Satellite-based Synthetic Aperture Radar Research Articles

SwinYOLOv7: Robust ship detection in complex synthetic aperture radar images

Using satellite-based SAR (Synthetic Aperture Radar) imagery to detect and track ships is a formidable challenge. However, accurate analysis is hampered by inherent difficulties such as obscured edges, multiple targets, varying dimensions and complex backgrounds. These factors contribute to a poor signal-to-noise ratio. Various artificial intelligence models have been developed to improve these problems, especially with YOLO-based models. To overcome the above challenges and achieve robust performance, this study presents a model called SwinYOLOv7, a novel fusion of the YOLOv7 framework with the Features Pyramid Network and the Swin Transformer using ground-breaking anchor-free detection algorithms. This innovative method aims to increase the accuracy of vessel detection while reducing the impact of background clutter. The proposed model improved by YOLOv7 examined three different datasets in detail: SRSDD-v1.0, HRSID, and SSDD to optimize the performance of the model. The training process was consistently verified using superior recall, precision, and F1-score values, which can be easily compared with previous studies. The results show that the model using the Swin Transformer attention mechanism and using an image size of 640×640 achieves the highest accuracy of 96.59%. Alternative attention mechanisms, including the Squeeze-and-Excitation Network (SEnet), the Convolutional Block Attention Module (CBAM), Channel Attention (CA) and Efficient Channel Attention (ECA), deliver poorer accuracy rates. The combination of YOLOv7 and Swin Transformer yielded encouraging results that enabled the proposed model to outperform the current benchmarking models. Therefore, the proposed model provides a compelling solution to ensure accurate vessel identification in complex search and rescue scenarios.

The potential of optical and SAR time-series data for the improvement of aboveground biomass carbon estimation in Southwestern China’s evergreen coniferous forests

ABSTRACT Accurate assessments of forest biomass carbon are invaluable for managing forest resources, evaluating effects on ecological protection, and achieving goals related to climate change and sustainable development. Currently, the integration of optical and synthetic aperture radar (SAR) data has been extensively utilized in estimating forest aboveground biomass carbon (AGC), while it is limited by using single-phase remote sensing images. Time-series data, which capture the interannual dynamic growth and seasonal variations of photosynthetic phenology in forests, can sufficiently describe forest growth characteristics. However, there remains a gap in research focusing on utilizing satellite-based time-series data for AGC estimation, especially for SAR sensors. This study investigated the potential of satellite-based optical and SAR time-series data for estimating AGC. Here, we undertook nine quantitative experiments of AGC estimation from Landsat 8 and Sentinel-1 and tested several regression algorithms (including multiple linear regression (MLR), random forests (RF), artificial neural network (ANN), and extreme gradient boosting (XGBoost)) to explore the contributions of spatiotemporal features to AGC estimation. The results suggested that the XGBoost algorithm was suitable for AGC estimation with explanatory solid power and stable performance. The temporal features representing forest growth trends and periodic change characteristics (such as coefficients of continuous wavelet transform) were more valuable for AGC estimation than spatial features for both sensor types, accounting for around 40% ~50% of the variance compared to 17% ~25%. The combination of optical and SAR time-series data produced the best performance (R2 = 0.814, RMSE = 18.789 Mg C/ha, rRMSE = 26.235%), compared with when utilizing optical or SAR time-series data alone (optical: R2 of 0.657 and rRMSE of 35.317%; SAR: R2 of 0.672 and rRMSE of 34.701%). Feature importance analysis also verified that temporal features of optical vegetation indices, SWIR 1/2 bands, and SAR backscatter from VV polarization were the most critical variables for AGC estimation. Furthermore, incorporating temporal features into the modeling is illustrated to be effective in reducing saturation effects within high-biomass forests. This study demonstrated the superiority of time-series data for forest carbon estimation. While the applicability of this methodology has only been investigated in evergreen coniferous forests, it may provide a viable approach needed to make full use of increasingly better and free satellite time-series data to estimate forest AGC with high accuracy, supporting policy making of forest management and sustainable development.

Finding Navigable Paths through Tidal Flats with Synthetic Aperture Radar

Tidal flats are some of the most dynamic coastal environments in the world, where traditional coastal mapping and monitoring provide insufficient temporal resolution to reliably map channels and sand flats. Satellite-based Synthetic Aperture Radar (SAR) enables regular cloud-penetrating detection of water flowing through channels within the tidal flats, referred to as tidal channels. This paper presents a method for detecting a path through tidal channels, using satellite imagery, that supports our understanding and safe exploitation of this valuable coastal environment. This approach is the first proposed to identify navigable paths in all conditions, with SAR images susceptible to variation due to weather and tidal conditions. Tidal channels are known to vary in SAR presentation, and we find that tidal flat presentation is also influenced by conditions. The most influential factor is the wind, with high winds causing an inversion in how both tidal flats and tidal channels present in SAR images. The presented method for the automatic detection of tidal channels accounts for this variability by using previous channel paths as a reference to reliably correct imagery and detect the latest path. The final algorithm produces paths with minor errors in 17.6% of images; the error rate increases to 71.7%, with an almost tenfold increase in errors, when the SAR image and paths are not adjusted to account for conditions. This capability has been used to support the Nith Inshore Rescue in attending call-outs from their base in Glencaple, UK, while the insights from monitoring tidal channels for a year demonstrate how periods of high river flow preceded major changes in the channel path.

A Lightweight SAR Image Ship Detection Method Based on Improved Convolution and YOLOv7

The airborne and satellite-based synthetic aperture radar enables the acquisition of high-resolution SAR oceanographic images in which even the outlines of ships can be identified. The detection of ship targets from SAR images has a wide range of applications. Due to the density of ships in SAR images, the extreme imbalance between foreground and background clutter, and the diversity of target sizes, achieving lightweight and highly accurate multi-scale ship target detection remains a great challenge. To this end, this paper proposed an attention mechanism for multi-scale receptive fields convolution block (AMMRF). AMMRF not only makes full use of the location information of the feature map to accurately capture the regions in the feature map that are useful for detection results, but also effectively captures the relationship between the feature map channels, so as to better learn the relationship between the ship and the background. Based on this, a new YOLOv7-based ship target detection method, You Only Look Once SAR Ship Identification (YOLO-SARSI), was proposed, which acquires the abstract semantic information extracted from the high-level convolution while retaining the detailed semantic information extracted from the low-level convolution. Compared to the deep learning detection methods proposed by previous authors, our method is more lightweight, only 18.43 M. We examined the effectiveness of our method on two SAR image public datasets: the High-Resolution SAR Images Dataset (HRSID) and the Large-Scale SAR Ship Detection Dataset-v1.0 (LS-SSDD-V1.0). The results show that the average accuracy (AP50) of the detection method YOLO-SARSI proposed in this paper on the HRSID and LS-SSDD-V1.0 datasets is 2.6% and 3.9% higher than that of YOLOv7, respectively.

Kinematic Evolution of a Large Paraglacial Landslide in the Barry Arm Fjord of Alaska

AbstractOur warming climate is adversely affecting cryospheric landscapes via glacial retreat, permafrost degradation, and associated slope destabilization. In Prince William Sound, Alaska, the rapid retreat of Barry Glacier has destabilized the slopes flanking the glacier, resulting in numerous landslides. The largest of these landslides (∼500 Mm3 in volume) is more than 2 km wide and has the potential to generate a tsunami that could affect nearby recreationists, marine traffic, infrastructure, natural and cultural resources, and the community of Whittier, located 60 km from the landslide. Here, we combine landslide structural and kinematic element mapping with data acquired from bi‐yearly airborne lidar, multi‐week satellite‐based synthetic aperture radar (SAR), sub‐hourly ground‐based SAR, and seismic monitoring from 2020 to 2022 to characterize this landslide and examine its evolution. While some methods serve as a snapshot in time that is a culmination of events, others emphasize the ever‐evolving nature of the landslide and associated hazards. Four major kinematic elements define the overall structure of the landslide, which vary in deformation type and rate, from creep (5 mm per day over several months) to episodic movement (2 m in 30 days) and landslide‐wide to localized events. In some areas of the landslide, short‐term deformation deviates from structures formed by cumulative movement, implying structural and kinematic evolution associated with glacier retreat. These insights are important for assessing landslide hazards and hazard evolution for large, slow‐moving bedrock landslides in actively deglaciating environments.

Assessment and mapping of landslides in steep mountainous terrain using PS-InSAR: A case study of Karimabad Valley in Chitral

Assessing and mapping landslides in rugged, mountainous terrain presents considerable challenges due to the high cost, risk, and complexity involved. High-resolution assessments of landslides are essential for mitigating landslide-relevant damage and understanding landslide kinematics. Recent studies have utilized satellite-based Synthetic Aperture Radar Interferometry (InSAR) to achieve millimeter-scale detection of ground surface deformation. Among these InSAR techniques, Persistent Scatterers Interferometry (PS-InSAR) is particularly useful in identifying landslide-prone areas, offering a means to map and detect slope movements and characterize the stability of potentially unstable slopes. This study focuses on the slow-moving unstable hillslopes of Karimabad Valley in Chitral. We conducted an analysis using Sentinel-1 SLC data over a three-year period, obtained in interferometric wide swath mode with vertical-vertical polarization. Our results, derived from both ascending and descending satellite passes, reveal that the hillslopes of Karimabad Valley exhibit an average velocity of −25 to +28 mm/yr. Based on the deformation pattern identified, we detected seven deformation zones, with particular concern for the Beshgram village area. To better understand terrain steepness, four topographic profile lines (AA', BB', CC', and DD') were extracted and correlated with the deformation time series. Furthermore, the deformation time series were examined in relation to rainfall, earthquake, and snowfall data. These analyses highlight the complex interplay of multiple factors contributing to the observed variations in deformation, including the region's topography characterized by steep slopes, the freezing and thawing of snow, and seismic events.

Real-time profile measurement method for a large-scale satellite antenna.

To improve the detection capability of satellite-based synthetic aperture radar, a large antenna array with a length scale of 100 meters is urgently needed. However, the structural deformation of the large antenna leads to phase errors, which will significantly reduce the antenna gain; hence, real-time and high-precision profile measurements of the antenna are essential for active compensation of the phase and thus improving the antenna gain. Nevertheless, the conditions of antenna in-orbit measurements are rather severe because of limited installation locations of measurement instruments, large areas, and long distance to be measured, and unstable measurement environments. To deal with the issues, we propose a three-dimensional displacement measurement method for the antenna plate based on laser distance measuring and digital image correlation (DIC). The proposed method uses the DIC method to retrieve the in-plane displacement information in combination with a laser range finder to provide depth information. A Scheimpflug camera is used to overcome the limitation of the depth of field of traditional cameras and enable clear imaging of the full field. Moreover, a vibration compensation scheme is proposed to eliminate the measurement error of the target displacement caused by the random vibration (within 0.01°) of the camera support rod. The results of the experiment in a laboratory setting show that the proposed method can effectively eliminate the measurement error caused by camera vibration (50mm) and reduce the displacement measurement error to within 1mm with a measurement range of 60m, which can meet the measurement requirements of next-generation large satellite antennas.

RETROSPECTIVE STUDY OF VERTICAL GROUND DEFORMATION IN COMO, NORTHERN ITALY: INTEGRATION OF LEVELLING AND PSI MEASUREMENTS

Abstract. Subsidence-related vertical ground deformation due to natural and anthropogenic factors may lead to considerable damages to structures and infrastructures, and may increase the occurrence probability of consequential events, such as floods, especially on sea and lake shores. Como city, placed in the north of Italy, adjacent to Como Lake, is subjected to significant subsidence phenomenon, which has been monitored by geodetic levelling networks. In this work the historical geodetic levelling measurements and satellite-based Synthetic Aperture Radar (SAR) data archive are integrated to assess the accuracy of Atmospheric Delay and Deformation Rate estimations obtained through Persistent Scatter Interferometry (PSI) techniques. Tree levelling measurement datasets acquired in 1990, 1997 and 2004 are used in order to obtain the precise deformation rate at the benchmarks for two periods of 1990–1997 and 1997–2004. For multi-temporal InSAR analysis, 106 SAR images (1992–2004) and 41 SAR images (1992–2004) in Ascending Track orbit from ERS-1/2 missions are used in this study. The assessment is performed through a statistical comparison between two sets of vertical land deformation rates obtained by integrated methods. The results of the validation represented a good consistency between deformation rates derived by both techniques. Also, this study has revealed the potential of SAR images acquired in gyro-less mode by ERS-2 mission (2001–2004) in terms of the estimation of ground deformation.

Highly Integrated Wideband Transmit/Receive Module for X-Band SAR Applications

The roadmaps of satellite-based synthetic aperture radar (SAR) systems show a trend that requires a continuous improvement of the active antenna in terms of operating bandwidth, scanning angle capability, and swath width, thus leading to the demand for increased RF power and better power efficiency of the RF transmitter. Moreover, compact size and light weight are relevant objectives for making the overall SAR instrument appealing for future applications. The transmit/receive module (TRM) shown in this paper was developed while combining all these requirements in a cost-effective approach. A careful design of all relevant RF interconnects and passive devices was performed to ensure the largest output power from the last high-power amplifier of the transmitting chain, the lowest noise figure at the input of the receiving chain, and the calibration capability for appropriately tuning the TX and RX signal. The TRM was manufactured and experimentally tested to verify its performances. The measurement results show the superior performances of the proposed compact high-power large-bandwidth TRM. The achieved target objectives make the design of the proposed TRM readily applicable for the development of a compact, high-power, and highly integrated AESAs to be used for next-generation satellite constellations for Earth observation.

Method for Frequent High Resolution of Optical Sensor Image Acquisition using Satellite-Based SAR Image for Disaster Mitigation

Method for frequent high resolution of optical sensor imagery data acquisition from satellite-based SAR (Synthetic Aperture Radar) image for disaster mitigation is proposed. The proposed method is based on Generative Adversarial Network: GAN-based super resolution and conversion method from a SAR imagery data to the corresponding optical sensor imagery data in order to increase observation frequency. Through experiments, it is found that it is possible to convert SAR imagery data to the corresponding optical sensor imagery data and also found that the spatial resolution of SAR imagery data is improved remarkably. Thus, initial stage of disaster (small scale of disaster) can be detected with resolution enhanced optical sensor imagery data derived from the corresponding SAR imagery data which results in prevention of secondary occurrence of relatively large scale of disaster. It is also found that 2.5 m of spatial resolution of optical sensor imagery data can be acquired every 2.5 days in the case that only Sentinel-1/SAR and Sentinel-2/MSI (Multi Spectral Imager) are used, for instance.

Revealing the global longline fleet with satellite radar

Because many vessels use the Automatic Identification System (AIS) to broadcast GPS positions, recent advances in satellite technology have enabled us to map global fishing activity. Understanding of human activity at sea, however, is limited because an unknown number of vessels do not broadcast AIS. Those vessels can be detected by satellite-based Synthetic Aperture Radar (SAR) imagery, but this technology has not yet been deployed at scale to estimate the size of fleets in the open ocean. Here we combine SAR and AIS for large-scale open ocean monitoring, developing methods to match vessels with AIS to vessels detected with SAR and estimate the number of non-broadcasting vessels. We reveal that, between September 2019 and January 2020, non-broadcasting vessels accounted for about 35% of the longline activity north of Madagascar and 10% of activity near French Polynesia and Kiribati’s Line Islands. We further demonstrate that this method could monitor half of the global longline activity with about 70 SAR images per week, allowing us to track human activity across the oceans.

Probabilistic SAR-based water segmentation with adapted Bayesian convolutional neural network

Geospatial resources, including satellite-based synthetic aperture radar (SAR) and optical data, have been instrumental in providing time-sensitive information about the extent and impact of natural hazard events, such as floods, to support emergency response and hazard management efforts. In effect, finite resources can be better optimized to support the needs of often extensively affected areas. However, the derivation of SAR-based flood information is not without its challenges and inaccurate flood detection can result in poor decision-making and non-trivial, adverse consequences. Reliable uncertainty quantification of flood extent estimates addresses this risk. In this context, our study presents the results of two probabilistic convolutional neural networks (CNNs) adapted for SAR-based water segmentation with freely available Sentinel-1 interferometric wide (IW) swath ground range detected (GRD) data. In particular, the performance of a variational inference-based Bayesian convolutional neural network (BCNN) is evaluated against that of a Monte Carlo dropout network (MCDN). In previous studies where uncertainty information has been generated along with segmentation results, MCDN has been more commonly applied as an approximation of Bayesian deep learning. Differences between the two approaches are highlighted with the application of a set of extended performance metrics. Both segmentation and uncertainty outputs are evaluated at data-, model-, tile- and scene-levels. The methods are demonstrated on the binary segmentation of a reference water dataset. We show that while different probabilistic techniques return comparable segmentation accuracies, they are differentiated based on their performance in assigning reliable probabilities. In particular, MCDNs characterized by more restrictive architectures generally lead to overconfident prediction intervals, whereas BCNNs have greater flexibility to learn the mean and the spread of the parameter posterior. Furthermore, we demonstrate that examining the (inaccurate, certain) metric is a better indicator of reliable uncertainty quantification and the BCNN is recommended to quantify uncertainties associated with SAR-based segmentation outputs. This information is especially valuable where the cost of inaccurate detections (false-positive and false-negative) is high. These preliminary findings highlight how the manner in which probabilities are properly assigned and their inclusion are instrumental and complementary to the production of flood masks and should be considered as a standard in the natural hazards domain.

Phase calibration for ideal wideband chirp in satellite-based synthetic aperture radar

With the development of launch vehicles, the number of satellite-based synthetic aperture radar (SAR) has been increasing. Satellite-based SAR can acquire high-resolution images at day and night regardless of the weather. Since SAR resolution is determined by a chirp bandwidth, corresponding a generation technology of wideband chirp is essentially required for high-resolution SAR images under consideration of relatively small and light weight SAR system. Therefore, an ideal chirp waveform generation, which is challenging task in actual hardware, is proposed by modifying a chirp equation for phase calibration, and is verified based on Sentinel-1 satellite SAR parameters. The proposed method is applied to our SAR hardware and we confirm the calibrated waveform. Evaluation results using SAR observation simulation show that the impulse response with the proposed method has better performances.

Improving above ground biomass estimates of Southern Africa dryland forests by combining Sentinel-1 SAR and Sentinel-2 multispectral imagery

Having the ability to make accurate assessments of above ground biomass (AGB) at high spatial resolution is invaluable for the management of dryland forest resources in areas at risk from deforestation, forest degradation pressure and climate change impacts. This study reports on the use of satellite-based synthetic-aperture radar (SAR) and multispectral imagery for estimating AGB by correlating satellite observations with ground truth data collected on forest plots from dryland forests in the Chobe National Park, Botswana. We undertook nineteen quantitative experiments with Sentinel-1 (S1), and Sentinel-2 (S2) and tested simple and multivariate regression including parametric (linear) and non-parametric (random forests) algorithms, to explore the optimal approaches for AGB estimation. The largest AGB value of 145 Mg/ha was found in northern Chobe while a large part of the study area (85%) is characterized by low AGB values (< 80 Mg/ha), with an average estimated at 51 Mg/ha. The results show that the AGB estimated using SAR backscatter values from vertical transmit receive (VV) polarization is more accurate than that based on horizontal receive (VH) polarization, accounting for 58% of the variance compared to 32%. Nevertheless, the combination of S1 SAR and S2 multispectral image data produced the best fit to the ground observations for dryland forests explaining 83% of the variance with an accuracy of 89%. Furthermore, the optimal AGB model performance was achieved with a random forest (RF) regression trees algorithm using S1 (SAR) and S2 (multispectral) image data (R2 = 0.95; RMSE = 0.25 Mg/ha). From the 11 vegetation indices tested, GNDVI, Normalized Difference Red Edge (NDRE1), and NDVI obtained the highest linear relationship with AGB (R2 = 0.71 and R2 = 0.56, p < 0.001), however, GNDVI and NDRE1 improved the AGB estimation at medium to high-density forests compared to NDVI. The GRVI and EVI were the least correlated with AGB (R2 = 0.09 and R2 = 0.31) at a significance level of p < 0.001, respectively. We show that NDVI saturates in areas with >80 Mg/ha AGB, whereas the inclusion of SAR backscatter and optical red edge bands (B5) significantly reduces saturation effects in areas of high biomass. GNDVI and red edge (B5) derived vegetation indices have more potential for estimating AGB in dryland forests than NDVI. Our results demonstrate that dryland AGB can be estimated with a reasonable level of precision from open access Earth observation data using multivariate random forest regression.

Integration of Satellite-Based Optical and Synthetic Aperture Radar Imagery to Estimate Winter Cover Crop Performance in Cereal Grasses

The magnitude of ecosystem services provided by winter cover crops is linked to their performance (i.e., biomass and associated nitrogen content, forage quality, and fractional ground cover), although few studies quantify these characteristics across the landscape. Remote sensing can produce landscape-level assessments of cover crop performance. However, commonly employed optical vegetation indices (VI) saturate, limiting their ability to measure high-biomass cover crops. Contemporary VIs that employ red-edge bands have been shown to be more robust to saturation issues. Additionally, synthetic aperture radar (SAR) data have been effective at estimating crop biophysical characteristics, although this has not been demonstrated on winter cover crops. We assessed the integration of optical (Sentinel-2) and SAR (Sentinel-1) imagery to estimate winter cover crops biomass across 27 fields over three winter–spring seasons (2018–2021) in Maryland. We used log-linear models to predict cover crop biomass as a function of 27 VIs and eight SAR metrics. Our results suggest that the integration of the normalized difference red-edge vegetation index (NDVI_RE1; employing Sentinel-2 bands 5 and 8A), combined with SAR interferometric (InSAR) coherence, best estimated the biomass of cereal grass cover crops. However, these results were season- and species-specific (R2 = 0.74, 0.81, and 0.34; RMSE = 1227, 793, and 776 kg ha−1, for wheat (Triticum aestivum L.), triticale (Triticale hexaploide L.), and cereal rye (Secale cereale), respectively, in spring (March–May)). Compared to the optical-only model, InSAR coherence improved biomass estimations by 4% in wheat, 5% in triticale, and by 11% in cereal rye. Both optical-only and optical-SAR biomass prediction models exhibited saturation occurring at ~1900 kg ha−1; thus, more work is needed to enable accurate biomass estimations past the point of saturation. To address this continued concern, future work could consider the use of weather and climate variables, machine learning models, the integration of proximal sensing and satellite observations, and/or the integration of process-based crop-soil simulation models and remote sensing observations.

Remote Monitoring of Civil Infrastructure Based on TomoSAR

Structural health monitoring and damage detection tools are extremely important topics nowadays with the civil infrastructure aging and deteriorating problems observed in urban areas. These tasks can be done by visual inspection and by using traditional in situ methods, such as leveling or using traditional mechanical and electrical sensors, but these approaches are costly, labor-intensive and cannot be performed with a high temporal frequency. In recent years, remote sensing has proved to be a very promising methodology in evaluating the health of a structure by assessing its deformation and thermal dilation. The satellite-based Synthetic Aperture Radar Tomography (TomoSAR) technique, based on the exploitation of a stack of multi-temporal SAR images, allows to remotely sense the movement and the thermal dilation of individual structures with a centimeter- to millimeter-level accuracy, thanks to new generation high-resolution satellite-borne sensors. In this paper, the effectiveness of a recently developed TomoSAR technique in assessing both possible deformations and the thermal dilation evolution of man-made structures is shown. The results obtained using X-band SAR data in two case studies, concerning two urban structures in the city of Naples (Italy), are presented.

Permanent Laser Scanner and Synthetic Aperture Radar Data: Correlation Characterisation at a Sandy Beach.

In recent years, our knowledge of coastal environments has been enriched by remotely sensed data. In this research, we co-analyse two sensor systems: Terrestrial Laser Scanning (TLS) and satellite-based Synthetic Aperture Radar (SAR). To successfully extract information from a combination of different sensors systems, it should be understood how these interact with the common environment. TLS provides high-spatiotemporal-resolution information, but it has high economic costs and limited field of view. SAR systems, despite their lower resolution, provide complete, repeated, and frequent coverage. Moreover, Sentinel-1 SAR images are freely available. In the present work, Permanent terrestrial Laser Scanning (PLS) data, collected in Noordwijk (The Netherlands), are compared with simultaneous Sentinel-1 SAR images to investigate their combined use on coastal environments: knowing the relationship between SAR and PLS data, the SAR dataset could be correlated to beach characteristics. Meteorological and surface roughness have also been taken into consideration in the evaluation of the correlation between PLS and SAR data. A generally positive linear correlation factor up to 0.5 exists between PLS and SAR data. This correlation occurs for low- or moderate-wind-speed conditions, whilst no particular correlation has been highlighted for high wind intensity. Furthermore, a dependence of the linear correlation on the wind direction has been detected.

Generating landslide density heatmaps for rapid detection using open-access satellite radar data in Google Earth Engine

Abstract. Rapid detection of landslides is critical for emergency response, disaster mitigation, and improving our understanding of landslide dynamics. Satellite-based synthetic aperture radar (SAR) can be used to detect landslides, often within days of a triggering event, because it penetrates clouds, operates day and night, and is regularly acquired worldwide. Here we present a SAR backscatter change approach in the cloud-based Google Earth Engine (GEE) that uses multi-temporal stacks of freely available data from the Copernicus Sentinel-1 satellites to generate landslide density heatmaps for rapid detection. We test our GEE-based approach on multiple recent rainfall- and earthquake-triggered landslide events. Our ability to detect surface change from landslides generally improves with the total number of SAR images acquired before and after a landslide event, by combining data from both ascending and descending satellite acquisition geometries and applying topographic masks to remove flat areas unlikely to experience landslides. Importantly, our GEE approach does not require downloading a large volume of data to a local system or specialized processing software, which allows the broader hazard and landslide community to utilize and advance these state-of-the-art remote sensing data for improved situational awareness of landslide hazards.

Monitoring deformations of infrastructure networks: A fully automated GIS integration and analysis of InSAR time-series

Ageing stock and extreme weather events pose a threat to the safety of infrastructure networks. In most countries, funding allocated to infrastructure management is insufficient to perform systematic inspections over large transport networks. As a result, early signs of distress can develop unnoticed, potentially leading to catastrophic structural failures. Over the past 20 years, a wealth of literature has demonstrated the capability of satellite-based Synthetic Aperture Radar Interferometry (InSAR) to accurately detect surface deformations of different types of assets. Thanks to the high accuracy and spatial density of measurements, and a short revisit time, space-borne remote-sensing techniques have the potential to provide a cost-effective and near real-time monitoring tool. Whilst InSAR techniques offer an effective approach for structural health monitoring, they also provide a large amount of data. For civil engineering procedures, these need to be analysed in combination with large infrastructure inventories. Over a regional scale, the manual extraction of InSAR-derived displacements from individual assets is extremely time-consuming and an automated integration of the two datasets is essential to effectively assess infrastructure systems. This paper presents a new methodology based on the fully automated integration of InSAR-based measurements and Geographic Information System-infrastructure inventories to detect potential warnings over extensive transport networks. A Sentinel dataset from 2016 to 2019 is used to analyse the Los Angeles highway and freeway network, while the Italian motorway network is evaluated by using open access ERS/Envisat datasets between 1992 and 2010, COSMO-SkyMed datasets between 2008 and 2014 and Sentinel datasets between 2014 and 2020. To demonstrate the flexibility of the proposed methodology to different SAR sensors and infrastructure classes, the analysis of bridges and viaducts in the two test areas is also performed. The outcomes highlight the potential of the proposed methodology to be integrated into structural health monitoring systems and improve current procedures for transport network management.

Environment Monitoring of Shanghai Nanhui Intertidal Zone With Dual-Polarimetric SAR Data Based on Deep Learning

Satellite-based synthetic aperture radar (SAR) can provide a low-cost, frequent environment monitoring for dynamic intertidal zones. The critical problem is to realize pixel-level classification of SAR images of the intertidal zones with excellent and robust performance. Recently, deep learning, in particular deep convolutional neural networks, has provided us with promising solutions to this problem. Based on a sophisticated deep learning-based pixel-level classification model U²-Net, we propose an MB-U²-ACNet model suitable for intertidal zone land cover classification using dual-polarimetric SAR data integrated with environmental information, such as wind speed and tide level information. The MB-U²-ACNet model has a multi-branch nested U-shaped encoding-decoding structure. We extract and fuse features from multiple data sources, including satellite remote sensing and environmental information, by establishing the multi-branch structure. Furthermore, we propose an asymmetric convolution residual U-block for each encoding-decoding stage to improve the model&#x2019;s feature extraction ability. Moreover, the model with attention mechanisms better distinguishes the importance of features from the channel&#x2019;s perspectives and spatial dimensions. We construct a dataset with 106 Sentinel-1 SAR images from 2016 to 2020 for environment monitoring in the intertidal zone of Shanghai Nanhui. On the dataset, the proposed model reaches the overall classification accuracy of 96.40% and the mean intersection over union score of 0.8307. The experiments show the advantages of the proposed model compared with the benchmarking models due to better feature extraction and multi-source information fusion. In addition, the contributions of every added sub-structure are analyzed systematically.