the dry season group (DS) were more active. • Fish from the RS group were motionless, less sociable and less aggressive. • AChE activity was higher in both the brain and the muscle in the DS group. • AChE activity was lower in the brain and higher in the muscle in the RS group.

Arsenic contamination of environmental systems has aroused a great deal of public interest due to its toxicological properties. This study investigates the efficiency of three novel approaches including: iron-assisted electrodeposition, lime, and adsorption using biochar and chitosan-based columns for arsenic removal from highly contaminated effluents generated from pilot soil washing tests of a contaminated mine site with 32,000 mg/kg arsenic in Northwest Territories, Canada. In a first-of-its-kind approach, Fe-assisted electrodeposition achieved 99% arsenic removal from soil washing effluents within 10 h using iron(III) sulfate, concurrently yielding FeAs solids on the cathode as a potential value-added byproduct. Similarly, application of lime resulted in 99% arsenic removal with precipitation of calcium arsenates. Amino-functionalized magnetic biochar exhibited maximum adsorption capacities of 26.74 mg g −1 for As(III) and 11.31 mg g −1 for As(V). A layered column of magnetic biochar, zeolite, activated carbon, and a chitosan-enhanced sand filtration system achieved almost complete arsenic removal (99%). The results of column tests demonstrated highly effective arsenic removal, alongside substantial co-removal of other toxic metals including lead, zinc, antimony, cadmium, uranium, and cobalt by up to 99% from soil washing effluents, underscoring the practical applicability of the proposed methods in real-world remediation scenarios. The combined techniques used in this study can be employed as a novel, eco-friendly, and cost-effective solution for arsenic remediation in wastewater generated from soil washing systems in mining-impacted environments. • Fe assisted-electrodeposition removed >99% As. • Fe As solids formed via iron-assisted electrodeposition. • Amino-functionalized magnetic biochar enhances As adsorption. • Chitosan-based columns reduced As and metals by up to 99%. • Lime applications dramatically reduced As and metals.

• A Novel Multi-Source Attention-Enhanced Transformer U-Net is proposed for SSS retrieval. • It can achieve accurate SSS retrieval without reliance on in-situ observations or salinity background fields. • The proposed method significantly outperformed the official SMAP 8-day SSS product without data assimilation. • It demonstrated robust retrieval performance across seasonal cycles and diverse spatial domains. Accurate mapping of sea surface salinity (SSS) is essential for diagnosing the marine hydrological cycle and enhancing coupled atmosphere–ocean forecasts. However, current L-band satellite products remain compromised by radio-frequency interference, land-sea contamination and imperfect roughness corrections. Therefore, this study developed a multi-source attention-enhanced Transformer U-Net (MSAtt-TransUNet), which synergistically exploits Soil Moisture Active Passive (SMAP) L-band brightness temperature (TB) and auxiliary environmental factors, adopting a physically guided loss function that preserves the salinity front to retrieve SSS in the dynamically complex Northwest Pacific. MSAtt-TransUNet reduced Root Mean Square Error (RMSE) by 12.6% relative to the baseline U-Net. Compared with in situ Argo data, the proposed model attained an RMSE of 0.235 psu and a Pearson correlation coefficient (R) of 0.844, markedly outperforming the operational SMAP Level-3 product (0.338 psu and 0.728). Validation against the BOA_Argo gridded dataset further confirms that MSAtt-TransUNet had better performance than SMAP official, while remaining robust across spatio-temporal domains. Notably, spatial diagnostics against the Hybrid Coordinate Ocean Model (HYCOM) analysis indicate that MSAtt-TransUNet effectively reconstructed salinity gradients across the Northwest Pacific. Shapley additive explanation and channel ablation experiments assigned the highest importance to dual-polarized TB, followed by wind speed and sea surface temperature. The model achieved these gains without assimilating any in situ measurements or model priors, offering a highly accurate, robust, and interpretable alternative for near real-time SSS monitoring and regional data-assimilation applications.

Partitioning soil organic carbon (SOC) into particulate (POC) and mineral-associated (MAOC) forms is fundamental to understanding carbon sequestration. Yet, how these pools change with depth and respond to climate in semiarid forests remains a critical uncertainty. Our study investigated 35 forest sites across a climatic gradient in northern China to reveal the vertical distribution (0-10 cm vs 40-60 cm) and climate sensitivity of POC and MAOC. We discovered a fundamental depth-dependent shift in both pool interactions and climate sensitivity. While POC and MAOC were strongly coupled in topsoil, they became decoupled in the subsoil. Crucially, subsoil MAOC exhibited significant negative correlations with increased temperature and aridity, a vulnerability absent in the topsoil. Structural equation modeling identified the mechanism behind this pattern: climate indirectly influences topsoil MAOC via POC and mineral protection, but transitions to a direct driver in the subsoil following carbon pool decoupling. This shift from indirect to direct control explains the heightened climate sensitivity of deep MAOC. This finding underscores that the stability of deep carbon is precarious, revealing a significant vulnerability of carbon stocks in semiarid forests to climate change. Our results provide a mechanistic basis for integrating these depth-specific dynamics into Earth system models, which is essential for accurate climate projection.

• Landsat time series reveal 33-year forest dynamics in Wuyishan National Park. • Recovery dominated post-disturbance; tea expansion drove anthropogenic pressure. • Time-series analysis confirms the 2015 National Park Pilot curbed forest loss. • Disturbance trajectories aligned with policy, economics, and climate extremes. Navigating complex human-nature interactions remains a critical challenge for protected areas globally. China’s establishment of the National Park System represents a fundamental governance shift toward ecological integrity, yet its long-term efficacy in economically vibrant regions remains underexplored. Focusing on Wuyishan National Park, we reconstructed forest disturbance dynamics from 1991 to 2023 using Landsat time series and the LandTrendr algorithm. Over this period, an estimated 6.0% of the park area (approximately 7,739.4 ha) experienced disturbances. Quantitative analysis revealed that the majority (58.8%) of disturbed forests eventually recovered, whereas tea plantation expansion (23.0%) exerted the primary permanent anthropogenic pressure, clustering selectively in non-protected Gaps and along administrative boundaries, while the Core Protected Area remained safeguarded. Disturbance trajectories aligned with policy evolution, economic incentives, and climatic extremes. Following the 2015 National Park Pilot, disturbances stabilized at minimal levels, coinciding with the effective containment of historical tea expansion. This study establishes a transferable remote sensing framework for evaluating conservation outcomes, highlighting the utility of continuous spatial monitoring in assessing protected areas confronting complex socio-economic and climatic pressures.

Most communities on carbonate islands face increasing freshwater scarcity because island hydrologic characteristics raise their vulnerability to variations in rainfall, sea level rise, and groundwater extractions. Engineered options for freshwater production can be challenging and expensive, leading many island communities to rely on groundwater for potable water supplies. Consequently, freshwater recharge magnitudes and rates provide critical information for accurate assessments of water resources. Recharge is commonly estimated with water balance models based on low frequency (monthly) precipitation (P) and evapotranspiration (ET) values, but such coarse analyses may miss short-term recharge (e.g., stochastic storm events) in high permeability carbonate island settings. To test the efficacy of water budget techniques and gain new insight on recharge dynamics in carbonate islands, we develop a recharge model framework that requires only high temporal frequency (15 min) groundwater level data to determine aquifer permeability, storage, and daily recharge. We apply the method on San Salvador Island, The Bahamas using 14-months of groundwater level data. We validate our aquifer characteristic assessments with high spatial resolution permeability and porosity measurements from two separate cores. Over the 14-month study period, recharge totaled 428 mm, or 37% of total precipitation, which is ∼2.5–3 times the monthly water budget estimate of ∼164 mm recharge or 12% of precipitation. Most recharge occurs within a day of precipitation, limiting the effectiveness of low-resolution estimates and emphasizing the need for high-resolution recharge for water management. Our recharge analysis also shows a power law relationship between increasing recharge fraction with decreasing precipitation amounts, highlighting limitations of using a single value of the fraction of precipitation that becomes recharge. A sensitivity analysis on the new model framework shows that recharge basin length is the primary source of uncertainty, but this uncertainty may be reduced with data available on small carbonate islands. Our results provide guidance on collecting data to estimate recharge in carbonate island settings and our model framework can improve freshwater management practices on carbonate island settings using easy to deploy and maintain water level loggers. • We have developed a new recharge framework to be used in data poor settings. • We analyzed high frequency groundwater level data to analyze recharge dynamics. • In carbonate islands, water budget techniques may produce erroneous recharge. • Our method can be applied to other remote settings to explore water budgets.

• Multi-disciplinary chronosequence analysis reveal synchronous physical–biological change. • Hillslope Development Index (HDI) quantifies long-term landscape evolution. • Key proxies enable efficient assessment of landscape maturity across timescales. Glacial retreat, volcanic eruptions, and erosion by landslides, rivers, and humans create barren landscapes where soils and vegetation develop over decadal to millennial timescales. Previous studies have documented the development of individual physical, chemical, and biological parameters in post-disturbance landscapes. However, the rates governing the co-evolution of soils, vegetation, and biogeochemical cycling remain obscure. Here, we present a multi-disciplinary analysis of 39 variables that reflect the storage and release of energy and matter on chronosequences of morainal hillslopes ranging in age from 30 to 14,000 years in two glacier forelands. We find a remarkable similarity in the timescales of physical and biological changes on these moraines. Based on this finding, we propose a ‘hillslope development index’ (HDI) that represents and quantifies the integrated hydrological, biogeochemical, and ecological state of development. In our glacier forelands, the HDI increases rapidly in the first 1000 years and then starts to asymptote, achieving a quasi-steady state at 5000 years. We find that a few key variables closely follow the HDI. These proxies allow us to assess the integrated hydrological, biogeochemical, and ecological state of hillslope development with limited experimental effort.

• Alternative practices enhance productivity and extend carbon uptake. • Carbon uptake period drives productivity consistently across sites. • Precipitation extends carbon uptake mainly under alternative management. Industrialized agriculture has dramatically increased crop yields over the last century, but this has led to soil erosion and reductions in soil organic matter, raising concerns about the long-term sustainability of agroecosystems under weather variability. The first step towards restoring soil organic matter in agroecosystems is achieving sustained net ecosystem carbon uptake, as reflected in positive net ecosystem productivity (NEP). Thus, we evaluated how prevailing and potentially more sustainable, alternative management practices influenced NEP across a range of agroecosystems in the United States. Specifically, we analyzed eddy covariance data from eleven Long-Term Agroecosystem Research cropland locations to compare gross primary productivity (GPP) and NEP dynamics in the two systems. At seven locations spanning diverse environments, mean annual NEP was higher under the alternative management. Five of these sites with cover crops exhibited longer carbon uptake period (CUP) than their prevailing management counterparts. Importantly, a longer CUP was associated with increases in both annual sum of GPP and NEP and their peak values. Increased precipitation at the alternative sites enhanced CUP length, which contributed to higher overall productivity while prevailing systems showed no sensitivity to precipitation. The relationship between annual NEP and CUP showed steeper slopes at drier sites, indicating potential benefits for NEP gains associated with extended CUP. Our findings demonstrate that sustainable management practices can strengthen carbon uptake capacity, highlighting their importance for maintaining agroecosystem productivity.

The Atlanto-Iberian continental shelf, located in the northern part of the Canary Current upwelling system (North Atlantic Ocean), is a highly productive and dynamic region (spatially and temporally), supporting a high diversity of marine organisms. However, current information on fish larval assemblages is absent. This study provides the first large-scale assessment of the composition and spatial distribution of ichthyoplankton communities along the western and southern Iberian coasts in spring (April/May). Fish larvae assemblages were compared between four regions with different oceanographic conditions: Northwest Iberia (NW), Southwestern Iberia (SW), Western Gulf of Cadiz (S) and Eastern Gulf of Cadiz (Cad). A total of 57 fish species and 12 genera, belonging to 26 families were recorded, with Alosidae, Gobiidae, Callionymidae, Sparidae, and Blenniidae being the most abundant families. The communities were dominated by small pelagic species ( Sardina pilchardus and Engraulis encrasicolus ) . A latitudinal gradient in community structure was observed, broadly reflecting the environmental temperature gradient. The eastern Gulf of Cadiz (Cad) presented a different larval community structure and the highest abundance of fish larvae. These patterns suggest that local spring conditions of productivity, temperature and retention favour fish reproduction, as evidenced by the high abundance of Engraulis encrasicolus , Sardina pilchardus and Diplodus sp.. Temperature, salinity and depth were identified as the main environmental drivers shaping larval community structure. This study provides valuable baseline information for future research to underpin biodiversity conservation and fisheries management efforts, especially in biogeographic transition zones where spawning patterns and population dynamics are likely to shift under climate change scenarios.

• A novel analytical approach to unpacking multiple senses of place in industrial decarbonisation. • Complex and ambivalent senses of place could exist in industrial towns other than pride in industry. • Divergent place visions between cluster stakeholders and host communities stem from different ontologies of industrial places. • Growing industry-community disconnection strengthens the need to go beyond job impacts for just transitions. The climate crisis and goals for ‘green growth’ have led governments worldwide to adopt cluster-based approaches to decarbonise carbon-intensive industries (e.g. steel, petrochemicals) and to revive declining industrial regions, a predominant concern for late capitalist societies. Research has begun to investigate the role played by sense of place in sustainability transitions in regions with fossil fuel-intensive industries. However, ways that sense of place may shape just industrial transitions have been underexplored. Adopting a novel analytical approach with the concept of spatial imaginaries, we explore contested senses of place in decarbonising industrial regions and their implications to just transitions. A multi-phase, multi-scalar qualitative methodology was used to investigate ‘net zero senses of place’ held by government and industry stakeholders and host communities in three emerging UK net zero industrial clusters. Our findings reveal an ontological contrast of industrial places existing between stakeholders and host communities. There was a gulf between the top-down, win–win, and hopeful place visions in policy/industry discourses (e.g. the idea of ‘SuperPlaces’) and the more ambivalent socio-cultural imaginaries of industrial places held by host communities, comprising bleak lived experiences of extractive place-industry relations alongside positive examples of community-led resilience. The contrasting senses of place underlay local scepticism about proclaimed benefits and visions of industrial decarbonisation, indicating the stakeholders’ deficiency in recognising local grievance and struggles, in engaging community in cluster development, and in supporting community wellbeing. We conclude by arguing that stakeholders must go beyond a narrow techno-economic ontology to view industrial decarbonisation as a form of inclusive place-making, working with host communities to address local needs and co-produce shared visions of net zero futures.