Search Results (174)

Urban safety is a critical concern for sustainable city development, with crime patterns often linked to localized environmental factors. Understanding the spatial dynamics of safety is critical for informed design and planning of urban environments. This study employs a Geographically Weighted Regression (GWR) approach to investigate how crime in Austin, Texas, correlates with Points of Interest (POIs) such as bars, transit stations, financial businesses, and public spaces, while accounting for localized socio-economic factors. Building on theoretical frameworks like Routine Activity Theory and Crime Pattern Theory, the analysis integrates crime data from the Austin Police Department (APD), POI datasets, and census variables to explore spatially varying relationships often overlooked by traditional global models (e.g., OLS). A novel adaptive geo-grid method refines spatial units by clustering high-density downtown areas into smaller zones and retaining larger grids in suburban regions, ensuring precision without over-fragmentation. Analysis of crime incidents and POI data reveals significant spatial non-stationarity in crime–environment associations. Transportation-related facilities demonstrate strong spatial correlation with crime citywide, particularly forming persistent crime hotspots around transit hubs in areas like Rundberg Lane, South Congress, and East Riverside. Alcohol-related establishments show a strong positive correlation with crime in entertainment districts (coefficient up to 13.5, p < 0.001) but a negligible association in suburban residential areas (coefficient close to 0, p > 0.05). The GWR model significantly outperforms traditional OLS regression, capturing critical local variations obscured by global models. Downtown Austin emerges as a complex hotspot for urban safety where multiple high-risk POI types overlap. This research advances urban design and planning knowledge by providing empirical evidence that environmental factors’ influence on safety is spatially conditional rather than universally consistent, aligning with Crime Pattern Theory and Routine Activity Theory. The findings support place-specific crime prevention strategies, offering policymakers data-driven insights for developing targeted design strategies for urban zones. Full article

Against the backdrop of urban–rural integrated development, special ecological function zones, as spatial carriers with significant regional ecological value and rural development functions, are confronted with a striking conflict between ecological conservation and regional advancement. This contradiction is comprehensively reflected in the interactions among land use functions (LUFs) that differ in nature and intensity. Therefore, exploring the trade-off and synergy (TOS) among regional LUFs is not only of great significance for optimizing territorial spatial patterns and advancing rural revitalization but also provides scientific evidence for the differentiated administration of regional land use. Taking 185 townships in the Funiu Mountain area of China as research units, this study constructs a land use assessment system based on the ‘Production–Living–Ecological’ (PLE) framework, utilizing multi-source datasets from 2000 to 2020. Spearman correlation analysis, geographically weighted regression (GWR), and bivariate local spatial autocorrelation methods are employed to examine the spatio-temporal dynamics of LUFs and the spatial non-stationarity of their TOSs. The findings indicate that, throughout the research period, the production function (PF) displayed a fluctuating declining trend, whereas the living function (LF) and ecological function (EF) demonstrated a fluctuating increasing trend. Notably, EF held an absolute dominant position in the overall structure of LUFs. This is highly consistent with the region’s positioning as a special ecological function zone and also a direct reflection of the effectiveness of continuous ecological construction over the past two decades. Spatially, PF is stronger in southern, eastern, and northern low-altitude townships, correlating with higher levels of economic development; LF is concentrated around townships near county centers; and high EF values are clustered in the central and western areas, showing an opposite spatial pattern to PF and LF. A synergistic relationship is observed between PF and LF, while both PF and LF exhibit trade-offs with EF. The TOSs between different function changes demonstrate significant spatial non-stationarity: linear synergy was the primary type for PF-LF, PF-EF, and LF-EF combinations, but each combination exhibited unique spatial characteristics in terms of non-stationarity. Notably, towns identified as having different types of trade-off relationships in the study of spatial non-stationarity are key areas for township spatial governance and optimization. Through the allocation of regional resources and targeted policy tools, the functional relationships can be adjusted and optimized to attain sustainable land use. Full article

Quantitative analysis of ecosystem services (ESs) supply–demand dynamics, and identifying its dominant drivers and the spatial non-stationarity of driving mechanisms, is a crucial prerequisite for effective regional ESs management and the formulation of scientific ecological conservation plans. Previous related studies have primarily focused on the supply–demand balance of specific ESs and the driving analysis of ESs supply. Comprehensive analysis of ESs supply–demand dynamics and research on their spatially heterogeneous response mechanisms remain relatively scarce. In this study, we assessed the supply, demand, and supply–demand matching relationships of ESs on the Loess Plateau (LP) from 1990 to 2023 using a land-cover-based ESs supply–demand quantitative matrix. We then employed Geodetector and Geographically weighted regression model to explore the dominant driving factors and their spatially varying effects on ESs supply–demand relationships. The results revealed that over the past three decades, the continuous decline in ESs supply coupled with the annual increase in ESs demand has led to a worsening trend in ESs supply–demand relationships towards deficit. Fortunately, the LP still maintained a supply-surplus state at present. The proportion of construction land, population density, GDP density, and the proportion of forestland and grassland were identified as key drivers of changes in ESs supply–demand relationships. The expansion of construction land was the most crucial driver of the deterioration in ESs supply–demand relationships on the LP, exhibiting a universally negative inhibitory effect. The proportion of forestland and grassland exerted a regionally wide positive spatial effect, highlighting the critical role of vegetation restoration in improving ESs relationships. The influences of population density and GDP density exhibited a coexistence of positive promoting and negative inhibitory effects across space. Our results emphasize that ESs management policies on the LP must account for the spatial heterogeneity of driving mechanisms, requiring more localized and targeted land use strategies and management policies to enhance ESs sustainability. Full article

Background: Brain–computer interfaces (BCIs) based on motor imagery (MI) offer promising solutions for motor rehabilitation and communication. However, electroencephalography (EEG) signals are often characterized by low signal-to-noise ratios, strong non-stationarity, and significant inter-subject variability, which pose significant challenges for accurate decoding. Existing methods often struggle to simultaneously model the spatial interactions between EEG channels, the local fine-grained features within signals, and global semantic patterns. Methods: To address this, we propose the graph attention-based hierarchical temporal network (GAH-TNet), which integrates spatial graph attention modeling with hierarchical temporal feature encoding. Specifically, we design the graph attention temporal encoding block (GATE). The graph attention mechanism is used to model spatial dependencies between EEG channels and encode short-term temporal dynamic features. Subsequently, a hierarchical attention-guided deep temporal feature encoding block (HADTE) is introduced, which extracts local fine-grained and global long-term dependency features through two-stage attention and temporal convolution. Finally, a fully connected classifier is used to obtain the classification results. The proposed model is evaluated on two publicly available MI-EEG datasets. Results: Our method outperforms multiple existing state-of-the-art methods in classification accuracy. On the BCI IV 2a dataset, the average classification accuracy reaches 86.84%, and on BCI IV 2b, it reaches 89.15%. Ablation experiments validate the complementary roles of GATE and HADTE in modeling. Additionally, the model exhibits good generalization ability across subjects. Conclusions: This framework effectively captures the spatio-temporal dynamic characteristics and topological structure of MI-EEG signals. This hierarchical and interpretable framework provides a new approach for improving decoding performance in EEG motor imagery tasks. Full article

Accurately predicting stock prices is crucial for investment and risk management, but the non-stationarity of the financial market and the complex correlations among stocks pose challenges to traditional models (ARIMA, LSTM, XGBoost), resulting in difficulties in effectively capturing dynamic patterns and limited prediction accuracy. To this end, this paper proposes the Financial Spatio-Temporal Graph Attention Network (FSTGAT), with the following core innovations: temporal modelling through gated causal convolution to avoid future information leakage and capture long- and short-term fluctuations; enhanced spatial correlation learning by adopting the Dynamic Graph Attention Mechanism (GATv2) that incorporates industry information; designing the Multiple-Input-Multiple-Output (MIMO) architecture of industry grouping for the simultaneous learning of intra-group synergistic and inter-group influence; symmetrically fusing spatio-temporal modules to construct a hierarchical feature extraction framework. Experiments in the commercial banking and metals sectors of the New York Stock Exchange (NYSE) show that FSTGAT significantly outperforms the benchmark model, especially in high-volatility scenarios, where the prediction error is reduced by 45–69%, and can accurately capture price turning points. This study confirms the potential of graph neural networks to model the structure of financial interconnections, providing an effective tool for stock forecasting in non-stationary markets, and its forecasting accuracy and industry correlation capturing ability can support portfolio optimization, risk management improvement and supply chain decision guidance. Full article

The spatiotemporal heterogeneity of hydrothermal conditions during the spring sowing period profoundly shapes cropping layouts and sowing strategies. Using NASA’s GLDAS remote sensing reanalysis, we developed a continuous agricultural climate risk index that integrates three remotely driven indicators—spring sowing window days (SWDs) derived from a “continuous suitable-day” logic, the hydrothermal coordination degree (D value), and a comprehensive suitability index (SSH_SI)—thus advancing risk assessment from single metrics to a multidimensional framework. Methodologically, dominant periodic structures of spring sowing hydrothermal risk were extracted via a combination of wavelet power spectra and the global wavelet spectrum (GWS), while spatial trend-surface fitting and three-dimensional directional analysis captured spatial non-stationarity. The index’s spatial migration trajectories and centroid-evolution paths were then quantified. Results reveal pronounced gradients along the Great Wall Belt: SWD displays a “central-high, terminal-low” pattern, with sowing windows restricted to only 3–6 days in northeastern Inner Mongolia and western Liaoning but extending to 11–13 days in the central plains of Inner Mongolia and Shanxi; SSH_SI and D values form an overall “south-west high, north-east low” pattern, indicating more favorable hydrothermal coordination in southwestern areas. Temporally, although SWD and SSH_SI show no significant downward trend, their interannual variability has increased, signaling rising instability, whereas the D value declines markedly in most regions, reflecting intensified hydrothermal imbalance. The integrated risk index identifies high-risk hotspots in eastern Inner Mongolia and northern North China, and low-risk zones in western provinces such as Gansu and Ningxia. Centroid-shift analysis further uncovers a dynamic regional adjustment in optimal sowing patterns, offering scientific evidence for addressing spring sowing climate risks. These findings provide a theoretical foundation and decision support for optimizing regional cropping structures, issuing climate risk warnings, and precisely regulating spring sowing schedules. Full article

Accurate short-term weather forecasting plays a vital role in disaster response, agriculture, and energy management, where timely and reliable predictions are essential for decision-making. Graph neural networks (GNNs), known for their ability to model complex spatial structures and relational data, have achieved remarkable success in meteorological forecasting by effectively capturing spatial dependencies among distributed weather stations. However, most existing GNN-based approaches rely on pairwise station connections, limiting their capacity to represent higher-order spatial interactions. Moreover, their dependence on supervised learning makes them vulnerable to spatial heterogeneity and temporal non-stationarity. This paper introduces a novel spatial–temporal pretraining framework, Hypergraph-enhanced Meteorological Pretraining (HyMePre), which combines hypergraph neural networks with self-supervised learning to model high-order spatial dependencies and improve generalization across diverse climate regimes. HyMePre employs a two-stage masking strategy, applying spatial and temporal masking separately, to learn disentangled representations from unlabeled meteorological time series. During forecasting, dynamic hypergraphs group stations based on meteorological similarity, explicitly capturing high-order dependencies. Extensive experiments on large-scale reanalysis datasets show that HyMePre outperforms conventional GNN models in predicting temperature, humidity, and wind speed. The integration of pretraining and hypergraph modeling enhances robustness to noisy data and improves generalization to unseen climate patterns, offering a scalable and effective solution for operational weather forecasting. Full article

This study analyzes air pollution time-series big data to assess stationarity, seasonal patterns, and the performance of machine learning models in forecasting PM2.5 concentrations. Fifty-two low-cost sensors (LCS) were deployed across Krakow city and its surroundings (Poland), collecting hourly air quality data and generating nearly 20,000 observations per month. The network captured both spatial and temporal variability. The Kwiatkowski–Phillips–Schmidt–Shin (KPSS) test confirmed trend-based non-stationarity, which was addressed through differencing, revealing distinct daily and 12 h cycles linked to traffic and temperature variations. Additive seasonal decomposition exhibited time-inconsistent residuals, leading to the adoption of multiplicative decomposition, which better captured pollution outliers associated with agricultural burning. Machine learning models—Ridge Regression, XGBoost, and LSTM (Long Short-Term Memory) neural networks—were evaluated under high spatial and temporal variability (winter) and low variability (summer) conditions. Ridge Regression showed the best performance, achieving the highest $R^2$ (0.97 in winter, 0.93 in summer) and the lowest mean squared errors. XGBoost showed strong predictive capabilities but tended to overestimate moderate pollution events, while LSTM systematically underestimated PM2.5 levels in December. The residual analysis confirmed that Ridge Regression provided the most stable predictions, capturing extreme pollution episodes effectively, whereas XGBoost exhibited larger outliers. The study proved the potential of low-cost sensor networks and machine learning in urban air quality forecasting focused on rare smog episodes (RSEs). Full article

Understanding the impacts of the spatial non-stationarity of environmental factors on surface soil moisture (SSM) in different seasons is crucial for effective environmental management. Yet, our knowledge of this phenomenon remains limited. This study introduces an interpretable machine learning framework that combines the SHapley Additive exPlanations (SHAP) method with two-step clustering to unravel the spatial drivers of SSM across Iran. Due to the limited availability of in situ SSM data, the performance of three global SSM datasets—SMAP, MERRA-2, and CFSv2—from 2015 to 2023 was evaluated using agrometeorological stations. SMAP outperformed the others, showing the highest median correlation and the lowest Root Mean Square Error (RMSE). Using SMAP, we estimated SSM across 609 catchments employing the Random Forest (RF) algorithm. The RF model yielded R² values of 0.89, 0.83, 0.70, and 0.75 for winter, spring, summer, and autumn, respectively, with corresponding RMSE values of 0.076, 0.081, 0.098, and 0.061 m³/m³. SHAP analysis revealed that climatic factors primarily drive SSM in winter and autumn, while vegetation and soil characteristics are more influential in spring and summer. The clustering results showed that Iran’s catchments can be grouped into five categories based on the SHAP method coefficients, highlighting regional differences in SSM controls. Full article

A geographically and temporally weighted regression (GTWR) model is an effective tool for dealing with spatial heterogeneity and temporal non-stationarity simultaneously. As an important characteristic of spatiotemporal data, spatiotemporal autocorrelation should be considered when constructing spatiotemporally varying coefficient models. The proposed local autoregressive geographically and temporally weighted regression (GTWRLAR) model can simultaneously handle spatiotemporal autocorrelations among response variables and the spatiotemporal heterogeneity of regression relationships. The two-stage weighted least squares (2SLS) estimation can effectively reduce computational complexity. However, the weighted least squares estimation is essentially a Nadaraya–Watson kernel-smoothing approach for nonparametric regression models, and it suffers from a boundary effect. For spatiotemporally varying coefficient models, the three-dimensional spatiotemporal coefficients (longitude, latitude, and time) inherently exhibit larger boundaries than one-dimensional intervals. Therefore, the boundary effect of the 2SLS estimation of GTWRLAR will be more serious. A local–linear geographically and temporally weighted 2SLS (GTWRLAR-L) estimation is proposed to correct the boundary effect in both the spatial and temporal dimensions of GTWRLAR and simultaneously improve parameter estimation accuracy. The simulation experiment shows that the GTWRLAR-L method reduces the root mean square error (RMSE) of parameter estimates compared to the standard GTWRLAR approach. Empirical analyses of carbon emissions in China’s Yellow River Basin (2017–2021) show that GTWRLAR-L enhances the adjusted $R^2$ from 0.888 to 0.893. Full article

Forecasting multivariate time series is a pivotal task in controlling multi-sensor systems. The joint forecasting of all channels may be too complex, whereas forecasting the channels independently may cause important spatial inter-dependencies to be overlooked. In this paper, we improve the performance of single-channel forecasting algorithms by designing an interpretable front-end that extracts the spatial–temporal components from the input multivariate time series. Specifically, the multivariate samples are first segmented into equal-sized matrix symbols. The symbols are decomposed into the frequency-separated Intrinsic Mode Functions (IMFs) using a 2D Empirical-Mode Decomposition (EMD). The IMF components in each channel are then forecasted independently using relatively simple univariate predictors (UPs) such as DLinear, FITS, and TCN. The symbol size is determined to maximize the temporal stationarity of the EMD residual trend using Bayesian optimization. In addition, since the overall performance is usually dominated by a few of the weakest predictors, it is shown that the forecasting accuracy can be further improved by reordering the corresponding channels to make more correlated channels more adjacent. However, channel reordering requires retraining the affected predictors. The main advantage of the proposed forecasting framework for multivariate time series is that it retains the interpretability and simplicity of single-channel forecasting methods while improving their accuracy by capturing information about the spatial-channel dependencies. This has been demonstrated numerically assuming a 64-channel EEG dataset. Full article

Methane (CH₄) emissions from dairy farming represent a substantial yet under-quantified share of agricultural greenhouse gas emissions. This study provides an in-depth, satellite-based fingerprinting analysis of methane emissions from Canada’s dairy sector, using Sentinel-5P/TROPOMI data. We utilized a robust quasi-experimental design, pairing 14 dairy-intensive zones with eight non-dairy reference regions, to analyze methane emissions from 2019 to 2024. A dynamic, region-specific baseline approach was implemented to remove temporal non-stationarity and isolate dairy-specific methane signals. Dairy regions exhibited consistently higher methane concentrations than reference areas, with an average methane anomaly of 17.4 ppb. However, this concentration gap between dairy and non-dairy regions notably narrowed by 57.23% (from 24.42 ppb in 2019 to 10.44 ppb in 2024), driven primarily by accelerated methane increases in non-dairy landscapes and a pronounced one-year contraction during 2022–2023 (−39.29%). Nationally, atmospheric methane levels rose by 3.83%, revealing significant spatial heterogeneity across provinces. Notably, an inverse relationship between the initial methane concentrations in 2019 and subsequent growth rates emerged, indicating spatial convergence. The seasonal analysis uncovered consistent spring minima and fall–winter maxima across regions, reflecting the combined effects of seasonal livestock management practices, atmospheric transport dynamics, and biogeochemical processes. The diminishing dairy methane anomaly suggests complex interplay of intensifying background methane emissions from climate-driven wetland fluxes, increasing fossil fuel extraction activities, and diffuse agricultural emissions. These findings underscore the emerging challenges in attributing sector-specific methane emissions accurately from satellite observations, highlighting both the capabilities and limitations of current satellite monitoring approaches. Full article

Climate change causes extreme weather events to occur frequently, such as drought, which may exacerbate forest fire regimes; as such, forest fire regimes may be closely related to drought characteristics. The spatial non-stationarity of factors affecting forest fires has not been fully clarified and needs further exploration. This study intends to address how drought characteristics affect forest fire regimes in China and whether spatial non-stationarity can improve the model prediction based on methods such as the run theory and GWR. Our results show that geographically weighted regression models perform better (AICc, AUC, R², RMSE, and MAE) than global regression models in modeling forest fire regimes. Although GWR improves accuracy, small sample sizes (vegetation zones, climatic zones) may affect its accuracy. Drought characteristics significantly affect (p < 0.05) the forest fire regimes, and the correlation is spatially non-static. At the grid scale, a positive correlation between the forest fire occurrence probability and drought characteristics is mostly distributed in the southwest and northwest regions. Our study is conducive to an in-depth understanding of the relationship between forest fire regimes and drought, aiming to provide a scientific basis for the development of forest fire management measures to mitigate drought stress according to local conditions. Full article

The Sylhet Division of Bangladesh, highly susceptible to monsoon flooding, requires effective flood risk management to reduce socio-economic losses. Flood frequency analysis is an essential aspect of flood risk management and plays a crucial role in designing hydraulic structures. This study applies regional flood frequency analysis (RFFA) using L-moments to identify homogeneous hydrological regions and estimate extreme flood quantiles. Records from 26 streamflow gauging stations were used, including streamflow data along with corresponding physiographic and climatic characteristic data, obtained from GIS analysis and ERA5 respectively. Most stations showed no significant monotonic trends, temporal correlations, or spatial dependence, supporting the assumptions of stationarity and independence necessary for reliable frequency analysis, which allowed the use of cluster analysis, discordancy measures, heterogeneity tests for regionalization, and goodness-of-fit tests to evaluate candidate distributions. The Generalized Logistic (GLO) distribution performed best, offering robust quantile estimates with narrow confidence intervals. Multiple Non-Linear Regression models, based on catchment area, elevation, and other parameters, reasonably predicted ungauged basin peak discharges (R² = 0.61–0.87; RMSE = 438–2726 m³/s; MAPE = 41–74%) at different return periods, although uncertainty was higher for extreme events. Four homogeneous regions were identified, showing significant differences in hydrological behavior, with two regions yielding stable estimates and two exhibiting greater extreme variability. Full article

Rural settlements in agro-pastoral ecotones reflect the complex interplay between natural constraints and human land use, particularly in ecologically sensitive and climatically transitional regions. This study investigated the agro-pastoral ecotone of eastern Inner Mongolia, a representative region characterized by environmental heterogeneity and competing land use functions. Landscape pattern indices, ordinary least squares (OLS) regression, and geographically weighted regression (GWR) were employed to analyze settlement morphology and its environmental determinants. The results reveal a distinct east–west spatial gradient: settlements are larger and more concentrated in low-elevation plains with favorable hydrothermal conditions, whereas those in mountainous and pastoral areas are smaller, sparser, and more fragmented. OLS regression revealed a strong positive correlation between arable land and settlement density (r > 0.8), whereas elevation and slope were significantly negatively correlated. GWR results further highlight spatial non-stationarity in the influence of key environmental factors. Average annual temperature generally shows a positive influence on settlement density, particularly in the central and eastern agricultural areas. In contrast, forest cover is predominantly negative, especially in the Greater Khingan Mountains. Proximity to water resources consistently enhances settlement density, although the magnitude of this effect varies across regions. Based on spatial characteristics and land use structure, rural settlements were categorized into four types: alpine pastoral, agro-pastoral transitional, river valley agricultural, and forest ecological. This study provides empirical evidence that natural factors (topography, climate, and hydrology) and land use variables (farmland, pasture, and woodland) collectively shape rural settlement patterns in transitional landscapes. The findings offer methodological and practical insights for targeted land management and sustainable rural development in agro-pastoral regions under ecological and socioeconomic pressures. Full article