Atmospheric and Oceanic Sciences

Study region Vilaine River basin, France. Study focus Phosphorus (P) concentrations have declined in Western countries since the 1970s due to reductions in point and diffuse sources, but recent concerns about a potential P return have emerged. We analyzed long-term trends in soluble reactive P (SRP) and total P (TP) across 18 subcatchments of a 10,137 km² temperate basin in western France. A discharge-normalization approach was used to separate the effects of hydrological variability from changes in P sources or forms. We also examined synchronies with chemical proxies and relationships with geographic variables to support attribution of observed trends. New hydrological insights From 2007–2024, 61% of subcatchments showed increasing SRP concentrations (0.3 – 7.7 μg P/L/year), whereas only 28% showed increasing TP (1.1 – 12.1 μg P/L/year). Because increases occurred mainly during baseflow conditions, trends in annual P loads were often not significant. Changes in discharge explained less than 10% of SRP trends. Instead, results point to remobilization of legacy P, associated with rising temperatures and decreasing dissolved oxygen, as the main driver. Remobilization occurs in both the surface water network (streams, lakes, ponds) and artificial systems such as riparian buffer strips and waste stabilization ponds, acting as sources of P to the stream network. Future work should identify where remobilization mainly occurs to improve targeted mitigation measures.

Introduction Satellite-based deforestation monitoring is critical for environmental sustainability and climate change mitigation. However, conventional remote sensing approaches face significant limitations in cloud-covered regions and are often inadequate for capturing temporal change dynamics, constraining their effectiveness in continuous forest surveillance. Methods We propose the Siamese Attention U-Net with Multimodal Temporal Fusion (SAU-MTF), a novel deep learning architecture that integrates optical (Sentinel-2) and Synthetic Aperture Radar (Sentinel-1) imagery within a tri-temporal framework. The model employs EfficientNet-based encoders and attention gates for discriminative feature extraction, alongside temporal context blocks designed to capture change dynamics across time steps. A multimodal fusion strategy is adopted to exploit the complementary strengths of SAR’s cloud-penetrating capability and the spectral richness of optical data. Results Evaluated on large-scale deforestation datasets, SAU-MTF achieves a classification accuracy of 94.7% and an Intersection over Union (IoU) of 0.93, outperforming existing state-of-the-art models across benchmark comparisons. Discussion These results demonstrate that the joint exploitation of temporal, spectral, and spatial information substantially enhances deforestation detection performance. The architecture proves especially effective in challenging environments characterized by persistent cloud cover and seasonal variability in remote forest regions, highlighting its potential for operational deployment in global forest monitoring systems.

Flooding compromises the microbiological quality of domestic and recreational water sources, increasing exposure to waterborne and opportunistic pathogens in affected communities. Following widespread flooding triggered by the September 2023 controlled spillage of the Akosombo and Kpong dams in Ghana, which led to massive flooding, we assessed the microbial quality of water bodies in communities downstream of the Volta River and associated public and ecosystem health implications. In July 2024, water was collected from multiple source types across four downstream townships (Asutsuare, Aveyime, Battor, and Mepe), including river sections, ponds, puddles, wells, boreholes, and canal-associated sites. Samples were filtered, microbial DNA was extracted, and bacterial community composition was profiled using 16S rRNA gene sequencing, with downstream bioinformatics analyses used to characterize microbial diversity, assess spatial variation across sites, and stratify detected taxa according to ecological origin and public health risk. Sequence data obtained from 17 out of 22 samples revealed communities dominated by Proteobacteria with frequent co-occurrence of Bacteroidetes and Firmicutes. Across all sites, 56 microbial taxa and 118 microorganisms were detected. Community composition varied by location and water source type, with bank and mid-river sites generally exhibiting higher richness than puddles and a canal drain. Thirteen samples contained site-specific taxa, indicating marked spatial heterogeneity within and between communities. Microorganisms of public health relevance were frequently detected, including enteric-associated bacteria such as Escherichia coli, Salmonella enterica, Shigella species, and Cronobacter sakazakii, as well as opportunistic and healthcare-associated bacteria including Acinetobacter baumannii, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Stenotrophomonas maltophilia. Higher-risk organisms were most widely distributed in Mepe and Battor, including in groundwater-associated sources. These findings provide molecular evidence of persistent, spatially heterogeneous microbial hazards months after flooding, underscoring continued exposure risks in downstream communities and the need for targeted water safety interventions and risk-informed public health responses.

Mining can significantly affect the spatial distribution and temporal evolution of groundwater chemistry. From July to August 2024, the research team collected 26 surface water and groundwater samples in the Shengli Coal Mine area of the Mongolian Plateau, conducting comprehensive hydrogeochemical analyses on surface water flowing through the mining area, groundwater within the mining area, seepage water, and groundwater outside the mining area. The results indicate that groundwater in this region is notably affected by human activities such as mining operations. Specifically, in surface water flowing through the mining area, concentrations of total dissolved solids (TDS), sulfates, nitrates, and nickel showed significant increases. Compared to groundwater systems in other areas of the Mongolian Plateau, nickel levels in the mining area’s groundwater were significantly higher, while nitrate levels exhibited the opposite trend. A significant positive correlation was observed between metal element concentrations in surface water and groundwater. The study found that abnormal distributions of heavy metals such as beryllium (Be), thallium (Tl), and tin (Sn) may originate from point-source pollution caused by mining activities. Furthermore, concentrations of manganese (Mn), arsenic (As), and antimony (Sb) in the groundwater of this area exceeded relevant regulatory limits, with arsenic being particularly prominent. The levels of arsenic in both surface water and groundwater may pose carcinogenic risks to human health. This study shows that nearly half of the sampled water bodies in the area require purification treatment to meet drinking water standards, highlighting the urgent need for further attention to water quality safety issues. The conclusions derived from this research provide theoretical support for understanding the long-term evolutionary mechanisms of groundwater in mining areas, while also offering important insights for improving groundwater environmental management and ensuring water resource security in mining regions.

Signal saturation in optical and synthetic aperture radar (SAR) data remains a critical challenge for accurate aboveground biomass (AGB) estimation in high-biomass forests. To address this, we developed a multi-source fusion framework integrating Landsat-8, Sentinel-2A, and ALOS-2 PALSAR data over the artificial coniferous forests of the Saihanba Forest in Hebei Province, China. The framework synergistically combines horizontal structure information from optical data with vertical structure information from SAR data, together with environmental factors. It systematically integrates three variable selection methods [Pearson correlation, Random Forest importance, and the least absolute shrinkage and selection operator (LASSO)] and three machine learning models (Random Forest, Support Vector Regression, Artificial Neural Network), evaluated under multiple cross-validation (CV) strategies (leave-one-out, repeated 10-fold, and repeated 5-fold). A novel method for quantifying the saturation point of combined variables is also introduced. Results demonstrate that horizontal structure explains AGB significantly better than vertical structure, and the proposed composite structural index simultaneously elevates the saturation point and improves estimation accuracy. The LASSO-ANN combination consistently achieves high performance across all three CV strategies (mean RMSE = 12.87 t/ha, MAE = 9.28 t/ha), attaining optimal accuracy under leave-one-out CV (RMSE = 5.14 t/ha, MAE = 2.44 t/ha) with a saturation point of 218.54 t/ha, effectively mitigating the saturation issue in medium-to-high-density coniferous forests. Coordinated optimization of feature selection, model architecture, validation strategy, and response variable is essential for improving AGB estimation accuracy and delaying saturation. The proposed framework provides a robust pathway for regional-scale forest biomass monitoring and carbon stock assessment, advancing the integration of multi-source remote sensing in carbon cycle research.

For small-satellite imagery, traditional image points derived from the direct back-projection of laser altimetry points onto the imagery often exhibit deviations of tens of pixels or more, owing to the relatively limited attitude and orbit determination accuracy of small-satellite platforms. Moreover, the diversity of imaging sensors, variations in image resolution, and inherently weak image geometric configurations further complicate the accurate acquisition of image-space coordinates for laser altimetry points. To facilitate the application of laser altimetry data for geometric positioning across multi-satellite, multi-sensor, and multi-resolution small-satellite imagery, this study proposes a measurement method for laser altimetry points tailored to small-satellite images and establishes a combined geometric positioning model that integrates virtual control points, laser altimetry points, and image-matching tie points. The framework comprises four key procedural components: (1) an image-point-measurement strategy for laser altimetry points; (2) the construction of a laser altimetry data-assisted geometric positioning model for small-satellite imagery; (3) the solution of the geometric positioning model using a total least squares approach based on the partial-EIV (errors-in-variables) models; and (4) a comprehensive accuracy assessment conducted under multiple image-combination scenarios, including single-satellite single-stereo, single-satellite multi-stereo, dual-satellite single-stereo, and multi-satellite multi-stereo imagery configurations. Experimental validation is carried out using Jilin-1 small-satellite panchromatic images (KF01A, GF02A, and GF02B) acquired over the Henan region of China. The experimental results demonstrate that, with the laser altimetry point-measurement method and the combined geometric positioning model, the vertical positioning accuracy is substantially improved across all tested image-combination scenarios. These findings further confirm the capability in enhancing the vertical geometric positioning performance of stereoscopic small-satellite imagery characterized by multi-satellite platforms, multi-sensors, and multi-resolutions over terrain conditions similar to those tested.

As an artificial–natural composite ecosystem, polders face increasingly prominent internal and surrounding water environment problems under sustained strong human activities. To systematically assess the water ecological health status of polders around Dongting Lake, this study took four typical polders around Dongting Lake as research subjects, and two seasonal surveys of benthic macroinvertebrates were carried out in July and December 2024. Through analyses of community structure characteristics and correlations with key environmental factors, the Trophic Level Index (TLI), Water Quality index (WQI), Shannon–Wiener index of benthic macroinvertebrates, and Benthic Macroinvertebrates Index of biotic integrity (B-IBI) were selected to establish an entropy-weighted Bayesian model for comprehensive evaluation of the current ecological health of Dongting Lake polders. In 2024, a total of 47 benthic macroinvertebrate species were found in Dongting Lake polders, including 25 species in the summer and 36 species in the winter. The main groups were Gastropoda, Insecta, and Bivalvia, and the community composition did not show significant seasonal or spatial differences. The benthic macroinvertebrate community structure in the summer was mainly influenced by TN/TP, NO3−-N, WD, Tur, and Chl.a, suggesting eutrophication as a critical concern, while in the winter it was mainly regulated by WSW, WT, Tur, and DO, highlighting hydrological conditions as pivotal. The entropy-weighted Bayesian assessment indicated an overall aquatic ecological health status of “moderate” in 2024, with summer conditions superior to those in the winter and notable spatial heterogeneity observed in winter sampling sites within ditches, among which Polder Chengxi exhibited the best condition, while Polder Junshan and Polder Xiangbin Nanhu showed degraded states. Notably, Polder Junshan (JS2 and JS3) displayed clear signs of ecological degradation during the winter, warranting immediate initiation of targeted restoration measures. This study provides a systematic diagnosis and scientifically grounded evaluation of aquatic ecological health in Dongting Lake polders, offering a robust theoretical framework and empirical data for future water environmental protection strategies and ecological restoration practices.

Abstract. The aroma compound ethyl butyrate (EB) and its methylated derivatives ethyl 2-methylbutyrate (EM), ethyl isovalerate (EI), and isopropyl butyrate (IB) are present in many consumer products. To evaluate the environmental and health impacts of these volatile organic compounds, a detailed understanding of their gas-phase photochemical reactivity is required. Here, we performed pulsed laser photolysis/laser-induced fluorescence (PLP-LIF) experiments to investigate the kinetics of their reactions with the hydroxyl radical (OH). Room temperature rate coefficients in units of 10-12molec.-1cm3s-1 with 2σ statistical errors were determined as: (5.5±0.2) for EB + OH, (7.0±0.3) for EM + OH, (11.2±0.4) for EI + OH, and (7.5±0.4) for IB + OH. All four reactions exhibited complex kinetics with distinct non-Arrhenius behaviour for temperatures up to about 400 K. This behaviour was attributed to pre-reaction complexes and is consistent with site-specific reactivities as predicted by an established structure-activity-relationship (SAR). In a second series of experiments, quasi-gas-phase UV-vis. spectroscopy and time-dependent density functional theory predictions were used to obtain absorption cross-sections. All four esters displayed an absorption band at around 213 nm (spin-forbidden π*←n transition), but did not absorb appreciably in the visible or UV-A part of the spectrum where light is abundant at ground level. Therefore, the reaction with OH was considered the main loss process, with lifetimes for tropospheric removal ranging from 22–45 h. Photochemical ozone creation potentials were estimated to be in a moderate range between 28 and 34.

The Kalamaili area in the Eastern Junggar, part of the Central Asian Orogenic Belt (CAOB), is a key locality for understanding the closure of the Paleo-Asian Ocean. This study focuses on the widely distributed Jiangbasitao Formation volcanic rocks in this region, which constitute a typical bimodal suite composed of alkali basalts and rhyolitic tuffs. Whole-rock geochemical analyses reveal that the alkali basalt exhibits a shoshonitic affinity and high TiO 2 contents, suggesting derivation from low-degree partial melting of an enriched lithospheric mantle source. In contrast, the rhyolitic tuff retains arc-like signatures, such as Nb-Ta depletions, indicating an origin primarily through crustal melting or fractional crystallization of mafic magmas. An integrated assessment of the geochemical characteristics demonstrates that this bimodal volcanic suite formed in a post-collisional extensional setting. Zircon U-Pb geochronological dating of a basaltic trachyandesite sample yields a weighted mean age of 322.8 ± 2 Ma (Serpukhovian, Early Carboniferous). This age provides a precise chronological constraint on the final closure of the Kalamaili oceanic basin. Our findings reveal a coherent tectonic evolutionary sequence: northward subduction of the Kalamaili oceanic basin beneath the Yemaquan block occurred from the Devonian to the Early Carboniferous, with oceanic closure culminating at approximately 321.9 Ma, followed by the onset of post-collisional extension. By integrating our new data with existing regional geological constraints (e.g., the ages of post-collisional granites and sedimentary sequences), this study refines the tectonic evolution of the Kalamaili region. The ca. 321.9 ± 2 Ma bimodal volcanism provides robust evidence for the initiation of post-collisional extension in the late Early Carboniferous, thereby supporting a model in which the Kalamaili Ocean had already closed prior to this time. This integrated framework offers new insights into the Paleozoic tectonic evolution of the Eastern Junggar and the CAOB.

A new study shows the pros and cons of different model training methods.

When nanoparticles are made from different materials, the precise control over the mixing process determines the number of hetero-contacts between the components which is one of the key characteristics that determines the quality of the final product. In this study, we develop a model in the context of a population balance modelling approach that captures the statistics of hetero-contact formation in a two-component system as a function of agglomerate size, agglomerate composition, space and time. The two components are characterized by varying primary particle diameters and current simulations include diameter size ratios between 1 and 6. Langevin Dynamics simulations, where every single primary particle is tracked, aid the model development and provide a database for validation. The new method accurately predicts both, the total number of hetero-contacts in the system and the size-specific average number of hetero-contacts. The model is valid in the entire transition regime from the free molecular to the continuum regime, and correctly captures increased mixing quality for higher gas Knudsen numbers, i.e. smaller primary particle sizes.

Ground-based optical observation is a key sensing modality for space-object monitoring, but reliable association becomes challenging when orbital priors are degraded, site-position information is imprecise, and each frame contains multiple point-source candidates. This paper proposes a robust framework for single-target association among multiple image candidates and recursive estimation under such conditions. The method first converts image-domain candidates into unit line-of-sight (LOS) directions and represents their local deviations in a prediction-aligned tangent plane. Angular measurement noise and site-position uncertainty are then propagated into the local covariance model. Based on this representation, a kinematic-photometric five-dimensional (5D) normalized innovation squared (NIS) gating statistic is constructed by jointly evaluating local position, pseudo-velocity, and photometric consistency. After association, a three-dimensional decoupled Kalman update is performed using only the single-frame position and photometric measurements. Experiments on four real ground-based optical satellite observation sequences, including two static scenarios and two dynamic scenarios, show correct association rates of 100.00%, 100.00%, 100.00%, and 85.86%, respectively. These results demonstrate that the proposed framework improves association reliability under degraded orbital priors and imprecise site-position information while maintaining stable recursive estimation.

Shadows in high-resolution urban remote sensing imagery significantly degrade radiometric and structural information, thereby limiting the performance of downstream tasks such as classification and object extraction. Therefore, effective shadow removal is essential for improving the reliability of urban remote sensing applications. Existing methods still exhibit limitations in accurately detecting complex shadows, especially small-scale shadows and ambiguous boundaries, and shadow compensation in umbra regions often suffers from under-correction due to inadequate illumination modeling. To address these challenges, a physics-guided shadow removal framework that integrates lightweight shadow detection with illumination-aware compensation is proposed. A lightweight U-Net (LSDU) is designed to efficiently capture multi-scale shadow features, while a modified illumination intensity ratio method (MIIRM) is developed to explicitly model illumination differences between umbra and penumbra. Furthermore, a dynamic penumbra compensation method (MDPCM) is introduced to alleviate over-compensation effects in transition regions and improve radiometric consistency. Experiments on the Aerial Imagery Shadow Dataset (AISD) demonstrate that the proposed method achieves over 96% overall accuracy in shadow detection and the lowest RMSE in shadow compensation among existing state-of-the-art methods, while maintaining strong robustness across diverse urban scenes.

Infrared UAV detection plays a crucial role in both security surveillance and military applications. However, under fast UAV movement or dynamic zooming scenarios, the rapid scale variation of targets poses severe challenges to existing detection models, especially on resource-constrained edge devices. To address this, a lightweight scale-adaptive multi-scale feature fusion model, termed LMF-IR, is proposed for efficient and accurate detection under sudden target size changes. The model integrates three key components: a Multi-Dilation Residual Block (MDRB) for enhanced multi-scale feature representation, an improved Channel Attention Model–Feature Fusion Pyramid Network (CAM-FPN) to boost adaptive feature fusion, and a modified P-WIoU loss function designed for precise bounding box regression under varying target sizes. The MDRB module effectively captures fine-grained features across multiple scales and reliably identifies targets of varying sizes. The CAM-FPN incorporates a channel attention mechanism, which can dynamically adjust the weights of features, enabling the model to focus on informative feature channels. The redesigned P-WIoU loss function is designed to account for the shape characteristics of UAV target bounding boxes. It includes centroid distance, overlap ratio, and aspect ratio, thereby improving localization accuracy under rapid scale changes. The experimental results on our self-built UAV–infrared dataset show that LMF-IR reduces 1.4 G in floating-point operations compared to the baseline model, and the parameter count is reduced to 62% of the baseline. At the same time, mAP@0.5:0.95 increases by 2.4%. Moreover, on the public ANTI-UAV dataset, our method increases mAP@0.5:0.95 by 4.8%, indicating that our method has excellent performance in real-time infrared UAV detection under rapid target scale changes.

This study investigates supervised machine learning techniques for identifying ship hydrodynamic coefficients from CFD-generated data from free-running simulations. Specifically, ordinary least squares and regularized regression methods are applied to Abkowitz-type manoeuvring models. Training and validation datasets are derived from URANS simulations of zig-zag and turning circle manoeuvres, which are validated against experimental benchmark data. The analysis evaluates the effects of coefficient set size, minimum training length required for predictive model training, and manoeuvre combinations on model performance. Results demonstrate the suitability of large-angle zig-zag manoeuvres for hydrodynamic system identification, provided that multicollinearity is addressed through appropriate coefficient selection, regression models, or input data variability. Larger coefficient sets offer greater model flexibility for variable conditions but are more prone to multicollinearity. Regularized regression techniques effectively mitigate multicollinearity and notably enhance prediction accuracy, as does incorporating more diverse manoeuvring data. Among tested models, Ridge regression provided the best compromise between computational efficiency and prediction accuracy.

OBJECTIVE: To describe the spatiotemporal distribution characteristics of colorectal cancer (CRC) mortality in Chinese cancer registration areas and to predict its future trends, thereby providing an epidemiological basis for CRC prevention and control in China. METHOD: This study used cancer registry data from 2005 to 2018, collected from the China Cancer Registry Annual Report (2008-2021), to comprehensively describe CRC mortality patterns. Joinpoint regression models were employed to calculate the annual percentage change (APC) and average annual percentage change (AAPC) to analyze temporal trends. The age-period-cohort model was used to disentangle the effects of age, period, and cohort. A Bayesian age-period-cohort (BAPC) model was constructed to predict future trends for 2019-2035. Global and local spatial autocorrelation methods were applied to describe the spatial distribution of CRC mortality and to identify high-risk areas. RESULTS: From 2005 to 2018, in Chinese cancer registration areas, the mortality of CRC was 14.45/100,000, and the age-standardized mortality rate (ASMR) was 11.13/100,000 [95% confidence interval (CI): 11.03, 11.27]. From 2005 to 2007, the mortality rate of CRC showed an upward trend, with an APC of 3.67% (95% CI: -0.77%, 8.33%); from 2007 to 2013, there was a significant downward trend with an APC of -2.45% (95%CI: -3.14%, -1.47%); from 2013 to 2018, there was an upward trend with an APC of 0.13% (95% CI: -0.88%, 1.17%). From 2007 to 2013, the CRC mortality rates of different genders showed a significant downward trend. In urban areas from 2009 to 2018, the CRC mortality rate showed a significant downward trend with an APC of -1.21% (95% CI: -1.71%, -0.71%), but in rural areas, it showed a significant upward trend with an AAPC of 0.59% (95% CI: 0.24%, 1.56%). The age effects indicated that the risk of CRC death increased with age. From 2019 to 2035, the ASMR of CRC in Chinese cancer registration areas will generally show a slow upward trend. Spatial autocorrelation analysis showed that the mortality rate of CRC in Chinese cancer registration areas exhibited spatial clustering characteristics, with hotspots concentrated in the eastern coastal and northeastern provinces. CONCLUSIONS: The mortality rate of CRC in Chinese cancer registration areas showed a trend of first increasing, then decreasing, and then increasing from 2005 to 2018. It was predicted that it will continue to rise from 2019 and still face a series of challenges such as urban-rural, gender, and regional inequality. In the future, more targeted prevention and control strategies should be adopted, combined with dietary and lifestyle interventions, to further reduce the burden of CRC.

Abstract. This study updates an article published in NHESS journal in 2015 and investigates long-term changes in landslide-triggering rainfall conditions in Calabria (southern Italy) over 1921–2020. A catalogue of 3006 rainfall events associated with landslides (RELs) was reconstructed using 9530 landslide records and daily rainfall measurements from 318 gauges. Rainfall thresholds were calculated for 15 30-year moving windows to investigate the triggering conditions of the RELs. Results show a marked increase in the number of RELs after 2009, shifts in seasonal occurrence, and decreasing rainfall duration and cumulative amounts. Triggering rainfall shows an overall decreasing trend over the years.

Accurate comparison of water-surface evaporation observations from different devices is essential for integrating long-term hydrological records. This study analyzed two years of synchronous daily observations from five co-located evaporation devices in Duyun, Guizhou, China: D20, E601, and three evaporation ponds with surface areas of 1, 5, and 20 m2 (P1, P5, and P20). Evaporation differences, conversion coefficients, correlations, and statistical variability were evaluated at annual, seasonal, monthly, and daily scales. The main findings were as follows: (1) among the three similarly constructed ponds, annual evaporation decreased from 946 mm for P1 to 896 mm for P5 and 874 mm for P20, whereas the annual totals measured by D20 and E601 were 842 and 861 mm, respectively; (2) the relationship among the three ponds varied within the year, reversing in July and August and becoming non-monotonic in May and June; (3) pairwise correlations were generally strongest in summer and weakest in winter, indicating pronounced seasonal variation in inter-device relationships; and (4) because D20 was installed above ground whereas E601 and the three ponds were buried, differences among all five devices reflected combined scale- and design-related effects, while comparisons among the three ponds primarily represented surface-area-related effects. These findings provide a site-specific basis for harmonizing evaporation records and improving their application in hydrological and water-balance studies.

Accurate prediction of effluent total nitrogen (TN) is important for early exceedance warning and operational control in wastewater treatment plants. Existing decomposition-based models may overestimate performance when full-series decomposition is performed before data splitting, causing potential temporal information leakage. To address this issue, this study compares noncausal and strictly causal Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise combined with Long Short-Term Memory (ICEEMDAN-LSTM) and Variational Mode Decomposition–Long Short-Term Memory network (VMD-LSTM) settings, and proposes a Level–Trend Persistence-Residual LSTM (LT-PR-LSTM) for univariate effluent TN prediction. The model uses Persistence as the short-term state baseline, extracts level features from historical TN, and introduces first- and second-order differences to learn residual corrections relative to the current state. Multi-model comparison, ablation experiments, stability tests, SHapley Additive exPlanations (SHAP) interpretation, supplementary dataset validation, and efficiency analysis were conducted. Results show that noncausal decomposition inflates predictive performance. LT-PR-LSTM achieves the best main-test performance, with RMSE 1.1273, MAE 0.6082, MAPE 7.5455%, and R2 0.8512, reducing RMSE, MAE, and MAPE by 6.73%, 7.64%, and 8.56% compared with Persistence. SHAP identifies TN(t−2h) as the dominant predictor, and the model requires only 0.5348 ms/sample, indicating potential for online TN early warning.

In Nigeria and mainly in Imo State, flood disasters continue to be one of the most devastating environmental threats to arable crops and to the livelihoods of farmers. However, there is still limited empirical data on the factors influencing arable crop farmers’ strategies to mitigate flood disasters in Imo State. Therefore, the study examined the strategies that arable crop farmers use for mitigating flood disasters, identified the factors that influence the implementation of flood mitigation strategies, and identified the barriers that hamper their effective application. A structured questionnaire was used to collect the primary data from three-hundred and twenty (320) farmers cultivating arable crops, selected specifically from the critical flood-prone farming communities of Imo State, Nigeria. The data were analyzed using multinomial logistic regression and descriptive statistics. The results revealed that arable crop farmers practiced various strategies to mitigate flood disasters, including the construction of floodways; dams; land filling; raising beds and ridges; and crop diversification. The multinomial logit result indicated that socio-economic factors significantly influenced farmers’ choice of flood mitigation strategies. Furthermore, arable crop farmers face major barriers in effectively implementing and practicing flood mitigation strategies in the area, which include the high cost of flood disaster strategies (95.00%) and inadequate knowledge of mitigation strategies (84.00%). The study concludes that strengthening farmers’ access to resources, extension services, and cooperative organizations would enhance the use of effective flood mitigation strategies and improve resilience to flood disasters among arable crop farmers in the study area.

The soil drilosphere is a critical biogeochemical hotspot, yet its role as a key interface for microplastic (MP) accumulation and impact remains poorly characterized. We investigated how polyethylene microplastics (<150 μm) affect the drilosphere compartments (gut, burrows, and casts) of two distinct earthworm ecotypes: epigeic Eisenia fetida and endogeic Pheretima guillelmi . Results showed that MPs significantly enrich in the drilosphere compared to bulk soil, with the endogeic species exhibiting greater accumulation. While earthworm activity typically stimulated nutrient characteristics, MP exposure disrupted these functions, significantly reducing total nitrogen (5.0–25.0%) and ammonium (28.5–62.1%). Ecotype-specific host damage emerged: E. fetida exhibited pronounced immune and oxidative stress responses, whereas P. guillelmi suffered severe digestive and metabolic impairments. These impacts were mediated by distinct microbiome reprogramming. MP-induced dysbiosis intensified progressively along the soil–drilosphere–gut continuum. Multivariate and transcriptomic analyses revealed that external-drilosphere microbiota shifts drove carbon–nitrogen characteristic alterations, while internal dysbiosis triggered host physiological stress. This study highlights that ecotype-specific restructuring of drilosphere microbiomes underpins the ecosystem-scale impacts of MP pollution, demonstrating that earthworm functional diversity is essential for comprehensive soil health risk assessments.

Introduction Against the backdrop of the accelerating global energy transition and the urgent need to address climate change, China, as the world’s largest greenhouse gas emitter, bears major responsibility for achieving its carbon peaking target by 2030 and carbon neutrality target by 2060. In this context, accelerating the development of renewable energy has become a crucial pathway toward sustainable development. Evaluating the impact of transportation infrastructure, particularly shipping development, on renewable energy development is therefore of important practical significance for achieving sustainable development goals. Methods Using panel data for 11 coastal provinces in China from 2013 to 2023, this study employs a two-way fixed effects model, mechanism analysis, and the System GMM approach to examine the impact of shipping development on the level of renewable energy development and its underlying mechanisms. Results The results show that, first, the level of shipping development in China’s coastal regions exhibited an overall fluctuating upward trend, while the regional stratification pattern remained relatively stable. Second, the baseline regression results indicate that shipping development significantly promotes renewable energy development. Third, the mechanism analysis reveals that shipping development mainly enhances renewable energy development indirectly through two channels: promoting industrial structure upgrading and facilitating technological progress. Fourth, the System GMM estimates further confirm the robustness of the above findings and suggest that renewable energy development exhibits significant dynamic inertia. Discussion This study enriches the research perspective on the relationship between transportation infrastructure and energy transition to some extent and provides empirical evidence as well as policy implications for coordinating high-quality shipping development and clean energy deployment in China’s coastal regions.

Introduction Maintaining river water quality remains a key environmental challenge, particularly in the context of pressures from economic activity, energy systems, and resource use. This study examined the development of biochemical oxygen demand (BOD) in rivers across selected European Union (EU) countries from 2000 to 2022 and assessed its statistical relationships with selected energy system characteristics, environmental pressures, and socio-economic factors within the harmonized analytical period 2005–2022. Methods The study combined distributional and trend analyses using country-level data for 17 EU countries. Pairwise correlation and regression analyses were conducted on a refined analytical panel of 16 countries following the exclusion of Cyprus. The initial set of explanatory variables included energy intensity of gross domestic product (GDP), energy productivity, final energy consumption per capita, total final energy consumption, the share of fossil fuels in gross available energy, environmental tax revenues, water exploitation index (WEI+), carbon dioxide (CO 2 ) emissions per capita, the share of energy from renewable sources, real GDP per capita, and urban population. Results The results indicated an overall decline in BOD levels over the period 2000–2022, accompanied by visible differences across countries. The highest BOD concentrations were observed in Romania and Bulgaria, while lower levels predominated in countries such as Ireland, Slovenia, and Croatia. Spain showed a visible upward temporal shift in BOD values. Several statistically significant relationships were identified. Positive associations were observed between BOD and WEI+ as well as energy intensity of GDP, while negative associations were found between BOD and environmental tax revenues, real GDP per capita, CO 2 emissions per capita, and the share of renewable energy. The identified patterns should be interpreted primarily as macro‐comparative associations within a longitudinal country‐level framework. The final regression model explained approximately 60% of the variability in BOD levels across the analyzed observations. Discussion The findings emphasize that river water quality in the EU is embedded within broader structural interactions between environmental pressures, energy system characteristics, and socio-economic development. These patterns highlight the importance of integrated policy approaches addressing energy efficiency, environmental taxation, cleaner energy systems, and broader water‐resource management, particularly in countries characterized by less favorable river water quality profiles.

Abstract UAV-based RGB imaging is increasingly used to track crop diseases and assess fungicide performance. However, standard disease severity measures may miss an important factor: fungicide-induced changes in canopy structure that modulate disease observability. Here we analyse a multi-site dataset of wheat yellow and leaf rust, integrating UAV-derived canopy cover and disease severity, and rater-based disease severity assessments. Results indicate that between visual and UAV imagery-based disease severity was high but fundamentally non-stationary, governed by canopy cover and epidemic intensity. Measurement error followed structured, canopy-dependent regimes. The canopy-normalized disease area, a metric that harmonizes severity with structural and spatial context, revealed ranking shifts in fungicide efficacy that are not captured through traditional metrics. This suggests that severity-based comparisons may combine true disease suppression with treatment-related changes in how visible symptoms are. Framing plant disease quantification as a measurement challenge, rather than only a detection, may support a more accurate assessment of treatment effects. Modelling the influence of canopy structure on symptom expression could provide a scalable and robust basis for disease monitoring and decision-making in precision agriculture.