JARS thanks the reviewers who served the journal in 2024.

Rapid damage assessment after disasters is crucial for humanitarian relief and emergency response. The abrupt and unpredictable nature of disasters causes variations in the time, location, and sensors used for image collection, resulting in significant data disparities in satellite imagery. This poses a significant challenge for assessment tasks. To enable a rapid response, training models from scratch using a sufficient amount of on-site data is impractical due to time constraints. Thus, in practical applications, a model with robust adaptability and generalization is essential for autonomously adjusting to data variations. However, the majority of current models are trained and tested on a singular dataset, neglecting the aforementioned issues. To address these challenges, this study created datasets from the Turkey earthquake with large and dense building features and introduced datasets from the Louisiana hurricane with small and sparse building features. These datasets exhibit significant style differences and cover a broader range of building characteristics, providing a comprehensive evaluation of the model’s adaptability. In the building damage assessment task, the model is trained on public datasets and validated using the newly introduced scenario data. Compared with existing assessment models, the U-net model demonstrates the highest adaptability performance in objectively evaluating damage levels.

Landslide identification is an important task in the field of geologic disaster monitoring and early warning, which is of great significance for improving social safety and mitigating the impact of disasters. With the development of computer vision, deep learning is widely used in landslide recognition research. We focus on segmenting landslides from high-resolution optical satellite images using convolutional neural network. Currently, deep learning semantic segmentation models still face issues such as neglecting small objects and incorrectly segmenting terrain features with similar shapes and pixel characteristics. Considering the unbalanced distribution of categories and large differences in scene styles during the extraction of key feature information from remote sensing images, landslides have diverse and complex backgrounds. We propose a fusion DeepLabv3+ and completed local binary pattern (CLBP) landslide image semantic segmentation method (CLBP-DeepLabv3+), using the improved inverted residual block as the core structure of backbone to extract different levels of image information, and after backbone extracts landslide image features, it connects the improved DenseASPP to fuse the different levels of features to better pay comprehensive attention to local and global features and obtain contextual information at different scales. Then, the texture and edge features of the image are extracted using CLBP, and the multi-level features are merged by introducing the feature aggregation module, which constitutes the CLBP-DeepLabv3+ model. Through ablation experiments and comparative tests on a self-made dataset, the experimental results show that the proposed method performs the best on the validation set, with a mean intersection over union (mIoU) of 88.62%, mean pixel accuracy (mPA) of 94.17%, recall rate of 90.17%, and an intersection over union (IoU) for landslides of 80.53%. Compared with the original DeepLabv3+ model, the improved DeepLabv3+ increased the mIoU by 3.15%, mPA by 3.99%, recall by 4.93%, and IoU by 4.97%. Compared with other semantic segmentation models, the improved DeepLabv3+ also achieved better segmentation accuracy in extracting landslide features.

With advancements in unmanned aerial vehicle (UAV) technology, UAV applications are rapidly growing, and their operations are becoming increasingly intelligent. Localization of UAVs commonly relies on global navigation satellite systems combined with inertial navigation systems through sensor fusion. However, this approach is vulnerable to significant risks, such as signal spoofing. In military conflicts, signal spoofing by hackers poses a severe security threat with potentially catastrophic outcomes. To address this issue, we propose a two-stage vision-based UAV localization method. This approach utilizes multi-category semantic segmentation and template matching to establish a connection between heterogeneous sensors. Experimental results demonstrate the method’s effectiveness in accurately identifying the UAV’s location within extensive geographical areas captured in remote sensing images. In addition, it achieves high precision in aligning UAV locations with Baidu maps, offering robust and accurate localization capabilities.

The digital elevation model (DEM) serves as a vital data source for surface 3D modeling. Due to the limitations in sampling conditions and cost constraints, we usually obtain unevenly distributed and relatively sparse sampling points. To reconstruct a complete DEM of the sampling area, we need to utilize spatial interpolation algorithms. However, traditional spatial interpolation methods typically have lower model complexity and often involve a large number of iterative calculations to approximate the points to be interpolated. This often results in significant interpolation errors and low real-time performance. We propose a multi-scale conditional generative adversarial network (multi-scale cGAN) with adaptive joint loss weights. In addition, during the model training process, we design a joint loss function that incorporates generator adversarial loss, content loss, and perceptual loss, with the ability to adaptively adjust the weight coefficients of each component, thereby optimizing model training and further improving its generalization and generation ability. The experimental results demonstrate that compared with the traditional spatial interpolation algorithm and other typical deep learning–based models, the interpolation error on typical land DEM data (including slopes, valleys, and ridges) is smaller, and the interpolated image has the highest clarity and similarity compared with the original image. Overall, our approach demonstrates high robustness and low error when dealing with DEM spatial interpolation tasks in complex terrain environments while also possessing the potential for expansion into various terrain environment DEM reconstruction applications.

Satellite products are widely used to understand and monitor the dynamics of the Peruvian coastal upwelling system (PCUS), one of the four major eastern boundary upwelling systems. They complement in situ observations obtained through surveys and oceanographic stations periodically operated by the Peruvian Marine Research Institute (IMARPE). Therefore, we mainly aim to assess the reliability of satellite products for sea surface temperature (SST) and surface chlorophyll-a (chl-a) using a unique in situ database from IMARPE. We evaluated SST data from MODIS (levels 2 and 3) and chl-a data from MODIS and SeaWiFS (both at level 3) due to their prior and continuous utilization. In the case of SST, our results demonstrated a strong correlation and good performance from MODIS compared with IMARPE data, although there is a tendency to slightly overestimate it, particularly during autumn. For surface chl-a, during their respective satellite periods, both sensors exhibited a favorable correlation and performance, with SeaWiFS outperforming MODIS. When analyzing matching points in terms of location and timing, both sensors tended to underestimate IMARPE data (autumn). Despite some differences between satellite and IMARPE data, the evaluated satellite products provide reliable information for understanding and monitoring the PCUS. However, for chl-a, it is important to consider the differences between sensors for retrospective studies.

To rapidly and accurately determine the spatial distribution of permanganate index (CODMn) concentrations in surface water, we employed multispectral unmanned aerial vehicle (UAV) remote sensing for water quality inversion monitoring. River extraction was conducted using the normalized difference water index (NDWI) in conjunction with a self-developed parameter a. Fifteen characteristic variables were derived through exponential calculation, and modeling was performed in Jiaobei River using both single-band variables and those with a correlation coefficient exceeding 0.3 and tested in Nanguan River. The model selection included the elastic net regression algorithm, random forest regression algorithm, and gradient boosting regression algorithm. The findings indicated that accurate river extraction is achieved when NDWI exceeds 0.2 and a remains below 0. The gradient boosting regression algorithm proved to be the superior method for constructing an inversion model of CODMn in the Jiaobei River. When applied to the Nanguan River, the model also yielded reliable inversion results. However, in the concentration range of 3 to 5mg/L, the inversion results were less accurate due to an insufficient sample size. A quantitative remote sensing monitoring model of water quality in Taizhou was established using a multispectral UAV, providing technical support for accurate water pollution control.

Using ground control points (GCPs) can improve positioning accuracy in the calibration process, but the acquisition of GCPs is always difficult and high cost, especially in mountain areas and abroad. Thus, we proposed an improved calibration algorithm assisted by the digital elevation model data to calibrate geometric parameters for the single-scene synthetic aperture radar (SAR) image with no need for GCPs by providing an alternative solution that introduces topographic slope as an index of applicability and reliability. The algorithm includes three main steps: SAR image simulation, registration parameters calculation between the real and simulated images using a very large search window, and time delay estimation in the range and azimuth directions based on the above registration results. Experiments on Gaofen-3 SAR images of different modes show that the proposed approach provides good performance without any GCPs, reducing the errors of checkpoints from 151.3 to 26.6 m. Besides, the orthoimages further demonstrate the effectiveness of the algorithm in the whole image, especially in mountainous areas where the layover is well compensated. Finally, the applicability of the proposed method is analyzed, proving its great suitability for SAR images with terrain slopes greater than 0.05.

Land cover changes are the main causes of carbon emissions and ecological damage, so accurate land cover data are essential for precise carbon emissions estimates. The Dempster–Shafer (D-S) evidence theory was used to fuse three sets of 30-m land cover data: China’s National Land Use and Cover Change (CNLUCC), Global 30-m Land Cover Classification with a Fine Classification System (GLC_FCS30), and the China Land Cover Dataset (CLCD), to product high-precision land cover data, and carbon emissions were calculated based on carbon emission factor method in the Yangtze River Delta (YRD) from 1990 to 2020. The results show that (1) the average annual accuracies of the three products, CNLUCC, GLC_FCS30, and CLCD, are 64%, 78%, and 76%, respectively, with CLC_FCS30 showing higher classification accuracy. (2) The D-S evidence theory as a fusion method effectively integrates multi-source land cover data, enhancing the overall accuracy of the fused data by 11% to 22% compared with the original data. (3) Carbon emissions in the YRD region reached 594.043 million tons by the end of 2020, marking an increase of 468.480 million tons since 1990. Notably, there is a significant spatial variability in carbon emissions, with coastal areas exhibiting higher emissions, and a shift toward inland regions. This can serve as a valuable reference for urban planning and environmental conservation in the YRD region.

Understanding wind conditions is critical for the NASA Advanced Air Mobility (AAM) mission. The types of aircraft, and the region of the atmosphere they operate in, make them highly susceptible to wind effects. We demonstrate the feasibility of using wind lidar to measure wind dynamics for the AAM mission. Wind vector measurements from two Doppler wind lidars using dual-Doppler techniques are compared with in situ measurements from a ground-based sonic anemometer and small uninhabited aircraft systems (sUAS). Both lidar beams intersected directly above a sonic anemometer and measurements were compared. The resulting root mean square error values between the two instruments’ speed and direction measurements were 1.72 m/s and 23.05 deg, respectively. Following this test, a dual-Doppler scan pattern which measured wind vectors along a vertical column was performed while a sUAS measured wind vectors along the scanned volume for comparison. The wind profiles from the two measurement techniques are consistent and demonstrate the potential of using Doppler lidar for AAM.

We explore barite-bearing lead–zinc mineral occurrences within Türkiye’s Central Taurus Orogenic Belt. Remote sensing technologies, such as ASTER, ALOS Palsar, and Planet data, were used to enhance the identification and evaluation of barite-bearing lead–zinc occurrences. The area, predominantly composed of carbonate rocks of different ages, presents geological complexities that challenge traditional exploration techniques. We employ various remote sensing methods, including band ratio techniques, minimum noise fraction, and support vector machine classifications, to differentiate lithological units, particularly distinguishing carbonate formations associated with mineralization. In addition, lineament analysis was conducted to identify structural features such as faults and fractures that control mineralization. The results demonstrate the effectiveness of remote sensing technology in identifying mineralized areas, despite the similar lithologies of the carbonate formations, and provide a reference for future exploration studies in the Taurus Orogenic Belt and similar geological settings.

The tides of the Qiantang River in eastern China are one of the three major tides in the world. Although these tides are spectacular, they seriously threaten the seawalls of the river. Rapid and accurate monitoring of ground deformations along the seawalls is important not only to the seawalls themselves but also to the vast amount of land behind them. We carried out comprehensive, unprecedented high-density mapping of ground deformations along the seawalls of the Qiantang River using the latest full scatterer (FS) interferometric synthetic aperture radar (InSAR) technique with 56 Sentinel-1 SAR images acquired from October 2020 to October 2022. The InSAR-derived deformations were then validated with quasi-synchronous leveling measurements. The results demonstrate that there are five subsidence centers on the northern seawalls and three on the southern seawalls, with a maximum subsidence rate of 96 mm/a and a measuring density of 1293 points/km2. The main reasons for the severe subsidence are the natural consolidation of soft soil in reclaimed areas, groundwater overexploitation due to aquaculture, and foundation pit drainage. We present, for the first time, a comprehensive mapping of ground deformations along the seawalls of the Qiantang River at ultrahigh density, which is very important for the health management of seawalls and the safety of the river on both sides.

The accurate positioning of the specular point of the spaceborne global navigation satellite system reflectometry (GNSS-R) receiver is a key step in using GNSS-R technology for geophysical parameter retrieval. As GNSS-R applications expand across diverse environments ranging from land and oceans to glaciers, traditional geometric positioning accuracy is no longer sufficient to meet the demands of these varied applications. Specifically, positioning algorithms must not only facilitate real-time spaceborne processing but also adapt to fluctuating surface conditions, all while sustaining high levels of accuracy and efficiency. In response to these demands, we propose a positioning algorithm tailored for spaceborne GNSS-R specular points across various surface environments. The proposed algorithm integrates actual surface height data by leveraging surface models for localized terrain fitting. It employs a regional grid traversal method, guided by geographic positioning criteria, to accurately search for specular points that satisfy specific conditions. Considering the impact of initial position accuracy on traversal speed, we developed an ellipsoidal reflection geometric model based on the unique properties of the ellipsoid. This model facilitates the swift, non-iterative determination of a high-precision initial position for specular points. Experimental data indicate that our proposed approach reduces the Euclidean distance error of the initial position by ∼87.5% and enhances grid traversal efficiency by a factor of more than eight times. To verify the robustness of the proposed method, this paper conducted multiple experiments in both ocean and land scenarios, including case studies in the Qinghai-Tibet Plateau region. The findings highlight that the normal vector angle analysis criterion offers superior reliability in complex terrains, positioning it as the optimal criterion for the positioning of specular points.

Laser image detection and ranging (LiDAR) data are used as an important source for monitoring the natural environment and environmental changes, forest management, transport, and urban development. Real-time LiDAR is especially useful in applications that require immediate feedback or decision-making, such as post-earthquake emergency analysis, autonomous vehicles, precision agriculture, military applications, and infrastructure monitoring. For instance, rapid and real-time assessment of structural damage following disasters like earthquakes ensures the effective use of resources by directing first responders to priority areas. We focused on minimizing the time losses in pre-flight preparation and post-flight data processing processes at the low-altitude LiDAR mapping. To eliminate time losses, LiDAR data, which can be accessed via a web browser, were visualized as a real-time 3D point cloud. The X3D file format was used in this visualization process. To realize the accessibility and sharing of the data, HTML5 technologies were used, and the data were transferred to the cloud system in real time. Thus, a user-friendly system was developed that can make these data available to researchers simultaneously. We designed and developed a platform-independent web-based real-time LiDAR (WebRT-LiDAR) mapping system. As an application, the performance of the WebRT-LiDAR mapping system was evaluated by integrating it into a drone (low-altitude Unmanned Aerial Vehicle) platform. The data received with this system can be stored in databases suitable for the use of artificial intelligence and machine learning–based systems. In this way, the system provides information, predictions, and capabilities that can be used in situations that require instant feedback or decision-making. In this context, the performance of the WebRT-LiDAR system can be further improved and tested in different scenarios, such as disaster management, war strategy, and efficient agriculture in future studies.

Long-term reconstructed solar-induced chlorophyll fluorescence (SIF) derived from raw gridded SIF has been used for the estimation of gross primary production (GPP), but the robustness of the spatial relationship may vary from location to location. We examined the often-used linear relationship between GPP and SIF in terms of R2 values for varied locations globally using three GPP datasets (FLUXCOM, VPM, PML) and three long-term reconstructed monthly SIF datasets (CSIF, SIF005, and RTSIF). The results show that the R2 value is a concave function of vegetation greenness level (NDVI) on an annual or seasonal basis. The average R2 is over 0.8 in areas where the annual average NDVI is in the range of 0.4 to 0.6, whereas the R2 is much lower where the annual average NDVI is less than 0.2 or greater than 0.8. Prediction of GPP or SIF by three methods from five major environmental variables revealed greater uncertainties in GPP and/or SIF at low or high greenness levels as an apparent cause of the low R2. The results offer useful insights into how global GPP may be effectively estimated from multi-satellite measured SIF.

Rice fallow areas present a significant opportunity to enhance food production and address the growing global food demand driven by population growth. Many lands remain fallow primarily due to insufficient water for cereal crops. To promote sustainable intensification, cultivating legumes in rice fallows offers a short-duration, water-efficient alternative that can improve soil health and potentially double farmers’ incomes. This study focuses on mapping rice fallow areas and soil moisture levels in West Bengal, a leading rice-producing state in India, where approximately 11.7 million hectares of land is left fallow in the rice cropping system. Utilizing K-means clustering and random forest algorithms, we effectively map these areas by exploiting optical and microwave satellite data from the Sentinel satellite. We also synchronize soil moisture data for the post-Kharif and post-Rabi periods. The methodology demonstrates the monthly dynamics of rice fallow and soil moisture from October to May during the 2019 to 2020 period, using high-resolution Sentinel-2 satellite imagery. The key findings of this study include (1) identification of a 1.97 million hectare Rabi rice fallow area in West Bengal; (2) high probabilities for crop intensification in 53% of post-Kharif and 43% of post-Rabi rice fallows; (3) development of high-resolution maps indicating optimal locations for legume or vegetable cultivation for sustainable intensification. We highlight that soil moisture rapidly depletes after rice harvesting, emphasizing the need for timely interventions. This innovative approach provides vital insights for decision-making regarding sustainable intensification in rice fallows through efficient legume cultivation practices.

In recent years, algorithms for fast and accurate recognition and detection of circular markers have become crucial in the field of high-speed videogrammetry. However, most existing techniques often necessitate a manually selected region of interest that encompasses the full information of the circular marker. This manual box selection method is inefficient and unsuitable for practical engineering applications. To address this issue, we propose a global automatic recognition and detection approach that employs multi-level constraints for identifying circular markers in high-speed videogrammetry. First, an edge detection method based on the Canny algorithm is employed to extract candidate edges containing all circular markers. Subsequently, two geometric constraints—general geometric condition and roundness metric constraint—are applied to eliminate a large number of non-circular marker edges. Finally, pseudo-edges of circular markers are removed, and the corresponding accurate edges are retained by applying an extrema point condition constraint. The performance of the proposed method is evaluated using several high-speed videogrammetry image datasets. Experimental results demonstrated that our method can accurately detect and recognize all circular markers, outperforming comparable methods. The proposed method holds promise for efficient and wide-ranging applications in the recognition and detection of circular markers in high-speed videogrammetry.

With the rapid development of artificial intelligence technology, deep learning has achieved significant advantages in synthetic aperture radar automatic target recognition (SAR-ATR). However, previous research showed that the addition of small perturbations not easily detected by the human eye can lead to SAR-ATR model recognition errors; that is, they are affected by adversarial attacks. To solve the problem of long computation time in existing SAR sparse adversarial attack algorithms, we propose a SAR fast sparse adversarial attack (FSAA) algorithm. First, an end-to-end sparse adversarial attack framework is developed based on the lightweight generator ResNet model using two different upsampling modules to control the amplitude and position of the adversarial perturbation. A loss function for the generator is then constructed, which mainly consists of the linear addition of the attack loss, the amplitude distortion loss, and the sparsity loss. Finally, the SAR image is mapped through the trained generator model in a one-step process to generate sparse adversarial perturbations quickly and effectively. Compared with the existing SAR sparse adversarial attack algorithm, the experimental results show that the generation speed of the proposed method is at least 30 times higher when the perturbation is less than 0.05% of the pixels in the entire image, and the recognition rate of the model is >13%.

For the synthetic aperture radar (SAR) target recognition task, feature representation accuracy determines the recognition performance. The previous mainstream deep learning–based SAR target recognition methods required a large amount of labeled samples to learn feature representation and made superior performance. However, only utilizing unlabeled samples to solve SAR target recognition is still a challenging task. To solve this problem, we propose an end-to-end multi-view feature enhancement-based contrastive clustering framework (MvECC) for unsupervised SAR target recognition. It utilizes the view of internal information and complementary information among adjacent views from multiple-view SAR images to learn discriminative target features. MvECC first augments the multi-view image sequences to build sequence pairs and feeds them into a pair of weight-sharing multi-view Vision Transformers (ViT) to extract features. The designed multi-view ViT has an intra-view and an inter-view Transformer layer, and it can capture and fuse the feature within each view and among different views from multi-view image sequences. Then, contrastive learning is performed at the sequence and class levels, which aims at optimizing pairwise similarity to learn feature representations and obtain clustering assignment results. The clustering accuracy of our method outperforms the state-of-the-art method by 7.87% on the moving and stationary target acquisition and recognition dataset. More experimental results on the synthetic and measured paired labeled experiment dataset show that MvECC has good robustness.

Very high-resolution remote sensing images are used to detect the changes with finer results. Accompanying with plenty of details, the environment around targets also becomes more complex, which poses challenges for change detection. In recent years, hybrid methods based on transformer and CNN have been widely used. The two methods usually are adopted to extract the change information because they can take advantage of global semantic relations and long-range spatial dependencies at the same time. However, within some complex environments, there is still information loss and inaccurate detection because the features cannot be fully integrated. We proposed a new cross-level hybrid feature aggregation network for change detection to improve the performance of change detection, especially within a complex environment. Within the new network, a parallel hybrid CNN-Transformer structure is adopted to model globally and locally, which extracts the features of different levels and produces rich semantic features. Meanwhile, the multi-branch feature interaction is used to implement interaction and fusion for multiscale feature information. Furthermore, multiscale feature aggregation was applied to remove redundancy. Subsequently, CNN-Transformer change feature enhancement is used to enhance the representation. Compared with several state-of-the-art methods on three available datasets, the accuracy is increased by 0.09%, 1.12%, and 2.62%, respectively. The experiments indicate that the method proposed in this paper detects the changed targets as continuous and complete objects with clear edges. Within a complex environment, it suppresses pseudo-changes and extracts more small changed targets.

Multi-label scene classification (MLSC) in remote sensing (RS) plays a crucial role in recognizing the intricate contents of RS images. However, the MLSC task is challenged by the complexity of label combinations and the diversity of visual content. Multispectral (MS) data provide valuable information for scene classification. To achieve accurate classification results, it is important to fully utilize the diverse information contained in MS data. To this end, we propose a multispectral transformer network (MST-Net), which leverages transformer-based architectures to capture the diverse information within MS images and the complex relationships between labels and features. Specifically, MST-Net consists of a feature fusion encoder and a semantic query decoder. Within the encoder, we have developed an MS deformable attention module based on a sampling strategy that reduces focus on redundant spectral areas, allowing for better integration of complementary MS information. In the decoder, geographic information is introduced as an inductive bias, utilizing the unique spatiotemporal characteristics of RS images to learn better class-related features. Extensive experiments were conducted on two RS multi-label datasets, namely, LSCIDMRv2 and BigEarthNet. Comparisons with several state-of-the-art multi-label classification methods demonstrate the effectiveness and superiority of MST-Net.

Hyperspectral unmixing (HU) is a critical technique for analyzing remote sensing images. Its primary goal is to extract endmembers and their corresponding abundances from complex hyperspectral images. In recent years, nonnegative matrix factorization (NMF)-based methods have gained significant attention in the field of HU. We propose a method called superpixel-guided manifold sparse NMF (SMSNMF) for HU. Unlike traditional NMF-based approaches, SMSNMF effectively utilizes spatial information within a regularized manifold, establishing a robust connection between the original image and its associated abundance maps. Furthermore, because neighboring pixels in hyperspectral images are closely related, superpixels can efficiently represent these relationships. Superpixel segmentation identifies meaningful homogeneous regions where pixels exhibit similar spatial properties. The primary advantage of SMSNMF lies in its ability to fully leverage superpixels for constructing an affinity graph, encoding spatial geometric information within the manifold. Moreover, SMSNMF integrates sparsity constraints to explore the sparse nature of abundance. These sparse constraints can also be combined with other methods to further enhance NMF performance. Experimental results, conducted on both synthetic and real hyperspectral datasets, demonstrate that the proposed method outperforms several state-of-the-art techniques in the field of HU.

Due to the limitations of hyperspectral satellite imaging systems, hyperspectral images (HSIs) have a low spatial resolution and spatial missing regions, which hinders their full potential for land cover classification. Through collaboration with the auxiliary high-resolution multispectral image (MSI), the high-resolution HSI (HHSI) can be obtained. Although image fusion deep learning methods have emerged as the predominant approach for enhancing the spatial resolution of HSI, few studies have solved the problem of spatial missing. Moreover, the reliance on the spectral response function and the point spread function limits the applicability and performance of these methods. This study proposes an HSI cross-region reconstruction framework, which can not only fuse the HSI and MSI in overlapping regions but also be used to reconstruct the missing information of HSI in non-overlapping regions. The core of this framework is an unsupervised deep end-to-end fusion network with adaptive attention based on the conception of endmember and abundance. Specifically, a multi-layer attention mechanism with an interaction module is designed to enhance the extraction of endmember and abundance features. In addition, the attached module addresses the dependency on the prior information about degradation models. To verify the practicability of the proposed framework, we conduct evaluations using both simulation datasets and real on-board data for experiments involving both overlapping and non-overlapping regions. The experimental results demonstrate that the proposed method is effective and the reconstructed HHSI is reliable.

Urban functional areas, with their intricate semantics, constitute complex ground scenes. Current methods for classifying urban functional areas require human-computer interaction for execution, and fully automated approaches are scarce. Most automatic classification methods rely on the traditional segmentation algorithm, performing semantic recognition only on the segmented image blocks, resulting in misrecognition. To enhance precision in automatic classification, we introduce an approach employing a graph convolutional network (GCN) to refine semantic features extracted by a convolutional neural network (CNN), termed the convolution-graph convolutional neural network (C-GCNN) model. To minimize the impact of irregular image segmentation CNN-based semantic extraction, regular image blocks are used for segmentation. The C-GCNN model takes a set of adjacent image blocks as input for the CNN and uses the output vector from the last fully connected layer as the semantic feature for the central image block. Each central pixel block serves as a data node, and the resulting graph data are input into GCN. The GCN is trained by transduction learning, and its output serves as the classification results for urban functional areas. Using both simulated and high-resolution remote sensing images for validation, the method achieved high total classification accuracies 96.07% for simulated images and 94.38% for real-world applications, as shown by the confusion matrix analysis. The C-GCNN processes semantic features through the GCN to achieve accurate classification of urban functional areas. The experimental results show that the C-GCNN’s effectiveness in classifying urban functional areas.

In recent years, solar-induced chlorophyll fluorescence (SIF) has been widely studied as an ideal “probe” to directly monitor vegetation photosynthesis. Satellite-retrieved SIF shows great potential for estimating gross primary productivity (GPP), but discontinuous space coverage limits its application. Although there are now some products to reconstruct satellite-scale SIF, it is mainly focused on established carbon satellites such as Orbiting Carbon Observatory-2 (OCO-2) and Tropospheric Monitoring Instrument (TROPOMI). Orbiting Carbon Observatory-3 (OCO-3), a carbon satellite launched by NASA in 2019, has more flexibility and observability than OCO-2, but there are still fewer SIF products for OCO-3. We generated a global contiguous SIF product SIFxco2 (8 days, 0.05 deg) over the period 2000 to 2021 from multivariate OCO-3 data and MODIS. The predictive model is developed by CatBoost. We innovatively added column-averaged atmospheric carbon dioxide concentrations (XCO2) as an explanatory variable, which is usually inversely related to SIF, to improve the model’s SIF explanation performance. SIFxco2 obtains good performance (R2=0.84, RMSE = 0.062) in estimating OCO-3 SIF, and it also correlates well with the flux tower GPP (R2=0.746). In addition, it accurately responds to the temporal and spatial variations of level 2 SIF on a global scale. Our SIFxco2 product holds value in evaluating terrestrial photosynthesis and ecosystem functioning, as well as serving as a benchmark for terrestrial biosphere and Earth system models.

Convolutional neural networks (CNNs) are widely used in hyperspectral image (HSI) classification due to their strong feature extraction capabilities. Nevertheless, CNN-based classification methods face challenges in capturing deep semantic features effectively, and as the number of layers increases, computational expenses escalate significantly. In addition, spatial locations and spectral bands in HSI data have varying discriminative power. To address these challenges, we propose a novel approach, DSNet, which integrates a multiway attention mechanism with a CNN-based deep semantic network to improve HSI classification. First, we extract shallow abstract features using a module that combines 3D and 2D convolution layers, followed by transformation through a Gaussian-weighted feature tokenizer. Second, we introduce an enhanced bifurcation attention mechanism with dynamically adjustable perceptual ranges to capture spatial semantic information from both horizontal and vertical perspectives. Third, a convolutional multilayer perceptron module is employed to extract high-level feature representations in both spatial and spectral dimensions, which are then refined by a light self-Gaussian attention module. Finally, a linear layer is used to determine the sample labels. Experiments conducted on three publicly available hyperspectral datasets illustrate that the proposed DSNet achieves superior performance compared with several contemporary state-of-the-art methods, even when trained with limited samples.

Due to issues of limited receptive fields, convolutional neural network-based change detection (CD) methods often fail to exploit the long-range spatial context information, resulting in sub-optimal CD performance. To overcome the limitations, a remote sensing image CD network with a multi-scale transformer (MSTransCDNet) is proposed, which follows a classic encoder-decoder architecture based on a multi-scale transformer (MSTransformer) unit. In particular, in MSTransformer, by introducing multi-scale key-value vectors, a multi-scale multi-head attention unit is constructed, allowing for the extraction of multi-scale features with rich spatial context information, thereby effectively addressing multi-scale targets and irregular geometric structures. In addition, a depthwise separable convolution (DWConv) layer is employed in the feed-forward network of MSTransformer, which significantly enhances the local information. The effectiveness of the proposed method is verified on change detection dataset (CDD) and learning, vision and remote sensing change detection (LEVIR-CD) datasets. Experimental results demonstrated the superiority of the proposed MSTransCDNet, which outperforms several state-of-the-art CD methods by a large margin, achieving an F1-score of 96.49% and 90.91% for CDD and LEVIR-CD, respectively.

Erratum corrects typos in Table 1.