Radar Data Research Articles

Human Activity Recognition Based on Deep Learning and Micro-Doppler Radar Data

Activity recognition is one of the significant technologies accompanying the development of the Internet of Things (IoT). It can help in recording daily life activities or reporting emergencies, thus improving the user’s quality of life and safety, and even easing the workload of caregivers. This study proposes a human activity recognition (HAR) system based on activity data obtained via the micro-Doppler effect, combining a two-stream one-dimensional convolutional neural network (1D-CNN) with a bidirectional gated recurrent unit (BiGRU). Initially, radar sensor data are used to generate information related to time and frequency responses using short-time Fourier transform (STFT). Subsequently, the magnitudes and phase values are calculated and fed into the 1D-CNN and Bi-GRU models to extract spatial and temporal features for subsequent model training and activity recognition. Additionally, we propose a simple cross-channel operation (CCO) to facilitate the exchange of magnitude and phase features between parallel convolutional layers. An open dataset collected through radar, named Rad-HAR, is employed for model training and performance evaluation. Experimental results demonstrate that the proposed 1D-CNN+CCO-BiGRU model demonstrated superior performance, achieving an impressive accuracy rate of 98.2%. This outperformance of existing systems with the radar sensor underscores the proposed model’s potential applicability in real-world scenarios, marking a significant advancement in the field of HAR within the IoT framework.

Advancements in Deep Learning for Autonomous Driving in Indian Road Conditions

Recent advancements in deep learning techniques tailored for autonomous driving in Indian road conditions are crucial for revolutionizing transportation systems. Indian roads present unique challenges, including unpredictable traffic patterns, diverse road infrastructures, and challenging weather conditions. Deep learning is pivotal in addressing these challenges, focusing on perception, decision-making, and control. Various architectures, including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Deep Reinforcement Learning (DRL), are analyzed for their efficacy in autonomous driving. Sensor fusion techniques enhance perception capabilities, such as lidar, radar, and camera data integration. Additionally, advancements in semantic segmentation and object detection algorithms improve scene understanding and obstacle recognition. Collaborative efforts between academia, industry, and government agencies are essential for accelerating the deployment of autonomous driving technology in Indian road conditions. In conclusion, leveraging cutting-edge deep learning methodologies enables safe and efficient navigation of autonomous vehicles, ushering in a transformative future in transportation systems.

New method for landslide susceptibility evaluation in alpine valley regions that considers the suitability of InSAR monitoring and introduces deformation rate grading

ABSTRACT Incorporating Interferometric Synthetic Aperture Radar (InSAR) data sources into the landslide susceptibility evaluation process has yielded favorable outcomes in some studies. However, there are fewer analyses of the applicability of InSAR data monitoring, and a framework of accurate and effective grading criteria needs to be developed. To overcome these limitations, this study presents a novel and precise approach for landslide susceptibility assessment in alpine valley regions. This method incorporates the suitability of InSAR monitoring and introduces a grading system based on deformation rates, enhancing accuracy and efficiency. Taking the Dongchuan district, the most typical high mountain valley area in southwest China, as the research object, the SAR is quantitatively simulated and analyzed in this study using the improved R-index shadow layover method. Then, the optimal Synthetic Aperture Radar (SAR) monitoring scheme is derived, calculate deformation rates using LiCSBA (small baseline subset with the automated sentinel-1 InSAR processor) technology, and accurately establish the extent of landslide hazards (including potential landslides) by incorporating high-resolution images. The criteria for grading susceptibility within the landslide hazard range are based on deformation rates. Evaluation factors for corresponding grid cells are obtained. The best evaluation factors are selected using covariance diagnosis and gray correlation analysis. The landslide susceptibility model is developed utilizing the Particle Swarm Optimization-Back Propagation (PSO-BP) algorithm. It includes evaluation techniques for regions without deformation rates. The study findings demonstrate that: (i) Analyzing SAR suitability in alpine and canyon areas is crucial. Complementary monitoring with SAR lift tracks may only sometimes resolve geometric distortion issues in all these regions. (ii) The InSAR deformation rate can be an essential evaluation factor for landslide susceptibility evaluation. (iii) The proposed method effectively addresses low coherence challenges in certain alpine valley regions, where grading based on deformation rate is complex. The landslide susceptibility evaluation model is validated using performance evaluation indexes (Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and R-Square (R2)), confirming its reliability and effectiveness. (iv) The proposed method improves the grading accuracy by 37.84 ~ 60.91%. Overall, our proposed new landslide susceptibility method brings a new way of evaluating landslide susceptibility in alpine valley regions.

Using SAR-based products to calculate potato carbon uptake in a tropical Andean region

ABSTRACT Gross primary productivity (GPP) is an essential parameter to estimate the efficiency of carbon transfer in terrestrial ecosystems. The daily GPP has been monitored in the past using mainly optical satellite imagery. However, GPP has never been monitored using satellite Synthetic Aperture Radar (SAR) imagery. We evaluate the possibility of using Sentinel-1 SAR data to estimate GPP and compare with field GPP measurements using eddy covariance (EC) systems in three commercial potato fields, located in the Andean region of Colombia, using three different water irrigation levels. Raw and processed radar data from Sentinel-1 were compared with the daily and accumulated GPP from the EC. Results indicate that SAR data has a high correlation with the accumulated GPP measured by the EC in the field. The normalized radar backscatter and radar brilliance coefficients with VH polarization show a good correlation with the EC accumulated GPP in irrigated potato fields (R2: 0.77–0.81), while radar vegetation indices show a good correlation with the EC accumulated GPP in potato fields with no irrigation (R2≈0.82). In particular, the accumulated GPP during the whole potato crop cycle was estimated with good accuracy (~5% error with irrigation and ~10% error with no irrigation).

Range-Doppler-Time Tensor Processing for Deep-Space Satellite Characterization Using Narrowband Radar

There is growing demand for the high-fidelity characterization of satellites in Geosynchronous Earth Orbit (GEO) to support Space Domain Awareness (SDA). This is particularly true for newly launched satellites, where it is necessary for satellite providers to ascertain whether components have deployed properly. Conventional wideband radar systems are capable of imaging satellites provided that (i) they have sufficient power aperture and bandwidth, and (ii) they observe enough target aspect change to generate a resolved image. While wideband radars are used routinely for characterizing satellites in Low-Earth Orbit (LEO), powerful radars with sensitivity sufficient for large GEO ranges (36,000 km or greater) are lacking. Thus, researchers often rely on more widely available high-power narrowband tracking radars for GEO characterization. In this paper, we present a novel range-Doppler-time (RDT) tensor processing technique for GEO characterization with narrowband radar. This technique encapsulates the strengths of previously proposed methods for narrowband-radar characterization at GEO, providing a generalized approach that can be applied in a variety of settings. The technique generates fully resolved 2D images of rotating GEO satellites in low-bandwidth scenarios. In cases where aspect change is limited, the technique provides detailed Doppler information for enhanced satellite status monitoring. This work presents a comprehensive quantitative analysis of the technique that considers the impact of key parameters on characterization performance. Simulated radar data, and radar data collected in a compact range on a scaled satellite model, are used to evaluate the technique.

Alpine topography of the Gamburtsev Subglacial Mountains, Antarctica, mapped from ice sheet surface morphology

Abstract. Landscapes buried beneath the Antarctic Ice Sheet preserve information about the geologic and geomorphic evolution of the continent both before and during the wide-scale glaciation that began roughly 34×106 years ago. Since the inception of this ice sheet, some areas have remained cold-based and non-erosive, preserving ancient landscapes remarkably intact. The Gamburtsev Subglacial Mountains in central East Antarctica are one such landscape, maintaining evidence of tectonic, fluvial and glacial controls on their distinctly alpine morphology. The central Gamburtsev Mountains have previously been surveyed using airborne ice-penetrating radar; however, many questions remain as to their evolution and their influence on the East Antarctic Ice Sheet, including where in the region to drill for a 1.5×106 year-long “oldest-ice” core. Here, we derive new maps of the planform geometry of the Gamburtsev Subglacial Mountains from satellite remote sensing datasets of the ice sheet surface, based on the relationship between bed roughness and ice surface morphology. Automated and manual approaches to mapping were tested and validated against existing radar data and elevation models. Manual mapping was more effective than automated approaches at reproducing bed features observed in radar data, but a hybrid approach is suggested for future work. The maps produced here show the detail of mountain ridges and valleys on wavelengths significantly smaller than the spacing of existing radar flightlines, and mapping has extended well beyond the confines of existing radar surveys. Morphometric analysis of the mapped landscape reveals that it constitutes a preserved (>34 Ma) dendritic valley network, with some evidence for modification by topographically confined glaciation prior to ice sheet inception. The planform geometry of the landscape is a significant control on locations of basal melting, subglacial hydrological flows and the stability of the ice sheet over time, so the maps presented here may help to guide decisions about where to search for oldest ice.

Ocean eddy detection based on YOLO deep learning algorithm by synthetic aperture radar data

Ocean eddies play a crucial role in the global energy cycle and significantly impact the transport of heat, salt, and nutrients in the global ocean. Spaceborne synthetic aperture radar (SAR), with its high spatial resolution (O(20 m)) and wide coverage, is an important tool for studying ocean eddies. In this paper, we propose a deep-learning network, named EOLO, to automatically detect ocean Eddies observed in C-band spaceborne SAR imagery, based on the You-Only-Look-Once (YOLO) deep learning algorithm. A Sentinel-1 (S1) SAR data-based ocean eddy dataset (named EddyDataset) is established to train the EOLO network. To effectively improve the performance of the EOLO network, we introduced a channel attention mechanism, adopted the up-sampling operator with the larger receptive field, and improved feature fusion method, anchor box size, and loss function. With the support of a high-performance detection network, the geographic information extraction module based on the affine geographic transformation model and a data preprocessing module were added, making EOLO an application-level framework. Our experiment results on EddyDataset demonstrate that EOLO achieves a high quality of eddy detection with 91.5% average precision. We further applied EOLO to entire scenes of S1 images to detect eddies in the Red Sea, the Baltic Sea, and the Western Mediterranean Sea, achieving 96.6%, 98.8% and 98.9% precision, respectively. Moreover, EOLO was used to extract size and location information of ocean eddies based on 6135 S1 scenes acquired in 2021 over the Western Mediterranean Sea. Their spatial characteristics were derived and compared with the eddies extracted from radar altimeter data, presenting interesting discrepancies between sub-mesoscale and mesoscale eddies in the ocean.

Correlated Decision Fusion Accompanied with Quality Information on a Multi-Band Pixel Basis for Land Cover Classification

Decision fusion plays a crucial role in achieving a cohesive and unified outcome by merging diverse perspectives. Within the realm of remote sensing classification, these methodologies become indispensable when synthesizing data from multiple sensors to arrive at conclusive decisions. In our study, we leverage fully Polarimetric Synthetic Aperture Radar (PolSAR) and thermal infrared data to establish distinct decisions for each pixel pertaining to its land cover classification. To enhance the classification process, we employ Pauli’s decomposition components and land surface temperature as features. This approach facilitates the extraction of local decisions for each pixel, which are subsequently integrated through majority voting to form a comprehensive global decision for each land cover type. Furthermore, we investigate the correlation between corresponding pixels in the data from each sensor, aiming to achieve pixel-level correlated decision fusion at the fusion center. Our methodology entails a thorough exploration of the employed classifiers, coupled with the mathematical foundations necessary for the fusion of correlated decisions. Quality information is integrated into the decision fusion process, ensuring a comprehensive and robust classification outcome. The novelty of the method is its simplicity in the number of features used as well as the simple way of fusing decisions.

Micro-Doppler Signature Analysis for Space Domain Awareness Using VHF Radar

The large quantity of resident space objects orbiting Earth poses a threat to safety and efficient operations in space. Radar sensors are well suited to detecting objects in space including decommissioned satellites and debris, whereas the more commonly used optical sensors are limited by daylight and weather conditions. Observations of three non-operational satellites using a VHF radar system are presented in this paper in the form of micro Doppler signatures associated with rotational motion. Micro Doppler signatures are particularly useful for characterising resident space objects at VHF given the limited bandwidth resulting in poor range resolution. Electromagnetic simulations of the micro Doppler signatures of the defunct satellites are also presented using simple computer-aided design (CAD) models to assist with interpretation of the radar observations. The simulated micro Doppler results are verified using the VHF radar data and provide insight into the attitude and spin axis of the three resident space objects. As future work, this approach will be extended to a larger number of resident space objects which requires a automated processing.

Influence of Assimilation of NEXRAD-Derived 2D Inner-Core Structure Data from Single Radar on Numerical Simulations of Hurricane Charley (2004) near Its Landfall

This study presents the first research that assimilates the ground-based NEXRAD observations-derived two-dimensional (2D), azimuthally averaged radar radial velocity and reflectivity within 60 km of radius from the hurricane center to examine their influence on the analysis and prediction of a hurricane near and after its landfall. The mesoscale community Weather Research and Forecasting (WRF) model and its four-dimensional variational (4D-VAR) data assimilation system are utilized to conduct data assimilation experiments for Hurricane Charley (2004). Results show that assimilation of the radar inner-core data leads to better forecasts of hurricane tracks, intensity, and precipitation. The improved forecast outcomes imply that the changes in dynamical, thermal, and moisture structures from data assimilations made more reasonable conditions for the hurricane development near and after its landfall. Overall results indicate that the assimilation of the radar-derived 2D inner-core structure could be a feasible way to utilize the radar data for improved hurricane prediction.

Prospecting Prediction for the Yulong Metallogenic Belt in Tibet Based on Remote Sensing Alteration Information and Structural Interpretation

The Yulong porphyry copper belt in eastern Tibet is located in the middle of Tethys–Himalayan metallogenic mega-province, which is one of the three major porphyry copper metallogenic mega-provinces. The Yulong copper belt belongs to the super porphyry copper belt and represents one of the most important copper mineralization prospecting areas in China. A significant quantity of research data shows that this study area belongs to the environment of intracontinental collision and compression, with a complex geological structure, magmatic rock development and excellent metallogenic geological background. However, because this area is located in an alpine and high-altitude area, it is difficult to carry out any traditional field geological surveys, and the existing studies of both prospecting and prediction are relatively weak. This study focused on information extraction for alteration minerals in the Yulong metallogenic belt and its surroundings based on multispectral data and hyperspectral data, establishing a spectral library of alteration minerals in this area. Based on Sentinel-1A radar data and Landsat-8 OLI color synthesis data, the linear structure of the study area was interpreted. On this basis, the information extraction results relating to alteration minerals obtained from multi-source remote sensing data, linear structure interpretation results and the geochemical exploration data of the study area were superimposed to comprehensively analyze the metallogenic geological conditions and mineralization characteristics in the area, establish remote sensing prospecting indicators there and optimize the potential areas for prospecting, providing technical support for the next step of prospecting and exploration in the area.

A MWCMLAI-Net method for LAI inversion in maize and rice using GF-3 and Lutan radar data

ABSTRACT This study aimed at alleviating the problems of unsatisfactory inversion accuracy and weak model stability in LAI remote sensing quantitative inversion. The properties and complex scattering mechanism of SAR data specify the polarization combinations and frequencies. This paper proposes an improved water cloud model combined with a deep neural network (MWCMLAI-Net) for high-precision inversion. The polarized GF-3 (C-band) and Lutan (L-band) were used to investigate the potential of SAR images to estimate LAI, a strong indicator of crop productivity. The study selected xiangfu district in the eastern part of Kaifeng City, Henan Province, as the test area and investigated the LAI of maize and rice. The RV I Freeman Model, backward scattering coefficient extracted by the modified cloud and water model (MWCM), and LAI obtained by the inversion of the MWCM were used as the inputs, and the MWCMLAI-Net inversion of the LAI was constructed. The results showed that the model’s inverted LAI fitting accuracies of maize and rice for the three fertility periods were better than the other models, with R 2 above 0.8516 and RMSE below 0.3999 m2/m2. The addition of noise did not affect the results.

The improvement of short-term quantitative precipitation forecast in mountainous areas by the assimilation of meteorological state variables retrieved by multiple-Doppler radar data

Abstract This study presents a sequential procedure formulated by combining a multiple-Doppler radar wind synthesis technique with a thermodynamic retrieval method, which can be applied to retrieve the three-dimensional wind, pressure, temperature, rainwater mixing ratio, and moisture over complex terrain. The retrieved meteorological state variables are utilized to re-initialize a high-resolution numerical model, which then carries out time integration using four different microphysical (MP) schemes, including the Goddard Cumulus Ensemble (GCE), Morrison (MOR), WRF single-moment 6-class (WSM6), and WRF double-moment 6-class (WDM6) schemes. It is found that through this procedure the short-term quantitative precipitation forecast (QPF) skill of a numerical model over mountainous areas can be significantly improved up to six hours. The moisture field plays a crucial role in producing the correct rainfall forecast. Since no specific microphysical scheme outperforms the others, a combination of various rainfall scenarios forecasted by different MP schemes is suggested in order to provide a stable and reliable rainfall forecast. This work also demonstrates that, with the proposed approach, radar data from only two volume scans are sufficient to improve the rainfall forecasts. This is because the unobserved meteorological state variables are instantaneously retrieved and directly used to re-initialize the model, thereby the model spin-up time can be effectively shortened.

Rock Layer Classification and Identification in Ground-Penetrating Radar via Machine Learning

Ground-penetrating radar (GPR) faces complex challenges in identifying underground rock formations and lithological structures. The diversity, intricate shapes, and electromagnetic properties of subsurface rock formations make their accurate detection difficult. Additionally, the heterogeneity of subsurface media, signal scattering, and non-linear propagation effects contribute to the complexity of signal interpretation. To address these challenges, this study fully considers the unique advantages of convolutional neural networks (CNNs) in accurately identifying underground rock formations and lithological structures, particularly their powerful feature extraction capabilities. Deep learning models possess the ability to automatically extract complex signal features from radar data, while also demonstrating excellent generalization performance, enabling them to handle data from various geological conditions. Moreover, deep learning can efficiently process large-scale data, thereby improving the accuracy and efficiency of identification. In our research, we utilized deep neural networks to process GPR signals, using radar images as inputs and generating structure-related information associated with rock formations and lithological structures as outputs. Through training and learning, we successfully established an effective mapping relationship between radar images and lithological label signals. The results from synthetic data indicate a rock block identification success rate exceeding 88%, with a satisfactory continuity identification of lithological structures. Transferring the network to measured data, the trained model exhibits excellent performance in predicting data collected from the field, further enhancing the geological interpretation and analysis. Therefore, through the results obtained from synthetic and measured data, we can demonstrate the effectiveness and feasibility of this research method.

Imaging Left-Lateral and Reverse Near-Surface Slip of the 2020 Mw 5.1 Sparta, North Carolina, Earthquake

ABSTRACT Coseismic deformation of the 9 August 2020 Mw 5.1 Sparta, North Carolina, earthquake was captured by an Advanced Land Observing Satellite 2 interferogram, a Sentinel 1A/B interferogram, and differential light detection and ranging (lidar). Owing to the moderate surface deformation compared to the inherent uncertainties of interferometric data, the interferograms suffered from low signal-to-noise ratios, exacerbated by spurious signals arising from the discrepancy in resolution between the interferograms and digital elevation model used during processing. Nevertheless, coseismic slip can be imaged using graph-guided fused LASSO, where LASSO is the least-absolute shrinkage and selection operator—an inverse method that combines sparsity constraints and smoothness regularization—to infer signal from noisy and/or incomplete data. We develop five coseismic slip models, one determined only from the Sentinel 1A/B interferogram, one determined from both interferograms, one determined from all the Interferometric Synthetic Aperture Radar (InSAR) data masking potentially spurious data corresponding to industrial parks, one from lidar 3D displacements only, and one constrained by the masked InSAR and the lidar data. All five models are coseismic slip on a single fault plane, trending northwest–southeast with dips ranging from 41° to 60° to the southwest. Coseismic slip is imaged to be shallow, with predominantly oblique reverse slip extending to about 2 km depth. We find that masking the interferometric data corresponding to the industrial parks yields a coseismic slip model in better concordance with ancillary information about the Sparta earthquake and that the lidar data yield additional constraints on near-surface coseismic slip. We show that our preferred coseismic slip model is consistent with a stress regime that does not depart from an Andersonian thrust regime to a large degree, constraining that the background stress is larger than 20–30 MPa.

Multi-Sensor Satellite Images for Detecting the Effects of Land-Use Changes on the Archaeological Area of Giza Necropolis, Egypt

The World Heritage Committee has been meeting to discuss the arrangements of existing World Heritage Sites, and, on 22–26 October, the area from Giza to the Dahshur was included in the list of World Heritage Sites. According to the Egyptian Antiquities Authority (EAA), the groundwater levels at the Pyramids Plateau are too shallow, which threatens the ancient Sphinx and Pyramids in Giza, Egypt. In addition, many geophysical studies have been carried out in the archaeological area of Giza, which prove that the area is facing the risk of a high level of groundwater, specifically threatening the Sphinx. Recent developments in Earth observation have helped in the field of land monitoring such as land use changes, risk observation, and the creation of models for protecting cultural heritage sites. This study aimed to examine the impact of land use changes on on the archaeological sites of the Giza Necropolis area by integrating various data sources including optical satellite imagery and SRTM data during the period of 1965–2019. A historical database of Corona 1965 and Landsat 2009 data was investigated along with the new acquisitions of Sentinel-2 2016 and Sentinel-1 2016 and 2019. In addition, the radar Sentinel-1 SLC data were collected and analyzed for calculating the land subsidence value in the area of interest through two periods between 6–30 July 2016 and 30 July–15 December 2016. Various methods were implemented, including cluster outliers, the Moran index, and spatial autocorrelation to examine the changes in urban masses. Additionally, the relationship between groundwater leakage and land subsidence in the region was investigated. The analysis was carried out using Envi5.3, ArcMap10.6.1, and SNAP6.0 software to extract spatial data from the raw data. The results from our investigation highlighted rapid changes in urban areas between 1965 and 2019. The data obtained and analyzed from optical and radar satellite imagery showed that changes in land use can cause changes in the topographic situation by decreasing the level of groundwater, which adversely affects Egyptian monumental pyramids and the Sphinx. Land use analysis showed that the urban area represented 7.63% of the total area of the study area in 1965, however it reached 32.72% in 2009, approximately half of the total area in 2016, and in 2019, the urban mass area increased to nearly two-thirds of the total area. The annual growth rate between 1965 and 2019 was estimated by nearly 0.642 km2/year. These land-use changes possibly affected the land subsidence value (−0.0138 m), causing the rising groundwater level close to the Sphinx. Using the information obtained from our RS- and GIS-based analysis, mitigation strategies have also been identified to support archaeological area preservation.

Three-Dimensional Human Pose Estimation from Micro-Doppler Signature Based on SISO UWB Radar

In this paper, we propose an innovative approach for transforming 2D human pose estimation into 3D models using Single Input–Single Output (SISO) Ultra-Wideband (UWB) radar technology. This method addresses the significant challenge of reconstructing 3D human poses from 1D radar signals, a task traditionally hindered by low spatial resolution and complex inverse problems. The difficulty is further exacerbated by the ambiguity in 3D pose reconstruction, as multiple 3D poses may correspond to similar 2D projections. Our solution, termed the Radar PoseLifter network, leverages the micro-Doppler signatures inherent in 1D radar echoes to effectively convert 2D pose information into 3D structures. The network is specifically designed to handle the long-range dependencies present in sequences of 2D poses. It employs a fully convolutional architecture, enhanced with a dilated temporal convolutions network, for efficient data processing. We rigorously evaluated the Radar PoseLifter network using the HPSUR dataset, which includes a diverse range of human movements. This dataset comprises data from five individuals with varying physical characteristics, performing a variety of actions. Our experimental results demonstrate the method’s robustness and accuracy in estimating complex human poses, highlighting its effectiveness. This research contributes significantly to the advancement of human motion capture using radar technology. It presents a viable solution for applications where precision and reliability in motion capture are paramount. The study not only enhances the understanding of 3D pose estimation from radar data but also opens new avenues for practical applications in various fields.

Research on Extracting Multiple Feature Impervious Surface Based on Fusion of Optics and Radar

Impervious surface is an important indicator for analyzing urban expansion, measuring the degree of urbanization, and characterizing the urban ecological environment. Accurately extracting impervious surface data is of great significance to regional economic development, disaster prediction, ecological restoration, and environmental assessment. In this paper, using multi-spectral, synthetic aperture radar (SAR) and surface temperature retrieval (LST) data, from the four perspectives of spectral features, time series features, SAR texture features, and coherence features, an index feature that highlights impervious surface information is constructed. The model generates a 10m impervious surface product in the urban area of Shaoguan, and verifies the accuracy of the real sample based on the 0.8m GF-2 data in the same period. The results indicate that the overall accuracy of extracting impermeable surfaces in the study area is 94%, with a Kappa coefficient of 0.92. The multi feature fusion random forest model combining optical data and SAR data has high extraction accuracy and applicability in the mountainous areas of southwestern China, which improves the misclassification of bare land and other land types as impermeable surfaces.

Sentinel-1-Based Soil Freeze–Thaw Detection in Agro-Forested Areas: A Case Study in Southern Québec, Canada

Nearly 50 million km2 of global land experiences seasonal transitions from predominantly frozen to thawed conditions, significantly impacting various ecosystems and hydrologic processes. In this study, we assessed the capability to retrieve surface freeze–thaw (FT) conditions using Sentinel-1 synthetic aperture radar (SAR) data time series at two agro-forested study sites, St-Marthe and St-Maurice, in southern Québec, Canada. In total, 18 plots were instrumented to monitor soil temperature and derive soil freezing probabilities at 2 and 10 cm depths during 2020–21 and 2021–22. Three change detection algorithms were tested: backscatter differences (∆σ) derived from thawed reference (Delta), the freeze–thaw index (FTI), and a newly developed exponential freeze–thaw algorithm (EFTA). Various probabilistic mixed models were compared to identify the model and predictor variables that best predicted soil freezing probability. VH polarization backscatter signals processed with the EFTA and used as predictors in a logistic model led to improved predictions of soil freezing probability at 2 cm (Pseudo-R2 = 0.54) compared to other approaches. The EFTA could effectively address the limitations of the Delta algorithm caused by backscatter fluctuations in the shoulder seasons, resulting in more precise estimates of FT events. Furthermore, the inclusion of crop types as plot-level effects within the probabilistic model also slightly improved the soil freezing probability prediction at each monitored plot, with marginal and conditional R2 values of 0.59 and 0.61, respectively. The model accurately classified observed binary ‘frozen’ or ‘thawed’ states with 85.2% accuracy. Strong cross-level interactions were also observed between crop types and the EFTA derived from VH backscatter, indicating that crop type modulated the backscatter response to soil freezing. This study represents the first application of the EFTA and a probabilistic approach to detect frozen soil conditions in agro-forested areas in southern Quebec, Canada.

Utilizing Polarization Diversity in GBSAR Data-Based Object Classification.

In recent years, the development of intelligent sensor systems has experienced remarkable growth, particularly in the domain of microwave and millimeter wave sensing, thanks to the increased availability of affordable hardware components. With the development of smart Ground-Based Synthetic Aperture Radar (GBSAR) system called GBSAR-Pi, we previously explored object classification applications based on raw radar data. Building upon this foundation, in this study, we analyze the potential of utilizing polarization information to improve the performance of deep learning models based on raw GBSAR data. The data are obtained with a GBSAR operating at 24 GHz with both vertical (VV) and horizontal (HH) polarization, resulting in two matrices (VV and HH) per observed scene. We present several approaches demonstrating the integration of such data into classification models based on a modified ResNet18 architecture. We also introduce a novel Siamese architecture tailored to accommodate the dual input radar data. The results indicate that a simple concatenation method is the most promising approach and underscore the importance of considering antenna polarization and merging strategies in deep learning applications based on radar data.