Search Results (991)

Sub-Saharan Africa (SSA) is a region vulnerable to extreme weather events due to its low level of adaptive capacity. In recent decades, SSA has been punctuated by more intense climatic phenomena that severely affect its population. Therefore, this study evaluates the performance of the ERA5 and CHIRPS datasets, and the spatio-temporal evolution of extreme weather indices and their potential relationship/response to climate variability modes in the Pacific, Indian, and Atlantic oceans, namely, the El Niño−Southern Oscillation, Indian Ocean Dipole, and Tropical Atlantic Variability (ENSO, IOD, and TAV). The CHIRPS dataset showed strong positive correlations with CPC in spatial patterns and similarity in simulating interannual variability and in almost all seasons. Based on daily CHIRPS and CPC data, nine extreme indices were evaluated focusing on regional trends and change detection, and the maximum lag correlation method was applied to investigate fluctuations caused by climate variability modes. The results revealed a significant decrease in total precipitation (PRCPTOT) in north−central SSA, accompanied by a reduction in Consecutive Wet Days (CWDs) and maximum 5-day precipitation indices (RX5DAYS). At the same time, there was an increase in Consecutive Dry Days (CDDs) and maximum rainfall in 1 day (RX1DAY). With regard to temperatures, absolute minimums and maximums (TNn and TXn) showed a tendency to increase in the center−north and decrease in the south of the SSA, while daily maximums and minimums (TXx and TNx) showed the opposite pattern. The IOD, TAV, and ENSO modes of climate variability influence temperature and precipitation variations in the SSA, with distinct regional responses and lags between the basins. Full article

The three-dimensional (3D) reconstruction of the hepatic duct tree is significant for the minimally invasive surgery of hepatobiliary stone disease, which can be influenced by the quantity of the CT scans of hepatic ducts. If insufficient CT scans with low inter-slice resolution are directly utilized for 3D reconstruction, discontinuities and gaps will emerge in the reconstructed hepatic duct tree. In this paper, a novel end-to-end deep learning framework is designed for the inter-slice super-resolution segmentation of the CT slices of hepatic ducts, which can improve the 3D reconstruction performance in the inter-slice dimension. Specifically, the framework cascades into an inter-slice super-resolution subnetwork and a segmentation subnetwork. A deep learning network is introduced as the inter-slice super-resolution subnetwork to generate an intermediate slice between two adjacent CT slices in the simulated CT scans with low inter-slice resolution. To capture the spatiotemporal correlation existing in the CT scans of hepatic ducts, the ConvLSTM is introduced into the U-Net-like segmentation subnetwork in the high-dimensional feature space. To further suppress the problems of discontinuities and gaps, a structure-aware loss function is proposed by incorporating the structural similarity index measure (SSIM) as a regulator to dynamically assign the contribution of the generated CT slice to the total loss of the designed framework. Experimental results demonstrate that our proposed framework performs better segmentations for hepatic ducts than several existing deep learning networks with the performance of 0.7690 DICE and a 0.7712 F1-score, which is beneficial for the 3D reconstruction of the hepatic duct tree. Full article

Golfers must modify their motor patterns when the demands of a putting task change. The objective was to compare joint angles and putter kinematics during putting at two distances and inclines. Recreational golfers (n = 14) completed putts over four conditions: 3-foot putts on flat and incline surfaces, and 7-foot putts on flat and incline surfaces. A Vicon motion capture system measured kinematic data. Joint angles, putter angles, and spatiotemporal variables were calculated. Analysis of variance compared spatiotemporal variables, and statistical parametric mapping compared angles between putts. There were faster putter head and ball velocities during longer and incline putts. The amplitude and time of backswing increased with longer putts. Longer putts resulted in increased trunk axial rotation during backswing, downswing, and follow-through, while incline putts only resulted in greater rotation during follow-through. There were minimal differences in shoulder angle. There was greater head rotation toward the hole during all putting phases for longer putts and during follow-through for incline putts. The trunk is the primary mechanism to increase putter head amplitude, and thereby velocity, when putting from longer distances. A similar strategy could be used when putting uphill. Additional work should confirm these results in highly skilled golfers. Full article

Arctic sea ice has a significant effect on global climate change, ship navigation, Arctic ecosystems, and human activities. Therefore, it is essential to produce high-resolution sea ice maps that accurately represent the geographical distribution of various sea ice types. Based on deep learning technology, many automatic sea ice classification algorithms have been developed using synthetic aperture radar (SAR) imagery over the last decade. However, sea ice classification faces two vital challenges: (1) it is difficult to distinguish sea ice types with close developmental stages solely from SAR images and (2) an imbalanced sea ice dataset has a significantly negative effect on ice classification model performance. In this article, a spatio-temporal deep learning model—the Dynamic Multi-Layer Perceptron (MLP)—is utilized to classify 10 sea ice types automatically. It consists of a SAR image branch and a spatio-temporal branch, which extracts SAR image features and spatio-temporal distribution characteristics of sea ice, respectively. By projecting similar image features to different positions in the spatio-temporal feature space dynamically, the Dynamic MLP model effectively distinguishes between similar sea ice types. Furthermore, to reduce the impact of data imbalance on model performance, the dynamic curriculum learning (DCL) method is used to train the Dynamic MLP model. Experimental results demonstrate that our proposed method outperforms the long short-term memory (LSTM) network approach in distinguishing between sea ice types with similar developmental stages. Moreover, the DCL training method can also effectively improve model performance in identifying minority ice types. Full article

Urban growth, a pivotal characteristic of economic development, brings many environmental and ecological challenges. Modeling urban growth is essential for understanding its spatial dynamics and projecting future trends, providing insights for effective urban planning and sustainable development. This study aims to assess the spatiotemporal patterns of urban growth and morphological evolution in mainland Shanghai from 2016 to 2060 using the SLEUTH model under multiple growth scenarios based on the Shanghai Urban Master Plan (2017–2035). A comprehensive set of urban growth metrics and quadrant analysis were employed to quantify the magnitude, rate, intensity, and direction of urban growth, as well as morphological evolution, over time. We found that (1) significant urban growth was observed across most scenarios, with the exception of stringent land protection. The most substantial growth occurred prior to 2045 with an obvious north–south disparity, where southern regions demonstrated more pronounced increases in urban land area and urbanization rates. (2) The spatiotemporal patterns of the rate and intensity of urban growth exhibited similar characteristics. The spatial pattern followed a “concave shape” pattern and displayed anisotropic behavior, with the high values for these indicators primarily observed before 2025. (3) The urban form followed a diffusion–coalescence process, with patch areas dominated by the infilling mode and patch numbers dominated by the edge-expansion mode. This resulted in significant alternating urban growth models in the infilling, edge-expansion, and leapfrog modes over time, influenced by varying protection intensities. These findings provide valuable insights for forward-looking urban planning, land use optimization, and the support of sustainable urban development. Full article

Qionghai Lake is an important freshwater source in the Yunnan–Guizhou Plateau. However, cyanobacterial blooms have been observed recently in Qionghai Lake, but their formation mechanism and control management are not well understood. Herein, phytoplankton, zooplankton, eutrophication, nutrients, and biochemical indices were measured in Qionghai Lake from May 2022 to April 2023. The results showed that cyanobacterial blooms in Qionghai Lake predominated in Anabaena sp. with a density of 1.11 × 10⁷–18.87 × 10⁷ cells/L. Anabaena blooms started in the northwestern area of Qionghai Lake in November 2022 and then expanded to the entire lake until it peaked and subsided in February 2023. Protozoa dominated in zooplankton while having no significant relationship with Anabaena blooms in Qionghai Lake. The trophic level index and chlorophyll a showed similar spatiotemporal trends with Anabaena sp. density, and water quality in the northwest of the Qionghai Lake was worse than in other parts. Total nitrogen (TN) and total phosphorus (TP) were 0.41–0.54 and 0.021–0.045 mg/L from November 2022 to February 2023. TN and TP were positively correlated with Anabaena sp. density, but TP was the most significant environmental factor affecting Anabaena bloom in Qionghai Lake. These findings might provide essential information for improving bloom control and water quality remediation in Qionghai Lake. Full article

Under regional environmental conditions such as open-pit mines and construction sites, there are usually fixed GNSS measurement points. Around these fixed stations, there are also mobile GNSS measurement modules. These mobile measurement modules offer advantages such as low power consumption, low cost, and large data volume. However, due to their low accuracy, these modules can only provide approximate positions as monitoring data, such as for vehicle management in open-pit mines. To extract more information from the existing large volume of low-accuracy data, it is necessary to process these low-accuracy data. Under conditions of the same time and space in a small area, factors affecting measurement accuracy can be comprehensively considered. By analyzing the temporal GNSS data within the same spatiotemporal small region and understanding the variation patterns of measurement errors, a general equation for measurement error variation can be formulated. Using filtering methods, the data quality can be improved. Through the analysis of the experimental data in this study, it was found that the variation patterns of measurement data obtained by devices of the same accuracy during the same time period are generally consistent. After applying filtering methods, the measurement accuracy of each station improved by up to approximately 95.9%, with a minimum improvement of approximately 84.4%. Under the condition of a 95% confidence level, the reliability increased by up to approximately 73.2%, with a minimum improvement of approximately 58.2%. These experimental results fully demonstrate that under regional spatiotemporal conditions, the temporal data obtained by GNSS measurement devices with similar accuracy exhibit similar error distribution patterns. Applying the same filtering method can significantly improve the accuracy and reliability of measurement data. Full article

This study is based on the Taihang Mountain Area in Southern Hebei Province, focusing on battlefield sites in She County. Interdisciplinary theories and methods such as field investigation, architectural typology, and GIS spatial analysis are used to establish the protection roadmap and geographic information database for battlefield sites. The aim is to reveal the architectural characteristics and spatiotemporal distribution patterns of battlefield sites, evaluate its current status and constraints, and propose targeted protection strategies and restoration measures. Research has found that battlefield sites combine the characteristics of traditional dwellings and village temples, presenting the characteristics of clear clustering distribution and low altitude with a system near the water in the spatial and temporal dimensions, with good natural adaptability and unique rural cultural landscapes. At the same time, in response to the constraints and challenges still faced by protection work, a protection strategy framework has been constructed from six aspects: strengthening daily maintenance, reducing human and natural damage, optimizing the property rights system, unifying responsibility subjects, comprehensive development and utilization, and international legislation and cooperation. Repair and protection guidelines have been established from five technical links: timber structure, joinery work, roof, wall, and ground. This study can lay the foundation for the cultural inheritance and sustainable protection of battlefield sites in the Taihang Mountain Area of Southern Hebei Province, and provide theoretical references and practical examples for similar research. Full article

Global climate fluctuations pose challenges not only to natural environments but also to the conservation and transmission of human cultural and historical heritage. World Heritage Sites are pivotal regions for studying climate change impacts and devising adaptation strategies, with remote sensing technology showcasing significant utility in monitoring these impacts, especially in the Mediterranean region’s diverse and sensitive climate context. Although existing work has begun to explore the role of remote sensing in monitoring the effects of climate change, detailed analysis of the spatial distribution and temporal trends of landscape stability remains limited. Leveraging remote sensing data and its derived products, this study assessed climate change impacts on the Causses and Cévennes Heritage Site, a typical Mediterranean heritage landscape. Specifically, this study utilized remote sensing data to analyze the trends in various climatic factors from 1985 to 2020. The landscape stability model was developed utilizing land cover information and landscape indicators to explore the landscape stability and its distribution features within the study area. Finally, we adopted the Geographical Detector to quantify the extent to which climatic factors influence the landscape stability’s spatial distribution across different periods. The results demonstrated that (1) the climate showed a warming and drying pattern during the study period, with distinct climate characteristics in different zones. (2) The dominance of woodland decreased (area proportion dropped from 76% to 66.5%); transitions primarily occurred among woodland, cropland, shrubland, and grasslands; landscape fragmentation intensified; and development towards diversification and uniformity was observed. (3) Significant spatiotemporal differences in landscape stability within the heritage site were noted, with an overall downward trend. (4) Precipitation had a high contribution rate in factor detection, with the interactive enhancement effects between temperature and precipitation being the most prominent. The present study delivers a thorough examination of how climate change affects the Causses and Cévennes Heritage Landscape, reveals its vulnerabilities, and offers crucial information for sustainable conservation efforts. Moreover, the results offer guidance for the preservation of similar Mediterranean heritage sites and contribute to the advancement and deepening of global heritage conservation initiatives. Full article

Establishing and managing nature reserves to mitigate wildlife habitat loss and fragmentation is challenging, particularly in the face of increasing human activity. To understand how wildlife coexists in environments affected by anthropogenic disturbances, we conducted a 19-month survey examining the Reeves’s pheasant (Syrmaticus reevesii) and Koklass pheasant (Pucrasia macrolopha) in the Anhui Tianma National Nature Reserve, China. Previous studies of large terrestrial birds focused primarily on livestock impacts, with less attention given to other human activities. We used occupancy models and performed daytime activity rhythm analysis based on camera trap data to examine the spatiotemporal responses of these species to human activities, livestock, and domestic dogs. The results showed that human disturbance activities within the reserve impact the distribution patterns of Reeves’s pheasant and Koklass pheasant, but the effect was not significant. In high-disturbance environments, both species adjusted their activity times to avoid direct or indirect interactions with humans, livestock, and dogs. These findings provide insights for replanning core and creation of buffer zones within the reserve and have broader implications for conservation strategies in similar habitats. Our study suggests that well-designed conservation objectives can balance species protection with sustainable human presence. Full article

A near-real-time change detection network can consistently identify unauthorized construction activities over a wide area, empowering authorities to enforce regulations efficiently. Furthermore, it can promptly assess building damage, enabling expedited rescue efforts. The extensive adoption of deep learning in change detection has prompted a predominant emphasis on enhancing detection performance, primarily through the expansion of the depth and width of networks, overlooking considerations regarding inference time and computational cost. To accurately represent the spatio-temporal semantic correlations between pre-change and post-change images, we create an innovative transformer attention mechanism named Sliding-Window Dissimilarity Cross-Attention (SWDCA), which detects spatio-temporal semantic discrepancies by explicitly modeling the dissimilarity of bi-temporal tokens, departing from the mono-temporal similarity attention typically used in conventional transformers. In order to fulfill the near-real-time requirement, SWDCA employs a sliding-window scheme to limit the range of the cross-attention mechanism within a predetermined window/dilated window size. This approach not only excludes distant and irrelevant information but also reduces computational cost. Furthermore, we develop a lightweight Siamese backbone for extracting building and environmental features. Subsequently, we integrate an SWDCA module into this backbone, forming an efficient change detection network. Quantitative evaluations and visual analyses of thorough experiments verify that our method achieves top-tier accuracy on two building change detection datasets of remote sensing imagery, while also achieving a real-time inference speed of 33.2 FPS on a mobile GPU. Full article

Leakage in water distribution systems (WDSs) causes a waste of water resources and increased carbon emissions. Rapid and accurate leakage localization to reduce the waste of water resources caused by leakages is an important way to overcome the problem. Using spatiotemporal correlation in monitoring data forms the basis of a leakage localization method proposed in a previous study. It is crucial to acknowledge that the chosen parameter settings significantly influence the localization performance of this method. This paper primarily seeks to optimize three essential parameters of this method: localization metrics weight (LMW), score threshold (ST), and the indicator of detection priority (IDP). LMW evaluates the similarity between simulated and measured pressure residuals. ST determines the size of the datasets involved in the spatial clustering, and IDP quantifies the likelihood of a true leakage within the candidate region. The leakage localization method is tested on a realistic full-scale distribution network to assess leakage flow rates and sensor noise. The results show that the optimized parameter settings could improve the efficiency and accuracy of leakage localization. Further, the findings indicate that the optimized parameter settings can enhance the effectiveness and precision of leakage localization. Full article

Over recent years, automated Human Activity Recognition (HAR) has been an area of concern for many researchers due to its widespread application in surveillance systems, healthcare environments, and many more. This has led researchers to develop coherent and robust systems that efficiently perform HAR. Although there have been many efficient systems developed to date, still, there are many issues to be addressed. There are several elements that contribute to the complexity of the task, making it more challenging to detect human activities, i.e., (i) poor lightning conditions; (ii) different viewing angles; (iii) intricate clothing styles; (iv) diverse activities with similar gestures; and (v) limited availability of large datasets. However, through effective feature extraction, we can develop resilient systems for higher accuracies. During feature extraction, we aim to extract unique key body points and full-body features that exhibit distinct attributes for each activity. Our proposed system introduces an innovative approach for the identification of human activity in outdoor and indoor settings by extracting effective spatio-temporal features, along with a Multi-Class Support Vector Machine, which enhances the model’s performance to accurately identify the activity classes. The experimental findings show that our model outperforms others in terms of classification, accuracy, and generalization, indicating its efficient analysis on benchmark datasets. Various performance metrics, including mean recognition accuracy, precision, F1 score, and recall assess the effectiveness of our model. The assessment findings show a remarkable recognition rate of around 88.61%, 87.33, 86.5%, and 81.25% on the BIT-Interaction dataset, UT-Interaction dataset, NTU RGB + D 120 dataset, and PKUMMD dataset, respectively. Full article

Video location prediction for handwritten digits presents unique challenges in computer vision due to the complex spatiotemporal dependencies and the need to maintain digit legibility across predicted frames, while existing deep learning-based video prediction models have shown promise, they often struggle with preserving local details and typically achieve clear predictions for only a limited number of frames. In this paper, we present a novel video location prediction model based on Convolutional Gated Recurrent Units (ConvGRU) that specifically addresses these challenges in the context of handwritten digit sequences. Our approach introduces three key innovations. Firstly, we introduce a specialized decoupling model using modified Generative Adversarial Networks (GANs) that effectively separates background and foreground information, significantly improving prediction accuracy. Secondly, we introduce an enhanced ConvGRU architecture that replaces traditional linear operations with convolutional operations in the gating mechanism, substantially reducing spatiotemporal information loss. Finally, we introduce an optimized parameter-tuning strategy that ensures continuous feature transmission while maintaining computational efficiency. Extensive experiments on both the MNIST dataset and custom mobile datasets demonstrate the effectiveness of our approach. Our model achieves a structural similarity index of 0.913 between predicted and actual sequences, surpassing current state-of-the-art methods by 1.2%. Furthermore, we demonstrate superior performance in long-term prediction stability, with consistent accuracy maintained across extended sequences. Notably, our model reduces training time by 9.5% compared to existing approaches while maintaining higher prediction accuracy. These results establish new benchmarks for handwritten digit video prediction and provide practical solutions for real-world applications in digital education, document processing, and real-time handwriting recognition systems. Full article

In the urbanization development trend, constructing inter-provincial metropolitan areas has gradually become an important pathway for China to implement its new urbanization strategy. Economic development in metropolitan areas inevitably leads to ecological degradation and fragmentation, threatening biodiversity. A reasonable assessment and optimization of the ecological network structure of inter-provincial metropolitan areas can effectively improve the stability of regional ecosystems and contribute to the sustainable development of urban and rural environments. In this study, we optimized the identification of ecological sources at the metropolitan area scale by constructing the ‘MIE’ model combined with the threshold analysis method, identified the core ecological sources in the Xuzhou Metropolitan Area, a typical cross-provincial metropolitan area in China, and then extracted the ecological corridors by using the Linkage Mapper tool based on this method. The spatiotemporal patterns and components of the ecological network of the Xuzhou metropolitan area from 1990 to 2020 were assessed, and an optimization analysis was performed based on the current ecological network. The results show that urban expansion has caused a decentralized layout of the ecological space in the study area across the provincial metropolitan area, with an uneven distribution of ecological sources and the development of ecological corridors in an east-dense and west-sparse pattern in 30 years. Increased interference by human activities has decreased the landscape dominance of the regional ecological resources and overall fragmentation of the habitats. The number of ecological corridors decreased, the average length of ecological corridors increased, the difficulty of biological flow increased, the connectivity of the ecological network structure decreased, and accessibility between source areas decreased. The α, β, and γ indices of the regional ecological network increased by 0.16, 0.46, and 0.1, respectively, after restoring important ecological sources and corridors. This study provides a framework for ecological network assessment and optimization in cross-provincial metropolitan areas, which is of great significance for studying metropolitan areas at similar stages of development. Full article