Change detection (CD) based on multi-temporal remote sensing imagery is a crucial step for various earth observation applications. While deep learning (DL) has revolutionized CD, its data-driven nature demands substantial labeled images for supervised model training, which is costly and time-consuming. This paper addresses the challenge of limited training samples by proposing a novel object-based change augmentation (OCA) method. Unlike conventional image-level augmentation methods that can introduce irrelevant contextual dependencies, OCA decomposes the augmentation process into few-shot object classification and foreground-background pasting, thereby generating in-distribution synthetic images with increased change diversity. An object-based training strategy is developed to create a high-confidence binary classifier for pseudo-semantic segmentation, facilitating the copy-paste operation. Experimental results on the very high-resolution remote sensing images demonstrate the superior performance of OCA compared to existing augmentation-based and generation-based methods. A comprehensive analysis of parameter sensitivity, adaptability to varying training data volumes, and compatibility with diverse CD methods validate its robustness. This approach provides a practical and effective solution for few-shot CD scenarios, advancing the applicability of DL-based CD methods in training data-limited environments. Codes and data are available: https://github.com/openrsgis/OCA.

A multi-scale methodology is designed to assess the baseline and future land subsidence risk scenarios in major urban areas of Italy. Ground deformation observations from multi-temporal satellite Interferometric Synthetic Aperture Radar (InSAR), hydrogeological, topographic and land use datasets are embedded into an innovative risk assessment workflow, and processed with advanced geostatistics to identify the main subsidence hotspots and drivers. Future subsidence risk is assessed accounting for various climate change scenarios, demographic and urban development. The results for the 15 metropolitan cities of Italy, and Emilia Romagna region showcase the potential of the developed methodology and its benefits to inform water resource management and decision making, towards sustainable use of groundwater resources and urban development.

Remote sensing has increasingly supported archaeological and cultural heritage applications over the past century, and satellite Synthetic Aperture Radar (SAR) has played a key role in advancing this application field. In this paper, two case studies from the wider Province of Rome (Italy) are exploited to investigate SAR imaging capabilities for archaeological prospection and heritage site protection. Medium to very high spatial resolution SAR data acquired by RADARSAT-2, Sentinel-1, ALOS-1 and COSMO-SkyMed are used to trial the detection of crop marks at (semi-)buried and sub-surface archaeological features in Ostia-Portus. Big data stacks of Sentinel-1 imagery are processed with the parallelized Small BAseline Subset (SBAS) Interferometric SAR (InSAR) method to monitor the stability of cultural heritage assets within the UNESCO World Heritage Site of Rome.

Ground subsidence, particularly in urban environments, poses a significant geohazard risk to private and public infrastructure. The subsidence can be caused by anthropogenic activity, such as excessive groundwater pumping, underground excavation, land reclamation, or geological forces (e.g., slow fault movement and earthquakes, and volcanic activity). Although terrestrial measurements using a Global Satellite Navigation Systems (GNSS) or in-situ leveling survey provide accurate ground displacement information with a variable temporal resolution, such measures are limited due to their sparse spatial sampling. Furthermore, terrestrial measurements are associated with a high cost to survey the ground deformation. We present time-series analyses of ground subsidence in the southeastern part of Korea using multi-frequency synthetic aperture radar (SAR) observations. We investigate if ground subsidence can be retrieved with recent SAR observations that will be processed using several approaches for Interferometric SAR (InSAR) time-series, including Small Baseline Subset (SBAS) and Persistent Scatterer Interferometry (PSI) techniques. Preliminary results using COSMO-SkyMed and ALOS-2 SAR observations will be presented.

The research results described in this paper have been obtained in the framework of the 2019–2022 ALGORITMI project between the Italian Space Agency (ASI) and the Institute of Applied Physics of the National Research Council (CNR-IFAC). The focus of the research was the development of innovative algorithms for the estimation of geophysical parameters of soil, snow, and vegetation with the aim of monitoring soil, snow cover and agricultural crop conditions. The estimation of soil moisture, vegetation biomass, snow water equivalent, and crop classification was improved by using retrieval algorithms based on machine- learning approaches and temporal series of SAR images from COSMO-SkyMed (CSK) and Sentinel-1 (S-1) missions, along with optical images from Sentinel-2. This paper provides an overview of the most recent and valuable results obtained during the project. In particular, the validation of soil moisture provided R=0.89 and RMSE=0.025 m3/m3by integrating data from S-1 and CSK and that one of snow water equivalent gave R=0.85 with RMSE=86.24 mm (CSK HIMAGE) and R=0.86 with RMSE=71.59 mm (CSK PP). Early mapping results showed an almost monotonic progression in overall accuracy over time higher than 90% by increasing the available images.

This study aims at estimating the dry snow water equivalent (SWE) by using X-band SAR data from the COSMO-SkyMed (CSK) satellite constellation. Time series of CSK acquisitions have been collected during the dry snow period in the Alto Adige test site, in the Italian Alps, during the winter seasons from 2013 to 2015 and from 2019 to 2021. The SAR data have been analyzed and compared with the in-situ measurements to understand the X-band SAR sensitivity to SWE, which has been further assessed by Dense Media Radiative Transfer (DMRT) model simulations. The sensitivity analysis provided the basis for addressing the SWE retrieval from the CSK data, by exploiting two different machine learning (ML) techniques, namely Artificial Neural Networks (ANN) and Support Vector Regression (SVR). To ensure a statistical independence of training and validation processes, the algorithms are trained and tested using SWE predictions of the fully distributed snow model AMUNDSEN as reference data and are subsequently validated on the experimental dataset. Due to its influence on the CSK estimates, the effect of forest canopy was accounted for in the analysis. Depending on the algorithm, the validation resulted in a correlation coefficient 0.78 ≤ R ≤ 0.91, and a Root Mean Square Error 55.5 mm ≤ RMSE ≤ 87.4 mm between estimated and in-situ SWE. Further analysis and validation are needed; however, the obtained results seem suggesting the Cosmo-SkyMed constellation as effective tool for the retrieval of the dry snow water equivalent in alpine areas.

This work provides Synthetic Aperture Radar (SAR) big data investigations based on ESA's Geohazards Exploitation Platform (GEP) and the Parallel Small BAseline Subset (P-SBAS) Interferometric SAR (InSAR) on-demand service. Six Sentinel-1 IW SAR stacks for a total of 981 scenes acquired in 2014–2020 were processed to generate advanced ground deformation products providing key geo-information on natural and anthropogenic processes affecting 4 study areas in the Mediterranean: Tunis (Tunisia), the town of Gela (Italy), Methana volcano (Greece), and Crotone and the Capo Colonna promontory (Italy). The identified geohazards comprise subsidence due to land drainage, reclamation and compaction, soil consolidation and infrastructure settlement following engineering works, groundwater pumping for irrigation and industrial use, hydrocarbon extraction, slow-moving landslides and erosion landforms.

In this paper, an algorithm based on Convolutional Neural Networks (CNNs) was developed to correctly classify an agricultural area in central Italy, by using SAR images. This preliminary step is vital for mastering the different influence of crop types in SAR data before the implementation of algorithms devoted to estimate of vegetation biomass. In situ data collected on the test site were used for validating the CNN algorithm-based classification. After the agricultural species recognition, a sensitivity analysis between C-band Sentinel-1 and X-band COSMO-SkyMed backscatter coefficients and crop biomass was carried out, laying the foundation for the implementation of algorithms able to estimate the biomass of different crop types.

The monitoring of snow conditions in Alpine areas to support water management and avalanche warning applications would require the estimate of snow parameters, such as the snow water equivalent (SWE). In this research, COSMO-SkyMed (CSK) X-band SAR data were exploited to estimate the SWE. In-situ snow measurements (depth, density, snow grain radius, temperature) collected in South Tyrol (Italy), were used to simulate the X-band backscatter with the Dense Medium Radiative Transfer (DMRT) electromagnetic model. Two SWE retrieval algorithms based on machine learning approach were implemented. The algorithms are based on Artificial Neural Networks (ANN) and Support Vector Regression (SVR) and have been trained with both experimental data and DMRT model simulations. These algorithms were applied to a selection of CSK StripMap HIMAGE HH-polarized scenes collected over the test area. The obtained results are promising and they confirm the potential of SAR data at X-band to retrieve snow parameters, although the algorithm validation should be improved in the future, with more consistent measurement dataset.

Vessel detection and their activities in the sea can provide updates on latest trends in maritime trade. Space-borne synthetic aperture radar (SAR) can aid in detecting vessels in (almost) all weather conditions. In this study high resolution SAR data are used to analyze the maritime traffic activities, especially the underscored independency in transport trends in Wuhan, the major port-hub on the central Yangtze river in China, before and during the COVID-19 pandemic. Time-series of COSMO-SkyMed SAR images covering Wuhan from 2018 to 2020 were exploited to detect vessels. We applied multi-mode feature and shape (MMFS) image enhancement for fast and accurate vessel detection. Variations in number of vessels were detected, especially a huge drop was observed during the COVID-19 lockdown. HwkEye360 radio frequency monitoring data were used to validate our results.

An unprecedented time series of 293 COSMO-SkyMed StripMap SAR images acquired in 2011–2020 is exploited to investigate land subsidence and vehicle traffic in Wuhan, China. Persistent Scatterer Interferometry using linear and non-linear deformation models suggests that the spatial and temporal evolution of subsidence relates with the dynamic urban development across the main city districts. Traffic patterns along bridges were captured by detecting vehicles based on their azimuth shift caused by their across-track motion, and identified by type based on their radar cross section and speed. The results of vehicle counting confirm an increasing number of vehicles over the last years, which is currently an urban challenge for Wuhan.

Long stacks of Copernicus Sentinel-1 IW SAR images acquired in 2014-2019 are processed with the Small Baseline Subset (SBAS) and Permanent Scatterers (PS) Interferometric SAR (InSAR) methods to retrieve present-day land deformation rates across major cities in central Mexico. InSAR-derived subsidence velocity reflects intense groundwater pumping from shallow and deep aquifers for public, agricultural and industrial use, and consequent water level drop and aquifer depletion. In the capital Mexico City, as well as in the valleys of Toluca and Tulancingo, which all belong to aquifers recognized by the National Water Commission as in deficit in 2018, vertical deformation rates are as high as 40, 8 and 6 cm/year, respectively. Most pronounced rates occur mainly on highly compressible, Quaternary clay and silt-rich deposits. Rates of 6.5 cm/year are also observed at well-defined subsiding zones in Puebla, in response to groundwater abstraction for public-urban and industrial use.

The 4 year-long Recovery Observatory project was triggered by the Committee on Earth Observation Satellites (CEOS) to define a sustainable vision for increased use of satellite EO in support of recovery after 2016 Hurricane Matthew struck southwestern Haiti. ESA's Geohazards Exploitation Platform (GEP) was exploited to develop a SAR-based workflow to access, process and generate value-added products with Sentinel-1, TerraSAR-X and COSMO-SkyMed imagery that Haitian end-users can use to support their decision-making processes and recovery progress monitoring. Sentinel-1 IW data were processed with SNAP and SNAC tools to generate change detection products (e.g. coherence and amplitude change maps to detect flooded areas). InSAR ground deformation products generated with PS-InSAR FASTVEL and P-SBAS tools allowed the identification of unstable areas in the town of Jérémie and along its western coastline, which highlight potential concern for urban development and reconstruction. Ground truth and evidence of land instability were collected in the field in mid-2019 to validate satellite observations.

Modern agriculture is facing new challenges about food production for a growing population in a sustainable manner. Crop mapping at local and regional scale could provide valuable information in support of agricultural policy. This paper describes a field mapping investigation in a populated area in Tuscany (Italy). Satellite images from Sentinel-1 C-band and COSMO-SkyMed X-band SAR and Sentinel-2 optical sensors are input of classifiers based on deep learning and convolutional neural networks. Results pinpointed that the use of optical images allowed the best overall classification accuracy (99.7%), nevertheless X-band SAR imagery, providing an accuracy of 94.6%, could be a good substitute of optical indices in case of lack of cloud-free multispectral data.

The characterization of snow conditions and the estimation of snow water equivalent (SWE) are the main goals of this paper, achieved through the exploitation of multi-frequency SAR data at both C- and X-bands from Sentinel-1 (S-1) and COSMO-SkyMed (CSK) satellites, respectively. Dry/wet snow conditions have first been assessed using C-band S-1 images. Subsequently, a sensitivity analysis was carried out by using datasets of in-situ snow measurements (i.e. snow depth, density, snow grain radius, temperature and wetness) collected in South Tyrol region, in north-eastern Italy. Simulations based on the Dense Medium Radiative Transfer (DMRT) forward electromagnetic model were considered to interpret and assess the experimental findings. Two retrieval algorithms for SWE estimation from X-band SAR data were implemented. These algorithms are based on machine learning approaches, i.e. Artificial Neural Networks (ANN) and Support Vector Regression (SVR). The training of the algorithms accounts for experimental data and DMRT model simulations and, then is applied to a selection of X-band CSK StripMap HIMAGE scenes collected over the test area. The results are promising, and pave the way for further analysis and validation to exploit the potential of SAR for snow parameter retrieval.

This work discusses the feasibility of monitoring ground stability and motion across the entire British landmass using satellite InSAR techniques. ERS-1/2 and ENVISAT archive data availability, topographic visibility and land cover constraints for multi-temporal InSAR techniques to succeed across Britain are analysed. Data volumes, hardware and software requirements for the generation of a nationwide InSAR product are discussed, with a view to both novel processing methods to extend InSAR across unfavourable land covers, and parallel and cloud computing systems to decrease InSAR processing time demands. The P-SBAS method implemented onto ESA's G-POD platform is tested for London and Newcastle using ERS-1/2 1992-2000 and ENVISAT 2002-2008 image stacks, revealing a decrease of the processing time demand to ~8 hours per image frame.

This work presents a Differential SAR Interferometry (DInSAR) analysis of slow-moving landslides in Costa della Gaveta (southern Italy) exploiting the improved spatial density of radar targets provided by the novel Intermittent SBAS (ISBAS) algorithm. Several landslides occurred in this area over the past decade, producing a ground displacement of several centimeters and causing unsafe road and rail traffic conditions. In the 2.4 km2study area, ISBAS analysis of TerraSAR-X data acquired in 2010–2011 has shown the presence of sixteen phenomena with a southeastern main direction of motion. The DInSAR results agree with both the magnitude and the deformation mechanisms that were mapped during field observations and reported in the geotechnical literature.

Stability of monuments and subsidence of residential quarters in Rome (Italy) are depicted based on geospatial analysis of more than 310,000 Persistent Scatterers (PS) obtained from Stanford Method for Persistent Scatterers (StaMPS) processing of 32 COSMO-SkyMed 3m-resolution HH StripMap ascending mode scenes acquired between 21 March 2011 and 10 June 2013. COSMO-SkyMed PS densities and associated displacement velocities are compared with almost 20 years of historical C-band ERS-1/2, ENVISAT and RADARSAT-1/2 imagery. Accounting for differences in image processing algorithms and satellite acquisition geometries, we assess the feasibility of ground motion monitoring in big cities and metropolises by coupling newly acquired and legacy SAR time series. Limitations and operational benefits of the transition from medium resolution C-band to high resolution X-band PS data are discussed, alongside the potential impact on the management of expanding urban environments.

36 ENVISAT ASAR images acquired in 2002 to 2010 along descending passes with nominal revisiting time of 35 days were processed over the whole region of Homs, western Syria, by implementing the low-pass Small Baseline Subset (SBAS) technique. More than 280,000 coherent pixels with ~100m ground resolution were obtained. We analysed pixel spatial distribution in respect of local geology and land use, to assess to what extent these factors can influence the performance of an interferometric deformation analysis in semi-arid environment. Filtering out the amount of pixels associated with the urban fabric of Homs and surrounding villages, it is apparent that limestone and marl units are less prone to generate coherent pixels if compared with the basalt units in the north-western sector of the processed region. The latter resulted in pixel density of ~50-60 pixels/km2, which is comparable with that found over urban settlements and man-made structures.

We model terrain visibility and topographic distortions to the ERS-1/2 SAR and ENVISAT ASAR IS2 satellite acquisition modes in Great Britain using the 5m NEXTMap DTM. Predictions of Persistent Scatterers (PS) densities identifiable over the landmass are drawn using the CORINE Land Cover 2006 dataset which is calibrated based on 6 PS datasets available for various areas of the UK. InSAR feasibility to monitor ground motions is discussed through the example of the Manchester area, with particular regard to landslide deposits in the Peak District.

This work illustrates the potential of Persistent Scatterer Interferometry (PSI) using X-band SAR (Synthetic Aperture Radar) data for a detailed detection and characterization of landslide ground displacements at local scale.

We present a subsidence study based on the analysis of displacement measurements from ENVISAT-ASAR Persistent Scatterer Interferometry and on the extraction of horizontal gradient maps. Land subsidence is analyzed for three cities in central Mexico: Chalco, Aguascalientes and Morelia. Even though stratigraphic and tectonic backgrounds of these areas impose different subsidence patterns as imaged through the interferometric analysis, the three case studies all reveal the necessity of a combined analysis of subsidence magnitudes and horizontal gradients for better definition of the areas with higher vulnerability to surface faulting. Such an analysis can be used as the basis for subsequent hazard and risk mapping.

This work shows the capabilities of InSAR time series analyses to support civil protection activities in the framework of geological risk management and mitigation. We discuss the outcomes from an integrated analysis of conventional in situ investigations and observations with advanced InSAR analyses carried out for the test sites of Agrigento and Naro (Italy), affected by ground instability respectively due to landsliding and tectonic forces. The study of past ground deformations provided valuable insights into the spatial and temporal patterns and behaviors of these phenomena, helping local civil protection authorities to focus resources on the areas of maximum need and to identify the most appropriate mitigation measures to reduce the impacts on elements at risk.