Search

Hyperspectral image (HSI) and light detection and ranging (LiDAR) data have gained significant attention due to their excellent complementarity, which can be combined to improve the accuracy of land cover classification. However, there are still many challenges, such as unrelated physical characteristics, different data structures, and a lack of labeled samples. Many methods fail to leverage the full potential of multi-source data, particularly hierarchical complementary information. To address these problems, a hierarchical coarse–fine adaptive (HCFA) fusion network with dynamic convolution and a transformer is proposed for multi-source remote sensing land cover classification. Although fusing hierarchical information can improve the classification accuracy of the model, improper hierarchical feature selection and optimization may negatively affect the classification results. Therefore, a coarse–fine mutual learning strategy is proposed to dynamically fuse hierarchical information. Additionally, the disparity between multi-source data continues to prevent the realization of effective fusion. To tackle this challenge, cross-tokenization and cross-token attention are implemented to enhance information interaction. Furthermore, to improve the model representation with limited computational cost, we combine the advantages of dynamic convolution with a transformer. Validation on three standard datasets demonstrates that HCFA achieves high accuracy with just 1% of the training set while maintaining low computational costs. [ABSTRACT FROM AUTHOR]

This paper addresses the challenge of integrating multiple energy sources into a single-domain microgrid, commonly found in urban buildings, while also providing a platform for energy management. A Lyapunov stability analysis of a simple boost converter was used as a basis for designing the dual control loop of the grid. The versatility of the developed control structure allows for the incorporation of an arbitrary number of sources hence achieving scalability. Next, the energy in the microgrid was separated into exogenous energy and actuator energy. This yielded a description of the system that quantified the condition of stability independent of the decision made by a would-be energy management system. This, in turns, liberates the process of designing an optimized energy management system from stability concerns. The acquired theoretical findings were then translated to a simulation model, where multiple components of the grid were simulated under a typical scenario of operation. Once the simulation phase was concluded, a prototype of the designed grid was constructed to emulate the theoretical results. The prototype exhibited promising performance, matching the simulation predictions to a reasonable degree. [ABSTRACT FROM AUTHOR]

Remote sensing images provide a valuable way to observe the Earth's surface and identify objects from a satellite or airborne perspective. Researchers can gain a more comprehensive understanding of the Earth's surface by using a variety of heterogeneous data sources, including multispectral, hyperspectral, radar, and multitemporal imagery. This abundance of different information over a specified area offers an opportunity to significantly improve change detection tasks by merging or fusing these sources. This review explores the application of deep learning for change detection in remote sensing imagery, encompassing both homogeneous and heterogeneous scenes. It delves into publicly available datasets specifically designed for this task, analyzes selected deep learning models employed for change detection, and explores current challenges and trends in the field, concluding with a look towards potential future developments. [ABSTRACT FROM AUTHOR]

Copyright of Machine Tool & Hydraulics is the property of Guangzhou Mechanical Engineering Research Institute (GMERI) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Multi-source water distribution systems (WDSs) are critical to solving the increasing demand for urban water supply. Appropriate management of limited resources necessitates optimization of water scheduling in order to reduce energy consumption. However, certain complexities of applying such systems bring severe challenges to optimal scheduling methods, exemplified in mountain regions, where larger elevation gradients make distribution more complicated than in plain regions. Therefore, this study attempts to present best practices in how to reduce the energy consumption of water supply, especially in complex mountainous regions, through innovation of optimal scheduling methods. Based on the seagull optimization algorithm (SOA), a systematic optimization scheduling method for multi-source WDSs is proposed. The optimization results are compared with those obtained from the genetic algorithm. A case study of such optimization in the mountainous region of C-County, China is presented. Power consumption prior and post optimization is compared. The results show that this optimization scheduling method is both effective and feasible. Annual power consumption can be reduced by significant amounts, savings of 23.3% in this case study, and the optimal solution can be deployed with 40 iteration steps. [ABSTRACT FROM AUTHOR]

Copyright of Geology & Exploration is the property of Geology & Exploration Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In this article, a solar and wireless energy harvesting (SWEH) Internet of Things network is presented. To address energy efficiency problems, we consider an amplify-and-forward relay node that harvests solar and wireless energy using the power splitting protocol to accomplish SWIPT. For the communication model under consideration, we derive the mathematical expressions for throughput and energy efficiency. We investigate the impact of both power transmitted by the source and power-splitting factor on the system's achievable energy efficiency and the energy harvested. To maximize the contradictory objectives of energy efficiency and energy harvested simultaneously, we propose to use the Pareto-based multi-objective ant lion optimization algorithm (MOALO) and jointly optimize the transmitted power and power-splitting factor. In-depth simulation is used to determine the Pareto optimal values for transmitted power and power-splitting factor and to assess how the proposed algorithm performs in comparison to leading-edge algorithms such as NSGA-II, MOPSO, MOWOA, and MOMVO. The results unambiguously show that the suggested MOALO offers the best Pareto front within a reasonable amount of time. Finally, the results corroborate that the proposed SWEH system outperforms in terms of harvested energy and attainable energy efficiency. [ABSTRACT FROM AUTHOR]

Alzheimer’s disease is the most common form of dementia and the early detection is essential to prevent its proliferation. Real data available has been of paramount importance in order to achieve progress in the automatic detection despite presenting two major challenges: Multi-source observations containing Magnetic resonance (MRI), Positron emission tomography (PET) and Cerebrospinal fluid data (CSF); and also missing values within all these sources. Most machine learning techniques perform this predictive task by using a single data modality. Nevertheless, the integration of all these sources of evidence could possibly bring a higher performance at different stages of disease progression. The Expectation Maximization (EM) algorithm has been successfully employed to handle missing values, but it is not designed for typical Machine Learning scenarios where an imputation model is created over training data and subsequently applied on a testing set. In this work, we propose EMreg-KNN, a novel supervised and multi-source imputation algorithm. Based on the EM algorithm, EMreg-KNN builds a regression ensemble model for the imputation of future data thus allowing the further utilization of any vector-based Machine Learning method to automatically assess the Alzheimer’s disease diagnosis. Using the ADNI database, the proposed method achieves significant improvements on F1, AUC and Accuracy measures over classical imputation methods for this database using four classification algorithms. Considering these classifiers in four different classification scenarios, our algorithm is experimentally superior in terms of the F measure, in nearly 82% of the cases under evaluation. [ABSTRACT FROM AUTHOR]

The frequency deviation from the nominal working frequency in power systems is a consequence of the imbalance between total electrical loads and the aggregate power supplied by production units. The sensitivity of energy system frequency to both minor and major load variations underscore the need for effective frequency load control mechanisms. In this paper, frequency load control in single-area power system with multi-source energy is analysed and simulated. Also, the effect of the photovoltaic system on the frequency deviation changes in the energy system is shown. In the single area energy system, the dynamics of thermal turbine with reheat, thermal turbine without reheat and hydro turbine are considered. The simulation results using Simulink/Matlab and model analysis using eigenvalue analysis show the dynamic behaviour of the power system in response to changes in the load. [ABSTRACT FROM AUTHOR]

The frequency deviation from the nominal working frequency in power systems is a consequence of the imbalance between total electrical loads and the aggregate power supplied by production units. The sensitivity of energy system frequency to both minor and major load variations underscore the need for effective frequency load control mechanisms. In this paper, frequency load control in single-area power system with multi-source energy is analysed and simulated. Also, the effect of the photovoltaic system on the frequency deviation changes in the energy system is shown. In the single area energy system, the dynamics of thermal turbine with reheat, thermal turbine without reheat and hydro turbine are considered. The simulation results using Simulink/Matlab and model analysis using eigenvalue analysis show the dynamic behaviour of the power system in response to changes in the load.

Exposure to ambient fine particulate matter (PM2.5) poses a significant global health challenge. However, a major obstacle for epidemiological studies and risk assessment lies in the absence of high-resolution spatiotemporal exposure estimates. Here, we present an integrated framework to achieve accurate estimations of population exposure to PM2.5 and assess related health risks by incorporating high temporal (hourly) and high spatial (1-km) PM2.5 concentrations and population density, using Beijing, China, in 2015 as an example. Firstly, hourly PM2.5 concentrations were estimated using a functional data model by integrating the gap-filled satellite-based aerosol optical depth data at 1-km resolution with hourly in-situ PM2.5 observations. Then, we calculated the population-level exposure to PM2.5 and associated health impacts incorporating both hourly PM2.5 and population information. We also investigated the bias of exposure and health impact assessment using coarse spatial or temporal PM2.5 and population data. Our findings revealed that exposure to PM2.5 resulted in 33,830, 21388, and 5,302 premature deaths in 2015 attributable to all-cause, cardiovascular, and respiratory diseases, respectively. A sensitivity analysis conducted underscored the critical importance of considering the high spatiotemporal heterogeneity of both PM2.5 concentrations and population density when investigating population-level PM2.5 exposure and related health impacts. Insights on more accurate PM2.5 exposure assessment from this study can help policymakers better assess improvements after clean air actions and analyze exposure inequality, and thereafter develop more effective pollution mitigation strategies.

ABSTRACTThe foundational step in constructing a digital twin involves transforming the physical world into an immersive 3D virtual environment. Unmanned aerial vehicle (UAV)-based photogrammetry, mobile mapping system (MMS), and terrestrial laser scanning (TLS) serve as powerful tools for 3D reconstruction of urban environments. Given the distinctive strengths and weaknesses of each sensor, a multi-source point cloud registration technique is imperative to achieve a detailed and precisely geolocated 3D urban model. Hence, we developed coarse-to-fine approach methods, called building exterior wall-based (BEWB) and building outline-based (BOB) algorithms, to register point clouds captured by diverse sensors in urban scenes. The BEWB algorithm coarsely registers urban point clouds by extracting building exterior walls, establishing correspondence points, and effectively removing outliers within this correspondence point set. The BOB algorithm precisely registers the urban point clouds acquired from multiple sensors by leveraging building outlines and points corresponding to the ground of the point clouds. To validate the proposed algorithms, we initially registered a UAV-based photogrammetry point cloud with the MMS point cloud and subsequently registered the TLS point clouds. Comprehensive quantitative and qualitative analyzes of the results demonstrate that our algorithms outperform existing methods in achieving precise registration. By successfully registering three distinct point clouds, we generated a comprehensive urban scene point cloud characterized by enhanced precision, point density, and geolocation accuracy.

Unsupervised domain adaptation (UDA) methods have made remarkable progress in histopathological image analysis and various cancer diagnosis domains. However, most current research focuses on transfer between single-source domains. The distribution of features between different cancer types is far away, and a well-trained model in one field may not be able to generalize well to data in other fields. To address the domain shift problem, this paper proposes a multi-source unsupervised domain adaptation method with domain mixing bridging. Using multiple source and target domains, feature representations of all domains are extracted, and latent relationships are captured. Afterward, the complementary information of the hybrid bridging intermediate domain is integrated to align the feature distribution. Additionally, we introduce a domain adversarial adaptation module to generate domain invariant features. We experimented on three different cancer pathology image datasets and achieved an average accuracy of 92.94 % classification performance. It is proved that compared with the existing deep transfer learning technology, the method in this paper has a better effect. Code will be available at: https://github.com/Ww994/MHDAN. [ABSTRACT FROM AUTHOR]

Due to technical progress in sensor technology and computer power, observation and model data are operationally available today with a high spatial and temporal resolution, suitable for thunderstorm detection and prediction. However, nowcasting the remaining lifetime of an observed thunderstorm is still a challenge to date. To improve nowcasting of deep convective cells, we developed the algorithm LOC-lifetime that predicts the remaining lifetime of thunderstorms based on life-cycle signatures present in satellite, radar, lightning and numerical weather prediction model data. We use the mathematical method “fuzzy logic” to combine this multi-source input and to categorize the thunderstorm evolution into the life-cycle sets growth and decay. We analyzed a data set of almost 1,800 thunderstorms that occurred during the summer months June 2016, May, June, July 2017 and June 2018. The data reveal highly variable life cycles which make it difficult to predict the remaining lifetime on basis of life-cycle statistics. Nevertheless, LOC-lifetime offers an improved nowcasting quality compared to a simpler nowcasting method as it increases the probability of correct prediction and reduces the root-mean-square error. Therefore, we propose the lifetime prediction via LOC-lifetime as a useful tool in combination with other existing algorithms to nowcast and forecast thunderstorms.

This paper addresses the challenge of integrating multiple energy sources into a single-domain microgrid, commonly found in urban buildings, while also providing a platform for energy management. A Lyapunov stability analysis of a simple boost converter was used as a basis for designing the dual control loop of the grid. The versatility of the developed control structure allows for the incorporation of an arbitrary number of sources hence achieving scalability. Next, the energy in the microgrid was separated into exogenous energy and actuator energy. This yielded a description of the system that quantified the condition of stability independent of the decision made by a would-be energy management system. This, in turns, liberates the process of designing an optimized energy management system from stability concerns. The acquired theoretical findings were then translated to a simulation model, where multiple components of the grid were simulated under a typical scenario of operation. Once the simulation phase was concluded, a prototype of the designed grid was constructed to emulate the theoretical results. The prototype exhibited promising performance, matching the simulation predictions to a reasonable degree.

Hyperspectral image (HSI) and light detection and ranging (LiDAR) data have gained significant attention due to their excellent complementarity, which can be combined to improve the accuracy of land cover classification. However, there are still many challenges, such as unrelated physical characteristics, different data structures, and a lack of labeled samples. Many methods fail to leverage the full potential of multi-source data, particularly hierarchical complementary information. To address these problems, a hierarchical coarse–fine adaptive (HCFA) fusion network with dynamic convolution and a transformer is proposed for multi-source remote sensing land cover classification. Although fusing hierarchical information can improve the classification accuracy of the model, improper hierarchical feature selection and optimization may negatively affect the classification results. Therefore, a coarse–fine mutual learning strategy is proposed to dynamically fuse hierarchical information. Additionally, the disparity between multi-source data continues to prevent the realization of effective fusion. To tackle this challenge, cross-tokenization and cross-token attention are implemented to enhance information interaction. Furthermore, to improve the model representation with limited computational cost, we combine the advantages of dynamic convolution with a transformer. Validation on three standard datasets demonstrates that HCFA achieves high accuracy with just 1% of the training set while maintaining low computational costs.

Remote sensing images provide a valuable way to observe the Earth’s surface and identify objects from a satellite or airborne perspective. Researchers can gain a more comprehensive understanding of the Earth’s surface by using a variety of heterogeneous data sources, including multispectral, hyperspectral, radar, and multitemporal imagery. This abundance of different information over a specified area offers an opportunity to significantly improve change detection tasks by merging or fusing these sources. This review explores the application of deep learning for change detection in remote sensing imagery, encompassing both homogeneous and heterogeneous scenes. It delves into publicly available datasets specifically designed for this task, analyzes selected deep learning models employed for change detection, and explores current challenges and trends in the field, concluding with a look towards potential future developments.

Wildfires play a crucial role in the transformation of forest ecosystems and exert a significant influence on the global climate over geological timescales. Recent shifts in climate patterns and intensified human–forest interactions have led to an increase in the incidence of wildfires. These fires are characterized by their extensive coverage, higher frequency, and prolonged duration, rendering them increasingly destructive. To mitigate the impact of wildfires on climate change, ecosystems, and biodiversity, it is imperative to conduct systematic monitoring of wildfire progression and evaluate their environmental repercussions on a global scale. Satellite remote sensing is a powerful tool, offering precise and timely data on terrestrial changes, and has been extensively utilized for wildfire identification, tracking, and impact assessment at both local and regional levels. The Canada Centre for Mapping and Earth Observation, in collaboration with the Canadian Forest Service, has developed a comprehensive National Burned Area Composite (NBAC). This composite serves as a benchmark for curating a bi-temporal multi-source satellite image dataset for change detection, compiled from the archives of Sentinel-2, Sentinel-1, and ALOS-2 PALSAR-2. To our knowledge, this dataset is the inaugural large-scale, multi-source, and multi-frequency satellite image dataset with 20 m spatial resolution for wildfire mapping, monitoring, and evaluation. It harbors significant potential for enhancing wildfire management strategies, building upon the profound advancements in deep learning that have contributed to the field of remote sensing. Based on our curated dataset, which encompasses major wildfire events in Canada, we conducted a systematic evaluation of the capability of multi-source satellite earth observation data in identifying wildfire-burned areas using statistical analysis and deep learning. Our analysis compares the difference between burned and unburned areas using post-event observation solely or bi-temporal (pre- and post-event) observations across diverse land cover types. We demonstrate that optical satellite data yield higher separability than C-Band and L-Band Synthetic Aperture Radar (SAR), which exhibit considerable overlap in burned and unburned sample distribution, as evidenced by SAR-based boxplots. With U-Net, we further explore how different input channels influence the detection accuracy. Our findings reveal that deep neural networks enhance SAR's performance in mapping burned areas. Notably, C-Band SAR shows a higher dependency on pre-event data than L-Band SAR for effective detection. A comparative analysis of U-Net and its variants indicates that U-Net works best with single-sensor data, while the late fusion architecture marginally surpasses others in the fusion of optical and SAR data. Accuracy across sensors is highest in closed forests, with sequentially lower performance in open forests, shrubs, and grasslands. Future work will extend the data from both spatial and temporal dimensions to encompass varied vegetation types and climate zones, furthering our understanding of multi-source and multi-frequency satellite remote sensing capabilities in wildfire detection and monitoring. [ABSTRACT FROM AUTHOR]

Analyzing the factors influencing traffic congestion is essential for urban planning and coordinated development. Previous research frequently focuses on the internal aspects of traffic systems, often overlooking the impact of external factors on congestion sources. Therefore, this study utilizes a geospatial dataset and mobile signaling data, firstly applying the Fuzzy C-Means (FCM) algorithm to identify congested roads of different levels and trace the localization of travelers' origins on regional congested roads. Furthermore, it employs the LightGBM method to study the influence of the built environment of various congestion sources on network-level congestion. The findings are as follows: (1) There is a positive correlation between traffic congestion and geographical location, with congestion predominantly caused by a few specific plots and demonstrating a concentrated trend in city centers. (2) Residential population density is the most critical factor, accounting for over 12% of the congestion contribution, followed by road density and working population density. (3) Both residential and working population densities show a non-linear positive correlation with congestion contribution, while the mixture of land use displays a non-linear V-shaped influence. Additionally, when residential population density is between 8000 and 11,000, it notably exacerbates congestion contribution. Significantly, by emphasizing land use considerations in traffic system analysis, these findings illuminate the intricate linkages between urban planning and traffic congestion, advocating for a more comprehensive approach to urban development strategies. [ABSTRACT FROM AUTHOR]

The fault diagnosis method based on domain adaptation is a hot topic in recent years. It is difficult to collect a complete data set containing all fault categories in practice under the same working condition, leading to fault categories knowledge loss in the single source domain. To resolve the problem, a domain adaptation method for bearing fault diagnosis using multiple incomplete source data is proposed in this study. First, the cycle generative adversarial network is used to learn the mapping between multi-source domains to complement the missing category data. Then, considering the domain mismatch problem, a multi-source domain adaption model based on anchor adapters is developed to obtain general domain invariant diagnosis knowledge. Finally, the fault diagnosis model is established by an ensemble of multi-classifier results. Extensive experiments on bearing data sets demonstrate that the proposed method in fault diagnosis with multiple incomplete source data is effective and has a good diagnosis performance. [ABSTRACT FROM AUTHOR]

Sensing urban vitality is a useful method for understanding urban development. However, the spatio-temporal characteristics of the association between the built environment and urban vitality in Shenzhen, the youngest mega-city in China, have not yet been explored. In this paper, we examined the effects of built environment indicators on urban vitality by using spatial regression models and multi-source geospatial data. The main research findings were as follows. Firstly, urban vitality displayed a consistent high–low pattern during both weekdays and weekends. Differences in the distribution of urban vitality with time between weekdays and weekends were more significant. Secondly, the effects of various built environment indicators on urban vitality exhibited significant temporal disparities. Within a day, population density, building density, bus station density, and distance to metro stations all exhibited positive effects, while distance to the central business district (CBD) exhibited negative effects, with pronounced diurnal differences. Moreover, the effects of road network density and functional mix on urban vitality were both positive and negative throughout the day. Thirdly, besides population density and building density, road network density, functional mix, bus stop density, and distance from metro stations exhibited positive and negative disparities within the study area. Overall, distance to the CBD had a negative effect on urban vitality. This concludes that planning for urban vitality should consider the spatio-temporal heterogeneity of the association between the built environment and urban vitality. [ABSTRACT FROM AUTHOR]

In order to solve the problem that it is difficult for a single sensor to accurately characterize the running state of rotating bearings under complex working conditions, this paper proposes a data-level fusion method based on multi-source isomorphic sensors to monitor spindle bearings. First, new vibration signals in the X,Y,Z direction were obtained through the process of decomposing, de-noise, and reconstructing. Second, the PCA algorithm was used to select the time-domain and frequency-domain features of the vibration signals, construct the feature matrix, and perform dimensionality reduction in the feature matrix. Finally, the entropy weight method was introduced to obtain the initial weights of the three directions as the inputs of the adaptive function. The chaotic particle swarm optimization algorithm proposed in this paper helps particles jump out of the local optimum. Chaotic mapping is used to initialize the velocity and position of the particles, which calculates globally optimal weights in three directions. In order to extract bearing signal features more accurately and efficiently, a DenseNet and Transformer (DAT) feature extraction model is proposed to deal with the complex changes and noise interference of bearing signals. Through the open data set of Jiangnan University and the data collected by our own experimental platform, the maximum accuracy of the DAT model was verified to be 100%. [ABSTRACT FROM AUTHOR]

The accuracy of ionospheric models estimated by ground‐based multiple global navigation satellite system ionospheric data over regions with sparse tracking stations is not ideal. To improve the accuracy of the estimated ionospheric model, different types of ionospheric data with different combinations were employed for previous studies. However, the ionospheric observational ranges for different types of ionospheric data are not the same. In this study, the accuracy of ionospheric maps generated by ground‐based ionospheric data (ground‐based strategy) and ground‐based ionospheric data combined with data provided by other geodetic measurements normalized by the single‐layer normalization method (multi‐source strategy) were studied. The results showed that the main differences between the ionospheric models estimated by the two strategies occur for data taken over the ocean, which mainly range from −1 to 0 total electron content unit (TECU). When assessed using Jason‐3 vertical total electron content data, the mean root mean square (RMS) value of the ionospheric model estimated by the multi‐source strategy was 5.03 TECU, which is approximately 15% smaller than that estimated by the ground‐based strategy. The maximum reduction in results using the multisource strategy was approximately 25% over different latitudes compared with that of the ground‐based strategy. Furthermore, the self‐consistency evaluation method was employed for evaluation. The results showed that the RMS of the ionospheric model estimated by the multi‐source strategy was 2.41 TECU, which is 3.60% better than that of the ground‐based strategy. The maximum reduction was 15% on different days. Key Points: The single‐layer normalization method is employed to make multi‐source data have the same observational range compared with those of ground‐based global navigation satellite system dataThe distribution of multi‐source data has better coverage, especially over the oceanic regionThe main differences between the ionospheric models estimated by the ground‐based strategy and the multi‐source strategy occur over the ocean [ABSTRACT FROM AUTHOR]

Due to the scarcity of observational data and the intricate precipitation–runoff relationship, individually applying physically based hydrological models and machine learning (ML) techniques presents challenges in accurately predicting floods within data-scarce glacial river basins. To address this challenge, this study introduces an innovative hybrid model that synergistically harnesses the strengths of multi-source remote sensing data, a physically based hydrological model (i.e., Spatial Processes in Hydrology (SPHY)), and ML techniques. This novel approach employs MODIS snow cover data and remote sensing-derived glacier mass balance data to calibrate the SPHY model. The SPHY model primarily generates baseflow, rain runoff, snowmelt runoff, and glacier melt runoff. These outputs are then utilized as extra inputs for the ML models, which consist of Random Forest (RF), Gradient Boosting (GDBT), Long Short-Term Memory (LSTM), Deep Neural Network (DNN), Support Vector Machine (SVM) and Transformer (TF). These ML models reconstruct the intricate relationship between inputs and streamflow. The performance of these six hybrid models and SPHY model is comprehensively explored in the Manas River basin in Central Asia. The findings underscore that the SPHY-RF model performs better in simulating and predicting daily streamflow and flood events than the SPHY model and the other five hybrid models. Compared to the SPHY model, SPHY-RF significantly reduces RMSE (55.6%) and PBIAS (62.5%) for streamflow, as well as reduces RMSE (65.8%) and PBIAS (73.51%) for floods. By utilizing bootstrap sampling, the 95% uncertainty interval for SPHY-RF is established, effectively covering 87.65% of flood events. Significantly, the SPHY-RF model substantially improves the simulation of streamflow and flood events that the SPHY model struggles to capture, indicating its potential to enhance the accuracy of flood prediction within data-scarce glacial river basins. This study offers a framework for robust flood simulation and forecasting within glacial river basins, offering opportunities to explore extreme hydrological events in a warming climate. [ABSTRACT FROM AUTHOR]

Internet of Things and growing popularity of connected devices are increasing the demand for wearables applications, which is driving the development of ultra-low power secure systems. Thus, having a system running at low voltage with low power consumption introduces the use of ambient energy harvesting. Besides, batteries have several limitations reducing their compatibility with wearables, including periodical need for recharge, limited life-time, heavy and large form factor and non-eco-friendly life-cycle. Therefore ambient energy harvesting technology gives a promising alternative for self-powered wearables. The integration of these solutions on the same silicon support is a real technological challenge. This paper proposes a multi-harvesting fully-integrated system architecture combining AC and DC sources. Based on a MATLAB/Simulink model, the proposed system allows the calibration of an optimal solution. It is designed and simulated in 40 nm CMOS process, demonstrating the use of an AC source as secondary input to start a harvesting system with very low DC source as main source. • Ambient energy harvesting technology gives a promising alternative for self-powered wearables. • Multi-harvesting fully-integrated system architecture combining AC and DC sources. • Energy harvesting system level modeling using MATLAB/Simulink. [ABSTRACT FROM AUTHOR]

Based on multi-source of radiation observations from CMA operational， Qinghai-Xizang Plateau Atmospheric Scientific Experiment and Chinese Academy of Sciences field stations， the integrated of multi-source radiation data over Qinghai-Xizang Plateau（IRDQTP）since 1993 has been constructed through quality control， splicing and integration.The percent of each radiation element actual records is more than 99.4%.The accuracy rate of photosynthetic effective radiation is less than 76%， and other elements are more than 96.7%.The correlation coefficient of downward shortwave radiation between atmospheric scientific experiment and meteorological operational observations at the same or similar sites exceeds 0.88.The Mbias （RMSE） of Ali （the same station site） between operational and experimental observations is -0.006 （0.141） MJ·m-2·h-1， with 99.8% distributes between -1.0 and 1.0 MJ·m-2·h-1， and Naqu （similar station site） is -0.028 （0.615） MJ·m-2·h-1， with 84.6% distributes between -1.0 and 1.0 MJ·m-2·h-1.The difference between the operational and experimental observations shows obvious seasonal and diurnal variations.It is little in autumn and winter while large in spring and summer.Affected by the difference of radiometer sensitivity and environmental response parameters， the difference gradually increases with the rapid increase （rapid decrease） of short-wave radiance in the sunrise period （sunset period） while decreases gradually with the increase of short-wave radiance during other times.It reaches the maximum when the change rate of short-wave radiation reaches the maximum.There are daily variations in downward shortwave radiation and net radiation with a single peak distribution.The downward shortwave radiation reaches the maximum at 12:00 （local average solar time， similarly hereinafter） in spring and autumn which is 2.71 and 2.21 MJ·m-2·h-1 respectively while at 13:00 in summer and winter which is 2.58 and 1.89 MJ·m-2·h-1 respectively.The net radiation reaches the maximum at 12:00 in spring and autumn which is 1.37 and 1.15 MJ·m-2·h-1 respectively while at 13:00 in summer and winter which is 1.53 and 0.80 MJ·m-2·h-1 respectively.Affected by climatic conditions and topographic factors， there are obvious spatial variations for downward shortwave radiation and net radiation distribution.Taking 85°E and 95°E as the dividing line， the downward shortwave radiation shows a gradually decreasing distribution pattern from west to east.It is 594.44， 582.56 and 461.95 MJ·m-2·mon-1 in western， central and eastern region respectively.The net radiation presents a low-high-low distribution pattern from west to east.It is 129.68， 207.13 and 173.56 MJ·m-2·mon-1 in western， central and eastern region respectively.

The occurrence of road traffic accidents has increased year by year.Driver inattention during driving is one of the major causes of traffic accidents.In this paper,we utilize multi-source data to detect driver distraction.However,the correlations derived from multi-source data will generate feature of high-dimensional entanglement.Existing methods perform similar processing for data of different sources or simply stick to concatenate multi-source features,which are not easy to catch the key feature of high-dimensional entanglement.And distracted driving can be affected by many factors.Supervised methods might cause misclassification when the type of driver distraction does not exist in the set of the known categories.Therefore,we propose a multi-dcale feature fusion network approach to tackle these challenges.Basically,it first learns low-dimensional representations from multi-source data through multiple embedding subnetworks,and then proposes a multi-scale feature Fusion method to aggregate these representations from the perspective of spatial-temporal correlation,thereby reducing the entanglement of feature.Finally,we utilize a ConvLSTM encoder-decoder model to detect driver distraction.Experimental results on a public loaded drive dataset show that the proposed method outperforms the existing methods.

Crop type mapping at high resolution is crucial for various purposes related to agriculture and food security, including the monitoring of crop yields, evaluating the potential effects of natural disasters on agricultural production, analyzing the potential impacts of climate change on agriculture, etc. However, accurately mapping crop types and ranges on large spatial scales remains a challenge. For the accurate mapping of crop types at the regional scale, this paper proposed a crop type mapping method based on the combination of multiple single-temporal feature images and time-series feature images derived from Sentinel-1 (SAR) and Sentinel-2 (optical) satellite imagery on the Google Earth Engine (GEE) platform. Firstly, crop type classification was performed separately using multiple single-temporal feature images and the time-series feature image. Secondly, with the help of information entropy, this study proposed a pixel-scale crop type classification accuracy evaluation metric, i.e., the CA-score, which was used to conduct a vote on the classification results of multiple single-temporal images and the time-series feature image to obtain the final crop type map. A comparative analysis showed that the proposed classification method had excellent performance and that it can achieve accurate mapping of multiple crop types at a 10 m resolution for large spatial scales. The overall accuracy (OA) and the kappa coefficient (KC) were 84.15% and 0.80, respectively. Compared with the classification results that were based on the time-series feature image, the OA was improved by 3.37%, and the KC was improved by 0.03. In addition, the CA-score proposed in this study can effectively reflect the accuracy of crop identification and can serve as a pixel-scale classification accuracy evaluation metric, providing a more comprehensive visual interpretation of the classification accuracy. The proposed method and metrics have the potential to be applied to the mapping of larger study areas with more complex land cover types using remote sensing. [ABSTRACT FROM AUTHOR]

Quantitative hazard analysis of rockfalls is a fundamental tool for sustainable risk management, even more so in places where the preservation of natural heritage and people's safety must find the right balance. The first step consists in determining the magnitude-frequency relationship, which corresponds to the apparently simple question: how big and how often will a rockfall be detached from anywhere in the cliff? However, there is usually only scarce data on past activity from which to derive a quantitative answer. Methods are proposed to optimize the exploitation of multi-source inventories, introducing sampling extent as a main attribute for the analysis. This work explores the maximum possible synergy between data sources as different as traditional inventories of observed events and current remote sensing techniques. Both information sources may converge, providing complementary results in the magnitude-frequency relationship, taking advantage of each strength that overcomes the correspondent weakness. Results allow characterizing rockfall detachment hazardous conditions and reveal many of the underlying conditioning factors, which are analyzed in this paper. High variability of the hazard over time and space has been found, with strong dependencies on influential external factors. Therefore, it will be necessary to give the appropriate reading to the magnitude-frequency scenarios, depending on the application of risk management tools (e.g., hazard zoning, quantitative risk analysis, or actions that bring us closer to its forecast). In this sense, some criteria and proxies for hazard assessment are proposed in the paper. [ABSTRACT FROM AUTHOR]

Fine-scale population map plays an essential role in numerous fields, including resource allocation, urban planning, disaster prevention and response. Point of Interest (POI) data is widely used for population spatialization, but the types of POI are ignored. Since different types of POI data have different impacts on population distribution, this paper used typed POI data and other multi-source data to map population distributions at fine scales. At the township level, three random forest models were used to generate the population maps of 150 m, 300 m, and 500 m in 2020, enabling the downscaling of county-level population distribution to the grid level. The main influencing factors of population distribution were extracted and analyzed based on the feature importance output from the model. Zhengzhou city was used as a case for study. The experiments show the results of population spatialization for all three scales in this study have better fitting accuracy than that of the GPWv4 and LandScan datasets. The coefficient of determination (R2) is 0.8333 for 150 m gridded population, 0.8295 for 300 m, and 0.8224 for 500 m; POI types related to residence information have greater contributions to population spatialization than other features; typed POI data are more conducive to population spatialization.

IntroductionUnsupervised domain adaptation (UDA) aims to adapt a model learned from the source domain to the target domain. Thus, the model can obtain transferable knowledge even in target domain that does not have ground truth in this way. In medical image segmentation scenarios, there exist diverse data distributions caused by intensity in homogeneities and shape variabilities. But multi source data may not be freely accessible, especially medical images with patient identity information.MethodsTo tackle this issue, we propose a new multi-source and source-free (MSSF) application scenario and a novel domain adaptation framework where in the training stage, we only get access to the well-trained source domain segmentation models without source data. First, we propose a new dual consistency constraint which uses domain-intra and domain-inter consistency to filter those predictions agreed by each individual domain expert and all domain experts. It can serve as a high-quality pseudo label generation method and produce correct supervised signals for target domain supervised learning. Next, we design a progressive entropy loss minimization method to minimize the class-inter distance of features, which is beneficial to enhance domain-intra and domain-inter consistency in turn.ResultsExtensive experiments are performed for retinal vessel segmentation under MSSF condition and our approach produces impressive performance. The sensitivity metric of our approach is highest and it surpasses other methods with a large margin.DiscussionIt is the first attempt to conduct researches on the retinal vessel segmentation task under multi-source and source-free scenarios. In medical applications, such adaptation method can avoid the privacy issue. Furthermore, how to balance the high sensitivity and high accuracy need to be further considered.

Analyzing the factors influencing traffic congestion is essential for urban planning and coordinated development. Previous research frequently focuses on the internal aspects of traffic systems, often overlooking the impact of external factors on congestion sources. Therefore, this study utilizes a geospatial dataset and mobile signaling data, firstly applying the Fuzzy C-Means (FCM) algorithm to identify congested roads of different levels and trace the localization of travelers’ origins on regional congested roads. Furthermore, it employs the LightGBM method to study the influence of the built environment of various congestion sources on network-level congestion. The findings are as follows: (1) There is a positive correlation between traffic congestion and geographical location, with congestion predominantly caused by a few specific plots and demonstrating a concentrated trend in city centers. (2) Residential population density is the most critical factor, accounting for over 12% of the congestion contribution, followed by road density and working population density. (3) Both residential and working population densities show a non-linear positive correlation with congestion contribution, while the mixture of land use displays a non-linear V-shaped influence. Additionally, when residential population density is between 8000 and 11,000, it notably exacerbates congestion contribution. Significantly, by emphasizing land use considerations in traffic system analysis, these findings illuminate the intricate linkages between urban planning and traffic congestion, advocating for a more comprehensive approach to urban development strategies.

• A novel method based on optimal transport achieves fine-grained alignment. • Class-aware sampling meets the necessary conditions to avoid the negative transfer. • Sample-reweighting module continuously optimize the transport plan. • Extensive experiments on several benchmarks show the advantages of the proposed method. Multi-Source Domain Adaptation (MSDA) techniques have attracted widespread attention due to their availability to transfer knowledge from multiple source domains to the unlabeled target domain. Optimal transport (OT) has recently been utilized to measure the distance between distributions in virtue of its robustness. This paper proposes a novel OT-based Class-Aware Sample Reweighting (CASR) method to achieve sample-level fine-grained alignment between multi-source and target. Technically, the class-aware sampling strategy ensures class-level conditional alignment during transport by explicitly selecting samples from each domain. Besides, the sample-reweighting module is designed to allocate specific mass to each transmitted sample, which considers the classification reliability and the spatial information correlation to obtain the alignment priority between target and multi-source and further optimize the transport plan. Extensive experiments conducted on several benchmarks show that CASR presents significant advantages compared with other MSDA methods, and the visualization analysis further demonstrates the effectiveness of each proposed module. [ABSTRACT FROM AUTHOR]

Copyright of Plateau Meteorology is the property of Plateau Meteorology Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)