Remote Sensing Image Data Research Articles

Processing Model of LUCC Temporal Characteristics in Urban and Rural Space Based on Remote Sensing Data Validation

The spatial land change of villages and towns can be analyzed through external spatio-temporal characteristics, which can realize the evaluation of historical development and assist the implementation of planning and design norms and the implementation of problem rectification. To explore the specific process and the overall pattern of land use change, it is necessary to analyze and predict all kinds of data in a unified way to grasp the law of development and the direction of future development. This paper takes the spatial development and planning of villages and towns as the research object, selects the remote sensing image data of regional historical development as the driving factor, combines with Gaussian function fitting to calculate the training samples, and constructs the regional spatial environment analysis system. According to the processing model of external characteristics of time series, the spatial distribution and evolution of Risk-Screening Environmental Indicators (RSEI) and its components are analyzed. The multi-dimensional remote sensing data is innovatively fused into the time series processing model. Besides, the curve which can characterize the spatial change law and is not completely symmetrical is obtained through the least square method fitting so that the data with unequal time series intervals can be flexibly processed. Finally, through the experimental analysis, the traditional Asymmetric Gaussian model (A-G) and Savitzky-Golay (S-G) are taken to compare and analyze the effect of various factors on the processing of time series characteristics. Tests show that the proposed algorithm is superior to the comparison method in noise removal of remote sensing data. The data fitting coefficient is 0.02, and the root mean square measurement error is more than 7% higher than the comparison method. It can predict the trend of land use change distribution, play a key role in assisting decision-making for different land use transformations and utilization in villages and towns, and make a scientific evaluation for the impact of social and economic development potential.

Building Damage Degree Recognition Based on Temporal Attention Features

Natural disasters pose significant harm to society. As an important place for social activities and economic development, the degree of damage to building areas is directly related to disaster loss assessment and emergency rescue. Remote sensing image data, characterized by its wide coverage and multi-temporal features, provides important data support for post-disaster loss assessment. However, imaging differences caused by factors such as shooting time, imaging angle, and different sensors can interfere with the extraction of damage features and loss assessment. This paper proposes a Dual-Exchange-Attention U-Net (DERU-Net) model, which transforms the identification of building damage levels into intra-class semantic change detection. The DFMA feature attention fusion module is introduced to enhance the ability of dual-temporal feature extraction and achieve end-to-end assessment of building damage. The proposed method is comprehensively evaluated and tested on the xBD dataset. Experimental results show that compared with other methods, the DERU-Net proposed in this paper exhibits better stability and evaluation accuracy in assessing the degree of building damage.

Spatial utilization of historical topographic map and its application in land reconstruction of ancient Chinese urban land use

The historical topographic map preserves rich geographic information and can provide direct assistance for the reconstruction of various geographic elements. Based on the historical data of cities throughout the Qing Dynasty, the land use scale data of cities across the country was obtained using GIS and urban perimeter conversion models. This study combines city information and city circumference records from the historical maps and archives of the late Qing Dynasty to quantitatively reconstruct the use patterns of ancient China’s urban land at a spatial resolution of 1° × 1°. Uncertainty analysis of the reconstruction results was conducted using modern remote sensing image data as the validation data set. The results showed the following. (1) During the late Qing Dynasty, the total area of urban land in the various provinces and regions was 1456.015 km2. The maximum value was 208.691 km2 in Beijing-Tianjin-Hebei, the minimum value was 1.713 km2 in Qinghai, and the average value was 56.001 km2. (2) The results of grid reconstruction show that among the 398 grids with urban land distribution, the maximum value is 64.099 km2/grid, the minimum value is 0.013 km2/grid, and the average value is 3.658 km2/grid. (3) Of all the grids with urban land, the urban land grid to the west of the Hu Line accounts for 12.5% and the east to 87.5%. (4) During the late Qing Dynasty, urban land use in China was primarily concentrated in agriculturally developed areas such as the North China Plain, the Central Plains, Jiangnan, and the Sichuan-Chongqing region. (6) The results of a kernel density estimation showed that there were obviously three core areas of urban land agglomeration in China during the late Qing Dynasty: the North China Plain-Central Plains, the Jiangsu-Shanghai-Zhejiang-Anhui area, and the Sichuan-Chongqing urban core area. This study provides basic data for urban land use during historical periods and provides a basis for the quantitative reconstruction of relevant urban land data for historical archives.

The relationship between landscape pattern and plant species diversity in east Dongting lake wetland based on different Eco-environment

Exploring the impact factors associated with biodiversity and the relationship between them has always been a concerned issue in recent years. However, the previous research mostly focus on theoretical layer. Accordingly, the relationship between landscape pattern and biodiversity is to be analyzed in this research. The landscape pattern determines the function and ecological process of the landscape, and affects the species flow, information flow and energy flow in the landscape. Land use patterns has inevitably left an impact on the landscape pattern. Landscape pattern determines the function and ecological process of landscape and thus plays a significant role in biodiversity. East Dongting Lake National Nature Reserve is taken as the research object of the paper, and the remote sensing image data of three different time periods are collected, including 2000, 2010 and 2020. With an interpretation of the vegetation landscape pattern changes inside the protected area to collect and analyze the vegetation coverage. By comparing landscape patterns and the dynamic changes of land use in different periods of time, the correlation between landscape pattern characteristics and regional biodiversity is to be analyzed. Research shows: (1) From 2000 to 2020, the vegetation coverage of East Dongting Lake increased, but the landscape shape, scale, diversity and uniformity index decreased to varying degrees. (2) At the class level of landscape type, the relationship between landscape index and biodiversity is different. A complex relationship between farmland landscape and biodiversity. There is a significant positive correlation between the index of grassland landscape type and the index of regional biodiversity. (3) The correlation analysis results at the landscape level show that the landscape characteristic index is positively correlated with the regional biodiversity index. The grassland landscape in the area is the main habitat of biological species. At the same time, as the main grain producing area, the impact of farmland landscape cannot be ignored. This study has certain theoretical guiding significance for the protection and management of biodiversity in the region in terms of maintaining landscape pattern in particular the grassland landscape area and increasing vegetation coverage in the process of land use.

Temporal and Spatial Response of Ecological Environmental Quality to Land Use Transfer in Nanling Mountain Region, China Based on RSEI: A Case Study of Longnan City

Nanling Mountain region is a typical southern hilly region, which plays an important ecological and environmental protection role in China’s overall land protection pattern. Based on the remote sensing image data of Longnan City in Nanling Mountain region in 2013, 2018 and 2023, this paper interpreted the land use type and analyzed the land use transfer situation by using land use transfer flow, and a land use transfer matrix. At the same time, based on the remote sensing ecological index (RSEI) model, the ecological environmental quality of Longnan City from 2013 to 2023 was retrieved. The temporal and spatial response model of the ecological environmental quality to land use transfer in Longnan City from 2013 to 2023 was discussed based on spatial autocorrelation and a geographical detector. The results show that from 2013 to 2023, the decrease of forest land (16.23 km2) and the increase of construction land (13.25 km2) were the main land use transfers in Longnan City. The ecological environment indexes of Longnan City in 2013, 2018 and 2023 were 0.789, 0.917 and 0.872, respectively, showing a trend of “first rising and then decreasing”. The ecological environmental quality in the north of Longnan City was significantly lower than that in the south, and the poor ecological quality area appeared in and around the northern main urban area, showing a trend of “inward contraction”. Forest land, garden land, grassland, cultivated land and water area have a positive impact on ecological environmental quality, while traffic land, construction land and other land have a negative impact on ecological environmental quality. The response of ecological environmental quality to different land use transfer modes is related to the change of the overall ecological environmental quality. The interaction between land use and land cover change (LUCC) and other factors had a great impact on the evolution of ecological environmental quality in Longnan City.

Research on the Joint Construction of a National Multi-source and Multi-resolution image Checkpoint Database

Abstract. In the process of quality inspection of Remote sensing image data results, the reuse of spatial location information of multiple units, multiple projects and multiple sources can not only overcome the problems of long time to obtain control information, high cost and difficulty in obtaining some areas, but also the basis for achieving efficient and high-precision geometric correction. From the perspective of reusability of checkpoints and saving the cost of quality inspection of remote sensing images, this paper discusses the necessity of joint construction of multi-source and multi-resolution image checkpoint database. And put forward the construction principle and management objectives of checkpoint database. At last, this paper briefly introduces and prospects the application of the national multi-source and multi-resolution image checkpoint database.

Ecological environment quality of the Shanxi section of the Yellow River Basin under different development scenarios

Shanxi Province holds an important strategic position in the overall ecological pattern of the Yellow River Basin. To investigate the changes of the ecological environment in the Shanxi section of the Yellow River Basin from 2000 to 2020, we selected MODIS remote sensing image data to determine the remote sensing ecological index (RSEI) based on the principal component analysis of greenness, humidity, dryness, and heat. Then, we analyzed the spatial and temporal variations of ecological quality in this region to explore the influencing factors. We further used the CA-Markov model to simulate and predict the ecological environment under different development scenarios in the Shanxi section of the Yellow River Basin in 2030. The results showed that RSEI had good applicability in the Shanxi section of the Yellow River Basin which could be used to monitor and evaluate the spatiotemporal variations in its ecological environment. From 2000 to 2020, the Shanxi section of the Yellow River Basin was dominated by low quality habitat areas, in which the ecological environment quality continued to improve from 2000 to 2010 and decreased from 2010 to 2020. The high quality habitat areas mainly located on the mountainous areas with superior natural conditions and rich biodiversity, while the low ecological quality areas were mainly in the Taiyuan Basin and the northern part of the study area, where the mining industry developed well. Climate factors were negatively correlated with ecological environment quality in the northern and central parts of the study area, and positively correlated with that in the mountainous area. Under all three development scenarios, the area of cultivated land, forest, water and construction land increased in 2030 compared to that in 2020. Compared to the natural development scenario and the cultivated land protection scenario, the ecological constraint scenario with RSEI as the limiting factor had the highest area of new forest and the lowest expansion rate of cultivated land and construction land. The results would provide a reference for land space planning and ecological environment protection in the Shanxi section of the Yellow River Basin.

Seasonal Monitoring Method for TN and TP Based on Airborne Hyperspectral Remote Sensing Images

Airborne sensing images harness the combined advantages of hyperspectral and high spatial resolution, offering precise monitoring methods for local-scale water quality parameters in small water bodies. This study employs airborne hyperspectral remote sensing image data to explore remote sensing estimation methods for total nitrogen (TN) and total phosphorus (TP) concentrations in Lake Dianshan, Yuandang, as well as its main inflow and outflow rivers. Our findings reveal the following: (1) Spectral bands between 700 and 750 nm show the highest correlation with TN and TP concentrations during the summer and autumn seasons. Spectral reflectance bands exhibit greater sensitivity to TN and TP concentrations compared to the winter and spring seasons. (2) Seasonal models developed using the Catboost method demonstrate significantly higher accuracy than other machine learning (ML) models. On the test set, the root mean square errors (RMSEs) are 0.6 mg/L for TN and 0.05 mg/L for TP concentrations, with average absolute percentage errors (MAPEs) of 23.77% and 25.14%, respectively. (3) Spatial distribution maps of the retrieved TN and TP concentrations indicate their dependence on exogenous inputs and close association with algal blooms. Higher TN and TP concentrations are observed near the inlet (Jishui Port), with reductions near the outlet (Lanlu Port), particularly for the TP concentration. Areas with intense algal blooms near shorelines generally exhibit higher TN and TP concentrations. This study offers valuable insights for processing small water bodies using airborne hyperspectral remote sensing images and provides reliable remote sensing techniques for lake water quality monitoring and management.

Evaluation of Ecosystem Services in Ruoergai National Park, China

This study utilizes ArcGIS10.8 and ENVI5.3 software and the InVEST model to analyze and operate field survey data and remote sensing image data from Ruoergai National Park. The work simulates the soil retention, carbon storage, water supply, and habitat quality of the park to evaluate and analyze its four major ecosystem services. Subsequently, important areas of ecosystem services are zoned based on the results, aiming to provide decision-makers with a theoretical and scientific basis for formulating ecological environment restoration, protection, and management measures in Ruoergai National Park. The results showed the following: (1) In the study area, the land use types, ranked from largest to smallest, are grassland, unused land, forest, water area, and construction land. (2) Soil retention and water supply show an increasing trend, while carbon storage shows a decreasing trend. Habitat quality remains relatively stable, with most areas maintaining a high level of quality. (3) The importance zoning of ecosystem services in the study area exhibited a trend of “four increases and one decrease”. Specifically, the areas classified as moderately important, highly important, and extremely important all increased, while the area designated as generally important decreased. The findings indicate that climate change, land use type changes, and human activities are the primary factors influencing changes in ESs. It is crucial to prioritize highly important and extremely important areas for protection and utilization within Ruoergai National Park. Moving forward, it will be essential to minimize human activities that disrupt the ecosystem, while also focusing on the conservation and sustainable use of forest and grassland.

PMNet: a multi-branch and multi-scale semantic segmentation approach to water extraction from high-resolution remote sensing images with edge-cloud computing

In the field of remote sensing image interpretation, automatically extracting water body information from high-resolution images is a key task. However, facing the complex multi-scale features in high-resolution remote sensing images, traditional methods and basic deep convolutional neural networks are difficult to effectively capture the global spatial relationship of the target objects, resulting in incomplete, rough shape and blurred edges of the extracted water body information. Meanwhile, massive image data processing usually leads to computational resource overload and inefficiency. Fortunately, the local data processing capability of edge computing combined with the powerful computational resources of cloud centres can provide timely and efficient computation and storage for high-resolution remote sensing image segmentation. In this regard, this paper proposes PMNet, a lightweight deep learning network for edge-cloud collaboration, which utilises a pipelined multi-step aggregation method to capture image information at different scales and understand the relationships between remote pixels through horizontal and vertical spatial dimensions. Also, it adopts a combination of multiple decoding branches in the decoding stage instead of the traditional single decoding branch. The accuracy of the results is improved while reducing the consumption of system resources. The model obtained F1-score of 90.22 and 88.57 on Landsat-8 and GID remote sensing image datasets with low model complexity, which is better than other semantic segmentation models, highlighting the potential of mobile edge computing in processing massive high-resolution remote sensing image data.

Application of Enhanced YOLOX for Debris Flow Detection in Remote Sensing Images

Addressing the limitations, including low automation, slow recognition speed, and limited universality, of current mudslide disaster detection techniques in remote sensing imagery, this study employs deep learning methods for enhanced mudslide disaster detection. This study evaluated six object detection models: YOLOv3, YOLOv4, YOLOv5, YOLOv7, YOLOv8, and YOLOX, conducting experiments on remote sensing image data in the study area. Utilizing transfer learning, mudslide remote sensing images were fed into these six models under identical experimental conditions for training. The experimental results demonstrate that YOLOX-Nano’s comprehensive performance surpasses that of the other models. Consequently, this study introduces an enhanced model based on YOLOX-Nano (RS-YOLOX-Nano), aimed at further improving the model’s generalization capabilities and detection performance in remote sensing imagery. The enhanced model achieves a mean average precision (mAP) value of 86.04%, a 3.53% increase over the original model, and boasts a precision rate of 89.61%. Compared to the conventional YOLOX-Nano algorithm, the enhanced model demonstrates superior efficacy in detecting mudflow targets within remote sensing imagery.

Exploring the Driving Forces of Vegetation Greening on the Loess Plateau at the County Scale

China has implemented several ecological projects in the Loess Plateau region to address severe land degradation and soil erosion. Accurately assessing ecological restoration and its driving factors remains challenging. Previous studies in this area concentrated on driving factors have mainly focused on natural factors at the regional or watershed scale, with limited consideration of socioeconomic factors at the county scale. In this study conducted in Huanglong County on the Loess Plateau, the focus was to fill the gaps in previous research and provide insights into the socioeconomic driving forces behind vegetation greening. Remote sensing image data (NDVI) from 1999 to 2019 were used to analyze vegetation greenness dynamics in the region. Five socioeconomic variables were considered, including afforestation intensity, deforestation intensity, agricultural intensity, village intensity, and road intensity layers, to characterize the impact of afforestation, agriculture, and urbanization policies. The RESTREND (residual trends) method was employed to assess the relative importance of climate and human activities on vegetation dynamics. This study found that temperature–NDVI relationships are more suitable for building RESTREND models than precipitation–NDVI relationships. Human activity was the main driver of vegetation dynamics, contributing 62% compared to 38% from climate change. Agricultural practices and afforestation were found to have a positive impact on ecological restoration, while deforestation and urbanization had no significant impact. These findings highlight a conceptual framework for understanding the intricate relationship between ecological restoration, climatic factors, and human activity on the Loess Plateau. This study suggests that significant progress has been made in ecological restoration through human efforts in combating land degradation. However, it emphasizes the need to strengthen natural conservation efforts and gradually transition toward restoration processes driven by natural forces for sustainable socioeconomic development. The methodology used in this study can be applied to explore the driving forces of ecological restoration in other regions facing human-driven land degradation.

Comparison of the CASA and InVEST models’ effects for estimating spatiotemporal differences in carbon storage of green spaces in megacities

Urban green space is a direct way to improve the carbon sink capacity of urban ecosystems. The carbon storage assessment of megacity green spaces is of great significance to the service function of urban ecosystems and the management of urban carbon zoning in the future. Based on multi-period remote sensing image data, this paper used the CASA model and the InVEST model to analyze the spatio-temporal variation and driving mechanism of carbon storage in Shenzhen green space and discussed the applicability of the two models to the estimation of carbon storage in urban green space. The research results showed that, from 2008 to 2022, in addition to the rapid expansion of construction land, the area of green space and other land types in Shenzhen showed a significant decrease trend. The estimation results of the carbon storage model showed that the carbon storage of green space shows a significant trend of reduction from 2008 to 2022, and the reduction amounts are 0.8 × 106 t (CASA model) and 0.64 × 106 t (InVEST model), respectively. The evaluation results of the model show that, in megacities, the spatial applicability of InVEST model is lower than that of CASA model, and the CASA model is more accurate in estimating the carbon storage of urban green space. The research results can provide a scientific basis for the assessment of the carbon sink capacity of megacity ecosystems with the goal of "dual carbon".

Archimedes optimisation algorithm quantum dilated convolutional neural network for road extraction in remote sensing images

Roads are closely intertwined with human existence, and the process of extracting road networks has emerged as the most prominent task in remote sensing (RS). The automated road interpretation process of remote sensing images (RSI) efficiently acquires road network data at a reduced expense in comparison to the traditional visual interpretation of RSI. However the manifestation of RSI is completely distinct because of the great difference in length, width, material, and shape of road networks in dissimilar areas. Thus, the extraction of road network data in RSI is still a complex issue. In recent times, DL-based approaches have projected a famous development in image segmentation outcomes, but a lot of them still could not retain boundary data and attain high-resolution road segmentation maps while processing the RSI. Traditional convolutional neural networks (CNNs) demonstrate impressive performance in road extract tasks; however, they frequently encounter difficulties in capturing intricate details and contextual information. The study introduces a novel method, named Archimedes Optimisation Algorithm, Quantum Dilated Convolutional Neural Network for Road Extraction (AOA-QDCNNRE), to tackle the challenges encountered in remote sensing images. The AOA-QDCNNRE technique aims to generate a high-resolution road segmentation map using DL with a hyperparameter tuning process. The AOA-QDCNNRE technique primarily relies on the QDCNN model, which integrates quantum technology (QC) with dilated convolutions to augment the network's capacity to capture local as well as global contextual information. In addition, the incorporation of the dilated convolutional technique effectively enhances the receptive field without sacrificing spatial resolution, enabling the extraction of precise road features. To develop the road extraction outcomes of the QDCNN approach, the AOA-based hyperparameter tuning process can be exploited. The AOA-QDCNNRE system's simulation results can be tested on benchmark databases, and the results indicate that the AOA-QDCNNRE method surpasses recent algorithms.

Development genetic and stability classification of seasonal glacial lakes in a tectonically active area—A case study in Niangmuco, east margin of the Eastern Himalayan Syntaxis

The Niangmuco region on the east margin of the Eastern Himalayan Syntaxis features a large number of glacial lakes. The development process and stability classification of glacial lakes is of great significance to the study of seasonal glaciers in the eastern Himalayan margin, with implications for economic development and disaster prevention. Based on Landsat remote sensing image data from 2000 to 2021, this study analyzed the development and change characteristics of glacial lakes in the Niangmuco region during the past 21 years, and classified the stability of lakes with areas >0.02 km2 using the fuzzy consistent matrix method. In this area, 126 glacial lakes were identified within an elevation range of 3044–4156 m with a total area of 10.94 km2. These lakes primarily included glacial erosion lakes, valley lakes, tectonic lakes, and landslide dam lakes. Specifically, glacial erosion lakes accounted for 88.9% of the total number of lakes and 60.3% of the total lake area, followed by valley lakes with 6.3% and 23.7%, respectively. From 2000 to 2010, the total area of glacial lakes decreased from 10.53 km2 to 10.09 km2, which may be attributable to climate fluctuations. Subsequently, the area of lakes increased significantly to 10.94 km2 in 2021, an increase of 0.41 km2. Compared with 2000, among the lakes with a growth rate of 0.019 km2/a in 21 years, glacial erosion lakes exhibited the largest change. Among the classified glacial lakes in the study area, 95.7% were stable and relatively stable, mostly comprising glacial erosion lakes at high altitudes between 3468 and 4156 m. Only 4 unstable and extremely unstable glacial lakes were identified, and they were located near a fault zone. The findings show that the development and the change of glacial lakes in the area are primarily controlled by temperature and precipitation, and the topography and fault activity have important influences on the stability of glacial lakes.

Pattern change and ecological risk analysis of Shilin World Geopark landscape

This study considered Shilin World Geopark as the research object and constructed a landscape ecological risk assessment model based on the landscape pattern index by using remote sensing image data during five periods between 2000 and 2020. In addition, it analyzed the spatial and temporal changes of landscape ecological risk in the region. Spatial autocorrelation analysis was utilized to study spatial differences in the landscape ecological risk in the park. The results showed that during the study period, (1) cultivated land, forest land, and rocky desertification land were the main landscape types, different landscape types differed, and the area of rocky desertification land and building land increased by 37.47 km2 and 14.29 km2, respectively, while the area of cultivated land and grassland decreased significantly, with changes of 34.11 km2 and 18.67 km2; (2) landscape ecological risk of the park showed significant spatial differences, the ‘high–high’ risk areas have been concentrated mainly in the central and northern parts of the park, the ‘low–low’ risk areas have been concentrated in the central part and the southwest-southeast area of the park; and (3) landscape ecological risk of the geopark has been increasing, with the degree of landscape ecological risk being spatially positively correlated. The results of the study are of great significance for maintaining ecosystem health of the Shilin World Geopark and optimizing the ecological risk management of the park.

Photoelectric Detection Technology Utilizing Communication Remote Sensing Image Data

The photoelectric detection system stands as a crucial instrument for target identification and task execution within diverse environments. The system is frequently confronted with rapid task alterations in the context of communication remote sensing observations. Conventional detection systems encounter challenges in swiftly adapting, underscoring the pressing necessity for an intelligent photoelectric detection system capable of multifaceted tasks and rapid alignment with communication remote sensing settings. This study undertakes a technical exploration of the intelligent photoelectric detection system, delineating the coexistence of semi-intelligent and fully intelligent modes. While the semi-intelligent mode is selected for specific task scenarios, the fully intelligent mode seamlessly takes precedence in the absence of specific tasks. Upon task assignment, the detection mode is designated, automatically calibrating system parameters (operating bands, aperture, integration time, gain, and focal length) in alignment with task requisites. The architecture comprises a detection module capable of seamlessly transitioning between imaging and spectral dimensions, complemented by an autonomous data processing module crafted through DSP+FPGA+ARM technologies. Grounded in this technological foundation, the study designs and employs an intelligent photoelectric detection system to procure communication remote sensing image data, focusing on underwater acoustic signal analysis. The system’s configuration facilitates the creation of a communication remote sensing photoelectric detection mechanism specifically tailored for underwater acoustic signals. Rigorous experimentation involving laser communication in air and sound waves in water culminates in the successful acquisition of communication remote sensing image data. Experimental findings affirm the system’s efficiency in effectively detecting underwater acoustic signals.

Retracted: Application of Remote Sensing Image Data Scene Generation Method in Smart City

Retracted: Application of Remote Sensing Image Data Scene Generation Method in Smart City

Evaluation on high-performance image compaction algorithms in spatio-temporal data processing

With the passage of time, the importance of spatio-temporal data (STD) is increasing day by day, but the spatiotemporal characteristics of STD bring huge challenges to data processing. Aiming at the problems of image information loss, limited compression ratio, slow compression speed and low compression efficiency, this method based on image compression. This article intended to focus on aircraft trajectory data, meteorological data, and remote sensing image data as the main research objects. The research results would provide more accurate and effective data support for research in related fields. The image compaction algorithm based on deep learning in this article consisted of two parts: encoder and decoder, and this method was compared with the JPEG (Joint Photographic Experts Group) method. When compressing meteorological data, the algorithm proposed in this paper can achieve a maximum compaction rate of 0.400, while the maximum compaction rate of the JPEG compaction algorithm was only 0.322. If a set of aircraft trajectory data containing 100 data points is compressed to 2:1, the storage space required for the algorithm in this paper is 4.2 MB, while the storage space required for the lossless compression algorithm is 5.6 MB, which increases the compression space by 33.33%. This article adopted an image compaction algorithm based on deep learning and data preprocessing, which can significantly improve the speed and quality of image compaction while maintaining the same compaction rate, and effectively compress spatial and temporal dimensional data.

A Multi-Modal Deep-Learning Air Quality Prediction Method Based on Multi-Station Time-Series Data and Remote-Sensing Images: Case Study of Beijing and Tianjin.

The profound impacts of severe air pollution on human health, ecological balance, and economic stability are undeniable. Precise air quality forecasting stands as a crucial necessity, enabling governmental bodies and vulnerable communities to proactively take essential measures to reduce exposure to detrimental pollutants. Previous research has primarily focused on predicting air quality using only time-series data. However, the importance of remote-sensing image data has received limited attention. This paper proposes a new multi-modal deep-learning model, Res-GCN, which integrates high spatial resolution remote-sensing images and time-series air quality data from multiple stations to forecast future air quality. Res-GCN employs two deep-learning networks, one utilizing the residual network to extract hidden visual information from remote-sensing images, and another using a dynamic spatio-temporal graph convolution network to capture spatio-temporal information from time-series data. By extracting features from two different modalities, improved predictive performance can be achieved. To demonstrate the effectiveness of the proposed model, experiments were conducted on two real-world datasets. The results show that the Res-GCN model effectively extracts multi-modal features, significantly enhancing the accuracy of multi-step predictions. Compared to the best-performing baseline model, the multi-step prediction's mean absolute error, root mean square error, and mean absolute percentage error increased by approximately 6%, 7%, and 7%, respectively.