Spatial Correlation Research Articles

A convolution neural network integrating climate variables and spatial-temporal properties to predict influenza trends

The spread of influenza is contingent upon a multitude of outbreak-related factors, including viral mutation, climate conditions, acquisition of immunity, crowded environments, vaccine efficacy, social gatherings, and the health and age profiles of individuals in contact with infected individuals. An epidemic in the region impacted by spatial transmission risk from adjacent regions. A few influenzas epidemic models start highlighting the spatial correlations between influenza patients and geographically adjacent regions. The proposed model is based on the concept of climatic, immunization, and spatial correlations which are represented by a convolution neural network (CNN) for influenza epidemic forecasting. This study presents an integration of three determinants for predicting influenza outbreaks, multivariate climate data, spatial data on influenza vaccination, and spatial-temporal data of historical influenza patients. The performance of three comparison models, CNN, recurrent neural network (RNN), and long short-term memory (LSTM) was compared by the root mean squared error metric (RMSE). The findings revealed that the CNN model represents human interaction at intervals of 12, 16, 20, 24, and 28 weeks resulting in the best effectiveness of the lowest RMSE=0.00376 with learning rate=0.0001.

Threshold-Free Symbol Decision Based on Spatial 4PPM Correlation for Uplink Screen Camera Communication Using Low-Luminance SDM Screen

Threshold-Free Symbol Decision Based on Spatial 4PPM Correlation for Uplink Screen Camera Communication Using Low-Luminance SDM Screen

Activity recognition based on spatio-temporal features with transfer learning

<span lang="EN-US">Human action recognition (HAR) has gathered significant attention recently due to its high demand in various application domains. We explore different CNN architectures for extracting Spatio-Temporal features. In addition to spatial correlation in 2D images, video actions also own the temporal domain attributes. Many research attempts addressed simple activity recognition, but complex activity recognition is still challenging. In the proposed work, the concept of transfer learning consists of two pre-trained networks, ResNet50 and VGG16, which are used to extract Spatiotemporal features and are later fused using the averaging method. Pre-processed frames are considered for the pre-trained network, the pre-processing consists of frame extraction, resizing, and contrast enhancement. The fused features are trained using a multiclass non-linear SVM classifier that achieved an accuracy of 89.71% on the UCF-101 database.</span>

Probabilistic microkinetic modeling: Species balance equations for a catalyst surface containing multiple short-range order parameters to capture spatial correlations.

Adsorbed molecules on a catalyst almost always arrange themselves in a manner that is far from perfectly random, which gives rise to spatial correlations. These correlations are a result of the interactions between the adsorbed species (adspecies) as well as elementary processes such as diffusion and reaction events that shape the adspecies arrangements. Despite their importance, spatial correlations are usually ignored while writing species balance equations for the modeling of heterogeneous catalytic systems. Recently, we have introduced a probabilistic microkinetic modeling (p-MKM) framework that aims at incorporating spatial correlations in the form of a short-ranged order (SRO) parameter into species balance equations. Here, we extend the approach to catalytic systems of higher complexity, namely, longer interactions and multiple species. This is made possible by including multiple pair probabilities in the p-MKM model for the first time. The interplay between different SRO parameters is probed. An important consideration is how many pair probabilities should be included to capture the underlying complexity with sufficient accuracy.

Healthcare-seeking behavior and spatial variation of internal migrants with chronic diseases: a nationwide empirical study in China.

Individuals migrating with chronic diseases often face substantial health risks, and their patterns of healthcare-seeking behavior are commonly influenced by mobility. However, to our knowledge, no research has used spatial statistics to verify this phenomenon. Utilizing data from the China Migrant Dynamic Survey of 2017, we conducted a geostatistical analysis to identify clusters of chronic disease patients among China's internal migrants. Geographically weighted regressions were utilized to examine the driving factors behind the reasons why treatment was not sought by 711 individuals among a population sample of 9272 migrant people with chronic diseases. The results indicate that there is a spatial correlation in the clustering of internal migrants with chronic diseases in China. The prevalence is highly clustered in Zhejiang and Xinjiang in north-eastern China. Hotspots were found in the northeast (Jilin and Liaoning), the north (Hebei, Beijing, and Tianjin), and the east (Shandong) and also spread into surrounding provinces. The factors that affect the migrants with no treatment were found to be the number of hospital beds per thousand population, the per capita disposable income of medical care, and the number of participants receiving health education per 1000 Chinese population. To rectify this situation, the local government should "adapt measures to local conditions." Popularizing health education and coordinating the deployment of high-quality medical facilities and medical workers are effective measures to encourage migrants to seek reasonable medical treatment.

Memory Augmentation and Non-Local Spectral Attention for Hyperspectral Denoising

In this paper, a novel hyperspectral denoising method is proposed, aiming at restoring clean images from images disturbed by complex noise. Previous denoising methods have mostly focused on exploring the spatial and spectral correlations of hyperspectral data. The performances of these methods are often limited by the effective information of the neighboring bands of the image patches in the spectral dimension, as the neighboring bands often suffer from similar noise interference. On the contrary, this study designed a cross-band non-local attention module with the aim of finding the optimal similar band for the input band. To avoid being limited to neighboring bands, this study also set up a memory library that can remember the detailed information of each input band during denoising training, fully learning the spectral information of the data. In addition, we use dense connected module to extract multi-scale spatial information from images separately. The proposed network is validated on both synthetic and real data. Compared with other recent hyperspectral denoising methods, the proposed method not only demonstrates good performance but also achieves better generalization.

Unraveling the collinearity in short-range order parameters for lattice configurations arising from topological constraints.

In multicomponent lattice problems, for example, in alloys and at crystalline surfaces and interfaces, atomic arrangements exhibit spatial correlations that dictate the kinetic and thermodynamic phase behavior. These correlations emerge from interparticle interactions and are frequently reported in terms of the short-range order (SRO) parameter. Expressed usually in terms of pair distributions and other cluster probabilities, the SRO parameter gives the likelihood of finding atoms/molecules of a particular type in the vicinity of other atoms. This study focuses on fundamental constraints involving the SRO parameters that are imposed by the underlying lattice topology. Using a data-driven approach, we uncover the interrelationships between different SRO parameters (e.g., pairs, triplets, and quadruplets) on a lattice. The main finding is that while some SRO parameters are independent, the remaining are collinear, i.e., the latter are dictated by the independent ones through linear relationships. A kinetic and thermodynamic modeling framework based on these constraints is introduced.

Spatial-temporal evolution and prediction of carbon storage in Mohe city by linking the logistic-CA-Markov and InVEST models

Carbon storage plays a vital role in the provision of ecosystem services, and the alteration of land use has a profound influence on the carbon storage capacity of ecosystems. Therefore, in this study, two scenarios of natural evolution scenario (NES) and ecological conservation scenario (ECS) were set up based on the development of Mohe City,China. Meanwhile, a coupled model of LUCC and ecosystem carbon storage was established using Logistic-CA-Markov and InVEST models, as well as optimal parameters_based geographical Detector and GeoDa, to predict the distribution and change of ecosystem carbon storage based on LUCC in the future. The effects of different influencing factors on the spatial differentiation of carbon storage were also explored. The results show that: (1) From 1980 to 2020, the land use type in Mohe City was dominated by the forest and its area decreased; the area of wetland increased. From 2020 to 2040, ecological benefits will be shown under the ECS, with an increase in forest and a slower expansion of built-up. (2) Carbon loss in Mohe City from 1980 to 2020 was 4.04Tg. Under the ECS carbon storage increased slightly by 0.2Tg. Soil carbon storage was the main carbon pool in Mohe City, and forest was the largest contributor. The carbon storage of Mohe city in 2030 and 2040 has a strong positive spatial correlation. Hot spots in more than a cold spots area, the high value area is concentrated in the east, low concentrated in urban areas. (3) Apart from LUCC, mean annual precipitation was the most significant factors affecting the spatial differentiation variability of carbon storage. The interactions of mean annual precipitation and population density with other factors exhibit a non-linear enhancement,which had a coefficient of 21.91%. This study contributes to a deeper understanding of the relationship between LUCC and carbon storage.

The social-ecological benefits of grain for green program based on coupled coordination network: Taking the China’s Loess Plateau as an example

The Grain for Green Program (GFGP) is the world’s largest payment for ecosystem services (PES). In the past 20 years, the Chinese government has implemented three stages of GFGP, which have profoundly affected the social-ecological benefits (SEB) of China’s Loess Plateau (LP), one of the regions with the most severe soil erosion and the most fragile ecological environment in the world. Therefore, we explored the changes and coupling coordination relationships of LP’s SEB in the past three stages of GFGP, and examined the spatial correlation network relationships between various units and their roles of GFGP, which was particularly important for the rationality and implementation direction of LP’s GFGP in the future. We found that in different GFGP stages, changes in LP’s SEB showed stage differences and spatiotemporal differences. The comprehensive social benefits (CSB) changed slowly (1.342 dropped to 1.232), and there was an obvious “northwest-southeast” regional division in space. The changes in comprehensive ecological benefits (CEB) fluctuated more obviously (1.481, 1.638, 0.597). After experiencing significant improvements of CEB in two stages, it entered a bottleneck period in the third stage. For a long time, the LP’s SEB have been in a state of coupling coordination, but the coupling coordination degree dropped to 0.493 in the third stage. With climate change, the continuous increase in potential grain yield has led to increasing losses in CSB. It was difficult to promote further coordination of LP’s SEB by accelerating the improvement of CEB. The coupling coordination degree has gradually moved from overall coupling coordination to mild imbalance. At the same time, the overall density of the SEB coupling coordination space network correlation was 0.218, which has shown an obvious network structure, and formed four major sectors: main beneficiary sector, broker sector, net overflow sector, and bidirectional overflow sector in each stage of GFGP. The role of each county has undergone regional changes in the entire GFGP, and there were recurring power transmission mechanisms and complex network formation factors. Based on this, we proposed to promote the coupling coordination and positive network transmission of SEB in the next stage of LP’s GFGP through “low-carbonization of agricultural production”, “further improvement of the compensation mechanism”, and “fusion of power transmission mechanism and network formation factors”.

Background-oriented schlieren experimental and simulation study of the effect of sidewalls on optical turbulence anisotropy in a thermal convection turbulence simulator

The anisotropic optical structure characteristics within the Rayleigh-Bénard (RB) turbulence simulator have been investigated through experiment and simulation. We collected data using background-oriented schlieren technology and extracted displacements using optical flow algorithms. The wavefront phase can be recovered using the conjugate gradient iterative algorithm. We analyzed the optical structural characteristics within the turbulence simulator under different temperature differences. It is observed that with an increased temperature difference, the spatial correlation of the refractive index field decreases at an accelerated rate as the separation distance increases, and the degree of anisotropy characterized by the phase structure function increases. Large eddy simulation and split-step beam propagation method complement experimental results. Numerical simulations demonstrate consistency with experimental data. The results show that the von Karman power spectral density fits the simulation and experimental phase structure function very well relative to the Kolmogorov power spectral density, which demonstrates that the simulation system combining dynamics simulation and optical simulation can achieve power spectral characteristics that are consistent with reality. The experimental and simulation results are crucial for understanding the flow field's optical structural properties inside the turbulence simulator. They will help to construct a suitable turbulence simulator for optical engineering applications and establish a rational scheme for analyzing its internal optical structural properties.

SPATIAL MODELING OF POVERTY IN BENGKULU PROVINCE WITH MIXED GEOGRAPHICALLY WEIGHTED REGRESSION

The percentage of poor people in Bengkulu Province is high from year to year. The poverty rate in Bengkulu Province also tends to fluctuate. If there is a decrease in the poverty rate, the decrease is relatively small. Poverty in the regions of Bengkulu Province also varies from district to district, subdistrict, and village to village, because poverty data is spatial data that varies regionally. The diversity of poverty data in Bengkulu Province is influenced by spatial effects, namely spatial dependency and spatial heterogeneity. Spatial dependency occurs due to spatial error correlation in cross section data, while spatial heterogeneity occurs due to random area effects, which is the difference between one region and another. Therefore, classical methods are not qualified enough to analyze the resulting diversity. This research will model the poverty of each district/city in Bengkulu Province using Mixed Geographically Weighted Regression (MGWR), because this method is quite complex in modeling data that contains spatial heterogeneity and variations in geospatial data. This modeling aims to identify and analyze poverty indicators in Bengkulu Province spatially, namely based on poverty data in each district/city in Bengkulu Province. The results showed that by using the MGWR method, the variables that locally influence the percentage of extreme poor people in each district/city in Bengkulu Province are Female Household Head Gender and not having a waterheater . Meanwhile, the variable that has a global effect on the percentage of the extreme poor in each district/city in Bengkulu Province is not having a flat screen television ().

Spatial impact of the digital economy on low-carbon logistics efficiency in RCEP countries

A thorough study of the spatial impact of the digital economy on low-carbon logistics efficiency would be greatly significant for Regional Comprehensive Economic Partnership (RCEP) countries to improve low-carbon logistics efficiency and achieve sustainable cooperation. This study constructs a theoretical framework from the perspective of spatial effects on the impact of the digital economy on low-carbon logistics efficiency in RCEP countries. The entropy method was used to measure the level of digital economic development. The super-efficiency SBM model was used to measure low-carbon logistics efficiency. Spatial feature analysis was conducted using kernel density estimation and Moran's index, followed by empirical analysis using spatial econometric models to examine the spatial impact of the digital economy on low-carbon logistics efficiency in RCEP countries. The results indicate that in RCEP countries, both low-carbon logistics efficiency and the level of digital economic development exhibit significant spatial positive correlation. Furthermore, the digital economy can promote low-carbon logistics efficiency in economically neighboring countries through spatial spillover effects. The improvement of domestic low-carbon logistics efficiency can also promote low-carbon logistics efficiency in neighboring countries. This conclusion was supported by endogeneity tests and a convergence analysis. Additionally, the mechanism analysis revealed that improving the level of green energy can enhance the spatial spillover effects of the digital economy and promote low-carbon logistics efficiency. Finally, countermeasures and suggestions was proposed to improve the low-carbon logistics efficiency of RCEP countries through the digital economy.

Traffic Flow Prediction with Random Walks on Graph and Spatiotemporal Bidirectional Attention Transformer

Traffic flow prediction is crucial in intelligent transportation systems. Considering the severe disruptions caused by traffic accidents or congestion, a time series model is developed for traffic flow prediction based on multiple random walks on graphs (MRWG) and the bidirectional spatiotemporal attention mechanism (BSAM), which can adapt to both normal and exceptional situations. The MRWG mechanism is applied to capture spatial features of urban areas during traffic accidents and congestion, especially the spatial dependencies among neighboring regions. Further, a local position attention module is applied to acquire the spatial correlations between different regions to investigate their impact on the global area, while a local temporal attention module is adopted to extract short-term periodic time correlations from traffic flow data. Finally, a spatiotemporal bidirectional attention module is applied to simultaneously extract both the temporal and spatial correlations of the historical traffic flow data in order to generate the output prediction. Experiments have been conducted on NYCTaxi and NYCBike datasets with abnormal events, and the results indicate that the developed model can efficiently predict traffic flow in abnormal events, especially short-term traffic disruptions, outperforming the baseline methods under both abnormal and normal conditions.

Mesoscale Eddy Effects on Vertical Correlation of Sound Field and Array Gain Performance

To solve the problem of array detection performance in environments with mesoscale eddies, this study utilizes the Gaussian eddy model to describe the sound speed structure disturbed by eddies. Through numerical simulations, the corresponding sound field is obtained, and the transmission loss influenced by the eddy is analyzed. Furthermore, to investigate the relation between the array gain and spatial correlation in the eddy environments, the differences in vertical correlation at different positions and their effects on the vertical array gain of conventional beamforming (CBF) are studied. When the source is around the eddy center, the conclusions drawn are as follows: (1) The presence of an eddy changes the turning-point depth and the sound field distribution, significantly affecting the direct sound region and the first convergence zone, while having a minor impact on the first shadow zone. (2) In different eddy-induced environments, the first convergence zone maintains a high vertical correlation, but the vertical correlation of the direct sound region is greatly influenced by the eddy. (3) The array gain of CBF is consistent with the vertical correlation. When the correlation between each element of the sound field is great, the array gain increases with the number of array elements.

Semi-supervised correction model for turbulence-distorted images

Significant progress has been made in addressing turbulence distortion in recent years, but persistent challenges remain. Firstly, existing methods heavily rely on fully supervised optimization strategies and synthetic datasets, posing difficulties in effectively utilizing unlabeled real data for training. Secondly, most approaches construct networks in a straightforward manner, overlooking the representation model of phase distortion and point spread function (PSF) in spatial and channel dimensions. This oversight restricts the potential for distortion correction. To address these challenges, this paper proposes a semi-supervised atmospheric turbulence correction method based on the mean-teacher framework. Our approach imposes constraints on the unlabeled data of student networks using pseudo-labels generated by teacher networks, thereby enhancing the generalization ability by leveraging information from unlabeled data. Furthermore, we introduce to use no-reference image quality assessment criterion to select the most reliable pseudo-label for each unlabeled sample by predicting physical parameters that indicating the level of degradation. Additionally, we propose to combine sliding window-based self-attention with channel attention to facilitate local-global context interaction. This design is inspired by the representation of phase distortion and PSF, which can be characterized by coefficients and basis functions corresponding to the channel-wise representation of convolutional neural network features. Moreover, the base functions exhibit spatial correlation, akin to Zenike and Airy disks. Experimental results show that the proposed method surpasses state-of-the-art models.

The effects and influencing variables based on “supply‐direction‐demand” flow processing: Water provisioning services of Inner Mongolia's ecological shelters

AbstractThe ecological shelter serves a dual purpose of protecting the target ecosystem from damage and benefitting surrounding ecosystems. One important ecosystem service provided by such areas is water provisioning service, which has significant impacts on the surrounding ecosystems due to its flow characteristics. This study aimed to simulate water resources in Inner Mongolia and its surrounding watersheds in 2010 and 2015 using a simplified service path attribution networks (SPANs) model. The InVEST software was utilized to simulate water provisions, and five types of water consumption were used to calculate the total water consumption. Probability statistics with BBNs were utilized to classify variables influencing water provisioning services and calculate state probabilities under different levels of water provisioning services. The results indicated that the remaining water resources in Inner Mongolia were 3.85 × 1010 m3 in 2010 and 1.92 × 1010 m3 in 2015. Additionally, the total net outflow from Inner Mongolia was 2.37 × 1010 m3 in 2010 and 1.58 × 1010 m3 in 2015. Inner Mongolia not only supplies water resources for its own needs but also supports the surrounding areas. Key variables influencing the amount of water‐provisioning services include precipitation, level of economic development, topography, and land use. This research highlights the considerable regional and transregional effects of water provisioning services provided by Inner Mongolia's ecological shelters by delineating the “supply‐flow‐demand” pathway, addressing the spatial correlation deficiency between supply and demand in the field of ecosystem service flow research. The findings of this study, focusing on a typical ecological shelter, can serve as a reference for future research in other similar shelters.

Dependency-aware deep generative models for multitasking analysis of spatial omics data.

Spatially resolved transcriptomics (SRT) technologies have significantly advanced biomedical research, but their data analysis remains challenging due to the discrete nature of the data and the high levels of noise, compounded by complex spatial dependencies. Here, we propose spaVAE, a dependency-aware, deep generative spatial variational autoencoder model that probabilistically characterizes count data while capturing spatial correlations. spaVAE introduces a hybrid embedding combining a Gaussian process prior with a Gaussian prior to explicitly capture spatial correlations among spots. It then optimizes the parameters of deep neural networks to approximate the distributions underlying the SRT data. With the approximated distributions, spaVAE can contribute to several analytical tasks that are essential for SRT data analysis, including dimensionality reduction, visualization, clustering, batch integration, denoising, differential expression, spatial interpolation, resolution enhancement and identification of spatially variable genes. Moreover, we have extended spaVAE to spaPeakVAE and spaMultiVAE to characterize spatial ATAC-seq (assay for transposase-accessible chromatin using sequencing) data and spatial multi-omics data, respectively.

Generating synthetic rainfall fields by R‐vine copulas applied to seamless probabilistic predictions

AbstractMany post‐processing methods improve forecasts at individual locations but remove their correlation structure. However, this information is essential for forecasting larger‐scale events, such as the total precipitation amount over areas like river catchments, which are relevant for weather warnings and flood predictions. We propose a method to reintroduce spatial correlation into a post‐processed forecast using an R‐vine copula fitted to historical observations. The method rearranges predictions at individual locations and ensures that they still exhibit the post‐processed marginal distributions. It works similarly to well‐known approaches, like the “Schaake shuffle” and “ensemble copula coupling.” However, compared to these methods, which rely on a ranking with no ties at each considered location in their source for spatial correlation, the copula serves as a measure of how well a given arrangement compares with the observed historical distribution. Therefore, no close relationship is required between the post‐processed marginal distributions and the spatial correlation source. This is advantageous for post‐processed seamless forecasts in two ways. First, meteorological parameters such as the precipitation amount, whose distribution has an atom at zero, have rankings with ties. Second, seamless forecasts represent an optimal combination of their input forecasts and may spatially shifted from them at scales larger than the areas considered herein, leading to non‐reasonable spatial correlation sources for the well‐known methods. Our results indicate that the calibration of the combination model carries over to the output of the proposed model, that is, the evaluation of area predictions shows a similar improvement in forecast quality as the predictions for individual locations. Additionally, the spatial correlation of the forecast is evaluated with the help of object‐based metrics, for which the proposed model also shows an improvement compared to both input forecasts.

STFEformer: Spatial–Temporal Fusion Embedding Transformer for Traffic Flow Prediction

In the realm of Intelligent Transportation Systems (ITSs), traffic flow prediction is crucial for multiple applications. The primary challenge in traffic flow prediction lies in the handling and modeling of the intricate spatial–temporal correlations inherent in transport data. In recent years, many studies have focused on developing various Spatial–Temporal Graph Neural Networks (STGNNs), and researchers have also begun to explore the application of transformers to capture spatial–temporal correlations in traffic data. However, GNN-based methods mainly focus on modeling spatial correlations statically, which significantly limits their capacity to discover dynamic and long-range spatial patterns. Transformer-based methods have not sufficiently extracted the comprehensive representation of traffic data features. To explore dynamic spatial dependencies and comprehensively characterize traffic data, the Spatial–Temporal Fusion Embedding Transformer (STFEformer) is proposed for traffic flow prediction. Specifically, we propose a fusion embedding layer to capture and fuse both native information and spatial–temporal features, aiming to achieve a comprehensive representation of traffic data characteristics. Then, we introduce a spatial self-attention module designed to enhance detection of dynamic and long-range spatial correlations by focusing on interactions between similar nodes. Extensive experiments conducted on three real-world datasets demonstrate that STFEformer significantly outperforms various baseline models, notably achieving up to a 5.6% reduction in Mean Absolute Error (MAE) on the PeMS08 dataset compared to the next-best model. Furthermore, the results of ablation experiments and visualizations are employed to clarify and highlight our model’s performance. STFEformer represents a meaningful advancement in traffic flow prediction, potentially influencing future research and applications in ITSs by providing a more robust framework for managing and analyzing traffic data.

Self-paced Gaussian-based graph convolutional network: predicting travel flow and unravelling spatial interactions through GPS trajectory data

ABSTRACT Spatial interaction research is particularly important for geographical analyses, as it plays a crucial role in extracting travel patterns. However, previous studies on spatial interactions have not adequately considered regional population variations over time, resulting in insufficiently precise travel predictions. Moreover, the threshold of spatial correlations is difficult to determine. Existing studies have assumed fully connected spatial correlation matrices, which is not realistic. To address these limitations, we proposed the Self-paced Gaussian-Based Graph Convolutional Network (SG-GCN) to automatically estimate the threshold of spatial correlations for travel flow predictions. It incorporates a temporal dimension into spatial relationship matrices to enhance the accuracy of vehicle flow predictions. In particular, Gaussian-based GCN identifies patterns in a time series of regional flows, enabling more precise capturing of spatial relationships while fusing node and edge features. Building on this model, self-paced contrastive learning automatically sets thresholds to determine the presence or absence of spatial relationships. The model's performance was verified through two empirical case studies conducted in New York City, USA, and Ningbo, China, using 2.8 million bicycle-sharing records and 1.25 million taxi trip records, respectively. The proposed model helps delineate mobility patterns in cities of varying scales and with different modes of transportation.