Remote Sensing doi: 10.3390/rs18121895

Authors: Guoyu Li Xuanchen Jiang Mingtao Xiang Jiaqi Liu Qing Wu Baihe Liang Mengran Ma Yangfei Huang

The three-dimensional (3D) built environment encodes critical information about urban form intensity, environmental exposure, and resource consumption. However, previous studies have often overlooked the integration of long-term analyses of both horizontal expansion and vertical growth. This study aims to identify the spatial differentiation, morphology types, and vertical growth trajectories of major Chinese megacities over the past three decades. Using high-resolution GABLE building data and time-series GAIA impervious surface data, we examine the evolution of urban 3D morphology across six major Chinese megacities from 1991 to 2023 through a retrospective analysis of building construction years combined with spatial gradient analysis. The results reveal that although the megacities exhibit distinct differences in vertical structure, shape complexity, and spatial compactness, they share a consistent center-to-periphery gradient across most 3D indicators. The most active volumetric growth was concentrated in a zone 8–14 km from city centers, which accounted for 23.6% of total new development, whereas the inner core within 6 km contributed less than 2.68%. In terms of temporal dynamics, Beijing, Shanghai and Guangzhou follow an inverted-V-shaped 3D expansion trajectory driven by mid-rise construction; Tianjin and Hangzhou show accelerated growth with a higher proportion of high-rise clusters; while Shenzhen demonstrates an early peak and a decelerated growth rate, accompanied by a pronounced polycentric pattern. While recent global-scale studies have suggested a shift from outward urban sprawl to vertical development, our findings indicate that horizontal expansion still dominates in the selected Chinese megacities, with outward sprawl exceeding vertical densification during the study period. The integrated approach provides a robust framework for mapping 3D urbanization and offers practical insights for policymakers seeking to manage horizontal expansion, guide vertical intensification, and optimize land-use efficiency in rapidly urbanizing megacities.

Remote Sensing doi: 10.3390/rs18121891

Authors: Zongduo Hao Xiaocui Dang Yanyu Zhang Jinshui Miao Zhiming Li Xiankai Lu

Foundation models such as the Segment Anything Model (SAM) have substantially advanced promptable object segmentation in remote sensing. However, extending these capabilities to infrared small target detection (IRSTD) remains highly challenging in the presence of severe background clutter and extremely low target visibility. In this paper, we propose IR-SAM2, an effective target enhancement framework for mask-level infrared small target segmentation in the IRSTD setting. Specifically, IR-SAM2 equips the SAM2 decoder with a dedicated frequency branch, facilitating simultaneous spatio-frequency learning and deep spatio-frequency fusion, while preserving SAM2’s pre-trained knowledge. Moreover, we introduce a target-centric loss to better guide the model in distinguishing small targets from complex backgrounds. Extensive experiments show that IR-SAM2 achieves highly competitive performance on the IRSTD-1k and NUDT-SIRST benchmarks, while striking an optimal balance between detection probability and false alarm rate on NUAA-SIRST. The results further demonstrate the effectiveness of spatio-frequency cues for complex-scene infrared small target segmentation. The source codes have been made publicly available to support reproducibility.

Remote Sensing doi: 10.3390/rs18121897

Authors: Alexei Kolgotin Detlef Müller

We present a novel algorithm for the retrieval of non-spherical particle microphysical parameters (PMP) from 3β + 2α + 3δ optical data taken with multiwavelength lidar. The 3β + 2α + 3δ optical datasets describe particle backscatter coefficients (β) at three wavelengths, λ = 355, 532, and 1064 nm, particle extinction coefficients (α) at two wavelengths, λ = 355 and 532 nm, and particle linear depolarization ratios (PLDR, δ) at three wavelengths, λ = 355, 532, and 1064 nm. The algorithm can be used for retrieving bimodal particle size distributions (PSDs). The PSDs can comprise mixtures of spheres and spheroids (SS). One or both modes can comprise spheroid-shaped particles or spherically shaped particles. The spheroids are used for approximating an arbitrary ensemble of non-spherical particles. The algorithm works on the basis of a combination of direct and analytical inversion methods. The algorithm uses the spheroid reference look-up table (RLUT) we developed and presented in part 1 of our research work. The algorithm uses constraints regarding the particle complex refractive index (CRI) and information on relative humidity (RH) in the atmosphere (in the case of aerosol lidar observation) for suppressing retrieval uncertainties. We carried out a numerical simulation study to evaluate the algorithm’s performance. In these numerical simulations, we considered perturbed synthetic 3β + 2α + 3δ optical data that mimic different organic carbon (OC)–dust (D) mixtures. Such mixtures are suitable examples for describing bimodal PSDs that consist of a fine mode of spherical particles and a coarse mode of non-spherical particles. The results of the numerical simulation show that (1) the PMPs of each mode of these particle mixtures can be found separately, (2) the mean retrieval errors of the effective radius, number, surface-area, and volume concentrations of these mixtures are 25%, 52%, 9%, and 28%, respectively, and (3) the mean retrieval error of single-scattering albedo (SSA) at 355 nm of these mixtures is as low as ±0.02. SSA retrieval accuracies at 532 and 1064 nm degrade because the complex refractive index (CRI) of OC and D particles depends on the measurement wavelength. In future studies, we will upgrade the algorithm such that it takes into account a spectrally dependent CRI. We also compare the results of our novel algorithm with our TiARA2.1 algorithm. The errors obtained from the TiARA2.1 algorithm are approximately three times larger compared to the errors we obtain with our novel ATLAS algorithm for the case of the OC-D mixtures considered in the present study. We explain the higher accuracy of the PMP retrievals by the use of three PLDRs and the extra constraints placed on CRI and RH.

Remote Sensing doi: 10.3390/rs18121896

Authors: Shuaiyong Lin Ding Nie Wangqiang Jiang Chuan Li

In the application of unmanned aerial vehicle (UAV) target perception in complex marine environments, the significant cyclostationarity of UAV radar echoes makes it highly suitable for extracting their signatures via cyclic spectral analysis. This method projects the signal onto the cyclic frequency dimension, exploiting the fundamental difference between the periodicity of the UAV’s micro-vibrations and the non-periodic randomness of sea clutter, enabling the effective and reliable extraction of the UAV’s target features. However, the sea-clutter background often masks the UAV signal, making it difficult to identify the target processing unit for cyclic spectral analysis rapidly. Autocorrelation processing excels at rapidly filtering out non-periodic components from the echo signal, thereby preserving and enhancing periodic components. It exploits the correlation between adjacent pulses to suppress slow clutter and enhance the echoes from moving targets, thereby establishing a target range for cyclic spectral analysis. Inspired by this, we first propose a novel method in this paper that innovatively employs autocorrelation-guided cyclic spectral fusion filtering, which effectively mitigates the short-term coherence and non-stationarity characteristics of strong sea-clutter background. Corresponding results with a measured strong sea-clutter background demonstrate that the proposed method effectively suppresses sea clutter and reliably extracts UAV target signals from other maritime targets. Compared with the classic moving target indicator (MTI) and the singular value decomposition (SVD) method, as well as their cascade processing, the proposed method achieves higher gain across various input signal-to-clutter-plus-noise ratios (SCNRs), demonstrating broad applicability and excellent detection performance.

Remote Sensing doi: 10.3390/rs18121894

Authors: Zizhen Jiang Yuxuan Liu Yuxin Wang Kai Chai Meimei Wang

The human–environment interaction on the Qinghai–Xizang Plateau determines the direction of global human sustainable development, making it necessary to propose a refined prediction for this relationship. Currently, there is a lack of a predictive method for human–environment relationships, especially at the grid scale. This study focuses on Qinghai Province and proposes a human–environment relationship simulation method based on cellular automata (CA), utilizing land-use data and a remote sensing-based ecological (RSEI) index. The method enables grid-scale explicit predictions of human–environment relationships. The results show that by 2030, the human–environment relationship in Qinghai Province will become more diverse, with the coordination ratio rising to 11% and the degradation ratio to 7%. The ecological protection scenario serves a defensive role, preventing 3835 km2 of land from degradation. In contrast, the urban development scenario plays a revitalizing role, achieving a coordinated area 2% larger than the business-as-usual scenario. By 2030, about 8956 km2 of land in Qinghai will be suitable for agricultural revitalization, and 54,340 km2 must be reserved for ecological protection. Due to the high-altitude environment, the human–environment relationship aligns only with the right half of the Environmental Kuznets Curve, namely, development brings greater harmony. We further discover the lag in the natural system’s response, for artificially increasing vegetation cover will not quickly improve habitat quality. Likewise, leapfrogging expansion in the urban development scenario may conceal long-term ecological risks behind short-term coordination. For stakeholders and policymakers, this study provides refined and differentiated governance measures at the grid scale, while highlighting the need to focus on underdeveloped regions and remain vigilant about the lag in human–environment relationship responses.

Remote Sensing doi: 10.3390/rs18121890

Authors: Caihong Ma Nian Liu Yi Zeng Kai Qin Xin Sui

Industrial heat source (IHS) radiation areas are key accumulation zones for fine particulate matter with an aerodynamic diameter of 2.5 μm or less (PM2.5) in heavily industrialized regions, but PM2.5 concentration reduction rates by vegetation types have not been systematically assessed. A systematic analysis of PM2.5 concentration reduction rates by vegetation types using IHS, PM2.5, land-cover, and digital elevation model (DEM) data was conducted to assess PM2.5 concentration reduction rates by vegetation types within IHS radiation areas. First, this study adopted the published IHS radiation-area dataset developed by Xin Sui et al. to define the spatial extent of industrially influenced areas. Second, PM2.5 concentrations were extracted within IHS radiation areas and areas covered by different vegetation types to support the calculation of PM2.5 concentration reduction rates. Third, PM2.5 concentration reduction rates by vegetation types were evaluated through masking and regional statistical analysis. Results for Beijing–Tianjin–Hebei (BTH) in 2015 and 2020 show that: (1) the average PM2.5 decreased from 71.70 to 39.60 µg/m3, corresponding to an overall reduction of 44.8%; (2) PM2.5 concentration reduction rates varied substantially among vegetation types; open deciduous broadleaved forest showed the highest reduction rate of 39.08%, while rainfed and irrigated cropland showed negative reduction rates of −9.35% and −6.71%; (3) city-scale and case analyses show denser vegetation in radiation zones generally lowers PM2.5 even under ongoing industrial activity. The study supports vegetation greening, IHS control, regional air quality improvement, and sustainable industrial development strategies.

Remote Sensing doi: 10.3390/rs18121893

Authors: Kai Liu Hongshuai Qi Hang Yin Feng Cai Gen Liu Shaohua Zhao Jixiang Zheng

The intertidal beach profile provides a fundamental representation of beach morphology and serves as a key indicator of shoreline morphodynamics. To enable frequent and accurate mapping of intertidal beach profiles, this study proposes an automated reconstruction framework that integrates single-pixel image columns with a stacked bidirectional long short-term memory (Bi-LSTM) network. Time-exposure imagery, commonly referred to as Timex imagery, acquired from a shore-based video monitoring station at Xisha Bay, China, is used as the primary data source, while wave records obtained from a wave buoy are incorporated to assign elevations to the detected waterline breakpoints, thereby enabling automatic beach profile reconstruction. The stacked Bi-LSTM network is trained for land–sea segmentation and waterline breakpoint localization. achieving the best performance among the tested methods, with precision, recall, accuracy, and F1 score values of 0.951, 0.894, 0.978, and 0.903, respectively, and a mean breakpoint localization error of 2.23 pixels. Breakpoint elevations were then estimated using a local slope–wave setup attribution model. Validation against field-measured topographic data from four fixed profiles and three survey periods showed good agreement between the reconstructed and measured profiles, with a period-based root mean square error (RMSE) of 0.212 ± 0.080 m. When all validation points were combined, the reconstructed elevations showed strong agreement with the measured elevations, with a coefficient of determination (R2) of 0.988 and an overall RMSE of 0.24 m. The profile comparisons further showed that the reconstructed profiles generally captured the overall profile shape and cross-shore morphological pattern of the measured profiles, although reconstruction accuracy varied among the four fixed profiles. These differences demonstrate that camera viewing angle, field-of-view position, camera-to-profile distance, and image quality are important factors influencing video-derived beach profile reconstruction. These results indicate that the proposed method can directly reconstruct fixed intertidal beach profiles from shore-based Timex imagery without generating a digital elevation model of the entire intertidal zone. It provides a practical tool for high-frequency monitoring of intertidal profile morphology and supports the quantitative analysis of beach erosion–accretion dynamics.

Remote Sensing doi: 10.3390/rs18121892

Authors: Guanyi Wang Weihua Bai Feixiong Huang Yueqiang Sun Junming Xia Xianyi Wang Xiangguang Meng Peng Hu Cong Yin Guangyuan Tan Ruhan Wu Yunlong Du Xiaofeng Meng

The Global Navigation Satellite System Reflectometry (GNSS-R) technique provides global ocean surface wind observations unaffected by rainfall with high spatiotemporal resolution. The Fengyun-3E (FY-3E) mission, as the first operational GNSS-R satellite in China, offers low-latency data suitable for numerical weather prediction (NWP). However, the dense along-track sampling of GNSS-R winds poses challenges for observation error specification in data assimilation. In this study, FY-3E GNSS-R winds are assimilated into the Weather Research and Forecasting (WRF) model to investigate the impacts of different observation error configurations. Both static and dynamic error specifications, with and without data thinning, are evaluated through a sensitivity experiment and subsequent Observing System Experiments (OSEs). The results indicate that using a static observation error of 6 m/s without data thinning achieves the best performance. Under this configuration, GNSS-R winds influence atmospheric analyses from the surface up to approximately 700 hPa in a single assimilation case, while cycling experiments further extend the impact vertically and spatially. These findings highlight the importance of appropriate observation error specification for dense GNSS-R data and provide a practical reference for their assimilation in WRF, with potential applicability to other NWP systems.

Remote Sensing doi: 10.3390/rs18121889

Authors: Jin-Woo Koh HyunSeong Sung

Image registration accuracy is strongly influenced by imaging geometry, yet this effect has not been explicitly characterized in a closed-form expression under practical plane-based image registration conditions. This paper establishes a geometry-induced lower bound on the image registration error within a plane-based image registration framework. While prior work has primarily focused on improving accuracy through algorithmic advancements, we show that the residual registration error under practical 2D registration conditions is fundamentally influenced by imaging geometry and height variation, regardless of the image registration performance. A closed-form expression of the lower bound is derived as a function of imaging geometry and height offset. The formulation explicitly characterizes how geometric configuration governs displacement in the reference image space or registration plane. The analysis reveals that, even under reliable matching conditions, residual errors may persist due to the inherent coupling between viewing geometry and elevation variation. The derived bound is validated using multiple image pairs acquired under different geometric configurations. The experimental results show that the formulation captures the dominant geometric effect. The proposed formulation provides a practical and interpretable geometric reference for analyzing registration accuracy under varying imaging configurations.

Remote Sensing doi: 10.3390/rs18121888

Authors: Shanhong Guo Ji Zhu Gao Chen Mu Yang Weixing Sheng

Target detection in synthetic aperture radar (SAR) images is constrained by three primary challenges. First, speckle noise overlaps heavily with the high-frequency features of target edges in the frequency domain, so standard convolutions cannot suppress noise without sacrificing edge texture. Second, the scattering signature of a SAR target varies markedly with viewing angle, and a fixed-parameter convolution kernel cannot accommodate this spatial non-stationarity. Third, deep and shallow levels of the feature pyramid differ in semantics and resolution, and a naive element-wise sum either introduces noise interference or loses small-target signals. We propose the Frequency–Spatial Detection Network (FSDNet), whose core FSDBlock cascades three operators to address these failure modes in turn. Wavelet Convolution (WTConv) projects features into Haar sub-bands and applies independent low- and high-frequency kernels prior to inverse-DWT reconstruction, suppressing noise while preserving edges. Receptive-Field Attention Convolution (RFAConv) generates location-conditional kernels and so adapts to non-stationary scattering. Spatial Context Self-Attention (SCSA) aggregates discrete scattering points into coherent target representations via long-range grouped attention. At the fusion stage, CGAFusion replaces FPN element-wise addition with a channel–spatial–pixel triple-attention soft switch that mitigates deep–shallow semantic mismatch. On HRSID, FSDNet attains mAP50 = 92.3% and mAP50:95 = 68.6%. On SSDD, it attains mAP50 = 98.7% and mAP50:95 = 74.2%. Both sets of results consistently surpass the baseline methods. Against the strongest YOLO baseline (YOLOv11n), FSDNet improves HRSID mAP50 by +1.7 percentage points (pp) and mAP50:95 by +2.3 pp, and SSDD mAP50 by +0.5 pp and mAP50:95 by +2.7 pp; against the capacity-fair YOLOv11s reference (∼51% more parameters), FSDNet still leads on mAP50, mAP50:95, recall, and F1. Ablation studies and power-spectral-density analyses corroborate the contribution of each module and confirm WTConv’s role in preserving high-frequency target features.

Remote Sensing doi: 10.3390/rs18121887

Authors: Maopeng Sun Wen Liu Xiaoming Li Shiyan Hong Renzhong Guo

Urbanization and rapid economic growth have exacerbated urban heat effects, increasing the frequency of heat-related disasters and intensifying human health risks. Urban traffic generates substantial carbon emissions and associated heat, which significantly alter roadside thermal environments and impact human activities. Numerous previous studies have investigated urban thermal environments and their influencing mechanisms. However, the relationships between road-level traffic carbon emission (TCE) and road surface temperature (RST) remain insufficiently explored. In this study, roadway segment-based TCE and RST were acquired by integrating hourly traffic flow information, localized vehicle carbon emission factors, high-resolution Landsat-8 remote sensing datasets, and the road network. Three commonly used linear regression models and an improved Random Forest (RF) model were utilized to assess the impact of TCE on RST for different grades of roads. The study showed that carbon emissions from road traffic exhibit a locally focused distribution pattern in space. Compared to other grades of roads, higher levels of TCE were observed in urban main roads. In summer, roads (e.g., minor arterials) with lower grades tended to have a higher thermal risk, with freeways having the lowest TCE and urban expressways experiencing the greatest TCE fluctuations. An improved RF model integrating the spatial weight matrix and Gaussian process could more efficiently identify the nonlinear effects of TCE on RST. The contributions of TCE to summer RST were 0.4, 0.37, 0.54, and 0.56 for freeways, urban expressways, main roads, and minor arterials, respectively. The relative impact of road TCE with lower grades on RST becomes more significant, while the impact of surrounding buildings and green areas tends to decrease. Our findings provide valuable insights for reducing urban carbon emissions and thermal risks.

Remote Sensing doi: 10.3390/rs18121886

Authors: Chang Zhong Yu Zhao Mengchun Pan Qi Zhang Xinxin Sui Li Chen Ning Wang Fan Bu

Lake depth is a fundamental parameter for estimating lake storage, analyzing basin morphology, and understanding the evolution of plateau lakes. Compared with typical shallow lakes, Tibetan Plateau lakes are characterized by high elevation, strong radiation, pronounced inter-lake and inter-annual variability, and in some cases considerable basin depth, which limits the accuracy, stability, and generalization ability of existing bathymetric inversion methods based on single-source optical imagery. Meanwhile, although ICESat-2 can provide sparse but high-precision along-track bathymetric constraints, a unified framework suitable for plateau-lake scenarios is still lacking. To address this issue, this study proposes TabKAN, a bathymetric inversion framework for Tibetan Plateau lakes under joint constraints from hyperspectral imagery and ICESat-2 data. TabKAN constructs tabular input features from hyperspectral reflectance, water indices, imaging geometry, and environmental variables; employs TabNet for feature selection and encoding; and introduces a KAN regression head to enhance nonlinear bathymetric mapping. A joint-supervision and bias-correction mechanism is further designed to incorporate ICESat-2 samples, thereby improving model robustness across lakes and acquisition dates. To enhance the temporal coverage of training samples, multi-year sample expansion based on stereo-mapping data is introduced, and a stripe-aware self-supervised learning strategy is developed for hyperspectral image restoration and pretraining. Experiments on five Tibetan Plateau lakes, including Anglaren Co, Caiduo Chaka, Cuoe, Geren Co, and Qixiang Co, show that the proposed method outperforms benchmark methods in both overall accuracy and depth-stratified evaluation, while providing more stable recovery of basin morphology and depth gradients. These results demonstrate that combining hyperspectral information, ICESat-2 laser constraints, and stripe-aware pretraining can effectively improve the accuracy and robustness of bathymetric inversion for Tibetan Plateau lakes and provide a new technical route for storage estimation and change monitoring of cold inland lakes.

Remote Sensing doi: 10.3390/rs18121885

Authors: Zhonghu Jiao Xihan Mu

High-resolution estimation of global near-surface air temperature (Ta) is essential for investigating microclimates, ecosystem processes, and agricultural suitability. However, sparse in situ observations do not capture local heterogeneity, whereas existing datasets lack fine-scale detail because of their coarse spatial resolution. To address this limitation, we developed an ensemble machine-learning framework using Landsat 8/9 data. Predictions from LightGBM, XGBoost, and CatBoost were combined through Bayesian model averaging (BMA), which assigns probabilistic weights to individual models to improve robustness. The models were trained using a globally distributed spatiotemporal matchup dataset that paired HadISD in situ Ta observations with MODIS/VIIRS products to support subsequent Landsat-based application. Key inputs included land surface temperature (LST), vegetation indices, elevation, solar zenith angle, and spatiotemporal features. The BMA ensemble achieved strong validation performance, with an RMSE of ~3 K, near-zero bias, and an R2 of 0.92. Feature-importance analysis identified LST as the dominant predictor, underscoring the primary role of surface thermal state in estimating Ta. The proposed method can generate robust global Ta fields at 90 m resolution, revealing fine-scale thermal patterns that have previously been difficult to resolve at the global scale. Unlike many regional models calibrated for single study area or dependent on dynamic external auxiliary fields, our Landsat-predominant application framework supports operational mapping of clear-sky and overpass-time Ta. Such detailed instantaneous data can advance climate research, improve assessments of ecological responses and climate impacts, and support applications such as urban heat island monitoring and precision agriculture.

Remote Sensing doi: 10.3390/rs18121884

Authors: Dong Wang Lijuan Miao Yutian Lu Hanyang Jiang Qiang Liu

The Leaf Area Index (LAI) serves as a vital biophysical parameter for quantifying vegetation dynamics and ecosystem functioning. While traditional LAI retrieval methods face challenges in handling nonlinear spectral-vegetation relationships, machine learning (ML) approaches offer promising alternatives through their data-driven adaptability. This study presents a comprehensive cross-site assessment of 13 ML algorithms for LAI estimation, leveraging ground observations from 98 sites worldwide. Our systematic assessment reveals three key findings: First, ensemble methods consistently outperformed other approaches, with Gradient Boosted Tree Regression (GBTR) achieving superior accuracy (R2 = 0.647, RMSE = 0.899) and robustness (ΔR2 < 0.05 beyond n = 69 training samples). Second, Gaussian Process Regression (GPR) illustrated exceptional stability across varying training sizes (R2 = 0.607 ± 0.012), highlighting its reliability for data-limited scenarios. Third, all tested ML models substantially outperformed operational LAI products, with the GBTR model demonstrating superior explanatory power (external validation R2 = 0.647) compared to MODIS; its R2 value had increased by 0.489. This optimal balance of accuracy, computational efficiency, and resistance to overfitting positions GBTR as a reasonable choice for large-scale LAI mapping. These findings underscore ML’s promising potential in vegetation monitoring while highlighting the need for hybrid approaches that combine physical principles with data-driven learning to address current limitations in extreme-value estimation and ecological generalizability.

Remote Sensing doi: 10.3390/rs18121881

Authors: Dongming Lu Zuchao Bao Zechen Tian Yifan Zhai Tingting Chen Jianpo Gao

Detecting infrared moving targets in low-altitude remote sensing scenes remains challenging due to strong clutter, scale inconsistency, and residual interference. Because these factors are often coupled in complex scenes and cannot be handled effectively by a single operation, a three-stage progressive Decoupled Separation, Enhancement, and Purification (DSEP) framework is proposed. The method integrates edge-preserving background decoupling, scale-consistent spatial screening, and residual-response purification into a non-iterative feedforward pipeline. Experiments on six representative self-collected infrared sequences and six selected scenes from the public SIRST dataset suggest that DSEP produces relatively compact and spatially continuous target responses while suppressing background interference. On the self-collected dataset, the method can achieve SCRG and BSF values up to 10.61 and 7.38, respectively, with a processing time of 0.009–0.016 s per frame. Compared with representative spatial filtering, local contrast, and low-rank decomposition methods, DSEP shows a relatively favorable balance between detection performance and low-latency processing efficiency. Although the performance gain becomes smaller in some SIRST scenes, the proposed method still shows generally stable detection performance across the evaluated scenes.

Remote Sensing doi: 10.3390/rs18121883

Authors: Asma Tanveer Juanle Wang Faith Ka Shun Chan

Glacial lake outburst floods (GLOFs) are a major cryosphere-related hazard in High Mountain Asia (HMA), where glacier mass loss and changing hydroclimatic conditions are reshaping glacial-lake systems and increasing the prevalence of potentially unstable lake–dam configurations. However, current knowledge remains fragmented across HMA. Therefore, this review synthesizes historical evidence, future changes, and risk-reduction strategies of GLOFs across HMA from a remote-sensing perspective. Historical evidence derived from satellite archives, multi-temporal lake inventories, geomorphological analyses, and documented event records indicate that reported GLOFs in HMA are strongly clustered by sub-region and dam type, with moraine-dammed lakes representing the dominant source of documented events, while ice-dammed lakes remain important in several mountain belts. The compiled record also shows that GLOFs have caused severe human, economic, geomorphic, and ecological losses. Future projections based on glacier evolution, glacial-lake expansion, and climate-sensitive hazard assessments indicate continued glacial-lake growth under global warming. However, reliable prediction of future GLOF event timing, magnitude, and frequency remains constrained by uncertainties in glacier evolution, dam stability, and triggering processes. This review further shows that effective GLOF risk reduction in HMA requires integrated systems that combine hazard and risk mapping, early warning, structural interventions, and non-structural measures. It also highlights the need to better link remote sensing with monitoring, assessment, and implementation frameworks, and proposes an integrated management cycle to support practical risk reduction. It concludes that the most urgent research priorities are harmonized multi-temporal lake inventories, targeted field observations, explicit consideration of heatwaves and compound extremes, transparent uncertainty propagation, and stronger operationalization of monitoring and warning systems to support durable climate adaptation and disaster risk reduction across HMA.

Remote Sensing doi: 10.3390/rs18121882

Authors: Yifei Liufu Lisiren Cao Jiali Yang Jiapei Li Fangyu Cao Yuqin Huang Jinyao Lin

Mitigating nitrogen dioxide (NO2) pollution is a critical objective for enhancing urban environmental quality. The spatial structure of urban tree canopies plays a crucial role in influencing NO2 diffusion and deposition. However, previous studies have focused mainly on the linear relationships between two-dimensional green spaces and NO2, while the associated nonlinear relationships and threshold effects of three-dimensional urban tree canopy (UTC) spatial structure remain underexplored. To address this gap, we leveraged 1 m resolution satellite-derived data and explainable machine learning (XGBoost, SHAP, PDP) to examine the nonlinear influences and threshold effects of three-dimensional UTC spatial structures on NO2 in Shenzhen. The results revealed that urban tree canopy spatial structure is associated with NO2 concentrations. Among the key metrics, the two-dimensional canopy coverage ratio (CCR) emerged as the primary canopy-related correlate of lower NO2 concentrations, while three-dimensional vertical structure metrics, particularly canopy height variability (CHV) and standard deviation of canopy height (SDCH), acted as critical secondary correlates in modulating the spatial distribution of pollutants. Based on these relationships, we identified potential threshold ranges for key metrics by comparing mathematically identified inflection points with practical urban planning constraints. In summary, this study advances the spatial analysis of “green spaces-NO2” interactions from a two-dimensional to a three-dimensional perspective. Our findings could provide quantitative guidance for optimizing green space structure in high-density urban areas to inform strategies potentially associated with improved NO2 outcomes.

Remote Sensing doi: 10.3390/rs18121880

Authors: Dejie Huang Xiaowen Zhang Fuquan Zhang

Affected by global climate change and complex environmental factors, the frequency and intensity of forest fires have been rising. Accurate early detection is crucial for disaster mitigation. Traditional methods (e.g., manual monitoring) suffer from low efficiency or limited coverage, while deep learning methods (e.g., YOLO (You Only Look Once), Faster RCNN (Region-based Convolutional Neural Networks)) perform well but are sensitive to degraded images (haze, low light), reducing accuracy. To address blurred smoke features and attenuated flame brightness in degraded images, this paper proposes CoDeF-Net (Collaborative Detection Framework Network), a collaborative detection framework integrating Retinex-BCE (Retinex-based Bright Channel Enhancement) image enhancement with YOLOv11 (You Only Look Once version 11) to improve robustness. Experiments on 1757 real forest fire images show that Retinex-BCE achieves an FSIMC (Full-Reference Image Quality Assessment Metric based on Structural Similarity and Contrast) index of 0.9611 and an LOE (Loss of Edge) value of 254.78, preserving image structure. CoDeF-Net reaches AP@0.5 (Average Precision at Intersection over Union threshold 0.5) of 87.9% (3.8% higher than original YOLOv11), with low missed detection of small flames and enhanced stability in extreme scenarios, providing a feasible solution for forest fire monitoring under degraded images.

Remote Sensing doi: 10.3390/rs18121878

Authors: Wei Chen Jie Song Wei Xiong

In maritime electronic support measures (ESMS), multipath-scattered pulses are often suppressed during pulse sorting, although their delay, amplitude, and angular differences may provide information for passive scatterer localization. This paper investigates a front-end path-classification task positioned after emitter-level clustering and before multipath-assisted passive localization. Pulses produced by the same non-cooperative emitter but received through different propagation paths are classified as direct-path or multipath-scattered pulses. The task is formulated as supervised binary classification over PDW sequences. Five representative solution families are evaluated under a common protocol: FCM, DBSCAN, temporal sequence analysis (TSA), Single-LSTM, and a residual two-layer unidirectional LSTM with residual fusion. The input features are RF, PA, PW, PRI, TOA, DOA, and ΔTOA; the recurrent models use class-weighted training to address the direct/scattered class imbalance. Across 36 coupled scenarios with pulse-loss rates from 0% to 50% and parameter-jitter levels from 0.0 to 1.0, the residual LSTM obtains the highest average macro-F1 score (0.8717), compared with Single-LSTM (0.7726), DBSCAN (0.7686), TSA (0.6511), and FCM (0.5917). Repeated training over four random seeds yields a validation macro-F1 of 0.9821 ± 0.0007 on the original validation set. The ablation results indicate that ΔTOA is the principal temporal cue in this setting, while LayerNorm, residual fusion, class weighting, and augmentation mainly contribute to optimization stability and perturbation robustness. Measured-data verification suggests that the learned temporal representation can provide usable inputs for subsequent scatterer localization. The current validation is limited to a one-emitter simulation and rule-assisted measured-data annotation; mixed-emitter validation and quantitatively calibrated localization evaluation remain subjects for future study.

Remote Sensing doi: 10.3390/rs18121879

Authors: Wei He Lianfa Li Haoxiong Wu Xilin Gao Yichen Yang Zixuan Zhang Xiaomei Yang Yong Ge

Deep learning has exhibited potential in land use and land cover (LULC) classification applications. However, the effectiveness of deep learning remains constrained by the availability and quality of annotated training data. The persistent scarcity of labeled samples and spatial heterogeneity of remote sensing imagery hinder the robustness and generalization of trained models. Meta-learning, commonly referred to as “learning to learn”, is a paradigm that trains models over a distribution of tasks to acquire transferable knowledge, enabling rapid adaptation to new tasks with only a few labeled samples. This cross-task learning capability makes meta-learning a promising solution to data scarcity and spatial heterogeneity in the remote sensing context. This paper provides a systematic review of meta-learning applications in LULC classification, identifying a total of 70 relevant studies between 2018 and 2025. Three mainstream meta-learning paradigms (memory-augmented, optimization-based, and metric-based) are reviewed, and the applications are analyzed across four core challenges in LULC remote sensing: label scarcity, cross-region and cross-domain distribution shifts, temporal dynamics modeling, and multimodal data integration. The review reveals that optimization-based and metric-based methods dominate current research, with MAML and its variants being the most widely adopted due to the model-agnostic property, while memory-augmented methods remain underexplored. A consistent finding is that meta-learning outperforms conventional pre-training followed by fine-tuning under significant domain shifts across multiple data modalities. Current limitations, including computational overhead, episodic training constraints, and the lack of standardized evaluation protocols, are discussed. Future directions in cross-domain generalization, integration with foundation models, novel architectures, and standardized benchmarks are identified.

Remote Sensing doi: 10.3390/rs18121877

Authors: Péter Farkas Gábor Timár

The continuous analysis of ground deformation is essential for both the assessment of natural hazards and the monitoring of human-induced activities. In this study, we present the results of a Persistent Scatterer Interferometry (PSI) analysis of ground deformations in the region of Cluj-Napoca, Romania. The PSI was performed using more than 10 years of Sentinel-1 ascending and descending Synthetic Aperture Radar data from 2014 to 2025, using a dual master approach. Results show significant displacements at many locations, including recently built-up areas at the edges of the city, often caused by the combined effect of anthropogenic activities and geological conditions. In this study, we highlight three case studies: the surroundings of a reclaimed mine, subsidence induced by dewatering, and a large-area, slow landslide, wherein we examined natural and anthropogenic influences. The accurately mapped and quantified ground deformations can be used for a better understanding of the geological processes and assessing the risk of the urban development in the area.

Remote Sensing doi: 10.3390/rs18121876

Authors: Jin Yan Binghui Chen Jin Xu Zekun Guo Minghao Yan Mengxin Sun Lin Qiao

Marine oil spill accidents pose a serious threat to the marine ecological environment. Therefore, efficient and accurate oil spill detection is of great significance for emergency response. To address the issues of blurred oil-slick boundaries, prominent co-frequency interference and severe speckle noise in shipborne radar images, this study proposed an oil spill detection method based on radar data collected from a real oil spill event at a terminal in Dalian Bay. The proposed method integrates an autoencoder, feature dimensionality reduction, pseudo-labeling, reinforcement learning and an improved intelligent optimization algorithm. First, an autoencoder was adopted to extract compact nonlinear local features from the radar images, and principal component analysis (PCA) was employed for feature dimensionality reduction. Subsequently, K-Means clustering was used to construct pseudo-labels, and the reduced features were discretized to build the state space for reinforcement learning. Based on this, the Q-learning algorithm was introduced to automatically extract the region of interest (ROI). Finally, for the ROI, an improved bat algorithm incorporating a dynamic weighting factor and a multi-constraint fitness function was designed to achieve fine segmentation of the oil-slick target. The experimental results showed that the proposed method outperformed classic intelligent optimization algorithms and the conventional bat optimization algorithm in oil-slick segmentation performance. Ablation experiments further verified the effectiveness of autoencoder-based feature learning, K-Means pseudo-labeling, and Q-learning-based ROI localization. This method may provide a new technical approach for timely offshore oil spill monitoring and emergency analysis.

Remote Sensing doi: 10.3390/rs18121875

Authors: Yuan Gao Jindong Zhao Shaobo Xia Sheng Nie Cheng Wang Xiaohuan Xi

Semantic segmentation of large-scale airborne point clouds traditionally relies on labor-intensive 3D manual annotations. While recent zero-shot methods attempt to alleviate this burden by distilling knowledge from 2D Vision–Language Models (VLMs) via 2D-to-3D projection, they suffer from performance degradation in complex urban environments. Specifically, lacking 3D geometric awareness, 2D VLMs frequently exhibit “semantic bleeding”, where large-scale background categories (e.g., ground) erroneously submerge small-scale targets (e.g., vehicles and street elements). To address this issue, we propose a geometry-constrained pseudo-label generation and purification framework. Our approach tackles the problem through a dual-branch design: extracting open-vocabulary semantics via SAM3-based multi-view projection while simultaneously deriving sharp, class-agnostic instances using SAM2 on Gamma-transformed elevation maps. By introducing a geometric–semantic consistency module, we evaluate the internal semantic purity and external spatial homogeneity of these instances, detecting and filtering out semantic misclassifications. The purified pseudo-labels are then used to supervise a 3D sparse convolutional network via a Masked Cross-Entropy Loss. Experiments on the H3D and Turin3D datasets demonstrate that our method recovers small-scale targets that are prone to being submerged, outperforming existing zero-shot baselines by improving mIoU from 52.15% to 63.45% on H3D and from 29.52% to 58.51% on Turin3D, thereby narrowing the performance gap with fully-supervised approaches.

Remote Sensing doi: 10.3390/rs18121874

Authors: Mohammad Marjani Masoud Mahdianpari Seyed Ehsan Khankeshizadeh Sahand Tahermanesh Amin Mohsenifar Ali Mohammadzadeh

Wildfires pose significant threats to ecosystems, economies, and human health. The integration of remote sensing (RS), geospatial information systems (GIS), and artificial intelligence (AI) has emerged as a powerful approach for addressing wildfire-related challenges. However, existing review studies typically focus on specific wildfire tasks and lack a comprehensive synthesis of how geospatial data and supervised AI techniques interact across the full wildfire management cycle. Therefore, this study aims to provide a meta-analysis review of the integration of RS, GIS, and supervised AI methods in wildfire science. This study follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework to systematically analyze 449 peer-reviewed journal articles published between 1994 and 2024. The review examines various wildfire-related tasks, data sources, algorithmic approaches, spatial scales, performance metrics, and other aspects used in wildfire geospatial AI (GeoAI) studies. The results reveal a strong concentration of research on tasks such as burned area mapping (BAM), wildfire detection, and susceptibility mapping, while critical areas, such as fuel mapping, wildfire vulnerability, and post-fire recovery, remain underexplored. The analysis also identifies a dominant use of traditional machine learning (ML) algorithms, such as Random Forest (RF), and an increasing adoption of deep learning (DL) models, particularly convolutional neural networks (CNNs). Furthermore, the geographic distribution of studies highlights significant global disparities, with most research conducted in high-income regions, while wildfire-prone areas in developing regions remain underrepresented. The review also reveals limited adoption of advanced AI techniques, including transfer learning, transformer architectures, Geo-foundation AI models, and explainable AI (XAI). These findings provide a comprehensive synthesis of GeoAI applications in wildfire management and highlight critical methodological, geographic, and application-level gaps. Addressing these gaps through improved data accessibility, adoption of advanced AI methods, and increased research focus on underrepresented wildfire tasks and regions will be essential for developing scalable, interpretable, and globally applicable wildfire management systems.

Remote Sensing doi: 10.3390/rs18121872

Authors: Zhuge Xia Huan Liu Kun Qian Qi Zhang Jiacheng Xiong Qihuan Huang Xiufeng He

Accurate time series forecasting is fundamental to geohazard early warning, yet remains a major challenge. Conventional in situ geotechnical monitoring remains costly and spatially constrained, whereas deep learning applied to remote sensing data has become increasingly prevalent but often suffers from opacity of model decision-making. To address this issue, we propose a Transformer-based forecasting framework, namely PatchTST-Fusion, adapted for multivariate Multi-Temporal Interferometric Synthetic Aperture Radar (MT-InSAR) time series. The framework integrates model interpretability analysis through TimeSHAP, providing temporal and feature-level attributions across the input sequence. Landslide deformation time series are first derived from Copernicus Sentinel-1 SAR data. Variational Mode Decomposition is then applied to decompose the non-linear signals into trend, seasonal, and noise components. The denoised displacement series are modeled and forecast using the proposed PatchTST-Fusion, which incorporates rainfall and reservoir water level fluctuations as feature-level drivers. Application to the Daping landslide cluster in the Three Gorges Reservoir Area in China demonstrates that our method captures both the long-term and transient non-linear coupling between deformation and its triggers, surpassing state-of-the-art models including CNN-BiGRU-Attention, Informer and original PatchTST with 7–55% improvements in MAE and 10–52% improvements in RMSE. Beyond predictive gains, feature attribution of environmental triggers via TimeSHAP reveals that rainfall and reservoir regulation exert temporally distinct influences on slope kinematics, with high relative importance concentrated in specific periods and characteristic lagged responses. This interpretable framework provides both enhanced forecasting accuracy and process-based insights, offering a broadly applicable tool for landslide early warning in reservoir regions.

Remote Sensing doi: 10.3390/rs18121873

Authors: Jingpan Bai Yan Shi Yuan Chen Houling Ji

To address the issues of low accuracy in transmission line anomaly detection and recognition caused by challenges such as multi-scale targets and partial occlusion, this paper proposes an optimized transmission line anomaly detection algorithm based on improved YOLOv11n. Firstly, an improved Cross Stage Partial with kernel size 2 (C3k2_DFF) module takes the place of the original C3k2 module in the backbone network. It adaptively fuses multi-scale local features and dynamically selects salient channel-wise and spatial features according to its global information during fusion to enhance the model’s feature representation. Secondly, a Separated and Enhancement Attention Module (SEAM) attention mechanism is introduced to enhance the unoccluded area feature response to compensate for the occluded area response deficit, suppressing the background features that interfere with the model and improving the model’s occluded target perception capability. Experimental results on our self-constructed dataset indicate that the proposed improved YOLOv11n model achieves precision, recall, mAP50, and mAP50-95 of 94.2%, 90.8%, 94.3%, and 68.0%, respectively. Compared with the baseline model, it represents improvements of 2.0%, 1.3%, 1.6%, and 2.9%, while the parameters and GFLOPs increase by only 6.6% and 4.8%, demonstrating superior detection performance.

Remote Sensing doi: 10.3390/rs18121871

Authors: Sofia Rossi Anna Balenzano Davide Palmisano Cinzia Albertini Francesco P. Lovergine Francesco Mattia Vanessa Paredes Gómez David Nafría García Giuseppe Satalino

The purpose of this work is to investigate the use of high-resolution (~100 m) surface soil moisture (SSM) maps derived from Sentinel-1 (S-1) and Sentinel-2 (S-2) data to identify irrigation events occurring in the Riaza irrigation district (Castilla y León region, Spain) from 2017 to 2021. The proposed method is based on the application of the Constant False Alarm Rate (CFAR) algorithm, which is an adaptive and unsupervised thresholding algorithm traditionally used for target detection in SAR images. This algorithm uses a sliding window approach that allows an adaptive threshold estimate for each pixel of the image, depending on the distribution of the surrounding pixels. The analysis was carried out on fields cultivated with maize, sugar beet and sunflower. Results show that the Overall Accuracy (OA) of the detection mainly depends on the time span (TS) between the S-1 passage and the irrigation event, the acquisition timing and the development stage of the vegetation. Indeed, the OA reaches a mean of 78% and 70%, respectively, for the 6 a.m. and 6 p.m. acquisitions, when the irrigation events occur within 36 h before the S-1 passage, and it follows a downward trend as the TS increases. On the other hand, when the vegetation reaches the mature stage, the mean OA decreases respectively to 56% and 52%. Stemming from the event detection, the study explored the estimation of the total irrigated area in the early growing season, showing promising agreement with in situ data, as evidenced by the low Relative Error (Er≈5.6%). Additionally, the analysis revealed a significant correlation between field-scale mean SSM and irrigation depths (R=0.89).

Remote Sensing doi: 10.3390/rs18121870

Authors: Oswaldo Maillard Robin L. Chazdon Sebastián Aguiar Bonifacio Mostacedo André Nunes Cristina Vidal-Riveros Roberto Vides-Almonacid

Fire is an essential component of species, ecosystems, and atmospheric dynamics. However, human activity has caused changes in fire regimes over the past two decades. In many cases, the spatial patterns of vegetation change after fire at the landscape scale remain unknown. The aim of this study was to evaluate spatial vegetation trends in burned areas across the Americas (2001–2024), using non-parametric tests and analyzing Normalized Difference Vegetation Index (NDVI) remote sensing products. Over a period of 24 years, fire activity burned a total area of 429.7 million hectares in 44 countries or territories and 269 ecoregions in the Americas. Regarding fire recurrence, the data indicates that 244.7 Mha (56.9%) burned only once (≤1), while 185.0 Mha (43.1%) burned multiple times (≥2), with certain regions experiencing up to 39 fires. The NDVI trend analysis showed that burned areas with increasing trends (p < 0.05) represented a total of 149.6 Mha (34.8%), primarily in Brazil (54.6 Mha, 12.7%), Argentina (17.8 Mha, 4.2%), the United States (14.4 Mha, 3.4%). In terms of decreasing NDVI trends (p < 0.05), these represented a total of 91.8 Mha (21.37%), primarily in Brazil (29.1 Mha, 6.8%), Canada (23.4 Mha, 5.4%), and the United States (14.2 Mha, 3.3%). The ecoregions with the largest areas showing increasing NDVI trends (p < 0.05) were the Cerrado (33.8 Mha, 7.8%), the Llanos (13.3 Mha, 3.1%) and the Humid Chaco (7 Mha, 1.6%). In contrast, the ecoregions with the largest areas showing decreasing NDVI trends (p < 0.05) were the Dry Chaco (9.2 Mha, 2.1%), the Cerrado (8.6 Mha, 2.0%), and the Boreal Shield (8.3 Mha, 1.9%). In terms of land cover types, savannas (37.2%) exhibited the highest proportions of increasing NDVI trends (p < 0.05), while decreasing trends were also present in savannas (28.0%) and grasslands (22.1%). Identifying spatiotemporal trends in vegetation change after fires is a fundamental step in implementing strategies and public policies to ensure ecological restoration. Moreover, given the high costs of restoration efforts, governments must work together to prevent these ecosystems from burning repeatedly.

Remote Sensing doi: 10.3390/rs18121869

Authors: Biao Wang Daochun Huang Yifeng Lin Xu He Zhengxian Guo Bo Hong

Early wildfire monitoring in power transmission corridors requires reliable detection of weak fire and smoke cues under complex field conditions and strict edge-computing constraints. To address these issues, this paper proposes an edge-deployable RGB–thermal framework based on visible and thermal infrared (TIR) imaging for unmanned aerial vehicle (UAV)-based corridor monitoring, including a spatial detector, YOLO-MMSC, and a temporal-enhanced version, YOLO-MMSC-T. The study also establishes a self-collected corridor-oriented RGB–thermal (RGB–T) dataset to complement public wildfire data. Unlike existing RGB–thermal wildfire datasets that mainly focus on forest or wildland fire scenes, the proposed dataset is specifically organized for complex-background power transmission-corridor monitoring, including continuous UAV sequences, nighttime conditions, smoke/vegetation occlusion, long-range small targets, and hard-negative interference. To the best of our knowledge, this is the first self-collected RGB–thermal wildfire dataset designed for this specific application scenario. The framework integrates a mobile inverted bottleneck convolution (MBConv) lightweight backbone, a Shallow Detail Fusion Module (SDFM) for shallow cross-modal alignment and denoising, a Content-Guided Attention (CGA) module for adaptive fusion, and normalized Wasserstein distance (NWD)-based box regression for long-range small-target localization. Experiments on public and self-collected datasets show that YOLO-MMSC achieves 94.6% mAP@0.5, 95.0% precision, and 93.9% recall while running at 60 FPS on Jetson Orin NX. With temporal fine-tuning, YOLO-MMSC-T reaches a continuous detection rate (CDR) of 95.6% with a jitter index of 2.8×10−3. Field experiments using a DJI Matrice 4T further indicate a practical operating altitude of 120–180 m. These results support lightweight RGB–thermal remote sensing for real-time wildfire monitoring in complex transmission-corridor environments.

Remote Sensing doi: 10.3390/rs18121868

Authors: Igor Yanovsky Nicholas LaHaye Olga V. Kalashnikova Derek J. Posselt William C. Porter

This paper evaluates a dense, total-variation-based optical flow method for retrieving wildfire smoke plume motion vectors from geostationary, deep-space, and airborne remote sensing imagery. Using multiple major fire events, we assess the robustness of the approach across a range of spatial resolutions and time intervals. The test cases include Geostationary Operational Environmental Satellite (GOES) observations of the 2025 Los Angeles Fires and the 2024 Park Fire, imagery from NASA’s Enhanced MODIS Airborne Simulator (eMAS) for the 2019 Sheridan and Williams Flats Fires, and a complementary Park Fire image pair from the Earth Polychromatic Imaging Camera (EPIC) aboard the Deep Space Climate Observatory (DSCOVR). Optical flow is computed directly on radiance fields, and smoke plumes are isolated using smoke masks derived from the Segmentation, Instance Tracking, and data Fusion Using multi-SEnsor imagery (SIT-FUSE) framework where available. Performance is evaluated by comparing the root mean square error (RMSE) between original image pairs and between the first image and the second image after warping with the retrieved motion field. RMSE is computed both globally and over smoke-only regions. Across GOES and eMAS cases, optical flow systematically reduces RMSE, often by more than a factor of two within smoke regions, indicating substantially improved frame-to-frame alignment of plume structures after motion correction. The DSCOVR/EPIC case, despite its coarser spatial resolution and longer temporal separation, also shows a marked reduction in global RMSE, demonstrating that the method remains informative under a broader range of observational conditions. For a selected subset of 10 consecutive GOES Park Fire pairs, we additionally compare the retrieved smoke motion vectors with collocated winds from the High-Resolution Rapid Refresh (HRRR) model and find the closest agreement in a broad lower-tropospheric layer centered near 875 hPa. These results show that dense optical flow can capture fine-scale plume evolution in high-temporal-resolution datasets while also providing useful motion estimates in coarser, global-view imagery. RMSE reduction is interpreted here as evidence of improved motion-compensated alignment, while the HRRR comparison provides initial physical context rather than independent validation. The resulting smoke motion vector fields provide a foundation for future comparison with model winds and for applications in plume analysis, fire hazard monitoring, and air quality studies.

Remote Sensing doi: 10.3390/rs18111866

Authors: Yuqi Li Bing Xue Houbing Chen Xiaobin Li Jingzhi Du Guoping Tang

Evapotranspiration (ET) is a key component of the terrestrial water and energy cycle, and its long-term dynamics are essential for regional hydrological assessment in subtropical China. In this study, two widely used satellite-based ET products, MOD16 and PML-V2, were selected for intercomparison because they provide consistent spatial (500 m) and temporal (8-day) resolutions. Validation against flux observations showed that PML-V2 performed better than MOD16 and was therefore used for subsequent analysis. Based on the 500 m, 8-day PML-V2 dataset, the spatiotemporal variation in ET in subtropical China during 2001–2020 was examined using the Theil–Sen slope estimator, Mann–Kendall test, and Hurst exponent. To identify the most relevant controls on ET variation, eXtreme Gradient Boosting (XGBoost) and SHapley Additive exPlanations (SHAP) were used to screen environmental factors and rank their relative importance. Multiple linear regression (MLR) was then applied only to the selected dominant factors to quantify their contributions. Residual analysis was used to distinguish climate–vegetation effects from residual influences, which could arise from human activities and unmodeled natural processes. The results showed that annual ET averaged 669 mm and increased significantly at a rate of 2.03 mm yr−1 from 2001 to 2020, with an accelerated increase after 2010. Spatially, ET exhibited clear gradients from south to north and from coastal to inland regions. Downward shortwave radiation (SWDown) and leaf area index (LAI) were the dominant drivers over most of the study area, although their controls varied geographically, with northern subregions being more energy-limited and southern subregions being jointly influenced by vegetation and temperature. Residual ET trends largely coincide with cropland and urbanising areas, indicating a partial influence of human activities, while in subregions such as XM, complex terrain and hydrological heterogeneity suggest that unmodeled natural processes may dominate. These findings enhance understanding of ET dynamics in subtropical China and demonstrate the value of high-resolution remote sensing products for regional hydrological monitoring and driver attribution.

Remote Sensing doi: 10.3390/rs18111867

Authors: Daniel Goldin Rajendra Bhatt Yolanda Shea

In this work, we discuss the impact of polarized scene radiances on the intercalibration of CPF and VIIRS reflective solar bands and the mitigation of these effects using empirical Polarization Distribution Models (ePDMs). The ePDMs, derived from multidirectional polarized reflectance measurements taken by the POLDER instrument, can provide the polarization state of the reflected solar radiation in terms of the Degree and Angle of Polarization, DOP and AOP, for each spatially, temporally, and angularly matched intercalibration footprint between CPF and VIIRS. The CPF science team will leverage these ePDMs to identify scenes with low polarization to reduce intercalibration uncertainties for specific VIIRS channels that are polarization-sensitive. The study also demonstrates that, in the absence of ePDM-based filtering of intercalibration samples, polarization-induced biases in VIIRS reflectance measurements for shortwave bands (e.g., M3 0.49 μm) can be as high as 2.4% for clear-sky over ocean scenes.

Remote Sensing doi: 10.3390/rs18111865

Authors: Bo Song Zhiyong Zhang Bo Li Zhili Chen Yun Chen Tao Yue Jianwu Jiang Zhen Cao Xing Zhang Qingyang Wang

Satellite remote sensing imagery has become an essential resource for large-scale surface water monitoring. Nevertheless, in karst regions, the elongated and fragmented morphology of water bodies, along with terrain shadows and vegetation interference, still leads to limitations in existing methods for small water body detection and accurate boundary delineation. To overcome the aforementioned issues, this paper proposes MESA-Net, a CNN–Mamba hybrid segmentation network for water body extraction in complex karst terrain. The network employs ResNet-18 as an encoder to extract shallow-level features. The decoder primarily consists of three modules: the Cross-Scale Adaptive Feature Fusion (CAFF) module, the Directional Gradient Histogram Edge-Guided Fusion (DGHEF) module, and the Omni-directional Global-Local Mamba Block (OGLMB). Among these, the CAFF module enhances the detection capability for small-scale water bodies by performing cross-scale feature fusion and dynamic weight allocation on the feature outputs from each level of the encoder. The OGLMB integrates an omnidirectional state space model with an 8-directional scanning mechanism and cross-attention guidance, effectively enhancing the ability to represent the structural continuity and global consistency of water bodies. The DGHEF utilizes directional gradient histograms to explicitly model multi-directional boundary information of water bodies, and combines this with a boundary guidance mechanism to enhance the representation of water body boundary features whilst suppressing spurious responses. In addition, the LJ-Water dataset has been constructed for the Lijiang River Basin in Guangxi, which is based on Sentinel-2 imagery. To validate the effectiveness and generalization capability of the method, comparative experiments were conducted on the self-built LJ-Water dataset as well as the publicly available Water-CD and LoveDA datasets. Experimental results demonstrate that MESA-Net consistently outperforms representative CNN-based, Transformer-based, and Mamba-based segmentation networks. On the LJ-Water dataset, it achieves 84.59% IoU and 91.65% F1, whilst on the Water-CD dataset, it attains 92.15% IoU and 95.91% F1, and 69.83% IoU and 82.24% F1 on the LoveDA dataset. Relative to the strongest baseline method, the proposed model achieved IoU gains of 1.51%, 2.34%, and 1.73% on the three datasets, respectively. In summary, MESA-Net demonstrates superior water segmentation performance under complex background conditions.

Remote Sensing doi: 10.3390/rs18111864

Authors: Xianyu Wu Hui Lin Xin Xiao

Urban flood susceptibility mapping is essential for disaster risk management in rapidly urbanizing regions. Although high-resolution Earth observation (EO) data provide detailed information for fine-scale flood analysis, existing studies are often limited by inadequate representation of drainage capacity, inappropriate spatial scales, and model uncertainty under sparse flood sample conditions. To address these issues, this study develops a multi-scale urban flood susceptibility mapping framework based on the Maximum Entropy (MaxEnt) model, integrating multi-source high-resolution geospatial data. A three-tier spatial unit system, including catchment, street, and grid scales, was constructed. Two models were developed at each scale using per capita drainage density (PCDD) and pipe density (PipeDen) as drainage capacity indicators. The results reveal significant scale-dependent differences in spatial autocorrelation, model performance, and variable responses. Compared with the PipeDen-based model, the standard deviation of AUC decreased by 37.5% and 25.0% at the catchment and street scales, respectively, and the model produced a more physically consistent relationship between drainage capacity and urban flood susceptibility. Considering the combined results of model performance, spatial autocorrelation, and response-curve analysis, the street scale PCDD-based model achieved the best overall performance among the six multi-scale models. Impervious area ratio, distance to roads, and annual maximum daily precipitation were identified as dominant factors influencing urban flood susceptibility. Based on the optimal street scale PCDD-based model, a 2 m resolution susceptibility map was generated, showing that high-susceptibility areas are mainly concentrated in highly urbanized central districts and along major transportation corridors. This study highlights the importance of spatial scale and drainage capacity representation in high-resolution urban flood susceptibility mapping.

Remote Sensing doi: 10.3390/rs18111863

Authors: Adolfo Molada-Tebar Natalia Nuño-Villanueva Alberto Morcillo-Sanz Diego González-Aguilera

Ground deformation monitoring is critical for safety and environmental management in modern mining. Active mining sites are highly exposed to terrain instabilities and subsidence, risking infrastructure integrity, disrupting operations, and posing hazards to communities. In this context, Differential Synthetic Aperture Radar Interferometry (DInSAR) techniques provide an effective and non-invasive tool capable of detecting millimetric surface displacements. This study implements the Small Baseline Subset (SBAS) technique through an open-source workflow based on the Python package hyp3_sbas, enabling semi-automated and reproducible interferometric processing by combining HyP3 with MintPy. The workflow is applied to the Björkdal gold mine (Sweden), a pilot site of the Horizon Europe XTRACT project focused on enhancing resilience in critical raw material supply chains. Integrating Sentinel-1 viewing geometries resolves the true vertical deformation field, yielding an overall mean velocity of −3.99 mm/year across the mining complex, with significant displacement rates concentrated below the 25th percentile (Q1) at −11.07 mm/year. Sector-specific analysis reveals localised subsidence accelerating over underground footprints and tailings storage facilities (mean velocities of −6.56 and −3.98 mm/year; Q1 thresholds near −13.00 mm/year), contrasting with the geomechanical stability observed at the open-pit area (mean: −0.45 mm/year). The proposed open-source framework shows strong potential for operational satellite-based monitoring, supporting predictive maintenance and early-warning strategies for risk management in mining environments while simplifying and standardising the interferometric processing workflow.

Remote Sensing doi: 10.3390/rs18111862

Authors: Yue Gong Jiaqiang Du Xiaoqian Zhu Lijuan Li Yushuo Li Xiaoshan Liu Jincao Han

Forests, as the largest terrestrial carbon sink, play a critical role in mitigating climate change. Accurately estimating forest vegetation carbon storage and identifying its drivers are essential for evaluating regional carbon sink functions and supporting carbon neutrality policies. However, long-term carbon storage estimation that simultaneously captures spatial non-stationarity and separately quantifies aboveground and belowground carbon pools at the provincial scale remains limited, and the spatial differentiation drivers and the temporal change drivers of carbon storage have rarely been disentangled through pixel-wise attribution. This study aimed to estimate forest vegetation carbon storage in Jiangxi Province, China, from 1990 to 2024, and to separately quantify the drivers of its spatial differentiation and the contributions of climate change and human activities to its temporal changes. A geographically weighted regression (GWR) model was constructed using field measurements and multi-source remote sensing data; the geographical detector and partial correlation analysis were applied for spatial differentiation attribution, and pixel-wise residual analysis was used for temporal change attribution. The results showed that: (1) total carbon storage fluctuated between 553.95 and 839.78 Tg C over the 35-year period and exhibited a significant increasing trend, with a cumulative carbon sequestration of approximately 122 Tg C; (2) the belowground carbon pool increased disproportionately (net gain 79.32 Tg C) compared with the aboveground pool (42.20 Tg C); (3) precipitation and solar radiation were the dominant drivers of the spatial differentiation of carbon storage; and (4) climate change contributed approximately 60% and human activities approximately 43% to the temporal changes in total carbon storage. These findings provide a scientific basis for delineating forest carbon sink conservation zones and formulating differentiated forest management strategies in subtropical China.

Remote Sensing doi: 10.3390/rs18111861

Authors: Alfredo Falconieri Roberto Colonna Vita Elena Di Leo Carolina Filizzola Giuseppe Mazzeo Nicola Pergola Carla Pietrapertosa Valerio Tramutoli

In this work, the portability of the Robust Satellite Techniques for FIRES detection and monitoring (RST-FIRES) has been preliminary experimented on the Spinning Enhanced Visible and InfraRed Imager (SEVIRI) aboard the Meteosat Second Generation (MSG) satellite in Rapid Scan Service (RSS) mode. Such a configuration offers 5 min of revisit time as compared with 15 min in the standard mode (0-degree). The impact in early fire detection has been assessed and quantified, also in comparison with the results of the RST-FIRES implemented on MSG/SEVIRI 0-degree data, using the official fire bulletins of the Calabria Region (Southern Italy) for the events occurred during July 2022, for which the official regional fire catalogue was available. The results obtained suggest that SEVIRI-RSS data could allow for a rather systematic earlier detection and a better sensitivity than SEVIRI 0-degree because of the improved temporal (and spatial) resolutions. These findings are remarkable in view of the next implementation of RST-FIRES on Meteosat Third Generation/Flexible Combined Imager (MTG/FCI) data, to exploit the improved spatial (2–1 km) and temporal (10–2.5 min) resolutions offered by such a new-generation geostationary mission, together with a more suitable dynamic range in the MIR spectral region (saturation at ~500 K @3.8 micron). The use of synthetic background reference fields would allow, in fact, for a straightforward RST-FIRES application to MTGI/FCI data allowing for a more effective fire early warning system.

Remote Sensing doi: 10.3390/rs18111860

Authors: Wei Zhou Yahui Guo Yongshuo H. Fu Fanghua Hao Xuan Zhang Le Xu Yuhong He

Lodging is a major agricultural hazard that can substantially reduce crop yields. Timely and accurate monitoring of winter wheat lodging is important for assessing potential yield losses, guiding field management, and mitigating further lodging damage. Recent advances in unmanned aerial vehicle (UAV) remote sensing and artificial intelligence have provided new opportunities for lodging assessment. In this study, a novel monitoring framework was proposed by integrating deep features extracted from UAV multi-spectral images with machine learning algorithms. Sensitivity analysis was conducted to identify vegetation indices (VIs), which are highly correlated with lodging. These sensitive VIs were combined with original multi-spectral bands, and YOLOv8, YOLO12, SAM1, and SAM2 were used for feature extraction. The SHAP method was applied to analyze feature importance and model interpretability. The results indicated that VARI, EXG, and MCARI were the most effective VIs for lodging monitoring. Furthermore, three feature representations, including a spectral feature set, deep features, and fused features, were evaluated. The highest accuracy was achieved using YOLO12 deep features combined with a BP classifier, reaching an accuracy of 98.20%, a precision of 98.38%, a recall of 98.56%, and an F1-score of 98.56%. Overall, incorporating deep features significantly improved monitoring performance. The proposed framework provides an accurate and effective approach for crop lodging monitoring using UAV multi-spectral imagery.

Remote Sensing doi: 10.3390/rs18111859

Authors: Wanyu Zheng Qingbiao Guo Zisu Cheng Lei Wang Sen Du Songbo Wu

Surface deformation induced by underground coal mining shows a clear time-lag effect, with persistent residual deformation in old goafs under strip pillar mining conditions, drawing significant research attention. This study focuses on the Gucheng mining area, where 210 Sentinel-1A SAR images (May 2017–January 2025) were processed using SBAS-InSAR to derive 7.5 years of time-series surface deformation. Based on these results, five strip pillar mining panels with different cessation times were selected. Through comparative analysis, a time-progressive sequence was constructed to identify post-mining residual deformation and stage-wise stabilization characteristics, and to reveal long-term deformation responses occurring years after cessation, thereby reconstructing the long-term evolution of surface deformation in old goafs. Furthermore, a stacking ensemble prediction model was developed to predict subsidence trends at representative feature points. The results indicate that subsidence mainly ranges from −20 to −10 mm/a, with a maximum of approximately −64 mm/a and cumulative subsidence of about −515 mm. Surface deformation follows a stage-wise evolution pattern of “residual subsidence—stage-wise stabilization—secondary subsidence—deformation stabilization”, with durations of approximately 2, 2, and 14 years, respectively, and overall stabilization occurring after approximately 18 years. The predicted results from the stacking model are highly consistent with the SBAS-InSAR monitoring data and can reliably describe the evolution trend of surface subsidence. The findings provide important evidence for understanding long-term surface deformation in old goafs of strip pillar mining areas.

Remote Sensing doi: 10.3390/rs18111858

Authors: Xiaotian Zhou Xin Wang Yan Tian Kai Jiang Min Guo Xuezheng Lian Lu Ding Quanyu Zhang Yaqi Xue

Recent years have witnessed the emergence of numerous U-shaped deep learning segmentation methods for infrared small-target detection (IRSTD). However, increasingly complex models still suffer from false and missed detections in challenging scenarios with cluttered backgrounds and weak targets while incurring escalating computational costs. To address these limitations, this paper proposes MCC-Net, a novel and efficient IRSTD framework that achieves superior detection performance with significantly reduced computational complexity. First, we integrate Magnitude-Aware Linear Attention (MALA) and Conditionally Parameterized Convolutions (CondConv) to replace conventional attention mechanisms in skip connections and standard convolutions, respectively, endowing the model with spatial contextual modeling and enhanced feature extraction capabilities at minimal computational overhead. Second, we design an innovative Conditional Cross-Channel Fusion (CondCCF) module that establishes a complementary spatial-channel dual-attention mechanism with MALA, enabling efficient multi-scale feature fusion. Extensive comparative and ablation experiments conducted on three public benchmarks—SIRST-v1, NUDT-SIRST, and IRSTD-1K—demonstrate that MCC-Net achieves state-of-the-art mIoU scores of 77.98%, 95.43%, and 70.46%, respectively, surpassing state-of-the-art methods by 1.07%, 1.95%, and 0.95%. MCC-Net also outperforms existing approaches across multiple evaluation metrics while maintaining substantially lower computational complexity.

Remote Sensing doi: 10.3390/rs18111855

Authors: Han Wang Haiyang Jiang Yangming Jiang Yuchen Wang Jing Zhao Liping Li Wenjiang Huang Tuo Wang

Regional image coverage retrieval is a critical problem in large-scale remote sensing data processing. However, traditional vector topology-based methods suffer from rapidly increasing computational costs when handling massive and highly overlapping datasets. This paper proposes a coverage retrieval approach based on the Discrete Global Grid System (DGGS), which transforms geometric topological operations into grid-encoding set operations, thereby reconstructing the coverage computation process under a discrete spatial indexing framework. A heuristic greedy strategy is integrated to achieve efficient coverage selection. Experimental results demonstrate that the proposed DGGS-based method achieves speedups ranging from approximately 1.5× to 11×, depending on dataset scale and coverage. Grid-level analysis indicates that level 7 grids generally provide a favorable balance between spatial accuracy and computational efficiency for near-complete coverage retrieval, whereas level 6 grids offer a more computationally efficient alternative for rapid coverage estimation and sparse-coverage scenarios, with only marginal accuracy loss. Furthermore, the method exhibits near-linear scalability with increasing data volume and maintains stable performance under incomplete coverage scenarios. The results confirm that DGGS-based discrete modeling significantly reduces computational complexity while preserving spatial reliability, providing an efficient and scalable solution for PB-scale remote sensing data processing.

Remote Sensing doi: 10.3390/rs18111857

Authors: Ke Yang Yanyan Huang Xin Lu

Accurate mapping of crop planting structures is critical for precision agriculture, yet it remains challenging in rugged terrain with fragmented fields, frequent cloud contamination, and limited high-quality training samples. This study evaluates an integrated framework combining recursive feature elimination (RFE) and the pretrained Tabular Prior-Data Fitted Network (TabPFN) for small-sample crop classification using Sentinel-2 time-series data in Yuxi City, located on the western margin of the Yunnan–Guizhou Plateau. A multidimensional feature set integrating spectral and temporal vegetation indices and textural and geospatial information was constructed and optimized via RFE. The TabPFN model achieved an overall accuracy (OA) of 96.27%, a kappa coefficient of 0.9558, and a macro-F1 score of 0.956 in the main validation. In repeated small-sample experiments, TabPFN maintained a mean OA of 90.60% at a 30% training-sample ratio and 82.89% at a 10% ratio. RF-guided feature ranking and ablation analyses suggested that temporal vegetation indices were important predictors, followed by early-season spectral characteristics, textural features, and supplementary geospatial information. Overall, these findings indicate that RFE-TabPFN is a feasible option for 10 m crop mapping in Yuxi under limited training samples, while its broader applicability still requires further testing across additional years, regions, and cropping systems.

Remote Sensing doi: 10.3390/rs18111856

Authors: Yulin Yao Mingxia Zhang Bopeng Yang Hang Zhao Qijin Han Minghui Hou

The Lutan-1 (LT-1) mission is the first civilian L-band differential interferometric synthetic aperture radar (SAR) system in China, with interferometry as its primary application. The system comprises two multi-polarimetric satellites, LT-1A and LT-1B. For the purpose of quantitative application from SAR images of Lutan-1 satellites, the relationship between the SAR image intensity and the backscattering coefficient of ground objects should be established by radiometric calibration. Field radiometric calibration provides absolute calibration constants, but it suffers from beam coverage. Internal on-board calibration, by contrast, tracks relative changes in radiometric performance but cannot yield absolute calibration constants. Therefore, we develop a method that combines on-board internal calibration with field radiometric calibration to monitor the radiometric performance of LT-1 satellites and to analyze the variation patterns revealed by both internal and field calibrations. We monitor the amplitude and phase trend of internal calibration, calculate absolute calibration constants from field calibration, and refine and evaluate the absolute calibration constants. We analyzed the internal calibration data and SAR calibration data of the LT-1 satellite from 2023 to 2025. The results show that the TRMs of the LT-1 satellite exhibit a slight decline over time, and the magnitude of the decrease in LT-1B is greater than that of LT-1A. The slight decrease in internal calibration has not yet led to visible changes in the absolute calibration constant for LT-1A, while the absolute calibration constants decrease slightly for LT-1B. After removing the calibration constant outliers and correcting the gain difference among the beams for the LT-1A satellite, absolute radiometric accuracy is improved from 0.40 dB (1σ) to 0.25 dB (1σ). The absolute radiometric accuracy of the LT-1B satellite is 0.38 dB (1σ). It gives a reference for radiometric performance monitoring of the SAR satellite over a long period.

Remote Sensing doi: 10.3390/rs18111854

Authors: Wenbin Liu Shu Li Chao Shi Hao Zhu Chao Huang Lichang Yin

Landslide mapping and susceptibility assessment are essential for hazard identification, infrastructure protection, and risk management. The middle and lower reaches of the Nujiang River have high relief, rapid geomorphic change, and fragile landscape conditions, which increase landslide susceptibility and hinder timely detection. To improve the spatiotemporal characterization of landslide activity, we developed a multi-source Earth observation framework for annual landslide mapping and susceptibility assessment. First, interannual embedding-change intensity maps were generated to guide the visual interpretation of landslide-related surface disturbances. Second, annual landslide and non-landslide samples were collected through field validation and visual interpretation. Third, annual 10 m landslide maps for 2017–2025 were generated using random forest on Google Earth Engine. Finally, 24 multidimensional environmental factors were incorporated into landslide susceptibility modeling. Landslides were concentrated mainly along the Nujiang River corridor and adjacent high-relief canyon slopes, with marked interannual variability but relatively stable hotspot regions. SHAP analysis further identified BSI_mean as the most important predictor, with a mean absolute SHAP value of 0.116, followed by NDVI_mean and terrain-related variables, indicating that bare-surface exposure, vegetation condition, and terrain dissection were strongly associated with mapped landslide occurrence. This study provides annual landslide inventories and susceptibility information for hazard mitigation and infrastructure planning.

Remote Sensing doi: 10.3390/rs18111853

Authors: Jiatao Zhang Weihua Ai Xianbin Zhao Jingjing Chen Xiong Hu

Hail is one of the most destructive warm-season severe convective hazards in China, yet the structure and microphysical evolution of hail-bearing clouds vary markedly among regions. Using GPM DPR/GMI observations together with ERA5 reanalysis during the warm seasons of 2020–2025, we identified 817 hail cloud systems across five representative hail-prone regions of China, namely Northeast China (NE), North China (NC), South China (SC), Southwest China (SW), and the Tibetan Plateau (TP), on the basis of the flagHail indicator. We then compared their macroscopic structure, vertical microphysical characteristics, organization scale, and environmental setting within a unified framework. The results reveal pronounced regional heterogeneity. Hail cloud systems in SC and SW exhibit higher echo-top heights and larger ice water paths, together with the strongest downward enhancement of reflectivity and particle size within the key ice-growth layer between 0 °C and −20 °C, indicating a deep moist-convective regime. By contrast, hail cloud systems in NC and NE more often develop into organized and horizontally extensive systems under stronger vertical wind shear, consistent with an organization-enhanced regime. Hail cloud systems over TP are characterized by high cloud tops, low hydrometeor content, and weak low-level growth, which together define a plateau-constrained regime. Environmental analyses indicate that these regional contrasts are jointly regulated by thermodynamic instability, vertical wind shear, and topographic forcing. These findings provide a physically consistent basis for satellite-based hail monitoring and region-specific hail warning over China.

Remote Sensing doi: 10.3390/rs18111852

Authors: Cong Liu Quanwei Gao Chenxi Song Bo Ouyang Ruyu Wang Hongtao Fan

Fine-grained aircraft recognition in optical remote sensing imagery remains highly challenging under cloud occlusion, as visibility degradation causes structural information loss and weakens discriminative representation learning. To address this issue, we propose DGR-MAE, a teacher–student masked image modeling framework based on posterior semantic correction for robust representation learning under incomplete observations. Unlike existing semantic-guided masking methods that modify token visibility during input construction, DGR-MAE preserves high-ratio stochastic masking in the student branch and introduces semantic correction after visibility degradation through teacher-guided differential reconstruction. Specifically, a semantic-aware teacher branch estimates patch-level importance to partition masked regions into semantic-critical and non-critical subsets, enabling region-dependent reconstruction prioritization. A collaborative feature refinement mechanism is further incorporated to enhance contextual consistency and structural reasoning during pretraining. To support controlled evaluation, we construct the ASRAir benchmark with hierarchical cloud occlusion levels. Experimental results show that DGR-MAE achieves 74.28% Top-1 accuracy on ASRAir-Occ and achieves the best Top-1 performance while maintaining competitive Top-5 accuracy compared with representative self-supervised baselines. In particular, it demonstrates substantially improved robustness under moderate-to-severe cloud occlusion, validating the effectiveness of posterior semantic correction for remote sensing representation learning under visibility degradation.

Remote Sensing doi: 10.3390/rs18111851

Authors: Stanislav Herber Tomáš Mikita Zdeněk Patočka Nikola Žižlavská

Accurate mapping of forest canopy height is fundamental to modern forestry, providing essential structural data for biomass estimation and monitoring forest health. This study evaluates the broad usability of global (25 m) and high-resolution (1 m) Canopy Height Models (CHMs) by comparing them against temporally aligned Airborne Laser Scanning (ALS) reference layers from 2018 and 2024. At the 25 m scale, we evaluated four global products: Global Forest Canopy Height (GFCH), Global Map of Tree Canopy Height (GMTCH), High-Resolution Canopy Height model of Earth (HRCH), and Europe Temporal Canopy Height (EUCH). These satellite-derived models exhibit significant height-dependent limitations, systematically underestimating mature forest canopies (>30 m) by more than 15 m due to signal saturation, though EUCH and GMTCH performed moderately better. Transitioning to 1 m high-resolution data revealed a dramatic recovery in structural fidelity. A photogrammetrically derived model (PALS) achieved an RMSE of 4.89 m and a Mean Error (ME) of 1.86 m, demonstrating remarkable vertical stability across complex topography, even on slopes >25°. While coniferous stands produced higher absolute errors (RMSE = 6.75 m) than deciduous stands (RMSE = 6.19 m) due to spire-like architectures, PALS effectively captured fine-scale canopy textures. Experimental deep learning architectures, specifically the ArcGIS Living Atlas model, showed promise with an RMSE of 8.90 m, though out-of-the-box implementations struggle without local calibration. For forest disturbance monitoring, a distinct performance trade-off emerged. High-resolution photogrammetry (PALS) provided the highest overall precision for identifying clear-cuts (F1 = 0.353) but was conservative, capturing only 51% of the reference area. In contrast, the global HRCH model captured the total spatial footprint (103.9% of area) despite its geometric inaccuracies. The Living Atlas deep learning model offered the most balanced sensitivity, detecting 118.6% of the area with a competitive F1 score of 0.326. Ultimately, digital aerial photogrammetry provides a cost-effective solution for frequent operational updates, such as the two-year national mapping cycle in the Czech Republic.

Remote Sensing doi: 10.3390/rs18111848

Authors: Tong Yan Kongwen Zhang Wuxue Cheng Jane Liu

Individual tree classification has a long history of diverse development, with recent trends focusing on the adoption of machine learning and deep learning approaches. It is a simple and powerful approach that allows the model to auto-pilot while reducing the need for physical characteristic understanding. Over more than a decade of research, we have focused on establishing a direct representation of individual trees that bridges 2D top-down imagery and true 3D models. In this study, we investigated the fundamental question of the influence of the input data on these ML/DL models. In 2024, we introduced a novel data transformation method, the Pseudo Tree Crown (PTC), which provides a pseudo-3D pixel-value perspective that enhances the informational richness of images and significantly improves classification performance. Our original implementation was successfully tested on urban and deciduous trees in 2024 and was later extended to Canadian natural conifer species under snow conditions in 2025. However, the original PTC relied on the green band, limiting its applicability to green-leaf species. In this study, we analyzed and compared the performance of different data variations and transformations, such as the Green–Red Vegetation Index (GRVI) and principal component analysis (PCA), as direct input and used their PTC forms. Classifications were conducted using Random Forest (RF), ResNet50, YOLOv10 and Segment Anything (SA). The results confirmed the effectiveness of the PTC, which consistently improves the classification accuracy by at least 5% without introducing additional computational time or complexity. Furthermore, PTC exhibits robust, consistent behavior across all data forms, demonstrating its strong resilience and reliability.

Remote Sensing doi: 10.3390/rs18111850

Authors: Nestor Erick Anibal Caal Suc Henry Antonio Pacheco Gil Martha Ruthilia Godoy Morales Víctor Manuel Lobos Morales Amado Adalberto López Bautista Carlos A. Rivas Rafael María Navarro-Cerrillo

The large-scale mobility restrictions implemented worldwide in response to the COVID-19 (SARS-CoV-2) pandemic led to short-term reductions in anthropogenic emissions, providing an opportunity to explore atmospheric pollutant responses to large-scale changes in human activity and mobility patterns. Although numerous studies have reported air quality improvements during lockdowns, most rely on ground-based monitoring networks and focus on developed regions, leaving gaps in less-studied areas such as Central America. This study evaluates spatiotemporal changes in tropospheric nitrogen dioxide (NO2) across Central America before, during, and after COVID-19 lockdowns using satellite-based remote sensing. High-resolution NO2 vertical column density (VCD) data from the TROPOMI instrument onboard Sentinel-5P were processed using Google Earth Engine. Percentage variations were calculated using the March–May 2020 lockdown period as a reference within the 2019–2021 analysis period. Results indicate reductions in NO2 across several high-density departments, particularly in Guatemala, El Salvador, and Honduras, with decreases of 20–30% and localized negative variations below −40%. In contrast, Nicaragua exhibited comparatively limited changes, while a gradual recovery in NO2 concentrations was observed during 2021. The observed patterns suggest a potential association between NO2 variability and changes in anthropogenic activity during the COVID-19 period, while also highlighting the importance of considering meteorological influences in regional atmospheric assessments. The results further demonstrate the potential of cloud-based Earth observation platforms for atmospheric monitoring in data-scarce tropical regions.

Remote Sensing doi: 10.3390/rs18111849

Authors: Juehui Xing Siyuan Yao Zhongyi Zhu Lingxin Zhang

Automatic building extraction plays an important role in various remote sensing applications, such as seismic disaster investigation, seismic hazard risk assessment, urban planning, and photogrammetry. Despite the substantial progress, state-of-the-art building extraction methods are still limited by two issues: (i) existing approaches leverage convolutional layers or non-local self-attention to encode the position-aware dependencies, while they cannot flexibly adapt to the complex background contexts and varied structure patterns of buildings; and (ii) the local details cannot be well preserved by existing hierarchical decoders due to the imperfect feature aggregation, yielding unsatisfactory segmentation outputs in the local adjacent region. To address these issues, we propose Relative Position Aggregated Transformer (RPAFormer), which is capable of modeling the relative position dependencies of buildings and producing accurate local details using a dual attention transformer network. Specifically, we propose a Relative Position-aware Self-attention (RPSA) framework to learn the token dependencies within the local window. A transformer decoder network consisting of multiple Cross Masked Attention (CMA) blocks is also introduced to fuse the multi-scale features. Extensive experiments demonstrate the superior performance of the proposed method and its great promise for real-world engineering deployment.

Remote Sensing doi: 10.3390/rs18111847

Authors: Junfang Jiang Wanjin Wang Xiaohui Lin Pingping Miao Lina Gao Mingzhu Xu

In recent years, salient object detection in optical remote sensing images (ORSI-SOD) has garnered increasing research attention. However, in practical applications, issues such as blurred target edges under low-contrast and complex background interference continue to restrict the accuracy and robustness of detection. To address these problems, this paper proposes the Phase Congruency-Guided Cross-Scale Contextual Fusion Network (PCFNet). Specifically, we design a novel Phase Congruency Enhanced Module (PCE) to solve the problem of low-contrast between targets and backgrounds. It acquire phase features via Fourier decomposition and employs them to generate a weighting map to modulate the shallow features via element-wise multiplication, thereby highlighting structurally significant regions. Meanwhile, we adopt a tailored loss weighting mechanism to weight phase congruency learning for better PCE adaptation. To address complex background interference, we design a novel Dynamic Residual Fusion (DRF) Module. It leverages dynamic spatial attention to generate sample-specific kernels that perform convolution to spatially weight features and uses consecutive residual connection, thereby refining multi-scale features to accurately capture effective targets under complex background interference. Experiments on ORSSD, EORSSD, and ORSI4199 benchmarks demonstrate that PCFNet achieves nine best performances and three second-best performances across the twelve core evaluation metrics, outperforming 23 state-of-the-art methods. Notably, the Fβ score is 1.16% higher than HFCNet on ORSSD and 0.85% higher than MCPNet on EORSSD.

Remote Sensing doi: 10.3390/rs18111843

Authors: Hao-Liang Li Xiu-Jun Dong Qiang Xu Ou Ou Yi-Shan Li Jie Liu Jing-Song Sima

In alpine mountain regions, significant seasonal surface changes reduce InSAR coherence over long time spans, hindering geohazard identification. This study proposes a method for geohazard detection based on InSAR seasonal coherence variation. First, time-series interferograms and coherence maps are generated from Sentinel-1 imagery. Each year is then partitioned into summer, transition, and winter seasons by analyzing the spatial migration of high-coherence zones. Interferometric pairs from the transition season are further screened and reassigned to summer or winter groups according to their coherence characteristics. Stacking-InSAR is applied separately to the summer and winter datasets to derive seasonal deformation rates; long-temporal-baseline pairs (60–120 days) that maintain sufficient coherence are selectively incorporated to improve the detectability of slow-moving slopes. Finally, geohazards are identified by combining the summer and winter deformation results. Applied in the eastern Himalayan syntaxis, the method showed that less than 19% of geohazards were detectable in both seasons, indicating seasonal variations in geohazard activity. Moreover, it identified approximately 29% more geohazards on average than traditional Stacking-InSAR using all interferograms. Thus, the proposed approach enables more accurate and effective geohazard detection in cold mountains, supporting disaster prevention and mitigation.

Remote Sensing doi: 10.3390/rs18111846

Authors: Jingjing Wang Rongbo Sun Haowei Duan Hao Chen Gang Yu Huaqiang Xu

Compressed sensing (CS) and matrix completion algorithms (MCA) have each introduced sparse and low-rank priors into synthetic aperture radar (SAR) imaging. However, their combined use reveals a fundamental zero-sum trade-off: enhancing spatial continuity tends to obscure weak targets, while strengthening sparse recovery amplifies off-grid artifacts. This inherent conflict is further exacerbated by static regularization, which imposes a rigid global compromise and prevents genuine synergy between the two priors. To overcome this limitation, this paper proposes a Spatio-Temporal Adaptive Synergistic Constraint Imaging (STASCI) algorithm, which dynamically balances the two priors in a scene-aware manner. The core of STASCI is a unified regularization framework. The low-rank constraint models’ spatial continuity in the background to suppress off-grid artifacts. The sparse constraint, enhanced by a non-convex Geman-McClure function, is employed to detect weak targets and compensate for detail loss. A key innovation is a spatio-temporal dual-dimensional regularization mechanism that employs Sobel operators to probe local spatial gradients and dynamically adjusts the strength of each prior according to regional scene characteristics. This enables adaptive synergy rather than a fixed trade-off. The optimization is solved via the alternating direction method of multipliers (ADMM), with the low-rank subproblem accelerated by randomized singular value decomposition (RSVD). Final imaging is performed using the Range Migration Algorithm (RMA). Experiments on real measurements and public datasets demonstrate that STASCI breaks the conventional detail-background trade-off. It effectively suppresses off-grid artifacts while retaining weak targets, leading to significant improvements in imaging accuracy and robustness across complex scenarios.

Remote Sensing doi: 10.3390/rs18111845

Authors: Akito Davis Kawamura Tomoya Kodama Takeo Tadono

Continuous monitoring of forest biomass is indispensable for establishing transparent carbon budgets and ensuring sustainable forest management toward achieving carbon neutrality. While satellite data has traditionally been used for wide-area biomass estimation, signal saturation in high-biomass regions has posed a significant challenge to accuracy. To address this saturation issue and enhance the precision of carbon budget estimations, this study develops a new methodology for estimating forest above-ground biomass (AGB). First, a training dataset was constructed by integrating airborne LiDAR data from across Japan with various satellite datasets, such as PALSAR-2 and Sentinel-2. Machine learning (XGBoost) was then employed to generate a nationwide canopy height map, achieving a high coefficient of determination (R2=0.594). Subsequently, allometric equations with parameters optimized for specific forest types (evergreen coniferous, evergreen broadleaf, deciduous coniferous, and deciduous broadleaf) were derived from the relationship between estimated canopy height and AGB to create a nationwide AGB map. Validation results indicated that the resulting AGB map demonstrated higher estimation accuracy (R2=0.265) compared to existing global products (ESA CCI Biomass), with significant improvements in mitigating underestimation (saturation) in high-biomass areas. By combining canopy height estimation with forest-type-specific allometry, this approach enables high-precision mapping that reflects the unique characteristics of Japanese forests and is expected to contribute to more reliable carbon budget assessments.

Remote Sensing doi: 10.3390/rs18111841

Authors: Zihan Xu Jiabao Wang Dongpan Chen Tianjie Lei Wei Su Weihua Xiao Yinlong Xu

Drought, one of the most severe natural disasters globally, has inflicted notable impacts on animal husbandry production, yet the current research on drought impact assessment in pastoral systems is plagued by obvious gaps, such as the lack of comprehensive quantitative evaluations integrating grassland ecosystem and livestock production indicators, unclear quantitative relationships between drought severity gradients and multi-level pastoral impacts, and the absence of validated quantitative assessment frameworks linking drought indices with actual pastoral economic losses. To fill these gaps, this study takes Inner Mongolia grasslands as the research area, analyzes the spatiotemporal characteristics of drought and its impacts on grassland net primary productivity (NPP) over the 50-year period from 1961 to 2012, and quantifies the differential impacts of three representative gradient drought events (1974 moderate, 1986 severe, and 1965 extreme) on grassland NPP, standard hay yield, sheep units and livestock economic losses. The long-term analysis shows that drought frequency in the study area decreases with increasing severity, with the typical steppe having the highest drought frequency and a “nine droughts in ten years” pattern in the central and western regions; drought intensity increases westward, and duration extends with rising severity, and its spatial distribution is highly consistent with the east–west precipitation gradient. Drought is the dominant driver of NPP variation, explaining up to 84% of NPP anomalies, with meadow steppe being the most sensitive to drought and desert steppe showing stronger drought resilience due to adaptive traits such as deeper root systems. The assessment of the three representative drought events reveals that drought impacts exhibit a linear amplification effect with severity, with extreme drought causing an average NPP loss 2.8 times greater, hay yield loss 1.1 times greater, and economic loss 4.4 times greater than those caused by moderate drought, and different grassland types show distinct response characteristics to drought of varying severity. The NPP loss spatial distribution is highly consistent with severe drought areas, and sheep unit loss is directly correlated with drought severity. Most importantly, the study validates a robust quantitative assessment framework (SPI→NPP→hay yield→sheep units→economic loss) with relative errors of less than 9% compared with historical disaster records, which systematically links drought indices with practical pastoral economic losses. This research clarifies the quantitative relationships between drought and multi-dimensional pastoral impacts, and provides actionable scientific insights for drought risk governance in arid and semi-arid pastoral areas such as Inner Mongolia.

Remote Sensing doi: 10.3390/rs18111844

Authors: Jilin Zheng Fanle Wan Yanlong Cai Junshuai Liu Dake Wang Xiaoyu Guo Bowei Chen

Gully erosion is a major driver of irreversible soil loss in Northeast China’s Mollisol belt, a region that supplies roughly one-quarter of the national grain output. Existing susceptibility assessments in this region have rarely combined multi-model comparison with spatially explicit cross-validation, and the predictive contribution of composite anthropogenic indicators such as the Human Footprint Index (HFI) has not been quantitatively benchmarked against conventional topographic variables. This study addresses these gaps for the Heihe region by combining an inventory of 4020 gully polygons supported by field checks in Xunke County, 16 VIF-screened environmental factors, three tree-based ensemble models and a logistic regression baseline. Under stratified random splitting, XGBoost achieved the highest discrimination (AUC = 0.95, κ = 0.74); under leave-one-district-out spatial cross-validation all tree-based models retained AUC above 0.83, confirming that random-split metrics overestimate discrimination by approximately 0.11 AUC units due to spatial autocorrelation and inter-district covariate shift. SHAP analysis identified LULC and HFI as the dominant predictors, exceeding all topographic variables, while slope gradient contributed least—consistent with the low-relief, intensively cultivated character of the study area. Susceptibility was highest in the southwestern agricultural lowlands. A one-factor sensitivity test in which only NDVI was increased by 20% suggested a reduction in modelled high-susceptibility area of approximately 12%, although co-occurring land-cover and hydrological changes were not simulated. The multi-model framework, integrating spatial cross-validation and post hoc interpretability, provides an explicit estimate of conventional evaluation optimism and supports spatially differentiated erosion management.

Remote Sensing doi: 10.3390/rs18111842

Authors: Chaoyang Tian Shijie Liu Hengxing Lan Langping Li

The Eastern Himalayan Syntaxis in the southeastern margin of the Tibetan Plateau is a tectonically active, deeply incised, high-relief region with frequent landslides. However, the long-term evolution of landslide susceptibility and the temporal behavior of its dominant conditioning factors remain insufficiently understood. This study compiled a 30-year inventory of 1350 landslides from multi-source remote-sensing data and divided it into three periods: P1 (1991–2000), P2 (2001–2010), and P3 (2011–2020). Period-specific random forest models were developed for susceptibility mapping, and Tree-SHAP was used to interpret temporal changes in dominant factors and their nonlinear responses. The models showed reliable performance, with AUC values of 0.887, 0.848, and 0.900, respectively. Susceptibility patterns showed broad temporal stability with localized reorganization, with unchanged areas accounting for 55.62%, 51.62%, and 58.51% of the P1–P2, P2–P3, and P1–P3 transitions, respectively. High and very high susceptibility zones were persistently concentrated along the Yarlung Tsangpo–Parlung Tsangpo–Yigong Tsangpo river system and major tributary junctions. SHAP results identified elevation, slope gradient, terrain curvature, NDVI, and annual precipitation as the persistent core factor group, whereas drainage proximity, the seismic disturbance proxy, and road proximity showed stronger period-dependent effects. Nonlinear SHAP responses revealed threshold-saturation, overall decreasing or distance-decay, threshold-transition, and inverted U-shaped patterns. These findings indicate that susceptibility evolution reflects the coupling between persistent geomorphic predisposition and stage-dependent environmental and disturbance-related modifiers, providing a basis for identifying persistent and stage-specific high-susceptibility zones in high-relief valley regions.

Remote Sensing doi: 10.3390/rs18111840

Authors: Maoping Zhang Pu Wang Jiajie He Shilin Zhou

High-precision cloud detection is essential for remote sensing applications such as agricultural monitoring and disaster response. However, thin clouds severely limit detection accuracy. The difficulty lies in their semi-transparent nature, which causes their reflected signals to couple with the reflectance of various underlying surfaces. This coupling leads to inconsistent cloud signatures and significant intra-class variability. To address this, we propose a Class Center Residual Attention Network (CCRANet), a radiative transfer theory-inspired framework that employs a class center approach to extract the intrinsic reflective characteristics of thin clouds. Specifically, the core of the network is the Class Center Attention (CCA) module, which extracts invariant intrinsic features of thin clouds, supplemented by the Class Center Residual (CCR) module to eliminate surface-induced interference. Experiments on three public datasets (Landsat-8, CSWV, and CloudS26) show that CCRANet achieves a mean Intersection over Union (mIoU) of 85.93% on the Landsat-8 dataset, outperforming the classic DeeplabV3+ baseline by 10.23 percentage points. In particular, it achieves 22.58 percentage point improvement in thin cloud IoU over DeeplabV3+ in snow/ice scenarios, significantly reducing false positive detections caused by surface spectral similarity.

Remote Sensing doi: 10.3390/rs18111839

Authors: Fei Ma Kangjie Yu Jianmei Zhang Jinran Zhang Wei Lian Qingbin Zhang Zhixing Zhao Haijun Zhang

Accurate monitoring of mining-induced three-dimensional surface deformation is critical for safety and environmental protection. Conventional InSAR often loses coherence in high-deformation areas and provides only one-dimensional measurements, while the Probability Integral Model (PIM) suffers from low accuracy at subsidence edges, caused by premature numerical convergence of its error-function-based mathematical formulation—the model prediction rapidly drops to zero and fails to capture subtle real-world deformations in marginal zones. This study developed a fusion method integrating multi-source InSAR (Sentinel-1A and SAOCOM), PIM, and the Covariance Matrix Adaptation Evolution Strategy (CMA-ES). Applied in the Yinying Mining Area, Shanxi Province, the approach combined ascending and descending SAR data processed via SBAS-InSAR, used CMA-ES to optimize PIM parameter inversion, and employed a zonal fusion strategy to reconstruct complete deformation fields. The method demonstrated substantial improvement in monitoring accuracy, with mean absolute errors in the vertical, north–south, and east–west directions reduced by more than 86% compared with the standalone PIM model in edge zones. The fusion approach effectively captured both large-magnitude center deformations and subtle edge displacements. Multi-source data fusion with intelligent optimization algorithms significantly enhances the accuracy of 3D deformation monitoring in mining areas, providing reliable technical support for safety management and environmental protection.

Remote Sensing doi: 10.3390/rs18111838

Authors: Yihao Chen Donghai Wang Xueting Zhang Enguang Li Lebao Yao Yangjinxi Ge Yuting Xue Rui Xie

The aim of this work was to reveal the differences in the macro- and microphysical characteristics and precipitation mechanisms of tropical precipitation systems (TPSs) in different regions. Based on the GPM satellite observation from 2014 to 2022, global TPSs were identified, and eight high-frequency areas were defined. Subsequently, their horizontal and vertical development, precipitation characteristics, and microphysical vertical structure were systematically analyzed. The results show that the horizontal development scale of TPSs is mostly between 104 and 105 km2, with vertical development exceeding 10 km. The convective area fraction (CAF) ranges from 20% to 60%, and TPSs have a higher CAF and lower vertical development over the ocean than over land. Continental TPSs exhibit significantly stronger vertical development and more intense precipitation in convective cores than oceanic TPSs. The stronger vertical development over land is mainly attributed to stronger updrafts associated with topographic lifting, which further enhances ice-phase microphysical processes and increases ice particle size. Meanwhile, the intensified updrafts also lead to higher collision–coalescence efficiency in the liquid layer, and temperature perturbations over land further enhance turbulent collision efficiency. Together, these processes result in stronger precipitation intensity in the convective cores of continental TPSs. Stratiform regions are characterized by weak precipitation dominated by raindrop breakup with small regional differences. These findings clarify the key land–ocean disparities in TPSs and provide critical observational evidence for optimizing cloud microphysical parameterization schemes in numerical models.

Remote Sensing doi: 10.3390/rs18111837

Authors: Wenqi Liu Yipeng Zhang Weijiang Yu

Submarine groundwater discharge (SGD) is an important pathway for water and solute exchange between coastal aquifers and the ocean, but its spatial detection remains challenging because field methods have limited coverage and remotely sensed anomalies may also reflect other coastal processes. This study developed a hydrodynamically constrained remote sensing framework for SGD identification by integrating optical and thermal indicators with hydrogeological constraints. Sentinel-2 imagery was used to derive the Normalized Difference Chlorophyll Index (NDCI) and Normalized Difference Turbidity Index (NDTI), while Landsat thermal data were used to quantify seasonal sea surface temperature variability using the 90th–10th percentile amplitude. These indicators were combined in a K-means clustering framework, and the classification results were further constrained using year-specific maximum offshore distances estimated from groundwater level observations with a Dupuit–Glover-based scaling approach and hydraulic time-lag correction. Applied to the north shore of Long Island, New York, the framework identified coherent nearshore SGD patches that were broadly consistent with field observation locations and showed both temporally persistent discharge zones and interannual variability in spatial extent. These results indicate that incorporating physically based constraints can improve the robustness and interpretability of remote sensing-based SGD detection.

Remote Sensing doi: 10.3390/rs18111836

Authors: Benjamin Britton Alec McLellan Nicholas Dunning

The detection of small archaeological artifacts in high-resolution aerial imagery is challenged by minimal target size and local spectral and geometric similarity to background soils. This study identifies a failure mode in end-to-end deep learning where radiometrically dominant chromatic signals destabilize gradient-based optimization, leading to rapid training collapse. Using UAV imagery of Maya archaeological sites in Belize, we examine fingernail-sized ceramic sherds characterized by a consistent reddish hue. A Hue-Weighted Loss Function (HWLF) is introduced as a diagnostic instrument. Under severe class imbalance, chromatic gradients suppress geometric feature learning, collapsing detection within 300 iterations. Motivated by this discovery, we propose a staged detection architecture that decouples geometric candidate generation from chromatic validation. Candidates are detected via a transformer-based object detector and validated using hue constraints derived from unmodified 16-bit HSV representations. This approach reduced the Phase I candidate pool (177,148 geometric detections) to 1647 prioritized detections—a 99.1% reduction—while retaining 97.8% of annotated targets (F1 = 0.731). Chromatic priors may be more effective as decoupled post-inference discriminants than as embedded end-to-end optimization signals under severe class imbalance, where their gradient influence risks suppressing geometric feature learning entirely.

Remote Sensing doi: 10.3390/rs18111835

Authors: Hanyuan Dong Ying Yu Xiguang Yang Guanran Wang Xuebing Guan Hang Xu

LAI is a critical parameter for forest management and global ecosystem monitoring. GEDI provides global-scale vegetation structure data, yet its L2B LAI product often exhibits systematic biases. This study investigates the Maoer Mountain forest in China, utilizing a total of 60 validated GEDI footprints as the primary dataset. To address the limitations of the standard GEDI L2B algorithm, which assumes a horizontally uniform canopy, we integrated a four-scale geometric optical model to characterize canopy clumping effects. This model was employed to simulate the geometric proportions of sunlit/shaded canopy and ground components within each footprint to derive a footprint-specific clumping index, thereby refining the gap rate estimates. The accuracy of the revised leaf area index was rigorously verified by using the measured data from the sample plots in the Maoer Mountain area. The results indicate that the original GEDI L2B data underestimates LAI, with a mean absolute error (MAE) of 1.79 m2/m2, a root mean square error (RMSE) of 1.47 m2/m2, and a bias of −1.25 m2/m2. After correcting for canopy clumping, accuracy improved significantly, reducing the MAE to 0.65 m2/m2 and the RMSE to 0.82 m2/m2, while effectively mitigating underestimation. These findings demonstrate that accounting for non-uniform canopy distribution effectively reduces errors, providing a robust methodological basis for high-precision LAI retrieval using spaceborne lidar. Despite these improvements, this method still has certain limitations: the model’s performance is constrained in extremely steep terrain due to waveform aliasing and in fragmented vegetation areas where sub-footprint heterogeneity is high. Future research should incorporate topographic corrections and multi-source data fusion to enhance the model’s robustness in complex landscapes.

Remote Sensing doi: 10.3390/rs18111834

Authors: Wei Shi Zhiwei Li Siyuan Chen Ning Wang Ronghong Cheng Tonghe Yang

In the reconstruction problem of compressed sensing, to address the challenge of adapting common sparse constraints to diverse data, we propose a data-driven inverse-weighted regularization for adaptive data matching to enhance the ability of sparse constraints. Specifically, we formulate a weighted regularization term based on the data transform domain, positing that higher values in the sparsity-promoting transform domain correspond to a greater probability of effective signals. Therefore, when solving sparse optimization problems, we inversely weight this portion based on the inverse relationship with the coefficient magnitude, thereby reducing its impact and mitigating damage to effective signals. However, recognizing that noise and other irrelevant signals are sparse and approximately uniformly distributed in the transform domain, we can increase the weight of this portion to boost the sparsity constraint in the transform domain, thereby enhancing noise suppression. Consequently, we presented the corresponding solution algorithm and convergence proof for inverse-weighted sparse regularization, along with an application example in the context of the Radon transform. Experimental data tests indicate that inverse-weighted sparse regularization enhances the capability of sparse constraints, protects effective signals, suppresses noise, and improves the recovery accuracy of compressive sensing algorithms, as demonstrated in natural image enhancement and seismic multiple suppression.

Remote Sensing doi: 10.3390/rs18111833

Authors: Yamei Shao Nan Wang Lijun Zhao Guohui Yao Yicong Chen Weilun Li Hao Wang Haidong Li

Revegetation serves as a critical ecological safeguard, while these interventions have added complexity to the evapotranspiration processes and water balance. Dalinor Lake basin (DLB), located in the southeast of Inner Mongolia Plateau, serves as a vital habitat for migratory birds and plays an important role in the ecological security of northern China. To enhance biodiversity, numerous ecological restoration projects have been carried out in this area in recent years. Dalinor Lake, a large inland lake within the basin, has experienced persistent shrinkage. Although existing studies have explored its driving factors, the potential influence of revegetation activities on lake shrinkage remains unclear. In this study, we used remote sensing imagery, combined with supervised classification and visual interpretation methods, to extract changes in the surface areas of lakes within the DLB (i.e., Dalinor Lake and Ganggeng Lake), and analyzed the effects of total terrestrial evapotranspiration (ETt), precipitation (PPT), runoff, soil moisture content, and the vapor pressure deficit on these changes. Results showed that the Dalinor Lake’s area decreased by 18.68% from 2000 to 2020, and was mainly influenced by ETt, with the Normalized Difference Vegetation Index (NDVI) contributing the most to ETt (54.02%). In contrast, Ganggeng Lake expanded by 5.68% and was strongly driven by PPT. Compared with Ganggeng Lake, there have been more revegetation activities around Dalinor Lake, resulting in significant increases in NDVI and ETt, together with widespread declines in soil moisture in its surrounding areas, suggesting that revegetation exerted non-negligible water pressure on Dalinor Lake. These findings can provide valuable information for policymakers to balance large-scale ecological restoration with sustainable water management in semi-arid regions.

Remote Sensing doi: 10.3390/rs18111832

Authors: Tianci Xie Jinquan Ai Ni Xie Man Qiao

Vegetation phenology refers to the cyclical growth patterns of vegetation in nature, which are influenced by climatic conditions, human activities, and genetic factors. It plays an irreplaceable role in regulating carbon cycling and energy flow within natural ecosystems. However, the combination of a cloudy and rainy climate with a landscape characterized by the interplay of land and water and fragmented patches has long posed challenges for remote sensing phenological monitoring data, including a scarcity of valid observations, frequent temporal gaps, and spectral distortion in mixed pixels. These issues make it difficult to reliably support the needs of wetland phenological inversion and mapping. To address this issue, this study uses vegetation inversion in the Poyang Lake wetlands as a case study and reconstructs high-spatiotemporal-resolution time-series kNDVI data based on multi-source remote sensing data. Methodologically, we propose an improved and enhanced spatiotemporal adaptive reflectance fusion model, IESTARFM. This model enhances the homogeneity of similar pixel selection through adaptive matching windows and land cover constraints. Additionally, it explicitly incorporates cloud probability and time-lag factors into the weighting structure to systematically downweight unreliable observations, and further employs quadratic term corrections to account for the nonlinear growth response of kNDVI. Using the reconstructed dataset, key phenological information is extracted by combining third-order harmonic analysis with a dynamic thresholding method, thereby enhancing the robust characterization of seasonal trajectories under conditions of missing data and noise. Accuracy evaluation results show that the 10m/8d high-frequency kNDVI dataset reconstructed by IESTARFM achieves at least a 12.61% improvement in fusion accuracy compared to classical methods such as ESTARFM, STARFM, and FSDAF, with a maximum reduction in RMSE of 0.026, and effectively restores details in areas with thin cloud cover. The reconstructed kNDVI series achieved a coefficient of determination R2 = 0.875 and RMSE = 0.066 relative to Sentinel-2 observations, indicating that the reconstructed series closely reproduces the reference imagery in both amplitude and spatial structure. The phenological parameters derived from kNDVI exhibit an RMSE of 4.81 days compared to field observations, demonstrating that the reconstructed time series reliably captures the timing of key phenological events. It should be noted that the proposed approach is designed for post-event time-series reconstruction and is not intended for real-time forecasting. In summary, this study collaboratively enhanced the reliability of high-resolution index time-series reconstruction and phenological identification in cloudy and rainy wetlands through three key aspects: cloud noise suppression, heterogeneous boundary preservation, and nonlinear growth characterization. It provides a generalizable technical foundation for dynamic monitoring of wetland vegetation, ecological restoration assessment, and refined management in regions with frequent cloud and rainfall.

Remote Sensing doi: 10.3390/rs18111831

Authors: Yunmin Zhang Youqiang Zhang Boshan Shi Bisheng Wang Qiqiong Yu Haitao Zhao

Existing noisy-label hyperspectral image classification (HSIC) methods usually address clean sample selection and representation regularization as separate problems, although the reliability of observed labels varies substantially across samples in hyperspectral data. This issue is amplified by mixed pixels, boundary ambiguity, spectral overlap, and limited labeled samples, which make hard clean samples difficult to distinguish from mislabeled ones. We therefore propose PDAM, a sample-reliability-guided training framework for noisy-label HSIC. The method first estimates feature-space class consistency by comparing each sample with the prototype of its observed class and converting this consistency into a reliability probability with a Gaussian mixture model. To reduce conservative false negatives, matched high-confidence selection is further used to recover hard but correctly labeled samples. The resulting reliability estimate then determines how strongly the observed label is trusted through target refinement and how strongly the input is perturbed through reliability-guided masking. Finally, masked reconstruction provides label-independent structural regularization so that uncertain samples can still contribute to spectral–spatial representation learning. Under the evaluated synthetic symmetric noise settings on the University of Pavia (UP), Salinas Valley (SV), and Kennedy Space Center (KSC) datasets, PDAM achieves the best OA and Kappa in most reported comparisons and improves robustness under both moderate and severe noise. At 30% noise, PDAM reaches 97.30% OA on UP, 98.13% OA on SV, and 95.37% OA on KSC. Ablation studies further support the necessity of reliability estimation, hard clean sample recovery, and reliability-guided supervision and regularization within this unified training mechanism.

Remote Sensing doi: 10.3390/rs18111830

Authors: Shangmin Zhao Xiangyu Li

The Northern Tibetan Plateau is among the most climate-sensitive alpine regions globally. To address the limited applicability of the traditional Remote Sensing Ecological Index (RSEI) in sparsely vegetated areas, this study developed a Soil-Adjusted Remote Sensing Ecological Index (SRSEI) tailored to cold and arid environments. The ecological quality of the Northern Tibetan Plateau from 2000 to 2025 was systematically evaluated and analyzed. The results indicate that: (1) The improved SRSEI achieved a first principal component (PC1) contribution of 72.76%, a significant enhancement over traditional models that effectively mitigates noise from soil backgrounds and anthropogenic features. (2) Between 2000 and 2025, ecological quality was predominantly moderate, following a characterized east-to-west declining spatial gradient. Overall mean SRSEI values fluctuated between 0.420 and 0.476, exhibiting a marginal downward trend. (3) Ecological degradation affected 50.17% of the region, with 26.14% facing risks of sustained decline. Conversely, 40.11% of the area displayed potential recovery trends, suggesting potential spatial divergence in future ecological trajectories. (4) Regional ecological dynamics are governed by a topographic-thermal compound driving mechanism. Elevation (DEM), temperature (TEMP), and surface shortwave radiation (SRAD) emerged as the dominant explanatory variables. Furthermore, dual-factor interactions exhibited significant enhancement effects, while the influence of anthropogenic factors was comparatively weak at the regional scale. These findings provide a scientific basis for the long-term monitoring of fragile alpine ecosystems and the strategic development of the Qiangtang National Park.

Remote Sensing doi: 10.3390/rs18111828

Authors: Wei Wang Shiqiang Liu Huijin Yang Ning Li Jianhui Zhao Wenfu Wu Wenkui Zheng

Rice is one of the world’s most important food crops, feeding over half of the global population and being crucial for food security. Accurate, timely mapping of rice fields is essential for precision agriculture, yet conventional methods relying on static samples fail to capture dynamic farmers’ planting decisions. To address this, we propose the Multi-Source Dynamic Sample Generation and Phenology-Guided Feature Selection Framework for In-Season Rice Identification (MSDF-RiceID) using multi-source remote sensing imagery. It incorporates two key innovations: (i) a rule-based sample updating mechanism based on historical rice maps and a dynamic threshold algorithm, and (ii) phenology-guided feature optimization through exponential weighting. Developed specifically to handle complex cropping patterns and high cloud cover in Hunan Province, MSDF-RiceID integrates these innovations within a grid-search-optimized Random Forest classifier to produce reliable monthly rice distribution maps. In-season samples corresponding to transplanting dates in April (DOY 100, 120), June (DOY 160), and July (DOY 184), differentiated as early-, middle-, and late-rice crops. The optimal feature set combined Sentinel-1 (PRI, VH, VH_VV), Sentinel-2 (NDYI, PSRI, NDBI, NDWI), and MODIS (NDVI, EVI, NDBI, LSWI) indices. Accuracy increased seasonally, with F1-score rising from 0.82 in May to 0.97 at harvest. Cross-region validation in Taishan (Guangdong) and Panjin (Liaoning) showed that the earliest identifiable stage (F1-score > 0.9) occurred earlier than in Hunan due to Hunan’s more complex triple-cropping phenology, highlighting the model’s strong transferability. Furthermore, MSDF-RiceID outperformed existing products (TWDTW-Rice and EARice10), increasing overall accuracy by 0.12–0.18, Kappa by 0.23–0.35, and F1-score by 0.09–0.15. These results demonstrate its effectiveness for in-season, large-scale, and dynamic rice mapping under persistent cloud cover, thereby providing direct support for precision agricultural management in heterogeneous cropping systems.

Remote Sensing doi: 10.3390/rs18111829

Authors: Guixin Zhang Jingpeng Liang Shanyou Zhu Yongming Xu

Accurate, high-resolution gridded air temperature forecasts are essential, particularly in mountainous areas with complex terrain. This study proposes a two-stage processing framework DOWN + BC to downscale Global Forecast System (GFS) temperature forecasts and correct their bias. The approach first employs a random forest (RF) model to geographically downscale 3-hourly 0.25° GFS forecasts to a 30 m resolution (DOWN), followed by bias correction (BC) using a first-order adaptive Kalman filter (AKF). The accuracy of the DOWN + BC-processed forecasts was evaluated against both automatic weather station (AWS) observations and high-resolution air temperature fields derived from an extreme gradient boosting model (XGB-derived). The results indicate that (1) the DOWN step effectively refines the spatial detail of temperature distribution, though it yields limited improvement in accuracy compared to the raw GFS forecasts; (2) the combined DOWN + BC method substantially enhances forecast accuracy. At AWS locations, the root mean square error (RMSE) of GFS forecasts decreased by 37.84% in January 2020 and 41.16% in July 2023. Relative to the XGB-derived temperature distribution, RMSE was reduced by 47.27% and 33.79% for the respective periods.

Remote Sensing doi: 10.3390/rs18111827

Authors: Shijun Sun Shuai Ma Xuyang Feng Chen Sun Baolong Ding Yaoyao Ran Yihong Zhang

In the field of remote sensing applications, multimodal object detection has emerged as an important technique for enhancing perception robustness in UAV-based scenarios. Nevertheless, RGB–IR UAV detection remains difficult: Degraded illumination destabilizes shallow representations and weakens local discriminative cues, while spatial inconsistencies and fluctuating modality reliability further hinder cross-modal interaction. In addition, existing methods, which often depend on global illumination estimation or simplistic fusion schemes, struggle to jointly maintain contextual stability, reliable cross-modal interaction, and compact discriminative representations in complex aerial scenes. To address these issues, this paper proposes LDSDet, an RGB–IR multimodal UAV object detector for challenging illumination conditions. Specifically, LDSDet integrates three complementary modules: a Long-range Aware Residual Convolution (LARC) module that enhances contextual perception and stabilizes shallow features; a Dynamic Attention-based Cross-modal Fusion (DACF) block that performs spatially adaptive RGB–IR interaction; and a lightweight SeqShuffleGate (SSG) module that suppresses redundant fusion responses to yield compact and discriminative multimodal representations. Extensive experiments on DroneVehicle, FLIR-Aligned, and LLVIP demonstrate the effectiveness of LDSDet, which achieves 85.2% mAP50, 45.3% mAP, and 67.1% mAP, respectively, showing strong robustness under day–night alternation, low-light environments, and complex illumination variations.

Remote Sensing doi: 10.3390/rs18111826

Authors: Md Abdullah Al Mazid Naphtali Rishe

Atmospheric correction is a critical preprocessing step in optical remote sensing, but repeated high-fidelity radiative transfer simulations remain computationally expensive for dense look-up-table generation, sensitivity analysis, retrieval support, and operational preprocessing. This study presents a physics-guided multi-fidelity surrogate framework for emulating atmospheric correction coefficients using paired 6S and libRadtran simulations. Atmospheric and geometric states are sampled using Latin Hypercube Sampling, and both radiative transfer models are evaluated under matched conditions for Sentinel-2 bands using spectral-response-function-aware coefficient generation. The high-fidelity targets are path reflectance, total transmittance, and spherical albedo. A physics-guided Kolmogorov–Arnold Network, termed pKANrtm, receives the atmospheric state and low-fidelity 6S coefficients, predicts the residual relative to libRadtran, and reconstructs the high-fidelity coefficients. The pKANrtm model uses an Efficient-KAN architecture and is trained with a physics-guided penalty applied in the original coefficient space. The proposed model is evaluated against state-of-the-art regression-based RTM surrogates. Across both standard and out-of-distribution (OOD) evaluation settings, pKANrtm achieves the strongest overall predictive performance among the compared models. Band-wise analysis shows that most Sentinel-2 bands are accurately emulated, while absorption-sensitive bands remain comparatively challenging. Runtime benchmarking demonstrates substantial acceleration relative to libRadtran, with GPU inference providing approximately four orders of magnitude single-sample speedup and batched inference reaching tens of thousands of samples per second. As an initial real-scene validation, the trained pKANrtm correction was applied to a Sentinel-2A acquisition over the Gobabeb RadCalNet site, demonstrating that the learned residual correction improves downstream surface-reflectance retrieval beyond synthetic RTM-to-RTM coefficient emulation. These results indicate that physics-guided multi-fidelity pKANrtm emulation provides an accurate, physically structured, computationally efficient, and practically useful strategy for atmospheric correction coefficient generation.

Remote Sensing doi: 10.3390/rs18111825

Authors: Xuan Zhao Qian Tan Yanpeng Cai

Planted forests are increasingly promoted to meet rising demand for forest products and restore degraded lands, but their extent and ecological implications are often misrepresented because tree crops (e.g., orchards, plantation agriculture) exhibit similar spectral and spatial signatures to planted forests. This study aims to improve differentiation between planted forests and tree crops within national-scale restoration programs. We combined Landsat-derived NDVI time series targeting disturbance-related phenological windows with the LandTrendr algorithm to map planting/clearcutting events and fused in situ spectral measurements with Sentinel-2 to develop a modified orchard flowering index (MOFI). Random forest models evaluated classification performance using combinations of spatiotemporal spectral features, biomass accumulation proxies, and the MOFI. Incorporating the MOFI improved discrimination of tree crops versus planted forests, raising the planted forest F1 from 0.751 to 0.843. The combination of the MOFI and spatiotemporal spectral features achieved the highest accuracy (F1 = 0.843). The results show tree crops are concentrated on plains and gentle mountain slopes, while plantations occur mostly on slopes > 15°, with tree crops comprising 27.1% of mapped planted tree area. These findings imply that many national planted forest map estimates may be biased without phenology- and biomass-informed methods and that integrating Landsat and Sentinel phenology metrics supports more accurate monitoring for management and policy.

Remote Sensing doi: 10.3390/rs18111824

Authors: Zhechen Yang Wangrui Zhang Qi Zhang Zongbao Hong Danjie Cheng Qiao Xu Yan Meng Yangjie Sun Yuxuan Liu

Waterbody extraction is of great significance for water resource investigation and monitoring. In addition to RGB bands, most common satellite images have a near-infrared (NIR) band. By combining these RGB-NIR bands, certain water, vegetation, and shadow indices can be calculated. The near-infrared band and these indices are very similar to the X modality in RGB+X data (common examples include RGB-D and RGB-Thermal). However, at present, no studies have thoroughly examined multimodal feature fusion from the RGB+X perspective in order to extract waterbodies with high precision. As a result, existing algorithms do not fully utilize satellite image information and have limited generalization ability. To overcome this limitation, we propose a dual-complexity backbone for waterbody extraction from the perspective of RGB+X data semantic segmentation. Its complex Transformer branch is used to extract RGB modality features, while its simple CNN branch is used to extract X modality features. This network structure can effectively capture multimodal, global, and local features in remote sensing images. It can also fully leverage the fact that the scale of RGB image datasets in computer vision is significantly larger than that of remote sensing waterbody extraction datasets. If a large pretrained model is used in the RGB branch, it is unnecessary to freeze the weights. Instead, both branches can be trained jointly, allowing the RGB branch to better adapt to the remote sensing waterbody extraction task without raising concerns that fine-tuning might undermine the pretrained model’s strong representation capability. We also propose two X modality configurations with strong generalization performance. To fully fuse multimodal features, we design a hybrid fusion module combining a CNN and a cross-attention mechanism. To integrate the multi-scale features, we employ a multi-scale Transformer structure in the RGB branch and design a multi-scale decoder. Our algorithm achieves state-of-the-art performance on the GID-5 dataset and competitive performance on the S1S2-Water dataset. Furthermore, it significantly outperforms existing methods in cross-dataset zero-shot transfer between the two datasets, with IoU/F1-score gains of 26.08%/27.33% on GID-5 and 38.74%/31.37% on S1S2-Water over previous SOTA methods. Our processing paradigm of modeling RGB-NIR remote sensing images as RGB+X data shows potential for generalization to other multi-modal remote sensing tasks. The dual-complexity backbone we design also has potential to be extended to other tasks that transfer large pretrained RGB models to remote sensing imagery with RGB-NIR four bands or even more spectral bands. We have open-sourced the code and trained models used in this research.

Remote Sensing doi: 10.3390/rs18111823

Authors: Zhenhua Cheng Yunchang Cao Linghao Zhou Hong Liang Kun Jing Panpan Zhao Yuyang Zhu Chenwei Yao Haifeng Yu

The key parameters of atmospheric convective stability, the K index (KI) and the Showalter index (SI), are important indicators for severe convective weather warnings. This study adopts a variational mode decomposition, bidirectional gated recurrent unit, and attention mechanism weighting combined model (VMD–BiGRU–Attention) to optimize core hyperparameters and verify model stability. Global Navigation Satellite System-derived precipitable water vapor (PWV) and relative humidity (RH) are incorporated as key parameters representing atmospheric water vapor conditions, thereby assisting VMD decomposition in accurately separating effective signals related to severe convection. The results show that the optimal VMD decomposition parameter K for the KI is 10 (minimum root mean square error [RMSE] = 3.96), whereas the optimal K for the SI is 11 (minimum RMSE = 1.87), verifying the applicability of VMD decomposition. In the validation using extreme rainfall events (2021–2025) at three meteorological stations in Guangxi (Baise, Nanning, and Hepu), the model, with the auxiliary contributions of PWV and RH, stably and accurately predicts the KI and SI for the next three hours, effectively capturing the critical characteristics of severe convection. The predicted results are consistent with the observed precipitation, demonstrating significant practical application value.

Remote Sensing doi: 10.3390/rs18111818

Authors: Patrick L. Brezonik Leif G. Olmanson

Pre-classification of lakes into optical water types (OWTs) is considered a useful step in analyzing satellite-based reflectance data. We used a dataset of 109 reflectance hyperspectra from Minnesota and Wisconsin lakes and rivers to evaluate the usefulness of pre-classification to improve the retrieval of water quality information from satellite data. Three OWT classes were derived from the dataset by K-means clustering using three integrative metrics of reflectance spectral shape and magnitude as clustering variables. Values of the three metrics can be determined from satellite reflectance data as well as hyperspectral data. The OWT classes had distinct water quality characteristics in terms of Secchi depth, chlorophyll-a, and colored dissolved organic matter (CDOM). Algorithms used to retrieve values of the variables from simulated Sentinel-2 band reflectance data usually yielded more accurate predictions when computed separately for each class than when computed for the entire dataset, although exceptions were found for some fitting metrics and models and results for chlorophyll-a were not definitive. The three water quality variables were related in distinct ways to the integrative shape metric of reflectance spectra, apparent visible wavelength (AVW), supporting its use to develop OWTs to organize waterbodies into water quality classes. AVW was correlated (r = 0.933) with the integrative metric, normalized difference index at green and red wavelengths (NDI). Based on that result, we found that OWTs developed using just two variables, AVW and a metric of spectral magnitude, were nearly the same as classifications using all three integrative metrics.

Remote Sensing doi: 10.3390/rs18111822

Authors: Yanru Yu Zejia Hao Letian Wen Jie Dong Mingsheng Liao

Large-scale linear water diversion infrastructures are highly susceptible to ground deformation induced by groundwater extraction, mining activities, and geological instability, posing potential risks to long-term operational safety. However, conventional SBAS-InSAR monitoring of ultra-long linear infrastructures is often constrained by extensive data volumes, computational burden, and uncertainty associated with empirical buffer selection. To address these issues, this study proposes a practical buffer optimization framework for deformation monitoring along the Middle Route Project (MRP) of the South-to-North Water Diversion Project (SNWDP), China. Using Sentinel-1A SAR images acquired from 2023 to 2024, multiple buffer scales were comparatively evaluated by jointly considering deformation inversion accuracy against leveling measurements and computational efficiency. The results indicate that a 5 km buffer achieves the optimal balance between monitoring reliability and processing efficiency. Validation against first-order leveling benchmarks shows high consistency, with an RMSE of 2.54 mm and an MAE of 2.08 mm. Spatial-temporal analysis reveals significant deformation heterogeneity along the MRP. Severe land subsidence was detected in the Tianjin section due to intensive groundwater exploitation, while localized uplift was observed in parts of Hebei Province, likely associated with groundwater recovery. In addition, pronounced subsidence related to mining activities was identified in Yuzhou, Henan Province. The proposed workflow provides a practical reference for deformation monitoring of large-scale linear water diversion infrastructures and demonstrates the potential applicability of buffer optimization strategies for similar long-distance engineering projects.

Remote Sensing doi: 10.3390/rs18111820

Authors: Yuehao Wang Haiqing Liu Mengmeng Zhang Lei Yu Dongfang Ma

UAV-based road traffic state monitoring and analysis have become a hotspot in current research, where road region detection serves as the prerequisite for the aforementioned applications. This paper proposes a UAV-driven edge-based lane-detection system, and a lane line semantic segmentation, modeling and road-region-detection method. Firstly, a lightweight lane line semantic segmentation model LSLNet is presented, where the strip- aware multi-branch depthwise operator (SMDO) and the Sobel-based feature-fusion scheme (SFFS) are used in conjunction to improve feature representation ability under low computational overheads. Furthermore, the segmented lane line mask is quantified into a parametric form and the lane-level road regions are constructed by lane line spatial geometric distribution. Finally, to evaluate the performance of the proposed method, an experiment is conducted using the self-constructed UAV-Laneline3K and UAV-Roadregion200 datasets. The experimental results show that LSLNet achieves 82.73% F1-score and 72.06% mIoU on the lane line semantic segmentation task, which runs at 82 FPS with merely 0.09M parameters and 13.0 GFLOPs. For road region detection, the mIoU and F1-score reach 97.62% and 98.86%, respectively. The results demonstrate that the proposed method enables accurate and robust road region detection in complex road environments with low computational costs.

Remote Sensing doi: 10.3390/rs18111821

Authors: Mohammed Alruqimi Abdelkader Riche Pierluigi Confuorto Mawloud Guermoui Silvia Bianchini Farid Melgani

Timely landslide detection and rapid qualitative assessment are fundamental to effective warning systems, hazard management, and risk mitigation. Yet, current practices that rely on on-site surveys and manual expert assessment remain risky, costly, and time-consuming. These limitations result in substantial delays between the event and the availability of actionable information. This study proposes a hybrid, multi-model framework that fuses RGB remote-sensing imagery with geospatial layers to enable timely landslide detection and actionable reporting. The pipeline couples an enhanced SegFormer (denoted as SDF-SegFormer-B2) model for landslide localization, a feature extraction technique for per-slide geo-attribute computation, and a lightweight instruction-tuned LLM (Mistral-7B-Instruct-v0.3) for structured, expert-style reporting. Although a few previous studies have explored landslide captioning, to our knowledge this is the first framework designed to generate structured technical reports enriched with terrain-context interpretation and qualitative intervention-priority indicators. Experiments use 26,758 georeferenced RGB tiles (64 × 64) with 3 m of spatial resolution from PlanetScope satellite imagery over Emilia–Romagna, Italy, with 68,592 annotated landslide boxes collected after the May 2023 rainfall events (~200 mm in 48 h on 1–3 May; 200–250 mm in 48 h on 16–17 May). The proposed SDF-SegFormer-B2 segmentation model achieved a precision of 85.54%, recall of 72.31%, and an F1-score of 78.39% on the unseen test dataset. To evaluate the quality of the generated landslide reports, 100 images were selected for domain-expert assessment. Among these, 58% of the reports were rated as “Very Good,” 30% as “Good,” 8% as “Acceptable,” and 4% as “Poor.” When considering only reports with complete and accurate inputs, 81.48% were rated “Very Good,” and 96.30% were rated either “Good” or “Very Good.” By integrating complementary models and modalities, the proposed approach automates localization-to-reporting and enables the generation of terrain-aware landslide summaries that may support preliminary decision-making and rapid post-disaster screening.

Remote Sensing doi: 10.3390/rs18111819

Authors: Yuanlin Qiu Ming Li Jianwei Jia Xiaohao Zhang Liangang Chen Zihe Tian Tao Wang Min Wan Wei Wang

The Yarlung Zangbo River Basin (YZRB) stores abundant solid water resources. These components are highly sensitive to climate warming and play a critical role in regulating downstream water availability. However, the spatiotemporal responses of the thermal state to ongoing climate change and its potential atmospheric forcing remain poorly understood. Here, we use satellite-based land surface temperature (LST) to characterize the thermal dynamics of the YZRB during 2000–2024. Further, a machine learning model combined with Shapley Additive Explanations (SHAP) is applied to quantify the pixel-level statistical contributions of meteorological variables to LST trends. The climatological LST exhibits pronounced spatial and seasonal heterogeneity, with lower temperatures in the northwestern and northeastern regions and higher temperatures in the central and southeastern regions. The intra-annual cycle follows a unimodal pattern, peaking in early summer, while downstream sub-basins show a delay in peaking times. Mean LST increases at a rate of 0.18 °C decade−1, while maximum LST warms at nearly twice this rate (0.40 °C decade−1) with widespread warming across the basin. However, minimum LST shows no significant long-term trend, mainly due to the polarization trend within the year. The warming signal shows strong season dependence, with the largest monthly warming trend exceeding 0.80 °C decade−1 for all three LST metrics. Attribution analysis identifies precipitation as the primary meteorological factor statistically associated with basin-scale LST trends. Wind speed may largely represent a response to increasing LST rather than a direct driving factor. Downward shortwave radiation, air temperature and specific humidity exhibit stronger influences in specific regions rather than at the basin scale. The dominant control of precipitation reflects strong monsoon influence on LST dynamics along the southern margin of the Tibetan Plateau.

Remote Sensing doi: 10.3390/rs18111816

Authors: Xiaoyun Jiang Xin Shen Kai Zhou Xiaoming Yang Lin Cao

The high-precision instance segmentation of tree saplings is a fundamental prerequisite for the high-throughput phenotypic analysis of individual seedlings in intelligent tree breeding and precision silviculture. However, sapling segmentation remains challenging because of blurred boundaries, object adhesion, missed detections, and inaccurate mask delineation in field environments. To improve sapling segmentation performance and address these challenges, this study proposes a multimodal Mask R-CNN framework in which RGB imagery was paired with one multispectral-derived vegetation index at a time to construct separate RGB-VI input combinations, taking ginkgo saplings as a representative case. A dataset of 400 saplings was constructed using a high-throughput field phenotyping platform. The backbone network was extended with an independent vegetation index branch, and three fusion strategies (early, multi-step, and late fusion) were designed within a feature pyramid network to enable multi-scale multimodal feature integration. The results showed that all multimodal models outperformed unimodal baselines in terms of segmentation accuracy and recall. Among them, the multi-step fusion strategy achieved the best performance, while the RGB-EVI multi-step fusion model achieved the highest strict-matching precision (AP@75 = 87.7%) and recall (71.3%), with superior performance in dense sapling delineation and background suppression. These findings indicate that multimodal feature fusion can effectively improve sapling instance segmentation and provide methodological support for high-throughput plant phenotyping.

Remote Sensing doi: 10.3390/rs18111817

Authors: Lei Liu Qi Li Jiaxin Lv Jiaxiang Wang

Existing multi-modal UAV tracking methods typically rely on fixed-interval dynamic template update strategies to capture diverse target appearances, together with predefined thresholds to select high-quality search regions for template update. However, due to the irregular motion of targets and the complexity of real-world scenarios, such passive update mechanisms suffer from notable limitations. Fixed sampling intervals often fail to adequately capture appearance variations, while fixed threshold-based selection is insufficient to accommodate diverse imaging conditions, leading to ineffective updates or the introduction of noisy templates, thereby degrading tracking robustness and accuracy. To address these issues, we propose a search region-guided adaptive dynamic template update framework for robust multi-modal UAV tracking, aiming to improve both scene adaptability and target matching capability. Specifically, we design a Guided Template Selection Transformer, which dynamically matches templates conditioned on the current search region, enabling the tracker to autonomously select the most suitable template for the target’s current state. Furthermore, we introduce a Dynamic Threshold Module that adaptively adjusts template selection criteria according to different tracking scenarios, ensuring the reliability and contextual relevance of candidate templates. In addition, we develop a Dynamic Template Memory Module to maintain an ordered repository of target templates under different target states, providing a structured and high-quality template pool for the proposed selection mechanism. Extensive experiments on a standard multi-modal UAV tracking benchmark demonstrate that the proposed method significantly outperforms existing approaches, effectively overcoming the limitations of conventional fixed update strategies. Moreover, the proposed approach exhibits strong generalization capability across three additional multi-modal tracking datasets from typical surveillance scenarios.

Remote Sensing doi: 10.3390/rs18111815

Authors: Yiping Song Liang Huang He Yang Shuo Li

Few-shot remote-sensing fine-grained ship classification (RS-FGSC) faces two coupled challenges: limited annotated samples and mixed-visibility imaging conditions caused by cloud occlusion. Although CLIP-based prompt learning provides useful transfer priors, conventional single-branch adaptation can encounter an over-correction dilemma: robust compensation applied globally may degrade clear samples, whereas clear-image optimization may fail on occluded samples. We propose CADP (Cloud-Aware Dual-Path Prompt Learning), which decouples clear and occluded processing through learnable routing. CADP contains three components: (1) a cloud detector (CloudDetector) trained with auxiliary cloud-state labels for instance-level routing, (2) a fine-grained adapter (FineGrainedAdapter) that preserves discriminative details for clear samples, and (3) a robust compensation path using occlusion-aware prompting (AOPD) from CARP for occluded samples. To evaluate mixed-visibility scenarios, we construct Mixed-Sky benchmarks by combining clear ship images with SeaCloud-Ship cloud-occluded samples introduced by CARP, using controlled cloud-mixed ratios (25%, 50%, and 75%) and a non-overlapping sampling strategy. Experiments from 1-shot to 16-shot show consistent gains over CoCoOp, CLIP-Adapter, and prior robust prompting methods. CADP achieves 35.49% average accuracy, improving the best-performing baseline in our protocol by +4.81 points (+15.7% relative improvement). Component ablations, routing controls, and attention visualizations indicate that explicit routing reduces negative transfer between clear and occluded samples.

Remote Sensing doi: 10.3390/rs18111814

Authors: Yan Zhang Xianghai Cao

Hyperspectral image (HSI) classification enables precise classification of land-cover types from rich spectral and spatial information. Recent methods combine convolutional neural network (CNN) and Mamba branches to exploit their complementary local and global modeling capabilities for HSI classification. However, most of these methods treat all spectral channels uniformly in feature fusion, failing to account for the discriminability differences across spectral bands. Moreover, most methods rely on a single classification head at the final layer, which may lead to vanishing gradients in shallow layers. To address these limitations, a spectral group-wise gated CNN–Mamba network with cross-stage mutual distillation, called SGGCMNet, is proposed. To address the first limitation, a CNN–Mamba spectral group-wise gating block (CMSB) is designed at the feature-fusion level. Specifically, the CMSB partitions channels into multiple sub-groups along the spectral dimension. Each sub-group learns its own fusion weights that balance local spectral–spatial cues produced by a CNN pathway with long-range context produced by a Mamba pathway. To address the second limitation, two loss-level optimization strategies are proposed jointly: A progressive deep supervision strategy with uncertainty-based dynamic weighting is proposed to attach classification heads at all network stages. A temperature-regulated cross-stage mutual-distillation mechanism is further designed to enable bidirectional knowledge transfer among classification heads at different stages. On three benchmark HSI datasets, SGGCMNet achieves state-of-the-art accuracy.

Remote Sensing doi: 10.3390/rs18111809

Authors: Ziyue Ma Cunjin Xue Chengbin Wu Chaoran Niu Zheng Xiang

In the geographic environment, mesoscale ocean eddies and similar phenomena exhibit continuous and gradual changes. However, due to limitations in remote sensing observation technology, the obtained observational data are discrete, which contradicts the continuously evolving characteristics of these phenomena. Although spatiotemporal interpolation is a key tool for bridging this gap, existing single-model methods fail to fully consider continuous process features, making it difficult to obtain consistent high-quality datasets. To solve this problem, this paper combines deep learning and geostatistics to propose an Evolutionary Process-embedded Marine Spatiotemporal Interpolation Model (EPMSIM). EPMSIM first applies Seasonal and Trend decomposition using Loess (STL) to decompose marine time-series fields into trend, seasonal, and evolutionary components. Then, a Convolutional Bidirectional Long Short-Term Memory (ConvBiLSTM) model is adopted to interpolate the trend and seasonal components. Meanwhile, a Process-based Spatiotemporal Dynamic Tracking Interpolation Method (PSDTIM) is designed to interpolate the evolutionary component. Finally, these components are combined through additive coupling to produce the final interpolation result. In case studies of mesoscale eddy interpolation using SST and SLA data, EPMSIM outperforms traditional geostatistical and deep learning baselines in RMSE, MAE, and SSIM. Experimental results confirm that the model achieves significant interpolation effects in marine environmental element fields with evolutionary characteristics, validating its effectiveness in capturing continuous evolution features of marine phenomena and its feasibility for generating high-temporal-resolution spatiotemporal datasets. This study provides a methodological reference for data interpolation of evolutionary process phenomena in marine information science, and this method can be extended to other similar marine environmental variables, serving research on marine ecological environments and dynamic processes.

Remote Sensing doi: 10.3390/rs18111813

Authors: Jingming Chen Zhaoyang Huang Feng Wang Zixiao Wen Jingxing Zhu Guangyao Zhou

Fine-Grained Few-Shot Classification (FG-FSC) in remote sensing has become a critical task, as the scarcity of high-quality annotated data severely restricts the performance of deep learning models in fine-grained classification. Although Contrastive Language-Image Pre-Training (CLIP) exhibits strong generalization ability in few-shot learning, it fails to generate discriminative text and image features when adapted to remote sensing tasks. In this paper, a framework is proposed to adapt CLIP to remote sensing FG-FSC from both visual and text aspects. First, we introduce a Distribution-AWare Adapter (DAWA) that adaptively fuses instance-level visual knowledge from few-shot samples with distribution-aware representations derived from Gaussian Discriminant Analysis based on the original CLIP zero-shot knowledge, leading to stable visual feature representations under various few-shot settings. A hybrid loss function that incorporates transductive and contrastive regularization is employed to further prevent overfitting and improve the discriminability of features. Furthermore, we generate category-level fine-grained text captions, optimizing the image–text alignment when extremely few training images are available. Experiments on multiple remote sensing and natural image datasets verify that the proposed framework achieves state-of-the-art few-shot fine-grained classification performance with a modest training cost, providing a practical solution for few-shot remote sensing image analysis.

Remote Sensing doi: 10.3390/rs18111812

Authors: Yuanmao Fan Shuanggen Jin Lei Hong

Accurate and continuous high-frequency water level monitoring is essential for flood control, water resource regulation, and hydrological studies in the Yellow River. However, traditional methods are often limited in complex inland river environments with insufficient temporal resolution, poor continuity, and weak robustness. In this study, high-frequency water level changes and their dynamic responses to water-sediment regulation were estimated at Huayuankou and Xiaolangdi stations based on a sliding window, variational mode decomposition (VMD), and multi-GNSS interferometric reflectometry (GNSS-IR). The results show that high-temporal-resolution water level series with a 5-minute interval were achieved at Huayuankou and Xiaolangdi. When compared with in situ gauge measurements, the GNSS-IR estimated water level has a root mean square error (RMSE) of 0.09 and 0.14 m, and coefficient of determination (R2) values with 0.92 and 0.96, respectively. These results demonstrated strong effects by reservoir regulation in both wide-and-shallow wandering reaches and canyon-controlled reaches. Water level responses to the 2025 water-sediment regulation operation showed that Xiaolangdi responded rapidly to upstream reservoir releases, whereas Huayuankou exhibited a delayed response, with the flood peak arriving about 21 h later and attenuating during downstream propagation. The proposed method shows strong potential for high-frequency water level monitoring and dynamic response analysis from GNSS-IR in complex inland rivers.

Remote Sensing doi: 10.3390/rs18111811

Authors: Heejun Park Seokhwan Hwang Jung Soo Yoon Narae Kang Sujong Lee

Quantifying changes in groundwater storage (GWS) remains a fundamental challenge in hydrology, given the sparsity of long-term in situ monitoring networks and the inherent difficulty of direct subsurface observation. Although GRACE and GRACE-FO satellite missions provide a means of tracking total terrestrial water storage at the continental scale, their coarse spatial resolution (~300 km) and monthly temporal sampling limit their direct applicability to regional groundwater studies. Here, we present a spatio-temporal disaggregation and data fusion framework for reconstructing daily GWS anomalies (GWSAs) across the Korean Peninsula, integrating GRACE/GRACE-FO Mascon solutions with the GLDAS Catchment Land Surface Model (CLSM). The approach leverages satellite-derived mass variations to constrain the model’s long-term anomaly structure while retaining the high-frequency temporal dynamics of land-surface modeling. The framework is evaluated against in situ bedrock monitoring well records from five sites: Seoul, Cheongyang, Uiseong, Imsil, and Wonju. Raw time-series correlations range from R = 0.14 to 0.70; upon removal of the monthly climatology to isolate non-seasonal variability, R improves to 0.49–0.72 across all sites, reaching 0.718 in Seoul and 0.707 in Cheongyang, with Cheongyang’s RMSE declining from 8.847 to 7.574 cm. These results indicate that the GRACE-CLSM fusion framework captures genuine sub-monthly groundwater dynamics beyond the dominant seasonal cycle. To our knowledge, this represents the first reconstruction of daily GWS changes for the Korean Peninsula with explicit preservation of spatial mass conservation, and the resulting dataset has direct utility for operational groundwater monitoring in a region subject to hydroclimatic variability.

Remote Sensing doi: 10.3390/rs18111810

Authors: Zhaoxu Zhang Lei Qian Yahan Wu Yujia Chen Yuanheng Sun Dan Wan

Intensive mining over recent decades has caused severe ground subsidence in mining regions, threatening safety and long-term sustainability. High-precision, continuous monitoring and prediction of subsidence are therefore urgently needed. Traditional methods—terrestrial surveying and GPS—offer limited coverage, sparse measurement points, high costs, and poor scalability, making them unsuitable for large-scale, long-term surface deformation monitoring. InSAR is widely used for ground deformation monitoring due to its wide-area coverage, long-term sampling, high spatial resolution, and millimeter-scale precision. However, conventional InSAR often fails in vegetated areas and under steep deformation gradients—common in mining zones. To overcome these limitations, this study applied SBAS-InSAR, a method better suited for large-magnitude, continuous subsidence monitoring in mining areas. This study proposed an enhanced hierarchical spatiotemporal dependency graph neural network (HSDGNN) integrated with a Long Short-Term Memory (LSTM) module to improve temporal feature representation. Using this model, this study predicted surface subsidence at the Dexing Copper Mine under environmental drivers. Key findings are as follows: (1) Surface subsidence exhibited pronounced spatial heterogeneity and strong temporal nonlinearity; major subsidence zones were localized in open-pit excavation areas and waste rock dumps, with peak subsidence rates reaching −126.121 mm/yr. (2) Precipitation and soil moisture emerged as the dominant environmental controls on subsidence, displaying distinct seasonal modulation and quantifiable lagged responses—up to several months—relative to subsidence onset. (3) The HSDGNN model achieved high predictive accuracy for both Mine 1 and Mine 2, attaining R2 values of up to 0.9950. This work establishes a robust, scalable, and operationally viable framework for high-precision subsidence monitoring and forecasting in geologically and anthropogenically complex mining environments.

Remote Sensing doi: 10.3390/rs18111808

Authors: Mingjian Li Liguo Liu

Accurate prediction of radio wave propagation in evaporation ducts is critical for radar systems but faces significant environmental uncertainties. This study presents an uncertainty quantification and global sensitivity analysis framework comparing three surrogate models: Polynomial Chaos Expansion, Ordinary Kriging, and Polynomial-Chaos Kriging. Using a parabolic equation solver, we quantify how five parameters—mean duct height, duct height slope, potential refractivity gradient, frequency, and root mean square (RMS) wave height—affect propagation loss. We assess predictive accuracy, perform Sobol-based sensitivity analysis, and explore how surrogate performance relates to the normalized frequency V, a parameter characterizing modal complexity. Results show that Kriging consistently outperforms the others: its local interpolation capability proves essential for capturing rapid spatial oscillations caused by multimode interference. We observe a statistically significant negative correlation between Kriging’s prediction error and V, suggesting that its local interpolation becomes increasingly advantageous as the modal complexity of the field (quantified by V) increases. This provides a physically interpretable, though not yet predictive, link between surrogate model choice and the underlying propagation physics. Sensitivity analysis reveals that mean duct height dominates uncertainty at short-to-medium ranges, while the potential refractivity gradient becomes increasingly influential at longer ranges. RMS wave height exhibits localized effects near multipath nulls, particularly at higher frequencies. These findings provide quantitative guidance for prioritizing environmental measurements and offer a physically interpretable basis for surrogate model selection in evaporation duct problems.

Remote Sensing doi: 10.3390/rs18111805

Authors: Zhen Fang Xu Ma

Recently, artificial neural networks (ANNs) have shown impressive performance in the compressive hyperspectral imaging (CHI) reconstruction task, but the high energy consumption limits their deployment on energy-constrained devices. This paper develops a novel spiking neural network (SNN), termed spiking spectral-weighting reconstruction network (SSWR-Net), to significantly improve the energy–efficiency ratio in CHI reconstruction. Firstly, a spiking spectral-weighting convolution block is proposed to adaptively modulate the spiking signals, enabling the SNN to fit continuous spectral correlation curves. Secondly, a residual feature reuse module with more direct connections is designed to achieve efficient and lightweight spatial–spectral feature extraction. Thirdly, customized feature scaling architectures are introduced to resolve the dimensional mismatch issue and enhance information flow. Finally, we propose a novel temporal-wise progressive training method to optimize the multi-timestep SSWR-Net, which can significantly improve both training efficiency and reconstruction quality. Both simulation and real experiments demonstrate the superiority of the proposed method in both CHI reconstruction performance and energy efficiency. Specifically, SSWR-Net outperforms its ANN-based counterpart by 0.87 dB at a 19.74% energy cost.

Remote Sensing doi: 10.3390/rs18111806

Authors: Yue Bai Wei Quan Xuming Shi Zeyi Yan Guoliang Yuan

Civil infrastructure plays a critical role in ensuring societal safety and economic development. However, structural damages such as cracks inevitably occur during long-term service. Traditional manual inspection methods are insufficient to meet the demands of large-scale and routine monitoring. Unmanned Aerial Vehicles (UAV) remote sensing has become an important approach for Structural Health Monitoring (SHM), owing to its high spatial resolution imaging capability and superior operational flexibility. Nevertheless, existing studies focus on optimizing individual algorithms, lacking a systematic analysis oriented toward multi-scenario engineering applications. Therefore, we present a comprehensive review of UAV-based crack detection techniques for infrastructure using remote sensing imagery. First, publicly available datasets, UAV platforms, and evaluation metrics are systematically summarized. Then a multi-level visual analysis framework for UAV inspection is established. The framework categorizes existing methodologies into five levels: image-level classification, object-level detection, pixel-level segmentation, geometric quantification, and three-dimensional (3D) reconstruction, followed by a systematic evaluation of representative methods. Furthermore, the applicability of different methods across diverse scenarios, including bridges, pavements, dams, building facades and wind turbine blades, is systematically explored. Finally, the key challenges and future research directions are discussed. This review aims to provide a systematic theoretical foundation and methodological reference for advancing UAV-based infrastructure crack inspection from algorithm development toward practical multi-scenario engineering applications.

Remote Sensing doi: 10.3390/rs18111807

Authors: Hangyu Cao Zhao Chen

High-resolution synthetic aperture radar (SAR) ship detection plays a pivotal role in maritime surveillance and ocean monitoring. However, it remains challenging in practice because of the single-channel imaging modality, severe multiplicative speckle noise, and the pronounced scale imbalance in which sparse large vessels are easily under-optimized compared to the dominant small and medium instances. In this paper, we propose a Texture Perception Convolution Network (TPCNet), a practical and reproducible detection framework to improve feature extraction robustness and high-IoU localization under a unified strict-COCO evaluation protocol. TPCNet begins with a lightweight texture perception convolution (TPConv) that augments the raw SAR intensity with a local fluctuation cue to stabilize early feature representations for SAR images affected by strong multiplicative speckle noise (speckle-rich imagery). To address scale skew during training without modifying dataset splits, a region-balanced sampler (RBS) is introduced to increase the sampling probability of images, thereby improving the effective exposure of informative large-target structures. A background similarity augmentation (BSA) is proposed to enrich medium and large instances while reducing unrealistic boundary artifacts via compatible background selection and soft blending. Beyond component-level designs, the high-IoU localization is highly sensitive to geometric perturbations. Accordingly, TPCNet adopts a two-stage localization-oriented training strategy that first learns robust multi-scale representations and then refines box regression by tightening translation and scale ranges during fine-tuning. Under strict-COCO settings, TPCNet achieves SOTA performance with an AP of 71.20% on HRSID and an AP of 72.7% on SSDD. Comprehensive ablation studies demonstrate that TPConv, RBS, BSA, and the proposed finetuning strategy contribute complementary gains, providing a transparent baseline and a strong recipe for future SAR ship detection research.

Remote Sensing doi: 10.3390/rs18111804

Authors: Xinheng Han Haoyuan Zhang Jiacheng Wang Jielei Xu Yingjie Lv Zunning Zhou Xiaoxue Feng Feng Pan

The rapid advancement of unmanned aerial vehicle (UAV) technology has made air-to-air UAV object detection increasingly essential. However, the model faces additional challenges including small target sizes, motion blur, illumination variations, and stringent real-time performance requirements under constrained computational resources. To address these challenges, this paper proposes A2A-YOLO, a specialized detection model that introduces LECA-Conv for local and channel feature enhancement to effectively mitigate motion blur and illumination variations while incorporating GhostModulev2 for efficient feature extraction and Tiny Detection Heads for improved small target recognition. The proposed LECA-Conv module operates on the principle that attention parameters need not directly modify original feature maps, a key insight validated through extensive experiments. Extensive evaluations on the Det-Fly dataset demonstrate A2A-YOLO’s superior performance with 85.0% precision (PP), 80.7% recall (PR), and 81.9% average precision (AP), outperforming YOLO11 by 0.9%, 8.4%, and 6.5%, respectively. The proposed method demonstrates outstanding performance across diverse backgrounds and challenging conditions including motion blur and illumination variations. The model achieves real-time detection at 15 FPS on RK3588 platform while delivering remarkable performance in infrared small target detection.

Remote Sensing doi: 10.3390/rs18111803

Authors: Guangtong Sun Ping Song Guohong Zhang

On 29 July 2025, an Mw 8.8 megathrust earthquake occurred offshore of the southeastern Kamchatka Peninsula, ranking among the ten largest earthquakes worldwide since 1900. Due to observational limitations, the rupture characteristics of large earthquakes along the Kamchatka subduction zone and the north–south contrast in earthquake magnitudes remain poorly understood. In this study, we combine InSAR data, GNSS displacements, and teleseismic waveforms to investigate the spatiotemporal evolution of the 2025 mainshock by constructing a curved fault geometry with along-strike and downdip variations and applying finite-fault inversion together with back-projection analysis. The inversion results show that the mainshock was characterized by unilateral rupture propagating from northeast to southwest, with a rupture length of about 560 km, a duration of about 200 s, and dominant slip concentrated at depths of 15–30 km, with a peak slip of about 10 m. Slip was weak during the initial nucleation stage near the hypocenter, whereas the main slip patch was located within a strongly locked region in the southern segment, and the rupture accelerated rapidly after entering that region. The back-projection results indicate that high-frequency radiation mainly migrated southwestward and was concentrated along the boundaries of the large-slip region and possible structural segmentation zones. These results indicate that the rupture behavior of the 2025 mainshock was jointly controlled by curved megathrust geometry and along-strike locking heterogeneity. The north–south contrast in earthquake size along the Kamchatka subduction zone may result from the combined effects of stronger locking and smoother megathrust geometry in the south, versus more complex fault geometry and submarine tectonic features in the north. This study provides new constraints on rupture processes, seismic cycle behavior, and regional seismic hazard along the Kamchatka subduction zone, and offers important implications for understanding the mechanisms and magnitude potential of future great earthquakes in the Kamchatka region.

Remote Sensing doi: 10.3390/rs18111802

Authors: Yiting Wang Zengnan Li Xin Zhang Donghui Xie

Forests in the Qinling Mountains play a critical role in maintaining regional ecosystem services, yet long-term, high-resolution forest disturbance datasets at the regional scale remain limited, particularly in mountainous and cloud-prone environments. Existing forest disturbance products are largely based on Landsat imagery and optimized for global-scale applications, which may constrain their performance at regional scales. In this study, we developed a 30 m resolution forest disturbance dataset for the Qinling Mountains spanning 1999–2025 by integrating Landsat and Sentinel-2 time-series imagery with the LandTrendr algorithm. Annual Normalized Burn Ratio time series were generated through multi-sensor fusion of Landsat and Sentinel observations, improving temporal continuity and data availability. Based on these annual composites, LandTrendr was applied to produce consistent annual forest disturbance maps. A comprehensive validation framework was implemented using 2000 visually interpreted disturbance sample points and 60 independently documented disturbance events. The results show strong temporal agreement between detected and reference disturbance years, with a regression slope of 0.89 and an R2 of 0.93. Spatial validation based on disturbance events yielded an overall accuracy of 90.95%. Comparative analyses indicate that the proposed dataset exhibits improved spatiotemporal consistency relative to existing forest disturbance products, including Global Forest Change (GFC) and the Forest Age Dataset of China (FAGE), particularly in complex mountainous terrain. This study provides a long-term, regionally optimized forest disturbance dataset for the Qinling Mountains and demonstrates the applicability of Landsat–Sentinel annual composites for reliable forest disturbance monitoring in mountainous regions.

Remote Sensing doi: 10.3390/rs18111801

Authors: Bertrand Rouet-Leduc Claudia Hulbert

Phase unwrapping, the recovery of a continuous signal from measurements known only modulo 2π, is a ubiquitous problem in coherent imaging, from medical MRI to radar remote sensing. In Interferometric Synthetic Aperture Radar (InSAR), phase unwrapping is both critical and computationally demanding: current methods require minutes to hours per interferogram and frequently fail on large images. We present FAUST-ADMM (Fast ADMM Unwrapping via Spectral Transforms), an algorithm that formulates phase unwrapping as a weighted L1 optimization and solves it efficiently on GPU using the Alternating Direction Method of Multipliers (ADMM). Each iteration reduces to a Poisson equation solved in closed form via the Discrete Cosine Transform, followed by element-wise soft thresholding, both trivially parallel. On 500 synthetic earthquake interferograms, FAUST-ADMM achieves 99% accuracy with reference-point correction, matching SNAPHU, MCF, and PUMA, while running 10 to 100× faster. On a full three-subswath Sentinel-1 interferogram of the 2019 Ridgecrest M7.1 earthquake (∼6500 × 8500 pixels), FAUST-ADMM agrees with SNAPHU on 99.7% of pixels in 35 s, a 74× speedup. Our method makes batch unwrapping of large InSAR time series practical on a single consumer GPU.

Remote Sensing doi: 10.3390/rs18111800

Authors: Yuanyuan Han Guoqing Zhu Kaixuan Wen Xinlong Tan Wanqing Wu Wenyan Guo Fei Xie

Tropospheric ozone (O3) is a critical air pollutant that poses significant risks to human health and ecosystems. While previous studies have primarily focused on O3 changes in Eastern China, limited attention has been given to Northwest China, where fragile but ecologically important systems may be vulnerable to O3 pollution. The temporal evolution and driving mechanisms of surface O3 in this region remain poorly understood. Using the European Centre for Medium-Range Weather Forecasts Reanalysis Version 5 (ERA5) datasets and simulations from the Community Atmosphere Model with Chemistry (CAM-Chem), we identified a significant increase in summer surface O3 concentrations across Northwest China from 1980 to 2020, with the most pronounced rise occurring during 1993–2010. This period accounts for the majority of the long-term upward trend, despite relative declines before and after. The increase in O3 during 1993–2010 is primarily attributed to rising surface temperatures, which reduce hydroperoxyl radical (HO2) concentrations and enhance nitrogen dioxide (NO2) production, leading to elevated nitrogen oxides (NOx) levels and promoting O3 formation. The warming trend is closely associated with a concurrent decrease in low cloud cover, which increases surface shortwave radiation and further contributes to surface warming. Further investigation reveals that warming sea surface temperature (SST) in the North Pacific influence atmospheric circulation through wave train processes, amplifying the regional geopotential height field. These circulation changes reinforce the reduction in low cloud cover and the associated increases in surface temperature and O3 concentrations over Northwest China. The decadal variability of North Pacific SST may therefore serve as an important indicator of long-term surface ozone variability in this region.

Remote Sensing doi: 10.3390/rs18111799

Authors: Luana O. Barros Luiz F. Sapucci Caroline Viezel Victor A. Ranieri Ivette H. Baños Carlos F. Bastarz Eder P. Vendrasco Thaisa G. Lopes Sindy S. S. Almeida João G. Z. de Mattos José A. Aravequia

Atmospheric Motion Vectors (AMVs) from geostationary satellites are a critical observational source for data assimilation, particularly in regions with sparse observations, such as the Southern Hemisphere. This study evaluates the impact of assimilating AMVs from the Geostationary Operational Environmental Satellite (GOES) series into the Numerical Modeling and Assimilation System (SMNA) used at the Center for Weather Forecasting and Climate Studies of the National Institute for Space Research (CPTEC/INPE). The SMNA consists of the Brazilian Global Atmospheric Model (BAM) coupled with the Gridpoint Statistical Interpolation (GSI) data assimilation system. Two experiments were conducted in February 2024: a control experiment that assimilated all conventional observations along with AMVs from GOES-16 and GOES-18 satellites, and a second experiment (data denial), in which the AMVs were excluded. This time period coincided with the formation of the tropical cyclone Akará offshore the southeast coast of Brazil. The diagnostic analysis of the assimilation process indicates a substantial increase in the relative contribution of wind observations to the cost function and a reduction in the differences between the background and the analysis, particularly in the mid and upper troposphere. Forecast verification showed that assimilating AMV data led to a reduction in RMSE and an increase in anomaly correlations for several variables, including wind and temperature at various vertical levels. The positive impact of GOES AMV data on the representation of the tropical cyclone Akará is evident in the improved positioning, intensity, and circulation structure of the cyclone, particularly during its intensification phase. With tropical cyclone events over South America becoming more frequent in recent years, results from this study indicate the critical need to assimilate AMV data to improve forecast skill. Furthermore, the assimilation of GOES AMVs significantly enhances the representation of atmospheric circulation over South America, particularly improving the predictability of large-scale events such as cyclones in the South Atlantic.

Remote Sensing doi: 10.3390/rs18111798

Authors: Qinghe Sun Hua Wang Shuai Wang Teng Yang Hui Zhao Xuewu Fan

With the rapid advancement of Earth observation technologies and the growing demand for intelligent remote sensing applications, high-resolution remote sensing imagery provides critical data support for a range of downstream applications, including land monitoring and disaster assessment. In this context, multi-label remote sensing image classification has become an important research task, because a single image may contain multiple ground-object categories with complex spatial distributions and semantic co-occurrence relationships. However, challenges such as the coexistence of multiscale objects, complex semantic dependencies, and long-tail category distributions impose significant limitations on existing methods in terms of feature representation capacity and class-balanced modeling. To address these challenges, a Multiscale Dynamic Reasoning Network (MSDR-Net) is proposed. Different from methods that focus on localized optimization for a single challenge, MSDR-Net establishes a task-driven modeling framework that jointly integrates multiscale feature extraction, label-aware semantic reasoning, and long-tail category optimization within an end-to-end architecture. The proposed network consists of three core modules. The Multiscale Feature Enhancement (MSFE) module incorporates a Feature Pyramid Network-based fusion mechanism, integrating deep semantic information with shallow, detailed features to effectively enhance the representation of multiscale objects. The Dynamic Semantic Reasoning (DSR) module introduces a Transformer-based global attention mechanism that models long-range dependencies among image features, enabling the capture of complex global semantic relationships. In the loss optimization stage, a Difficulty-Weighted Loss (DW-Loss) is introduced, which jointly incorporates category frequency weights and prior difficulty coefficients to dynamically regulate the contributions of rare classes and hard samples during training, thereby mitigating bias induced by class imbalance. Experiments conducted on the large-scale Detection in Optical Remote Sensing Images dataset demonstrate that the proposed method achieves superior performance. Ablation studies validate the effectiveness of each component, while comparative experiments indicate that MSDR-Net achieves a mean Average Precision of 95.88%, outperforming existing state-of-the-art methods. An improvement of approximately 1.74% is observed over the strongest baseline, MSCA, with consistent advantages demonstrated across Overall F1 and Class-wise F1 metrics. By unifying multiscale feature extraction, global semantic reasoning, and balanced loss optimization within a single framework, MSDR-Net provides a robust and efficient solution for multi-label classification in complex remote sensing scenarios.