International journal of applied earth observation and geoinformation : ITC journal最新文献_第5页

Fine-grained building function recognition with street-view images and GIS map data via geometry-aware semi-supervised learning

IF 7.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-02-24 DOI: 10.1016/j.jag.2025.104386

Weijia Li , Jinhua Yu , Dairong Chen , Yi Lin , Runmin Dong , Xiang Zhang , Conghui He , Haohuan Fu

The diversity of building functions is vital for urban planning and optimizing infrastructure and services. Street-view images offer rich exterior details, aiding in function recognition. However, street-view building function annotations are limited and challenging to obtain. In this work, we propose a geometry-aware semi-supervised method for fine-grained building function recognition, which effectively uses multi-source geoinformation data to achieve accurate function recognition in both single-city and cross-city scenarios. We restructured the semi-supervised method based on the Teacher–Student architecture into three stages, which involve pre-training for building facade recognition, building function annotation generation, and building function recognition. In the first stage, to enable semi-supervised training with limited annotations, we employ a semi-supervised object detection model, which trains on both labeled samples and a large amount of unlabeled data simultaneously, achieving building facade detection. In the second stage, to further optimize the pseudo-labels, we effectively utilize the geometric spatial relationships between GIS map data and panoramic street-view images, integrating the building function information with facade detection results. We ultimately achieve fine-grained building function recognition in both single-city and cross-city scenarios by combining the coarse annotations and labeled data in the final stage. We conduct extensive comparative experiments on four datasets, which include OmniCity, Madrid, Los Angeles, and Boston, to evaluate the performance of our method in both single-city (OmniCity & Madrid) and cross-city (OmniCity - Los Angeles & OmniCity - Boston) scenarios. The experimental results show that, compared to advanced recognition methods, our method improves mAP by at least 4.8% and 4.3% for OmniCity and Madrid, respectively, while also effectively handling class imbalance. Furthermore, our method performs well in the cross-categorization system experiments for Los Angeles and Boston, highlighting its strong potential for cross-city tasks. This study offers a new solution for large-scale and multi-city applications by efficiently utilizing multi-source geoinformation data, enhancing urban information acquisition efficiency, and assisting in rational resource allocation.

{"title":"Fine-grained building function recognition with street-view images and GIS map data via geometry-aware semi-supervised learning","authors":"Weijia Li , Jinhua Yu , Dairong Chen , Yi Lin , Runmin Dong , Xiang Zhang , Conghui He , Haohuan Fu","doi":"10.1016/j.jag.2025.104386","DOIUrl":"10.1016/j.jag.2025.104386","url":null,"abstract":"<div><div>The diversity of building functions is vital for urban planning and optimizing infrastructure and services. Street-view images offer rich exterior details, aiding in function recognition. However, street-view building function annotations are limited and challenging to obtain. In this work, we propose a geometry-aware semi-supervised method for fine-grained building function recognition, which effectively uses multi-source geoinformation data to achieve accurate function recognition in both single-city and cross-city scenarios. We restructured the semi-supervised method based on the Teacher–Student architecture into three stages, which involve pre-training for building facade recognition, building function annotation generation, and building function recognition. In the first stage, to enable semi-supervised training with limited annotations, we employ a semi-supervised object detection model, which trains on both labeled samples and a large amount of unlabeled data simultaneously, achieving building facade detection. In the second stage, to further optimize the pseudo-labels, we effectively utilize the geometric spatial relationships between GIS map data and panoramic street-view images, integrating the building function information with facade detection results. We ultimately achieve fine-grained building function recognition in both single-city and cross-city scenarios by combining the coarse annotations and labeled data in the final stage. We conduct extensive comparative experiments on four datasets, which include OmniCity, Madrid, Los Angeles, and Boston, to evaluate the performance of our method in both single-city (OmniCity & Madrid) and cross-city (OmniCity - Los Angeles & OmniCity - Boston) scenarios. The experimental results show that, compared to advanced recognition methods, our method improves mAP by at least 4.8% and 4.3% for OmniCity and Madrid, respectively, while also effectively handling class imbalance. Furthermore, our method performs well in the cross-categorization system experiments for Los Angeles and Boston, highlighting its strong potential for cross-city tasks. This study offers a new solution for large-scale and multi-city applications by efficiently utilizing multi-source geoinformation data, enhancing urban information acquisition efficiency, and assisting in rational resource allocation.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"137 ","pages":"Article 104386"},"PeriodicalIF":7.6,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143474903","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An integrated graph-spatial method for high-performance geospatial-temporal semantic query

IF 7.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-02-22 DOI: 10.1016/j.jag.2025.104437

Zichen Yue , Wei Zhu , Xin Mei , Shaobo Zhong

Knowledge graphs (KGs) have gained significant attention in the GIS community as a cutting-edge technology for linking heterogeneous and multimodal data sources. However, the efficiency of semantic querying of geospatial-temporal data in KGs remains a challenge. Graph databases excel at handling complex semantic associations but exhibit low efficiency in geospatial analysis tasks, such as topological analysis and geographic calculations, while relational databases excel at geospatial data storage and computation but struggle to efficiently process association analysis. To address this issue, we propose GraST, a geospatial-temporal semantic query optimization method that integrates property graphs and relational databases. GraST stores complete geospatial-temporal objects in a relational database (using built-in or extended spatial data engines), and employs spatiotemporal partitioning and indexing to enhance query efficiency. Simultaneously, GraST stores lightweight geospatial-temporal nodes in the graph database and links them to multi-granularity time tree and Geohash encoding nodes to enhance spatiotemporal aggregation capabilities. During query processing, user queries are broken down into graph semantic searches and geospatial calculations, pushed down to the graph and relational database for execution. Additionally, GraST adopts the two-phase commit protocol for cross-database data synchronization. We implemented a GraST prototype system by integrating PostGIS and Neo4j, and conducted performance evaluations and case studies on large-scale real-world datasets. Experimental results demonstrate that GraST shortens query response times by 1–2 orders of magnitude and offers flexible support for diverse geospatial-temporal semantic queries.

{"title":"An integrated graph-spatial method for high-performance geospatial-temporal semantic query","authors":"Zichen Yue , Wei Zhu , Xin Mei , Shaobo Zhong","doi":"10.1016/j.jag.2025.104437","DOIUrl":"10.1016/j.jag.2025.104437","url":null,"abstract":"<div><div>Knowledge graphs (KGs) have gained significant attention in the GIS community as a cutting-edge technology for linking heterogeneous and multimodal data sources. However, the efficiency of semantic querying of geospatial-temporal data in KGs remains a challenge. Graph databases excel at handling complex semantic associations but exhibit low efficiency in geospatial analysis tasks, such as topological analysis and geographic calculations, while relational databases excel at geospatial data storage and computation but struggle to efficiently process association analysis. To address this issue, we propose GraST, a geospatial-temporal semantic query optimization method that integrates property graphs and relational databases. GraST stores complete geospatial-temporal objects in a relational database (using built-in or extended spatial data engines), and employs spatiotemporal partitioning and indexing to enhance query efficiency. Simultaneously, GraST stores lightweight geospatial-temporal nodes in the graph database and links them to multi-granularity time tree and Geohash encoding nodes to enhance spatiotemporal aggregation capabilities. During query processing, user queries are broken down into graph semantic searches and geospatial calculations, pushed down to the graph and relational database for execution. Additionally, GraST adopts the two-phase commit protocol for cross-database data synchronization. We implemented a GraST prototype system by integrating PostGIS and Neo4j, and conducted performance evaluations and case studies on large-scale real-world datasets. Experimental results demonstrate that GraST shortens query response times by 1–2 orders of magnitude and offers flexible support for diverse geospatial-temporal semantic queries.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"137 ","pages":"Article 104437"},"PeriodicalIF":7.6,"publicationDate":"2025-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143464027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Spatiotemporal masked pre-training for advancing crop mapping on satellite image time series with limited labels

IF 7.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-02-22 DOI: 10.1016/j.jag.2025.104426

Xiaolei Qin , Haonan Guo , Xin Su , Zhenghui Zhao , Di Wang , Liangpei Zhang

Accurate crop mapping plays a critical role in optimizing agricultural monitoring and ensuring food security. Although data-driven deep learning methods have demonstrated success in crop mapping with satellite image time series (SITS) data, their promising performances heavily depend on labeled training samples. Nevertheless, the difficulty of annotating crop types often results in labeled data scarcity, leading to a decline in the model’s performance. Self-supervised learning (SSL) is a novel technique for crop mapping with limited labels. However, the existing SSL methods applied to SITS data typically explore masking solely on temporal dimension, which cannot guarantee strong spatial representation and therefore hinders the accurate prediction of complex crop fields. Furthermore, these methods sequentially extract spatial and temporal information without fully integrating information across different dimensions. In this study, we propose a spatiotemporal masking strategy for pre-training a SpatioTemporal Collaborative Learning Network (STCLN) to extract informative spatial and temporal representations from SITS data. Additionally, we design a SpatioTemporal Attention (STA) module in STCLN that integrates representations from spatial and temporal dimensions. The experimental results on two crop type mapping benchmarks encompassing various crop types demonstrate the outperformance of our proposed method. STCLN_wp outperforms the previous state-of-the-art (SOTA) methods with 6.49% higher mIoU on PASTIS dataset and 4.04% higher mIoU on MTLCC dataset. The ablation experiments on pre-training, masking strategies, and the STA module validate the effectiveness of our methodological design. Additionally, experiments conducted under varying sizes of the training set highlight the superior generalization ability of our method for crop type mapping in label-scarce situations. The code of our method is available at https://github.com/XiaoleiQinn/STCLN.

{"title":"Spatiotemporal masked pre-training for advancing crop mapping on satellite image time series with limited labels","authors":"Xiaolei Qin , Haonan Guo , Xin Su , Zhenghui Zhao , Di Wang , Liangpei Zhang","doi":"10.1016/j.jag.2025.104426","DOIUrl":"10.1016/j.jag.2025.104426","url":null,"abstract":"<div><div>Accurate crop mapping plays a critical role in optimizing agricultural monitoring and ensuring food security. Although data-driven deep learning methods have demonstrated success in crop mapping with satellite image time series (SITS) data, their promising performances heavily depend on labeled training samples. Nevertheless, the difficulty of annotating crop types often results in labeled data scarcity, leading to a decline in the model’s performance. Self-supervised learning (SSL) is a novel technique for crop mapping with limited labels. However, the existing SSL methods applied to SITS data typically explore masking solely on temporal dimension, which cannot guarantee strong spatial representation and therefore hinders the accurate prediction of complex crop fields. Furthermore, these methods sequentially extract spatial and temporal information without fully integrating information across different dimensions. In this study, we propose a spatiotemporal masking strategy for pre-training a SpatioTemporal Collaborative Learning Network (STCLN) to extract informative spatial and temporal representations from SITS data. Additionally, we design a SpatioTemporal Attention (STA) module in STCLN that integrates representations from spatial and temporal dimensions. The experimental results on two crop type mapping benchmarks encompassing various crop types demonstrate the outperformance of our proposed method. STCLN_wp outperforms the previous state-of-the-art (SOTA) methods with 6.49% higher mIoU on PASTIS dataset and 4.04% higher mIoU on MTLCC dataset. The ablation experiments on pre-training, masking strategies, and the STA module validate the effectiveness of our methodological design. Additionally, experiments conducted under varying sizes of the training set highlight the superior generalization ability of our method for crop type mapping in label-scarce situations. The code of our method is available at <span><span>https://github.com/XiaoleiQinn/STCLN</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"137 ","pages":"Article 104426"},"PeriodicalIF":7.6,"publicationDate":"2025-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143464028","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Detecting mass wasting of Retrogressive Thaw Slumps in spaceborne elevation models using deep learning

IF 7.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-02-22 DOI: 10.1016/j.jag.2025.104419

Kathrin Maier , Philipp Bernhard , Sophia Ly , Michele Volpi , Ingmar Nitze , Shiyi Li , Irena Hajnsek

Climate change has led to stronger warming in the Arctic, causing higher ground temperatures and extensive permafrost thaw. Retrogressive Thaw Slumps (RTSs) represent one of the most rapid and considerable geomorphological changes in permafrost regions, occurring when ice-rich permafrost is exposed and thaws. However, large-scale quantification of RTS-related mass wasting in Arctic permafrost landscapes is currently lacking, despite its importance to understand impacts on local environments and the global permafrost carbon cycle. Generating differential digital elevation models (dDEMs) from TanDEM-X single-pass Interferometric SAR (InSAR) observations enables us to quantify volume changes induced by rapid permafrost thaw. To extend this capability across the entire Arctic permafrost region, automation in data processing and RTS detection is essential. This study introduces a method that employs deep learning on InSAR-derived dDEMs to map RTSs and quantify volume changes from RTS activity. We chose eleven study sites with a total area of 71 400 km² to reflect the diverse character of Arctic environments for model training, testing, and inference. Our trained UNet++ model delivers a scalable solution for mapping RTSs and quantifying mass wasting towards a pan-Arctic scale, achieving segmentation accuracies of 0.58 (Intersection over Union) and classification accuracies of 0.75 (F1) on previously unseen test sites, with volume change estimates from model predictions being within

\pm

20% of the actual values. We found a total of almost 5000 RTSs active between 2010 and 2021 with volume change rates between 40.75 m³yr⁻¹km

^{2}

for sites in the Siberian to 1164.11 m³yr⁻¹km

^{2}

in the Canadian Arctic.

{"title":"Detecting mass wasting of Retrogressive Thaw Slumps in spaceborne elevation models using deep learning","authors":"Kathrin Maier , Philipp Bernhard , Sophia Ly , Michele Volpi , Ingmar Nitze , Shiyi Li , Irena Hajnsek","doi":"10.1016/j.jag.2025.104419","DOIUrl":"10.1016/j.jag.2025.104419","url":null,"abstract":"<div><div>Climate change has led to stronger warming in the Arctic, causing higher ground temperatures and extensive permafrost thaw. Retrogressive Thaw Slumps (RTSs) represent one of the most rapid and considerable geomorphological changes in permafrost regions, occurring when ice-rich permafrost is exposed and thaws. However, large-scale quantification of RTS-related mass wasting in Arctic permafrost landscapes is currently lacking, despite its importance to understand impacts on local environments and the global permafrost carbon cycle. Generating differential digital elevation models (dDEMs) from TanDEM-X single-pass Interferometric SAR (InSAR) observations enables us to quantify volume changes induced by rapid permafrost thaw. To extend this capability across the entire Arctic permafrost region, automation in data processing and RTS detection is essential. This study introduces a method that employs deep learning on InSAR-derived dDEMs to map RTSs and quantify volume changes from RTS activity. We chose eleven study sites with a total area of 71 400km<sup>2</sup> to reflect the diverse character of Arctic environments for model training, testing, and inference. Our trained UNet++ model delivers a scalable solution for mapping RTSs and quantifying mass wasting towards a pan-Arctic scale, achieving segmentation accuracies of 0.58 (Intersection over Union) and classification accuracies of 0.75 (F1) on previously unseen test sites, with volume change estimates from model predictions being within <span><math><mo>±</mo></math></span> 20% of the actual values. We found a total of almost 5000 RTSs active between 2010 and 2021 with volume change rates between 40.75m<sup>3</sup>yr<sup>−1</sup>km<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span> for sites in the Siberian to 1164.11m<sup>3</sup>yr<sup>−1</sup>km<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span> in the Canadian Arctic.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"137 ","pages":"Article 104419"},"PeriodicalIF":7.6,"publicationDate":"2025-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143464026","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Estimating two-decadal variations of global oceanic particulate organic carbon using satellite observations and machine learning approaches

IF 7.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-02-21 DOI: 10.1016/j.jag.2025.104438

Wenyue Jiao , Shengqiang Wang , Deyong Sun , Shuyan Lang , Yongjun Jia , Lulu Wang

Particulate organic carbon (POC) is fundamental to the marine carbon cycle, yet accurately estimating its concentration from satellite data remains challenging. In this study, we developed a novel machine learning framework that incorporates multiple data streams, covering apparent and inherent optical properties, biological indicators, and environmental variables, to improve global POC retrieval. Our model achieved high accuracy, with a Spearman’s correlation coefficient of 0.92, root-mean-square error of 68.46 mg m⁻³, and median absolute percentage error of 25.01 %, outperforming conventional algorithms that rely solely on remote sensing reflectance. Applying our approach to a long-term satellite dataset (1997–2023), we identified four major seasonal variation patterns across different oceanic regions: In high-latitude regions (Type 1), POC peaks in summer due to increased light availability, while mid-latitudes (Type 2) exhibit a stable pattern with a spring peak driven by water mixing and favorable sea surface temperatures. In the Equatorial Atlantic and Indian Oceans (Type 3), a spring trough and autumn peak suggest potential significant wind-driven nutrient inputs, whereas the Equatorial Pacific (Type 4) maintains high POC levels year-round, likely influenced by persistent upwelling and nutrient dynamics. These findings highlight the advantages of integrating machine learning for improved POC estimations and provides deeper insights into global POC dynamics.

{"title":"Estimating two-decadal variations of global oceanic particulate organic carbon using satellite observations and machine learning approaches","authors":"Wenyue Jiao , Shengqiang Wang , Deyong Sun , Shuyan Lang , Yongjun Jia , Lulu Wang","doi":"10.1016/j.jag.2025.104438","DOIUrl":"10.1016/j.jag.2025.104438","url":null,"abstract":"<div><div>Particulate organic carbon (POC) is fundamental to the marine carbon cycle, yet accurately estimating its concentration from satellite data remains challenging. In this study, we developed a novel machine learning framework that incorporates multiple data streams, covering apparent and inherent optical properties, biological indicators, and environmental variables, to improve global POC retrieval. Our model achieved high accuracy, with a Spearman’s correlation coefficient of 0.92, root-mean-square error of 68.46 mg m<sup>−3</sup>, and median absolute percentage error of 25.01 %, outperforming conventional algorithms that rely solely on remote sensing reflectance. Applying our approach to a long-term satellite dataset (1997–2023), we identified four major seasonal variation patterns across different oceanic regions: In high-latitude regions (Type 1), POC peaks in summer due to increased light availability, while mid-latitudes (Type 2) exhibit a stable pattern with a spring peak driven by water mixing and favorable sea surface temperatures. In the Equatorial Atlantic and Indian Oceans (Type 3), a spring trough and autumn peak suggest potential significant wind-driven nutrient inputs, whereas the Equatorial Pacific (Type 4) maintains high POC levels year-round, likely influenced by persistent upwelling and nutrient dynamics. These findings highlight the advantages of integrating machine learning for improved POC estimations and provides deeper insights into global POC dynamics.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"137 ","pages":"Article 104438"},"PeriodicalIF":7.6,"publicationDate":"2025-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143455047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

LLM-enhanced disaster geolocalization using implicit geoinformation from multimodal data: A case study of Hurricane Harvey

IF 7.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-02-21 DOI: 10.1016/j.jag.2025.104423

Wenping Yin , Yong Xue , Ziqi Liu , Hao Li , Martin Werner

Timely and accurate geolocalization of natural disasters is crucial for effective emergency response, which is foundational for risk mitigation and resilience development. Although social media texts have been widely used to recognize and resolve disaster geolocations, the implicit geoinformation in social media images remains largely underexplored. In this paper, we propose a novel large language model (LLM)-enhanced disaster geolocalization method that considers both explicit and implicit geoinformation from multimodal data. Based on the recognition of geolocations in disaster-related images and texts, geolocalization results were obtained by combining LLMs with map services. The selection of a geolocalization strategy depends on the available geoinformation modality and the presence of spatial relationships. A multimodal dataset of 1,000 images and 1,000 texts from the Hurricane Harvey Twitter dataset was constructed to evaluate geolocalization accuracy through error distance. The results demonstrated that the proposed method achieves significant improvements over baseline geocoding and toponym retrieval methods, with overall accuracies of 81.45%, 78.40%, 74.60%, 65.20%, and 44.95% within 161, 100, 50, 10, and 1 km, respectively. These findings confirm the potential of LLMs in enhancing geolocalization by considering implicit geoinformation from multimodal data for future disaster response and broader GeoAI applications.

{"title":"LLM-enhanced disaster geolocalization using implicit geoinformation from multimodal data: A case study of Hurricane Harvey","authors":"Wenping Yin , Yong Xue , Ziqi Liu , Hao Li , Martin Werner","doi":"10.1016/j.jag.2025.104423","DOIUrl":"10.1016/j.jag.2025.104423","url":null,"abstract":"<div><div>Timely and accurate geolocalization of natural disasters is crucial for effective emergency response, which is foundational for risk mitigation and resilience development. Although social media texts have been widely used to recognize and resolve disaster geolocations, the implicit geoinformation in social media images remains largely underexplored. In this paper, we propose a novel large language model (LLM)-enhanced disaster geolocalization method that considers both explicit and implicit geoinformation from multimodal data. Based on the recognition of geolocations in disaster-related images and texts, geolocalization results were obtained by combining LLMs with map services. The selection of a geolocalization strategy depends on the available geoinformation modality and the presence of spatial relationships. A multimodal dataset of 1,000 images and 1,000 texts from the Hurricane Harvey Twitter dataset was constructed to evaluate geolocalization accuracy through error distance. The results demonstrated that the proposed method achieves significant improvements over baseline geocoding and toponym retrieval methods, with overall accuracies of 81.45%, 78.40%, 74.60%, 65.20%, and 44.95% within 161, 100, 50, 10, and 1 km, respectively. These findings confirm the potential of LLMs in enhancing geolocalization by considering implicit geoinformation from multimodal data for future disaster response and broader GeoAI applications.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"137 ","pages":"Article 104423"},"PeriodicalIF":7.6,"publicationDate":"2025-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143464025","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Augmenting estuary monitoring from space: New retrievals of fine-scale CDOM quality and DOC exchange

IF 7.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-02-20 DOI: 10.1016/j.jag.2025.104389

Alana Menendez , Maria Tzortziou

Fueled by both terrestrial and marine inputs, estuaries worldwide are important biogeochemical reactors, strongly susceptible to natural and anthropogenic, episodic, or compounding disturbances. Satellite sensors provide a unique vantage point to capture estuarine processes at scales not feasible with in situ sampling alone; yet, remote sensing retrievals of estuarine biogeochemical dynamics remain challenging. Here, we developed new algorithms for high spatial resolution satellite retrievals of colored dissolved organic matter (CDOM) and dissolved organic carbon (DOC)—two key indicators of estuarine water quality and biogeochemical state. CDOM algorithms were optimized for Long Island Sound—one of the world’s most heavily urbanized estuaries—and included retrievals of CDOM absorption at 300 nm (a_CDOM(300)), a proxy for CDOM amount, as well as absorption spectral slope in the 275–295 nm range (S_275-295), a proxy for CDOM quality. Algorithms were specifically designed for Sentinel-2A/2B MSI and Landsat-8/9 OLI, a constellation of high spatial resolution sensors that provides coverage of the most dynamic zones for estuary exchange. MSI and OLI a_CDOM(300) and S_275-295 were most successfully retrieved using machine learning (ML) random forest regression, with the input features of remote sensing reflectance bands, band ratios, and month of acquisition. Satellite retrieval of DOC concentrations relied on a tight, widely applicable relationship between a_CDOM(300) and S_275-295. Timeseries generated for Long Island Sound, and its most impaired water quality region in the Western Narrows, revealed strong CDOM spatiotemporal dynamics associated with seasonal freshwater discharge, tidal wetland carbon export, recurring wastewater pollution, and episodic extreme events. Results highlight the value of these new ecosystem-scale observations for enhanced, sustainable estuarine management.

{"title":"Augmenting estuary monitoring from space: New retrievals of fine-scale CDOM quality and DOC exchange","authors":"Alana Menendez , Maria Tzortziou","doi":"10.1016/j.jag.2025.104389","DOIUrl":"10.1016/j.jag.2025.104389","url":null,"abstract":"<div><div>Fueled by both terrestrial and marine inputs, estuaries worldwide are important biogeochemical reactors, strongly susceptible to natural and anthropogenic, episodic, or compounding disturbances. Satellite sensors provide a unique vantage point to capture estuarine processes at scales not feasible with <em>in situ</em> sampling alone; yet, remote sensing retrievals of estuarine biogeochemical dynamics remain challenging. Here, we developed new algorithms for high spatial resolution satellite retrievals of colored dissolved organic matter (CDOM) and dissolved organic carbon (DOC)—two key indicators of estuarine water quality and biogeochemical state. CDOM algorithms were optimized for Long Island Sound—one of the world’s most heavily urbanized estuaries—and included retrievals of CDOM absorption at 300 nm (a<sub>CDOM</sub>(300)), a proxy for CDOM amount, as well as absorption spectral slope in the 275–295 nm range (S<sub>275-295</sub>), a proxy for CDOM quality. Algorithms were specifically designed for Sentinel-2A/2B MSI and Landsat-8/9 OLI, a constellation of high spatial resolution sensors that provides coverage of the most dynamic zones for estuary exchange. MSI and OLI a<sub>CDOM</sub>(300) and S<sub>275-295</sub> were most successfully retrieved using machine learning (ML) random forest regression, with the input features of remote sensing reflectance bands, band ratios, and month of acquisition. Satellite retrieval of DOC concentrations relied on a tight, widely applicable relationship between a<sub>CDOM</sub>(300) and S<sub>275-295</sub>. Timeseries generated for Long Island Sound, and its most impaired water quality region in the Western Narrows, revealed strong CDOM spatiotemporal dynamics associated with seasonal freshwater discharge, tidal wetland carbon export, recurring wastewater pollution, and episodic extreme events. Results highlight the value of these new ecosystem-scale observations for enhanced, sustainable estuarine management.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"137 ","pages":"Article 104389"},"PeriodicalIF":7.6,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143445878","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An enhanced image stacks method for mapping long-term retrogressive thaw slumps in the Tibetan Plateau

IF 7.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-02-20 DOI: 10.1016/j.jag.2025.104414

Jiapei Ma, Genxu Wang, Shouqin Sun, Chunlin Song, Jinlong Li, Linmao Guo, Kai Li, Peng Huang, Shan Lin

Retrogressive thaw slumps (RTSs) are severe manifestations of permafrost degradation with profound implications for regional environments and ecosystems. Previous studies heavily rely on high-resolution imagery and deep learning methods for RTS mapping. However, the acquisition of high-resolution imagery and the extensive computation of the deep learning-based method present challenges for long-term and large-scale monitoring. The image stacks method can overcome the defects of deep learning but is not sensitive in low-productivity ecosystems. This study proposes a feature-enhanced image stacks method. Instead of utilizing the Normalized Difference Vegetation Index (NDVI) directly in time-series change detection, the method employs the ratio of NDVI to its background value to enhance weak RTS signals caused by little change in NDVI or climate variation. A case study applied in the source region of the Yangtze River (SRYR), Tibetan Plateau, indicates that the method can amplify the RTS signal by more than 50 %, yielding accuracy slightly lower than the deep learning methods based on high-resolution imagery, but with a speed advantage of nearly an order of magnitude. The overall precision is 0.74, the F1 score is 0.73, and the maximum Intersection over Union (IOU) is 0.8. The delineation of RTSs takes about half an hour for the entire study area (158,000 km²), even with relatively low hardware specifications. Besides, the experiment conducted in the Horton Delta in the Arctic also demonstrates a good generalization of the method, with signal enhancement exceeding 80 %. This study confirms the feasibility of using medium-resolution data for long-term and large-scale RTS monitoring and will contribute to understanding the impact of climate change on permafrost dynamics in cold regions.

{"title":"An enhanced image stacks method for mapping long-term retrogressive thaw slumps in the Tibetan Plateau","authors":"Jiapei Ma, Genxu Wang, Shouqin Sun, Chunlin Song, Jinlong Li, Linmao Guo, Kai Li, Peng Huang, Shan Lin","doi":"10.1016/j.jag.2025.104414","DOIUrl":"10.1016/j.jag.2025.104414","url":null,"abstract":"<div><div>Retrogressive thaw slumps (RTSs) are severe manifestations of permafrost degradation with profound implications for regional environments and ecosystems. Previous studies heavily rely on high-resolution imagery and deep learning methods for RTS mapping. However, the acquisition of high-resolution imagery and the extensive computation of the deep learning-based method present challenges for long-term and large-scale monitoring. The image stacks method can overcome the defects of deep learning but is not sensitive in low-productivity ecosystems. This study proposes a feature-enhanced image stacks method. Instead of utilizing the Normalized Difference Vegetation Index (NDVI) directly in time-series change detection, the method employs the ratio of NDVI to its background value to enhance weak RTS signals caused by little change in NDVI or climate variation. A case study applied in the source region of the Yangtze River (SRYR), Tibetan Plateau, indicates that the method can amplify the RTS signal by more than 50 %, yielding accuracy slightly lower than the deep learning methods based on high-resolution imagery, but with a speed advantage of nearly an order of magnitude. The overall precision is 0.74, the F1 score is 0.73, and the maximum Intersection over Union (IOU) is 0.8. The delineation of RTSs takes about half an hour for the entire study area (158,000 km<sup>2</sup>), even with relatively low hardware specifications. Besides, the experiment conducted in the Horton Delta in the Arctic also demonstrates a good generalization of the method, with signal enhancement exceeding 80 %. This study confirms the feasibility of using medium-resolution data for long-term and large-scale RTS monitoring and will contribute to understanding the impact of climate change on permafrost dynamics in cold regions.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"137 ","pages":"Article 104414"},"PeriodicalIF":7.6,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143445879","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Estimation of fractional cover based on NDVI-VISI response space using visible-near infrared satellite imagery

IF 7.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-02-19 DOI: 10.1016/j.jag.2025.104432

Zhaoyang Han , Qingjiu Tian , Jia Tian , Tianyu Zhao , Chenglong Xu , Qing Zhou

Remote sensing observations of green vegetation (GV), impervious surface (IS), and bare soil (BS) fractional cover are essential for understanding climate change, characterizing ecosystem functions, monitoring urbanization process. As an important indicator of urbanization, the continuous increase of impervious surfaces alters the radiative transfer process at the surface, causing a series of environmental problems. Therefore, timely and accurate monitoring of the spatial and temporal changes in impervious surfaces and their impact on the ecological environment is of great significance for a comprehensive understanding of the process of urbanization as well as for the planning and construction of future cities. This study aims to propose a generalized method for the accurate estimation of GV, IS, and BS coverage. In this study, the visible impervious surface index (VISI),

{(B}_{r} - B_{g}) / {(B}_{r} + B_{g})

, was developed using measured spectral data of GV, IS, and BS, and analyzing their spectral characteristics to determine the spectral bands where they can be distinguished. Furthermore, the VISI combined with the NDVI was utilized to establish a triangular space for linear unmixing of the satellite image data to estimate the coverage of its GV, IS, and BS. Finally, the generalizability of this method was verified using UAV and satellite image data, with pearson correlation coefficient > 0.69. The results demonstrate that the VISI index proposed in this study is feasible for long-term series of multispectral imagery and large-scale coverage estimation.

{"title":"Estimation of fractional cover based on NDVI-VISI response space using visible-near infrared satellite imagery","authors":"Zhaoyang Han , Qingjiu Tian , Jia Tian , Tianyu Zhao , Chenglong Xu , Qing Zhou","doi":"10.1016/j.jag.2025.104432","DOIUrl":"10.1016/j.jag.2025.104432","url":null,"abstract":"<div><div>Remote sensing observations of green vegetation (GV), impervious surface (IS), and bare soil (BS) fractional cover are essential for understanding climate change, characterizing ecosystem functions, monitoringurbanization process. As an important indicator of urbanization, the continuous increase of impervious surfaces alters the radiative transfer process at the surface, causing a series of environmental problems. Therefore, timely and accurate monitoring of the spatial and temporal changes in impervious surfaces and their impact on the ecological environment is of great significance for a comprehensive understanding of the process of urbanization as well as for the planning and construction of future cities. This study aims to propose a generalized method for the accurate estimation of GV, IS, and BS coverage. In this study, the visible impervious surface index (VISI), <span><math><mrow><msub><mrow><mo>(</mo><mi>B</mi></mrow><mi>r</mi></msub><mo>-</mo><msub><mi>B</mi><mi>g</mi></msub><mrow><mo>)</mo><mo>/</mo><msub><mrow><mo>(</mo><mi>B</mi></mrow><mi>r</mi></msub><mo>+</mo><msub><mi>B</mi><mi>g</mi></msub><mo>)</mo></mrow></mrow></math></span>, was developed using measured spectral data of GV, IS, and BS, and analyzing their spectral characteristics to determine the spectral bands where they can be distinguished. Furthermore, the VISI combined with the NDVI was utilized to establish a triangular space for linear unmixing of the satellite image data to estimate the coverage of its GV, IS, and BS. Finally, the generalizability of this method was verified using UAV and satellite image data, with pearson correlation coefficient > 0.69. The results demonstrate that the VISI index proposed in this study is feasible for long-term series of multispectral imagery and large-scale coverage estimation.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"137 ","pages":"Article 104432"},"PeriodicalIF":7.6,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143445877","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

PolSAR image classification using complex-valued multiscale attention vision transformer (CV-MsAtViT)

IF 7.6 Q1 REMOTE SENSING

International journal of applied earth observation and geoinformation : ITC journal

Pub Date : 2025-02-18 DOI: 10.1016/j.jag.2025.104412

Mohammed Q. Alkhatib

This paper Introduces a novel method for Polarimetric Synthetic Aperture Radar (PolSAR) image classification using a Complex-Valued Multiscale Attention Vision Transformer (CV-MsAtViT). The model incorporates a complex-valued multiscale feature fusion mechanism, a complex-valued attention block, and a Complex-Valued Vision Transformer (CV-ViT) to effectively capture spatial and polarimetric features from PolSAR data. The multiscale fusion block enhances feature extraction, while the attention mechanism prioritizes critical features, and the CV-ViT processes data in the complex domain, preserving both amplitude and phase information. Experimental results on benchmark PolSAR datasets, including Flevoland, San Francisco, and Oberpfaffenhofen, show that CV-MsAtViT achieves superior classification accuracy, with an overall accuracy (OA) of 98.35% on the Flevoland dataset, outperforming state-of-the-art models like PolSARFormer. The model also demonstrates efficient computational performance, minimizing the number of parameters while preserving high accuracy. These results confirm that CV-MsAtViT effectively enhances the classification of PolSAR images by leveraging complex-valued data processing, offering a promising direction for future advancements in remote sensing and complex-valued deep learning.

The codes associated with this paper are publicly available at https://github.com/mqalkhatib/CV-MsAtViT.

{"title":"PolSAR image classification using complex-valued multiscale attention vision transformer (CV-MsAtViT)","authors":"Mohammed Q. Alkhatib","doi":"10.1016/j.jag.2025.104412","DOIUrl":"10.1016/j.jag.2025.104412","url":null,"abstract":"<div><div>This paper Introduces a novel method for Polarimetric Synthetic Aperture Radar (PolSAR) image classification using a Complex-Valued Multiscale Attention Vision Transformer (CV-MsAtViT). The model incorporates a complex-valued multiscale feature fusion mechanism, a complex-valued attention block, and a Complex-Valued Vision Transformer (CV-ViT) to effectively capture spatial and polarimetric features from PolSAR data. The multiscale fusion block enhances feature extraction, while the attention mechanism prioritizes critical features, and the CV-ViT processes data in the complex domain, preserving both amplitude and phase information. Experimental results on benchmark PolSAR datasets, including Flevoland, San Francisco, and Oberpfaffenhofen, show that CV-MsAtViT achieves superior classification accuracy, with an overall accuracy (OA) of 98.35% on the Flevoland dataset, outperforming state-of-the-art models like PolSARFormer. The model also demonstrates efficient computational performance, minimizing the number of parameters while preserving high accuracy. These results confirm that CV-MsAtViT effectively enhances the classification of PolSAR images by leveraging complex-valued data processing, offering a promising direction for future advancements in remote sensing and complex-valued deep learning.</div><div>The codes associated with this paper are publicly available at <span><span>https://github.com/mqalkhatib/CV-MsAtViT</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"137 ","pages":"Article 104412"},"PeriodicalIF":7.6,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143428850","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0