International journal of applied earth observation and geoinformation : ITC journal最新文献

{"title":"Structural stability of vegetation greening across India from multi-season MODIS EVI: a time–frequency earth observation framework","authors":"Dibyendu Dutta","doi":"10.1016/j.jag.2026.105576","DOIUrl":"10.1016/j.jag.2026.105576","url":null,"abstract":"<div><div>Large-scale vegetation greening has been widely reported across India; however, national assessments have largely relied on monotonic trend metrics without explicitly evaluating seasonal variability, spatial heterogeneity, or the structural persistence of vegetation dynamics. This study presents a physiography-resolved, multi-season assessment of vegetation change using a 24-year (2001–2024) MODIS 16-day Enhanced Vegetation Index (EVI; 250 m) dataset spanning 14 physiographic subdivisions. Seasonal Mann–Kendall analysis and Sen’s slope estimation reveal widespread but spatially heterogeneous greening, with all physiographic subdivisions exhibiting significant positive winter trends and the strongest greening rates (0.0032–0.0051 yr⁻¹) over the Central Highlands, Northern Deccan, and Western Coastal Plains. Pixel-scale analysis indicates significant greening over 62.47 % of India during the pre-monsoon and 61.46 % during winter, declining to 34.57 % during the monsoon before partially recovering (∼43 %) during the post-monsoon, whereas browning remains limited (&lt;2.2 %) and spatially restricted. Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) further reveals that vegetation dynamics across most physiographic subdivisions are dominated by sub-interannual (6–24 month) variability, indicating that long-term greening is embedded within persistent low-frequency oscillations rather than short-term fluctuations. In contrast, high-frequency variability is concentrated over climatically extreme and hydroclimatically stressed regions. Seasonal Intrinsic Mode Function 3 (IMF3) energy anomalies identify monsoon-concentrated, off-monsoon-amplified, and mixed variability regimes, while a persistence-based stability metric demonstrates that statistically significant greening does not necessarily imply structural resilience. The Western Plain exhibits the highest structural persistence, whereas plateau, mountain, and humid eastern regions remain dominated by temporal instability despite positive greening trends. By integrating seasonal trend analysis, frequency-resolved decomposition, physiographic stratification, and land-use differentiation, this study provides a scalable Earth observation framework for distinguishing persistent vegetation regime from variability-dominated change across heterogeneous landscapes.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"153 ","pages":"Article 105576"},"PeriodicalIF":8.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148933337","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}

{"title":"Tracking diurnal aerosol dynamics and smoke transport over East Asia using multi-geostationary satellite fusion (GK-2B, Himawari-9, FY-4B)","authors":"Tong Sha,&nbsp;Zhenan Yao,&nbsp;Yueming Cheng,&nbsp;Tie Dai,&nbsp;Jing Wei,&nbsp;Guangyu Shi","doi":"10.1016/j.jag.2026.105595","DOIUrl":"10.1016/j.jag.2026.105595","url":null,"abstract":"<div><div>Continuous, high-frequency observations of aerosol optical depth (AOD) are crucial for understanding aerosol dynamics and regional air quality. However, individual geostationary (GEO) AOD products suffer from distinct spatiotemporal gaps and uncertainties driven by cloud screening, surface-reflectance assumptions, and strict algorithmic quality controls. To address these limitations, we propose a deep-learning fusion framework synergizing the spectral, temporal, and geometric complementarities of three GEO satellites (GK-2B, Himawari-9, and FY-4B) over East Asia. Central to our framework is a TabNet-based dynamic feature selection mechanism that adaptively weights heterogeneous multi-sensor inputs to suppress local retrieval errors. To improve spatial generalization, we implement a two-stage transfer learning strategy: pre-training with spatially extensive MODIS data to establish large-scale spatial priors, followed by fine-tuning with ground-based AERONET observations. Compared to single-satellite baselines, the fused AOD product from May 2024 to April 2025 improves accuracy and spatiotemporal coverage. Specifically, it yields an overall cross-validation correlation coefficient (R) of 0.94, reduces root-mean-square error (RMSE) by 47 %, and increases valid hourly matchups by 33.5 %. Independent-site validation shows consistent generalizability (R = 0.88). Furthermore, evaluations of extreme events, including severe biomass-burning and dust transport episodes, demonstrate that the fused AOD product can reconstruct hourly plume gradients during rapid transport. By improving the temporal characterization of complex aerosol environments, this deep learning-driven fusion approach provides a useful foundational dataset for future regional air-quality analyses and chemical transport model assimilation.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"153 ","pages":"Article 105595"},"PeriodicalIF":8.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148980149","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}

{"title":"A novel method for estimating urban building heights using bi-temporal high resolution satellite imagery and reprojected nDSM","authors":"Xiaoxue Feng,&nbsp;Li Tian,&nbsp;Peijun Li,&nbsp;Fang Chen,&nbsp;Changyong Dou","doi":"10.1016/j.jag.2026.105585","DOIUrl":"10.1016/j.jag.2026.105585","url":null,"abstract":"<div><div>Information on multi-temporal building heights in urban areas is crucial for understanding vertical urbanization patterns. In particular, acquiring building height data for historical periods is of great significance for studying long-term urban land changes. However, existing deep learning-based methods for building height estimation from monocular (single-view) imagery face two key limitations that hinder the retrieval of historical building heights: the difficulty in directly deriving building heights from off-nadir imagery (i.e., images acquired at a tilted viewing angle), and the substantial degradation in accuracy when applied to cross-temporal height estimation, where no height data for the target time period are available. This study proposes a novel method for estimating building heights for a target temporal period using bi-temporal (target and reference dates) off-nadir high spatial resolution (HSR) satellite images and a local reprojected normalized Digital Surface Model (nDSM) from reference date. The method consists of three main stages: generation of bi-temporal Line-of-Sight nDSMs (Los-nDSMs), estimation of target-temporal building heights using a dual-attention-based conditional Generative Adversarial Network (DA-cGAN), and height calibration. The proposed method was evaluated in two study areas using bi-temporal HSR images and local height data derived from LiDAR point cloud data for a reference date. For the first study area, the main urban center of Beijing, WorldView-2 images were used, while for the second, Tongzhou New City, Beijing, Pleiades images were used. Averaged over the two study areas, the proposed method outperforms two state-of-the-art methods relying on single temporal imagery and nDSM. It reduces the mean absolute errors (MAEs) by 3.42 m and 2.24 m, and the root mean square errors (RMSEs) by 4.95 m and 4.35 m, respectively, relative to the two competitors, while achieving a zero-mean normalized cross-correlation (ZNCC) of 0.92. The proposed method provides an effective solution for estimating building heights across different temporal periods in urban areas, using HSR images with varying viewing angles and from different sensors. It is particularly valuable in scenarios where historical height data are unavailable.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"153 ","pages":"Article 105585"},"PeriodicalIF":8.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148980351","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}

{"title":"Viewing-geometry effects on UAV-based tree height retrieval using BRDF-derived vegetation indices","authors":"Zhihui Mao,&nbsp;Lei Deng,&nbsp;Qinghua Qiao","doi":"10.1016/j.jag.2026.105537","DOIUrl":"10.1016/j.jag.2026.105537","url":null,"abstract":"<div><div>Tree height is a key descriptor of forest canopy structure, yet a single nadir view captures only part of the directional structural information contained in canopy reflectance. This study investigated whether viewing geometry provides directional information useful for UAV-based optical tree-height retrieval in a <em>Larix principis-rupprechtii</em> plantation. Multi-angular reflectance observations extracted from overlapping UAV multispectral images were used to fit the RossThick–LiSparseR (RTLSR) kernel-driven bidirectional reflectance distribution function (BRDF) model. The fitted model was then used to reconstruct directional reflectance at 27 prescribed viewing geometries along the solar principal plane, while UAV-LiDAR data provided grid-level reference tree heights. Nine vegetation indices were derived and evaluated using linear regression, random forest, and gradient boosting regressor under single- and multi-angle schemes. DVI, EVI, GDVI, TVI, and GOSAVI exhibited consistent angle-dependent sensitivity to tree height, with their strongest correlations and lowest retrieval errors generally occurring within the intermediate forward-viewing sector (30°–45°) in the BRDF-derived dataset. For example, the absolute correlation between GDVI and tree height increased from 0.429 at nadir to 0.507 at 40°. Among the tested index–scheme–model combinations, the lowest validation error was obtained using GDVI under the Forward + 0° scheme with random forest, reducing RMSE from 2.19 m at nadir to 1.79 m (RMSE% = 10.25 %). However, the relative performance of the tested angular schemes varied among models, indicating that angular configuration and retrieval algorithm jointly affected performance, while increasing the number of angular features alone did not consistently reduce retrieval error. Overall, the findings indicate that viewing geometry can provide useful directional information for within-site optical tree-height retrieval and that intermediate forward-viewing information may complement nadir observations.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"153 ","pages":"Article 105537"},"PeriodicalIF":8.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148852863","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}

{"title":"A multi-task collaborative method for individual tree segmentation and multi-dimensional attribute extraction in mountainous areas using deep learning technologies","authors":"Guangbin Lei,&nbsp;Ainong Li,&nbsp;Jinhu Bian,&nbsp;Zhengjian Zhang,&nbsp;Xiaohan Lin,&nbsp;Xi Nan,&nbsp;Amin Naboureh,&nbsp;Yulan Chen,&nbsp;Yi Deng","doi":"10.1016/j.jag.2026.105539","DOIUrl":"10.1016/j.jag.2026.105539","url":null,"abstract":"<div><div>Accurate mapping and monitoring of forest structure and composition in complex mountainous regions remain challenging because of crown occlusion, crown adhesion, and pronounced topographic variation. Moreover, current monitoring methods often focus on single-attribute retrieval and lack an integrated framework for collaborative multidimensional attribute extraction. To address these challenges, this study proposes a multi-task collaborative method for individual tree segmentation and integrated multidimensional attribute extraction in mountainous forests using deep learning and unmanned aerial vehicle (UAV) RGB imagery. A two-stage segmentation method combining a U-Net++ deep learning model with a marker-controlled watershed algorithm was developed, followed by a sequential transfer learning strategy to facilitate knowledge transfer among individual tree segmentation, tree height prediction, and forest type classification. The proposed method was evaluated over a 5 km × 5 km area within the Wanglang National Nature Reserve (WNNR) using manually delineated tree crowns and canopy height models (CHM) derived from UAV LiDAR data. The results showed that the two-stage segmentation method combined the semantic representation capability of deep learning with the geometric partitioning ability of the watershed algorithm, thereby reducing under-segmentation caused by crown occlusion and adhesion in mountainous forests. The omission rate decreased from 9.12 % to 4.67 %, while the F1-score increased from 0.9298 to 0.9537. Sequential transfer learning also improved individual-tree attribute prediction, with the RMSE of tree height prediction decreasing from 7.1504 m to 3.4825 m and the overall accuracy of forest type classification increasing from 81.68 % to 92.51 %. The proposed method enables integrated mapping of individual-tree location, crown width, tree height, and forest type in complex mountainous forests, providing detailed information for precision forest management and forest carbon assessment.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"153 ","pages":"Article 105539"},"PeriodicalIF":8.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148855274","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}

{"title":"Tropoformer-S: iTransformer-based spatiotemporal reconstruction of three-dimensional tropospheric delay fields integrating ERA5 reanalysis and GNSS observations","authors":"Fangxin Hu,&nbsp;Bo Chen,&nbsp;Mingyuan Zhang,&nbsp;Ran Lu,&nbsp;Pengfei Xia,&nbsp;Tengfeng Wang,&nbsp;Xiaowei Zou,&nbsp;Yuanfan Deng","doi":"10.1016/j.jag.2026.105553","DOIUrl":"10.1016/j.jag.2026.105553","url":null,"abstract":"<div><div>Spatially continuous three-dimensional tropospheric delay fields are essential for Earth observation and atmospheric sensing. The ERA5 reanalysis offers continuous coverage but is delayed by several days, whereas Global Navigation Satellite System (GNSS) observations are accurate and near-real-time but pointwise. This study develops Tropoformer-S, an iTransformer-based framework that reconstructs the regional delay field by integrating the two sources. The framework forecasts the ERA5-derived level-wise Zenith Tropospheric Delay (ZTD) at 15 pressure levels from 1000 to 550 hPa for 1 to 120 h, where the variate-token attention explicitly models the spatial correlation among grid points. The forecast field is converted into station ZTD through a physically guided retrieval chain and refined by GNSS-informed corrections along two complementary paths, harmonic self-correction for stations with local GNSS records and spatial reconstruction of the bias field for stations without. Over mainland China, the forecasts attain a level-mean RMSE of 2.84 cm against the ERA5-derived ZTD and outperform five alternative deep-learning architectures. The overall RMSE is 3.20 cm in 2022 and 3.23 cm in 2023 against GNSS, on average 18.64 % and 33.10 % below GPT3 and HGPT2, and the radiosonde evaluation shows a near-zero bias. The harmonic correction path lowers the RMSE to 2.79 and 2.98 cm and reduces the network bias to near zero, while the spatial correction path reaches 2.10 to 2.66 cm at withheld stations under nine network densities and largely removes the lead dependence. The framework thus fills the ERA5 latency gap with continuous tropospheric delay information for geodetic and atmospheric applications.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"153 ","pages":"Article 105553"},"PeriodicalIF":8.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148895346","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}

{"title":"Improving text–image alignment for referring remote sensing image segmentation guided by vision foundation model","authors":"Yijie Zhang,&nbsp;Jian Cheng,&nbsp;Shimiao Fan,&nbsp;Changjian Deng,&nbsp;Ziying Xia,&nbsp;Nyima Tashi","doi":"10.1016/j.jag.2026.105439","DOIUrl":"10.1016/j.jag.2026.105439","url":null,"abstract":"<div><div>The task of referring remote sensing image segmentation (RRSIS) aims to predict segmentation masks of target objects based on text descriptions. Effective alignment between textual and visual modalities is crucial for achieving high performance. However, most existing RRSIS methods overlook the inherent domain gap between the two modalities and naively fuse visual and text features. Moreover, they lack guidance from domain-specific priors during alignment, often leading to inaccurate segmentation results. In this paper, we propose a novel method for RRSIS, namely TIANet. This method combines implicit and explicit alignment approaches to achieve fine-grained image-text alignment. Furthermore, to compensate for the lack of domain priors during alignment, we leverage the Remote Sensing Vision Foundation Model (RS-VFM) to guide both alignment strategies. On one hand, the Prior-Guided Visual-Text Feature Alignment Module (PVTAM) leverages domain priors to fuse textual and visual features in a stage-wise manner, thereby achieving cross-modal implicit alignment at the feature level. On the other hand, the Dynamic Bidirectional Alignment and Prior Distillation (DBAPD) loss is adopted to implement explicit cross-modal alignment during training. Furthermore, to enhance the perception ability of small targets, the Kernel-Adaptive Small Target Enhancement Module (KASTEM) is used in the decoder stage. Extensive experiments on the RefSegRS and RRSIS-D datasets show that our TIANet achieves excellent performance compared with other advanced methods, with a 1.22% mIoU improvement over existing state-of-the-art approaches on the RefSegRS dataset.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"153 ","pages":"Article 105439"},"PeriodicalIF":8.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148895347","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}

{"title":"GICL-Net: A graph-interactive collaborative learning network for joint classification of HSI and LiDAR","authors":"Hao Zhu,&nbsp;Yulin Xie,&nbsp;Xiaotong Li,&nbsp;Huanhuan Wu,&nbsp;Zhongle Ren,&nbsp;Zhen Cao,&nbsp;Biao Hou","doi":"10.1016/j.jag.2026.105525","DOIUrl":"10.1016/j.jag.2026.105525","url":null,"abstract":"<div><div>Hyperspectral and LiDAR images, as a class of multi-source remote sensing data, have their own advantageous features. HSI contains rich spectral information, compared to LiDAR, which is more capable of reflecting the height information of the target area. However, determining the intrinsic correlation characteristics between HSI channels and fully exploiting the complementary characteristics of HSI and LiDAR to improve classification accuracy remains a challenging issue. In this paper, we propose a graph-interactive collaborative learning network (GICL-Net) for joint classification of HSI and LiDAR, which consists of a multi-scale spectral-spatial graph feature extractor (MSGE) and a cross-modality adjacency matrix fusion (CAMF) module. Specifically, MSGE is divided into two branches. One branch starts from the spectral dimension, defining each channel of HSI as a node to construct a graph network and mine potential correlations at the channel level, thereby capturing the intrinsic spectral correlation features of the target region. The other branch focuses on the height information of the target area, constructing a graph network at the geometric spatial level to extract representative height features. The CAMF module leverages the fused adjacency matrix to establish a cross-modal graph interaction mechanism, enabling the spectral features and the spatial features to achieve deep synergy through iterative updates in graph convolution. This generates cross-modal fusion features with fine spectral discrimination capability and spatial dimensional structural integrity. This method has been validated for effectiveness across multiple HSI and LiDAR datasets. In particular, on the Houston2013 dataset, the overall accuracy (OA) reached 98.91%, with a Kappa coefficient of 0.9882. Furthermore, compared with the baseline network (i.e., the conventional two-branch graph network, the overall accuracy improved by 6.58%. Our code is available at the link: <span><span>https://github.com/Xidian-AIGroup190726/GICLNet</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"153 ","pages":"Article 105525"},"PeriodicalIF":8.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148933344","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}

{"title":"Tree-neighborhood scale structure–radiation coupling analysis and modeling using multi-source remote sensing and ensemble learning","authors":"Guanjun Lian,&nbsp;Huaiqing Zhang,&nbsp;Hua Liu,&nbsp;Tingdong Yang,&nbsp;Hanqing Qiu,&nbsp;Yang Liu,&nbsp;Hongyan Yu,&nbsp;Xianying Wang","doi":"10.1016/j.jag.2026.105565","DOIUrl":"10.1016/j.jag.2026.105565","url":null,"abstract":"<div><div>Forest structure regulates the spatial heterogeneity of absorbed photosynthetically active radiation (APAR), thereby shaping ecosystem functioning and productivity. However, how neighborhood spatial configuration and competition dynamics control APAR distribution remains poorly understood, particularly in structurally heterogeneous and topographically complex natural forests. To address this, we proposed FASAR (Forest Architecture–Solar Absorption Response), a structure–radiation coupling framework for complex-terrain natural forests. FASAR establishes an integrated workflow for structure–radiation coupling analysis by integrating multi-source LiDAR-based structural parameterization, Adaptive Leaf Synthesis Algorithm (ALSA)-assisted 3D forest reconstruction, 3D radiative transfer simulation, and structure-stratified ensemble modeling. It converts heterogeneous forest architecture into plot-, neighborhood-, and tree-level descriptors and links these descriptors with spatially explicit APAR patterns. Applied to 144 plots (5.76 ha, 3,417 trees) in a natural Qinghai spruce forest in western China, FASAR revealed that large trees absorb more light with vertically stable patterns, while canopy stratification, competition, and size disparity intensify vertical displacement of light and suppress central-tree absorption. Structural dominance and near-natural spatial configurations enhanced APAR interception (ρ = 0.08–0.49, <em>p</em> &lt; 0.001), while moderate aggregation and openness improved light scattering (ρ = 0.06–0.19, <em>p</em> &lt; 0.001). Slope (ρ = –0.37, p &lt; 0.001) and gap density (ρ = –0.18, p &lt; 0.05) showed dual facilitation–inhibition effects on mean APAR. Structure-specific Bagging models achieved moderate performance (R² = 0.457–0.546). Variable importance analysis identified crown area (CA) as the dominant driver across all structural types (contribution: 68.95–81.78 %), followed by tree height, spatial density, and other structural heterogeneity indices. This framework provides an operational pathway for quantifying how neighborhood-scale forest architecture regulates individual-tree light absorption under complex-terrain conditions. The structure–radiation mechanisms revealed in this study may also provide physically based constraints for APAR/fAPAR remote sensing retrieval and related research. FASAR has the potential to be extended across forest types and successional stages, with its performance further evaluated in future studies.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"153 ","pages":"Article 105565"},"PeriodicalIF":8.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148933348","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}

{"title":"SFMNet: Joint semantic and structural modeling via spatial–frequency collaboration for remote sensing scene segmentation","authors":"Wenwu Xiong,&nbsp;Wei Chen,&nbsp;Zhou Huang,&nbsp;Hongzhao Tang,&nbsp;Kevin Tansey,&nbsp;Józef Hernik,&nbsp;Binbin Lu,&nbsp;Shichao Cui,&nbsp;Yida Chen","doi":"10.1016/j.jag.2026.105541","DOIUrl":"10.1016/j.jag.2026.105541","url":null,"abstract":"<div><div>As the spatial resolution of remote sensing imagery continues to improve, Earth observation scenes exhibit greater diversity in spatial and spectral characteristics, together with increasingly complex surface structures across multiple scales. In such scenarios, global land-cover patterns and local high-frequency details are often strongly coupled, posing significant challenges for accurate scene interpretation and boundary delineation. From a feature representation perspective, spatial-domain features are effective for modeling global context and high-level semantics, whereas frequency-domain representations facilitate the separation of structural components at different scales and enhance fine boundary details, but have limited capability in explicitly capturing semantic relationships. Consequently, most existing semantic segmentation methods operate in a single domain, either spatial or frequency, which restricts their ability to jointly model semantic consistency and diverse spatial structures. To address this issue, we propose a semantic segmentation network termed Spatial–Frequency Collaborative Modeling Network (SFMNet), which jointly learns complementary representations in both spatial and frequency domains. By leveraging spatial-domain semantic priors to guide frequency-domain structural enhancement, SFMNet enables coordinated modeling of land-cover semantics and multi-scale surface structures. In addition, cross-scale interaction between low- and high-frequency components and adaptive multi-level feature fusion are introduced to improve structural consistency across hierarchical feature representations. Extensive experiments on the ISPRS Potsdam, Vaihingen, and LoveDA datasets demonstrate that SFMNet consistently outperforms existing methods in both overall segmentation accuracy and boundary delineation quality. Specifically, SFMNet achieves mean Intersection-over-Union scores of 81.30%, 70.23%, and 55.07%, and Boundary mean Intersection-over-Union scores of 65.42%, 56.76%, and 43.74% on the three datasets, respectively.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"153 ","pages":"Article 105541"},"PeriodicalIF":8.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148884169","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}