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

{"title":"Bridging the visual-computational gap: transforming UAV reality meshes into volumetric geological models via explicit-implicit integration","authors":"Zhibin Liu,&nbsp;Jiateng Guo,&nbsp;Yuxin Peng,&nbsp;Zirui Zhang,&nbsp;Xulei Wang,&nbsp;Hongwei Li,&nbsp;Luyuan Wang,&nbsp;Liangkai Cheng,&nbsp;Kangning Li","doi":"10.1016/j.jag.2026.105420","DOIUrl":"10.1016/j.jag.2026.105420","url":null,"abstract":"<div><div>Targeting complex geological environments characterized by high-steep cliffs, such as Danxia landforms, traditional geological modeling often relies on sparse subsurface detection data and low-precision Digital Elevation Models (DEM), making it difficult to accurately depict the complete 3D morphology of both the surface and the subsurface. To overcome this limitation, this paper introduces large-area, high-precision UAV Oblique Photogrammetry (OSGB) remote sensing data and proposes an OSGB-enhanced explicit–implicit integrated modeling framework. This method aims to bridge the “Visual-Computational” gap, where raw OSGB data possesses excellent visual fidelity but essentially consists of unstructured, non-manifold surface meshes, posing severe topological challenges for direct Boolean operations and volumetric calculations. Rather than merely applying existing algorithms, the methodological innovation lies in formulating a coupled architecture where explicit high-precision surface meshes mathematically constrain the implicit volumetric fields, bridging the distinct mathematical spaces of computer graphics and geostatistics. The approach first converts the OSGB reality model into high-quality explicit geometric constraints through topology repair and mesh optimization, combining Hermite Radial Basis Functions (HRBF) to establish a high-precision peak model. Subsequently, the Skinning Method is adopted to represent rivers and Quaternary thin layers, and a Dual Implicit Function managed by a binary tree is introduced to accurately resolve complex fault networks. Finally, within a unified tetrahedral grid space, a Boolean cutting process strictly following the geological evolutionary sequence is established. Application to the core area of Mount Danxia demonstrates that this framework successfully transforms large-area photogrammetric meshes into 3D geological models, realizing a transition from traditional DEM-constrained modeling to the integration of OSGB-derived, geometrically detailed surface constraints with deep geological attributes. Cross-section analysis indicates a high degree of consistency between the model and stratigraphic data. This study provides a universal technical paradigm for high-precision refined modeling in complex orogenic belts and special geomorphic areas.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"151 ","pages":"Article 105420"},"PeriodicalIF":8.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148289457","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":"AFDINet: Adaptive frequency disentanglement and interaction in hybrid Mamba–CNN for optical-SAR land cover classification","authors":"Bo Ren,&nbsp;Jin Gu,&nbsp;Chen Yang,&nbsp;Biao Hou,&nbsp;Qianfang Wang,&nbsp;Zitong Wu,&nbsp;Licheng Jiao","doi":"10.1016/j.jag.2026.105406","DOIUrl":"10.1016/j.jag.2026.105406","url":null,"abstract":"<div><div>The synergistic integration of optical and Synthetic Aperture Radar (SAR) imagery enables robust Earth observation by overcoming the limitations of single-sensor modalities, such as atmospheric occlusion and spectral deficiency. However, prevailing fusion paradigms typically treat multi-modal integration as a homogeneous spatial operation, neglecting the distinct physical properties inherent in different frequency domains. This oversight often results in the loss of high-frequency (HF) details, leading to fragmented boundaries in fine-grained land cover classification (LCC). To bridge this gap, this study introduces the Adaptive Frequency Disentanglement and Interaction Network (AFDINet), a hybrid architecture fusing CNN and Mamba. We devise a Learnable Frequency Disentanglement (LFD) module that reformulates wavelet decomposition as a data-driven process, adaptively separating structural consensus from modality-specific textures. Subsequently, a High–Low Frequency Fusion (HLF<span><math><msup><mrow></mrow><mrow><mn>3</mn></mrow></msup></math></span>) module implements a role-aware interaction strategy, calibrating low-frequency (LF) geometry while enhancing HF details. Extensive empirical evaluations on the Xi’an, YYX-OPT-SAR, and WHU-OPT-SAR datasets demonstrate that AFDINet achieves state-of-the-art (SOTA) performance in terms of OA, mIoU, and Kappa, while exhibiting exceptional resilience to cloud interference.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"151 ","pages":"Article 105406"},"PeriodicalIF":8.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148289458","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":"Construction of high spatiotemporal resolution beach SBT-DEM using multi-source data fusion","authors":"Zhuocheng Liu ,&nbsp;Lianqiang Shi ,&nbsp;Junli Guo ,&nbsp;Qinghua Ye ,&nbsp;Haijiang Liu ,&nbsp;Wenliang Zhang ,&nbsp;Yu Peng ,&nbsp;Huangzhe Qi ,&nbsp;Tianshui Cao","doi":"10.1016/j.jag.2026.105359","DOIUrl":"10.1016/j.jag.2026.105359","url":null,"abstract":"<div><div>Sandy coastlines are increasingly threatened by severe erosion, necessitating high-resolution topographic data for effective coastal management. Traditional ground surveys are limited by high costs and inefficiency, while remote sensing inversion techniques often suffer from low resolution and accuracy. This study presents a multi-platform collaborative observation and data fusion approach to construct a seamless bathymetric and digital elevation model (SBT-DEM) with enhanced accuracy and resolution. The approach integrates terrestrial laser scanning, shore-based video imaging, and an unmanned surface vessel with single-beam echo sounders to capture topography across the backshore, foreshore, and submerged zones. Reliability was assessed using Global Navigation Satellite System Real-Time Kinematic measurements. A random forest regression further refined the foreshore topography, reducing the root mean square error (RMSE) by about 0.1 m. The resulting SBT-DEM achieves high spatial–temporal resolution and accuracy (mean absolute error = 0.106 m, RMSE = 0.141 m), providing a robust dataset for coastal monitoring and erosion control.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"150 ","pages":"Article 105359"},"PeriodicalIF":8.6,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148179569","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}

{"title":"Aridity thresholds Reveal transitions in dominant drivers of partitioned Water–Carbon coupling across Central Asian drylands","authors":"Shiran Song ,&nbsp;Xi Chen ,&nbsp;Akylbek Kurishbayev ,&nbsp;Philippe De Maeyer ,&nbsp;Chuan Wang ,&nbsp;Chanjuan Zan ,&nbsp;Shukhrat Shokirov ,&nbsp;Tie Liu","doi":"10.1016/j.jag.2026.105344","DOIUrl":"10.1016/j.jag.2026.105344","url":null,"abstract":"<div><div>Intensifying global aridification is reshaping the water–carbon balance in dryland ecosystems, yet when and how this coupling reaches critical thresholds remains poorly understood—posing major uncertainties for predicting ecosystem resilience under climate change. Using Central Asia as a representative dryland region, we partitioned ecosystem water-use efficiency (WUE = GPP/ET) into transpiration efficiency (WUEc = GPP/Ec) and the transpiration-to-evapotranspiration ratio (TET = Ec/ET) to probe threshold responses. We identified distinct aridity thresholds associated with nonlinear shifts in water–carbon coupling: WUE and WUEc declined and then rebounded beyond aridity levels of 0.75 and 0.81, respectively, indicating tighter coupling between carbon gain and water loss under severe climatic stress. In contrast, ET, Ec, GPP, and TET declined sharply at thresholds of 0.80, 0.85, 0.86, and 0.88, indicating broader constraints on ecosystem carbon assimilation and hydrological functioning. SHAP-based machine learning revealed threshold-dependent shifts in ecosystem controls, with vegetation traits dominating carbon–water coupling under lower aridity, and hydroclimatic stressors—including soil moisture, precipitation, temperature, and vapor pressure deficit—emerging as dominant regulators beyond critical thresholds. Notably, temperature was positively associated with WUE and WUEc, while VPD exerted persistent negative effects post-threshold. These findings characterize nonlinear, threshold-dependent changes in dryland carbon–water coupling and highlight the need to evaluate apparent efficiency gains in the context of declining ecosystem fluxes under future drought intensification.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"150 ","pages":"Article 105344"},"PeriodicalIF":8.6,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148179565","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}

{"title":"Bridging temporal and spatial gaps in passive microwave observations for global snow depth estimation","authors":"Qingyu Gu ,&nbsp;Yan Huang ,&nbsp;Jiahui Xu ,&nbsp;Tao Che ,&nbsp;Jiming Liu ,&nbsp;Zhan Pan ,&nbsp;Jianping Wu ,&nbsp;Shengli Wu ,&nbsp;Bailang Yu","doi":"10.1016/j.jag.2026.105369","DOIUrl":"10.1016/j.jag.2026.105369","url":null,"abstract":"<div><div>Establishing a long-term, consistent snow depth record from multiple passive microwave (PMW) sensors is critical for hydrological modeling and climate change research. However, integrating data from different PMW sources is challenging due to variations in sensor design and observation configurations. This study proposes a novel cross-sensor calibration and fusion framework that harmonizes brightness temperatures (TBs) from the Microwave Radiation Imager (MWRI) onboard the Fengyun-3 (FY-3) satellites with AMSR-E/AMSR2 data. Using AMSR2 as a reference, a Simultaneous Conical Overpass (SCO) calibration method is applied to align FY-3B and FY-3D MWRI data. The framework incorporates orbital differences to synergistically fuse cross-calibrated TBs, enhancing temporal continuity and filling data gaps. A spatiotemporal sample adaptation strategy is introduced to match TB observations from cross-calibrated orbital PMW products with hourly in situ snow depth measurements, generating a reliable training dataset for Random Forest (RF) model development. The trained model is subsequently applied to a consistent multi-sensor gridded TB dataset to produce a global snow depth record from 2002 to 2023. Validation results show excellent inter-sensor consistency, with R² values exceeding 0.99 across all TB channels. The incorporation of FY-3 data significantly improves spatial coverage: after 2012, the annual land coverage of ascending-orbit TBs increased from 74.80% to 88.56%, and descending-orbit TBs from 73.93% to 87.94%. The resulting snow depth product demonstrates high accuracy, broad spatial coverage, and strong temporal continuity. This study offers a scalable framework for fusing heterogeneous PMW data, providing a high-quality, globally consistent snow depth product for climate and hydrological studies.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"150 ","pages":"Article 105369"},"PeriodicalIF":8.6,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148179566","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}

{"title":"Joint monument–subsurface thermoelastic modelling of temperature-induced displacement at global GNSS reference stations","authors":"Ran Lu ,&nbsp;Kai Liu ,&nbsp;Yan Chen ,&nbsp;Mingyuan Zhang ,&nbsp;Xiaowei Zou ,&nbsp;Yanming Feng","doi":"10.1016/j.jag.2026.105360","DOIUrl":"10.1016/j.jag.2026.105360","url":null,"abstract":"<div><div>Global Navigation Satellite System (GNSS) height time series are essential for monitoring vertical land motion and maintaining terrestrial reference frames. However, temperature-driven thermoelastic deformation remains difficult to correct globally because it depends on both above-ground monument expansion and subsurface deformation. Here, we develop a full-spectrum coupled monument–subsurface thermoelastic model to estimate site-specific vertical displacement at 3,279 GNSS stations worldwide during 2000–2023. The model combines monument thermal expansion driven by hourly ERA5 2 m air temperature with layered subsurface thermoelastic deformation driven by ERA5 soil-temperature records, using gridded subsurface-property profiles, representative thermo-mechanical parameters, and monument type and height information from GNSS station logs. Annual thermoelastic amplitudes reach about 2–3 mm in many mid- and high-latitude regions, but are generally less than 1 mm at low latitudes. The monument and subsurface components commonly show annual phase differences of about 50–90°, leading to regional reinforcement or partial cancellation in the total response. Multi-year temperature variability can further introduce a potential thermoelastic bias of about ± 0.2 mm yr⁻¹ in vertical rates in some continental regions. Thermoelastic correction reduces the scatter of GNSS height residuals by about 3–8% at most stations and by up to about 12% at selected stations. When combined with non-tidal loading corrections, the reduction in residual scatter can reach about 70% at stations with the largest improvements. These results indicate that thermoelastic deformation is a physically distinct temperature-driven contribution that should be considered together with loading-related effects in GNSS height time series interpretation.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"150 ","pages":"Article 105360"},"PeriodicalIF":8.6,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148179570","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}

{"title":"UAV multispectral to hyperspectral reconstruction based on deep learning and radiative transfer models for crop nitrogen monitoring","authors":"Jie Wang ,&nbsp;Anirudh Belwalkar ,&nbsp;Sebastian T. Meyer ,&nbsp;Fei Li ,&nbsp;Ittai Herrmann ,&nbsp;Kang Yu","doi":"10.1016/j.jag.2026.105364","DOIUrl":"10.1016/j.jag.2026.105364","url":null,"abstract":"<div><div>Accurate and timely monitoring of canopy nitrogen (N) status is essential for precision farming and understanding N dynamics in vegetation systems. Uncrewed aerial vehicle (UAV)-based multispectral sensors are widely used for canopy monitoring due to their high spatial and temporal resolutions. However, their broad spectral bandwidths and the lack of short-wave infrared (SWIR) bands limit their effectiveness in advanced canopy nitrogen estimation models. In this study, we propose a novel framework of multispectral to hyperspectral reconstruction method (M2H–SWIR) by integrating the physically-based radiative transfer modeling (PROSAIL-PRO) with convolutional neural networks (CNN) to reconstruct full-spectrum spectral reflectance (400–2500 nm) from the visible to near infrared broad bands (400–900 nm) multispectral UAV images. The reconstructed hyperspectral data were validated against ground-based spectrometer-measured canopy reflectance data, where averaging strategies were used to reduce footprint and scale mismatches between UAV multispectral imagery and ground-based spectrometer-measured canopy reflectance observations. Our results demonstrated that the LUT-M2H–SWIR approach outperformed traditional PROSAIL-PRO inversion methods, particularly in reproducing the nitrogen-sensitive SWIR bands. Despite this success, extrapolating the short-range multispectral data to the long-range SWIR bands remains challenging, primarily due to spectral absorption overlapping from water absorption and canopy structural effects. Additional validation through canopy nitrogen estimation using the reconstructed hyperspectral reflectance reinforced the robustness of our proposed approach, suggesting its potential for operational canopy nitrogen mapping.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"150 ","pages":"Article 105364"},"PeriodicalIF":8.6,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148179567","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}

{"title":"LSFC-Net: Lightweight Spatial-Frequency Collaboration Network with sparse interaction for remote sensing change detection","authors":"Qi He ,&nbsp;Guangyue Lu ,&nbsp;Tao Lei ,&nbsp;Hailong Ning ,&nbsp;Wuxia Zhang ,&nbsp;Asoke K. Nandi","doi":"10.1016/j.jag.2026.105352","DOIUrl":"10.1016/j.jag.2026.105352","url":null,"abstract":"<div><div>Deep learning-based remote sensing change detection (RSCD) models have shown significant progress, yet they still face two major challenges. First, most methods struggle to effectively cooperate spatial-frequency information and generate complementary dual-domain features, thereby weakening the structural integrity of change objects. Additionally, the abundance of irrelevant semantic information in remote sensing scenes hinders spatiotemporal difference modeling, and the generation of redundant difference information can result in false positives. To this end, this study proposes a Lightweight Spatial-Frequency Collaboration Network (LSFC-Net) with sparse interaction for RSCD. Specifically, a Spatial-Frequency Context Aggregator (SFCA) is developed to decompose the original input into spatial and frequency streams. The former further refines spatial features, while the latter captures frequency discrepancies and texture details, finally integrating the dual-domain information in a complementary manner. Furthermore, a Feature Sparse Interactor (FSI) is designed to highlight genuine changes and suppress the influence of irrelevant semantics. FSI performs cross-temporal interaction on bi-temporal images through a sparse interaction attention mechanism, dynamically selecting the tokens most relevant to genuine changes. Experimental results show that LSFC-Net outperforms most representative RSCD methods in accuracy and efficiency.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"150 ","pages":"Article 105352"},"PeriodicalIF":8.6,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148179571","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}

{"title":"UAV-based built environment perception: Progress, challenges, prospects, and regulatory contexts","authors":"Xinya Kong ,&nbsp;Filip Biljecki ,&nbsp;Jing Yao ,&nbsp;Fan Zhang ,&nbsp;Yan Zhang ,&nbsp;Shensheng Chen ,&nbsp;Yuzhen Tang ,&nbsp;Wenhui Xu ,&nbsp;Pengyuan Liu ,&nbsp;Zexin Lei ,&nbsp;Junjie Luo","doi":"10.1016/j.jag.2026.105324","DOIUrl":"10.1016/j.jag.2026.105324","url":null,"abstract":"<div><div>Uncrewed Aerial Vehicles (UAVs) have proven to be a transformative technology for the fine-scale perception of the built environment. The recent proliferation of new platforms, sensors, and algorithms creates unprecedented opportunities to understand a complex environment. However, effectively harnessing these opportunities requires a systematic assessment of the emerging methodological practices to address challenges concerning the comparability, reproducibility, and generalizability of the knowledge being produced. Therefore, this study aims to systematically map the dominant methodological workflows in UAV-based built environment perception and critically assess their implications for scientific knowledge production. We conducted a systematic review of 201 peer-reviewed articles in the last decade (2015–2025), complemented by the construction of a novel global dataset of UAV flight policies across 80 countries, to deconstruct the dominant research workflows and to synthesize the progress and challenges across key application domains. Our analysis, leveraging a novel method that integrates PRISMA, machine learning, and Large Language Models, reveals a pronounced convergence in research practices, which stands in contrast to the apparent diversity of available technologies. We determine that the state of the art is characterized by: (i) a geographical concentration of studies in the Global North, correlated with permissive regulatory environments; (ii) a technological path dependency on a ‘standardized toolkit’ of multirotor UAVs and RGB sensors; and (iii) a methodological reliance on self-collected data (91%) that often remains non-public, fostering a research ecosystem of quantitative, computer vision-based analysis. By diagnosing these dominant patterns and their associated challenges, we propose a forward-looking agenda centered on fostering open science, diversifying technologies, and expanding methodological horizons to build a more integrated, robust, and equitable research future.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"150 ","pages":"Article 105324"},"PeriodicalIF":8.6,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148179574","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}

{"title":"Coupled Human–Climate drivers of groundwater storage decline in an arid inland river basin","authors":"Yuxin Lei ,&nbsp;Xiaohui Jiang ,&nbsp;Sinuo Fan ,&nbsp;Tong Nie ,&nbsp;Mengzhi He","doi":"10.1016/j.jag.2026.105370","DOIUrl":"10.1016/j.jag.2026.105370","url":null,"abstract":"<div><div>Understanding the evolution of groundwater storage and its drivers is essential for groundwater conservation and sustainable management in arid regions. However, a comprehensive assessment of long-term groundwater storage changes in the Heihe River Basin (HRB), especially before and after the artificial water diversion scheme, remains lacking. Here, we developed a downscaling framework to produce a 1 km groundwater storage anomaly (GWSA) dataset spanning the past three decades. A random forest model and a modified interactive geographical and temporal weighted regression (mIGTWR) model were then used to characterize spatiotemporal evolution patterns and identify the driving mechanisms. The results showed that: (1) groundwater level anomalies from monitoring wells were reasonably consistent with the reconstructed GWSA, with a maximum Pearson correlation coefficient of 0.746; (2) groundwater storage in the HRB declined significantly during 1990–2019, at rates of −0.20 cm/yr, −0.21 cm/yr, and −0.25 cm/yr for the entire basin, the midstream, and the downstream, respectively; (3) nighttime light intensity (NTL) was identified as the dominant temporal driver of GWSA variations; and (4) spatially, the interactions TEM ∩ NTL (temperature × NTL), PET ∩ NTL (potential evapotranspiration × NTL), NTL ∩ NPP (NTL × net primary productivity), together with NDVI (Normalized Difference Vegetation Index), were the key explanatory factors of GWSA change. These results highlight the coupled human–climate impacts on groundwater dynamics and provide critical implications for adaptive basin management and sustainable groundwater use in arid regions.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"150 ","pages":"Article 105370"},"PeriodicalIF":8.6,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148179568","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}