Search Results (567)

Urban air pollution, especially the problem of PM_2.5, is one of the major health challenges facing the planet today. To provide accurate PM_2.5 predictions despite data noise and missing data, the authors proposed a deep learning model. We constructed a Stacked Autoencoder–Convolutional Neural Network–Bidirectional Long Short-Term Memory–Long Short-Term Memory (SAE-CNN-BiLSTM-LSTM) model that (1) utilises convolutional layers to extract spatial features from the input data, (2) employs bidirectional LSTM layers to capture long-term temporal dependencies, and (3) utilises an autoencoder to learn latent representations of the data to mitigate the effects of missing data. The model was trained on a large dataset of hourly measurements of air quality and meteorological parameters collected between 2018 and 2020 in Klang, Malaysia. The performance of the model on data that were not used during training was evaluated using a range of metrics. The SAE-CNN-BiLSTM-LSTM model achieved a test RMSE of approximately 11.97 µg/m³ and an R² statistic of approximately 0.85 for PM_2.5 concentrations, outperforming the other models tested on the same datasets. The additional metrics of MAE, MAPE, Mean Bias Error, and Index of Agreement confirmed the model’s accuracy and low bias in the prediction of air pollution levels. Statistical tests, such as the Diebold–Mariano test, confirmed the significance of the model’s accuracy over the CNN-LSTM models. These findings indicate that the proposed model effectively captures the dynamics of the air pollution data. The proposed model structure efficiently achieved an accurate and lightweight model for urban air pollution forecasting. Full article

The growing availability of genomic resources is changing how genetic diversity is studied in Vitis vinifera L. At the same time, it has become increasingly clear that a single reference genome cannot fully represent the complexity of a species characterised by high heterozygosity, clonal propagation and a long history of diversification. Recent grapevine pangenomes, super-pangenomes and graph-based resources have revealed forms of variation that are often overlooked in conventional reference-based analyses, including structural variants and gene presence–absence variation. Rather than providing another inventory of available datasets, this review examines how continued reliance on a single reference genome may influence the interpretation of grapevine diversity and what can be gained from a broader pangenomic perspective. Drawing on recent studies in grapevine and other crops, we discuss how these approaches are beginning to improve the representation of genetic diversity, uncover biologically relevant variation and strengthen links between genomic information and adaptive traits. We also examine the challenges that still limit their practical use, particularly the integration of genomic resources with functional studies and breeding programmes. In the end, the value of pangenomics will probably depend not only on generating additional genomic resources, but also on how effectively these can be translated into tools that support grapevine conservation, climate adaptation and varietal improvement. Full article

LiDAR-based place recognition is critical for long-term autonomous navigation in Global Navigation Satellite System (GNSS)-denied environments, yet existing methods struggle to balance accuracy and efficiency under substantial yaw rotations. This paper proposes a robust framework based on a multi-channel polar bird’s-eye-view (BEV) representation. Under yaw-dominated revisits, the polar BEV image transforms yaw rotation into cyclic column shifts, providing a useful structural prior for rotation-equivariant feature extraction. Raw point clouds are projected onto polar BEV grids encoding density, height, and intensity. A rotation-equivariant feature extractor comprising a Radial Compression Module and a rotation-equivariant Transformer module captures long-range azimuthal dependencies via Conditional Positional Encoding and Circular Relative-Position Bias. The equivariant features are aggregated by NetVLAD into a compact global descriptor, trained end-to-end with a hard-example mining triplet loss. Extensive experiments on the public KITTI and NCLT datasets, as well as our self-constructed LiDAR Place Recognition Revisit (LPRR) dataset, demonstrate competitive performance on KITTI and superior performance on NCLT and LPRR among the compared methods. The proposed framework achieves a favorable trade-off between performance and computational cost, and shows promising cross-dataset generalization on the evaluated NCLT and LPRR datasets without fine-tuning. Full article

Understanding ecosystem carbon processes relies heavily on the reliable assessment of gross primary productivity (GPP) yet remains challenging in the Inner Mongolia grasslands due to data scarcity and high uncertainty among existing products. We developed a ConvTransformer-based framework that exploits complementary information from satellite observations and meteorological datasets to enhance the representation of complex spatiotemporal dependencies in grassland ecosystems. Grounded in leave-one-site-out cross-validation across six eddy covariance sites, the model achieved average performance metrics of R² = 0.59, RMSE = 1.40 g C m⁻² d⁻¹, Bias = −0.31 g C m⁻² d⁻¹, and KGE = 0.46, outperforming traditional machine learning models (RF, GBRT, and SVR) as well as the light use efficiency model (EC-LUE) in both accuracy and robustness. Using this framework, we generated a daily GPP dataset at spatial granularity of 1 km for the Inner Mongolia grasslands from 2003 to 2018. The results reveal a clear spatial gradient, with GPP decreasing from southeast to northwest. Comparisons with established products, including FLUXCOM, BESS V2, and PML V2, show strong spatial consistency and reduced discrepancies, supporting the reliability of the estimates. Overall, the proposed framework provides an effective approach for characterizing regional carbon dynamics and supports long-term ecological monitoring in semi-arid regions. Full article

Deterministic radio wave propagation models provide high accuracy in complex outdoor environments but remain computationally impractical for large-scale network planning and spectrum management. In contrast, empirical and hybrid models offer low complexity at the expense of reduced accuracy, systematic bias, and limited terrain sensitivity. This paper proposes a unified delta learning framework that enhances fast baseline propagation models by learning a data-driven correction toward a deterministic Parabolic Equation Modeling (PEM) reference. A key novelty lies in a compact, physics-informed feature representation that replaces the full terrain profile with an 18-dimensional vector combining local geometric descriptors, global terrain characteristics, and baseline responses, enabling accurate correction with low-dimensional input. The study also provides the first systematic investigation of delta-based correction across multiple widely used propagation models. The framework is evaluated for free-space propagation, ITU-R P.1546, ITU-R P.1812, and ITU-R P.452 using ridge regression, kernel ridge regression, gradient boosting regression trees, and a neural network model. Model performance is assessed in terms of error reduction, bias mitigation, robustness across learning algorithms, and profile-level generalization to previously unseen propagation paths within the considered terrain categories. Results show substantial error reduction, with up to twofold improvement for simpler baseline models and consistent gains for hybrid models, while preserving computational efficiency. Full article

Global climate change is causing an increase in extreme rainfall events, which impacts the risk of hydrometeorological disasters. To support disaster mitigation and early warning systems, accurate and reliable rainfall predictions are required. Although ensemble forecasting is widely used to model atmospheric uncertainty, raw ensemble results often exhibit insufficient bias and dispersion. Therefore, post-processing techniques are needed to improve the quality of probabilistic predictions. The most commonly used calibration method is Bayesian Model Averaging (BMA). This study conducted a scoping review of peer-reviewed papers on ensemble forecast calibration using BMA, based on the PRISMA-ScR framework. Furthermore, this study presents a comprehensive bibliometric analysis involving co-authorship networks of productive authors and bibliometric maps with clustered terms. A total of 35 relevant articles were identified from 49 screened publications. The bibliometric analysis revealed that “ensemble forecasting” and “Gaussian distribution” are the most dominant terms in the research network, indicating that Gaussian-based approaches remain more widely used in ensemble forecast calibration studies. In contrast, studies explicitly applying Gamma-based approaches are still relatively limited despite their relevance for modeling asymmetric rainfall data. The results obtained in this study highlight the importance of developing and integrating more appropriate probability distributions, such as those within the Extreme Value Theory framework, into BMA models. These findings suggest that the selection of appropriate probabilistic distributions in BMA-based calibration frameworks plays an important role in improving forecast reliability and the representation of uncertainty in rainfall prediction. Furthermore, the development of more suitable probability distributions, including Extreme Value Theory (EVT)-based distributions, has strong potential to enhance probabilistic calibration performance for asymmetric rainfall data. This approach is expected to improve the accuracy and reliability of extreme rainfall predictions. The findings of this study provide an important contribution to the development of early warning systems for hydrometeorological disasters and support the achievement of Sustainable Development Goals (SDGs). Full article

To improve the path-following performance of an underactuated autonomous underwater vehicle (AUV) under varying path geometries and desired velocities, this study proposes a direct X-rudder control method based on Task-Informed Inductive-Bias Conservative Soft Actor–Critic (TIB-CSAC). The proposed method directly learns the X-rudder control policy from the path-following information of the current and subsequent path segments in a data-driven way, thereby avoiding the complex design and manual tuning of guidance laws and attitude controllers for rudder command generation. To support such two-segment policy learning, a task-informed inductive-bias encoder is proposed to construct structured and conditioned state representations, thereby improving sample efficiency and overall training quality. In addition, given the long-tail characteristics of task difficulty in agent training, a multi-head conservative value evaluation mechanism is incorporated to mitigate return drawdowns induced by challenging tasks in the tail stage of training and to enhance tail-stage convergence stability. The path-following performance is validated in three representative scenarios with different path pitch, path heading variations, and desired surge velocity conditions. The results show that, compared with the baseline soft actor–critic (SAC) method, TIB-CSAC improves multiple vertical and horizontal error metrics, including maximum absolute error, mean absolute error, tail error, and error threshold exceedance ratio. These results indicate that TIB-CSAC not only improves overall adherence to the reference path, but also more effectively suppresses extreme errors and tail errors, thereby demonstrating stronger path-following robustness and reliability. Full article

Forests of Lualaba Province (DR Congo) form a compositionally complex mosaic of dry dense forest, gallery forest, and Miombo woodland. Yet, categorical land-cover maps impose discrete boundaries on these inherently continuous vegetation gradients, systematically discarding subpixel compositional information critical for forest monitoring and carbon accounting. The magnitude of this information loss at the landscape scale, however, remains largely unquantified. In this study, we train a Multi-Output Neural Network (MONN) using Sentinel-2 spectral and textural predictors (2025) to estimate the proportional cover of three forest types across the province. Model performance is benchmarked against a normalised Random Forest (RF) using spatial block cross-validation. Categorical information loss is quantified pixel-wise using two complementary metrics, dominant class proportion and Shannon compositional entropy, alongside a derived interpretive quantity, categorical information loss. The MONN slightly outperformed RF (R² = 0.648 vs. 0.630; RMSE = 0.224 vs. 0.229), yet the results reveal a fundamentally heterogeneous landscape structure. The mean dominant-class proportion was only 56.2%, indicating that categorical maps discard, on average, 43.8% of compositional information per pixel. Only 7.9% of forested pixels exceeded the 75% dominance threshold, while Shannon entropy reached 74.1% of its theoretical maximum, indicating that forest types coexist in near-equal proportions across most pixels. This renders categorical attribution structurally inadequate for most of the forested landscape. Across 92.1% of forested pixels, no single forest type achieved clear dominance. These results show that compositional mixing is the dominant structural condition of the landscape, and that compositional mapping is essential for representing tropical forest structure in heterogeneous drylands. By formally quantifying categorical information loss at the landscape scale, this study shows that continuous compositional mapping converts this structural ambiguity into a spatially explicit ecological signal, with direct implications for monitoring vegetation dynamics and biodiversity, suggesting a structural source of error in carbon stock estimation in tropical dry forests that warrants empirical validation. Full article

Effectively integrating graph topology and node attributes, while assigning nodes with both semantic similarity and structural closeness to the same community, remains a key challenge in attributed graph community detection. To address this challenge, this study proposes TVHCL-CD, a two-view hierarchical contrastive learning-driven method for community detection. The proposed method constructs an attribute view and a modularity view from the node attribute matrix and the modularity matrix, respectively, to model attribute semantics and high-order community structure priors. Structure-aware two-view representations are then learned in parallel through dual-view graph attention encoders incorporating multi-order neighborhood priors. Furthermore, a structure-enhanced Graph Transformer fusion module is designed to achieve node-level adaptive fusion of the two-view representations by introducing a learnable adjacency bias into global self-attention and a view-aware gating mechanism into the feed-forward network. To align the optimization objective with community semantics, a hierarchical contrastive learning strategy is further developed. Specifically, view-level consistency contrastive learning constructs modularity-guided augmented views to improve representation robustness, while community-level semantic contrastive learning incorporates partial ground-truth labels to enhance intra-community compactness and inter-community separation. Finally, clustering is performed on the fused representations to obtain community partitions. Experimental results on eight real-world attributed graphs and the generated tree-like attributed graph Tree-2500 indicate that TVHCL-CD achieves competitive performance under the semi-supervised transductive setting, while ablation results support the contributions of its main components. Full article

Cross-scene Hyperspectral Image (HSI) classification faces substantial domain shifts caused by sensor heterogeneity, acquisition variation, and scene diversity. While benchmark annotations are assigned to individual center pixels, local patches often contain implicit multi-label semantics due to spectral mixing and spatial overlap. This mismatch distorts prediction structure, exacerbates generalization errors, and limits the effectiveness of standard domain generalization (DG) techniques focused solely on feature or prediction invariance. We propose Class-Aware Semantic Calibration (CASC), a systematic semantic structure calibration framework that addresses three complementary distortions induced by mismatched patch supervision: (i) Balance corrects class frequency bias via reweighted supervision; (ii) Separability enhances boundary decision stability through margin-based logit calibration; and (iii) Independence reduces domain-specific spurious co-occurrence via prediction covariance decorrelation. To preserve calibrated semantics under pseudo-source shift, we further introduce a complementary DualAlign (DA) module, which jointly aligns feature statistics and prediction distributions, enforcing consistency at both representation and semantic levels. Extensive experiments on three cross-scene benchmarks (Houston, Pavia, and WHU-Hi) demonstrate that CASC-DA consistently improves performance over strong baselines, achieving an average gain of 3.0% in overall accuracy and 4.9% in Kappa coefficient compared with the best-performing baseline on each dataset. These results underscore the importance of semantic structure calibration for domain-generalized HSI classification. Full article

Real-time precise point positioning (PPP) is increasingly supported by open satellite-broadcast state-space representation (SSR) services, yet standalone operation with a single service remains vulnerable to limited constellation support, correction outages, latency variations, and service-dependent modeling inconsistencies. In the Asia-Pacific region, MADOCA-PPP and PPP-B2b provide two publicly accessible and complementary SSR sources, but their consistent fusion before user-level PPP estimation remains insufficiently investigated. This paper proposes a correction-domain fusion framework that combines MADOCA-PPP and PPP-B2b orbit and clock corrections before PPP estimation, rather than merging final positioning solutions. Inter-service discrepancies and unknown cross-correlations are handled by a bias-state-aware structured covariance intersection strategy, in which the relative weighting is derived from the respective correction information (inverse variance), preserving statistical consistency and avoiding overconfident fusion. A unified multi-GNSS PPP scheme further supports signal-priority harmonization, broadcast-ephemeris adaptation, correction-age control, and GLONASS inter-frequency and differential code bias handling. Static-station per-epoch (pseudo-kinematic) and offshore kinematic experiments validate the framework. In the static-station test, fusion raised the mean number of valid satellites from 21.98 and 14.98 to 26.56 and improved the horizontal RMS to 0.033 m—better than either standalone service (0.037 m, 0.079 m)—confirming a genuine combination rather than source selection, while the 3D RMS (0.068 m) matched the best standalone service (0.066 m). In the offshore test, fusion achieved the best overall accuracy (0.232 m horizontal, 0.290 m 3D, versus 0.332 m and 0.313 m for the standalone services) and the most satellites (25.4). It also degraded most slowly with increasing elevation cut-off, outperforming both services about threefold at 40°. A normalized-innovation-squared check confirmed the fused covariance is consistent and not overconfident (median ≈ 1.1; within the 99% bound in 100% of epochs). Under single-service outages from 30 s to 600 s, fusion maintained 100.0% availability, confirming its advantage in redundancy, continuity, and resilience. Full article

Direct flood loss estimation for industrial complexes is jointly sensitive to terrain representation, rainfall magnitude, infiltration assumptions, and depth–damage function selection, yet these uncertainties are rarely evaluated together. We quantify their combined effects for the Gumi National Industrial Complex (GNIC), South Korea, using five DEM resolutions (0.5–10 m), six rainfall return periods (10–200 years plus the observed July 2024 event), and three infiltration regimes (5, 10, 20 mm h⁻¹), yielding 90 hydrodynamic realisations from a GPU-accelerated 2D shallow-water model. Each was combined with a harmonised inventory of 16,463 buildings (replacement value 43.07 trillion KRW) and three vulnerability-function families (HAZUS-MH, JRC Huizinga, Korean MD-FDA), producing 270 loss estimates under a common dimensionless transformation. A three-way ANOVA on log-transformed damage confirmed highly significant main effects of resolution, rainfall, and infiltration across all functions, more than an order of magnitude larger than interactions, and robust to heteroscedasticity-consistent and permutation tests. Coarsening the DEM from 0.5 to 10 m reduced expected annual loss (EAL) by 55–57%, while inter-function depth–damage divergence exceeded four-fold at shallow inundation. Validation against the July 2024 event gave the best skill at 2 m resolution (critical success index 0.80, accuracy 0.86). Multi-family residential and heavy industry accounted for 83–89% of total EAL. These results show that terrain resolution and damage-function selection are first-order, statistically independent controls on industrial flood loss, and that omitting any sensitivity axis can bias EAL by more than two-fold. Full article

Spaceborne precipitation products such as NASA IMERG V07 provide sub-hourly data required for hydrological modelling, but systematic biases in wet-event frequency and extreme-intensity representation limit their reliability for applications sensitive to precipitation extremes. This study develops a three-stage calibration pipeline combining probability-of-precipitation frequency adaptation, empirical quantile mapping of the distribution body, and Generalised Pareto Distribution tail modelling with constrained blending. The approach is calibrated against 202 Roshydromet stations using 3-hourly observations and evaluated on 15 spatially independent stations over a 9-year validation period. At the station-optimal blending weight, the proposed pipeline reduces median absolute percentage bias at the P99 quantile from 43.9% to 10.2%, while maintaining comparable volume balance (|PBIAS| 6.5%). To suppress a disaggregation artefact arising from amplification of multi-hour accumulations, the operational gridded R-factor product instead adopts a more conservative blend (|PBIAS@P99| = 24.9%) together with an empirically constrained accumulation cap, although the absence of sub-hourly calibration data remains the principal limitation. The calibrated dataset is applied to derive a 24-year (2001–2024) rainfall erosivity climatology for the Volga region, yielding a domain-mean R-factor of 254 ± 55 MJ mm ha⁻¹ h⁻¹ yr⁻¹ with no detectable monotonic trend. The proposed framework improves the representation of precipitation extremes and provides a transferable preprocessing approach for hydrological modelling applications. Full article

As AI permeates daily life, algorithmic platforms increasingly function as complex sociotechnical systems that shape public perception and societal attitudes. Addressing concerns that AI text-to-image models and search engines reinforce stereotypes, this study focuses on China, a context marked by traditional gender norms and a vast technological ecosystem, examining how algorithmic systems perpetuate gender power structures through occupational representations. Using algorithmic audits of 60 occupations, Z-tests, and QAP network analysis, this study compares platform gender representations with national census data, systematically distinguishing “generative bias” in AI platforms (Doubao Seedream 3.0, Jimeng Image 3.0) from “retrieval bias” in search engines (Baidu, Sogou). Findings reveal that search engines reinforce stereotypes by over-representing dominant genders and obscuring non-mainstream ones. Generative AI exhibits more radical distortions. The specialized AI Jimeng shows a strong gender polarization feature, while the general AI Doubao shows an ideal balanced gender presentation tendency, balancing representation yet creating an equally false reality. Compared to search engines, AI platforms have greater creativity in representing occupational gender. This study reveals a mutually reinforcing bias cycle among audiences, media, and algorithms, offering a crucial non-Western perspective for feminist technology studies and significant implications for equitable AI governance. Full article

Imaging-based spatial transcriptomics has advanced from low-plex single-molecule fluorescence in situ hybridization to a diverse set of highly multiplexed platforms, with recent multimodal and pathology-compatible capabilities. Despite major differences in chemistry, coding, and imaging strategies across different platforms, their biological interpretation often converges on a few notable computational biology problems. This review examines imaging-based spatial transcriptomics through the lens of data interpretation and applications, focusing on the analytical framework that converts raw fluorescence signals or accompanying in situ sequencing data into molecule-, cell-, and tissue-level representations. We discuss the key challenges in preprocessing, registration, restoration, feature detection, barcode decoding, molecule calling, cell segmentation, transcript assignment, probabilistic cell typing, spatial-domain inference, and atlas integration. We also highlight how optical crowding, tissue thickness, panel bias, and multimodal complexity increase computational difficulty. Finally, we summarize applications of imaging-based spatial transcriptomics techniques, ranging from subcellular RNA localization to atlas-scale and pathology-aware spatial analysis. Full article