Search Results (106)

Spatial data has distinctive properties that differentiate it from non-spatial data. One prominent characteristic is spatial autocorrelation (SA). When machine learning techniques are applied for spatial data modeling, they require spatially explicit consideration. If these inherent spatial structures are ignored, models may produce biased predictions. However, integrating this property into the model yields additional spatial insight, thereby enhancing learning and improving predictive accuracy. This study examines spatially explicit $K$-nearest neighbors (SE-KNN) by integrating SA as a spatially explicit property, $\lambda$, into the learning algorithm. The innovation of SE-KNN lies in its alignment with the principle of spatial autocorrelation, as KNN’s core learning assumption—that similar observations tend to have similar outcomes—naturally parallels spatial dependence. The proposed SE-KNN is applied to a house price prediction model using house sales data from Franklin County, Ohio to demonstrate a spatially dependent, data-rich, and real-world problem. The results show that SE-KNN achieved the best prediction accuracy compared to mean of absolute error (MAE) of three other machine learning approaches (i.e., standard KNN, linear regression, and artificial neural networks). The proposed method effectively captures the spatial structures in the housing market and leaves only a trace amount of SA in the residuals. Full article

The Jiangsu Yangtze River city cluster is a key growth pole of the Yangtze River Economic Belt, yet substantial disparities in development levels persist across cities, and the role of rail transit investment in fostering regional economic coordination remains insufficiently understood. This study aims to reveal the dynamic mechanisms through which railway transportation investment influences regional economic growth via population migration and service industry agglomeration, and to quantify the economic multiplier effects under different investment scenarios. Considering the close economic linkages among cities, spatial autocorrelation analysis is applied to assess intercity economic dependence, which provides the basis for developing a system dynamics model that links the rail transit system with the regional economy. Using data from eight core cities over the period 2014–2023, the model is employed to simulate long-term economic responses under different investment scenarios. The results indicate that increasing the rail transit investment ratio from 0.0077 to 0.02 is associated with an estimated 13.2% increase in regional GDP by 2030, with a corresponding economic multiplier of approximately 1.8, while simulation errors remain within 4.1–16.2% compared with historical data. The findings suggest that rail transit investment promotes regional growth through improved accessibility, factor agglomeration, and industrial upgrading, and that coordinated planning at the urban agglomeration scale is more effective than isolated city-level strategies. By integrating spatial dependence analysis with system dynamics modeling, this study offers a dynamic perspective on the regional economic impacts of rail transit investment. Full article

Understanding urban deprivation and its impact on health is crucial for addressing inequalities in cities. Using Quito as a case study, we developed a census-based socioeconomic urban deprivation index and analyzed three health outcomes: fatal injuries, COVID-19 deaths, and maternal mortality. Spatial patterns were examined using Local Moran’s I, and regression analyses included OLS, spatial lag, spatial error, and GWR models, applied at two spatial scales, census sectors and census zones, with deprivation as the independent variable. Most of the regression models indicated that deprivation does not explain health outcomes, with the exception of fatal injuries and COVID-19 deaths at the census zone scale when spatial error models are applied. Our results also revealed MAUP effects, as spatial patterns and associations between the studied variables vary depending on spatial scale. Spatial models improved explanatory power compared to OLS, uncovering spatial dependence and heterogeneity in the relationships between deprivation and health outcomes. Our findings underscore the importance of multiscale and spatially explicit approaches in urban health research and provide actionable evidence for targeted interventions and urban planning that account for both local and structural patterns of deprivation. Full article

This study proposes a multi-station high-frequency microtremor surface-wave exploration method for high-resolution characterization of shallow subsurface structures in coastal engineering environments. Three representative layered geological models were established, and Rayleigh-wave theoretical dispersion curves were calculated using a fast vector transfer algorithm to analyze dispersion characteristics associated with different stratigraphic conditions. Five array geometries were then employed to acquire high-frequency ambient-noise data, and dispersion curves were extracted using the Extended Spatial Autocorrelation (ESPAC) method. Comparative analysis revealed that the rectangular, triangular, and circular arrays provided the most stable and accurate dispersion imaging, with mismatch errors below 0.5%, and their inverted S-wave velocity structures closely matched theoretical models. Field application on the Dongzhou Peninsula in Fujian Province further demonstrated the effectiveness of the proposed method. The inverted shear-wave (S-wave) velocity profiles from three survey lines successfully delineated the original and reclaimed coastlines, showing strong agreement with known geological boundaries. These results demonstrate that the proposed approach provides a non-invasive, cost-effective, and high-resolution tool for evaluating geological conditions in coastal engineering settings. It shows substantial potential for broader application in coastal site characterization and marine engineering development. Full article

This study analyzed the relationships between land-cover types and the distribution of alien and invasive plant species using data from the 5th National Ecosystem Survey of Korea (2019–2023). A total of 711,557 plant occurrence records were collected across 780 map sheets, resulting in the identification of 3842 vascular plant species, including both alien and invasive taxa. To evaluate spatial patterns and environmental drivers, multiple linear regression and spatial regression models—specifically the Spatial Lag Model (SLM) and Spatial Error Model (SEM)—were applied. The results revealed that alien and invasive species exhibited non-random, spatially clustered distributions influenced by habitat type and disturbance intensity. Alien species were more abundant in agricultural areas and wetlands, whereas forests and grasslands acted as resistant ecosystems. In contrast, invasive species were concentrated in bare lands and urbanized drylands, highlighting the importance of habitat openness and human disturbance in facilitating invasion. Spatial autocorrelation analyses (Moran’s I = 0.0777 for alien species; 0.1933 for invasive species) and the strong spatial dependence in the Spatial Error Model (λ = 0.7405 and 0.6428) confirmed that invasion patterns are shaped by spatial connectivity and environmental continuity. These findings demonstrate that invasion processes in Korea are driven by both anthropogenic disturbance and spatial dependency. Effective management therefore requires habitat-specific, spatially coordinated strategies, emphasizing early detection and rapid control in high-risk areas while reinforcing the ecological buffering capacity of forests to maintain biodiversity and ecosystem stability. Full article

Urban river restoration provides significant ecological and social benefits, yet its market valuation remains underexamined in rapidly urbanizing inland cities. This study estimates the economic value of integrated blue–green spaces generated by the Bai River Ecological Restoration Project in Nanyang, China, using a spatially explicit hedonic pricing framework that links geocoded resale transactions with NDVI-based vegetation measures. Properties located within blue–green zones—areas jointly characterized by restored waterways and enhanced riparian greening—command an average price premium of 17.9% (CNY 1509/m²). Visual accessibility further increases housing values, although interaction effects indicate diminishing marginal premiums where multiple amenities co-occur. Quantile regressions show stronger capitalization effects in lower- and middle-priced segments, suggesting that ecological improvements may yield broad-based rather than elite-focused benefits. Spatial dependence diagnostics confirm significant autocorrelation, and Spatial Error Model estimates remain consistent with the baseline results. Overall, the findings provide robust evidence of supra-additive blue–green synergies and demonstrate the utility of combining NDVI with spatial econometric hedonic modeling. The study offers a transferable framework for supporting nature-based urban planning and informing cost–benefit evaluations of integrated ecological restoration initiatives. Full article

The Ordovician No. 6 fault zone reservoir in the Shunbei Oilfield exhibits ultra-deep-burial, high-pressure, and high-temperature conditions. Its pronounced tectonic control and significant heterogeneity render traditional in situ stress prediction methods—based on linear elasticity and anisotropy assumptions—inadequate for accurately characterizing the evolution and uncertainty of carbonate reservoir stiffness. Therefore, quantitatively predicting the development patterns and distribution characteristics of the Shunbei No. 6 structural fault zone is crucial for the exploration and development of Ordovician carbonate reservoirs in the Shunbei region. This study integrates wave impedance inversion with high-confining-pressure PFC particle flow biaxial test results to establish a constitutive calibration system consistent with seismic and experimental data. It introduces a nonlinear weakening function incorporating higher-order derivative constraints to fuse structural fracture and effective stress weakening effects, enabling dynamic correction of elastic parameters. This approach establishes a novel in situ stress prediction model. Simulation results indicate a predicted range for maximum horizontal principal stress between 201 and 261 MPa, with minimum horizontal principal stress ranging from 124 to 173 MPa. Predicted stress values for three key wells exhibit measurement errors within 6.92% compared to actual logging data, displaying a zoned spatial distribution consistent with regional tectonic stress evolution patterns. Simultaneously, sensitivity analysis reveals that the Young’s modulus fitting accuracy improved from 0.89 to 0.95, with a 43% reduction in mean square error, with the proportion of outliers reduced to below 1%. This significantly enhances response continuity and numerical stability in high-gradient disturbance zones and stiffness drop regions. The new model explicitly incorporates the nonlinear coupling between fracture geometry and pore pressure disturbance into the parameter field, eliminating systematic bias along fracture zones. Higher-order derivative constraints suppress numerical oscillations in high-gradient areas, stabilizing variance and preventing anomaly propagation. Residual distributions exhibit enhanced symmetry and reduced spatial autocorrelation, effectively suppressing numerical oscillations and divergence in complex fracture zones while significantly improving stress prediction accuracy for the study area. Overall, this research provides novel methodologies for predicting in situ stresses in ultra-deep carbonate reservoirs, offering engineering guidance and parameterization references for scheme deployment in complex fractured karst systems. Full article

This paper investigates within-scene path loss prediction at 1.8 GHz in a smart-campus micro-urban environment using multivariate machine-learning (ML) models. We leverage an open measurement campaign from Covenant University (Nigeria) comprising three routes with per-sample geospatial predictors—longitude, latitude, altitude, elevation, Tx–Rx distance, and clutter height—and train Random Forests (RF), Gradient Boosting (GB), Support Vector Regression (SVR), Gaussian Processes (GP), and a shallow neural network (NN). A unified pipeline with 5-fold cross-validation (CV), seeded reproducibility, and Optuna-driven hyperparameter search is adopted; performance is reported as RMSE/MAE/R² (mean ± sd). To contextualize feature reliability, we include Pearson correlation heatmaps and Variance Inflation Factors (VIFs), a systematic ablation of predictors, and TreeSHAP beeswarm analyses on held-out splits. We also evaluate spatially aware validation (blocked CV within route and leave-one-route-out checks) to mitigate optimism due to spatial autocorrelation. Results show that multivariate ML consistently outperforms classical empirical formulas (COST-231, ECC-33) in this campus setting, with RF achieving the lowest errors across routes (RMSE ≈ 2.14/2.16/2.95 dB for X/Y/Z, respectively), while GB ranks second and kernel methods (SVR/GP) and the NN trail closely behind. Ablation confirms that distance plus coordinates drive the largest gains, with terrain/clutter providing route-dependent refinements. SHAP analyses align with these findings, highlighting stable, interpretable contributions of geospatial covariates. Spatial CV increases absolute errors moderately but preserves model ranking, supporting the robustness of conclusions. Overall, scenario-aware, multivariate ML yields material accuracy gains for smart-campus planning at 1.8 GHz. Full article

This study evaluates the influence of land use and water stress on ecosystem resilience, using panel data for thirty-three European countries from 2007 to 2024, following the identification of a research gap in the literature on this topic. The dependent variable is the bioclimatic ecosystem resilience index (BER), and the explanatory variables are Agricultural Land Share (ALS), Forest Land Share (FLS), and the Level of Water Stress (WS). The estimated models are a fixed-effects panel regression with Driscoll-Kraay standard errors, robust to autocorrelation, heteroscedasticity, and spatial dependence, and a kernel-based regularized least squares model, which offers a new, nonlinear, heterogeneous, and sensitive to local contexts perspective on ecosystem resilience. The results indicate a significant positive effect of FLS on ecosystem resilience, ALS has a mixed influence, while WS has a negative impact. Robustness checks using cluster-robust standard errors and alternative model specifications confirmed the stability and direction of the estimated coefficients. The conclusions support the promotion of forest conservation policies, sustainable water resource management, and ecosystem-friendly agriculture practices as main directions for enhancing the capacity of ecosystems to respond to human and climate pressures. Full article

A high-quality set of typical scenarios is significant for power grid planning. Existing construction methods for typical scenarios do not account for the spatiotemporal correlations among renewable energy plant outputs, failing to adequately reflect the distribution characteristics of original scenarios. To address the issues mentioned above, this paper proposes a construction method for typical scenarios considering spatiotemporal correlations, providing high-quality typical scenarios for power grid planning. Firstly, a symmetric spatial correlation matrix and a temporal autocorrelation matrix are defined, achieving quantitative representation of spatiotemporal correlations. Then, distributional differences between typical and original scenarios are quantified from multiple dimensions, and a scenario reduction model considering spatiotemporal correlations is established. Finally, the genetic algorithm is improved by incorporating adaptive parameter adjustment and an elitism strategy, which can efficiently obtain high-quality typical scenarios. A case study involving five wind farms and fourteen photovoltaic plants in Belgium is presented. The rate of error between spatial correlation matrices of original and typical scenario sets is lower than 2.6%, and the rate of error between temporal autocorrelations is lower than 2.8%. Simulation results demonstrate that typical scenarios can capture the spatiotemporal correlations of original scenarios. Full article

The spectral variation hypothesis (SVH) posits that the complexity of spectral information in remote sensing imagery can serve as a proxy for regional biodiversity. However, the relationship between spectral diversity (SD) and biodiversity differs for different environmental conditions. Previous SVH studies often overlooked these differences. We utilized species data from field surveys in Inner Mongolia and drone-derived multispectral imagery to establish a quantitative relationship between SD and biodiversity. A geographically weighted regression (GWR) model was used to describe the SD–biodiversity relationship and map the biodiversity indices in different experimental areas in Inner Mongolia, China. Spatial autocorrelation analysis revealed that both SD and biodiversity indices exhibited strong and statistically significant spatial autocorrelation in their distribution patterns. Among all spectral diversity indices, the convex hull area exhibited the best model fit with the Margalef richness index (Margalef), the coefficient of variation showed the strongest predictive performance for species richness (Richness), and the convex hull volume provided the highest explanatory power for Shannon diversity (Shannon). Predictions for Shannon achieved the lowest relative root mean square error (RRMSE = 0.17), indicating the highest predictive accuracy, whereas Richness exhibited systematic underestimation with a higher RRMSE (0.23). Compared to the commonly used linear regression model in SVH studies, the GWR model exhibited a 4.7- to 26.5-fold improvement in goodness-of-fit. Despite the relatively low R² value (≤0.59), the model yields biodiversity predictions that are broadly aligned with field observations. Our approach explicitly considers the spatial heterogeneity of the SD–biodiversity relationship. The GWR model had significantly higher fitting accuracy than the linear regression model, indicating its potential for remote sensing-based biodiversity assessments. 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

Effective detection of center-pivot irrigation systems is crucial in understanding agricultural activity and managing groundwater resources for sustainable uses, especially in semi-arid regions such as North Dakota, where irrigation primarily depends on groundwater resources. In this study, we have adopted YOLOv11 to detect the center-pivot irrigation systems using multiple remote sensing datasets, including Landsat 8, Sentinel-2, and NAIP (National Agriculture Imagery Program). We developed an ArcGIS custom tool to facilitate data preparation and large-scale model execution for YOLOv11, which was not included in the ArcGIS Pro deep learning package. YOLOv11 was compared against other popular deep learning model architectures such as U-Net, Faster R-CNN, and Mask R-CNN. YOLOv11, using Landsat 8 panchromatic data, achieved the highest detection accuracy (precision: 0.98; recall: 0.91; and F1-score: 0.94) among all tested datasets and models. Spatial autocorrelation and hotspot analysis revealed systematic prediction errors, suggesting a need to adjust training data regionally. Our research demonstrates the potential of deep learning in combination with GIS-based workflows for large-scale irrigation system analysis, adopting precision agricultural technologies for sustainable water resource management. Full article

In InSAR processing, optimizing baselines by selecting appropriate interferometric pairs is crucial for ensuring interferogram quality and improving InSAR monitoring accuracy. However, in multi-temporal InSAR processing, the quality of interferometric pairs is constrained by spatiotemporal baseline parameters and surface scattering characteristics. Traditional selection methods, such as those based on average coherence thresholding, consider only a single factor and do not account for the interactions among multiple factors. This study introduces a principal component analysis (PCA) method to comprehensively analyze four factors: temporal baseline, spatial baseline, NDVI difference, and coherence, scientifically setting weights to achieve precise selection of interferometric pairs. Additionally, the GACOS (Generic Atmospheric Correction Online Service) atmospheric correction product is applied to further enhance data quality. Taking the Haikou Phosphate Mine area in Kunming, Yunnan, as the study area, surface deformation information was extracted using the SBAS-InSAR technique, and the spatiotemporal characteristics of subsidence were analyzed. The research results show the following: (1) compared with other methods, the PCA-based interferometric pair optimization method significantly improves the selection performance. The minimum value decreases to 0.248 rad, while the mean and standard deviation are reduced to 1.589 rad and 0.797 rad, respectively, effectively suppressing error fluctuations and enhancing the stability of the inversion; (2) through comparative analysis of the effective pixel ratio and standard deviation of deformation rates, as well as a comprehensive evaluation of the deformation rate probability density function (PDF) distribution, the PCA optimization method maintains a high effective pixel ratio while enhancing sensitivity to surface deformation changes, indicating its advantage in deformation monitoring in complex terrain areas; (3) the combined analysis of spatial autocorrelation (Moran’s I coefficient) and spatial correlation coefficients (Pearson and Spearman) verified the advantages of the PCA optimization method in maintaining spatial structure and result consistency, supporting its ability to achieve higher accuracy and stability in complex surface deformation monitoring. In summary, the PCA-based baseline optimization method significantly improves the accuracy of SBAS-InSAR in surface subsidence monitoring, fully demonstrating its reliability and stability in complex terrain areas, and providing a solid technical support for dynamic monitoring of surface subsidence in mining areas. Full article

To support China’s “3060” dual carbon targets, this study quantitatively evaluates the spatial–temporal characteristics and influencing factors of carbon emission performance (CEP) across administrative levels. While prior research has examined CEP patterns, a systematic comparison of factor contributions at different levels—particularly from global and local perspectives—is lacking. This study addresses this gap by analyzing CEP in 31 provinces and 333 prefecture-level cities (2003–2020) using a coupling coordination degree model to measure CEP, spatial autocorrelation indices (Moran’s I) to assess global/local dependence, static/dynamic Spatial Durbin model (SDM/DSDM) with two-way fixed effects to compare global impacts, and geographically and temporally weighted regression (GTWR) to quantify spatiotemporal heterogeneity. The results show the following: (1) CEP showed consistent growth at both levels with positive spatial autocorrelation, revealing significantly richer clustering patterns at the prefectural rather than provincial level. (2) From a global perspective, influencing factors’ contributions to CEP vary significantly between levels. Provincially, dominant factors rank as time-lagged CEP(CEP_lag)> proportion of built-up land(P_built) > spatial lag of CEP(W×CEP) > fractional vegetation coverage (lnFVC); while prefecturally, CEP_lag > spatial error coefficient(rho) > W×CEP > P_built, with the proportion of secondary industry in GDP (GDP2)/proportion of tertiary industry in GDP (GDP3) gaining greater significance. (3) Local regression results reveal significant spatiotemporal heterogeneity in CEP influencing factors. lnFVC and W×CEP show the most distinct differences between levels, while land-use factors like P_built and nighttime light index (NTL) exhibit unstable spatiotemporal effects. The study underscores the need for scale-specific policies addressing spatial spillovers and local heterogeneity, providing actionable insights for China’s carbon mitigation strategies. Full article