Objective This study investigates the effects of green manure and straw incorporation, along with different regulators (goethite, ethephon, and sodium dichloroethyl sulfonate) on soil carbon and nitrogen pool management indices of the super hybrid high-yield rice variety u201CJingliangyou Huazhanu201D. The findings hold significant implications for rational resource utilization, enhancing soil fertility, and expanding carbon and nitrogen pools. Methods A field experiment was conducted from 2022 to 2024 in Gaoqiao Town, Changsha County, Hunan Province. Six treatments were set up, including chemical fertilizer + rice straw returning (CK), chemical fertilizer + rice straw and green manure returning (M), 30% nitrogen reduction + rice straw and green manure returning (R), chemical fertilizer + rice straw and green manure returning + goethite (F), chemical fertilizer + rice straw and green manure returning + ethephon (E), and chemical fertilizer + rice straw and green manure returning + sodium dichloroethyl sulfonate (S). The soil carbon and nitrogen pool management indices were calculated, and their key influencing factors were analyzed. Results 1) carbon and nitrogen reservoir management indices (CPMI and NPMI) exhibited significant seasonal and interannual variations across treatments. CPMI showed pronounced fluctuations, while NPMI increased annually with the third-year NPMI exceeding the previous two years by 43.63% and 36.18%, respectively; 2) compared to the control (CK) treatment (straw return plus chemical fertilizer alone), all other treatments generally increased both CPMI and NPMI. Treatment M showed the most significant enhancement effects for CPMI and NPMI (199.08 u00B1 44.60 and 96.29 u00B1 9.19, respectively), while treatments with three regulators showed 4.24%u201215.27% lower CPMI and 6.20%u201214.08% lower NPMI than M. Notably, treatment S significantly reduced NPMI (p u0026lt; 0.05); 3) random forest and structural equation modeling analyses indicated that DOC and TDN were primary factors influencing CPMI, while TDN was the main factor affecting NPMI. Conclusion The combined application of chemical fertilizer and rice straw green manure to fields is more effective in enhancing the soil carbon and nitrogen reservoir management index of super rice, representing the optimal choice for improving soil fertility and expanding soil carbon and nitrogen reservoirs.
Objective Fine particulate matter (PM2.5) pollution has become one of the most important environmental challenges worldwide due to its significant impacts on human health, ecosystems, and regional climate systems. Long-term exposure to elevated PM2.5 concentrations has been associated with respiratory and cardiovascular diseases, premature mortality, and increased risks of hospitalization. Atmospheric PM2.5 can also affect visibility, cloud properties, and regional climate processes. However, obtaining spatially continuous and temporally consistent PM2.5 observations remains challenging because ground-based monitoring networks are sparse and unevenly distributed, especially in regions with complex terrain and heterogeneous emission sources. Satellite remote sensing provides an effective approach for large-scale PM2.5 estimation, and aerosol optical depth (AOD) has been widely used as an important proxy for ground-level PM2.5 concentrations. Nevertheless, existing AOD-based inversion methods still have several limitations. Traditional statistical and machine learning models generally focus on either spatial feature extraction or temporal variation analysis, while lacking the ability to simultaneously characterize complex spatiotemporal relationships. Moreover, most deep learning approaches are purely data-driven and do not explicitly incorporate atmospheric physical mechanisms, which may reduce their generalization capability and lead to physically inconsistent predictions across different temporal or spatial conditions. Therefore, this study develops a novel physics-constrained U-Net-LSTM deep learning framework for high-resolution spatiotemporal reconstruction of ground-level PM2.5 concentrations. Methods The proposed framework integrates a U-Net encoder-decoder architecture with a long short-term memory (LSTM) network to simultaneously capture spatial patterns and temporal evolution of PM2.5 pollution. The U-Net architecture is employed to extract multiscale spatial features and preserve detailed spatial information through skip connections, while the LSTM network is used to model temporal dependencies and dynamic variations in air pollution processes. To improve feature representation and model adaptability, convolutional block attention modules (CBAMs) are embedded in the encoder and decoder pathways, enabling the model to automatically learn channel and spatial attention weights. This mechanism allows the network to focus on critical environmental variables and important geographic regions while reducing the influence of irrelevant information. Furthermore, a physics-based residual constraint derived from the atmospheric advection-diffusion equation is incorporated into the training loss function. This constraint introduces atmospheric mechanisms, including horizontal wind-driven transport, turbulent diffusion, and pollutant removal processes, into the deep learning framework. By combining data-driven feature learning with physical knowledge, the model is encouraged to generate PM2.5 predictions that are not only statistically accurate but also consistent with atmospheric dynamics, thereby improving prediction reliability and generalization capability. The proposed model was trained and evaluated using observations from 97 air quality monitoring stations across the Sichuan Basin from 2015 to 2020. The Sichuan Basin was selected because of its complex topography, severe pollution conditions, and frequent haze events. The input dataset included multisource environmental variables at a spatial resolution of 1 km, including satellite AOD products, ERA5 meteorological parameters (wind speed, temperature, relative humidity, surface pressure, and boundary layer height), a digital elevation model (DEM), the normalized difference vegetation index (NDVI), land-use classification, and geographic coordinates. A total of 85205 spatiotemporal samples were generated. Three validation strategies, namely sample-based, time-based, and site-based validation, were used to evaluate model accuracy and generalization capability. Model performance was assessed using the coefficient of determination (R2), root mean square error (RMSE), and mean absolute error (MAE). Results The proposed physics-constrained U-Net-LSTM model achieved superior performance compared with benchmark models, including CNN, LSTM, and ConvLSTM, under all validation scenarios. In sample-based validation, the model achieved an R2 of 0.874, an RMSE of 8.12 u03BCgu00B7m-3, and an MAE of 8.01 u03BCgu00B7m-3. In time-based and site-based validations, the model maintained strong predictive capability, achieving R2 values of 0.86 and 0.83, respectively, with prediction errors reduced by approximately 5%u201225% compared with baseline models. These results demonstrate the strong temporal transferability and spatial generalization capability of the proposed framework. Ablation experiments further confirmed the effectiveness of each model component. The baseline U-Net-LSTM model without CBAM or physical constraints achieved an R2 of 0.79, with an RMSE of 16.62 u03BCgu00B7m-3 and an MAE of 14.63 u03BCgu00B7m-3. After introducing CBAM, model performance improved to R2 = 0.82, demonstrating the contribution of the attention mechanism to the extraction of important environmental features. With the further incorporation of the physics-based residual constraint, the model achieved R2 = 0.87, RMSE = 11.21 u03BCgu00B7m-3, and MAE = 8.83 u03BCgu00B7m-3, with more than 80% of monitoring stations achieving R2 values greater than 0.7. These findings confirm that physical guidance significantly enhances model accuracy and robustness. Permutation importance analysis indicated that AOD was the most influential predictor, contributing for 40.63% of the total permutation importance, followed by surface pressure, air temperature, DEM, and relative humidity. These results highlight the combined effects of aerosol loading, meteorological conditions, and topographic characteristics on PM2.5 accumulation and dispersion in the Sichuan Basin. The reconstructed PM2.5 maps successfully captured regional pollution patterns and long-term temporal trends. The annual mean PM2.5 concentration decreased from 38.95 u03BCgu00B7m-3 in 2015 to 25.91 u03BCgu00B7m-3 in 2020, indicating the effectiveness of regional air pollution control measures. Spatially, a persistent pattern of high concentrations in the center and low concentrations at the periphery was observed, with the Chengdu Plain and surrounding urban agglomerations consistently exhibiting higher PM2.5 levels. Seasonal analysis showed that winter had the highest PM2.5 concentration (44.71 u03BCgu00B7m-3), mainly because of unfavorable atmospheric dispersion conditions, temperature inversion, and increased heating emissions, whereas summer had the lowest concentration (15.95 u03BCgu00B7m-3) because of enhanced atmospheric ventilation and precipitation scavenging. Additionally, the proposed framework successfully reproduced a severe haze episode in December 2020, accurately capturing pollutant accumulation, rapid enhancement, and dissipation following the intrusion of a cold front. The results further revealed the influence of regional transport processes and identified major urban clusters as persistent pollution hotspots. Conclusion This study proposes a physics-constrained deep learning framework that effectively integrates spatiotemporal feature learning, attention mechanisms, and atmospheric physical constraints for high-resolution PM2.5 reconstruction. The proposed physics-constrained U-Net-LSTM model achieves improved accuracy, generalization capability, and interpretability compared with conventional data-driven approaches. By bridging the gap between deep learning prediction and physical understanding of atmospheric processes, this framework provides an effective solution for generating continuous, high-resolution PM2.5 distributions in regions with complex terrain and offers valuable support for air quality assessment, pollution control strategies, and environmental management.
Objective This study aims to assess the value of ecological products in rare earth mine restoration areas, explore the factors influencing changes in ecological product value, and provide a scientific basis for optimizing mine restoration models and enhancing the value of ecological products in specific regional units undergoing rare earth mine ecological restoration. Methods Taking Xunwu County, Jiangxi Province, as a case study, we constructed a theoretical framework and accounting system for the value of ecological products in the rare earth mine restoration areas based on statistical and survey data from 2010, 2016 and 2022. The functional value method was used to estimate the value of ecological products in the rare earth mine restoration areas of Xunwu County and to reveal their spatiotemporal evolution characteristics, while the geographic detectors were used to analyze the influencing factors. Results 1) The value of ecological products varied significantly among different restoration areas. The value of ecological products in the u201Crestoration + industryu201D area decreased by 11.8485 million yuan because of industrial park construction, whereas the values in the comprehensive resource utilization area and ecological greening area increased by 227.6748 million yuan and 5.0726 million yuan, respectively; 2) During the study period, the growth in ecological product value showed clear stage-specific characteristics. From 2010 to 2016, constrained by the scale of ecological restoration project implementation, the value of ecological products increased by only 12.7162 million yuan. From 2016 to 2022, the total value increased by 208.18260 million yuan, accompanied by coordinated improvements in ecological, economic, and social benefits; 3) The duration and area of ecological restoration projects were the dominant factors affecting the realization of ecological product value in rare earth mining areas, while the interaction between capital investment per unit area and ecological restoration project area was the most significant. Conclusion Rare earths are key strategic resources for modern industry; however, intensive mining has caused systemic ecological problems in mining areas, including forest degradation, soil erosion and microclimate deterioration. Ecological restoration in typical rare earth mining areas has significantly enhanced ecosystem service functions through comprehensive measures such as terrain consolidation, water-system management, and vegetation restoration, thereby increasing the supply capacity of ecological products. To further improve ecological governance in rare earth mining areas, a diversified mechanism for realizing the value of ecological products should be established to increase the economic value of the ecological restoration areas; nature-based solutions should be adopted to optimize restoration pathways and value-realization mechanisms; and synergies between land development and urbanization should be strengthened to promote coordinated economic and ecological development.
Objective The biogeobattery effect is widely observed in freshwater and marine environments. Previous studies have demonstrated that organic matter undergoes distinct transformation processes under the biogeobattery effect, which can substantially influence overall carbon turnover. As the most mobile and reactive fraction of organic matter in sediments, dissolved organic matter (DOM) is a major contributor to carbon turnover and plays a pivotal role in regulating contaminant fate through adsorption, complexation, photosensitization, and redox reactions, thereby profoundly influencing sediment ecological health. However, the transformation characteristics of DOM under the biogeobattery effect and the underlying mechanisms remain poorly understood, resulting in the limited knowledge of the ecological functions and carbon turnover processes in sediments associated with this widespread natural phenomenon. Methods In this study, we constructed a series of reactors to characterize DOM properties and elucidate the mechanisms driving DOM transformation in coastal sediments under the biogeobattery effect. Three experimental groups were established: coastal sediment-only reactors (CK) as the control group; reactors without an electrical connection between the sediment and overlying water (SO), representing the absence of a biogeobattery effect; and reactors with an electrical connection between the sediment and overlying water (SC), representing the formation of a biogeobattery system. Results During the 84-day operation period, the electric potential of sediments across all layers in the SC group ranged from u2212193 to 319 mV (vs. Ag/AgCl), which was consistently higher than that in the CK and SO groups (u2212335 to 162 mV and u2212321 to u221215.1 mV, respectively). These results clearly demonstrate that the artificial biogeobattery effect increased sediment electric potential. Moreover, the maximum output voltage across a 100 u03A9 external resistor in the SC group was 57.3 mV, while the maximum mean self-potential of the sediment (with the water-sediment interface defined as zero potential) reached 93.9 mV. Compared with previous studies, the lower output voltage observed in the SC group suggests a relatively weaker biogeobattery effect in this study. Under this relatively weak biogeobattery effect, DOM consumption efficiencies in the CK, SO, and SC groups were 51.5% u00B1 19.3%, 37.1% u00B1 25.4%, and 36.4% u00B1 9.1%, respectively. These results suggest that the biogeobattery effect may, to some extent, reduce DOM loss and enhance carbon retention in coastal sediments. Fluorescence analysis identified three major components in all DOM samples: a long-wavelength humic-like component, a fulvic-acid-like component, and a tyrosine-like component. After 84 days of operation, the tyrosine-like component became the dominant contributor to DOM fluorescence in all sediment samples, accounting for 68.40%, 74.93%, 73.98%, and 75.92% of the total fluorescence intensity in the initial sediment (IS), CK, SO, and SC groups, respectively. These results indicate a transformation trend from long-wavelength humic-like and fulvic-acid-like components toward the tyrosine-like component in the CK, SO, and SC groups, with the most pronounced transformation observed in the SC group. Owing to the relatively high abundance of tyrosine-like components, DOM in the SC group exhibited the highest electron-transfer capacity, reaching (5.20 u00B1 0.27) mmolu00B7gu22121, compared with (4.01 u00B1 0.36) and (3.70 u00B1 0.08) mmolu00B7gu22121 in the CK and SO groups, respectively. This enhanced electron-transfer capacity may facilitate electron-transfer processes within the sediment under the biogeobattery effect. Further analysis revealed that the SC group exhibited the lowest Fe(u2161) concentration and Fe(u2161)/total Fe ratio, while the SO42u2212 concentration remained comparable to that in the initial sediment. These altered iron and sulfur redox states may have increased sediment electric potential and reshaped microbial community structure, thereby promoting DOM transformation under the biogeobattery effect. After 84 days of operation, Proteobacteria exhibited a high relative abundance in the anodic biofilms of both the SO and SC groups, particularly on the SC anode, where they accounted for 62.16% of the microbial community. The enrichment of Proteobacteria in anodic biofilms suggests that this phylum may include key electroactive microorganisms (EAMs) involved in establishing the artificial biogeobattery effect. Furthermore, Proteobacteria showed significant negative correlations with the long-wavelength humic-like component, fulvic-acid-like component, DOM concentration, and UV indices, while exhibiting a significant positive correlation with the tyrosine-like component. These relationships suggest that Proteobacteria may contribute to the degradation and transformation of humic-like and fulvic-acid-like components into tyrosine-like substances, potentially explaining their enrichment on the SC anode. Among the 18 dominant genera identified in this study, Bacillus, Paenisporosarcina, Acinetobacter, Romboutsia, and Microbulbifer exhibited significant differences in their distribution patterns among the experimental groups. Specifically, the relative abundance of Bacillus increased significantly in both sediments and anodic biofilms of the SO and SC groups compared with that in the initial sediment (IS). Paenisporosarcina and Acinetobacter were enriched in anodic biofilms, accounting for 1.19% and 56.13% of the microbial community in the SC group, respectively. Bacillus, which has been reported to participate in the synthesis of aromatic compounds, and Paenisporosarcina, which is associated with the degradation of protein-like substances, both showed significant positive correlations with DOM content. Moreover, the tyrosine-like component exhibited significant positive correlations with Bacillus, Paenisporosarcina, and Acinetobacter, suggesting that these genera, particularly Acinetobacter, may contribute to the production or transformation of tyrosine-like components, thereby enhancing the electron-transfer capacity of DOM under the biogeobattery effect. Conclusion Collectively, these results suggest that microbial community composition and relative abundance are governed by sediment environmental conditions, particularly the availability and characteristics of electron donors and acceptors. The interplay among these factors regulates sediment electric potential and ultimately determines the molecular composition and electrochemical properties of DOM. This study provides a foundation for advancing our understanding of the ecological functions of coastal sediments exhibiting the biogeobattery effect and establishes a scientific basis for developing strategies to regulate carbon turnover. However, the results of this study represent a preliminary investigation based on a single sampling time point. Future studies incorporating observations at multiple time points are needed to elucidate the temporal dynamics of the biogeobattery effect and associated DOM transformation processes, thereby enabling a more comprehensive understanding of the ecological functions and carbon cycling processes associated with this natural phenomenon in coastal sediments.
Objective To explore the feasibility of using low-altitude unmanned aerial vehicle (UAV) hyperspectral remote sensing to estimate soil copper (Cu) content in contaminated areas and to provide technical support for the rapid assessment and remediation of contaminated sites. Methods An abandoned non-ferrous metal processing plant in Yinzhou District, Tieling City, Liaoning Province, was selected as the study area. A quantitative inversion method for soil Cu content was developed by integrating spectral transformation, feature-band selection, and spectral indices. Three modeling strategies were compared: four two-dimensional spectral indices (2D), four three-dimensional vegetation spectral indices (3D SI), and four newly constructed three-dimensional spectral indices (3D BI). The Newton-Raphson optimized eXtreme Gradient Boosting (NRBO-XGBoost) algorithm was used to establish the Cu content inversion model. Results Spectral transformation significantly improved the correlations between spectral variables and Cu content. The first-order derivative transformation yielded the strongest correlation, with ABI3 (850, 885, 520) exhibiting an r value of -0.570. Under the same preprocessing conditions, the 3D vegetation spectral indices and newly constructed 3D spectral indices showed better correlations with Cu content and better modeling performance than the 2D spectral indices. The NRBO-XGBoost model combined with the FD-3D SI dataset achieved the best performance (R2 = 0.933, u03C3RMSE = 75 mgu00B7kg-1), substantially outperforming the conventional XGBoost model (R2 = 0.833, u03C3RMSE = 116 mgu00B7kg-1). Conclusion The UAV hyperspectral estimation model based on NRBO-XGBoost can rapidly estimate soil Cu contamination and generate regional distribution maps. These results demonstrate the potential of UAV hyperspectral remote sensing for monitoring soil heavy metal contamination and provide an effective technical approach for the efficient assessment and remediation of contaminated sites.
