International journal of applied earth observation and geoinformation : ITC journal最新文献_第7页

CAFTrans: Content-Aware Fusion Transformer for ground-based remote sensing cloud detection CAFTrans：用于地面遥感云检测的内容感知融合变压器

IF 8.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2026-01-06 DOI: 10.1016/j.jag.2025.105044

Shuang Liu , Zeyu Yu , Ying Liu , Zhong Zhang , Chaojun Shi , Baihua Xiao

Recently, Transformer structures have been applied to ground-based remote sensing data to detect clouds, and the vanilla self-attention as the core component of Transformer utilize all tokens in the process of modeling long-range dependencies, resulting in involving less semantically correlated tokens. In this paper, we propose a novel Transformer network named Content-Aware Fusion Transformer (CAFTrans) for the ground-based remote sensing cloud detection task, which could effectively select the relevant tokens according to the content information of cloud sample. To this end, we propose the Content-Aware Selection Attention (CASA) in the Transformer encoder where we first construct the token-to-token similarity matrix with a learnable weight matrix and then dynamically select the tokens with high semantic relevance for each query according to the similarity matrix. Meanwhile, we introduce the Multi-Scale Fusion Mechanism (MSFM), which is built upon CASA and designed to capture long-range dependencies across multiple feature scales. To facilitate model training and evaluation, we present the Large-Scale Occluded Cloud Detection Dataset (LOCDD)—the first ground-based remote sensing dataset to consider three categories: clouds, sky, and occlusions. Comprehensive quantitative results and qualitative visualizations on the LOCDD dataset demonstrate the robust performance of the proposed CAFTrans model. The source code is freely available at: https://github.com/shuangliutjnu/CAFTrans.

近年来，Transformer结构已被应用于地面遥感数据的云检测中，而Transformer的核心组件vanilla self-attention在远程依赖关系建模过程中利用了所有标记，导致涉及较少的语义相关标记。本文针对地面遥感云检测任务，提出了一种基于内容感知融合变压器（CAFTrans）的新型变压器网络，该网络可以根据云样本的内容信息有效地选择相关令牌。为此，我们在Transformer编码器中提出了内容感知选择注意（Content-Aware Selection Attention， CASA），首先用一个可学习的权重矩阵构造令牌到令牌的相似性矩阵，然后根据相似性矩阵为每个查询动态选择具有高语义相关性的令牌。同时，我们引入了基于CASA的多尺度融合机制（MSFM），该机制旨在捕获跨多个特征尺度的远程依赖关系。为了便于模型训练和评估，我们提出了大规模遮挡云检测数据集（LOCDD），这是第一个考虑云、天空和遮挡三大类的地面遥感数据集。LOCDD数据集的综合定量结果和定性可视化显示了所提出的CAFTrans模型的鲁棒性。源代码可以在https://github.com/shuangliutjnu/CAFTrans免费获得。

{"title":"CAFTrans: Content-Aware Fusion Transformer for ground-based remote sensing cloud detection","authors":"Shuang Liu , Zeyu Yu , Ying Liu , Zhong Zhang , Chaojun Shi , Baihua Xiao","doi":"10.1016/j.jag.2025.105044","DOIUrl":"10.1016/j.jag.2025.105044","url":null,"abstract":"<div><div>Recently, Transformer structures have been applied to ground-based remote sensing data to detect clouds, and the vanilla self-attention as the core component of Transformer utilize all tokens in the process of modeling long-range dependencies, resulting in involving less semantically correlated tokens. In this paper, we propose a novel Transformer network named Content-Aware Fusion Transformer (CAFTrans) for the ground-based remote sensing cloud detection task, which could effectively select the relevant tokens according to the content information of cloud sample. To this end, we propose the Content-Aware Selection Attention (CASA) in the Transformer encoder where we first construct the token-to-token similarity matrix with a learnable weight matrix and then dynamically select the tokens with high semantic relevance for each query according to the similarity matrix. Meanwhile, we introduce the Multi-Scale Fusion Mechanism (MSFM), which is built upon CASA and designed to capture long-range dependencies across multiple feature scales. To facilitate model training and evaluation, we present the Large-Scale Occluded Cloud Detection Dataset (LOCDD)—the first ground-based remote sensing dataset to consider three categories: clouds, sky, and occlusions. Comprehensive quantitative results and qualitative visualizations on the LOCDD dataset demonstrate the robust performance of the proposed CAFTrans model. The source code is freely available at: <span><span>https://github.com/shuangliutjnu/CAFTrans</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105044"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926085","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Increased far-field landslide hazards due to postseismic acceleration by the 2017 Mw 6.4 Nyingchi earthquake 2017年林芝6.4级地震的震后加速导致远场滑坡危害增加

IF 8.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2026-01-06 DOI: 10.1016/j.jag.2025.105077

Shuang Zhu , Chuang Song , Chen Yu , Zhenhong Li , Yi Chen , Zhenjiang Liu , Kui Liu , Guangqian Zou , Jianbing Peng

Landslides represent a major type of earthquake-triggered geohazard, with their dynamic evolution influenced by the complex interaction of postseismic environmental factors. The 2017 Mw 6.4 Nyingchi earthquake triggered numerous landslides, yet research has largely focused on coseismic mapping, and there remains a lack of cataloging and characterization of postseismic landslides accelerated by the earthquake (i.e., Earthquake Accelerated Landslides, EALs). These landslides remain a serious hazard to surrounding communities, with impacts that persist well beyond the earthquake. This study employs Sentinel-1 time-series analysis to reveal the spatial distribution and postseismic deformation mechanisms of EALs. A total of 299 EALs were identified and cataloged, which are primarily distributed in high mountainous regions west of the epicenter and show a distinct linear clustering along active faults. By comparing near-field and far-field EALs, we found that far-field landslides may experience greater acceleration despite weaker ground shaking. Time-series analysis indicated a rise in EAL deformation velocity from 1.56 cm/yr (preseismic) to 5.08 cm/yr (postseismic), reflecting a strong accelerating effect of the earthquake. In addition to the deformation acceleration triggered by the mainshock, postseismic landslide deformation also exhibited seasonality correlated to precipitation and land surface temperature, as well as aftershock-induced variations. Exponential modeling further indicates a decreasing deformation trend of EALs, with stabilization occurring approximately 6.8 years after the mainshock. This research systematically examined landslide evolution after the Nyingchi earthquake, providing a basis for post-earthquake landslide hazard assessment, with particular emphasis on the increased far-field landslide hazards.

滑坡是地震诱发的主要地质灾害类型，其动态演变受震后环境因素复杂相互作用的影响。2017年林芝6.4级地震引发了大量滑坡，但研究主要集中在同震制图上，并且仍然缺乏地震加速的震后滑坡的编目和特征（即地震加速滑坡，EALs）。这些山体滑坡对周围社区仍然是一个严重的威胁，其影响远远超出了地震。本文利用Sentinel-1时间序列分析揭示了地震带的空间分布及其震后变形机制。共发现并编目了299个地震震源，主要分布在震中以西的高山区，沿活动断层呈明显的线状聚集。通过比较近场和远场EALs，我们发现远场滑坡可能经历更大的加速度，尽管地面震动较弱。时间序列分析表明，地震前EAL变形速度从1.56 cm/yr（震前）上升到5.08 cm/yr（震后），反映了地震的强烈加速作用。除主震引起的变形加速外，震后滑坡变形还表现出与降水、地表温度以及余震相关的季节性变化。指数模型进一步表明，主震后约6.8年出现稳定，EALs变形呈减小趋势。本研究系统考察了林芝地震后滑坡的演变，为震后滑坡危险性评价提供了依据，重点研究了远场滑坡危险性的增加。

{"title":"Increased far-field landslide hazards due to postseismic acceleration by the 2017 Mw 6.4 Nyingchi earthquake","authors":"Shuang Zhu , Chuang Song , Chen Yu , Zhenhong Li , Yi Chen , Zhenjiang Liu , Kui Liu , Guangqian Zou , Jianbing Peng","doi":"10.1016/j.jag.2025.105077","DOIUrl":"10.1016/j.jag.2025.105077","url":null,"abstract":"<div><div>Landslides represent a major type of earthquake-triggered geohazard, with their dynamic evolution influenced by the complex interaction of postseismic environmental factors. The 2017 Mw 6.4 Nyingchi earthquake triggered numerous landslides, yet research has largely focused on coseismic mapping, and there remains a lack of cataloging and characterization of postseismic landslides accelerated by the earthquake (i.e., Earthquake Accelerated Landslides, EALs). These landslides remain a serious hazard to surrounding communities, with impacts that persist well beyond the earthquake. This study employs Sentinel-1 time-series analysis to reveal the spatial distribution and postseismic deformation mechanisms of EALs. A total of 299 EALs were identified and cataloged, which are primarily distributed in high mountainous regions west of the epicenter and show a distinct linear clustering along active faults. By comparing near-field and far-field EALs, we found that far-field landslides may experience greater acceleration despite weaker ground shaking. Time-series analysis indicated a rise in EAL deformation velocity from 1.56 cm/yr (preseismic) to 5.08 cm/yr (postseismic), reflecting a strong accelerating effect of the earthquake. In addition to the deformation acceleration triggered by the mainshock, postseismic landslide deformation also exhibited seasonality correlated to precipitation and land surface temperature, as well as aftershock-induced variations. Exponential modeling further indicates a decreasing deformation trend of EALs, with stabilization occurring approximately 6.8 years after the mainshock. This research systematically examined landslide evolution after the Nyingchi earthquake, providing a basis for post-earthquake landslide hazard assessment, with particular emphasis on the increased far-field landslide hazards.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105077"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926447","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Local effects of pattern interactions in driving urbanization 模式相互作用在城市化驱动中的局部效应

IF 8.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2026-01-06 DOI: 10.1016/j.jag.2025.105072

Yanfang Sun , Guosheng Wu , Yongze Song , Haiyang Liu , Lin Wang , Zehua Zhang , Jiao Hu

Spatial association and spatial interaction are fundamental to understanding geographical phenomena and regional development disparities, with broad applicability across disciplines. Existing spatial heterogeneity analysis face significant challenges in capturing pattern interactions and local variability. This study develops a local pattern interaction (LPI) model that integrates local complexity patterns or geocomplexity of spatial data, the interaction of patterns, and their locally varied power of determinants (PD). LPI is implemented in assessing the PD of local variables and pattern interactions on the spatial distributions of urbanization using statistical data, remote sensing imagery, and open geospatial data. The results show that LPI effectively identifies the local PD of interactions involving the geocomplexity patterns of urbanization-related explanatory variables. Model performance is evaluated by comparison with the optimal-parameters geographical detector (OPGD), a widely used spatial heterogeneity–based PD identification model. The model validation shows that LPI provides advantages over OPGD by capturing spatially varying interaction patterns and local effects, whereas OPGD assesses only global interaction effects. For example, the LPI-derived PD for the interaction between total retail sales and the geocomplexity pattern of tertiary-industry output averages 0.610 [0.336,0.783], indicating critical spatial variation in both local PD values and their significance, while the OPGD-derived PD yields a single global estimate of 0.537 (p < 0.01). This research advances theoretical understanding of spatial association and interaction, while providing an innovative analytical tool and decision-support capability for regional development, urban planning, and resource allocation.

空间关联和空间相互作用是理解地理现象和区域发展差异的基础，具有广泛的跨学科适用性。现有的空间异质性分析在捕获模式相互作用和局部变异方面面临重大挑战。本研究开发了一个局部模式相互作用（LPI）模型，该模型集成了空间数据的局部复杂性模式或地理复杂性、模式的相互作用及其局部变化的决定因素（PD）。利用统计数据、遥感图像和开放地理空间数据，对城市化空间分布的局部变量和模式相互作用的PD进行了评估。结果表明，LPI能够有效识别城市化相关解释变量地理复杂性模式相互作用的局部PD。通过与最优参数地理探测器（OPGD）（一种广泛使用的基于空间异质性的PD识别模型）的比较来评估模型的性能。模型验证表明，LPI通过捕获空间变化的相互作用模式和局部效应而优于OPGD，而OPGD仅评估全局相互作用效应。例如，基于lpi的零售总额与第三产业产出地理复杂性模式之间相互作用的PD平均为0.610[0.336,0.783]，表明局部PD值及其重要性存在关键的空间差异，而基于opgd的PD产生的单一全球估计值为0.537 （p < 0.01）。本研究推进了空间关联与相互作用的理论认识，为区域发展、城市规划和资源配置提供了创新的分析工具和决策支持能力。

{"title":"Local effects of pattern interactions in driving urbanization","authors":"Yanfang Sun , Guosheng Wu , Yongze Song , Haiyang Liu , Lin Wang , Zehua Zhang , Jiao Hu","doi":"10.1016/j.jag.2025.105072","DOIUrl":"10.1016/j.jag.2025.105072","url":null,"abstract":"<div><div>Spatial association and spatial interaction are fundamental to understanding geographical phenomena and regional development disparities, with broad applicability across disciplines. Existing spatial heterogeneity analysis face significant challenges in capturing pattern interactions and local variability. This study develops a local pattern interaction (LPI) model that integrates local complexity patterns or geocomplexity of spatial data, the interaction of patterns, and their locally varied power of determinants (PD). LPI is implemented in assessing the PD of local variables and pattern interactions on the spatial distributions of urbanization using statistical data, remote sensing imagery, and open geospatial data. The results show that LPI effectively identifies the local PD of interactions involving the geocomplexity patterns of urbanization-related explanatory variables. Model performance is evaluated by comparison with the optimal-parameters geographical detector (OPGD), a widely used spatial heterogeneity–based PD identification model. The model validation shows that LPI provides advantages over OPGD by capturing spatially varying interaction patterns and local effects, whereas OPGD assesses only global interaction effects. For example, the LPI-derived PD for the interaction between total retail sales and the geocomplexity pattern of tertiary-industry output averages 0.610 [0.336,0.783], indicating critical spatial variation in both local PD values and their significance, while the OPGD-derived PD yields a single global estimate of 0.537 (p < 0.01). This research advances theoretical understanding of spatial association and interaction, while providing an innovative analytical tool and decision-support capability for regional development, urban planning, and resource allocation.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105072"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925515","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

GDiT: A graph-prior-guided diffusion transformer for semantic-controllable remote sensing image synthesis ggdit：一种用于语义可控遥感图像合成的图形优先引导扩散变压器

IF 8.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2026-01-06 DOI: 10.1016/j.jag.2025.105038

Kai Deng , Xiangyun Hu , Yibing Xiong , Aokun Liang , Jiong Xu

Semantic image synthesis (SIS) is essential for remote sensing, particularly in generating high-quality training data for scarce annotated datasets. While existing SIS methods have advanced pixel-wise mappings between semantic maps and images, they often overlook spatial priors, such as relationships between geographic objects (e.g., road-building adjacency), leading to structural inconsistencies in synthesized images. To address this, we propose the graph-prior diffusion transformer (GDiT) for semantically controllable remote sensing image synthesis. We first convert semantic maps into semantic graphs, encoding geographic objects as nodes with structured spatial interactions. To capture spatial and semantic relationships, we propose the Geometric-Semantic Aware Module (GSAM), which integrates CLIP-extracted semantics and geometric attributes for a more context-aware representation. Furthermore, we design the Graph Diffusion Transformer (GDiT) Block, which employs graph-to-image cross-attention to refine spatial structures, ensuring topological coherence and semantic fidelity in synthesized images. Experiments on the landcover and landuse dataset show that GDiT achieves competitive performance by incorporating text prompts to enable multilevel control across global, object and pixel dimensions, generating high-fidelity images while using only 38.9% of the parameters compared to GeoSynth, significantly improving efficiency and accuracy. The code and dataset will be released at https://github.com/whudk/GDiT.

语义图像合成（SIS）对于遥感至关重要，特别是在为稀缺的注释数据集生成高质量的训练数据方面。虽然现有的SIS方法在语义图和图像之间具有先进的逐像素映射，但它们往往忽略了空间先验，例如地理对象之间的关系（例如，道路建筑相邻性），从而导致合成图像中的结构不一致。为了解决这个问题，我们提出了用于语义可控遥感图像合成的图优先扩散转换器（GDiT）。我们首先将语义图转换为语义图，将地理对象编码为具有结构化空间交互的节点。为了捕获空间和语义关系，我们提出了几何语义感知模块（GSAM），该模块集成了clip提取的语义和几何属性，以实现更具上下文感知的表示。此外，我们设计了图形扩散转换器（GDiT）块，该块采用图与图像的交叉关注来细化空间结构，确保合成图像的拓扑一致性和语义保真度。在土地覆盖和土地利用数据集上的实验表明，GDiT通过结合文本提示来实现跨全局、对象和像素维度的多级控制，与GeoSynth相比，仅使用38.9%的参数就能生成高保真图像，显著提高了效率和准确性，从而取得了具有竞争力的性能。代码和数据集将在https://github.com/whudk/GDiT上发布。

{"title":"GDiT: A graph-prior-guided diffusion transformer for semantic-controllable remote sensing image synthesis","authors":"Kai Deng , Xiangyun Hu , Yibing Xiong , Aokun Liang , Jiong Xu","doi":"10.1016/j.jag.2025.105038","DOIUrl":"10.1016/j.jag.2025.105038","url":null,"abstract":"<div><div>Semantic image synthesis (SIS) is essential for remote sensing, particularly in generating high-quality training data for scarce annotated datasets. While existing SIS methods have advanced pixel-wise mappings between semantic maps and images, they often overlook spatial priors, such as relationships between geographic objects (e.g., road-building adjacency), leading to structural inconsistencies in synthesized images. To address this, we propose the graph-prior diffusion transformer (GDiT) for semantically controllable remote sensing image synthesis. We first convert semantic maps into semantic graphs, encoding geographic objects as nodes with structured spatial interactions. To capture spatial and semantic relationships, we propose the Geometric-Semantic Aware Module (GSAM), which integrates CLIP-extracted semantics and geometric attributes for a more context-aware representation. Furthermore, we design the Graph Diffusion Transformer (GDiT) Block, which employs graph-to-image cross-attention to refine spatial structures, ensuring topological coherence and semantic fidelity in synthesized images. Experiments on the landcover and landuse dataset show that GDiT achieves competitive performance by incorporating text prompts to enable multilevel control across global, object and pixel dimensions, generating high-fidelity images while using only 38.9% of the parameters compared to GeoSynth, significantly improving efficiency and accuracy. The code and dataset will be released at <span><span>https://github.com/whudk/GDiT</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105038"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925517","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Cross-platform super-resolution: A diffusion model approach for enhancing satellite imagery with aerial data 跨平台超分辨率：利用航空数据增强卫星图像的扩散模型方法

IF 8.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2026-01-06 DOI: 10.1016/j.jag.2025.105046

Zhe Wang , Carmen Galaz García , Benjamin S. Halpern

Monitoring Earth’s dynamic systems benefits from high spatial and temporal resolution imagery, a combination rarely available from a single sensor. This study addresses the novel challenge of cross-platform super-resolution (SR), aiming to enhance high-frequency satellite data with high-resolution aerial data. We employed a diffusion-based deep learning model, Super-Resolution via Iterative Refinement (SR3), to upscale 3-m Planet satellite imagery to the 60-cm resolution of the National Agriculture Imagery Program (NAIP) images, a fivefold enhancement. The findings reveal that the “domain gap” between aerial and satellite data is a significant obstacle. While the model performed robustly on single-source data (peak signal-to-noise-ratio, or PSNR, of 27.28 dB), its cross-platform performance was substantially lower (best PSNR of 16.85 dB). Interestingly, models trained from scratch consistently outperformed fine-tuned models, suggesting negative transfer due to the differences between the aerial and satellite data sources. Furthermore, environmental metrics like the Normalized Difference Vegetation Index (NDVI) proved to be more effective performance indicators than standard computer vision metrics (such as PSNR) and structural similarity index measure (SSIM), showing better preservation of critical spectral information for vegetation analysis. This work demonstrates both the potential and the distinct challenges of using diffusion models for super-resolution across different remote sensing platforms. Our findings underscore the importance of tailored approaches in super-resolution and provide insights into leveraging state-of-the-art deep learning techniques for ecological monitoring and resource management.

监测地球动态系统得益于高空间和时间分辨率的图像，这是单个传感器很少能获得的组合。该研究解决了跨平台超分辨率（SR）的新挑战，旨在用高分辨率航空数据增强高频卫星数据。我们采用了一种基于扩散的深度学习模型，即通过迭代细化的超分辨率（SR3），将3米行星卫星图像提升到国家农业图像计划（NAIP）图像的60厘米分辨率，提高了5倍。研究结果表明，航空和卫星数据之间的“领域差距”是一个重大障碍。虽然该模型在单源数据（峰值信噪比为27.28 dB）上表现良好，但其跨平台性能明显较低（最佳PSNR为16.85 dB）。有趣的是，从头开始训练的模型始终优于微调模型，这表明由于空中和卫星数据源之间的差异，负迁移。此外，归一化植被指数（NDVI）等环境指标被证明是比标准计算机视觉指标（如PSNR）和结构相似性指数（SSIM）更有效的性能指标，可以更好地保存植被分析的关键光谱信息。这项工作展示了在不同遥感平台上使用超分辨率扩散模型的潜力和独特的挑战。我们的研究结果强调了超分辨率定制方法的重要性，并为利用最先进的深度学习技术进行生态监测和资源管理提供了见解。

{"title":"Cross-platform super-resolution: A diffusion model approach for enhancing satellite imagery with aerial data","authors":"Zhe Wang , Carmen Galaz García , Benjamin S. Halpern","doi":"10.1016/j.jag.2025.105046","DOIUrl":"10.1016/j.jag.2025.105046","url":null,"abstract":"<div><div>Monitoring Earth’s dynamic systems benefits from high spatial and temporal resolution imagery, a combination rarely available from a single sensor. This study addresses the novel challenge of cross-platform super-resolution (SR), aiming to enhance high-frequency satellite data with high-resolution aerial data. We employed a diffusion-based deep learning model, Super-Resolution via Iterative Refinement (SR3), to upscale 3-m Planet satellite imagery to the 60-cm resolution of the National Agriculture Imagery Program (NAIP) images, a fivefold enhancement. The findings reveal that the “domain gap” between aerial and satellite data is a significant obstacle. While the model performed robustly on single-source data (peak signal-to-noise-ratio, or PSNR, of 27.28 dB), its cross-platform performance was substantially lower (best PSNR of 16.85 dB). Interestingly, models trained from scratch consistently outperformed fine-tuned models, suggesting negative transfer due to the differences between the aerial and satellite data sources. Furthermore, environmental metrics like the Normalized Difference Vegetation Index (NDVI) proved to be more effective performance indicators than standard computer vision metrics (such as PSNR) and structural similarity index measure (SSIM), showing better preservation of critical spectral information for vegetation analysis. This work demonstrates both the potential and the distinct challenges of using diffusion models for super-resolution across different remote sensing platforms. Our findings underscore the importance of tailored approaches in super-resolution and provide insights into leveraging state-of-the-art deep learning techniques for ecological monitoring and resource management.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105046"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926084","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multi-Platform geodetic synergy of InSAR, UAV, optical, and HD-ERT constrains kinematic evolution of the Jungong landslide (Yellow River Basin) InSAR、无人机、光学和HD-ERT多平台协同测量对黄河准公滑坡运动演化的约束

IF 8.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2026-01-05 DOI: 10.1016/j.jag.2025.105082

Xiaoyu Liu , Wu Zhu , Yuxin Zhou , Jiewei Zhan , Zhanxi Wei , Jing Wu , Haixing Shang , Chao Du

Following the September 20, 2019 instability event, the Jungong landslide—a large-scale red-bed feature in the upper Yellow River Basin—has exhibited persistent creep, necessitating systematic kinematic analysis to constrain deformation drivers. In this context, we conducted a multidisciplinary approach integrating interferometric synthetic aperture radar (InSAR), unmanned aerial vehicle (UAV) surveys, optical satellite remote sensing, and high-density electrical resistivity tomography (HD-ERT) to investigate its kinematic evolution. Firstly, interferometric processing of SAR imagery from ALOS/PALSAR-1, ALOS/PALSAR-2 and Sentinel-1 systems (March 2007-August 2024) revealed continuous creeping with maximum deformation velocity reaching −129 mm/yr in descending Sentinel-1. Based on morphological and deformation characteristics, the slope was divided into four secondary zones. Through digital image correlation (DIC) of optical images, horizontal displacements exceeding 20 m induced by instability were detected at the front edge of Zone I. The three-dimensional (3D) deformation field was then inverted by combining multi-orbit InSAR observations and a topography-constrained model, revealing significant spatial heterogeneity of displacement characteristics. The maximum velocities in the eastward, northward, and vertical directions were −107, 53, and −71 mm/yr, respectively. Additionally, the internal structure along two profiles was detected using HD-ERT. Finally, a method combining Singular Spectrum Analysis (SSA) and wavelet transform was proposed to quantitatively analyze the temporal relationship between periodic displacements and rainfall. Different zones exhibited varying degrees of correlation with rainfall, with a time lag of approximately 45 days in Zone I. This multidisciplinary approach enhances our understanding of the kinematic behavior of the Jungong landslide, providing critical reference for future hazard assessment.

在2019年9月20日的不稳定事件之后，黄河上游的大型红层特征军公滑坡表现出持续的蠕变，需要系统的运动学分析来约束变形驱动因素。在此背景下，我们采用了多学科方法，结合干涉合成孔径雷达（InSAR）、无人机（UAV）测量、光学卫星遥感和高密度电阻率层析成像（HD-ERT）来研究其运动学演变。首先，对ALOS/PALSAR-1、ALOS/PALSAR-2和Sentinel-1系统（2007年3月- 2024年8月）的SAR图像进行干涉处理，发现Sentinel-1在下降过程中连续爬行，最大变形速度达到- 129 mm/yr。根据边坡的形态和变形特征，将其划分为4个次生带。通过光学图像的数字图像相关（DIC），在i区前缘检测到由失稳引起的超过20 m的水平位移，并结合多轨道InSAR观测和地形约束模型反演三维变形场，发现位移特征具有明显的空间异质性。东、北、垂直方向最大流速分别为- 107、53和- 71 mm/yr。此外，利用HD-ERT检测了沿两条剖面的内部结构。最后，提出了一种结合奇异谱分析（SSA）和小波变换的方法来定量分析周期性位移与降水的时间关系。不同区域与降雨的相关程度不同，i区滞后时间约为45 天。该多学科方法增强了我们对君公滑坡运动学行为的理解，为未来的灾害评估提供了重要参考。

{"title":"Multi-Platform geodetic synergy of InSAR, UAV, optical, and HD-ERT constrains kinematic evolution of the Jungong landslide (Yellow River Basin)","authors":"Xiaoyu Liu , Wu Zhu , Yuxin Zhou , Jiewei Zhan , Zhanxi Wei , Jing Wu , Haixing Shang , Chao Du","doi":"10.1016/j.jag.2025.105082","DOIUrl":"10.1016/j.jag.2025.105082","url":null,"abstract":"<div><div>Following the September 20, 2019 instability event, the Jungong landslide—a large-scale red-bed feature in the upper Yellow River Basin—has exhibited persistent creep, necessitating systematic kinematic analysis to constrain deformation drivers. In this context, we conducted a multidisciplinary approach integrating interferometric synthetic aperture radar (InSAR), unmanned aerial vehicle (UAV) surveys, optical satellite remote sensing, and high-density electrical resistivity tomography (HD-ERT) to investigate its kinematic evolution. Firstly, interferometric processing of SAR imagery from ALOS/PALSAR-1, ALOS/PALSAR-2 and Sentinel-1 systems (March 2007-August 2024) revealed continuous creeping with maximum deformation velocity reaching −129 mm/yr in descending Sentinel-1. Based on morphological and deformation characteristics, the slope was divided into four secondary zones. Through digital image correlation (DIC) of optical images, horizontal displacements exceeding 20 m induced by instability were detected at the front edge of Zone I. The three-dimensional (3D) deformation field was then inverted by combining multi-orbit InSAR observations and a topography-constrained model, revealing significant spatial heterogeneity of displacement characteristics. The maximum velocities in the eastward, northward, and vertical directions were −107, 53, and −71 mm/yr, respectively. Additionally, the internal structure along two profiles was detected using HD-ERT. Finally, a method combining Singular Spectrum Analysis (SSA) and wavelet transform was proposed to quantitatively analyze the temporal relationship between periodic displacements and rainfall. Different zones exhibited varying degrees of correlation with rainfall, with a time lag of approximately 45 days in Zone I. This multidisciplinary approach enhances our understanding of the kinematic behavior of the Jungong landslide, providing critical reference for future hazard assessment.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105082"},"PeriodicalIF":8.6,"publicationDate":"2026-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145919767","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

DA-MiTUNet: A Mix Transformer with dual attention embedding in unet for Land-Sea segmentation of remote sensing images DA-MiTUNet：用于遥感图像陆海分割的双注意力嵌入混合变压器

IF 8.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2026-01-05 DOI: 10.1016/j.jag.2025.105066

Jiawei Wu , Zijian Liu , Qixiang Tong , Zhipeng Zhu , Hui He , Xinghui Wu , Haihua Xing

Automatic extraction of coastlines from remote sensing images is of great practical importance for coastal risk assessment, ecological environmental protection, and marine economic development. However, the highly dynamic nature of coastlines and the complex, diverse characteristics of land–sea boundaries make precise coastline extraction a challenging task. Although traditional deep learning methods have demonstrated good performance in this respect, they still face numerous shortcomings when dealing with high computational costs and the need to fully utilize multiscale features. In this paper, to address these problems, we propose a novel and efficient land–sea segmentation model for remote sensing imagery based on a classical U-shaped network structure, named DA-MiTUNet. On the one hand, we introduce the convolutional block attention module into the Mix Transformer (MiT), forming a dual-attention encoder in conjunction with an efficient self-attention mechanism. This integration ensures comprehensive extraction of global context and local information, thereby enabling more precise determination of complex land–sea boundary features. On the other hand, we propose an adaptive feature fusion module to further promote the effective fusion of features across different hierarchical levels, achieving more refined land–sea boundary segmentation. Experimental results on the Gaofen-1 Hainan Coastline Dataset (GF–HNCD) and Benchmark Sea–Land Dataset (BSD) datasets demonstrated that the proposed DA-MiTUNet model outperforms other comparative models in terms of both the average F1 score and the mean Intersection over Union value, while achieving excellent segmentation results with relatively low computational complexity, thereby reflecting the potential of our model for dynamic coastal monitoring during extreme sea level events.

海岸带遥感影像自动提取对海岸带风险评估、生态环境保护和海洋经济发展具有重要的现实意义。然而，海岸线的高度动态性质和陆海边界的复杂多样特征使得精确的海岸线提取成为一项具有挑战性的任务。尽管传统的深度学习方法在这方面表现良好，但在处理高计算成本和需要充分利用多尺度特征时，它们仍然面临许多缺点。为了解决这些问题，本文提出了一种基于经典u型网络结构的新型高效遥感影像陆海分割模型DA-MiTUNet。一方面，我们将卷积块注意模块引入到Mix Transformer （MiT）中，结合高效的自注意机制形成双注意编码器。这种整合确保了对全球背景和局部信息的全面提取，从而能够更精确地确定复杂的陆海边界特征。另一方面，提出自适应特征融合模块，进一步促进不同层次特征的有效融合，实现更精细的陆海边界分割。在高分一号海南海岸线数据集（GF-HNCD）和基准海-地数据集（BSD）数据集上的实验结果表明，所提出的DA-MiTUNet模型在平均F1得分和平均Intersection over Union值方面都优于其他比较模型，同时在较低的计算复杂度下获得了良好的分割结果，从而体现了我们的模型在极端海平面事件下动态海岸监测的潜力。

{"title":"DA-MiTUNet: A Mix Transformer with dual attention embedding in unet for Land-Sea segmentation of remote sensing images","authors":"Jiawei Wu , Zijian Liu , Qixiang Tong , Zhipeng Zhu , Hui He , Xinghui Wu , Haihua Xing","doi":"10.1016/j.jag.2025.105066","DOIUrl":"10.1016/j.jag.2025.105066","url":null,"abstract":"<div><div>Automatic extraction of coastlines from remote sensing images is of great practical importance for coastal risk assessment, ecological environmental protection, and marine economic development. However, the highly dynamic nature of coastlines and the complex, diverse characteristics of land–sea boundaries make precise coastline extraction a challenging task. Although traditional deep learning methods have demonstrated good performance in this respect, they still face numerous shortcomings when dealing with high computational costs and the need to fully utilize multiscale features. In this paper, to address these problems, we propose a novel and efficient land–sea segmentation model for remote sensing imagery based on a classical U-shaped network structure, named DA-MiTUNet. On the one hand, we introduce the convolutional block attention module into the Mix Transformer (MiT), forming a dual-attention encoder in conjunction with an efficient self-attention mechanism. This integration ensures comprehensive extraction of global context and local information, thereby enabling more precise determination of complex land–sea boundary features. On the other hand, we propose an adaptive feature fusion module to further promote the effective fusion of features across different hierarchical levels, achieving more refined land–sea boundary segmentation. Experimental results on the Gaofen-1 Hainan Coastline Dataset (GF–HNCD) and Benchmark Sea–Land Dataset (BSD) datasets demonstrated that the proposed DA-MiTUNet model outperforms other comparative models in terms of both the average F1 score and the mean Intersection over Union value, while achieving excellent segmentation results with relatively low computational complexity, thereby reflecting the potential of our model for dynamic coastal monitoring during extreme sea level events.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105066"},"PeriodicalIF":8.6,"publicationDate":"2026-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925924","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

WECMD: A multisensor dataset for wearable event cameras in the age of embodied intelligence WECMD：具身智能时代可穿戴事件相机的多传感器数据集

IF 8.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-12-19 DOI: 10.1016/j.jag.2025.105019

Xueli Guo , Zhichao Wen , Jianzhu Huai , Jian Kuang , Yizhou Xue , Xuanxuan Zhang , Bisheng Yang , You Li

In the era of embodied intelligence, wearable datasets for pedestrian navigation are essential. However, publicly available multisensory datasets tailored to such scenarios remain scarce. Traditional sensors such as RGB cameras and LiDAR often struggle to capture the fast and irregular dynamics of human motion. To address this gap, we introduce a large-scale pedestrian-wearable dataset primarily recorded using event cameras. The dataset includes event camera, RGB cameras, LiDAR sensors, a tactical-grade IMU, and GNSS, covering a wide range of indoor and outdoor environments with diverse motion types and illumination conditions. High-precision ground truth is obtained using motion capture systems indoors and GNSS/IMU integration with bidirectional smoothing outdoors. The dataset is structured into 23 subsets categorized by motion dynamics and lighting, supporting the development and evaluation of robust localization and SLAM algorithms. Benchmarking with state-of-the-art frameworks reveals notable performance degradation under highly dynamic or low-light conditions, highlighting the dataset’s value for advancing pedestrian navigation and event-based perception. The dataset and tools are publicly available at: https://github.com/xueli-guo/WECMD.git.

在具身智能时代，用于行人导航的可穿戴数据集至关重要。然而，针对这些场景的公开可用的多感官数据集仍然很少。传统的传感器，如RGB相机和激光雷达，往往难以捕捉人体运动的快速和不规则动态。为了解决这一差距，我们引入了一个主要使用事件相机记录的大规模行人可穿戴数据集。该数据集包括事件相机、RGB相机、LiDAR传感器、战术级IMU和GNSS，涵盖了广泛的室内和室外环境，具有不同的运动类型和照明条件。室内采用运动捕捉系统，室外采用双向平滑GNSS/IMU集成系统，获得高精度的地面真值。该数据集分为23个子集，按运动动力学和光照分类，支持鲁棒定位和SLAM算法的开发和评估。使用最先进的框架进行基准测试显示，在高动态或低光照条件下，性能显著下降，突出了数据集在推进行人导航和基于事件的感知方面的价值。数据集和工具可在：https://github.com/xueli-guo/WECMD.git上公开获取。

{"title":"WECMD: A multisensor dataset for wearable event cameras in the age of embodied intelligence","authors":"Xueli Guo , Zhichao Wen , Jianzhu Huai , Jian Kuang , Yizhou Xue , Xuanxuan Zhang , Bisheng Yang , You Li","doi":"10.1016/j.jag.2025.105019","DOIUrl":"10.1016/j.jag.2025.105019","url":null,"abstract":"<div><div>In the era of embodied intelligence, wearable datasets for pedestrian navigation are essential. However, publicly available multisensory datasets tailored to such scenarios remain scarce. Traditional sensors such as RGB cameras and LiDAR often struggle to capture the fast and irregular dynamics of human motion. To address this gap, we introduce a large-scale pedestrian-wearable dataset primarily recorded using event cameras. The dataset includes event camera, RGB cameras, LiDAR sensors, a tactical-grade IMU, and GNSS, covering a wide range of indoor and outdoor environments with diverse motion types and illumination conditions. High-precision ground truth is obtained using motion capture systems indoors and GNSS/IMU integration with bidirectional smoothing outdoors. The dataset is structured into 23 subsets categorized by motion dynamics and lighting, supporting the development and evaluation of robust localization and SLAM algorithms. Benchmarking with state-of-the-art frameworks reveals notable performance degradation under highly dynamic or low-light conditions, highlighting the dataset’s value for advancing pedestrian navigation and event-based perception. The dataset and tools are publicly available at: <span><span>https://github.com/xueli-guo/WECMD.git</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105019"},"PeriodicalIF":8.6,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145791194","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A fully automatic and label-free Sentinel-1 SAR framework for green-tide mapping 用于绿潮制图的全自动无标签Sentinel-1 SAR框架

IF 8.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-12-19 DOI: 10.1016/j.jag.2025.105036

Pengfei Tang , Peijun Du , Shanchuan Guo , Lu Qie , Wei Zhang , Peng Zhang , Mathias Réus , Jocelyn Chanussot

Green tides in the Yellow Sea are recurrent hazardous algal blooms whose optical monitoring is often hindered by cloud cover, while existing SAR approaches remain sensitive to sea state and look-alike targets and frequently require per-scene tuning or curated labels, limiting transferability and temporal consistency. To address this, we develop a label-free, fully automatic Sentinel-1 workflow that operationalizes three empirical signatures of green tides: spatial anomaly pre-location using local standard deviation on VV, edge-guided intensity separation via edge-balanced Otsu, and temporal anomaly screening using Z-scores with adaptive thresholding; an automatic object-level filter then removes non-algal marine targets. Implemented on Google Earth Engine at 10 m resolution, the pipeline delivers rapid processing without manual parameters. Validation shows high mapping accuracy: with a global stratified sample set, F1 equals 0.96 in 2019 and 0.97 in 2021; with a local edge validation set, F1 equals 0.94; in an all-pixel assessment over more than 2.5 billion pixels against a baseline, overall F1 equals 0.91. Qualitative comparisons likewise show fewer omissions of low-contrast filaments and fewer perforations within mats than GA-Net and UDNet. An ablation analysis clarifies the role of each module: the spatial pre-locator supplies contiguous candidates; the edge-guided intensity module sharpens boundaries and limits leakage; the temporal module suppresses transient bright seawater and consolidates persistent mats. Used jointly, the three constraints provide complementary information that yields the most stable cross-year performance and a favorable balance between precision and recall. Overall, the framework offers a simple, scalable, and operational pathway for fine-scale, all-weather monitoring and consistent multi-year assessment of green tides in the Yellow Sea.

黄海的绿潮是反复出现的有害藻华，其光学监测经常受到云层覆盖的阻碍，而现有的SAR方法对海况和相似目标仍然敏感，并且经常需要对每个场景进行调整或管理标签，限制了可转移性和时间一致性。为了解决这个问题，我们开发了一个无标签的全自动Sentinel-1工作流，该工作流可操作绿潮的三个经验特征：使用VV的局部标准差进行空间异常预定位，使用边缘平衡Otsu进行边缘引导强度分离，以及使用自适应阈值的z分数进行时间异常筛选；自动对象级过滤器然后删除非藻类海洋目标。该管道在谷歌Earth Engine上实现，分辨率为10米，无需手动参数即可进行快速处理。验证表明，在全局分层样本集上，F1在2019年等于0.96，在2021年等于0.97；对于局部边缘验证集，F1 = 0.94；在超过25亿像素的全像素评估中，总体F1等于0.91。定性比较同样表明，与GA-Net和UDNet相比，低对比度细丝的遗漏和垫内穿孔较少。消融分析阐明了每个模块的作用：空间预定位器提供连续候选；边缘引导强度模块锐化边界，限制泄漏；时间模块抑制了短暂的明亮海水并巩固了持久的垫。联合使用，这三个约束提供了互补的信息，产生最稳定的跨年性能，并在精度和召回率之间取得了有利的平衡。总体而言，该框架为黄海绿潮的精细、全天候监测和持续多年评估提供了一个简单、可扩展和可操作的途径。

{"title":"A fully automatic and label-free Sentinel-1 SAR framework for green-tide mapping","authors":"Pengfei Tang , Peijun Du , Shanchuan Guo , Lu Qie , Wei Zhang , Peng Zhang , Mathias Réus , Jocelyn Chanussot","doi":"10.1016/j.jag.2025.105036","DOIUrl":"10.1016/j.jag.2025.105036","url":null,"abstract":"<div><div>Green tides in the Yellow Sea are recurrent hazardous algal blooms whose optical monitoring is often hindered by cloud cover, while existing SAR approaches remain sensitive to sea state and look-alike targets and frequently require per-scene tuning or curated labels, limiting transferability and temporal consistency. To address this, we develop a label-free, fully automatic Sentinel-1 workflow that operationalizes three empirical signatures of green tides: spatial anomaly pre-location using local standard deviation on VV, edge-guided intensity separation via edge-balanced Otsu, and temporal anomaly screening using Z-scores with adaptive thresholding; an automatic object-level filter then removes non-algal marine targets. Implemented on Google Earth Engine at 10 m resolution, the pipeline delivers rapid processing without manual parameters. Validation shows high mapping accuracy: with a global stratified sample set, F1 equals 0.96 in 2019 and 0.97 in 2021; with a local edge validation set, F1 equals 0.94; in an all-pixel assessment over more than 2.5 billion pixels against a baseline, overall F1 equals 0.91. Qualitative comparisons likewise show fewer omissions of low-contrast filaments and fewer perforations within mats than GA-Net and UDNet. An ablation analysis clarifies the role of each module: the spatial pre-locator supplies contiguous candidates; the edge-guided intensity module sharpens boundaries and limits leakage; the temporal module suppresses transient bright seawater and consolidates persistent mats. Used jointly, the three constraints provide complementary information that yields the most stable cross-year performance and a favorable balance between precision and recall. Overall, the framework offers a simple, scalable, and operational pathway for fine-scale, all-weather monitoring and consistent multi-year assessment of green tides in the Yellow Sea.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105036"},"PeriodicalIF":8.6,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145791192","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Desertification expansion significantly suppresses photosynthetic peak capacity of arid ecosystems at the global scale 在全球尺度上，沙漠化扩张显著抑制了干旱生态系统光合峰值容量

IF 8.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-12-18 DOI: 10.1016/j.jag.2025.105042

Kaiyang Qiu , Qingbin Zhang , Yingzhong Xie , Mingjie Shi , Chengyun Wang , Tong Dong , Jun Ma , Panxing He

Arid ecosystems occupy about two-fifths of the global land surface, and fluctuations in their productivity play a pivotal role in global carbon sequestration and ecosystem service provision. However, the global-scale effect of desertification expansion on the annual maximum photosynthetic peak has not yet been systematically quantified. In this study, 30-m high-resolution desert cover data (GLCLUC) and multi-source remote-sensing photosynthetic indicators were integrated, using a space-for-time substitution framework to establish a global desertification scenario classification system. We quantitatively evaluated the influence of diverse desert expansion and contraction scenarios on the ecosystem photosynthetic peak (GPP_max). Results indicate that the average GPP_max in high-intensity expansion regions (HIEs) is 8.23 g C m⁻² 8d⁻¹, whereas medium- to low-intensity expansion regions (MIEs) show a value of 8.95 g C m⁻² 8d⁻¹. By contrast, medium- to low-intensity contraction regions (MIRs) and high-intensity contraction regions (HIRs) demonstrate markedly higher GPP_max values of 10.64 g C m⁻² 8d⁻¹ and 17.64 g C m⁻² 8d⁻¹, respectively. Regarding the photosynthetic peak difference (ΔGPP_max), expansion scenarios (HIEs, MIEs) significantly decrease ecosystem photosynthetic potential, with average ΔGPP_max reductions of 1.19–3.95 g C m⁻² 8d⁻¹ relative to contraction scenarios (HIRs, MIRs). The most pronounced losses occur in South America, North America, and Eurasia, with South America exhibiting reductions exceeding 6 g C m⁻² 8d⁻¹. Additionally, ecosystems with initially higher photosynthetic potential experience greater GPP_max declines under intense desert expansion. This study provides the first global-scale evidence revealing how different desertification pathways modify ecosystem photosynthetic peaks and their regional disparities, offering critical scientific support for ecological restoration, carbon sequestration strategies, and land management across arid landscapes.

干旱生态系统约占全球陆地面积的五分之二，其生产力的波动在全球固碳和提供生态系统服务方面发挥着关键作用。然而，在全球尺度上，沙漠化扩张对年最大光合峰值的影响尚未得到系统的量化。本研究将30 m高分辨率沙漠覆盖数据（GLCLUC）与多源遥感光合指标相结合，采用时空替代框架建立全球沙漠化情景分类体系。定量评价了不同荒漠扩张收缩情景对生态系统光合峰值（GPPmax）的影响。结果表明，高强度膨胀区（HIEs）的平均GPPmax为8.23 g C m−2 8d−1，中低强度膨胀区（MIEs）的平均GPPmax为8.95 g C m−2 8d−1。相比之下，中低强度收缩区（MIRs）和高强度收缩区（HIRs）的GPPmax值分别为10.64 g C m−2 8d−1和17.64 g C m−2 8d−1。在光合峰值差（ΔGPPmax）方面，扩张情景（HIEs, MIEs）显著降低了生态系统光合潜力，相对于收缩情景（HIRs, MIRs），平均ΔGPPmax降低1.19-3.95 g C m−2 8d−1。最显著的损失发生在南美洲、北美洲和欧亚大陆，南美洲的减少量超过6 g C m−28d−1。此外，具有较高光合潜力的生态系统在剧烈的沙漠扩张下会经历更大的GPPmax下降。该研究首次提供了全球尺度的证据，揭示了不同沙漠化途径如何改变生态系统光合峰值及其区域差异，为干旱景观的生态恢复、碳固存策略和土地管理提供了重要的科学支持。

{"title":"Desertification expansion significantly suppresses photosynthetic peak capacity of arid ecosystems at the global scale","authors":"Kaiyang Qiu , Qingbin Zhang , Yingzhong Xie , Mingjie Shi , Chengyun Wang , Tong Dong , Jun Ma , Panxing He","doi":"10.1016/j.jag.2025.105042","DOIUrl":"10.1016/j.jag.2025.105042","url":null,"abstract":"<div><div>Arid ecosystems occupy about two-fifths of the global land surface, and fluctuations in their productivity play a pivotal role in global carbon sequestration and ecosystem service provision. However, the global-scale effect of desertification expansion on the annual maximum photosynthetic peak has not yet been systematically quantified. In this study, 30-m high-resolution desert cover data (GLCLUC) and multi-source remote-sensing photosynthetic indicators were integrated, using a space-for-time substitution framework to establish a global desertification scenario classification system. We quantitatively evaluated the influence of diverse desert expansion and contraction scenarios on the ecosystem photosynthetic peak (GPP<sub>max</sub>). Results indicate that the average GPP<sub>max</sub> in high-intensity expansion regions (HIEs) is 8.23 g C m<sup>−2</sup> 8d<sup>−1</sup>, whereas medium- to low-intensity expansion regions (MIEs) show a value of 8.95 g C m<sup>−2</sup> 8d<sup>−1</sup>. By contrast, medium- to low-intensity contraction regions (MIRs) and high-intensity contraction regions (HIRs) demonstrate markedly higher GPP<sub>max</sub> values of 10.64 g C m<sup>−2</sup> 8d<sup>−1</sup> and 17.64 g C m<sup>−2</sup> 8d<sup>−1</sup>, respectively. Regarding the photosynthetic peak difference (ΔGPP<sub>max</sub>), expansion scenarios (HIEs, MIEs) significantly decrease ecosystem photosynthetic potential, with average ΔGPP<sub>max</sub> reductions of 1.19–3.95 g C m<sup>−2</sup> 8d<sup>−1</sup> relative to contraction scenarios (HIRs, MIRs). The most pronounced losses occur in South America, North America, and Eurasia, with South America exhibiting reductions exceeding 6 g C m<sup>−2</sup> 8d<sup>−1</sup>. Additionally, ecosystems with initially higher photosynthetic potential experience greater GPP<sub>max</sub> declines under intense desert expansion. This study provides the first global-scale evidence revealing how different desertification pathways modify ecosystem photosynthetic peaks and their regional disparities, offering critical scientific support for ecological restoration, carbon sequestration strategies, and land management across arid landscapes.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105042"},"PeriodicalIF":8.6,"publicationDate":"2025-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145791193","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0