最新刊期
卷 30 ， 期 4 ， 2026
-
摘要：Land subsidence is one of the significant geohazards for urban development, and continuous monitoring and research are essential. The application of Interferometric Synthetic Aperture Radar (InSAR) to urban land subsidence monitoring has advanced markedly across data acquisition, methodologies, accuracy validation, and various applications. This study systematically reviews the development trajectory of urban subsidence monitoring by building on more than 2 decades of the authors’ time-series InSAR work in Shanghai—from exploratory application to engineering practice. The reviewed development trajectory covers (i) the initial experimental verification stage, (ii) a method development stage based on limited datasets, (iii) infrastructure monitoring enabled by high-resolution SAR data, and (iv) the subsequent expansion toward multisource, multiscale InSAR monitoring. On the basis of the research conducted in Shanghai, the group has extended the study to monitor land subsidence in other regions, such as the Beijing Plain. At present, urban land subsidence in Shanghai remains minor and manageable, with annual subsidence kept within 6 mm. This situation places rigorous demands on the detection of small magnitude deformation. In the future, several research directions are expected to advance InSAR’s development and application further in urban settings, including high-precision deformation monitoring in dense high-rise buildings, improved understanding and modeling of radar scattering mechanisms in complex urban environments, scientific interpretation of deformation under multifactor coupling, and Artificial Intelligence-supported information mining and early-warning frameworks.关键词：urban land subsidence;SAR;InSAR;time series InSAR;small dataset method;high-resolution SAR;infrastructure monitoring;multi-scale and multi-sensor monitoring601|2015|0更新时间：2026-04-24 -
摘要：Interferometric synthetic aperture radar (InSAR), as an all-weather and 24-hour operational spaceborne remote sensing technology, has been widely utilized in applications such as terrain reconstruction, surface deformation detection, and geophysical parameter inversion. As an extension of traditional InSAR technology, SAR split-spectrum interferometry techniques (including SAR range split-spectrum interferometry, azimuth split-spectrum interferometry, and TOPS mode burst overlap interferometry) have shown significant potential in fields such as interferometric phase unwrapping, three-dimensional surface deformation extraction, and ionospheric error correction in recent years. This study aims to provide a comprehensive overview of the development status of satellite-borne SAR split-spectrum interferometry techniques, with a particular emphasis on their principles and data processing methodologies. A total of 112 scientific papers published between 1992 and 2024, focusing specifically on the topic of split-spectrum interferometry, were reviewed. The review content was organized into four main parts: SAR sensors, current research progress in SAR split-spectrum interferometry, applications and advancements of SAR split-spectrum interferometry, and challenges and future perspectives of SAR split-spectrum interferometry. The paper offers a comprehensive summary and insights into the development of split-spectrum interferometry through analysis. First, this study summarizes the current SAR satellites, including commercial platforms. Then, it introduces the fundamental principles of SAR split-spectrum interferometry and systematically reviews its latest advancements in areas such as high-precision terrain mapping, surface deformation monitoring in regions with steep gradients, multidimensional surface deformation extraction, and ionospheric error correction within InSAR frameworks. Additionally, the study explores optimal parameter configuration schemes for high-precision surface deformation monitoring using SAR split-spectrum interferometry. Future trends in the development of this technology are also outlined in this paper. Continuous innovations in SAR technology enable the advancement of satellite-borne SAR systems toward broad bandwidths, multiangle observations, multifrequency capabilities, diverse polarizations, and satellite constellations. Techniques such as range split-spectrum interferometry and azimuth split-spectrum interferometry are expected to enhance accuracy in deformation measurements and ionospheric error correction, whereas burst overlap interferometry may significantly enhance the spatial coverage of deformation monitoring. Ultimately, these advancements are poised to establish a high-precision, multidimensional deformation monitoring system that integrates InSAR with SAR split-spectrum interferometry, thereby driving innovations in large-scale terrain mapping, geological hazard monitoring, and global environmental change assessments.关键词：InSAR;SAR interferometry;split-Spectrum interferometry;multiple-aperture InSAR;ionospheric correction;ground deformation;topography mapping0|0|0更新时间：2026-04-24 -
摘要：As an important Earth observation technology, Interferometric Synthetic Aperture Radar (InSAR) has been widely applied in surface deformation monitoring, such as urban subsidence, mining activities, and geological hazards, owing to its advantages of wide spatial coverage, millimeter-level precision, and long-term temporal observations. In recent years, its applications have extended to the safety monitoring of transportation infrastructure. Roads, railways, bridges, and airports are characterized by wide spatial distribution, linear and elongated structures, dynamic operation, and complex service environments, which pose high demands on conventional monitoring methods. InSAR provides a promising alternative solution. This study focuses on the applications of InSAR in transportation infrastructure monitoring by systematically reviewing research progress and development trends. First, based on current studies, typical applications of InSAR for different types of infrastructure, including roads, railways, bridges, and airports, are summarized to demonstrate its potential in subsidence monitoring, deformation identification, and structural safety assessment. Second, the major challenges of current InSAR applications are outlined by considering the particularities of transportation scenarios. These challenges include insufficient coherence due to weak scattering targets, discontinuous deformation of bridges and similar facilities that violate conventional models, and severe interference of atmospheric turbulence with millimeter-level monitoring precision. Several methodological improvements are discussed to address these challenges. These improvements include phase optimization, localized inversion with external constraints, atmospheric delay modeling, and integration with auxiliary datasets, which collectively enhance the reliability and applicability of InSAR results. Finally, the study incorporates representative application cases from the authors’ team on various types of infrastructure, thereby highlighting the potential of time-series InSAR technology in this field and further outlining future research directions. The application of InSAR technology in transportation infrastructure monitoring has already demonstrated broad prospects; however, continuous methodological and application-oriented improvements are still required to support transportation safety management and risk mitigation adequately.关键词：InSAR;Transportation infrastructure;deformation monitoring;Road subsidence;Bridge monitoring;Airport subsidence161|732|0更新时间：2026-04-24
Research Review
-
摘要：Atmospheric effect is one of the major sources of errors in Synthetic Aperture Radar interferometry (InSAR), and high-precision correction is essential for improving the reliability of InSAR applications. Atmospheric correction using external data, such as imaging spectrometer measurements or meteorological reanalysis products, is the most widely used technical means at present. Based on China’s first-generation global atmosphere and land reanalysis product (CRA), the latest generation of China Meteorological Association Global Atmospheric Reanalysis (CMA-RA V1.5, abbreviated as CRA1.5) provides hourly updated, global 3D atmospheric parameters with a horizontal resolution of 10 km covering the period from 1979 to the present. In this study, the CRA1.5 product is applied to InSAR atmospheric correction for the first time, and its performance is systematically evaluated. First, high-resolution vertical interpolation of geopotential height, temperature, and specific humidity parameters from reanalysis data is implemented by the hierarchical integration method, enabling the accurate estimation of Zenith Tropospheric Delay (ZTD). Second, the discrete ZTD values are interpolated into a high-spatial-continuum delay field by using an iterative tropospheric decomposition model, which is then employed to correct atmospheric signals in the interferometric phase.A total of 870 interferograms from Sentinel-1 satellite data acquired between 2021 and 2023 are constructed over two study areas, namely, Shandong in China and California in the USA. The performance of CRA1.5 is evaluated through a comprehensive set of metrics, including phase statistics, spatial structure, and terrain correlation. CRA1.5 is then compared with the widely used ERA5 and Generic Atmospheric Correction Online Service for InSAR (GACOS). Results show that CRA1.5 substantially reduces the phase standard deviation of interferograms (with the average reduction exceeding 30% in both regions), effectively suppresses long-wavelength atmospheric signals and elevation-related errors, and delivers an overall performance that is comparable to that of ERA5 and GACOS. CRA1.5 demonstrates excellent spatiotemporal consistency and reliability in mitigating InSAR atmospheric delay errors. It provides critical data support for the development of independent, high-precision atmospheric correction technologies in China and underscores the value of domestic reanalysis data for quantitative application of remote sensing.关键词：InSAR;zenith total delay;atmospheric correction;CRA1.5;statistical evaluation;standard deviation;spatial structure function128|144|0更新时间：2026-04-24 -
摘要：Karst sinkholes represent a significant geohazard characterized by complex spatiotemporal evolution and causative mechanisms. This study aims to enhance the applicability of Distributed Scatterer InSAR (DS-InSAR) in low-coherence karst terrains and clarify the hydrological controls governing sinkhole development along the Dead Sea. We propose a Dynamic Hypothesis Test of Confidence Interval (D-HTCI) algorithm for homogeneous pixel identification to address the limitations of inaccurate homogeneous pixel selection and unstable phase estimation. This method iteratively updates the reference mean and confidence interval per pixel to mitigate estimation biases. We took the Dead Sea karst region as a case study and integrated D-HTCI with a sequential phase-linking strategy to reconstruct a continuous, high-density deformation field using 242 Sentinel-1A scenes acquired between 2016 and 2024. The proposed approach retrieved 832,000 monitoring points, representing a substantial increase of 367,000 and 153,000 points over PS-InSAR and conventional DS-InSAR, respectively. Deformation rates primarily ranged from -120 mm/a to 20 mm/yr, with maximum cumulative subsidence exceeding 800 mm in the southwestern collapse zone. Time-series analysis revealed a strong linear coupling with synchronous Dead Sea water-level changes (R2>0.98). This result quantitatively supports a hydrogeological driver whereby sustained water-level decline lowers the fresh-saline interface, promotes salt dissolution, and triggers collapse. The integration of D-HTCI with sequential phase optimization significantly improves the feasibility and accuracy of DS-InSAR for monitoring natural terrains. This study provides compelling quantitative evidence and a robust framework for understanding the spatiotemporal evolution of karst sinkholes.关键词：Karst collapse;Dynamic Confidence Intervals;Homogeneous pixel identification;DS-InSAR;deformation monitoring;Water level decline244|198|0更新时间：2026-04-24 -
摘要：The LuTan-1 (LT-1) satellite, China’s first L-band interferometric SAR constellation, offers significant advantages for large-scale geological hazard monitoring because of its long wavelength and high spatiotemporal resolution. However, early-stage limitations such as insufficient real-time orbit accuracy, irregular image frames, and low effective image overlap rates significantly constrain its potential for wide-area applications. This study aims to develop a wide-area LT-1 data processing strategy that explicitly accounts for image overlap rate, thereby improving processing efficiency and deformation monitoring accuracy. The proposed approach classifies images by orbital path and then, within each path, constructs interferometric networks optimized by constraints on both spatiotemporal baselines and overlap ratios. A combination of interferogram stacking and regional network adjustment is employed to achieve seamless mosaicking of deformation results from different paths. On this basis, the small baseline subset interferometric synthetic aperture radar technique performs refined time-series deformation inversion for key areas. The Chongqing section of the Three Gorges Reservoir Area (including Yunyang, Fengjie, and Wushan counties) serves as the experimental area. The methodology is evaluated against a conventional LT-1 processing workflow to quantify improvements in data utilization and deformation accuracy. Compared with the traditional frame-based network construction strategy, the proposed method improves image utilization by 62% and achieves a deformation monitoring accuracy of 6.1 mm/a, outperforming the 10.7 mm/a obtained using the conventional frame-based network construction approach. Landslide detection results indicate that LT-1 ascending and descending track data identify 57 and 63 more potential landslide hazard sites, respectively, than Sentinel-1 ascending track data. The findings highlight the distinctive capability of L-band SAR satellites for landslide detection in mountainous terrain with complex surface conditions. The proposed wide-area LT-1 data processing method effectively addresses challenges of low efficiency and reduced accuracy caused by irregular image coverage and low overlap rates. It significantly enhances LT-1’s applicability for large-scale geological hazard monitoring and demonstrates superior landslide detection performance over C-band systems in complex terrain. This method is expected to increase the precision and reliability of deformation monitoring for wide-area hazard assessment further as the LT-1 data archives grow and the orbital accuracy improves.关键词：wide-area InSAR;Lutan-1;overlap-rate constraint;deformation monitoring;landslide detection;Three Gorges Reservoir Area;Stacking;L-band484|591|0更新时间：2026-04-24 -
摘要：Fucheng-1, China’s first commercial SAR satellite with interferometric capability, can operate in constellation with the Shenqi satellite, providing a new technical approach for high-precision monitoring of glaciers and geohazards. However, systematic evaluations of its velocity monitoring capability remain limited. This study utilizes Synthetic Aperture Radar (SAR) data acquired by Fucheng-1 (October-November 2023) and Shenqi (January-March 2025) over Sedongpu Valley in Nyingchi City, Tibet Autonomous Region, in combination with concurrent Sentinel-1 imagery. The Pixel Offset Tracking (POT) method and the differential SAR interferometry (DInSAR) technique are employed to extract glacier surface velocity in different scenarios and conduct a systematic comparative analysis. POT monitoring results indicate that because of their high spatial resolution, Fucheng-1 and Shenqi substantially outperform Sentinel-1 in POT displacement monitoring of small-scale glaciers, successfully identifying six glacier movement signals with a maximum velocity of up to 90 cm/day. By contrast, low-resolution Sentinel-1 data fail to provide sufficient information for accurate offset identification when a window suitable for small-scale glaciers is used. In terms of DInSAR, the spatial baseline of Shenqi satellite imagery’s repeat-pass interferometric pair is only 13.89 m, which is much lower than the concurrent Sentinel-1 baseline length; given its high-resolution characteristic, it exhibits superior interferometric quality. However, during periods of slow glacier movement, interferometric performance can still be improved when short-baseline Sentinel-1 data are employed. In typical glacier velocity monitoring, the boundary clarity and displacement characteristics obtained by Fucheng-1 and Shenqi are consistent with actual glacier dynamics, indicating their remarkable potential for refined monitoring of glacier movement. This study systematically evaluates the comprehensive performance of Fucheng-1 and Shenqi in glacier monitoring and confirms their effectiveness and application potential in fine glacier movement monitoring. This work provides an important basis for the development of high-precision glacier remote sensing monitoring systems.关键词：Fucheng-1;Shenqi;Sentinel-1;glacier velocity;Sedongpu;pixel offset tracking0|0|0更新时间：2026-04-24 -
摘要：Homogeneous pixel selection constitutes a fundamental preprocessing stage in distributed scatterer interferometric synthetic aperture radar (DS-InSAR), exerting a direct influence on the precision and dependability of subsequent phase unwrapping procedures. Conventional methodologies encounter limitations, including diminished performance with restricted image stacks and inherent difficulties in reconciling Type I/Type II error trade-offs. This research presents a novel algorithmic framework specifically designed to enhance the robustness and accuracy of homogeneous pixel identification across variable sample sizes and heterogeneous scene characteristics.The proposed Dynamic Center Growing Selector (DCGS) algorithm synergistically combines region-growing methodology with rigorous statistical hypothesis testing protocols. The computational workflow comprises three sequential stages. At the first stage, locally homogeneous pixel subsets are identified within 7×7 analysis windows through likelihood ratio testing (F-test). The mean intensity value derived from this preliminary subset subsequently replaces the central pixel magnitude to establish a robust seed point. At the second stage, adaptive region expansion is implemented using gamma distribution testing. Commencing from the validated seed pixel, eight-connected neighborhood pixels undergo iterative evaluation via breadth-first search methodology. A dynamically updated statistical parameter enables real-time adaptation during region growth, ensuring precise boundary delineation. At the third stage, the fully developed spatial region constitutes the definitive homogeneous pixel set. This algorithmic architecture minimizes the dependency on potentially skewed initial reference values while optimizing discrimination capability in regions exhibiting subtle heterogeneity.Monte Carlo simulations conducted under six sample sizes (N=10 to 60) demonstrated that DCGS achieved an average standard deviation of 0.014, outperforming the generalized likelihood ratio test, Kolmogorov-Smirnov test, Baumgartner-Weiss-Schindler test, and Hypothesis Test of Confidence Interval (HTCI) by 68.4%, 63.2%, 67.9%, and 10.7%, respectively. In real-data experiments utilizing 22 Sentinel-1 images from the Xiong’an area, Hebei Province, DCGS generated superior homogeneous pixel maps with distinct differentiation of roads, vegetation, and buildings. Quantitative assessment of phase optimization quality revealed that DCGS attained minimal values in the sum of phase differences, phase standard deviation, and residue point number, exceeding HTCI’s performance by 2.6%, 8.9%, and 18.4%, respectively. Furthermore, DCGS-based DS-InSAR processing enhanced the point density by a factor of 10.2 relative to the StaMPS method and detected deformation signals within ranges (i.e., -128 mm/a to -100 mm/a and 80 mm/a to 100 mm/a) undetectable by alternative approaches.The proposed DCGS algorithm substantially enhances accuracy and robustness in homogeneous pixel selection for DS-InSAR applications. It exhibits exceptional performance in low-contrast, small-sample scenarios through improved control of Types I and II errors. Although the proposed algorithm is computationally less efficient than parametric methods, such as HTCI, its substantial improvements in selection precision, phase optimization quality, and deformation monitoring resolution justify this trade-off. This methodology demonstrates considerable potential for reliable surface deformation monitoring in complex environments characterized by diverse scattering mechanisms.关键词：Distributed scatterers;DS-InSAR;homogeneous pixel selection;region growing;non-parametric hypothesis test;deformation monitoring111|246|0更新时间：2026-04-24 -
摘要：With the operation of large-scale hydraulic infrastructures, many slopes located in reservoir areas are prone to deformation and failure under the influence of fluctuations in the reservoir water level, thereby seriously threatening the safe operation of hydraulic facilities and possibly leading to more severe secondary disasters. Therefore, conducting accurate landslide susceptibility assessment is of great significance for landslide risk prevention and mitigation in reservoir regions. In this study, a refined dynamic landslide susceptibility assessment was conducted for the Zigui-Badong reservoir bank section of the Three Gorges Reservoir area. First, a comprehensive landslide inventory was established on the basis of long-term field investigations and historical records. Subsequently, 16 conditioning factors, including elevation, slope, and lithology, were selected to construct a landslide susceptibility evaluation index system, and an initial static susceptibility assessment was performed using ensemble machine learning methods. On this basis, long-term surface deformation information was retrieved using the small baseline subset interferometric synthetic aperture radar (SBAS-InSAR) technique. An optimization matrix that integrates susceptibility classes and SBAS-derived deformation rates was developed to couple static predictions with dynamic deformation information, thereby obtaining dynamic landslide susceptibility results. (1) Very high- and high-susceptibility zones are mainly distributed along the Yangtze River and its tributaries, and elevation, vegetation coverage, distance to rivers, and rainfall are the dominant controlling factors for landslide spatial distribution; (2) compared with individual base models, the ensemble learning model more effectively integrates their advantages and achieves the highest prediction accuracy (AUC = 0.954); (3) the dynamic susceptibility results coupled with time-series InSAR data can effectively correct false-negative and false-positive errors in static susceptibility assessments. This study improves the accuracy and timeliness of landslide susceptibility modeling and provides a valuable reference for the dynamic management of landslide hazards in reservoir areas.关键词：landslide;dynamic susceptibility;machine learning;ensemble learning;time series InSAR;Three Gorges Reservoir Area0|0|0更新时间：2026-04-24 -
摘要：Monitoring urban ground and infrastructure deformation is essential for the safe operation of cities. Interferometric Synthetic Aperture Radar (InSAR) has become a key technology for this purpose. However, in complex urban environments, conventional persistent scatterer InSAR heavily relies on prior constraints, and advanced variants are often limited by high computational demands. These limitations lead to insufficient arc coverage and reduced robustness in time series analysis, especially when deformation is localized and nonlinear.To address these challenges, we propose exhaustive pairs InSAR (EP-InSAR), a time series InSAR method that is based on dense arc linking. By sequentially connecting baseline pairs, the linear deformation term is embedded into the phase component of the temporal coherence to reduce the dimensionality of arc parameter estimation. A dual-threshold iterative network expansion strategy is introduced to identify high-quality point pairs in an approximately exhaustive manner, mitigating the reliance on amplitude-based priors. In the implementation, conventional InSAR preprocessing is followed by sequential interferogram formation, 1D spectral search for arc parameters, and robust global least-squares adjustment of height residuals and thermal dilation coefficients. Deformation time series are retrieved from a rectified low-frequency interferogram stack, and elevation and thermal components are removed.Experiments using TerraSAR-X data over Shanghai demonstrate that EP-InSAR increases point coverage, improves arc network connectivity, and reduces global adjustment errors. In a 3000×3000 pixel processing window, the method extracts ~910,000 measurement points connected by ~60,000,000 high-quality arcs, providing substantial redundancy for parameter estimation. Among all the accepted points, 29.3% have an amplitude dispersion index larger than 0.6, indicating that many structurally reliable scatterers are discarded under amplitude-based screening alone. The estimated height residual and thermal dilation fields form coherent spatial patterns that match building footprints and material distributions, and the derived time series reveal localized nonlinear deformation that is consistent with construction histories visible in optical satellite imagery.The EP-InSAR workflow is parallelized and optimized for large-scale processing, showing strong potential for applications in urban structural health monitoring. EP-InSAR provides a practical strategy to approximate exhaustive arc exploration without a prohibitive computational cost, relaxes the dependence on strict amplitude-based screening, and increases sensitivity to local nonlinear deformation under complex scattering conditions. These properties indicate that EP-InSAR can support future large-area urban deformation surveys and routine monitoring of critical infrastructure and building stability.关键词：Urban Deformation Monitoring;Interferometric Synthetic Aperture Radar (InSAR);Exhaustive Pairs InSAR (EP-InSAR);arc network expansion;local nonlinear deformation;persistent scatterer0|0|0更新时间：2026-04-24 -
摘要：Synthetic Aperture Radar Tomography (TomoSAR), which is based on multi-baseline interferometry, utilizes repeat-pass monostatic or single-pass multistatic data to synthesize an aperture in the elevation direction. This technique enables 3D imaging and the high-resolution reconstruction of forest structures. However, residual phase errors between tracks frequently lead to spectral energy dispersion and image blurring, significantly compromising the accuracy and stability of forest height retrieval. Moreover, existing phase error calibration methods require a high computational load for the tomographic imaging of large-scale forest scenarios.To address the aforementioned issues, this study proposes a novel phase error calibration method, called Subarea Fast Contrast Optimization Autofocusing (SFCOA), which enhances the efficiency and accuracy of TomoSAR imaging. This method leverages the spatial invariance of phase errors within local subareas to reduce the dimensionality of the optimization problem. By transforming traditional pixel-by-pixel optimization into an efficient contrast maximization task that is performed over a limited number of subareas, the method significantly lowers computational complexity. In addition, a chain initialization strategy based on spatial continuity between adjacent subareas minimizes the required iterations and substantially accelerates convergence.The proposed SFCOA method is validated using airborne P-band TomoSAR data acquired over Saihanba Forest in Northern China. SFCOA is then compared with several representative phase calibration methods, including Phase Derivative Constrained Optimization (PDCO), Phase-Shifting Interferometry (PSI), Phase Gradient Autofocus (PGA), and Network Construction (NC)-PGA. After phase error calibration, tomograms are generated using a Capon beamformer, and the resulting vertical profiles are evaluated against LiDAR measurements. The qualitative analysis of representative azimuth lines demonstrates that SFCOA achieves the narrowest main lobes, the lowest side lobes, and smooth vertical continuity in the reconstructed tomograms. By contrast, PDCO results exhibit vertical discontinuities. PSI only partially suppresses interference, with residual side lobes indicating incomplete phase error calibration. PGA achieves reasonable focus but suffers from contrast degradation near segment transitions, blurring the separation between canopy and ground reflections. Although NC-PGA reconstructs a continuous forest vertical structure, it remains inferior to SFCOA in terms of tomographic sharpness and peak localization accuracy.Quantitative evaluations confirm the computational efficiency of SFCOA, which requires only 12.36 s to process the study area on a 2.5 GHz Intel CPU with 16 GB RAM, compared with 16.7 h for PDCO, 388.97 s for PSI, 28.25 s for PGA, and 45.39 s for NC-PGA. In forest height estimation, SFCOA also achieves superior accuracy, improving R2 from 0.516 to 0.609, mean absolute error from 0.509 m to 0.435 m, mean absolute percentage error from 3.152% to 2.820%, and root mean square error from 0.627 m to 0.526 m relative to NC-PGA.In summary, SFCOA provides an accurate and computationally efficient phase error calibration solution for TomoSAR in forested environments. By combining subarea processing, derivative-constrained contrast optimization, and chain initialization, it achieves an effective balance between processing speed and reconstruction quality. The method exhibits strong potential for supporting high-precision 3D forest mapping and sustainable forest management.关键词：SAR Tomography (TomoSAR);phase error calibration;space-invariant assumption;subarea segmentation;autofocus;tomogram;forest vertical structure0|0|0更新时间：2026-04-24 -
摘要：Forest height is a crucial parameter of the forest vertical structure and plays a notable role in forest resource management, carbon cycle research, and climate change studies. Traditional Time-Frequency (TF) analysis-based Interferometric Synthetic Aperture Radar (TF-InSAR) techniques employ the random volume over ground model with a limited number of discrete sub-look complex coherence observations. However, given the reliance on limited observations, these methods often struggle to acquire the characteristic points used to describe the geometric distribution of sub-look complex coherence coefficients, resulting in low accuracy in forest height inversion. To address this issue, this study proposes a TF-InSAR forest height inversion method that incorporates the geometric distribution of the sub-look coherence set. First, a multi-aperture interferometric coherence matrix is constructed through sub-aperture decomposition. Second, the concept of a sub-look coherence set is introduced, and its geometric characteristics are analyzed. On this basis, the major axis of the coherence set is utilized to obtain complex observations and a surface phase that are close to pure-volume decorrelation. Finally, the forest height product is generated using a fixed extinction coefficient inversion approach. The experimental data are acquired from the airborne fully polarimetric P-band SAR data obtained during the AfriSAR campaign conducted by the European Space Agency in Gabon, Africa, in 2016, covering the Mabounie test site. Validation is performed using full-waveform LiDAR data from the Land Vegetation and Ice Sensor. Experimental results indicate that at the 40 m×40 m plot scale, the forest height estimation accuracy achieved by the proposed single-polarization TF analysis method (root mean square error RMSE=6.70 m) shows a 34.44% improvement over the conventional method (RMSE=10.22 m). Conclusion Using only single-polarization SAR data, the proposed TF-based analysis method can achieve a forest height estimation accuracy comparable to that of full polarimetric data-based methods (RMSE=8.54 m), providing a reliable and cost-effective technical solution for large-scale forest monitoring.关键词：forest height;Interferometric SAR (InSAR);Time-Frequency (TF) analysis;sub-aperture decomposition;coherence set;Random Volume over Ground (RVoG)0|0|0更新时间：2026-04-24 -
摘要：Compared to conventional time-series InSAR (TS-InSAR) techniques, phase gradient stacking offers advantages such as immunity to atmospheric and phase unwrapping errors, fast computational speed, and robustness in low-coherence scenarios, making it well-suited for large-area landslide identification. However, complex mountainous terrain significantly distorts phase gradient stacking results. Firstly, topographic variations cause severe geometric distortions in SAR images, leading to interferometric phase artifacts. Secondly, the external Digital Elevation Model (DEM) used in differential interferometry inevitably contains errors, and the residual topographic phase in interferograms introduces false deformation signals in the phase gradient results, hindering accurate landslide detection. This study provides an in-depth analysis of topographic effects in phase gradient stacking and proposes corresponding correction methods. A divide-and-conquer approach is adopted to address the two types of terrain-induced errors. To mitigate geometric distortions, the actual ground distance between adjacent pixels is calculated and used to correct phase-gradient estimates, thereby unifying the spatial reference for gradient computation across the entire interferogram and reducing foreshortening effects. For residual topographic phases, a strict theoretical derivation based on their generation mechanism and the stacking algorithm demonstrates that their magnitude is proportional to the absolute value of the cumulative perpendicular baseline. Based on this relationship, a reference threshold for the cumulative baseline is derived, and a greedy algorithm is designed to efficiently select an optimal baseline subset that meets the threshold requirement and minimizes DEM-induced errors. Experimental results in the Guxue section of the Jinsha River demonstrate that the improved phase gradient stacking method effectively eliminates foreshortening effects with only 120 interferograms and significantly enhances deformation detection capability. In contrast, conventional methods still exhibit distortions even with over 200 interferograms, confirming the effectiveness of the proposed geometric distortion correction. Further validation using the Daguangbao landslide triggered by the Wenchuan earthquake shows a clear positive correlation between the residual topographic phase error and the absolute cumulative vertical baseline but largely independent of the number of interferograms. Reducing the absolute cumulative vertical baseline effectively suppresses the impact of DEM errors. Correcting the phase gradient using the actual distance between adjacent pixels unifies the spatial reference for gradient calculation across the interferogram, effectively alleviating foreshortening and improving deformation identification capability in complex mountainous areas. The experimental results also indicate that, in practical applications, more than 180 interferograms should be used for stacking to ensure the completeness of the detection results. The residual topographic phase error is proportional to the absolute cumulative vertical baseline. By optimizing baseline selection to control the cumulative vertical baseline, this error can be significantly reduced, enhancing the accuracy and stability of phase gradient stacking results.关键词：InSAR;phase gradient stacking;identification of landslide hazards;geometric distortion;DEM Error;residual topographic phase;surface deformation detection0|0|0更新时间：2026-04-24 -
摘要：Urban deformation poses considerable threats to residents’ lives, property safety, and social stability, making its regular monitoring critical. Time-series Interferometric Synthetic Aperture Radar (InSAR) enables large-scale urban deformation monitoring with high precision and has been employed in numerous studies and engineering applications. Synthetic Aperture Radar (SAR) images with different resolutions exhibit distinct advantages in urban deformation monitoring. However, comparative analyses on this topic remain insufficient. Additionally, discussions on the scale effects of SAR images with different resolutions on urban deformation monitoring are scarce.Deformation monitoring results and scale effects across SAR images with different resolutions are compared and analyzed by using low-, medium-, and high-resolution SAR images covering the same area and monitoring period and Persistent Scatterer InSAR (PSInSAR).Findings indicate that as SAR image resolution decreases, the observed targets become increasingly blurry because of scale effects. This degradation leads to the reduced capability for precisely characterizing deformation information from the InSAR interferometric phase, a decline in coherence, a decrease in the maximum deformation detected by PSInSAR, and a weakened correspondence between SAR pixels and building structures, overall diminishing the capability of PSInSAR in fine-scale urban deformation monitoring. Compared with medium- and high-resolution SAR images, Sentinel-1 data demonstrate weaker deformation monitoring performance for urban buildings, particularly in detecting localized and large deformations. However, in large-scale deformation identification and monitoring, Sentinel-1 images yield results comparable to those provided by high-resolution SAR images.关键词：Urban Deformation Monitoring;PSInSAR Technology;scale effects;SAR Image Resolution;InSAR Deformation Monitoring223|229|0更新时间：2026-04-24 -
摘要：Urban security is the essential foundation of stable and sustainably developing cities. Fucheng-1, a domestic commercial SAR satellite with interferometric capability, can play an important role in urban deformation monitoring, for which geometric rectification is a critical step. High-precision geometric rectification establishes accurate correspondence between image pixels and real-world objects. Such a correspondence is essential for the semantic analysis of imagery, attributing deformation anomalies, and understanding deformation mechanisms. Given that the performance of geometric rectification depends on the geolocation capability of satellites and efficacy of precision geometric rectification methods, this study first assesses the geolocation accuracy of Fucheng-1 in stripmap mode by deploying corner reflectors. This work proposes a two-stage road-vector matching algorithm that integrates structure-tensor features with phase congruency to address the high cost and limited accuracy of conventional geometric rectification methods in urban scenes. By decoupling range- and azimuth-direction errors, performing unidirectional offset searches, applying density-based clustering, and conducting local fine matching, the proposed method achieves the precise geometric rectification of Fucheng-1 imagery in urban areas. Results show that Fucheng-1’s geolocation accuracy is greater than 50 m. Furthermore, compared with other methods, such as simulated SAR-image matching, model-based error correction, and optical-SAR matching, the proposed approach yields the best performance, improving geolocation accuracy to approximately 6 m in the study area. This work provides a practical approach to enhance the geolocation accuracy of Fucheng-1 and supports the application of domestic SAR data in urban environments.关键词：SAR;Fucheng-1;road-vector matching;geolocation;precision geometric rectification;urban monitoring;structure-tensor;phase congruency0|0|0更新时间：2026-04-24 -
摘要：Airborne interferometric synthetic aperture radar (InSAR) plays a key role in high-precision topographic mapping and surface deformation monitoring. However, phase unwrapping, a core step in InSAR data processing, continues to be a challenging ill-posed problem, particularly in complex environments. When interferograms suffer from high fringe density, large-area decorrelation, or dense phase residues, traditional phase unwrapping algorithms based on single-feature weighting methods often fail to evaluate phase quality accurately. This inadequacy leads to significant unwrapping errors; for optimization-based algorithms in particular, such a defect inevitably results in incorrect cost function assignments, thereby triggering severe regional unwrapping failures across the entire interferogram. To address these limitations, this paper proposes a novel multifeature fusion weighted phase unwrapping method tailored for airborne InSAR systems, with the aim of providing robust constraints for the global optimization model and significantly improving the reliability of the unwrapped phase in complex scenarios. The proposed method establishes a comprehensive cross-domain weighting framework that uniformly integrates SAR image intensity features and interferometric phase features. First, a normalized representation model is constructed for multisource dimensional features to overcome physical dimension differences. Instead of relying on a single data source, the algorithm adaptively regulates the weight allocation strategy by combining the structural characteristics of the interferometric phase with the scattering properties of the targets. Specifically, the actual fusion weight is mathematically designed using a power-law multiplication strategy, which enhances the contrast between high-quality and low-quality regions. Furthermore, the proposed fusion weight is embedded into the Iterative Reweighted Least Squares (IRLS) phase unwrapping framework. An iterative direction weight is derived and introduced to further refine the constraint mechanism of the global cost function and prevent local anomalous gradients from diffusing within the solution network. Unlike traditional fixed-parameter methods, this direction weight is dynamically updated and adaptively assigned on the basis of its own weight distribution during the iteration process, thereby significantly enhancing the numerical stability and global convergence of phase unwrapping. The performance of the proposed algorithm was rigorously evaluated using measured airborne InSAR datasets that represent four typical complex scenarios: large-area decorrelation, high residue density, low signal-to-noise ratio, and high fringe density. Quantitative and qualitative analyses were conducted to evaluate the unwrapping quality and computational complexity. As detailed in the comprehensive statistics shown in Table 2, the proposed multifeature fusion weighting method consistently outperforms both the stand-alone image feature and phase feature weighting methods in terms of unwrapping accuracy across all tested conditions. Rather than yielding localized improvements, the fusion weight systematically achieves the lowest root mean square error in every scenario. Furthermore, the statistical distribution of the unwrapping residuals confirms that the proposed method tightly constrains the mean error ( -
摘要：As one of the most important ecological indicators of wetlands, water levels directly reflect hydrological processes and ecological patterns. Therefore, their efficient and accurate monitoring is critical for wetland conservation and restoration. Interferometric Synthetic Aperture Radar (InSAR), with its advantages of wide coverage, all-day/all-weather imaging capability, and high precision, has been successfully applied in monitoring wetland water levels. However, considerable variations in internal hydrological connectivity within wetlands lead to challenges, particularly cross-hydrological-unit phase unwrapping errors and inconsistent water level reference baselines, for conventional InSAR techniques during large-scale water level retrieval. We propose a hydrological-unit-based small baseline subset InSAR (SBAS-InSAR) method for monitoring absolute wetland water level changes to address the above issue. First, multitemporal Sentinel-1 SAR imagery and global land cover data are used to analyze hydrological connectivity comprehensively and divide the study area into multiple independent hydrological units. Then, within each unit, a short baseline interferometric network is constructed to retrieve time-series relative water level changes, which are calibrated by using in situ observations of water levels from hydrological stations. Finally, least-squares estimation is applied to obtain the spatiotemporal distribution of absolute water level changes. The results obtained by taking the floodplain of Louisiana, the USA, as the subject of this case study demonstrate that (1) hydrological unit division significantly improves the reliability of time-series inversion, reducing the overall root mean square error from 13.20 cm to 4.03 cm. (2) Hydraulic barriers, such as levees and urban areas, significantly alter the spatial continuity of variations in wetland water levels. (3) C-band coherence in wetlands exhibits pronounced seasonal differences, being highest in late winter to early spring and lowest in late summer to early autumn, and is mainly affected by vegetation phenology and inundation. Overall, our proposed method enables the centimeter-level, large-scale monitoring of changes in wetland water levels, providing technical support for wetland water resource management and ecological protection.关键词：wetlands;hydrological unit;SBAS-InSAR;multi-reference point calibration;absolute water level0|0|0更新时间：2026-04-24
Theoretical and Methodological Innovation
-
摘要：The Sichuan-Tibet Transportation Corridor has a complex geological environment and widespread slope hazards, posing significant safety risks to the construction and operation of highways and railways.In response to the multitiered and differentiated demands for survey, construction, and maintenance in transportation engineering, this study proposes an integrated space-ground Interferometric Synthetic Aperture Radar (InSAR)-based monitoring scheme designed to provide stereoscopic and dynamic observation capabilities. This approach aims to address the diverse information needs of various stakeholders across three spatial scales—corridor zones, critical engineering sections, and key hazard sites—and temporal scales ranging from annual surveys and quarterly inspections to continuous monitoring. Specifically, a 3D deformation calculation framework that combines multitrack time-series InSAR data and a precise slope deformation extraction model based on the unscented Kalman filter were constructed. These models aim to combine the wide-area coverage of satellites with the high precision and high sampling rate of ground-based radar, thereby enhancing information support and technical assurance for early identification and risk prevention of engineering hazards in complex mountainous areas.To validate the scheme’s feasibility, we applied multiplatform satellite time-series InSAR methods along the G318 Highway and under-construction railway routes within the corridor to conduct large-area deformation monitoring and identify potential slope hazards. Particular attention was given to the Luding earthquake (Mw 6.6) epicentral area for interpretation and time-series monitoring of secondary landslides. For disaster mitigation at key bridge and tunnel engineering sites, we extracted 3D time-series deformation patterns for high-risk slopes near the Jinsha River bridge site and jointly employed ground-based InSAR to conduct multiscale monitoring applications and verification analysis on slope creep in critical tunnel portal areas.Results demonstrate that satellite and ground-based InSAR exhibit complementary advantages in spatiotemporal coverage and perspective for monitoring subtle deformations. The proposed comprehensive observation scheme and the multiscale application results in the Sichuan-Tibet Transportation Corridor provide a reference for basic geological surveys and disaster prevention and reduction research in the rugged mountainous regions of Western China.关键词：satellite-based InSAR;ground-based InSAR;Sichuan-Tibet Transportation Corridor;comprehensive monitoring scheme;geological hazard;three-dimensional deformation;Kalman Filter0|0|0更新时间：2026-04-24 -
摘要：Time-series Interferometric Synthetic Aperture Radar (InSAR), which has the advantage of an all-weather, 24-hour operation with high precision and high spatial resolution, has been widely applied to bridge structural health monitoring. However, numerous expansion joints and pronounced thermal expansion and contraction effects identified during large-scale bridge monitoring often cause phase discontinuities. This limitation undermines the spatial continuity assumption required by conventional phase unwrapping methods and hinders accurate deformation inversion. This paper presents an adaptive segmented unwrapping time-series InSAR algorithm that does not rely on prior model assumptions. Phase discontinuities at expansion joints and the interferometric network partitioned by setting arc error thresholds are automatically identified. Then, subnet reference points are selected based on bridge mechanical characteristics to enable segmented unwrapping, thereby recovering the displacement time series of coherent points. The method was validated using 13 scenes of PAZ X-band satellite imagery over the Hangzhou Bay Bridge. Results show that the proposed approach effectively locates expansion joints and achieves stable unwrapping, with better adaptability and robustness than the minimum cost flow algorithm. Further analysis reveals that the overall bridge deformation is highly correlated with temperature (maximum correlation coefficient of 0.988). After the thermal effects are removed, 97.1% of coherent points exhibit residual deformation rates below 2 mm/year, with only a few sections showing potential nonthermal deformation. These findings provide new insights for the health monitoring of large bridges and contribute to enhancing the reliability and applicability of time-series InSAR in infrastructure monitoring.关键词：InSAR;segmented phase unwrapping;bridge deformation monitoring;thermal expansion and contraction;Hangzhou Bay Bridge0|0|0更新时间：2026-04-24 -
摘要：Aiming at the 2015 catastrophic landslide at the Hong’ao Dump in Guangming District, Shenzhen, existing post event studies have predominantly relied on optical imagery, field investigations, and numerical simulations to analyze causal factors. However, traditional creep survey methods (e.g., manual monitoring stakes, inclinometers) fail to capture high temporal resolution deformation data for sudden landslides. People’s understanding of its deformation process is still not deep enough, especially the early stage creep characteristics. This study seeks to address this limitation by comprehensively characterizing the full disaster incubation and evolution process especially the presliding creep and its driving factors of the Hong’ao Dump. Radar satellite imagery (35 COSMO-SkyMed images from 2013 to 2016 with 3 m spatial resolution) with all weather and high revisit rate advantages was adopted for time-series monitoring of the Hong’ao Dump, covering the entire filling stage (from 2013 to 2015) and post landslide recovery period (2016) Combined with Small Baseline Subset InSAR (SBAS-InSAR), Shape From Shading SAR (SAR-SFS), and a depth-integrated continuum model, time-series analysis and numerical simulation of surface deformation before and after the landslide were conducted. Specifically, SAR-SFS derived multi-temporal Digital Elevation Models (DEMs, 5 m spatial resolution) which helped restore the presliding creep process and reduce topographic errors in subsequent simulations. Based on optical images, we calculated the Normalized Difference Vegetation Index (NDVI) and Fractional Vegetation Cover (FVC) to delineate the dump’s filling stages and verify the engineering operation timeline; The depth-integrated continuum model coupled with the Coulomb model was applied to simulate the fluid state of the high speed and long runout landslide, with the pore water pressure coefficient (λ) iteratively adjusted within 0.3—0.8 to match onsite observations. The SBAS-InSAR analysis showed that the study area’s surface deformation rate ranged from -40.7 mm/a to 64.3 mm/a, with significant settlement concentrated at the S7—S9 area (landslide trailing edge, accounting for 32% of the total dump area), where the maximum settlement rate reached -40.4 mm/a, and two obvious acceleration events (October 2014 and March 2015) were observed, coinciding with the two fastest filling phases (1.2×10⁵ m³/month). Numerical simulations indicated that when λ reached 0.6, the model best matched onsite observations: The sliding mass above the slip surface had a volume of about 3.60×10⁶ m³, a projected area of 1.03×10⁵ m², and a maximum filling thickness of 62.7 m. The landslide exhibited strong fluidity: The main sliding phase completed in 50 s (average speed 18.2 m/s), followed by 120 s of residual sliding, generating a collision momentum of 114006.3 kg·m/s, which is far exceeding safety thresholds for human survival (1200 kg·m/s) and building structures (44000 kg·m/s). From the dual perspectives of radar remote sensing and numerical simulation, this study identifies the coupled effect of excessive and rapid dump filling (filling rate 0.8-1.2×10⁵ m³/month, without adequate compaction) and elevated pore water pressure (λ=0.6, reducing soil shear strength by 35%) as the primary cause of the Hong’ao landslide. This work verifies the effectiveness of integrated radar remote sensing and numerical simulation in monitoring dump landslides, and further provides critical insights for predisaster risk assessment and early warning of similar dump sites.关键词：Guangming Landslide;SBAS-InSAR;Time-series Deformation;Numerical Simulation;Pre-failure Creep;Landslide Mechanism;risk analysis87|97|0更新时间：2026-04-24 -
摘要：China faces significant contradictions in natural gas supply-demand and substantial peak-shaving requirements. This scenario makes underground gas storage facilities a critical infrastructure for ensuring a stable gas supply. However, salt rock creep induced by high-pressure operations in salt cavern gas storage can trigger pronounced ground deformation, which poses severe threats to the safe production of storage facilities. Thus, accurately mapping fine ground deformation of gas storage facilities has become a prerequisite for ensuring their safe and stable operation. This study aims to characterize the spatiotemporal patterns of ground deformation in a large-scale salt cavern gas storage using the advanced time series Interferometric Synthetic Aperture Radar (InSAR) technique. It also aims to validate the feasibility of high-precision monitoring for hazard prevention.This research employs the latest full scatterer InSAR (FS-InSAR) technique. This technique can separate the temporal low-frequency deformation phase from the temporal high-frequency phases, such as atmospheric delay contribution, to process 79 Sentinel-1 C-band SAR images acquired from August 2021 to August 2024. The dataset covers a large-scale salt cavern gas storage in eastern China. The study retrieves millimetric-level ground deformation rates and cumulative deformation through time-series SAR processing, including interferogram generation, phase unwrapping, dual-scale temporal low-pass filtering, and deformation inversion. Additionally, leveling measurements collected synchronously are used to validate the accuracy of FS-InSAR results.FS-InSAR technology can obtain high-precision and high-density ground deformation in gas storage areas with abundant water and farmland. The monitoring density reaches up to 1,121 points/km², enabling full-pixel deformation monitoring for all areas except water bodies. Validation by 28 synchronously leveling data shows that the standard deviation of the difference between the FS-InSAR and leveling measurements is 2.9 mm/a.The new and old cavities in this gas storage exhibit significantly differentiated deformation characteristics. Severe land subsidence has occurred in the old cavity area, with a maximum subsidence rate of up to 82 mm/year. Subsidence in the new cavity area is relatively minor, mostly with subsidence rates of 10—20 mm/a. This difference may be related to the application of advanced technologies, such as natural gas resistance dissolution and redissolution cavity construction in new cavities. These technologies can effectively control cavity shape, reduce stress concentration, and enhance stability.Time-series analysis shows that the periodic surface deformation of the gas storage is highly correlated with gas injection and production activities. During the gas production stage, the release of reservoir pore pressure causes elastic retraction of the rock skeleton, thereby leading to ground subsidence. Conversely, ground uplift is observed during the gas injection stage. Therefore, the pressure inside the cavity should be closely monitored during the operation of the gas storage to prevent the risk of rock creep caused by excessively high or low pressures.This study highlights the effectiveness of FS-InSAR in providing high-resolution, continuous monitoring of surface deformation in salt cavern gas storage facilities. The identified spatial heterogeneity and periodic deformation patterns underscore the importance of integrating radar remote sensing with operational data for real-time safety assessment. The technology enables early detection of high-risk zones (e.g., old caverns with severe subsidence) and facilitates adaptive management of gas storage operations. The findings contribute to the development of intelligent monitoring systems for underground energy storage infrastructure. This development supports a sustainable and secure gas supply in China’s energy transition.关键词：InSAR;full scatterer (FS);salt cavern gas storage;surface deformation;gas injection and production542|584|0更新时间：2026-04-24 -
Deformation monitoring and analysis of Zoige Alpine Freeze-Thaw Peatlands using time-series InSAR AI导读
摘要：The Zoige Wetland, which has an important ecological carbon sink function, covers the largest alpine permafrost peatland in Eurasia. Both the freeze-thaw process of permafrost and the carbon cycle of peatlands can cause surface deformation. Monitoring and analyzing deformation can provide important evidence for studying the freeze–thaw and carbon cycle processes. However, current research on surface deformation in this area is relatively scarce. This study is the first to take the Zoige peatland as a research area. The deformation in this area was monitored using the Small Baseline Subset Interferometric Synthetic Aperture Radar (SBAS-InSAR). The spatiotemporal characteristics, evolution trends, and driving factors of the deformation were comprehensively studied. The health status of peatlands was evaluated by considering the deformation distribution and evolution trend.This study obtained 89 ascending and 83 descending Sentinel-1 SAR images from January 2020 to December 2022. The radar Line-Of-Sight (LOS) deformation in the peatland area was extracted by SBAS-InSAR. This deformation was verified by comparing it with the deformation results from the adjacent orbits. The vertical and east-west deformation were obtained by LOS deformation decomposition. The vertical linear cumulative deformation and the seasonal amplitude were extracted through deformation component modeling. Moreover, the historical deformation trend was obtained using the Mann-Kendall trend test and Theil-Sen estimation methods, whereas the future deformation trend was estimated based on the Hurst index. The spatial and temporal characteristics, change trends of the deformations, and health status of the peatland were explored in depth by combining diverse information, such as land cover type, surface temperature, and precipitation.The correlation coefficient of the LOS velocity of the overlapping areas between adjacent orbits reaches 0.74, and the root mean square error is ±0.55 mm/a. Vertical and east-west velocities in the study area range from -45 mm/a to 45 mm/a and from -25 mm/a to 25 mm/a, respectively. The vertical deformation is mainly distributed in the peat areas, particularly around the wetlands, water bodies, and the western high-altitude area. The nonpeat area in the northwest has relatively evident eastward deformation affected by elevation and aspect. Seasonal deformation, with a maximum amplitude of 16.9 mm, is mostly concentrated in the peat areas, particularly around the Cuorewajian Lake and the Manrima Township. The trend test results demonstrate that the areas with significant uplift and subsidence trends account for 51.95% and 26.10% of the total area, respectively, whereas the remaining areas with uplift and subsidence trends account for 8.38% and 5.88%, respectively. Moreover, 75.72% of the area may show an anticontinuity trend in the future, and 24.28% of the area may maintain the current trend.The deformation of the Zoige peatland exhibits complex characteristics and distribution patterns of subsidence and uplift, linear accumulation, and seasonal changes, as well as vertical and horizontal components coexisting. This deformation is mainly related to factors such as the freeze-thaw process, carbon cycle, land surface temperature, and precipitation. Moreover, the deformation trends of different land types vary significantly, mainly because of the differences in driving factors. Overall, the uplift area of the Zoige peatland is larger than the subsidence area, thereby indicating a good carbon sink function. However, local significant subsidence phenomena occur in areas such as the Cuorewajian Lake surroundings and are accompanied by a high Hurst index. This observation indicates a significant subsidence persistence. Thus, the peatland in these areas may face degradation risks. This study is the first to reveal the complex deformation characteristics, change trends, and influencing factors of the alpine permafrost peatland in Zoige, thereby providing scientific reference for the assessment of ecological functions and vulnerability in this region. It also verifies the effectiveness of the SBAS-InSAR technology in monitoring the surface deformation of large-scale permafrost peatlands.关键词：SBAS-InSAR;Freeze-thaw Peatland;deformation monitoring;Spatiotemporal Deformation Characteristics;Seasonal Variation;Driving Factors of Deformation;Assessment of Peatland Health Status217|1149|0更新时间：2026-04-24 -
摘要：Landslide detection and deformation monitoring are critical for geological hazard prevention and risk mitigation. This study proposes an integrated framework for landslide detection and three-dimensional (3D) deformation monitoring using Interferometric Synthetic Aperture Radar (InSAR). The proposed framework comprises three key components: (1) time-series InSAR data processing, (2) landslide detection based on InSAR results, optical images, and C-index, and (3) 3D deformation monitoring. The proposed framework is applied to Hualong County in Qinghai Province, a landslide-prone area. Initially, we employed multitemporal InSAR to analyze the ascending and descending Sentinel-1 satellite images acquired between January 2021 and June 2023 over Hualong County, Qinghai Province. Then, deformation rate maps were generated and cross-validated with field measurements from the Global Navigation Satellite System station. Landslide identification was performed by integrating InSAR-derived deformation signals, high-resolution Google Earth imagery, and the C-index of each potential deformation area. Subsequently, we applied the Aspect Parallel Flow Model (APFM) to calculate the 3D displacement field of representative landslides. The deformation results from InSAR and on-site equipment exhibit strong agreement, and the standard deviation of the obtained average deformation rates is 5 mm/a for the ascending and descending images. This performance shows the high reliability of the obtained InSAR deformation results. We detected 334 landslides by integrating InSAR deformation rate maps and Google Earth imagery. Among these, 233 landslides were discernible using ascending data, 265 using descending data, and 164 using both ascending and descending datasets. The total area of the detected landslides is approximately 95.56 km2. The detected landslides have slope gradients ranging from 5° to 40°, with 184 landslides posing direct threats to infrastructure (e.g., buildings, roads) and natural features (e.g., rivers). Fewer landslides were detected in the near north-south direction than in other orientations. This observation suggests that InSAR may exhibit reduced sensitivity to deformations associated with landslides occurring along this axis. In theory, observational errors can notably influence the 3D displacement field obtained from APFM, particularly in the context of landslides occurring in a nearly north-south direction. The 3D deformation time series of the Anjuhu landslide in Chuma Township, which encompasses the largest area, was calculated by utilizing APFM. Analysis reveals that the horizontal deformation significantly outweighs the vertical deformation, with the maximum cumulative horizontal displacement surpassing 1 m. The landslide presents a threat to two villages, a provincial road, and agricultural areas. Thus, monitoring this landslide is necessary. This study demonstrates the effectiveness of InSAR for regional-scale landslide detection and 3D deformation monitoring while highlighting its limitations, particularly in areas with unfavorable slope orientations. The results illustrate the benefits and limitations of InSAR in landslide monitoring, thereby offering practical examples that can guide county-level efforts in landslide identification and 3D deformation monitoring while providing technical support.关键词：InSAR;SBAS;landslide detection;C-index;landslide monitoring;three-dimensional deformation;Hualong County;ascending and descending295|539|0更新时间：2026-04-24 -
摘要：Under global warming, glacier retreat and meltwater accumulation drive the expansion of glacial lakes. As a result, the risk of glacial lake outburst floods that threaten downstream infrastructure and communities increases significantly. Continuous and accurate monitoring of glacial lakes is essential for disaster prevention. Synthetic aperture radar (SAR) has become widely applied in complex mountainous glacial lake studies because of its all-weather, continuous capability and cloud-penetrating advantage, with high temporal resolution enabling capture of seasonal dynamics. However, identification errors caused by seasonal freeze-thaw processes affecting SAR backscatter characteristics remain inadequately explored, thereby limiting extraction accuracy. This study aims to investigate the response mechanisms of SAR features to glacial lake freeze-thaw cycles systematically and develop methods to improve monitoring reliability under seasonal ice conditions.This study selected two adjacent glacial lakes with distinct freeze-thaw characteristics—Gongcuo and Langcuo in Basu County, Tibet—as test sites. We analyzed lake area changes throughout 2020 by combining Sentinel-1 SAR imagery with multisource optical data (Sentinel-2, Landsat-8, and PlanetScope). SAR preprocessing included radiometric calibration, multilooking, and geocoding using GAMMA software. Lake boundaries were extracted using a simple noniterative clustering algorithm applied to dual-polarization Vertical-Vertical (VV) and Vertical-Horizontal (VH) intensity data. Interferometric coherence was computed from temporally adjacent SAR image pairs with 12-day intervals to assess surface stability. The dark channel prior algorithm was employed for cloud masking of optical imagery. ERA5-Land reanalysis climate data were integrated to correlate SAR feature variations with temperature, precipitation, and wind conditions through Pearson correlation analysis.Analysis revealed that Langcuo rarely experiences seasonal freezing, with lake areas extracted from optical and SAR imagery showing good agreement. By contrast, Gongcuo exhibited pronounced seasonal freeze-thaw differences; in particular, SAR-derived areas fluctuated between 1.85 and 2.65 km2, whereas optical imagery showed stable areas from 2.6 km2 to 2.8 km2. Combined analysis of SAR intensity and interferometric coherence revealed that Gongcuo develops relatively stable lake ice from November through March annually. This stable ice is the primary cause of discrepancies between SAR and optical lake extraction. Backscatter intensity exhibited a strong negative correlation with temperature (r=-0.75), thereby confirming that warming promotes ice thinning and reduces effective scattering surfaces. Coherence analysis demonstrated that stable thick ice maintains high coherence values (mean ~0.2; upper quartile 0.25—0.3), whereas open water and thin ice show low coherence (mean ~0.15). An empirical threshold classification was established: open water (VH≤-24 dB; coherence≤0.15), thin ice (-20 dB ≤VH≤-24 dB; coherence ≤ 0.15), and stable ice (VH≥-20 dB; coherence≥0.15).This study demonstrates that SAR imagery is highly suitable for monitoring glacial lakes during warm ice-free seasons or for lakes without significant seasonal freeze-thaw cycles. Moreover, combining SAR intensity with interferometric coherence provides effective evidence for identifying the presence of stable lake ice. The findings reveal that seasonal ice cover is the core source of SAR extraction errors in glacial lake monitoring. The proposed intensity-coherence joint threshold method offers a feasible technical approach for dynamic monitoring under seasonal freeze-thaw conditions; however, further validation across the diverse glacial lakes of the Tibetan Plateau is needed to optimize broad applicability.关键词：Glacial lake;Area evolution;synthetic aperture radar;Backscatter intensity;Polarization characteristics;Interferometric coherence;Stable ice detection99|250|0更新时间：2026-04-24 -
摘要：The Baihetan Hydropower Station, as a national mega-project, is located in a geologically complex reservoir area where slope stability monitoring after impoundment is crucial. Traditional InSAR techniques mainly rely on deformation rate indicators and thus face limitations in capturing the full temporal evolution of slope deformation. This study aims to overcome these limitations by developing an intelligent framework for time-series deformation pattern recognition to improve hazard identification and risk assessment in the reservoir area.Sentinel-1 SAR data spanning April 2021 to March 2024 were processed using the SBAS-InSAR technique to retrieve the reservoir-wide deformation field. A transfer learning-enhanced deep learning architecture was proposed, in which synthetic samples were used for pretraining and real samples for fine-tuning. Based on this framework, a six-class taxonomy of deformation patterns—stable, linear, step-like, piecewise linear, power-law, and undefined—was established. A Temporal Convolutional Network (TCN) was then applied to classify deformation modes across the Baihetan reservoir.(1) Among 97377 deformation points, the dominant patterns were stable (45.6%) and piecewise linear (24.4%), with stable deformation strongly associated with low velocity and high coherence. (2) The Shimenkan landslide exhibited a delayed uplift response regulated by reservoir water level, lagging peak impoundment by about two months.(3) The Yezhutang slope showed rainfall-triggered step-like deformation and a piecewise linear response under coupled water level-precipitation forcing.The findings demonstrate that transfer learning effectively reduces labeling costs, while multidimensional deformation pattern analysis provides a new paradigm for dynamic risk assessment of reservoir-induced geohazards.关键词：time-series deformation classification;TCN;transfer learning;Baihetan reservoir area;SBAS-InSAR;deep learning;reservoir bank landslide;geohazard monitoring0|0|0更新时间：2026-04-24 -
摘要：On January 1, 2024, a magnitude Mw7.5 earthquake struck Japan’s Noto Peninsula. To investigate the seismogenic mechanism and tectonic characteristics of the event, this study integrated data from ALOS-2 PALSAR-2 and Sentinel-1/2 satellites. Geodetic techniques, including differential interferometric synthetic aperture radar (D-InSAR), SAR pixel offset tracking (POT), and optical image correlation (OIC), were employed to derive coseismic surface displacement. The 3D coseismic deformation field was resolved using the least-squares method. Fault geometry was determined by employing the Okada dislocation model. Then, distributed slip inversion was performed to resolve the detailed slip distribution on the fault plane, and the associated coseismic Coulomb stress changes were calculated. The principal findings of this study are summarized as follows: (1) The resolved 3D deformation field revealed substantial crustal movement, with the city of Wajima experiencing uplift of up to 5 m. Substantial north-south deformation of approximately 1 m was detected southwest of Wajima, and the east-west component of the deformation field reached a maximum coseismic displacement of 2 m. (2) Slip distribution inversion indicated that the 2024 Noto Earthquake was a major shallow reverse faulting event. The rupture occurred primarily on a northeast-southwest-striking, southeast-dipping reverse fault located between the Suzuyama and Wajima rupture segments. Seismic rupture was predominantly concentrated at depths of 0—10 km, with a maximum slip of 15 m. The estimated moment magnitude was Mw 7.35. (3) Analysis of coseismic Coulomb stress changes suggested that the subsequent aftershock activity was predominantly stress-triggered. The coastal rupture segments of the Noto Peninsula exhibited a state of substantial stress loading, indicating considerable potential for future seismogenesis. Future research should integrate multisource data, such as those from global navigation satellite systems and seismic waves, to enhance the reliability of deformation results. Deep learning techniques can be incorporated to improve the accuracy of 3D deformation resolution and fault geometric parameter estimation. Moreover, dynamic models that couple physical mechanisms should be developed to simulate the cascading evolution process of secondary hazards triggered by coseismic deformation. This endeavor will ultimately provide a quantitative basis for post-earthquake early warning and risk mitigation strategies.关键词：2024 Noto earthquake;InSAR;co-seismic deformation field;pixel offset tracking;slip distribution inversion0|0|0更新时间：2026-04-24 -
摘要：Accurately estimating long-term and seasonal deformation in permafrost regions and analyzing its driving factors are of great importance for assessing changes in active layer thickness and permafrost degradation. This work proposes a framework integrating InSAR, RobustSTL, and geographical detectors for extracting the seasonal deformation of permafrost and analyzing its driving factors. First, time-series InSAR technology is utilized to derive surface deformation over permafrost areas. Subsequently, the RobustSTL algorithm is employed to extract seasonal deformation and analyze its variation characteristics. Finally, geographical detectors are applied to explore the driving factors behind the seasonal deformation of permafrost. Sentinel-1 data from 2017 to 2021 in the source region of the Yellow River were collected for this study. Experimental results demonstrate that the surface deformation rate in the Yellow River’s source region ranges from -30 mm/a to 20 mm/a, with seasonal deformation varying between 2 and 30 mm. Analysis using geographical detectors indicates that NDVI, MAGT, and slope are significant factors influencing surface deformation. The effects of different factors on deformation are not independent but instead interact to influence the distribution of surface deformation jointly.关键词：permafrost;InSAR;Yellow River Source region;deformation;driving factors;mean annual ground temperature;precipitation0|0|0更新时间：2026-04-24 -
摘要：Fucheng-1 is China’s first commercial SAR satellite with high-resolution C-band capability for interferometric measurement, but its performance in mining subsidence monitoring lacks a comprehensive comparative evaluation. This study uses the Shanxi Datong Coalfield as the research area and employs Small Baseline Subset Interferometric Synthetic Aperture Radar (SBAS-InSAR) technology to process time-series data from 11 scenes of Fucheng-1, comparing the results with those from Sentinel-1 and LuanTan-1 satellites. The results show that, in terms of time–series InSAR measurement, Fucheng-1 identified 57 subsidence areas, providing a comprehensive depiction of the mining subsidence distribution. For the detection of small-scale deformation areas (<0.1 km²), Fucheng-1 detected 8, whereas Sentinel-1 failed to identify them, demonstrating its advantage in small-scale mining subsidence detection due to its high-resolution imaging. Moreover, when comparing deformation velocities, both data sets showed good consistency in low velocity deformation areas. However, for high velocity deformation areas, Fucheng-1, with its higher resolution and larger detectable deformation gradient, captured more accurate deformation signals, especially for small-scale rapid deformations, where Fucheng-1 demonstrated superior measurement accuracy. Multi-satellite differential InSAR comparison results revealed that Fucheng-1 consistently detected larger deformation values than Sentinel-1 and captured approximately 23 mm more deformation in a small deformation area compared to LuanTan-1, indicating that Fucheng-1, with its higher resolution, can detect larger deformation levels, while L-band LuanTan-1 is more sensitive to small deformation values. In terms of coherence, Fucheng-1 achieved 12.53% of pixels with coherence values between 0.6 and 1.0, higher than Sentinel-1’s 7.19%, showing better coherence performance under the same C-band conditions, but lower than LuanTan-1’s 22.43% with its longer L-band. In conclusion, Fucheng-1 exhibits a sensitive detection advantage for small-scale deformation in mining subsidence monitoring, with the ability to monitor both large deformation levels and small deformations. This study provides a reference for the future operational applications of domestic SAR satellites in mining subsidence monitoring.关键词：Fucheng-1;high-resolution commercial SAR satellite;differential InSAR (DInSAR);time-series InSAR;monitoring performance evaluation;subsidence monitoring0|0|0更新时间：2026-04-24 -
摘要：Large-scale landslide disasters often evolve from active landslides. Therefore, early and accurate identification of such active landslides is a key step to reducing disaster risks effectively. The core goal is to avoid casualties and minimize significant economic losses. At present, landslide disasters occur frequently in China. However, traditional manual interpretation methods have problems, such as low efficiency and insufficient identification accuracy. As a result, the actual needs of large-scale geological disaster monitoring cannot be adequately met. Accordingly, this study is dedicated to developing an automatic identification method suitable for large-scale active landslides. Moreover, Hualong County in Qinghai Province, which is severely affected by landslide disasters, is selected as a typical research area. Accurate and efficient identification of active landslides in the region is achieved by constructing a deep learning model. This approach ultimately provides a solid scientific basis for early warning and prevention of regional geological disasters and helps improve the overall ability of China’s mountainous cities and towns to prevent and control geological disaster risks.This study comprehensively utilizes the small baseline subset interferometric synthetic aperture radar and spatial analysis methods to extract effective deformation areas on the ground surface automatically. It also constructs a comprehensive criterion for identifying hidden dangers related to active deformation, geology, terrain, environment, meteorology, and human engineering activities. This criterion covers landslide movement, disaster breeding, and disaster-causing characteristics. A deep learning AMRetNet algorithm that can fully explore the multiscale nonlinear relationship between disaster breeding, disaster-causing characteristics, and landslide deformation is established to capture the global context and local detail features effectively through the retention mechanism. An arithmetic module is also introduced to learn multiscale nonlinear relationships adaptively. The establishment and introduction of the aforementioned algorithm and module significantly improve the accuracy and robustness of landslide identification in complex geological environments.In the case of Hualong County, this study successfully identified 178 active landslides, including 48 newly discovered ones. This result fully demonstrates the effectiveness of the method. The model exhibits excellent performance in various evaluation metrics in the test area (884 km²), with accuracy, precision, F1-score, AUC, and kappa coefficients reaching 99.05%, 90.21%, 0.7332, 0.9803, and 0.7286, respectively. It also has a low missed detection rate of 1.85%. Comparisons with mainstream algorithms, such as transformer, U-Net, CART, and SVM, as well as ablation experiments, showed that AMRetNet performed optimally in various performance metrics. This result demonstrates AMRetNet’s significant progressiveness and reliability.This study confirms the superiority of the proposed automatic identification method for active landslides in accurately identifying active landslides in vast areas. The technical bottleneck of high false alarms and high missed detections in traditional methods is successfully solved by constructing a comprehensive identification index set and a deep learning AMRetNet algorithm that can fully explore the multiscale spatial nonlinear dependence between active landslide disaster breeding, disaster-causing characteristics, and deformation. The research results can provide reliable technical support for the prevention and control of geological disaster risks in mountainous towns in China. They also have important theoretical significance and promotion application value.关键词：InSAR;active landslide;geoenvironmental and disaster-triggering factors;active deformation;automatic identification;deep learning;Hualong County0|0|0更新时间：2026-04-24 -
Spatiotemporal analysis of land subsidence in the Beijing Plain area based on factor decomposition AI导读
摘要：Land subsidence in the Beijing Plain has become one of the most prominent geological hazards affecting urban safety and sustainable development. Therefore, investigating the spatiotemporal distribution characteristics of land subsidence is crucial for understanding its evolution patterns and supporting the implementation of effective risk prevention and groundwater management. This study aims to characterize the spatial distribution and temporal evolution of land subsidence in the Beijing Plain from 2017 to 2020 and quantitatively analyze its long-term trends and seasonal deformation behaviors using a time series decomposition approach.A total of 95 Sentinel-1 Synthetic Aperture Radar (SAR) images acquired between 2017 and 2020 were collected to cover the Beijing Plain. Land subsidence time series were retrieved using the Permanent Scatterer Interferometric Synthetic Aperture Radar (PS-InSAR) technique. Using factor decomposition theory, a time-series model of land subsidence, referred to as the Time-Series model of Land Subsidence based on Factor Decomposition (TMLS-FD), was constructed to decompose the subsidence time series into long-term trend and seasonal fluctuation components. The performance of model fitting was evaluated using residual statistics, including the expectation and distribution characteristics of residuals. The spatiotemporal evolution characteristics of land subsidence were further analyzed at the regional scale.Result Results indicate that land subsidence in the Beijing Plain exhibits substantial spatial heterogeneity. Two major subsidence funnels were identified, centered in the Heizhuanghu-Jinzhan area of Chaoyang District and the northwestern part of Tongzhou District. The TMLS-FD model demonstrates a high fitting accuracy, revealing residual expectations close to zero and residuals approximately following a normal distribution, thereby indicating the absence of substantial systematic bias. Temporally, land subsidence demonstrates an overall continuous development trend, superimposed with a stable annual periodic fluctuation. Different subsidence zones exhibit distinct dominant deformation mechanisms: severe subsidence areas are mainly controlled by long-term subsidence trends, whereas slight subsidence and non-subsidence areas are largely governed by seasonal fluctuations. The seasonal deformation displays a clear annual cycle, with subsidence troughs occurring predominantly between July and September, accounting for approximately 71% of all observation points. Regions such as southeastern Shunyi District and Pinggu District show relatively large seasonal amplitudes, revealing a maximum amplitude of 20.13 mm.Through the integration of PS-InSAR observations with a factor decomposition-based time series model, this study reveals the differentiated spatiotemporal evolution characteristics of land subsidence in the Beijing Plain. The proposed TMLS–FD model effectively captures long-term subsidence trends and seasonal deformation behaviors, providing a clear and interpretable framework for land subsidence analysis. Overall, the findings enhance the understanding of land subsidence dynamics in the Beijing Plain, offering scientific support for land subsidence monitoring, risk assessment, and groundwater resource management.关键词：Beijing Plain Area;land subsidence;Persistent Scatterer Interferometric Synthetic Aperture Radar (PS-InSAR);factor decomposition;time-series model of land subsidence;seasonal deformation;spatiotemporal distribution characteristics49|263|0更新时间：2026-04-24
InSAR Innovative Applications
- Postal code：100190
- Tel：010-58887052 Email：nrsb@aircas.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备20021838号-9
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0