The bidirectional reflectance factor (BRF) characterizes the directional radiative properties of terrestrial surfaces. However, existing three-dimensional (3D) radiative transfer models require complex scene construction and computationally intensive radiative transfer solvers, limiting efficient generation of multi-angle hyperspectral reflectance imagery. 3D Gaussian Splatting (3DGS) offers an efficient framework for neural scene representation and novel view synthesis, but its low-order spherical harmonics representation is insufficient for complex directional reflectance, while the high dimensionality and inter-band quality differences of hyperspectral data introduce additional challenges. To address these challenges, we propose BRF-GS, a 3DGS-based framework for BRF modeling and hyperspectral reflectance image generation. BRF-GS introduces a hybrid BRDF-driven kernel to represent complex directional reflectance, selects geometry-reliable spectral bands for robust 3D scene initialization, and adopts a two-stage training strategy that decouples geometry optimization from spectral modeling. We further construct the AIR-BRF dataset, a multi-angle hyperspectral directional reflectance dataset comprising three scenes with diverse natural and artificial targets. Experiments demonstrate that BRF-GS achieves superior spatial and spectral fidelity and accurately reproduces characteristic view-dependent BRF responses. The proposed framework provides an efficient data-driven approach for BRF modeling and multi-angle hyperspectral reflectance image generation in remote sensing scenes.

Hyperspectral images generally suffer from low spatial resolution, limiting their utility in fine-scale applications. To address the limitations of existing unsupervised fusion methods in adapting to spatial heterogeneity and scale effects that often lead to blurred details and reduced target discrimination, we propose an unsupervised adaptive scale-aware detail feature extraction network (UASNet), which introduces a structure-adaptive mechanism and degradation-aware modeling to effectively harmonize the representation consistency of multiscale ground objects without paired training samples. The network consists of three stages: the prior information mining stage, the spectral channel mapping stage, and the detail feature fusion stage. Specifically, an adaptive scale-aware convolution is embedded within a reversible detail extraction module to capture features of objects with varying scales and geometries, thereby preserving fine textures and structural integrity. Furthermore, driven by the requirements of real-world application scenarios, spectral and spatial branches are constructed to learn the respective degradation priors, which are integrated into the loss function to realize a fully unsupervised fusion framework. Extensive experiments on both simulated and real datasets demonstrate the superior performance of the proposed method compared with that of state-of-the-art approaches. Moreover, even without ground truth data, the downstream classification results indirectly validate that UASNet exhibits strong potential for real-world applications.

Joint classification of multi-modal remote sensing data, such as hyperspectral image (HSI) and Light Detection and Ranging (LiDAR), is important for Earth observation. Early deep learning-based methods usually adopt a single network to extract the features of HSI and LiDAR data, respectively. However, this kind of single-network architecture has limited ability to represent complex features, which may deliver a suboptimal classification result. Moreover, many existing hybrid network frameworks often rely on simple feature fusion operations, such as concatenation, cascading, or summation, and fail to explicitly model the complex complementary relationships across modalities. To address this issue, this paper proposes a spatial-spectral multi-scale Mamba fusion network (S2MSMamba) for HSI and LiDAR data classification. In the proposed framework, the Mamba branch models the global long-range dependencies with linear computational complexity, while the convolutional neural network (CNN) branch extracts local spatial textures and edge information. S2MSMamba jointly exploits the complementary strengths of Mamba and CNN by effectively mining the global long-range dependencies and the local spatial structure features of ground objects. In addition, a multi-scale hierarchical alignment mechanism is developed to address scale variations of ground objects. Moreover, an adaptive decision-level fusion strategy is further adopted to integrate uni-modal feature to refine classification performance. Experiments conducted on the Houston, MUUFL, and Trento benchmark datasets show that the proposed S2MSMamba outperforms existing methods on several evaluation metrics, demonstrating its effectiveness for land-cover classification.

Novel View Synthesis (NVS) can reconstruct scenes from multi-view images and synthesize novel images from new viewpoints, which provides technical support for tasks such as target recognition and environmental perception. Aerial remote sensing can conveniently capture a wealth of multi-view images with just a few flights. However, the challenges brought by large distances and sparse viewing angles during collection can cause the model to easily produce floaters and overgrowth issues due to geometric estimation errors. This results in low visual quality and a lack of precise geometric estimation capabilities. Therefore, this study presents ARSGaussian, an innovative novel view synthesis (NVS) method for aerial remote sensing. The method incorporates LiDAR point cloud as constraints into the 3D Gaussian Splatting approach, adaptively guiding the Gaussians to grow and split along geometric benchmarks, thereby addressing the overgrowth and floaters issues. Additionally, considering the geometric distortions arising from data acquisition, coordinate transformations with distortion parameters are integrated to replace the simple pinhole camera model parameters to achieve pixel-level alignment between LiDAR point cloud and multi-view optical images, facilitating the accurate fusion of heterogeneous data and achieving the high-precision geo-alignment. Moreover, depth, normal and scale consistency losses are introduced into the regularization process to guide Gaussians toward real depth and plane representations, significantly improving geometric estimation accuracy. To address the current lack of dense airborne hybrid datasets, we have established and released AIR-LONGYAN, an open-source dataset containing a dense LiDAR point cloud (8 pts/m2) and multi-view optical images captured by airborne scanners and cameras in diverse scenes. Experimental results demonstrate that our method achieves more realistic NVS compared to other stateof-the-art (SOTA) methods tailored for large-scale scenes (e.g., VastGaussian, Momentum-GS, CityGaussianV2). On the public UrbanScene3D dataset, our approach attains a PSNR of 26.75 dB, showcasing superior rendering fidelity. Furthermore, it significantly improves geometric structure estimation accuracy, reducing the RMSE from 1.626 m (suboptimal LetsGO baseline) to 0.327 m, marking a 79.88 % enhancement in geometric estimation precision. Our self-build AIR-LONGYAN dataset and code will be available at https://github.com/WenjuanZhang -aircas/ARSGaussian.

Low-resolution satellites, due to their wide coverage and fast data acquisition, are commonly used in large-scale studies. However, these optical remote sensing data are often limited by weather conditions and sensor system issues during acquisition, which leads to missing information. For example, MODIS data, as a typical representative of low-resolution satellites, often encounter issues of small-region data loss, which corresponds to a large area on the surface of the earth due to the relatively large spatial scale of the pixels, thereby limiting the high-quality application of the data, especially in building datasets for deep learning. Currently, most missing data restoration methods are designed for medium-resolution data. However, low-resolution satellite data pose greater challenges due to the severe mixed-pixel problem and loss of texture features, leading to suboptimal restoration results. Even MNSPI, a typical method for restoring missing data based on similar pixels, is not exempt from these limitations. Therefore, this study integrates four-temporal phase characteristic information into the existing MNSPI algorithm. By comprehensively utilizing temporal–spatial–spectral information, we propose an algorithm for restoring small missing regions. Experiments were conducted under two scenarios: areas with complex surface types and areas with homogeneous surface types. Both simulated and real missing data cases were tested. The results demonstrate that the proposed algorithm outperforms the comparison methods across all evaluation metrics. Notably, we statistically analyzed the optimal restoration range of the algorithm in cases where similar pixels were identified. Specifically, the algorithm performs optimally when restoring regions with connected pixel areas smaller than 1936 pixels, corresponding to approximately 484 km2 of missing surface area. Additionally, we applied the proposed algorithm to global surface reflectance data restoration, further validating its practicality and feasibility for large-scale application studies.

Since the 1960s,remote sensing science and technology has emerged as a competitive high-tech field,with major countries striving to advance their capabilities.It has become a fundamental tool for human research in the earth system science and the comprehensive application of aerospace information across multiple domains.Recently,two significant developments warrant attention:First,in 2022,the Ministry of Education of China officially recognized Remote Sensing Science and Technology as a first-level interdisciplinary discipline within the graduate education framework,thereby strengthening foundational research in remote sensing and broadening its application areas.Second,the rise of artificial intelligence technologies,particularly deep learning,has ushered in a new paradigm for data-driven analysis and application of remote sensing data.While remote sensing fundamentally belongs to the domain of electromagnetic radiation physics,the associated physical models have been indispensable for the development of quantitative remote sensing.Nevertheless,the data-driven deep learning paradigm has introduced transformative ideas and methodologies to the field.Moving forward,the synergy between physical models and artificial intelligence will undoubtedly shape the future trajectory of remote sensing research and applications.In this context,a deeper exploration of the core concepts and fundamental issues in remote sensing science is crucial for achieving significant technological breakthroughs and scientific discoveries within this discipline. This article begins by examining the physical origins of remote sensing science,focusing on the interaction between ground objects and electromagnetic waves,which produces spectral radiation images under specific conditions.It explores the characteristics of various remote sensing methods across the electromagnetic spectrum,including solar reflected radiation in the visible to shortwave infrared remote sensing,daylight-induced chlorophyll fluorescence(SIF)remote sensing,laser remote sensing,both medium and longwave infrared remote sensing,and microwave remote sensing.The fundamental theoretical issues in remote sensing science are categorized into three primary characteristics:radiative,spectral,and temporal characteristics,along with five major effects:scale,atmospheric,angular,adjacent,and transfer effects.The former pertains to the intrinsic physical and chemical properties of ground objects within the electromagnetic spectrum,while the latter relates to factors such as imaging scale,atmospheric conditions,observation angle,and background environment.This discussion includes the expression and variation patterns of remote sensing features of land objects formed under diverse observation modes and conditions. The radiative characteristics reflect overall difference in term of radiation across different electromagnetic bands for various land covers,closely tied to geophysical and chemical properties.The spectral characteristics of land cover manifest as variations in the intensity of reflected and emitted signals with wavelength,highlighting significant differences in absorption,reflection,and emission behaviors among different materials,known as spectral characteristics.Temporal characteristics pertain to the systematic changes in spectral reflection or emission over time,aiding in remote sensing identification or feature inversion of land cover.The scale effect refers to the changes in remote sensing observation characteristics due to variations in pixel area size,influenced by spatial resolution or point scanning density(e.g.,laser scanning spot density).The atmospheric effect describes how electromagnetic waves are impacted by the absorption,scattering,and emission from atmospheric particles during remote sensing imaging,leading to radiation distortion in image data.The angular effect highlights the directional nature of the interaction between land cover and electromagnetic waves,resulting in significant anisotropic characteristics and variations in radiation values based on the angles of incident radiation,remote sensing observation,and electromagnetic wave wavelength.The adjacent effect refers to the influence of spatial structure heterogeneity among land features,which can create cross-radiation contributions from non-target pixels to target pixels,dependent on spatial distribution and remote sensing observation mode.Finally,the transfer effect encompasses the changes in imaging quality after the electromagnetic signal of the ground objects entering the remote sensing system,including the processes such as photoelectric conversion,signal transmission,and digital recording. The review and discussion presented in this article on the fundamental issues of remote sensing science aim to deepen theoretical research in the field,particularly in the context of artificial intelligence.This exploration is intended to foster innovative methods in remote sensing technology and applications,promote the collaborative evolution of AI for Science and Science for AI in remote sensing,and encourage profound cross-disciplinary integration between remote sensing and other fields.

Optical remote sensing images, as a significant data source for Earth observation, are often impacted by cloud cover, which severely limits their widespread application in Earth sciences. Synthetic aperture radar (SAR), with its all-weather, all-day observation capabilities, serves as a valuable auxiliary data source for cloud removal (CR) tasks. Despite substantial progress in deep learning (DL)-based CR methods utilizing SAR data in recent years, challenges remain in preserving fine texture details and maintaining image visual authenticity. To address these limitations, this study proposes a novel diffusion-based CR method called the Dual-branch Multimodal Conditional Guided Diffusion Model (DMDiff). Considering the intrinsic differences in data characteristics between SAR and optical images, we design a dual-branch feature extraction architecture to enable adaptive feature extraction based on the characteristics of the data. Then, a cross-attention mechanism is employed to achieve deep fusion of the multimodal feature extracted above, effectively guiding the progressive diffusion process to restore cloud-covered regions in optical images. Furthermore, we propose an image adaptive prediction (IAP) strategy within the diffusion model, specifically tailored to the characteristics of remote sensing data, which achieves a nearly 20 dB improvement in PSNR compared to the traditional noise prediction (NP) strategy. Extensive experiments on the airborne, WHU-OPT-SAR, and LuojiaSET-OSFCR datasets demonstrate that DMDiff outperforms SOTA methods in terms of both signal fidelity and visual perceptual quality. Specifically, on the LuojiaSET-OSFCR dataset, our method achieves a remarkable 17% reduction in the FID metric over the second-best method, while also yielding significant enhancements in quality assessment metrics such as PSNR and SSIM.

Due to the narrow swath width of hyperspectral remote sensing images, the limited availability of such data is insufficient to meet the demands of large-scale applications. To address this issue, spectral superresolution leveraging the extensive coverage of existing multispectral images offers a promising solution. However, current research methods overlook the issue of heterogeneous land cover in real-world data, thereby limiting the practical applicability of deep neural networks. To overcome these challenges, we propose a spectral-spatial residual attention U-Net (SSU-Net) to improve spectral superresolution performance in heterogeneous land cover scenarios. Specifically, a weight-adaptive allocation module is integrated at the model's front to address inconsistent spectral reconstruction accuracy caused by varying land cover distributions. Moreover, given the long-range spectral dependencies and local spatial correlations in hyperspectral images, our model incorporates a dual-branch design, consisting of a spectral branch and a spatial branch, to effectively extract spectral and spatial features separately. Considering the scarcity of large-scale datasets, we also constructed two real-world datasets to validate the effectiveness of the proposed network. Extensive experiments demonstrate that the proposed SSU-Net achieves state-of-the-art performance, exhibiting strong practical applicability for large-scale, real-world scenarios.

Hyperspectral remote sensing, which can acquire data in both spectral and spatial dimensions, has been widely applied in various fields. However, the available data are limited by factors such as revisit time, imaging width, and weather conditions. Three-dimensional (3D) hyperspectral simulation based on ray tracing can overcome these limitations by enabling physics-based modeling of arbitrary imaging geometries, solar conditions, and atmospheric effects. This type of simulation offers advantages in acquiring multi-angle and multi-condition quantitative results. However, the 3D hyperspectral simulation requires substantial computational resources. With the development of hardware, a graphics processing unit (GPU) offers a potential way to accelerate it. This paper proposes a 3D hyperspectral simulation model based on GPU-accelerated ray tracing, which is realized by modifying and using a common graphics API (OpenGL). Through experiments, we demonstrate that this model enables 600-band hyperspectral simulation with a computational time of just 2.4 times that of RGB simulation. Furthermore, we analyzed the balance between calculation efficiency and accuracy, and carried out a correlation analysis between ray count and accuracy. Additionally, we verified the accuracy of this model by using UAV-based data. The results demonstrate over 90% spectral curve similarity between simulated and UAV-acquired images. Finally, based on this model, we conducted additional simulation experiments under different environmental variables and observation conditions to analyze the model’s ability to characterize different situations. The results show that the model effectively captures the effects of environmental variables and observation conditions on the hyperspectral characteristics of vehicles.

Due to the all-time and all-weather characteristics of synthetic aperture radar (SAR) data, they have become an important input for optical image restoration, and various cloud removal datasets based on SAR-optical have been proposed. Currently, the construction of multi-source cloud removal datasets typically employs single-polarization or dual-polarization backscatter SAR feature images, lacking a comprehensive description of target scattering information and polarization characteristics. This paper constructs a high-resolution remote sensing dataset, AIR-POLSAR-CR1.0, based on optical images, backscatter feature images, and polarization feature images using the fully polarimetric synthetic aperture radar (PolSAR) data. The dataset has been manually annotated to provide a foundation for subsequent analyses and processing. Finally, this study performs a performance analysis of typical cloud removal deep learning algorithms based on different categories and cloud coverage on the proposed standard dataset, serving as baseline results for this benchmark. The results of the ablation experiment also demonstrate the effectiveness of the PolSAR data. In summary, AIR-POLSAR-CR1.0 fills the gap in polarization feature images and demonstrates good adaptability for the development of deep learning algorithms.

Optical remote sensing images play a crucial role in the observation of the Earth's surface. However, obtaining complete optical remote sensing images is challenging due to cloud cover. Reconstructing cloud-free optical images has become a major task in recent years. This paper presents a two-flow Polarimetric Synthetic Aperture Radar (PolSAR)-Optical data fusion cloud removal algorithm (PODF-CR), which achieves the reconstruction of missing optical images. PODF-CR consists of an encoding module and a decoding module. The encoding module includes two parallel branches that extract PolSAR image features and optical image features. To address speckle noise in PolSAR images, we introduce dynamic filters in the PolSAR branch for image denoising. To better facilitate the fusion between multimodal optical images and PolSAR images, we propose fusion blocks based on cross-skip connections to enable interaction of multimodal data information. The obtained fusion features are refined through an attention mechanism to provide better conditions for the subsequent decoding of the fused images. In the decoding module, multi-scale convolution is introduced to obtain multi-scale information. Additionally, to better utilize comprehensive scattering information and polarization characteristics to assist in the restoration of optical images, we use a dataset for cloud restoration called OPT-BCFSAR-PFSAR, which includes backscatter coefficient feature images and polarization feature images obtained from PoLSAR data and optical images. Experimental results demonstrate that this method outperforms existing methods in both qualitative and quantitative evaluations.

Context: The Loess Plateau plays an important role in ensuring food security in China. In the past few decades, the plastic film mulching (PFM) system has been widely used in the Loess Plateau, but whether PFM technology can maintain the stability of yield and the sustainability of soil water in the Loess Plateau remains uncertain. Objective: To clarify whether PFM technology can maintain yield stability and soil water sustainability in the Loess Plateau, and reveal the effects of PFM on water use, yield, and regional water balance. Methods: Using a 12-year field experiment (2012-2023), meta-analysis, and data-driven modeling, we demonstrated the following. Results and conclusions: (1) In the Loess Plateau, compared to the CK treatment, PFM and film mulching with nutrient management (PFM+N) significantly improved both water use efficiency (WUE; +28 %, +31 %, respectively) and yield (+21 %, +33 %) in maize, and PFM+N synergistically enhanced yield stability (coefficient of variation 32.36 %) and sustainability (Sustainability Yield Index 0.5). (2) Future climate scenario simulations revealed that from 2021 to 2040, both CK and PFM treatments showed an increasing trend in the proportion of water-deficit areas across the Loess Plateau. However, PFM demonstrated critical compensatory advantages by improving precipitation use efficiency and optimizing annual soil moisture redistribution, effectively alleviating regional water imbalance. (3) Spatial analysis using data-driven models further indicated that PFM mitigated water deficits in the central, western, and northern regions (which have relatively low precipitation) through reduced evapotranspiration losses, whereas in southern areas characterized by relatively high precipitation, non-mulching practices proved more ecologically sustainable because excessive PFM use would disrupt natural hydrological cycles. while non-mulching practices proved more ecologically sustainable in southern areas where excessive PFM use could disrupt natural hydrological cycles. Significance: Overall, the PFM system effectively alleviated the overall water consumption of the Loess Plateau, significantly increased WUE and maize yield, and, of great significance to sustainable development, the effect was best in low rainfall areas (Northwest of the Loess Plateau. Growing season precipitation: <250 mm; non-cropped precipitation:<200 mm). In the long run, PFM could effectively ensure water resource sustainability in the Loess Plateau, thereby laying a solid foundation for sustainable agricultural development in the region.

Hyperspectral images (HSI) inherently face a trade-off between spatial resolution and spectral resolution due to the limitations of imaging principles. To rapidly obtain remote sensing images with both high spatial and high spectral resolution, unsupervised deep learning methods for fusing HSI and multispectral images (MSI) have achieved remarkable progress in recent years. However, existing studies often overlook the issue of high-frequency information loss in the fused images. To address this limitation, we propose a Prior-based three-stage unsupervised Invertible neural Fuse Network (PIFNet). Specifically, the framework consists of three key modules: prior information extraction, spectral channel mapping, and detail feature fuse. In particular, the detail feature fuse module leverages an invertible neural network to prevent information loss through mutual generation of input and output features. Experimental results on simulated datasets demonstrate that the PIFNet outperforms existing unsupervised deep learning methods, highlighting its potential and effectiveness in HSI-MSI fusion tasks.

3D Gaussian Splatting (3D GS) achieves impressive results in novel view synthesis for small, single-object scenes through Gaussian ellipsoid initialization and adaptive density control. However, when applied to large-scale remote sensing scenes, 3D GS faces challenges: the point clouds generated by Structure-from-Motion (SfM) are often sparse, and the inherent smoothing behavior of 3D GS leads to over-reconstruction in high-frequency regions, where have detailed textures and color variations. This results in the generation of large, opaque Gaussian ellipsoids that cause gradient artifacts. Moreover, the simultaneous optimization of both geometry and texture may lead to densification of Gaussian ellipsoids at incorrect geometric locations, resulting in artifacts in other views. To address these issues, we propose PSRGS, a progressive optimization scheme based on spectral residual maps. Specifically, we create a spectral residual significance map to separate low-frequency and high-frequency regions. In the low-frequency region, we apply depth-aware and depth-smooth losses to initialize the scene geometry with low threshold. For the high-frequency region, we use gradient features with higher threshold to split and clone ellipsoids, refining the scene. The sampling rate is determined by feature responses and gradient loss. Finally, we introduce a pre-trained network that jointly computes perceptual loss from multiple views, ensuring accurate restoration of high-frequency details in both Gaussian ellipsoids geometry and color. We conduct experiments on multiple datasets to assess the effectiveness of our method, which demonstrates competitive rendering quality, especially in recovering texture details in high-frequency regions.

Single-image super-resolution (SISR) for high-resolution (HR) remote sensing image (RSI) acquisition is becoming increasingly valuable and important, and convolutional neural networks (CNNs) have produced considerable progress in this field. In RSIs, many similar geo-objects recur within the same scene, maintaining the same positions in both low resolution (LR) and HR. Based on this observation, we found that these similar geo-objects could be utilized to reconstruct texture details in LR by exploiting the consistent non-local relationships between these geo-objects in LR and HR, thereby improving the quality of SISR. Therefore, we propose a novel graph convolutional network (GCN) for SISR including a dynamic graph attention mechanism to learn the in-scale and cross-scale non-local features of RSIs. In scale, we propose a dynamic graph attention block (DGAB) that adaptively determines non-local patches upon the scene correlation derived from RSIs and further fuses patch-wise non-local information weighed by the attention scores of topological relationships and radiation characteristics in RSIs. Across different scales, we also introduce a dynamic graph attention mixing block (DGAMB) to upsample LR non-local information to HR non-local information. Most SISR methods have the upsampling blocks at the end of the network, ignoring feature extraction in high-dimensional space. To address this problem, DGAMB was designed as an upsampler in the middle of the model, enhancing the level of high-dimensional information extraction from the model. The experiments based on the WHU Building and UC Merced datasets show that our proposed method outperforms state-of-the-art methods. Our code is available at https://github.com/WenjuanZhang-aircas/NSGCN.

High-precision radiometric calibration is the basis for quantitative applications of hyperspectral remote sensing. Cross-calibration facilitates the cross-comparison and radiation reference transfer between multi-source hyperspectral equipment and normalizes different remote sensors to a common radiometric baseline. In the collaborative use of different unmanned aerial vehicle (UAV) hyperspectral observations, cross-calibration helps to eliminate the differences in the radiometric and spectral scales of the multi-source remote sensors, improve the radiometric quality and interpretation consistency of the imaging from different remote sensors. However, a significant portion of the error in cross-calibration between UAV hyperspectral instruments using radiation transfer modeling comes from the assumption of aerosol type. When using the irradiance method for calculations, it is important to consider the case that the uplink radiation transfer from the UAV remote sensors passes through only a portion of the atmosphere. Therefore, cross-calibration is necessary to improve the radiation transfer model with its own characteristics. In this paper, we propose the cross-calibration method for UAV hyperspectral to address the above problems. A full set of data such as multi-gray level target images, atmospheric aerosol, water vapor content data, etc. are collected in our experiment. The method improves the traditional irradiance calibration method by combining the measured atmospheric diffuse-to-global ratio, and effectively reduces the error caused by the aerosol assumption by taking into account the special characteristics of the uplink radiation transmission path of the UAV. At the same time, considering that it is difficult to satisfy the need of cross-calibration of the whole response interval by using a single reflectance feature, the experiment adopts six kinds of targets with different gray levels for cross-calibration. Finally, the accuracy and impact of different response intervals are analyzed. The results demonstrate that the method proposed in this paper can ensure the cross-calibration accuracy more reliably, especially when the aerosol type is difficult to be determined, and it is very suitable for cross-radiometric calibration between UAV sensors.

The timely and accurate prediction of winter wheat yields is of importance in maintaining food security. However, existing deep-learning methods used for crop yield prediction are limited. While most methods utilize recurrent neural networks (RNNs) to interpret crop time series data, they struggle to learn geographical spatial information from input data and often prove challenging to interpret with prior knowledge. In this study, hyperspectral images were used as the input to a BS-Nets network for band selection, and a graph-based RNN framework GT-long-short-term memory (LSTM) [two channels based LSTM-graph neural network (GNN)], was proposed for predicting the winter wheat yield at the county level. Using the BS-Nets for hyperspectral bands selection at the field scale, we obtained the top 4 bands in selection results for all six stages, the results were band 46 (791 nm), 50 (825 nm), 66 (954 nm), and 161 (2484 nm). Based on the results of hyperspectral bands selection, at the county scale, the similar wavelength bands of Sentinel-2 red-edge 3 (783 nm), NIR (834 nm), red-edge 4 (865 nm), SWIR2 (2190 nm) were chosen as inputs for the GT-LSTM county-level estimates of winter wheat yield. When only remote sensing data were used, the highest prediction accuracy ( $R^{2}= 0.688$ , RMSE = 0.54 t/ha) was obtained for DOY135 (30 days before harvest). The incorporation of the meteorological data improved the accuracy by 7% ( $R^{2}= 0.714$ , RMSE = 0.50 t/ha), and the optimal time for predicting wheat yield was at DOY115 (50 days before harvest). Further addition of GNN layers to the model improved the accuracy of the results by an additional 14% ( $R^{2}= 0.757$ , RMSE = 0.43 t/ha), and the best prediction results were then obtained at DOY105 (60 days before harvest).

Winter wheat constitutes approximately 20% of China's total cereal production. However, calculations of total production based on multiplying the planted area by the yield have tended to produce overestimates. In this study, we generated sample points from existing winter wheat maps and obtained samples for different years using a temporal migration method. Random forest classifiers were then constructed using optimized features extracted from spectral and phenological characteristics and elevation information. Maps of the harvested and planted areas of winter wheat in Chinese eight provinces from 2018 to 2022 were then produced. The resulting maps of the harvested areas achieved an overall accuracy of 95.06% verified by the sample points, and the correlation coefficient between the CROPGRIDS dataset is about 0.77. The harvested area was found to be about 13% smaller than the planted area, which can primarily be attributed to meteorological hazards. This study represents the first attempt to map the winter wheat harvested area at 10-m resolution in China, and it should improve the accuracy of yield estimation.

The ultraviolet spectrum has significant applications in the fields of global auroral detection， marine oil spill， atmospheric glow， etc. Surface reflectance is important background data in the research. However， the existing satellite data resources are relatively insufficient to meet the application needs. In this study， a machine learning-based on XGBoost algorithm is proposed for simulating surface reflectance data in the Near-Ultraviolet （N-UV） （350 nm – 400 nm） spectral channel. Firstly， Sentinel-2 MSI 2， 3 and 4 channels are selected as the data source and the spectral of vegetation， water， soil and other typical features are extract based on the USGS spectral database， then equivalently calculated to the corresponding channels. Secondly， the correlation analysis between the data source and the channel to be simulated is carried out. The correlation coefficients between Sentinel-2 MSI 2， 3， and 4 channels and the channel to be simulated are all greater than 0.88， which indicates that the N-UV surface reflectance simulation can be carried out based on this data source. Thirdly， based on the typical spectral data set after the equivalent calculation constructed XGBoost regression model to simulate the N-UV channel surface reflectance. Results indicate that the coefficient of determination (R 2 ) of all the channel models is above 0.91， the root mean square error （RMSE） is less than 0.076， the mean absolute error percentage （MAPE） is within 20%， and the standard deviation of the above three accuracy indicators for different categories of samples is within 0.0212， which shows that the model has high accuracy and robustness. Finally， based on the Sentinel-2 MSI 2， 3 and 4 channels image data， the simulated images of surface reflectance at 355 nm， 365 nm， 375 nm， 385 nm and 395 nm were generated， and the images better reflected the spectral characteristics of the surface.