Search Results (13,942)

To address the limited accuracy of power demand prediction for hybrid electric vehicles under complex and dynamic driving conditions, this paper proposes a hybrid prediction approach based on the cascade correction of Long Short-Term Memory networks and Least Squares Support Vector Machines (LSTM-LSSVM-CC). The proposed method adopts a stage-wise modeling framework that exploits the least-squares optimality of LSSVM for low-frequency steady-state signals and the dynamic compensation capability of LSTM for high-frequency non-stationary residuals, thereby achieving complementary feature representation in the frequency domain. Specifically, an LSSVM is first used to construct a baseline regression model that captures stationary components, followed by an LSTM network that performs deep temporal modeling of the residual sequence to correct nonlinear prediction errors. Extensive experiments conducted on three standard driving cycles—CLTC-P, WLTP, and UDDS—demonstrate that the proposed model consistently outperforms conventional methods including LSSVM, RNN, ELMAN, and Random Forest in multi-step predictions, achieving an average RMSE reduction of 28–52% and maintaining correlation coefficients (R²) between 0.87 and 0.99. Particularly under highly dynamic and abrupt load conditions, the model exhibits superior real-time performance and stability while significantly mitigating cumulative prediction errors. These results demonstrate that the proposed LSTM-LSSVM-CC model achieves robust modeling performance of non-stationary time series while balancing prediction accuracy and computational efficiency, providing an effective technical foundation for hybrid vehicle energy management optimization and offering a transferable theoretical framework for time-series prediction in complex systems. Full article

Morphological traits, as core components of functional traits, are fundamental in determining environmental adaptability. However, under climate warming, the adaptive morphological changes and associated ecological risks of locust populations migrating to higher altitudes remain poorly understood. Here, we investigated Calliptamus italicus, the dominant locust species in the desert steppes of the Ili River Basin, to explore the response patterns of its morphological functional traits along an altitudinal gradient and their relationships with environmental factors. Morphological measurements revealed that forewing area, width, and length, as well as hindwing width, exhibited highly significant positive correlations with altitude (p < 0.01); in contrast, body length, head width, head height, pronotum length, pronotum width, hind femur length, and hind tibia length displayed significant negative correlations with altitude (p < 0.05). All morphological indicators presented highly significant sexual dimorphism (p < 0.001). Ratio analysis showed that the pronotum width-to-head width ratio (M/C), pronotum height-to-head width ratio (H/C), and forewing length-to-hind tibia length ratio (E/F) were significantly positively correlated with the altitudinal gradient (p < 0.05), with all ratios exhibiting significant sexual differences (p < 0.05). Random Forest analysis showed that PC1 (75.5% of variation) reflected traits for feeding, jumping, and reproduction, whereas PC2 (5.6%) represented flight-related traits, with significant sexual dimorphism. This study demonstrates that trait variation in C. italicus along an altitudinal gradient is closely linked to environmental factors. Our findings provide critical data for predicting habitat adaptation responses in locust populations, thereby enhancing the precision and efficacy of locust plague management and contributing to the conservation and restoration of desert steppe ecosystems. Full article

Fine-scale crop type information is essential for agricultural monitoring, irrigation management, and food security assessment. This study mapped three major crops—wheat, corn, and sunflower—in the Hetao Irrigation District, China, using multi-temporal Sentinel-2 optical imagery and Sentinel-1 SAR observations at the parcel scale. A multi-source feature set, including spectral bands, vegetation and red-edge indices, moisture-related variables, radar backscatter coefficients, and derived radar features, was constructed from the full growing season. An LSTM network was used to learn temporal representations of crop phenological dynamics, and the resulting embeddings were then combined with traditional machine learning classifiers, including Random Forest (RF), Support Vector Machine (SVM), and Extreme Gradient Boosting (XGBoost), for final classification. The results show that the hybrid framework substantially improves classification performance compared with the corresponding non-LSTM classifiers. Among all tested models, XGBoost + LSTM achieved the best performance, with an overall accuracy of 93.61%, a Kappa coefficient of 91.66%, and a mean IoU of 87.41%. The class-wise F1-scores were 85.61% for wheat, 97.22% for corn, and 87.27% for sunflower. Additional experiments further confirmed the advantages of parcel-based aggregation in improving spatial consistency and reducing mixed-field noise. The proposed framework provides a promising parcel-scale workflow for crop type mapping in fragmented irrigation districts, while its transferability across years and regions still requires further validation. Full article

Soil health is critical for the sustainability of tropical plantation ecosystems, However, the ecological factors driving productivity gradients remain inadequately understood. This study investigated rubber plantations on Hainan Island with varying yield levels to assess soil health and its underlying ecological mechanisms using a minimum data set (MDS) approach. Twenty-seven soil physical, chemical, and biological indicators were analyzed at two depths (0–20 cm and 20–40 cm). Principal component analysis identified seven key indicators for the MDS: soil organic matter (OM), alkaline-hydrolyzable nitrogen (AN), cation exchange capacity (CEC), dissolved organic carbon (DOC), microbial biomass phosphorus (MBP), acid phosphatase activity (ACP), and microbial diversity (Shannon-Wiener index, SHDI). The soil health indices derived from the MDS showed strong correlations with those generated from the total data set (TDS) (p < 0.001), confirming the reliability of the MDS framework. Overall, soil health levels were rated low to moderate with no significant differences across low-yield plantations (≤900 kg·ha⁻¹), medium-yield plantations (900–1200 kg·ha⁻¹), and high-yield plantations (≥1200 kg·ha⁻¹)., suggesting a decoupling of soil health and rubber productivity under uniform management practices. Random forest analysis identified microbial-driven phosphorus cycling, particularly MBP and ACP, as the primary determinant of soil health across soil layers, with DOC and SHDI also contributing significantly. These findings highlight the critical role of microbial-mediated nutrient cycling in maintaining soil health in rubber plantations and suggest that current management practices prioritize short-term yields over long-term soil ecological stability. Enhancing microbial activity and increasing organic matter inputs may be essential for improving soil health and ensuring the sustainability of rubber production in tropical agroecosystems. Full article

Many cities adopt greening strategies to reduce contamination from combined sewer overflows (CSOs). Nonetheless, quantifying the impact of urban greening on CSO-affected water quality at the city scale remains challenging. To address this challenge, this work leveraged supervised learning to link water swimmability with the greening of a CSO shed (the drainage area of a CSO outfall), using New York City (NYC) as a case study. Random forest classification models were built to predict water swimmability after rainfall at 46 sites in NYC water bodies impacted by CSOs. A 14-feature model (AUROC =0.81, accuracy = 0.78) revealed that greening improved local water quality. However, water flow speed, antecedent rain depth, and CSO shed area were also influential. A simplified four-feature model (AUROC = 0.8, accuracy = 0.75) explored links between levels of greening and the probability of non-swimmable waters ($P_{ns}$) following different 18 h rainfall depths. Increased greening was found to be most impactful in reducing $P_{ns}$ for CSO sheds discharging to water bodies with flow speeds < 6 cm/s. For CSO sheds discharging to water bodies with flow speeds $\ge$ 14.7 cm/s, urban greening had no impact on $P_{ns}$. The work illustrates the utility of supervised learning in supporting citywide decisions regarding urban greening investments. Full article

Mongolians have high overweight prevalence linked to their nomadic lifestyle and diet, but gut microbiota studies in this population are scarce. This study used fecal metagenomic and serum metabolomic analyses of 96 Mongolian participants (normal-weight n = 55, overweight n = 41) to characterize gut microbiome alterations and identify weight-related biomarkers. The analyses revealed that Parabacteroides distasonis, Barnesiella intestinihominis, and Alistipes onderdonkii were significantly reduced in overweight individuals (p < 0.05). Concurrently, the metabolites such as beta-cryptoxanthin, p-cresol, and ribothymidine were significantly down-regulated in the overweight group (p < 0.05). Random forest models from the three datasets showed a strong diagnostic ability for microbial families (AUC > 0.70). A subsequent integrated multi-kingdom classifier that combined microbiota and metabolite data achieved the highest performance (AUC = 0.818). Key features with high predictive contributions were identified, including Lactobacillus crispatus, Alistipes onderdonkii, and Parabacteroides distasonis, and metabolites, such as beta-cryptoxanthin, p-cresol, and picolinic acid. These results show the random forest model has high predictive value for distinguishing normal weight and overweight individuals. In summary, this study identified specific gut microbiota and serum metabolomic profiles linked to overweight in Mongolians. Multi-omics integration established a diagnostic biomarker model, laying a theoretical basis for microbiome-targeted weight management interventions. Full article

Urban trees are increasingly deployed as nature-based infrastructure to mitigate heat and manage stormwater, yet quantitative guidance on how species traits and site context shape transpiration remains fragmented. We conducted a systematic metadata analysis of seven field studies that measured daily transpiration rate in urban settings using heat-pulse methods. The units and spatial scales reported were harmonized with the sap flow density across active sapwood (J_s, g H₂O/${\mathrm{cm}}^2$/day) by converting reported stand transpiration and the outer 2 cm of sapwood sap flux using established Gaussian radial distribution functions for angiosperms and gymnosperms, which account for the non-linear decline in sap flux from the vascular cambium to the heartwood boundary. We then summarized distributions and tested group differences with Kruskal–Wallis and Dunn post hoc comparisons across wood anatomy, climate, soil texture, and planting configuration. Conifers exhibited significantly lower median $J_s$ (39.76 g/${\mathrm{cm}}^2$/day) than angiosperms, while the ring-porous group (median $J_s$ = 92.25 g/${\mathrm{cm}}^2$/day) and diffuse-porous groups (median $J_s$ = 96.70 g/${\mathrm{cm}}^2$/day) had similar distributions overall. Climate-modulated responses within wood anatomy groups differed, with diffuse-porous species exhibiting the highest median $J_s$ (152.59 g/${\mathrm{cm}}^2$/day) in semi-arid regions, ring-porous species maintaining comparatively stable median $J_s$ across climates (varying slightly between 80.72 and 99.32 g/${\mathrm{cm}}^2$/day), and conifers reaching their highest median $J_s$ (69.90 g/${\mathrm{cm}}^2$/day) in humid continental sites. Soil texture effects were consistent with moisture availability: sandy loam generally reduced $J_s$ relative to loam or silt loam for conifers and diffuse-porous species. Across anatomies, single trees transpired more than clustered trees or closed canopies. For example, planting as single trees increased median $J_s$ by 86% in conifers (from 33.01 to 61.37 g/${\mathrm{cm}}^2$/day) and by 45% in diffuse-porous species (from 81.31 to 118.25 g/${\mathrm{cm}}^2$/day). These results provide actionable ranges and contrasts to inform species selection and planting design for urban greening and runoff reduction, while highlighting data gaps for future research. Ultimately, by matching specific wood anatomies and planting configurations to local soil and climatic conditions, urban planners and ecohydrologists can strategically optimize urban forests to maximize targeted ecosystem services. Full article

Ledum palustre (L. palustre) is widely used in drug development because of its antibacterial and analgesic effects. However, wild L. palustre is often affected by wildfires, resulting in unstable yields. Forest fires represent a major disturbance in northern forest ecosystems and profoundly affect shrub vegetation and its associated rhizosphere microbial communities. In this study, we investigated a fire chronosequence (1991, 2004, 2012, 2017, and 2020) to systematically examine the morphological traits of L. palustre, rhizosphere soil physicochemical properties, and microbial community characteristics and to identify the key drivers underlying these patterns. The results revealed that postfire recovery time significantly influenced the morphological traits of L. palustre. The biomass, branch number, basal diameter, and plant height of the shrubs at the 1991 burned site increased by 270.49%, 36.11%, 79.32%, and 191.36%, respectively (p < 0.05). From unburned soils, 29 bacterial and 29 fungal isolates were obtained, with Bacillus sp. and Oidiodendron sp. being the dominant culturable bacterial and fungal taxa, respectively. With increasing postfire recovery time, soil moisture, total nitrogen, ammonium, nitrate, soil organic carbon, acid phosphatase (AP) and N-acetyl-β-D-glucosaminidase (NAG) activity significantly decreased. Early fire disturbance markedly altered soil microbial abundance and community composition, leading to an overall decrease in bacterial α diversity. The bacterial community structure at the 2020 burn site and the fungal community structure at the 2012 burn site significantly differed. Mantel tests revealed significant positive correlations between branch number and basal diameter (p < 0.01) and significant negative correlations between plant height and stem density (p < 0.001). Soil carbon and hydrolysable nitrogen were significantly positively correlated with AP and NAG activities (p < 0.001). Moreover, soil physicochemical properties significantly shaped soil microbial community structures, with bacterial communities in early postfire sites driven by total carbon and nitrogen (p < 0.05), whereas fungal communities in the 2012 burned site were influenced primarily by β-N-acetylglucosaminidase (BG) activity (p < 0.05). Fire disturbance drives successional changes in the rhizosphere microbial community structure and function by altering the soil nutrient status and enzyme activity, which in turn influences the morphological traits of L. palustre. This study provides a theoretical basis for improving the yield of L. palustre by exploring the variation in rhizosphere microorganisms. Full article

The conventional window for ultrasonic pregnancy diagnosis in sows is 22–25 days post-insemination, which often results in missed opportunities for the optimal re-insemination of non-pregnant sows and elevated production costs. This present study aimed to establish an early pregnancy detection method for sows at 12–18 days post-insemination, thereby providing a reference for efficient reproductive management. Saliva, urine and vaginal secretions were collected from sows during this period, and seven metabolites were quantified. Seven machine learning models were employed for data analysis, after which the optimal combination was determined, and the detection protocol was refined using recursive feature elimination. The results revealed that the majority of metabolites in saliva and urine differed significantly between pregnant and non-pregnant groups (p < 0.05). Among the models evaluated, the random forest algorithm exhibited the best predictive performance, with accuracy ranging from 0.59 to 1.00. Saliva sampled at 17 days post-insemination was identified as the optimal diagnostic medium, and 100% prediction accuracy was achieved by measuring only three metabolites: Glc, Ste, and Xan. The diagnostic approach established in this study allows pregnancy detection 5–8 days earlier than conventional methods, with the additional benefits of non-invasive sampling and minimal stress to sows. Accordingly, it provides a novel reference for enhancing the efficiency of swine production. Full article

We analyzed wildfire hazard profiles across the Hyrcanian temperate forests of northern Iran (Guilan Province) by simulating a large set of wildfires with FlamMap MTT. We first derived geospatial data on terrain, fuel models, weather conditions, and historical wildfire occurrence (1992–2022) for the study area. We stratified fire weather conditions and fuel moisture based on the bioclimatic classification of the study area, considering observed extreme fire weather, as well as observed and random fire ignition locations for the simulations. The wildfire simulations were used to estimate burn probability (BP), conditional flame length (CFL), fire size (FS), and crown fire probability (CFP). BP ranged from 0 to 5.0 × 10⁻², with mean values of 1.3 × 10⁻³ and 1.1 × 10⁻³ for observed and random scenarios, respectively. The mean value of CFL from random ignition simulations (0.78 m) was substantially higher than that obtained in the observed ignition simulations (0.54 m), ranging from 0 to 6.75 m. We evidenced significant differences between observed and random ignition simulations for all wildfire hazard metrics. The highest wildfire hazard profiles were observed in the Cold-Mountainous bioclimatic zone under the random ignition simulations. On average, the annual number of anthropic structures threatened by wildfires ranged from 97 (observed scenario) to 123 (random scenario). This research provides detailed and spatially explicit fire hazard and exposure maps to inform fire modeling, land management, and policy actions. Full article

The once-extensive coastal forested wetlands (CFWs) of the Mississippi River Delta (MRD) are declining under the combined pressures of pervasive hydrologic change, unregulated harvesting, relative water level rise (due to the combination of geological subsidence and sea-level rise—SLR), and climate change. We synthesize here over 50 years of research conducted in the MRD to examine the history of the CFWs and their management, their ecosystem functions and services, and the nature, extent, and severity of ongoing changes. Seedling recruitment failure and increasing salinity levels are the most immediate threats to forest persistence, necessitating management that restores hydrologic function and sediment and nutrient supply to allow seedling survival and minimizes saltwater intrusion. Collectively, the evidence indicates that managed inflows can bolster accretion and sustain forest function, and long-term resilience requires hydrologic restoration at landscape scales coupled with site-level actions that secure recruitment and address local degradation trajectories. These include freshwater and sediment introduction, protection from herbivory, and, in some cases, planting. Our research findings have important implications for worldwide CFWs, and tidal freshwater ecosystems in general, which occur mainly in tropical deltas. Full article

Wildfires are among the key drivers of transformation in boreal ecosystems; however, the mechanisms of postfire recovery in the arid regions of Eurasia remain insufficiently understood. The aim of this study was to identify the role of burn severity and associated edaphic and hydrological factors in shaping the natural regeneration trajectories of Scots pine forests in the Kostanay region of northern Kazakhstan. This study is based on the integration of field data on seedling regeneration and soil conditions with the analysis of long-term satellite-derived indices (NDVI, NDMI, and NBR). Sample plots were grouped according to fixed burn severity classes, which enabled a consistent statistical comparison and reduced the interpretative ambiguity that has characterized previous studies in the region. The results indicate that pine forest regeneration is most successful under low and moderate burn severity, where seed sources are preserved and favourable moisture conditions are maintained. In contrast, high burn severity leads to a reduction in regenerative potential and a shift in recovery trajectories toward deciduous or sparsely vegetated communities. The spectral indices derived from the remote sensing data strongly agreed with the field-based indicators, confirming their suitability for assessing postfire vegetation dynamics. Soil properties act as important modifying factors of recovery processes, particularly under conditions of limited water availability. These findings enhance the current understanding of postfire recovery mechanisms in the arid part of the boreal zone and highlight the need for differentiated management of postfire landscapes. The integration of field observations with remote sensing data provides a robust framework for monitoring and forecasting recovery processes under an increasingly intensified fire regime. Full article

Background/Objectives. Head and neck mucosal melanoma (HNMM) is a rare, aggressive malignancy with poor outcomes and limited evidence to guide prognostication and treatment. A detailed assessment of long-term survival and prognostic factors is needed to inform clinical management and staging. This work aimed to describe outcomes and prognostic factors in HNMM patients treated over 45 years. Methods. This was a retrospective observational cohort study of consecutive patients treated at a tertiary referral center in Italy from 1975 to 2020. Random-forest-based screening informed covariate selection for Cox models. Endpoints were overall survival (OS), disease-free survival (DFS), and post-recurrence DFS (prDFS). Associations with clinical and pathological variables were evaluated using Kaplan–Meier estimates, log-rank tests, and multivariable Cox regression. Results. Among 112 patients (median follow-up, 121.1 months), 3-/5-year OS was 42.8%/28.0%, DFS 20.5%/13.2%, and 1-/3-year prDFS 36.7%/10.9%. Ulceration was associated with worse OS (HR 2.12; 95% CI 1.05–4.26) and DFS (HR 2.23; 95% CI 1.16–4.28). Male sex showed a trend toward poorer OS and DFS. Regional lymph-node treatment correlated strongly with OS and prDFS (overall p < 0.001), with neck dissection indicating unfavorable risk (OS HR 5.22; 95% CI 2.39–11.40). Conclusions. HNMM remains a high-mortality disease with frequent recurrence. Ulceration and nodal involvement were key adverse prognostic factors, while surgery was associated with improved survival. The findings support incorporating ulceration into future staging and highlight the potential for durable control through salvage surgery. Further investigation of treatment intensification, biomarkers, and multimodal strategies is warranted. Full article

Using a coordinated multi-year monitoring dataset collected during the 2020–2024 partial-opening management period, we examined benthic diatom assemblages across the Sejong, Gongju, and Baekje weirs in the Geum River, Republic of Korea. Seasonal surveys at eight stations were used to evaluate spatiotemporal variation in water quality and benthic diatom community structure under this hydrological management regime. Annual basin-wide averages showed gradual interannual changes in water quality, including declines in total phosphorus, total nitrogen, chlorophyll-a, turbidity, and biochemical oxygen demand after 2021, accompanied by increased dissolved oxygen. Diatom community indices based on relative-abundance data showed corresponding temporal variation, with decreased dominance and increased Shannon diversity, evenness, and taxon richness. Ordination analyses indicated gradual differentiation between the earlier (2020–2021) and later (2022–2024) monitoring groups within the study period, whereas random forest models showed limited explanatory power and were treated as exploratory. Overall, the results support benthic diatoms as sensitive descriptors of ecological change in flow-regulated monsoonal rivers while underscoring the value of long-term monitoring where true pre-intervention biological baselines are unavailable. Full article

Sustainable forest management is increasingly recognized as an important climate change mitigation strategy because forests capture and store large amounts of carbon. This study presents a regional framework that integrates LiDAR data, forest cartography, and Life Cycle Assessment (LCA) to quantify biomass-related carbon dynamics and greenhouse gas emissions associated with forest management operations. The methodology was applied to the Autonomous Community of La Rioja (Spain) for the period 2010–2016 using public LiDAR campaigns, the Forest Map of Spain, and inventory data for reforestation and logging operations. Results show that above-ground biomass increased from 4,537,956 t in 2010 to 7,092,890 t in 2016, which corresponds to an increase of 1,200,819 t C in above-ground carbon stock. A complementary first-order estimate based on IPCC default root/shoot ratios suggests that total living biomass carbon (above- plus below-ground) increased by approximately 1,495,269 t C during the same period. In parallel, LCA results indicate that logging has substantially higher operational impacts than reforestation, particularly in terms of global warming potential. Even under a conservative scenario in which part of the carbon removed through logging is returned to the atmosphere, the regional balance remains net negative in CO₂-equivalent terms, indicating a net sink over the analyzed period. However, the approach has important limitations, including the absence of independent field validation, stand-age stratification, and explicit soil-carbon accounting. Full article