Search Results (285)

Regional landslide susceptibility screening can support sustainable land management by improving the allocation of limited geological-survey and field-verification resources toward locations requiring further investigation. An uncertainty-aware framework integrating ensemble learning, spatial validation, inter-model agreement, and probability uncertainty was proposed for regional landslide investigation prioritization in the Hualong–Xunhua region of the Upper Yellow River Basin, China. Using 281 manually interpreted landslide locations, 281 pseudo-absence samples, and twelve conditioning factors, four tree-based models (Random Forest, XGBoost, LightGBM, and CatBoost) were developed, and their probability outputs were combined through arithmetic averaging. Model performance was evaluated using stratified random validation and 10 km spatial block validation, while agreement and probability divergence were incorporated to define screening priorities. Random-validation AUC values ranged from 0.885 to 0.898, and spatial-validation AUC values ranged from 0.852 to 0.873; the Mean Ensemble achieved AUC values of 0.8961 and 0.8637, respectively. NDVI showed the highest mean normalized permutation importance (0.754), although sensitivity analysis demonstrated that other factors retained useful discrimination after removing NDVI and land-cover information. The final priority screening zone covered 862.72 km² (19.04% of the valid mapped area). The framework provides a transparent decision-support approach for sustainable land management in mountainous regions by identifying investigation priorities while accounting for model consistency and uncertainty. Full article

High-precision, multi-tier categorical soil classification faces critical bottlenecks, including the neglect of spatial context by conventional pixel-based models, the error cascade propagation phenomenon in multi-level classification networks, and the disruption of geophysical directional anisotropy by traditional geometric data augmentation. To address these challenges, in this study, we propose a multi-level soil classification model based on MTSC-ResNet-Trans, which organically couples residual convolutional blocks with a 3-layer Transformer encoder to model long-range spatial dependencies. The framework integrates a geospatial-safe data augmentation pipeline to preserve the topological fidelity of absolute geographic coordinates alongside four decoupled parallel multi-task classification heads to substantially suppress inter-level error propagation. Evaluated in the Southern Foothills of Qilian Mountains using 18 environmental covariates, the framework achieves an Overall Accuracy of 0.8931 at the Great Group level under conventional random splitting. Under a distance-stratified spatial evaluation—which isolates the contribution of spatial autocorrelation to accuracy estimates—the framework maintains robust performance, with MTSC-ResNet-Trans consistently outperforming pixel-based baselines (Random Forest) by approximately 3.7 percentage points even at spatial separation distances exceeding 400 m. This protocol transparently decomposes predictive accuracy into a component attributable to spatial proximity and a component reflecting reduced spatial proximity performance. Across the four taxonomic levels, accuracy decay is suppressed to 5.11%. Although spatial-block cross-validation indicates a lower regional extrapolation accuracy (OA = 0.7389 ± 0.0671), the decoupled parallel framework provides an effective and robust baseline for high-resolution regional digital soil mapping. Full article

Quantifying groundwater depletion and aquifer–stream interactions in intensively irrigated basins remains a critical challenge for sustainable water management. This study numerically examines these issues in the San Joaquin River Basin (SJRB) using the coupled SWAT+ gwflow framework. To manage computational demands across the 28,435 km² basin, upland contributing areas were represented as inlet point sources at five gauging stations at the outlets of each of the upstream subbasins, reducing computational cost by ~45% while preserving hydrological fidelity. A fine-resolution (300 m × 300 m) model was developed for the downstream valley floor, with release from reservoirs and mountain-block runoff incorporated as a surface boundary condition. We applied the water allocation module gwflow to withdraw water from the aquifer to satisfy crop water stress based on user-defined crop water coefficient values and compared the simulated and observed spatially distributed groundwater head values. Results show a persistent decline in groundwater storage and water levels from 2006 to 2022, with extraction exceeding recharge by ~40%, resulting in an average annual storage loss of 450,175 acre-feet. Intensive pumping also caused hydraulic disconnection between groundwater and some streams, where pronounced cones of depression developed in Turlock, Merced, and western Madera subregions. These findings indicate that targeted managed aquifer recharge and reductions in groundwater extraction in the most severely depleted subregions, are essential to restore aquifer–stream connectivity and ensure long-term water security in the SJRB. Overall, integrating gwflow with SWAT+ provides critical insights into groundwater dynamics in California’s Central Valley and supports sustainable groundwater management. Full article

This study takes the Ankang area in the Qinling–Dabashan Mountains on the northern margin of the Yangtze Block as an example to investigate whether effective shale gas preservation conditions can exist in large-scale thrust nappe belts at convergent plate margins—a critical scientific question. The aim is to provide new concepts and models for shale gas exploration in tectonically complex regions. An integrated approach combining surface geological mapping, geophysical surveying (2D seismic and wide-field electromagnetic method), calibration of a key borehole (ZBDR01), and geochemical analysis was employed to reconstruct the deep geological structure and evaluate the hydrocarbon generation potential and reservoir characteristics of the target shale interval. The results reveal a relatively gentle, weakly deformed “structural stability window” beneath the Zhongbao Fault, a major thrust nappe surface. Within this window, strata dip at low angles and faults are sparse, exhibiting a significant stress-shielding effect. The Lower Cambrian Niutitang Formation shale within this window is well preserved, characterized by high total organic carbon (average TOC: 4.26%) and moderate thermal maturity (average R_o = 3.02%), falling within the effective shale gas generation window. In contrast, the Lujiaping Formation shale in the hanging wall of the fault, though widely distributed, shows excessive thermal maturity and poor reservoir properties. The study demonstrates that the “stress-shielding” effect is the core mechanism controlling the formation of this stability window and proposes a new “tectonic shielding” accumulation model. This model elucidates how the thrust nappe body itself acts as a thick regional caprock, which together with lateral sealing by the fault zone forms a composite seal-cap system, ensuring in situ preservation of shale gas under a strongly tectonic background. It is concluded that local preservation units can form in the footwalls of thrust nappe belts at convergent plate margins due to stress shielding, challenging the conventional view that intensely deformed zones are unfavorable for shale gas preservation. This research not only provides a new direction and model for shale gas exploration in the tectonically complex Qinling–Dabashan region but also offers important theoretical and technical insights for unconventional hydrocarbon exploration in similar tectonic settings globally. Full article

Objective: To address small bolt defect targets, complex background interference, and limited edge deployment in hilly and mountainous rail transport environments, a detection method balancing accuracy, lightweight design, and real-time performance was proposed. Methods: A track image dataset containing missing bolts, loose bolts, and missing nuts was constructed. Based on YOLOv11m, HFERBC3K2 was developed by replacing the standard bottleneck in C3K2 with a High-Frequency Enhancement Residual Block to strengthen edge, texture, and local structural feature extraction. A Spectral Enhanced Feed-Forward module was introduced into C2PSA to form SEFFNC2PSA, enhancing defect-related frequency components and suppressing background interference through adaptive frequency-domain modulation. The integrated model was named HFSE-YOLOv11. Channel-level structured pruning was then applied, and the model with a pruning ratio of 0.5 was named LHFSE-YOLOv11. Results: On the validation set, HFSE-YOLOv11 achieved 91.7% precision, 93.4% recall, 91.3% mAP@0.5, and 80.2% mAP@0.5:0.95, improving upon YOLOv11m by 3.6, 2.2, 1.2, and 3.1 percentage points, respectively. After pruning, LHFSE-YOLOv11 had 15.9 M parameters, 53.8 GFLOPs, and a 32.5 MB model size, representing reductions of 16.3%, 14.3%, and 11.7%, while mAP@0.5 and mAP@0.5:0.95 decreased by only 0.3 and 0.9 percentage points. On the independent test set, it achieved 91.7% precision, 93.4% recall, 91.0% mAP@0.5, 79.3% mAP@0.5:0.95, and 81.5 FPS, outperforming all compared models in the four detection metrics. Conclusion: LHFSE-YOLOv11 balances accuracy, efficiency, and model size, supporting deployment on vehicle-mounted inspection terminals and resource-constrained edge devices. Full article

To gain a deeper understanding of the lower-atmospheric dynamic characteristics in the transition zone on the northeastern margin of the Tibetan Plateau, high-resolution wind profile data collected by a Doppler wind lidar (DWL) at Yinchuan Hedong International Airport from 2021 to 2023 were used to analyze the vertical structure, seasonal variations, and diurnal characteristics of the low-height wind field and wind shear in this region. The results indicate that (1) the data acquisition rate (DAR) below 1.5 km is generally high, exceeding 90% during most periods, and decreases monotonically with height; the 90% DAR contour height exhibits clear seasonal and diurnal variations, with the largest diurnal amplitude in summer and the smallest in winter. (2) The middle- and low-height wind fields are jointly modulated by topographic forcing and local circulations. Below 0.4–0.7 km, north–northeast and south–southwest winds prevail across all seasons, which is consistent with the blocking and splitting effects of the Helan Mountains. At 42 m, the wind direction shows a marked diurnal transition that may reflect the combined influence of the Helan Mountains’ bypass flow, mountain–plain circulation, and thermal contrasts between the Yellow River and surrounding desert/plain surfaces. (3) Horizontal wind speeds are predominantly concentrated below 6 m s⁻¹, and the development height of this low-wind-speed zone varies seasonally. The vertical velocity statistics show weak positive values in parts of the observed layer, but these signals are interpreted cautiously because vertical-velocity retrieval is subject to additional uncertainty. (4) The low-level wind shear intensity reaches its peak below 100 m and generally exhibits a U-shaped vertical distribution; severe wind shear below 100 m occurs most frequently from nighttime to early morning during May–October, whereas its occurrence frequency is lowest in winter. These findings provide observational evidence for aviation meteorological support and boundary-layer studies in semi-arid regions of Northwest China. Full article

This study utilizes Omani limestone waste powder (CaCO₃) sourced from the mountain ranges of the Sultanate of Oman as an economical natural reinforcement for scrap-based aluminum matrix composites (AMCs) using stir casting. The recycling of aluminum alloy from end-of-life automotive engine cylinder blocks was strengthened with limestone at volume fractions of 2.5%, 5.0%, and 7.5%. Mechanical characterization was conducted in accordance with ASTM standards (E8/E8M, E18, E23). Statistical significance (p < 0.05) was calculated using one-way ANOVA. The optimum 5.0 vol.% reinforcing fraction showed tensile strength, Rockwell hardness and Charpy impact energy of 130.9 MPa (+16.9%), 91 HRF (+19.7%) and 7.2 J (+28.6%) compared to the unreinforced scrap alloy (112.0 MPa, 76 HRF, 5.6 J). The results of Scanning Electron Microscopy (SEM) research showed that the composite of 5.0 vol.% had a uniform distribution of CaCO₃ particles and very low porosity, while the composite of 7.5 vol.% had a significant porosity (2–8 μm), interconnected microcracks and particle agglomeration. The porosity was increased with the increase of the content of the reinforcement as shown by the density experiments based on Archimedes’ principle. The maximum deviation of the experimental density from the predicted one was at 7.5 vol. % reinforcement. X-ray diffraction (XRD) confirmed the stability of the aluminum matrix structure as well as stable CaCO₃ phases without any evidence of harmful interfacial reaction products (Al₄C₃ or CaAl₂O₄). The findings confirm the optimal reinforcement ratio of 5.0 vol.% of Omani limestone, tackling both the environmental load of limestone quarrying waste and the expensive synthetic reinforcements, in accordance with the circular economy goals of Oman Vision 2040. Full article

Permanent grassland management commonly includes moderate mineral N fertilization, which can differentially affect botanical composition, dry matter (DM) yield and forage quality of grasses, legumes and forbs. We hypothesized that increasing N would enhance grass dominance and DM yield but raise fiber and reduce metabolizable energy (ME) content. We evaluated long-term N fertilization, cutting time and their interaction on DM yield and ME of grasses, legumes and forbs in a permanent mountain grassland. The field experiment, established in 1967, used a randomized block design (4 replicates, 3 cuts/year) with five treatments: unfertilized (Control), PK without N (N0), and PK plus 80 (N80), 120 (N120) or 180 (N180) kg N/ha per year; forage was sampled in 2014–2016. Increasing N raised grasses to ~92% of biomass at N180 and reduced legumes and forbs (p < 0.001). Mean DM yield rose from 2.5 (Control) to 9.8 t/ha (N180; p < 0.001). High N (N120, N180) lowered ME of grasses in later cuts (8.9 vs. 10.1 MJ/kg DM in the third cut), whereas legumes and forbs maintained high ME. Moderate N fertilization (80–120 kg N/ha) best balanced DM yield, legume and forbs contribution and ME content. Full article

Wind power integration into mountainous power grids amplifies probabilistic forecasting challenges arising from strong non-stationarity, multi-source meteorological redundancy and frequent curtailment events. To address the limitations of existing approaches, this paper proposes a multi-horizon probabilistic forecasting framework integrating multi-perspective entropy-weighted fusion, permutation-entropy-guided decomposition, and residual-anchored probability modelling. First, an Entropy-Weighted Multi-criteria Permutation Feature Importance (EW-MPFI) module fuses KSG mutual information, Tree-SHAP, and elastic-net permutation importance through entropy-based weighted aggregation, distilling 23-dimensional meteorological inputs into eight informative features while suppressing single-criterion bias. Then, a three-stage decomposition strategy applies ICEEMDAN primary decomposition, permutation-entropy and sample-entropy guided band reconstruction, and SSA secondary refinement on high-frequency components, achieving complexity-aligned multi-scale separation. Finally, a decomposition-aware patch-based Transformer backbone (DPC-Former) generates three-quantile point forecasts, upon which an NGBoost residual layer models the conditional distribution via natural-gradient optimization in the information-geometric parameter space. Case studies on a 130 MW mountainous wind farm in Sichuan, China, covering 8736 15-min samples with 566 curtailment samples (6.48% of the dataset), show that, under the partition-wise offline batch-evaluation protocol, the proposed framework achieves an NMAE of 5.21%, an NCRPS of 3.74%, and a PICP80 of 0.84 across forecasting horizons from 15 min to 4 h. Ablation analysis attributes NMAE improvements of 25.36% and 24.57% to the decomposition and feature-selection modules, respectively, while 50-seed ensembling further reduces NCRPS, NMAE, and NRMSE by 7.40%, 7.00%, and 13.70% relative to single-seed training. A fixed-checkpoint test-block diagnostic further shows limited sensitivity at approximately weekly and three-day decomposition cadences, but a material degradation at a one-day cadence. The reported metrics should therefore be interpreted as offline best-case results rather than as performance under strictly causal real-time deployment. Full article

Steep rock slopes are widely distributed in the karst mountainous regions of southwestern China, where structurally controlled rockfalls frequently threaten transportation infrastructure and human safety. Accurate identification of unstable rock blocks (URs) and quantitative assessment of their post-failure hazards remain major challenges because of complex discontinuity networks and fragmentation during rockfall motion. Taking the Zuojiaying steep rock slope in Guizhou Province as a representative case, this study integrates high-resolution UAV photogrammetry, automatic discontinuity identification, unstable rock block detection, and three-dimensional rockfall simulation to investigate the formation mechanisms and hazard characteristics of discontinuity-controlled rockfalls. A high-resolution three-dimensional terrain model was reconstructed from UAV imagery, and six dominant discontinuity sets were automatically identified using the I-MinPts-constrained DBSCAN algorithm. Combined with the Rock Occurrence Kinematic Analysis (ROKA) algorithm and Block Theory, 54 unstable rock blocks were identified, with wedge failure and toppling failure representing the dominant instability modes. The results indicate that discontinuity combinations govern both rock mass segmentation and unstable rock block geometry. Specifically, discontinuity sets J1, J3, and J5 mainly control wedge-shaped blocks, and J2 and J4 dominate columnar toppling blocks, whereas J6 further promotes the formation of isolated unstable rock blocks. Three-dimensional RockGIS simulations considering fragmentation reproduced the complete rockfall process from detachment to final deposition. The maximum travel distance, kinetic energy, and bounce height reached 395 m, 748.5 kJ, and 40.1 m, respectively. Fragmentation increased the number of rock blocks from 54 to 1013, substantially enlarging the potential impact area. A raster-based Rockfall Hazard Index (RHI) further revealed that the middle–lower slope and slope toe constitute the principal high-hazard zones, and under extreme scenarios, high-energy fragments may reach the G246 National Highway and adjacent infrastructure. This study revealed the formation mechanisms and hazard characteristics of unstable rock blocks controlled by discontinuity combinations in the study area, providing a case reference for rockfall hazard identification and mitigation on similar high-steep rock slopes in karst mountainous regions. Full article

Rockfall hazards are frequently observed within mountainous road networks. Significant uncertainties regarding the optimal selection of evaluation units and spatial modeling scales are still being identified in this field. A comprehensive comparative framework for rockfall susceptibility mapping is presented in this study, using Wufeng County as the empirical study area. Five evaluation scenarios were constructed to systematically isolate the independent predictive contributions of the spatial domain, the mapping unit morphology, and the physics-informed engineering proxy. These scenarios included a whole-county macro-scale raster; three multi-scale road buffers with widths of 1 km, 2 km, and 3 km; and an object-oriented vector road evaluation unit (REU) framework. To parameterize localized engineering-induced risks, a physics-informed feature defined as the theoretical slope-cutting height ($H_{\mathrm{cut}}$) was structurally introduced into the vector-based assessment. Thirteen representative machine learning, deep learning, and statistical algorithms—including Random Forest, LightGBM, and TabNet—were systematically cross-examined under both unconstrained splits and strict Leave-One-Road-Corridor-Out Validation (LORCOV) protocols. The empirical multi-metric sensitivity analysis explicitly decouples the three structural effects. First, isolating the effect of the spatial domain reveals that restricting the validation extent from a broad countywide area to a narrow road corridor purges unperturbed background terrain noise, shifting the focus from easy negatives to geomorphological hard negatives. Second, evaluating the independent effect of the evaluation unit demonstrates that transitioning from continuous raster pixels to homogeneous vector REUs successfully resolves the terrain smoothing effect, precisely characterizing sharp geomechanical gradients adjacent to cut slopes. Third, isolating the effect of adding $H_{\mathrm{cut}}$ proves that this engineering indicator drives the primary descriptive gain, enabling tree-based ensembles to achieve a peak baseline AUC of 0.7763 and maintain a robust spatial validation AUC of 0.6129 under strict geographic block constraints, whereas legacy deep learning architectures exhibit an inductive bias mismatch on small-scale tabular records. Rather than asserting a single optimal paradigm, this coordinated feature–unit matching framework provides transport authorities with a highly calibrated, target-tiered decision matrix to optimize localized public works safety budgets and protect critical linear infrastructure assets. Full article

Accurate delineation of current landslide extent is important for hazard assessment, sustainable watershed management, and disaster risk reduction in tectonically active mountainous regions. This study presents a pilot machine learning framework for pixel-level landslide classification in the Laonung (Laonong) Creek Watershed, southern Taiwan, using very high-resolution digital elevation model (DEM) derivatives and SPOT-6 multispectral imagery. Thirteen geomorphometric and spectral features, including slope, curvature, and six spectral indices derived from SPOT-6 bands, were extracted from 96 landslide-containing tiles within a pilot subregion of the watershed; no landslide-free tiles were included in model training or evaluation. Landslide annotations followed a geomorphic-unit delineation protocol in which optical imagery provided the primary evidence of current activity and DEM-derived hillshade supported boundary refinement. Three classifiers were evaluated using column-quartile spatially blocked four-fold cross-validation, with each fold comprising a geographically contiguous range of columns, to reduce spatial leakage: logistic regression (LR), random forest (RF), and XGBoost. All three models substantially outperformed the no-skill baseline for the resampled evaluation dataset (average precision, AP $=0.250$), achieving mean AP values of $0.854\pm 0.040$, $0.858\pm 0.033$, and $0.846\pm 0.035$ for LR, RF, and XGBoost, respectively. The convergence of linear and nonlinear model performance suggests that the dominant discriminatory signal is largely captured by relatively simple spectral and topographic predictors within this pilot dataset, rather than reflecting a general property of landslide classification. SHapley Additive exPlanations (SHAP) analysis across all four spatial folds identified SPOT-6 Band 3 (Red) as the dominant predictor in every fold, with NDVI a robust secondary predictor, consistent with the spectral characteristics of fresh bare-soil landslide surfaces and with the optical cues used in the annotation protocol. The results are interpreted in the context of the pilot dataset’s limited spatial extent, the resampled class distribution used for model evaluation, and unquantified label uncertainty. This study provides a transferable methodological baseline for future, larger-scale landslide classification analysis in the Laonung Creek Watershed and highlights the potential contribution of spatially explicit landslide mapping to sustainability-oriented disaster management. Full article

Assessing earthquake-induced landslide (EQIL) susceptibility is essential for hazard mitigation in mountainous regions. While background hydrological variations influence slope stability, long-term mean soil moisture is rarely incorporated into deep learning-based landslide susceptibility mapping (LSM). This study proposes a hybrid Convolutional Neural Network and Swin Transformer (CNN-SwinT) framework that integrates long-term mean soil moisture as a static covariate to represent persistent background moisture conditions. The model couples the local spatial feature extraction of CNNs with the hierarchical contextual representation of Swin Transformers to capture multi-scale spatial dependencies. Using Minxian County of China as the study area, thirteen conditioning factors were selected via multicollinearity and information gain ratio analyses. The dataset was split into training (70%) and validation (30%) sets. Performance comparison against standalone CNN and SwinT models revealed that the hybrid CNN-SwinT achieved the highest accuracy (0.856) and AUC (0.95), with predicted high-susceptibility zones closely aligning with historical inventories. However, these reported metrics reflect a random, spatially non-independent split, and spatial block cross-validation is recommended for future operational deployment. The results demonstrate that incorporating long-term mean soil moisture provides critical complementary hydrological information that enhances predictive performance. These findings indicate that the proposed hybrid framework is reliable and effective for high-resolution EQIL susceptibility mapping. Full article

East Asian spring dust activity has generally weakened since the early 2000s (Theil-Sen trend −1.07 × 10⁻⁶ yr⁻¹, significant over 62% of the domain), but severe events continue to occur when synoptic forcing, source-region dryness, and terrain-guided transport are favorably coupled. This study places the 19–23 March 2023 East Asian dust storm within this 2000–2024 background and provides an integrated three-dimensional analysis of its transport, vertical structure, and topographic controls. The event developed as a dual-source relay-convergence process: Taklamakan Desert dust was emitted first on 19 March and transported southeastward along the Hexi Corridor, while Mongolian Plateau dust intensified on 21 March and mainly affected North China. Independently calibrated, PM10-cross-validated FLEXPART-WRF trajectory arrays (R = 0.74–0.81) show Taklamakan contributed 100% of the calibrated near-surface dust mass at Lanzhou and Mongolian 98% at Beijing during each receptor’s event peak window. Four independent dynamical diagnostics quantify topographic control, showing the Helan Mountains attenuate westward-approaching Taklamakan dust by 23% across the range. TROPOMI AAI, CALIPSO, ground PM10, and CAMS EAC4 jointly corroborate multi-level cold-vortex/trough-frontal coupling and terrain blocking as the controlling mechanisms, demonstrating that extreme dust episodes can still occur under a weakening long-term background when dynamic lifting, dual-source activation, and topographic channeling act together. Full article

Environmental heterogeneity can promote niche differentiation among closely related species, yet ecological segregation may not be mirrored by traits measured during a single field campaign. We tested whether edaphic differentiation exceeded morphological and ecophysiological divergence in Rosa acicularis, R. alberti, and R. spinosissima across the Altai and Tarbagatai Mountains of eastern Kazakhstan. The dataset comprised 40 field records representing 25 geographic coordinates and 30 species–site combinations. Plant height, crown width, chlorophyll fluorescence, chlorophyll-related outputs, and soil pH, electrical conductivity, total nitrogen, and total carbon were analysed using FDR-controlled univariate tests with Dunn post hoc comparisons, PERMANOVA, reduced-variable PCA, partial redundancy analysis, correlation analysis, and small-sample explanatory models with leave-one-out validation. Within Tarbagatai, species identity explained 48.6% of multivariate soil variation but only 9.2% of plant-trait variation. Omnibus differences in pH, nitrogen, and carbon remained significant after correction, and the species–site analysis also supported EC differentiation. The soil block explained no significant independent component of plant-trait variation after species and mountain system were taken into account. Rosa spinosissima was taller in the Altai at the observation level, and a species-plus-region model showed moderate cross-validated performance for height; models for other traits had little predictive value. These results support edaphic niche differentiation accompanied by limited divergence in field-measured traits. Because the study was observational and temporally restricted, the findings demonstrate environmental association rather than adaptation. Still, they identify testable hypotheses and support environmentally stratified conservation of wild Rosa genetic resources. Full article