Sentinel-1A Synthetic Aperture Radar Research Articles

Enhancing precision of flood estimation in EOS-04 SAR imagery: a statistical approach

Floods pose a recurrent and devastating natural disaster in India, and accurate and timely assessment of their extent is crucial for effective disaster management and mitigation efforts. In this regard, this study proposes a novel approach for flood inundation estimation through the statistical analysis of two key geospatial datasets: the Normalized Difference Flood Index (NDFI) and temporal Standard Deviation images. To establish a baseline under ordinary conditions, two products were derived. The first product is the mean image during the non-flood season, and the second is the temporal Standard Deviation image for the entire year. These images are derived from Sentinel-1A Synthetic Aperture Radar (SAR) data using Google Earth Engine (GEE). The NDFI is calculated by comparing the mean image to the image obtained during flood event, allowing the identification of flood-affected areas. The process involves constructing Kernel Density Estimation (KDE) plots for NDFI and Standard Deviation, from which a 95% density ellipse is generated using the covariance matrix and eigenvalues. Multiple SAR scenes from diverse regions in India are analyzed individually, yielding density ellipses for each location. The innovative step lies in synthesizing these individual ellipses into the best fit ellipse through the convex hull method. This best fit ellipse encapsulates the general flood characteristics observed across disparate geographic regions, providing a unified representation. The derived best fit ellipse serves as a powerful tool for generating near-real-time flood inundation maps, with an overall accuracy of over 98.2%. Thus, the fusion of statistical insights and historical SAR information enables rapid near-real-time flood mapping, with enhanced accuracy of flood extent predictions.

Comparative analysis of leaf area index and maize yield estimation assimilating remote sensing and DSSAT crop simulation model

Maize is a global staple crop, impacting food security, economic development, and agricultural sustainability. This study investigates the integration of Sentinel-1A Synthetic Aperture Radar (SAR) data with the DSSAT CERES-Maize crop simulation model to estimate Leaf Area Index (LAI) and rabi maize yield in Belagavi district, Karnataka. Field data, including LAI, days to anthesis, silking, grain filling, and farmers' field practices, were collected for model calibration and validation, supplemented by crop-cutting experiments (CCE) to determine actual yields. The study revealed strong correlations between LAI values obtained from remote sensing (RS) and field observations, with RS-derived LAI showing an average agreement of approximately 96.07% compared to field measurements. The DSSAT model exhibited slightly better performance, averaging 97.09%. Statistical analysis for LAI showed an R² value of 0.853 for RS and 0.864 for DSSAT, indicating strong correlations with observed LAI values. For maize yield estimation, the DSSAT model demonstrated higher accuracy with an average yield of 8129 kg/ha, compared to RS-derived yield averages of 7533.9 kg/ha and CCE yield averages of 8096.6 kg/ha. The average concordance between DSSAT and CCE yields was 94.19%, while RS and CCE yields had an average concordance of 92.29%. Statistical analyses revealed coefficients of determination of 0.854 for DSSAT-CCE and 0.867 for RS-CCE comparisons. The study underscores the value of combining RS data with DSSAT for comprehensive and accurate crop yield forecasting, highlighting the potential for improved agricultural assessments and decision-making.

Observation of radial velocity along the Yangtze River based on sentinel-1A

ABSTRACT Estuaries and coasts runoff plays a crucial role in the dynamics of the ocean system. However, the current emphasis on synthetic aperture radar (SAR) retrieved current velocities has been on studying strong boundary currents and large-scale ocean circulation, resulting in relatively limited observational research on estuarine current velocities. This study aims to observe the current velocities in the Yangtze River estuary area using Sentinel-1A SAR data. The findings indicate a high agreement between the current velocities retrieved by Sentinel-1A and those observed at hydrographic stations, showcasing the characteristics of current velocity distribution associated with seafloor topography. This highlights the significant potential of SAR in observing estuarine current velocities, contributing to the analysis of benthic topography and the statistics of river runoff.

Synergistic use of SAR satellites with deep learning model interpolation for investigating of active landslides in Cuenca, Ecuador

Among the most intense geological disasters, landslides frequently occur throughout the world. These phenomena have been studied using space geodetic techniques, including Global Navigation Satellite Systems (GNSS) and Multi-Temporal Interferometric Synthetic Aperture Radars (MT-InSAR). Nevertheless, complete mapping and analysis of landslides’ surface deformation in most areas can be complicated due to a large diversity in kinematics, such as periods of quiescence and acceleration in the toe and crown. One of these landslides is the Cuenca landslides in Ecuador, where the geological investigation revealed that the toe of the landslides was located in urban areas, with more noticeable deformation effects. In contrast, its crown was located mainly in a rural and green land area. In this study, we show the potential of a synergistic use of COSMO-SkyMed (CSK) and Sentinel-1A (S1A) synthetic aperture radar (SAR) data for comprehensively monitoring the Cuenca landslides. To this aim, we have used Long-Short Term Memory (LSTM) and Convolutional Neural Networks (CNN) as two different Deep Learning Algorithms (DLAs) to integrate results in the temporal and spatial domain, respectively. A cross-comparison of the results was made with the nine GPS-derived deformations and the visual effects (i.e. crack width and pattern) on the field. This validation against GPS observation reveals that the RMSEs of the final MT-InSAR-derived velocity after applying the synergic double band SAR dataset decrease at more than 73% of nine GPS stations. Highlights Synergic MT-InSAR approach for studying landslide deformation in diverse kinematic areas. Utilized DLAs (LSTM and CNNs) for effective temporal and spatial interpolation of InSAR results. Findings emphasize the potential of multi-sensor SAR and DLAs for landslide monitoring regarding improving the RMSE at nine stations with an average of 73%.

Mapping Ratoon Rice Fields Based on SAR Time Series and Phenology Data in Cloudy Regions

Ratoon rice (RR) has emerged as an active adaptation to climate uncertainty, stabilizing total paddy rice yield and effectively reducing agriculture-related ecological environmental issues. However, identifying key remote sensing parameters for RR under cloudy and foggy conditions is challenging, and existing RR monitoring methods in these regions face significant uncertainties. Here, given the sensitivity of synthetic aperture radar (SAR) backscattering signals to the crop phenological period, this paper introduces a threshold model utilizing Sentinel-1A SAR data and phenological information for mapping RR. The Yongchuan District of Chongqing, which is often cloudy and foggy, was selected as a specific study region where VH-polarized backscatter coefficients of Sentinel-1 images were obtained at 10 m spatial resolution in 2020. Based on the proposed threshold model, the RR extraction overall accuracy was up to 90.24%, F1 score was 0.92, and Kappa coefficient was 0.80. Further analysis showed that the extracted RR boundaries exhibited high consistency with true Sentinel-2 remote sensing images and the RR extracted area was in good agreement with the actual planted area situation. This threshold model demonstrated good applicability in the studied cloudy and foggy region, and successfully distinguished RR from other paddy rice types. The methodological framework established in this study provides a basis for extensive application in China and other significant RR-producing regions globally.

An Advanced Quality Assessment and Monitoring of ESA Sentinel-1 SAR Products via the CyCLOPS Infrastructure in the Southeastern Mediterranean Region

The Cyprus Continuously Operating Natural Hazards Monitoring and Prevention System, abbreviated CyCLOPS, is a national strategic research infrastructure devoted to systematically studying geohazards in Cyprus and the Eastern Mediterranean, Middle East, and North Africa (EMMENA) region. Amongst others, CyCLOPS comprises six permanent sites, each housing a Tier-1 GNSS reference station co-located with two calibration-grade corner reflectors (CRs). The latter are strategically positioned to account for both the ascending and descending tracks of SAR satellite missions, including the ESA’s Sentinel-1. As of June 2021, CyCLOPS has reached full operational capacity and plays a crucial role in monitoring the geodynamic regime within the southeastern Mediterranean area. Additionally, it actively tracks landslides occurring in the western part of Cyprus. Although CyCLOPS primarily concentrates on geohazard monitoring, its infrastructure is also configured to facilitate the radiometric calibration and geometric validation of Synthetic Aperture Radar (SAR) imagery. Consequently, this study evaluates the performance of Sentinel-1A SAR by exploiting the CyCLOPS network to determine key parameters including spatial resolution, sidelobe levels, Radar Cross-Section (RCS), Signal-to-Clutter Ratio (SCR), phase stability, and localization accuracy, through Point Target Analysis (PTA). The findings reveal the effectiveness of the CyCLOPS infrastructure to maintain high-quality radiometric parameters in SAR imagery, with consistent spatial resolution, controlled sidelobe levels, and reliable RCS and SCR values that closely adhere to theoretical expectations. With over two years of operational data, these findings enhance the understanding of Sentinel-1 SAR product quality and affirm CyCLOPS infrastructure’s reliability.

Detection and monitoring of mining-induced subsidence with Gaofen-3 and Sentinel-1A SAR datasets

ABSTRACT The mining-induced subsidence has caused severe casualties and poses a threat to the safe operation of coal extraction, especially under the roads, buildings and rivers. This alarming situation demands the large-area subsidence detection and the large-gradient deformation monitoring to ensure safety. Although the traditional Interferometric Synthetic Aperture Radar (InSAR) technology can monitor surface deformation in high accuracy, unfortunately, it often fails to recover the sufficient deformation when facing the effects of decorrelation and large gradient displacement. Therefore, this study firstly proposes a Stacking-InSAR and coherence gradient fusion method to detect the mining-induced subsidence over large area. Secondly, the robust offset tracking method is applied to retrieve the large-gradient deformation in Buertai mining site. Both Gaofen-3 and Sentinel-1A Synthetic Aperture Radar (SAR) datasets are acquired over the border of Shaanxi and Inner Mongolia, China, for experimental test. The results show that 179 mining-induced subsidence sites are detected with the proposed method, and the complete spatial temporal deformation fields are obtained by both Gaofen-3 and Sentinel-1A offset tracking method with maximum deformation of 3 m in range direction. This study provides a valuable tool for detecting the fine boundary of mining subsidence and recovering the accurate deformation time series.

Automated Extraction of Textural Features From Segmented Sentinel-1ASynthetic Aperture Radar Satellite Image Using Grey Level Co-Occurrence Matrix

This paper focuses on the automated extraction of textural features from segmented Sentinel-1A Synthetic Aperture Radar (SAR) imagery, which was captured on 25th August 2017. Textural properties play a pivotal role in discerning regions of interest in satellite imagery, essential for precise classification in diverse applications, including remote sensing, medical image analysis, biometric analysis, and document image analysis. To achieve this, the Local Adaptive Threshold (LAT) technique was employed for segmenting the foreground and background areas of the pre-processed vertical transmit-vertical receive (VV) polarized Sentinel-1A SAR data. The approach relies on second-order statistics, specifically co-occurrence measures, which evaluate the relationships between pairs of pixels within their local neighbourhood’s in the input image. Four fundamental second-order statistical measures, namely correlation, energy, homogeneity, and contrast, were computed using the Grey Level Co-occurrence Matrix (GLCM). The GLCM was generated by quantizing the foreground region of the segmented grey-scale image into different levels, including 8, 16, and 64 grey levels, considering two inter-pixel distances, d=1 and d=2, and adopting an omni-directional orientation (θ = 00, 450, 900, 1350).The results underscore the effectiveness of the GLCM-based approach in computing second-order statistical texture measures from the segmented SAR image. Notably, the findings highlight that quantizing the image to Ng=8 with d=1 and considering omni-directional statistical measures (θ = 00, 450, 900, 1350) significantly enhances performance.

Fusion of sentinel-1 SAR and sentinel-2 MSI data for accurate Urban land use-land cover classification in Gondar City, Ethiopia

Effective urban planning and management rely on accurate land cover mapping, which can be achieved through the combination of remote sensing data and machine learning algorithms. This study aimed to explore and demonstrate the potential benefits of integrating Sentinel-1 SAR and Sentinel-2 MSI satellite imagery for urban land cover classification in Gondar city, Ethiopia. Synthetic Aperture Radar (SAR) data from Sentinel-1A and Multispectral Instrument (MSI) data from Sentinel-2B for the year 2023 were utilized for this research work. Support Vector Machine (SVM) and Random Forest (RF) machine learning algorithms were utilized for the classification process. Google Earth Engine (GEE) was used for the processing, classification, and validation of the remote sensing data. The findings of the research provided valuable insights into the performance evaluation of the Support Vector Machine (SVM) and Random Forest (RF) algorithms for image classification using different datasets, namely Sentinel 2B Multispectral Instrument (MSI) and Sentinel 1A Synthetic Aperture Radar (SAR) data. When applied to the Sentinel 2B MSI dataset, both SVM and RF achieved an overall accuracy (OA) of 0.69, with a moderate level of agreement indicated by the Kappa score of 0.357. For the Sentinel 1A SAR data, SVM maintained the same OA of 0.69 but showed an improved Kappa score of 0.67, indicating its suitability for SAR image classification. In contrast, RF achieved a slightly lower OA of 0.66 with Sentinel 1A SAR data. However, when the datasets of Sentinel 2B MSI and Sentinel 1A SAR were combined, SVM achieved an impressive OA of 0.91 with a high Kappa score of 0.80, while RF achieved an OA of 0.81 with a Kappa score of 0.809. These findings highlight the potential of fusing satellite data from multiple sources to enhance the accuracy and effectiveness of image classification algorithms, making them valuable tools for various applications, including land use mapping and environmental monitoring.

Tracking tsunami propagation and Island’s collapse after the Hunga Tonga Hunga Ha’apai 2022 volcanic eruption from multi-space observations

The quantity and accuracy of satellite-geodetic measurements have increased over time, revolutionizing the monitoring of tectonic processes. Global Navigation Satellite System (GNSS) and satellite radar signals provide observations beyond ground deformation, including how earthquake and tsunami processes affect variations in the ionosphere. Here, we study the Hunga Tonga Hunga Ha’apai (HTHH) volcanic eruption 2022 and its associated tsunami propagation with the analysis GNSS derived Total Electron Content (TEC), Synthetic Aperture Radar (SAR) Sentinel-1 data, complemented with tide gauge observations. We utilize GNSS sites data within a ~ 5000 km radius from the volcanic eruption for estimating the ionospheric perturbation as Vertical TEC. We give evidence on the detection of acoustic gravity, internal gravity, and atmospheric Lamb waves signatures in the TEC perturbation. In particular, the internal gravity waves that concentrated in the southwest of Tonga, directly correlates with the observed tsunami propagation direction as accounted by the tide gauge measurements. However, the acoustic gravity wave signature in the TEC is dominant in the north direction suggesting a surface deformation, which could be verified using Sentinel-1A SAR amplitude data. The analysis presented herein shows that within 5 h of the volcanic eruption, the central part of the HTHH island landscape disappeared with the biggest explosion. The unprecedented detail resolved by integrating satellite data yields previously unknown details of the deformation of the 2022 HTHH volcano eruption.

Impact of SAR-based vegetation attributes on the SMAP high-resolution soil moisture product

NASA's Soil Moisture Active Passive (SMAP) mission employs L-band radiometer observations with a spatial resolution of ~33 km to produce global soil moisture products at various spatial resolutions and regular intervals. Among these products is the SMAP-Sentinel active-passive high-resolution soil moisture, available at spatial resolutions of 1 km and 3 km. The SMAP-Sentinel product combines data from the SMAP L-band radiometer and the Copernicus Sentinel-1A and Sentinel-1B C-band Synthetic Aperture Radar (SAR) observations. The SMAP-Sentinel high-resolution soil moisture product enables various hydrological and agricultural applications that were previously beyond the capability of coarser resolution radiometer-only products. However, the assessment of SMAP-Sentinel product over SMAP core validation sites (CVS) has revealed high errors and biases, particularly in agricultural regions. We suspect that the discrepancy may be related to the vegetation attribute parameters (e.g., vegetation optical depth, τ) used in the SMAP-Sentinel baseline Single-Channel Algorithm (SCA) microwave emission model (i.e., τ-ω model) for soil moisture retrievals, which might be out of sync due to reliance on NDVI climatology in the τ estimation. In this study, we leveraged and evaluated the advantage of using vegetation attribute information inherently available through the SMAP ratiometer and Sentinel- 1A/1B SAR backscatter as an alternative to NDVI climatology-derived τ in the SCA to enhance soil moisture retrievals. The SAR cross-polarization (vh) backscatter potentially contains vegetation attribute information that can be correlated with the coarse resolution τ derived from SMAP observations. Therefore, we hypothesized that including Sentinel- 1A/1B vh observations to derive high-resolution (1 km and 3 km) τ, and subsequently incorporating it into the SCA, can significantly improve the accuracy of the SMAP-Sentinel soil moisture product. We established a statistical relationship between Sentinel- 1A/1B SAR backscatter and the SMAP Dual-Channel Algorithm (DCA)-based τ retrieval at ~33 km, for a specific region and different landcover classes at a global extent, using seven years of observations (2015–2022). Further, we developed an algorithm to downscale the SMAP DCA τ of ~33 km to 1 km and 3 km resolutions based on the developed statistical relationships, enabling us to capture the actual temporal variability in vegetation attributes (i.e., τ) at higher resolutions. The downscaled τ values were then incorporated into the SCA to retrieve SMAP-Sentinel high-resolution (1 km and 3 km) soil moisture. The validation analysis of high-resolution soil moisture retrievals demonstrates that the inclusion of the downscaled τ approach significantly improved the performance of the SMAP-Sentinel product by reducing both bias and error standard deviation compared to the retrievals based on τ derived from the NDVI climatology.

Mapping Rice Area in the Cauvery Delta Zone of Tamil Nadu Using Sentinel 1A Synthetic Aperture Radar (SAR) Data

Since remote sensing based crop inventory provides accurate and timely information as compared to the conventional survey methods of estimating area, Multi-temporal Sentinel 1A Synthetic Aperture Radar data was used for the estimation of rice area during Samba season 2022 in the Cauvery delta zone comprising Thanjavur, Thiruvaru, Mayiladuthurai, and Nagapatnam districts of Tamil Nadu. SAR data was preferred over optical satellite data due to excess cloud cover during cropping the major season in Tamil Nadu. Temporal back-scatter (dB) signature of rice crop was generated from the multi-temporal processed SAR data utilizing the modules of a fully automated MAPscape software aiding the discriminating of the crop from others. The signatures revealed that the dB levels to be the lowest during agronomic floods, reached the highest during maximum tillering stage and started declining thereafter. Multi-temporal feature extraction module of Mapscape was used to estimate rice area and validated for accuracy using ground truth data collected during survey. A total of 3.05 lakh ha of rice area was estimated with an overall accuracy of 90.8 % and 0.82 kappa coefficient. Largest area of 1.12 lakh ha was recorded in Thanjavur followed by Thiruvarur and Mayiladuthurai with 0.95 and 0.51 lakh ha respectively.

MACHINE LEARNING UNET-BASED SEGMENTATION OF SENTINEL-1A SATELLITE IMAGES

Image segmentation is an area of study which has generated a lot of interest in engineering in the last few years. Segmentation of Sentinel 1A satellite images is an area that is not as common as that of its optical counterparts as a result of the appearance of its images. The focus of this paper is the segmentation of Sentinel-1A Synthetic Aperture Radar (SAR) satellite images using UNet, a Deep learning technique designed specifically for semantic segmentation of images. The model was developed with Python programming language in keras, which is a deep learning Python API running on the Tensorflow framework. The satellite images were acquired from the European Space Agency’s (ESA) Copernicus satellite acquisition hub. The performance evaluation was carried out with Jaccard Index or Intesection over Union (IoU) and Accuracy metrics. The IoU metric measures the number of pixels common between the target and prediction masks divided by the total number of pixels present across both masks while the Accuracy metric reports the percent of pixels in the image which were correctly classified. The results recorded in the evaluation had IoU as 0.42 and an Accuracy of 95%. This show a high performance of the model in segmenting images.

The Precursory 3D Displacement Patterns and Their Implicit Collapse Mechanism of the Ice-Rock Avalanche Events Occurred in Sedongpu Basin Revealed by Optical and SAR Observations

Quantifying the kinematic evolution patterns of mountain glaciers near Yarlung Tsanpo River performs a major role in evaluating the glacial instability and the secondary disasters. For the Sedongpu Basin near the Yarlung Tsanpo River Valley, the dramatic geomorphic landscape variations triggered by the ice-rock avalanche events were visually identified as the dominant texture deficiencies in time-series optical images. To improve the image correlation quality broken by these image texture deficiencies, the Landsat-8/Sentinel-2 optical images were divided into different groups, then a stepwise combination strategy was innovatively proposed to derive the glacier time-series displacement velocities in different temporal stages. The standard deviations (STD) of the optical measurements in the stable area maintained around 0.04 m/yr for the normalized displacement velocity and maintained from 0.6 to 1.7 m for the cumulative displacement time series. The obvious variations in glacier displacement velocity were identified before each collapse event. Subsequently, the offset-tracking procedures were performed on 7 Sentinel-1A Synthetic Aperture Radar (SAR) images to acquire the range and azimuth displacement velocities. To better reveal the dynamic mechanism of the glacier activity, the three-dimensional (3D) glacial displacement velocity was also derived by using optical and SAR results. The precipitation, temperature, and seismic activities were assumed as the main triggering factors of controlling the glacial dynamic mechanism and final collapse events. Additionally, the dynamic mechanism of the active glaciers in Sedongpu Basin conformed to a power law, which was limited by the changes of the internal stress friction force on the sliding base surface. The aim of this study is to shed a light on interpreting the precursory displacement patterns and their implicit failure mechanism of these ice-rock avalanche events with the conventional freely optical and SAR observations.

Near real-time mapping of jute (Corchorus sp.) area using multi-temporal Sentinel-1 intensity data over the central part of West Bengal, India

ABSTRACT The present study aimed to develop a methodology to estimate jute (Corchorus sp.) cultivated areas in the Indian subcontinent using sentinel-1A Synthetic aperture radar (SAR) C band dual polarization data. We developed a hierarchical decision tree (DT) model to estimate jute area and compared it with other supervised and unsupervised methods. The proposed DT model showed improved overall accuracy of 82–83% (Kappa coefficient, k = 0.78–0.79) over supervised or unsupervised methods and validated against Directorate of Economics and Statistics (DES) data, Government of India during 2016–2019. The VV backscatter record improved accuracy over VH backscatter.

Local Climate Zone Classification by Seasonal and Diurnal Satellite Observations: An Integration of Daytime Thermal Infrared Multispectral Imageries and High-Resolution Night-Time Light Data

Accurate, rapid, and automatic local climate zone (LCZ) mapping is essential for urban climatology and studies in terms of urban heat islands. Remotely sensed imageries incorporated with machine learning algorithms are widely utilized in LCZ labeling. Nevertheless, large-scale LCZ mapping is still challenging due to the complex vertical structure of underlying urban surfaces. This study proposed a new method of LCZ labeling that uses a random forest classifier and multi-source remotely sensed data, including Sentinel 1A Synthetic Aperture Radar (SAR), Sentinel 2 Multispectral Instrument, and Luojia1-01 night-time light data. In particular, leaf-on and -off imageries and surface thermal dynamics were utilized to enhance LCZ labeling. Additionally, we systematically evaluated how daytime and night-time features influence the performance of the classification procedure. Upon examination, the results for Beijing, China, were confirmed to be robust and refined; the Overall Accuracy (OA) value of the proposed method was 88.86%. The accuracy of LCZs 1–9 was considerably increased when using the land surface temperature feature. Among these, the Producer Accuracy (PA) value of LCZ 3 (compact low-rise) significantly increased by 16.10%. Notably, it was found that NTL largely contributed to the classification concerning LCZ 3 (compact low-rise) and LCZ A/B (dense trees). The performance of integrating leaf-on and -off imageries for LCZ labeling was better than merely uses of leaf-on or -off imageries (the OA value increased by 4.75% compared with the single use of leaf-on imagery and by 3.62% with that of leaf-off imagery). Future studies that use social media big data and Very-High-Resolution imageries are required for LCZ mapping. This study shows that combining multispectral, SAR, and night-time light data can improve the performance of the random forest classifier in general, as these data sources capture significant information about surface roughness, surface thermal feature, and night-time features. Moreover, it is found that incorporating both leaf-on and leaf-off remotely sensed imageries can improve LCZ mapping.

Synthetic Aperture Radar Remote Sensing for Crop Classification

This Study proposes the approach for crop classification using the Grey Level Co-occurrence Matrix feature of Synthetic Aperture Radar (SAR) images. The method utilizes the SAR Images acquired by Sentinel 1A SAR Data and extract textural features using GLCM. In this study, we investigate the potential of Grey Level Co-occurrence Matrix (GLCM)-based texture information for horticulture crop classification with SAR images in Kharif and cloud weather condition. A study on Synthetic Aperture Radar (SAR) satellite imagery was conducted in Chhattisgarh with the objective to evaluate the potential of different texture parameters among crop. The SAR data were pre-processed for textural analysis having entire angle and equal distance quantization. The results were categorized among different parameters showing significant variation for horticulture crops for Contrast, Dissimilarity, Homogeneity, ASM, Energy, Entropy and GLCM Mean. The statistical analysis was done for fruit crop along with major kharif crop of study area. The results shows that mean backscatter value was lowest for banana (99.12 dB) and highest for Mango (198.26 dB) regarding contrast textural property in VH Channel whereas mean backscatter value in VH Channel w.r.t to energy was maximum for banana (0.60 dB) followed by papaya (0.49 dB) and guava (0.45 dB) and least for mango (0.44 dB). The mean backscatter value for GLCM mean textural property in VH channel was shown maximum by banana (51.24 dB) followed by papaya (41.96 dB) and mango (32.98 dB). These results indicate the usefulness of texture information for classification of SAR images, particularly when acquisition of optical images is difficult in Kharif and cloud weather condition for crop classification. Thus GLCM feature of SAR Data proven to be significant for the classification of horticulture crops.

HA-Unet: A Modified Unet Based on Hybrid Attention for Urban Water Extraction in SAR Images

Urban water plays a significant role in the urban ecosystem, but urban water extraction is still a challenging task in automatic interpretation of synthetic aperture radar (SAR) images. The influence of radar shadows and strong scatters in urban areas may lead to misclassification in urban water extraction. Nevertheless, the local features captured by convolutional layers in Convolutional Neural Networks (CNNs) are generally redundant and cannot make effective use of global information to guide the prediction of water pixels. To effectively emphasize the identifiable water characteristics and fully exploit the global information of SAR images, a modified Unet based on hybrid attention mechanism is proposed to improve the performance of urban water extraction in this paper. Considering the feature extraction ability and the global modeling capability in SAR image segmentation, the Channel and Spatial Attention Module (CSAM) and the Multi-head Self-Attention Block (MSAB) are both introduced into the proposed Hybrid Attention Unet (HA-Unet). In this work, Resnet50 is adopted as the backbone of HA-Unet to extract multi-level features of SAR images. During the feature extraction process, CSAM based on local attention is adopted to enhance the meaningful water features and ignore unnecessary features adaptively in feature maps of two shallow layers. In the last two layers of the backbone, MSAB is introduced to capture the global information of SAR images to generate global attention. In addition, two global attention maps generated by MSAB are aggregated together to reconstruct the spatial feature relationship of SAR images from high-resolution feature maps. The experimental results on Sentinel-1A SAR images show that the proposed urban water extraction method has a strong ability to extract water bodies in the complex urban areas. The ablation experiment and visualization results vividly indicate that both CSAM and MSAB contribute significantly to extracting urban water accurately and effectively.

Improving Typical Urban Land-Use Classification with Active-Passive Remote Sensing and Multi-Attention Modules Hybrid Network: A Case Study of Qibin District, Henan, China

The study of high-precision land-use classification is essential for the sustainable development of land resources. This study addresses the problem of classification errors in optical remote-sensing images under high surface humidity, cloud cover, and hazy weather. The synthetic aperture radar (SAR) images are sensitive to soil moisture, and the microwave can penetrate clouds, haze, and smoke. By using both the active and passive remote-sensing data, the Sentinel-1A SAR and Sentinel-2B multispectral (MS) images are combined synergistically. The full-band data combining the SAR + MS + spectral indexes is thus constructed. Based on the high dimensionality and heterogeneity of this data set, a new framework (MAM-HybridNet) based on two-dimensional (2D) and three-dimensional (3D) hybrid convolutional neural networks combined with multi-attention modules (MAMs) is proposed for improving the accuracy of land-use classification in cities with high surface humidity. In addition, the same training samples supported by All bands data (SAR + MS + spectral index) are selected and compared with k-Nearest Neighbors (KNN), support vector machine (SVM), 2D convolutional neural networks, 3D convolutional neural networks, and hybridSN classification models to verify the accuracy of the proposed classification model. The results show that (1) fusion classification based on Sentinel-2B MSI and Sentinel-1A SAR data produce an overall accuracy (OA) of 95.10%, a kappa coefficient (KC) of 0.93, and an average accuracy (AA) of 92.86%, which is better than the classification results using Sentinel-2B MSI and Sentinel-1A SAR images separately. (2) The classification accuracy improves upon adding the spectral index, and the OA, KC, and AA improve by 3.77%, 0.05, and 5.5%, respectively. (3) With the support of full-band data, the algorithm proposed herein produces better results than other classification algorithms, with an OA of 98.87%, a KC of 0.98, and an AA of 98.36%. These results indicate that the synergistic effect of active-passive remote-sensing data improves land-use classification. Additionally, the results verify the effectiveness of the proposed deep-learning classification model for land-use classification.

Flood Detection in Dual-Polarization SAR Images Based on Multi-Scale Deeplab Model

The proliferation of massive polarimetric Synthetic Aperture Radar (SAR) data helps promote the development of SAR image interpretation. Due to the advantages of powerful feature extraction capability and strong adaptability for different tasks, deep learning has been adopted in the work of SAR image interpretation and has achieved good results. However, most deep learning methods only employ single-polarization SAR images and ignore the water features embedded in multi-polarization SAR images. To fully exploit the dual-polarization SAR data and multi-scale features of SAR images, an effective flood detection method for SAR images is proposed in this paper. In the proposed flood detection method, a powerful Multi-Scale Deeplab (MS-Deeplab) model is constructed based on the dual-channel MobileNetV2 backbone and the classic DeeplabV3+ architecture to improve the ability of water feature extraction in SAR images. Firstly, the dual-channel feature extraction backbone based on the lightweight MobileNetV2 separately trains the dual-polarization SAR images, and the obtained training parameters are merged with the linear weighting to fuse dual-polarization water features. Given the multi-scale space information in SAR images, then, a multi-scale feature fusion module is introduced to effectively utilize multi-layer features and contextual information, which enhances the representation of water features. Finally, a joint loss function is constructed based on cross-entropy and a dice coefficient to deal with the imbalanced categorical distribution in the training dataset. The experimental results on the time series of Sentinel-1A SAR images show that the proposed method for flood detection has a strong ability to locate water boundaries and tiny water bodies in complex scenes. In terms of quantitative assessment, MS-Deeplab can achieve a better performance compared with other mainstream semantic segmentation models, including PSPNet, Unet and the original DeeplabV3+ model, with a 3.27% intersection over union (IoU) and 1.69% pixel accuracy (PA) improvement than the original DeeplabV3+ model.