International journal of applied earth observation and geoinformation : ITC journal最新文献_第6页

MapMate: A framework bridging natural language interaction and map design through large language models MapMate：一个通过大型语言模型连接自然语言交互和地图设计的框架

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.2025.105073

Zihao Tang , Songshan Yue , Fangzhuo Mu , Yucheng Shu , Zhuo Sun , Yongning Wen

Map design is fundamental to geographic information communication and applications, yet it remains an expertise-intensive task involving complex operational procedures that limit accessibility for diverse users despite the proliferation of digital platforms. This study introduces MapMate, a large language model (LLM)-based framework that enables map design through natural language interaction. MapMate addresses the gap between modification goals expressed in natural language and technical map design configurations by integrating a hierarchical map design knowledge base with platform-specific specifications. The framework comprises four core components: a request validator that ensures cartographic parameter validity and operational feasibility, a map design task planner that decomposes goal-oriented requirements into executable operations, a context information retriever that maintains project coherence, and a map design tool router that orchestrates map design functions. To overcome LLM memory limitations in multi-round interactions, MapMate implements a persistence strategy that maintains records of operational history and map-wide design states. Three case studies validate the framework’s effectiveness across single-layer refinement, cross-layer design refinement, and context-aware map design. Results demonstrate that MapMate successfully bridges natural language interaction and map design, providing a human-AI collaborative environment that reduces technical barriers while maintaining cartographic integrity. This framework represents a promising advancement toward the development of intelligent, accessible map design systems for integration with AI-enhanced GIS applications.

地图设计是地理信息交流和应用的基础，但它仍然是一项专业知识密集型任务，涉及复杂的操作程序，尽管数字平台激增，但它限制了不同用户的可访问性。本研究介绍了MapMate，一个基于大型语言模型（LLM）的框架，通过自然语言交互实现地图设计。MapMate通过将分层地图设计知识库与特定于平台的规范集成在一起，解决了用自然语言表达的修改目标与技术地图设计配置之间的差距。该框架包括四个核心组件：确保地图参数有效性和操作可行性的请求验证器，将面向目标的需求分解为可执行操作的地图设计任务规划器，维护项目一致性的上下文信息检索器，以及编排地图设计功能的地图设计工具路由器。为了克服LLM在多轮交互中的内存限制，MapMate实现了一种持久化策略，该策略维护了操作历史记录和地图范围内的设计状态。三个案例研究验证了该框架在单层细化、跨层设计细化和上下文感知地图设计方面的有效性。结果表明，MapMate成功地在自然语言交互和地图设计之间架起了桥梁，提供了一个人类-人工智能协作环境，在保持地图完整性的同时减少了技术障碍。该框架代表了智能、可访问的地图设计系统与人工智能增强的GIS应用程序集成的一个有希望的进步。

{"title":"MapMate: A framework bridging natural language interaction and map design through large language models","authors":"Zihao Tang , Songshan Yue , Fangzhuo Mu , Yucheng Shu , Zhuo Sun , Yongning Wen","doi":"10.1016/j.jag.2025.105073","DOIUrl":"10.1016/j.jag.2025.105073","url":null,"abstract":"<div><div>Map design is fundamental to geographic information communication and applications, yet it remains an expertise-intensive task involving complex operational procedures that limit accessibility for diverse users despite the proliferation of digital platforms. This study introduces MapMate, a large language model (LLM)-based framework that enables map design through natural language interaction. MapMate addresses the gap between modification goals expressed in natural language and technical map design configurations by integrating a hierarchical map design knowledge base with platform-specific specifications. The framework comprises four core components: a request validator that ensures cartographic parameter validity and operational feasibility, a map design task planner that decomposes goal-oriented requirements into executable operations, a context information retriever that maintains project coherence, and a map design tool router that orchestrates map design functions. To overcome LLM memory limitations in multi-round interactions, MapMate implements a persistence strategy that maintains records of operational history and map-wide design states. Three case studies validate the framework’s effectiveness across single-layer refinement, cross-layer design refinement, and context-aware map design. Results demonstrate that MapMate successfully bridges natural language interaction and map design, providing a human-AI collaborative environment that reduces technical barriers while maintaining cartographic integrity. This framework represents a promising advancement toward the development of intelligent, accessible map design systems for integration with AI-enhanced GIS applications.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105073"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926330","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

BuildingMultiView:powering multi-scale building characterization with large language models and Multi-perspective imagery BuildingMultiView：通过大型语言模型和多视角图像为多尺度建筑特征提供动力

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.2025.105034

Zongrong Li , Yunlei Su , Filip Biljecki , Wufan Zhao

Buildings play a crucial role in shaping urban environments, influencing their physical, functional, and aesthetic characteristics. However, urban analytics is frequently limited by datasets lacking essential semantic details as well as fragmentation across diverse and incompatible data sources. To address these challenges, we conducted a comprehensive meta-analysis of 6,285 publications (2019–2024). From this review, we identified 11 key visually discernible building characteristics grouped into three branches: satellite house, satellite neighborhood, and street-view. Based on this structured characteristic system, we introduce BuildingMultiView, an innovative framework leveraging fine-tuned Large Language Models (LLMs) to systematically extract semantically detailed building characteristics from integrated satellite and street-view imagery. Using structured image–prompt–label triplets, the model efficiently annotates characteristics at multiple spatial scales. These characteristics include swimming pools, roof types, building density, wall–window ratio, and property types. Together, they provide a comprehensive and multi-perspective building database. Experiments conducted across five cities in the USA with diverse architecture and urban form, San Francisco, San Diego, Salt Lake City, Austin, and New York City, demonstrate significant performance improvements, with an F1 score of 79.77% compared to the untuned base version of ChatGPT’s 45.66%. These results reveal diverse urban building patterns and correlations between architectural and environmental characteristics, showcasing the framework’s capability to analyze both macro-scale and micro-scale urban building data. By integrating multi-perspective data sources with cutting-edge LLMs, BuildingMultiView enhances building data extraction, offering a scalable tool for urban planners to address sustainability, infrastructure, and human-centered design, enabling smarter, resilient cities.

建筑在塑造城市环境方面发挥着至关重要的作用，影响着城市的物理、功能和美学特征。然而，城市分析经常受到缺乏基本语义细节的数据集以及不同和不兼容的数据源的碎片化的限制。为了应对这些挑战，我们对6285篇出版物（2019-2024年）进行了全面的荟萃分析。从这篇综述中，我们确定了11个关键的视觉上可识别的建筑特征，分为三个分支：卫星住宅、卫星社区和街景。在此结构化特征系统的基础上，我们引入了BuildingMultiView，这是一个利用微调大语言模型（llm）的创新框架，可以从综合卫星和街景图像中系统地提取语义详细的建筑特征。该模型采用结构化的图像提示标签三元组，在多个空间尺度上有效标注特征。这些特征包括游泳池、屋顶类型、建筑密度、墙窗比和财产类型。它们共同提供了一个全面和多角度的建筑数据库。在美国五个建筑和城市形态各异的城市（旧金山、圣地亚哥、盐湖城、奥斯汀和纽约市）进行的实验表明，性能得到了显著改善，F1得分为79.77%，而ChatGPT未调优的基本版本的得分为45.66%。这些结果揭示了不同的城市建筑模式以及建筑与环境特征之间的相关性，展示了该框架分析宏观尺度和微观尺度城市建筑数据的能力。通过将多角度数据源与先进的法学硕士相结合，BuildingMultiView增强了建筑数据提取，为城市规划者提供了可扩展的工具，以解决可持续性、基础设施和以人为本的设计问题，从而实现更智能、更有弹性的城市。

{"title":"BuildingMultiView:powering multi-scale building characterization with large language models and Multi-perspective imagery","authors":"Zongrong Li , Yunlei Su , Filip Biljecki , Wufan Zhao","doi":"10.1016/j.jag.2025.105034","DOIUrl":"10.1016/j.jag.2025.105034","url":null,"abstract":"<div><div>Buildings play a crucial role in shaping urban environments, influencing their physical, functional, and aesthetic characteristics. However, urban analytics is frequently limited by datasets lacking essential semantic details as well as fragmentation across diverse and incompatible data sources. To address these challenges, we conducted a comprehensive <em>meta</em>-analysis of 6,285 publications (2019–2024). From this review, we identified 11 key visually discernible building characteristics grouped into three branches: satellite house, satellite neighborhood, and street-view. Based on this structured characteristic system, we introduce BuildingMultiView, an innovative framework leveraging fine-tuned Large Language Models (LLMs) to systematically extract semantically detailed building characteristics from integrated satellite and street-view imagery. Using structured image–prompt–label triplets, the model efficiently annotates characteristics at multiple spatial scales. These characteristics include swimming pools, roof types, building density, wall–window ratio, and property types. Together, they provide a comprehensive and multi-perspective building database. Experiments conducted across five cities in the USA with diverse architecture and urban form, San Francisco, San Diego, Salt Lake City, Austin, and New York City, demonstrate significant performance improvements, with an F1 score of 79.77% compared to the untuned base version of ChatGPT’s 45.66%. These results reveal diverse urban building patterns and correlations between architectural and environmental characteristics, showcasing the framework’s capability to analyze both macro-scale and micro-scale urban building data. By integrating multi-perspective data sources with cutting-edge LLMs, BuildingMultiView enhances building data extraction, offering a scalable tool for urban planners to address sustainability, infrastructure, and human-centered design, enabling smarter, resilient cities.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105034"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926448","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

Canopy height mapping in the Western Himalayas, Pakistan: A deep learning approach using GEDI and Sentinel-2 fusion 巴基斯坦西喜马拉雅地区的冠层高度测绘：使用GEDI和Sentinel-2融合的深度学习方法

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.2025.105030

Adeel Ahmad , Srikumar Sastry , Aayush Dhakal , Subash Khanal , Alex Levering , Hammad Gilani , Nathan Jacobs

The western Himalayas in Pakistan, characterized by a diverse range of conifer species at higher elevations, represent a critical biodiversity hotspot and habitat for numerous species. Accurate spatial assessments of canopy height are essential for improving estimates of aboveground biomass, carbon sequestration, and associated forest ecosystem services in this region. In this study, we estimated canopy heights in the western Himalayas in Pakistan using a U-Net CNN approach, fusing data from the Global Ecosystem Dynamics Investigation Mission (GEDI) with multi-band Sentinel-2 (S2) imagery. We produced a canopy height map at a 10 m resolution for 2020. To ensure accurate measurements across various canopy height groups, we implemented a stratified training approach that optimized the representation of GEDI data throughout the training, validation, and testing phases. We trained multiple models using varying thresholds and assigned different weights to taller trees to improve accuracy between different canopy height groups in the study region. Our best model achieved a Root Mean Square Error (RMSE) of 7.52 m and a Mean Absolute Error (MAE) of 5.71 m in the test set, significantly outperforming existing global canopy height models in this topographically complex region. We further validate our predictions against field inventory plots, achieving a coefficient of determination (R

^{2}

) of 0.49 for plots containing at least 15 trees. The resulting tree canopy height map, designated as the Western Himalaya Canopy Height Map (WHiCH Map), is publicly available.

巴基斯坦西喜马拉雅地区海拔较高，针叶树种类繁多，是一个重要的生物多样性热点和众多物种的栖息地。准确的冠层高度空间评估对于改善该地区地上生物量、碳固存和相关森林生态系统服务的估算至关重要。在这项研究中，我们使用U-Net CNN方法，融合了来自全球生态系统动力学调查任务（GEDI）的数据和多波段Sentinel-2 （S2）图像，估计了巴基斯坦西喜马拉雅地区的冠层高度。我们制作了一张2020年10米分辨率的冠层高度图。为了确保各种冠层高度组的准确测量，我们实施了一种分层训练方法，在整个训练、验证和测试阶段优化GEDI数据的表示。我们使用不同的阈值训练多个模型，并为较高的树木分配不同的权重，以提高研究区域不同冠层高度组之间的准确性。我们的最佳模型在测试集中实现了7.52 m的均方根误差（RMSE）和5.71 m的平均绝对误差（MAE），显著优于该地形复杂地区现有的全球冠层高度模型。我们进一步验证了我们的预测与实地调查样地的关系，对于至少包含15棵树的样地，其决定系数（R2）为0.49。生成的树冠高度图，被命名为西喜马拉雅树冠高度图（哪个地图），是公开的。

{"title":"Canopy height mapping in the Western Himalayas, Pakistan: A deep learning approach using GEDI and Sentinel-2 fusion","authors":"Adeel Ahmad , Srikumar Sastry , Aayush Dhakal , Subash Khanal , Alex Levering , Hammad Gilani , Nathan Jacobs","doi":"10.1016/j.jag.2025.105030","DOIUrl":"10.1016/j.jag.2025.105030","url":null,"abstract":"<div><div>The western Himalayas in Pakistan, characterized by a diverse range of conifer species at higher elevations, represent a critical biodiversity hotspot and habitat for numerous species. Accurate spatial assessments of canopy height are essential for improving estimates of aboveground biomass, carbon sequestration, and associated forest ecosystem services in this region. In this study, we estimated canopy heights in the western Himalayas in Pakistan using a U-Net CNN approach, fusing data from the Global Ecosystem Dynamics Investigation Mission (GEDI) with multi-band Sentinel-2 (S2) imagery. We produced a canopy height map at a 10 m resolution for 2020. To ensure accurate measurements across various canopy height groups, we implemented a stratified training approach that optimized the representation of GEDI data throughout the training, validation, and testing phases. We trained multiple models using varying thresholds and assigned different weights to taller trees to improve accuracy between different canopy height groups in the study region. Our best model achieved a Root Mean Square Error (RMSE) of 7.52 m and a Mean Absolute Error (MAE) of 5.71 m in the test set, significantly outperforming existing global canopy height models in this topographically complex region. We further validate our predictions against field inventory plots, achieving a coefficient of determination (R<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span>) of 0.49 for plots containing at least 15 trees. The resulting tree canopy height map, designated as the Western Himalaya Canopy Height Map (WHiCH Map), is publicly available.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105030"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925513","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

High spatial resolution GLASS FAPAR (version 2) product from Landsat imagery: Algorithm development using a knowledge transfer strategy 基于Landsat图像的高空间分辨率GLASS FAPAR（版本2）产品：使用知识转移策略的算法开发

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.2025.105051

Yuting Qiao , Huaan Jin , Tao He , Shunlin Liang , Feng Tian , Wei Zhao , Zhouyang Liu

The fraction of absorbed photosynthetically active radiation (FAPAR) is a critical parameter for measuring the vegetation photosynthetic capacity. The Hi-resolution Global LAnd Surface Satellite (Hi-GLASS) FAPAR product (version 1, V1) from Landsat imagery has been successfully applied to the ecosystem productivity modeling; however, this product algorithm still exhibits some limitations, including the poor adaptability to heterogeneous surfaces and limited physical interpretability, due to the absence of real-world knowledge guidance. To address these issues, we integrated deep transfer learning and radiative transfer models to update the Hi-GLASS FAPAR algorithm and generate the corresponding product (i.e., version 2, V2). A long short-term memory (LSTM) model was pre-trained on Soil-Leaf-Canopy (SLC) simulations and then optimized using physical knowledge-guided transfer learning, which was used to generate the new FAPAR product from Landsat image series. Validation results demonstrated that the Hi-GLASS FAPAR V2 (R² = 0. 95, RMSE = 0.08) significantly outperformed V1 (R² = 0.94, RMSE = 0.11), with notable improvements in various vegetation categories and sensors. The greatest improvement of FAPAR was found over multiple forest types, where different forest categories showed substantial gains, with R² increasing by 2 % − 11 % and RMSE decreasing by 15 % − 55 %, confirming the improved adaptability of our proposed method to heterogeneous canopies. Moreover, the Hi-GLASS V2 product preserved better spatial details than MODIS、GLASS、GEOV2 products, and its temporal dynamics were more closely aligned with field measurements than the V1 product. These advancements highlight the potential of Hi-GLASS FAPAR V2 as valuable data for supporting terrestrial ecosystem studies.

光合有效辐射吸收分数（FAPAR）是衡量植被光合能力的一个重要参数。基于Landsat影像的高分辨率全球陆地表面卫星（Hi-GLASS） FAPAR产品（版本1，V1）已成功应用于生态系统生产力模型；然而，由于缺乏现实世界的知识指导，该乘积算法仍然存在一定的局限性，包括对异构表面的适应性差和物理可解释性有限。为了解决这些问题，我们整合了深度迁移学习和辐射迁移模型来更新Hi-GLASS FAPAR算法并生成相应的产品（即版本2，V2）。在土壤-叶-冠（Soil-Leaf-Canopy， SLC）模拟中对长短期记忆（LSTM）模型进行预训练，并利用物理知识引导迁移学习对其进行优化，利用Landsat影像序列生成新的FAPAR产品。验证结果表明，Hi-GLASS FAPAR V2 (R2 = 0。95, RMSE = 0.08)显著优于V1 (R2 = 0.94, RMSE = 0.11)，在各种植被类别和传感器上均有显著改善。FAPAR在多种森林类型上的改善最大，其中不同森林类型均有显著提高，R2增加2% ~ 11%，RMSE降低15% ~ 55%，证实了我们提出的方法对异质林冠的适应性提高。与MODIS、GLASS、GEOV2产品相比，Hi-GLASS V2产品保留了更好的空间细节，其时间动态比V1产品与现场测量更接近。这些进展突出了Hi-GLASS FAPAR V2作为支持陆地生态系统研究的宝贵数据的潜力。

{"title":"High spatial resolution GLASS FAPAR (version 2) product from Landsat imagery: Algorithm development using a knowledge transfer strategy","authors":"Yuting Qiao , Huaan Jin , Tao He , Shunlin Liang , Feng Tian , Wei Zhao , Zhouyang Liu","doi":"10.1016/j.jag.2025.105051","DOIUrl":"10.1016/j.jag.2025.105051","url":null,"abstract":"<div><div>The fraction of absorbed photosynthetically active radiation (FAPAR) is a critical parameter for measuring the vegetation photosynthetic capacity. The Hi-resolution Global LAnd Surface Satellite (Hi-GLASS) FAPAR product (version 1, V1) from Landsat imagery has been successfully applied to the ecosystem productivity modeling; however, this product algorithm still exhibits some limitations, including the poor adaptability to heterogeneous surfaces and limited physical interpretability, due to the absence of real-world knowledge guidance. To address these issues, we integrated deep transfer learning and radiative transfer models to update the Hi-GLASS FAPAR algorithm and generate the corresponding product (i.e., version 2, V2). A long short-term memory (LSTM) model was pre-trained on Soil-Leaf-Canopy (SLC) simulations and then optimized using physical knowledge-guided transfer learning, which was used to generate the new FAPAR product from Landsat image series. Validation results demonstrated that the Hi-GLASS FAPAR V2 (R<sup>2</sup> = 0. 95, RMSE = 0.08) significantly outperformed V1 (R<sup>2</sup> = 0.94, RMSE = 0.11), with notable improvements in various vegetation categories and sensors. The greatest improvement of FAPAR was found over multiple forest types, where different forest categories showed substantial gains, with R<sup>2</sup> increasing by 2 % − 11 % and RMSE decreasing by 15 % − 55 %, confirming the improved adaptability of our proposed method to heterogeneous canopies. Moreover, the Hi-GLASS V2 product preserved better spatial details than MODIS、GLASS、GEOV2 products, and its temporal dynamics were more closely aligned with field measurements than the V1 product. These advancements highlight the potential of Hi-GLASS FAPAR V2 as valuable data for supporting terrestrial ecosystem studies.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105051"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925519","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

STAMI: A machine learning-based small-scale tomography-aided multibaseline (Pol)InSAR forest height inversion framework STAMI：基于机器学习的小尺度层摄影辅助多基线InSAR森林高度反演框架

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.2025.105054

Tianyi Song , Jie Yang , Weidong Sun , Lei Shi , Changcheng Wang , Pingxiang Li , Haiqiang Fu , Lingli Zhao , Peng Shen , Pingping Huang

The complementary measurements of model-based multibaseline (polarimetric) synthetic aperture radar interferometry ((Pol)InSAR) and data-driven SAR tomography (TomoSAR) have become the development trend of forest height mapping missions. Their common core is the characterization of forest vertical structure, which promotes the introduction of TomoSAR reconstructed vertical structure function into (Pol)InSAR forest height inversion model. However, to provide knowledge of vertical structure, TomoSAR measurements have to cover the entire forested area. The high data acquisition cost and multibaseline observation errors hinder its application in large-scale wall-to-wall forest height mapping. Therefore, based on machine learning, this study proposes a small-scale tomography-aided multibaseline (Pol)InSAR (STAMI) framework to derive vertical structure knowledge at low data burden and estimate forest height with high accuracy. The synergy inversion is reformulated as the cross-scale clustering and classification task of the vertical structure knowledge. We conduct TomoSAR only in a small-scale forested area, and establish the mapping relationship between interferometric features and vertical structure with spectral clustering and Support Vector Machine (SVM). In large-scale mapping area, we obtain vertical structure information with only three-baseline (Pol)InSAR, which is used to construct different forest height inversion models. Finally, a reweighted method is applied to resist observation errors to different vertical structures for improving accuracy of three-baseline inversion. Airborne L- and P-band single and fully polarimetric datasets covering boreal and tropical forests are used to demonstrate the superiority of the proposed framework. This research is valuable for data acquisition planning and algorithm application of multibaseline (Pol)InSAR missions aimed at global forest mapping.

基于模型的多基线（偏振）合成孔径雷达干涉测量（Pol）InSAR)与数据驱动的SAR层析成像（TomoSAR）的互补测量已成为森林高程制图任务的发展趋势。它们的共同核心是森林垂直结构的表征，这促进了将TomoSAR重建的垂直结构函数引入（Pol）InSAR森林高度反演模型。然而，为了提供垂直结构的知识，TomoSAR测量必须覆盖整个森林地区。数据采集成本高，多基线观测误差大，阻碍了该方法在大范围墙对墙森林高度制图中的应用。因此，本研究基于机器学习，提出了一种小尺度层摄影辅助多基线InSAR （STAMI）框架，以低数据负担获得垂直结构知识，并以高精度估计森林高度。将协同反演重新表述为垂直结构知识的跨尺度聚类和分类任务。我们只在小尺度林区进行TomoSAR，利用光谱聚类和支持向量机（SVM）建立干涉特征与垂直结构的映射关系。在大尺度作图区，仅利用三基线（Pol）InSAR获取垂直结构信息，构建不同森林高度反演模型。最后，采用加权方法抵抗不同垂直结构的观测误差，提高三基线反演精度。覆盖北方森林和热带森林的机载L波段和p波段单波段和全偏振数据集证明了所提出框架的优越性。该研究对全球森林制图的多基线InSAR任务的数据采集规划和算法应用具有一定的参考价值。

{"title":"STAMI: A machine learning-based small-scale tomography-aided multibaseline (Pol)InSAR forest height inversion framework","authors":"Tianyi Song , Jie Yang , Weidong Sun , Lei Shi , Changcheng Wang , Pingxiang Li , Haiqiang Fu , Lingli Zhao , Peng Shen , Pingping Huang","doi":"10.1016/j.jag.2025.105054","DOIUrl":"10.1016/j.jag.2025.105054","url":null,"abstract":"<div><div>The complementary measurements of model-based multibaseline (polarimetric) synthetic aperture radar interferometry ((Pol)InSAR) and data-driven SAR tomography (TomoSAR) have become the development trend of forest height mapping missions. Their common core is the characterization of forest vertical structure, which promotes the introduction of TomoSAR reconstructed vertical structure function into (Pol)InSAR forest height inversion model. However, to provide knowledge of vertical structure, TomoSAR measurements have to cover the entire forested area. The high data acquisition cost and multibaseline observation errors hinder its application in large-scale wall-to-wall forest height mapping. Therefore, based on machine learning, this study proposes a small-scale tomography-aided multibaseline (Pol)InSAR (STAMI) framework to derive vertical structure knowledge at low data burden and estimate forest height with high accuracy. The synergy inversion is reformulated as the cross-scale clustering and classification task of the vertical structure knowledge. We conduct TomoSAR only in a small-scale forested area, and establish the mapping relationship between interferometric features and vertical structure with spectral clustering and Support Vector Machine (SVM). In large-scale mapping area, we obtain vertical structure information with only three-baseline (Pol)InSAR, which is used to construct different forest height inversion models. Finally, a reweighted method is applied to resist observation errors to different vertical structures for improving accuracy of three-baseline inversion. Airborne L- and P-band single and fully polarimetric datasets covering boreal and tropical forests are used to demonstrate the superiority of the proposed framework. This research is valuable for data acquisition planning and algorithm application of multibaseline (Pol)InSAR missions aimed at global forest mapping.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105054"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925539","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

Cloud-based big data analytics for monitoring invasive plants in groundwater-dependent ecosystems of Nuwejaars catchment, South Africa 基于云的大数据分析用于监测南非Nuwejaars流域地下水依赖生态系统中的入侵植物

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.2025.105053

Mmasechaba L. Moropane , Cletah Shoko , Timothy Dube , Dominic Mazvimavi

Groundwater-dependent ecosystems (GDEs) provide crucial ecological and hydrological stability but are increasingly threatened by groundwater-dependent invasive plants (GDIPs), particularly in regions with limited water resources. Although GDEs have been widely studied, long-term quantitative assessments of how invasive plants alter these ecosystems remain limited. Hence, this study evaluated the impacts of invasive plants within the GDEs of the Nuwejaars Catchment, South Africa, by monitoring their spatial and temporal dynamics and quantifying the extent to which they displace native plants. Landsat-8 imagery, a Random Forest classifier, and Explainable Artificial Intelligence (XAI) techniques were integrated to map and quantify the annual distribution of GDIPs over a 12-year period. XAI interpretability techniques including SHapley Additive exPlanations (SHAP), partial dependence plots (PDPs), and recursive feature elimination (RFECV) were applied to identify key environmental conditions influencing GDIP occurrence. Spatial-temporal analysis revealed that GDIPs expanded from 40.9 % (1060 ha) in 2013 to 63.9 % (1660 ha) in 2024, displacing large areas of native fynbos vegetation. Inter-annual change analysis showed accelerated GDIP growth following the extreme 2015–2018 drought, which reduced groundwater availability for native species with shallow roots. Elevation, slope, and moisture vegetation indices emerged as the most influential predictors for classification, with PDPs revealing that GDIPs favoured lower elevations and steep slopes. Classification accuracy improved over time, with F1-Scores and overall accuracies ranging between 68.4 % to 82.5 % from 2013 to 2024. Overall, these findings highlight the persistent spread of GDIPs and their potential to transform GDEs in semi-arid areas. This study demonstrates the value of integrating remote sensing and interpretable machine learning to support ecological monitoring and targeted invasive species management.

地下水依赖生态系统（GDEs）提供了至关重要的生态和水文稳定性，但日益受到地下水依赖入侵植物（gdip）的威胁，特别是在水资源有限的地区。尽管对gde进行了广泛的研究，但对入侵植物如何改变这些生态系统的长期定量评估仍然有限。因此，本研究评估了入侵植物对南非Nuwejaars流域GDEs的影响，通过监测它们的时空动态并量化它们取代本地植物的程度。将Landsat-8图像、随机森林分类器和可解释人工智能（XAI）技术整合在一起，绘制和量化了12年期间gdip的年度分布。应用SHapley加性解释（SHAP）、部分依赖图（pdp）和递归特征消除（RFECV）等XAI可解释性技术识别影响GDIP发生的关键环境条件。时空分析表明，GDIPs从2013年的40.9% （1060 ha）扩大到2024年的63.9% (1660 ha)，大面积取代了原生态植被。年际变化分析显示，在2015-2018年极端干旱之后，GDIP增长加速，这减少了浅根本地物种的地下水可用性。海拔、坡度和湿度植被指数是最具影响力的分类预测指标，pdp显示gdip倾向于低海拔和陡坡。分类精度随着时间的推移而提高，从2013年到2024年，f1分数和总体准确率在68.4%到82.5%之间。总的来说，这些发现突出了gdip的持续传播及其在半干旱地区转化gde的潜力。该研究证明了将遥感和可解释机器学习结合起来支持生态监测和有针对性的入侵物种管理的价值。

{"title":"Cloud-based big data analytics for monitoring invasive plants in groundwater-dependent ecosystems of Nuwejaars catchment, South Africa","authors":"Mmasechaba L. Moropane , Cletah Shoko , Timothy Dube , Dominic Mazvimavi","doi":"10.1016/j.jag.2025.105053","DOIUrl":"10.1016/j.jag.2025.105053","url":null,"abstract":"<div><div>Groundwater-dependent ecosystems (GDEs) provide crucial ecological and hydrological stability but are increasingly threatened by groundwater-dependent invasive plants (GDIPs), particularly in regions with limited water resources. Although GDEs have been widely studied, long-term quantitative assessments of how invasive plants alter these ecosystems remain limited. Hence, this study evaluated the impacts of invasive plants within the GDEs of the Nuwejaars Catchment, South Africa, by monitoring their spatial and temporal dynamics and quantifying the extent to which they displace native plants. Landsat-8 imagery, a Random Forest classifier, and Explainable Artificial Intelligence (XAI) techniques were integrated to map and quantify the annual distribution of GDIPs over a 12-year period. XAI interpretability techniques including SHapley Additive exPlanations (SHAP), partial dependence plots (PDPs), and recursive feature elimination (RFECV) were applied to identify key environmental conditions influencing GDIP occurrence. Spatial-temporal analysis revealed that GDIPs expanded from 40.9 % (1060 ha) in 2013 to 63.9 % (1660 ha) in 2024, displacing large areas of native fynbos vegetation. Inter-annual change analysis showed accelerated GDIP growth following the extreme 2015–2018 drought, which reduced groundwater availability for native species with shallow roots. Elevation, slope, and moisture vegetation indices emerged as the most influential predictors for classification, with PDPs revealing that GDIPs favoured lower elevations and steep slopes. Classification accuracy improved over time, with F1-Scores and overall accuracies ranging between 68.4 % to 82.5 % from 2013 to 2024. Overall, these findings highlight the persistent spread of GDIPs and their potential to transform GDEs in semi-arid areas. This study demonstrates the value of integrating remote sensing and interpretable machine learning to support ecological monitoring and targeted invasive species management.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105053"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925609","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

Precise urban tree species identification and biomass estimation using UAV–Handheld LiDAR Synergy and YOLOv11 deep learning 基于无人机-手持激光雷达协同和YOLOv11深度学习的精确城市树种识别和生物量估算

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.2025.105049

Zhaohan Huo , Lei Fang , Yukai Chu , Shuo Dang , Jian Yang , Lin Li , Xuan Li , Shilong Ren , Jinyue Chen , Yanbo Peng , Guoqiang Wang , Qiao Wang

Accurate estimation of individual tree biomass in urban landscapes is critical for carbon stock assessment and urban forest management but remains challenging because of the structural complexity and species diversity of urban trees. This study presents an integrated methodological framework that combines deep learning-based tree species identification with LiDAR-derived structural parameter estimation to enable rapid and precise biomass mapping at the individual tree level. Using multiplatform LiDAR data (UAV-borne and handheld mobile laser scanning), we developed a lightweight sample generation method derived from side-view tree projections (SVP) to efficiently construct a species-adaptive training library and proposed a novel individual tree identification approach optimized with the YOLOv11 deep learning algorithm. Our framework systematically evaluated the performance of single-source versus fused LiDAR point clouds across three key metrics: species classification accuracy (achieving 87.3 % on independent test data), structural parameter retrieval (R² = 0.925 for DBH, R² = 0.844 for height), and biomass estimation fidelity (86.3 % agreement with field measurements). The results demonstrated that compared with single-source alternatives, data fusion reduces parameter estimation errors by 4.8–56.2 %, while the SVP strategy enables computationally efficient species-specific allometric model matching. This work advances urban forest monitoring by providing a scalable solution that balances scientific rigor with operational practicality, addressing critical gaps in high-resolution biomass mapping for heterogeneous urban ecosystems.

准确估算城市景观中单个树木生物量对碳储量评估和城市森林管理至关重要，但由于城市树木的结构复杂性和物种多样性，仍然具有挑战性。本研究提出了一个集成的方法框架，将基于深度学习的树种识别与激光雷达衍生的结构参数估计相结合，以实现在单个树木水平上快速精确的生物量映射。利用多平台激光雷达数据（无人机机载和手持移动激光扫描），开发了一种基于侧视树投影（SVP）的轻量级样本生成方法，以高效构建物种自适应训练库，并提出了一种基于YOLOv11深度学习算法优化的新型个体树识别方法。我们的框架系统地评估了单源激光雷达点云与融合激光雷达点云在三个关键指标上的性能：物种分类精度（在独立测试数据上达到87.3%）、结构参数检索（胸径R2 = 0.925，高度R2 = 0.844）和生物量估计保真度（与现场测量一致度为86.3%）。结果表明，与单源方案相比，数据融合可将参数估计误差降低4.8 - 56.2%，而SVP策略可实现计算效率高的物种特异性异速生长模型匹配。这项工作通过提供一种可扩展的解决方案来促进城市森林监测，该解决方案平衡了科学严谨性和操作实用性，解决了异质性城市生态系统高分辨率生物量制图的关键空白。

{"title":"Precise urban tree species identification and biomass estimation using UAV–Handheld LiDAR Synergy and YOLOv11 deep learning","authors":"Zhaohan Huo , Lei Fang , Yukai Chu , Shuo Dang , Jian Yang , Lin Li , Xuan Li , Shilong Ren , Jinyue Chen , Yanbo Peng , Guoqiang Wang , Qiao Wang","doi":"10.1016/j.jag.2025.105049","DOIUrl":"10.1016/j.jag.2025.105049","url":null,"abstract":"<div><div>Accurate estimation of individual tree biomass in urban landscapes is critical for carbon stock assessment and urban forest management but remains challenging because of the structural complexity and species diversity of urban trees. This study presents an integrated methodological framework that combines deep learning-based tree species identification with LiDAR-derived structural parameter estimation to enable rapid and precise biomass mapping at the individual tree level. Using multiplatform LiDAR data (UAV-borne and handheld mobile laser scanning), we developed a lightweight sample generation method derived from side-view tree projections (SVP) to efficiently construct a species-adaptive training library and proposed a novel individual tree identification approach optimized with the YOLOv11 deep learning algorithm. Our framework systematically evaluated the performance of single-source versus fused LiDAR point clouds across three key metrics: species classification accuracy (achieving 87.3 % on independent test data), structural parameter retrieval (R<sup>2</sup> = 0.925 for DBH, R<sup>2</sup> = 0.844 for height), and biomass estimation fidelity (86.3 % agreement with field measurements). The results demonstrated that compared with single-source alternatives, data fusion reduces parameter estimation errors by 4.8–56.2 %, while the SVP strategy enables computationally efficient species-specific allometric model matching. This work advances urban forest monitoring by providing a scalable solution that balances scientific rigor with operational practicality, addressing critical gaps in high-resolution biomass mapping for heterogeneous urban ecosystems.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105049"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925518","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

ASAI: A general and training-free artificial surfaces anomaly index using post-disaster single-temporal and high-resolution imagery ASAI：基于灾后单时间和高分辨率图像的通用和无需训练的人工地表异常指数

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.2025.105050

Shoujia Ren , Yaozhong Pan , Chuanwu Zhao , Yuan Gao , Gelilan Ma

Efficient detection of anomalous artificial surfaces is vital after disasters to support rapid emergency response. This study proposes a novel, training-free texture index method to accurately identify damaged artificial surfaces using post-disaster imagery. The method was tested across diverse disaster sites, including earthquake-affected areas in Turkey (Sites A–C), and tsunami- and tornado-damaged regions in Palu and Joplin (Sites D–I), covering a range of surface types and building structures. Sites A–C were chosen to develop and assess the efficacy of the proposed texture index method. Primarily, the three-dimensional texture features (3DTF) comprised of Contrast, Gabor wavelets, and secondary texture extraction (Con_Gabor), were amalgamated with a K-means classifier to delineate post-disaster artificial surface areas without prior knowledge. Given the discernible texture discrepancies between normal and anomalous artificial surfaces post-disaster, Homogeneity and Entropy texture features derived from Worldview-3 images at each site were leveraged to construct the artificial surface anomaly index (ASAI) for automatically extracting anomalous artificial surfaces. The findings demonstrated high overall accuracies of detecting anomaly artificial surfaces, ranging from 90.07 % to 91.76 % in Sites A–C. Notably, the ASAI outperformed Artificial Neural Network (ANN), Random Forest (RF), and the U-Net model. The overall accuracy (OA) of the ASAI method is 10 %–20 % higher than that of the U-Net model, showing superior automatic performance without necessitating training samples. Furthermore, the effectiveness of ASAI was validated in the Palu and Joplin sites, affirming its utility across diverse disaster scenarios. The damages of steel-tiled houses, middle-level houses and roads were effectively identified as anomaly artificial surfaces. The overall accuracies of the ASAI for identifying anomaly artificial surfaces were 89.55 %–92.5 %. These findings indicate the identification anomaly artificial surface using the ASAI was robust in different types of anomaly artificial surface caused by different disaster. The method of using ASAI to automatically identify anomaly artificial surfaces in post-disaster and single-temporal images has the potential for wide applicability.

在灾害发生后，对异常人工表面的有效检测对于支持快速应急响应至关重要。本研究提出一种新的、无需训练的纹理指数方法，利用灾后图像准确识别受损人工表面。该方法在不同的灾难现场进行了测试，包括土耳其的地震灾区（地点a - c），帕卢和乔普林的海啸和龙卷风受灾地区（地点D-I），涵盖了一系列表面类型和建筑结构。选取A-C位点开发并评估所提出的纹理指数方法的有效性。首先，将由对比度、Gabor小波和二次纹理提取（Con_Gabor）组成的三维纹理特征（3DTF）与K-means分类器相结合，在没有先验知识的情况下描绘灾后人工表面区域。考虑到灾后正常和异常人工地表存在明显的纹理差异，利用Worldview-3图像的均匀性和熵性纹理特征构建人工地表异常指数（ASAI），实现异常人工地表的自动提取。结果表明，A-C站点异常人工地表的检测精度在90.07% ~ 91.76%之间，总体精度较高。值得注意的是，ASAI优于人工神经网络（ANN）、随机森林（RF）和U-Net模型。ASAI方法的总体准确率（OA）比U-Net模型高10% - 20%，在不需要训练样本的情况下表现出优越的自动性能。此外，ASAI的有效性在Palu和Joplin现场得到了验证，证实了其在各种灾难场景中的实用性。钢瓦房屋、中层房屋和道路的破坏被有效地识别为异常人工路面。ASAI识别异常人工表面的总体准确率为89.55% ~ 92.5%。结果表明，在不同灾害类型的异常人工地表中，ASAI识别异常人工地表具有较强的鲁棒性。利用ASAI自动识别灾后和单时相影像中异常人工地表的方法具有广泛的适用性。

{"title":"ASAI: A general and training-free artificial surfaces anomaly index using post-disaster single-temporal and high-resolution imagery","authors":"Shoujia Ren , Yaozhong Pan , Chuanwu Zhao , Yuan Gao , Gelilan Ma","doi":"10.1016/j.jag.2025.105050","DOIUrl":"10.1016/j.jag.2025.105050","url":null,"abstract":"<div><div>Efficient detection of anomalous artificial surfaces is vital after disasters to support rapid emergency response. This study proposes a novel, training-free texture index method to accurately identify damaged artificial surfaces using post-disaster imagery. The method was tested across diverse disaster sites, including earthquake-affected areas in Turkey (Sites A–C), and tsunami- and tornado-damaged regions in Palu and Joplin (Sites D–I), covering a range of surface types and building structures. Sites A–C were chosen to develop and assess the efficacy of the proposed texture index method. Primarily, the three-dimensional texture features (3DTF) comprised of Contrast, Gabor wavelets, and secondary texture extraction (Con_Gabor), were amalgamated with a K-means classifier to delineate post-disaster artificial surface areas without prior knowledge. Given the discernible texture discrepancies between normal and anomalous artificial surfaces post-disaster, Homogeneity and Entropy texture features derived from Worldview-3 images at each site were leveraged to construct the artificial surface anomaly index (ASAI) for automatically extracting anomalous artificial surfaces. The findings demonstrated high overall accuracies of detecting anomaly artificial surfaces, ranging from 90.07 % to 91.76 % in Sites A–C. Notably, the ASAI outperformed Artificial Neural Network (ANN), Random Forest (RF), and the U-Net model. The overall accuracy (OA) of the ASAI method is 10 %–20 % higher than that of the U-Net model, showing superior automatic performance without necessitating training samples. Furthermore, the effectiveness of ASAI was validated in the Palu and Joplin sites, affirming its utility across diverse disaster scenarios. The damages of steel-tiled houses, middle-level houses and roads were effectively identified as anomaly artificial surfaces. The overall accuracies of the ASAI for identifying anomaly artificial surfaces were 89.55 %–92.5 %. These findings indicate the identification anomaly artificial surface using the ASAI was robust in different types of anomaly artificial surface caused by different disaster. The method of using ASAI to automatically identify anomaly artificial surfaces in post-disaster and single-temporal images has the potential for wide applicability.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105050"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925604","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

Million clandestine gravesites over southeastern China’s land surfaces revealed by satellite images 卫星图像显示，中国东南部陆地表面有数百万个秘密墓地

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.2025.105063

Lei Xu , Zhongliang Wang , Yongxi Wang , Ting Xiong , Zhenni Ye

Gravesite occupies a nonnegligible fraction of land resources with growing population in China. Apart from public cemeteries, the clandestine gravesites built by local residents distributed over various land surfaces are yet not to be understood. Here, we use high-resolution satellite images and deep learning methods to detect prevailing informal gravesites in highly vegetated land surfaces of southeastern China. The deep learning gravesite detection model is trained, validated and tested using 20,349 manually labeled samples and achieves an average precision (AP) of 0.905 during the testing stage (precision = 0.830, recall = 0.840, F1 score = 0.835). Using 31 million satellite image tiles, we detect nearly one million (996,074) clandestine gravesites over the land surfaces of Zhejiang, Fujian and Guangdong provinces, especially in forests and croplands. Further spatial attribution analysis suggests that the detected mass informal gravesites are closely related to gross domestic product (GDP) per capita, elevation, vegetation cover, climatic factors and the distances to rivers and coastlines. These findings and conclusions may provide meaningful references for gravesite occupied land use monitoring and management at regional to national scales.

随着中国人口的不断增长，墓地占用了不可忽视的土地资源。除了公共墓地外，当地居民建造的秘密墓地分布在各个地面上，目前还不清楚。在这里，我们使用高分辨率卫星图像和深度学习方法来检测中国东南部高度植被覆盖的地表上普遍存在的非正式墓地。深度学习墓地检测模型使用20,349个人工标记样本进行训练、验证和测试，测试阶段的平均精度（AP）为0.905（精度= 0.830，召回率= 0.840，F1得分= 0.835）。利用3100万张卫星图像，我们在浙江、福建和广东三省的地面上发现了近100万（996,074）个秘密墓地，特别是在森林和农田里。进一步的空间归因分析表明，发现的大规模非正式墓地与人均国内生产总值、海拔、植被覆盖、气候因素以及到河流和海岸线的距离密切相关。这些发现和结论可为区域乃至全国范围内墓地占用土地利用监测和管理提供有意义的参考。

{"title":"Million clandestine gravesites over southeastern China’s land surfaces revealed by satellite images","authors":"Lei Xu , Zhongliang Wang , Yongxi Wang , Ting Xiong , Zhenni Ye","doi":"10.1016/j.jag.2025.105063","DOIUrl":"10.1016/j.jag.2025.105063","url":null,"abstract":"<div><div>Gravesite occupies a nonnegligible fraction of land resources with growing population in China. Apart from public cemeteries, the clandestine gravesites built by local residents distributed over various land surfaces are yet not to be understood. Here, we use high-resolution satellite images and deep learning methods to detect prevailing informal gravesites in highly vegetated land surfaces of southeastern China. The deep learning gravesite detection model is trained, validated and tested using 20,349 manually labeled samples and achieves an average precision (AP) of 0.905 during the testing stage (precision = 0.830, recall = 0.840, F1 score = 0.835). Using 31 million satellite image tiles, we detect nearly one million (996,074) clandestine gravesites over the land surfaces of Zhejiang, Fujian and Guangdong provinces, especially in forests and croplands. Further spatial attribution analysis suggests that the detected mass informal gravesites are closely related to gross domestic product (GDP) per capita, elevation, vegetation cover, climatic factors and the distances to rivers and coastlines. These findings and conclusions may provide meaningful references for gravesite occupied land use monitoring and management at regional to national scales.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105063"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925606","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

Daily seamless 30-m fractional snow cover mapping via an adaptive Time-Series approach 通过自适应时间序列方法绘制每日30米的无缝积雪覆盖地图

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.2025.105068

Cheng Zhang , Lingmei Jiang , Jinmei Pan , Jianwei Yang , Jian Wang , Zongyi Jin

Accurate daily mapping of 30-m fractional snow cover (FSC) is critical for hydrological modeling and disaster assessment. Frequent cloud cover and satellite revisit cycles create significant data gaps in high-resolution optical imagery (e.g., Landsat, Sentinel-2), hindering the continuous monitoring of rapid snow dynamics. To address these limitations, we propose the Time-series-based Adaptive snow-Fraction Fusion (TAFF) framework for generating seamless daily 30-m FSC over large scales. The core of TAFF is a dual-path fusion strategy that adapts to the physical state of the snowpack. First, a time-series-based snow stability assessment gauges the magnitude of temporal FSC change. This assessment then directs the fusion process: stable snow is processed using time-weighted fusion, while rapidly changing snow is handled by a pixel-level regression. Evaluated over the Qinghai-Tibet Plateau, TAFF demonstrates robust improvements over established spatiotemporal fusion algorithms, particularly under cloudy conditions. Independent validation against 215 Landsat 8 images yielded strong performance (R² = 0.76, RMSE = 19.58 %). Further validation against 46 in-situ snow depth stations indicated a high binary classification accuracy of 0.91. As a robust and practical method for large-scale FSC mapping, TAFF shows promise for integrating additional data sources, such as geostationary and microwave sensors, to enhance the high-resolution monitoring of ephemeral snow.

30米积雪（FSC）的精确日制图对水文建模和灾害评估至关重要。频繁的云层覆盖和卫星重访周期在高分辨率光学图像（例如Landsat、Sentinel-2）中造成了重大的数据缺口，阻碍了对快速降雪动态的连续监测。为了解决这些限制，我们提出了基于时间序列的自适应雪分数融合（TAFF）框架，用于在大尺度上生成无缝的每日30米FSC。TAFF的核心是一种适应积雪物理状态的双路径融合策略。首先，基于时间序列的雪稳定性评估测量了时间FSC变化的幅度。然后，这个评估指导融合过程：稳定的雪使用时间加权融合处理，而快速变化的雪使用像素级回归处理。在青藏高原上进行的评估表明，TAFF比现有的时空融合算法有了强大的改进，特别是在多云条件下。对215幅Landsat 8图像的独立验证获得了较好的效果（R2 = 0.76, RMSE = 19.58%）。进一步对46个现场雪深站进行验证，二元分类精度达到0.91。作为一种强大而实用的大规模FSC测绘方法，TAFF有望整合其他数据源，如地球静止和微波传感器，以增强对短暂降雪的高分辨率监测。

{"title":"Daily seamless 30-m fractional snow cover mapping via an adaptive Time-Series approach","authors":"Cheng Zhang , Lingmei Jiang , Jinmei Pan , Jianwei Yang , Jian Wang , Zongyi Jin","doi":"10.1016/j.jag.2025.105068","DOIUrl":"10.1016/j.jag.2025.105068","url":null,"abstract":"<div><div>Accurate daily mapping of 30-m fractional snow cover (FSC) is critical for hydrological modeling and disaster assessment. Frequent cloud cover and satellite revisit cycles create significant data gaps in high-resolution optical imagery (e.g., Landsat, Sentinel-2), hindering the continuous monitoring of rapid snow dynamics. To address these limitations, we propose the Time-series-based Adaptive snow-Fraction Fusion (TAFF) framework for generating seamless daily 30-m FSC over large scales. The core of TAFF is a dual-path fusion strategy that adapts to the physical state of the snowpack. First, a time-series-based snow stability assessment gauges the magnitude of temporal FSC change. This assessment then directs the fusion process: stable snow is processed using time-weighted fusion, while rapidly changing snow is handled by a pixel-level regression. Evaluated over the Qinghai-Tibet Plateau, TAFF demonstrates robust improvements over established spatiotemporal fusion algorithms, particularly under cloudy conditions. Independent validation against 215 Landsat 8 images yielded strong performance (R<sup>2</sup> = 0.76, RMSE = 19.58 %). Further validation against 46 in-situ snow depth stations indicated a high binary classification accuracy of 0.91. As a robust and practical method for large-scale FSC mapping, TAFF shows promise for integrating additional data sources, such as geostationary and microwave sensors, to enhance the high-resolution monitoring of ephemeral snow.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"146 ","pages":"Article 105068"},"PeriodicalIF":8.6,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925668","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