We present in this study an enhancement to our previous work, in which an automated method, based on Convolutional Neural Networks (CNNs), has been developed for identifying co-seismic ground deformation patterns within the Differential SAR Interferometry (DInSAR) maps generated through the EPOSAR service of the European Plate Observing System (EPOS) Research Infrastructure. The implemented improvements have been achieved through two main lines of actions. Firstly, the generation of the synthetic dataset, used to train the developed CNN, has been enhanced; this concerns the improvement of the simulation of possible atmospheric disturbance within the DInSAR interferograms, and the introduction of a simulator of phase unwrapping errors. Secondly, the original CNN-based deformation pattern detector layout, performing a binary classification, has been modified to account for a multiclass deformation pattern classification. This allows us extending the capability of the system which, in addition to detect co-seismic deformation patterns, may also provide information on the earthquake source characteristics. The presented solution will be included in the EPOSAR DInSAR processing chain.

We present in this work the results of an extensive, full resolution Parallel SBAS (P-SBAS) processing activity carried out on large multi-temporal SAR data archives of COSMO-SkyMed first (CSK) and second (CSG) generation SLC images, relevant to some of the main cities of the Italian territory, acquired from ascending and descending orbits since 2009. In particular, we first summarize the key points of the exploited, full resolution P-SBAS processing chain. Subsequently, we also describe the main algorithmic developments aimed to exploit large temporal sequences of CSK and CSG SLC data relevant to wide areas. Finally, we present several results highlighting, through some selected examples, that the use of high resolution SAR images, such as those of the CSK-CSG constellation, processed with techniques like the full resolution P-SBAS technique approach, may provide valued-added information that, properly integrated with in-situ measurements, can be very relevant for the assessment of the built-up environment health conditions, at the national scale.

Over the past 30 years it has been widely demonstrated the effectiveness of the Differential Synthetic Aperture Radar Interferometry (DInSAR) technique to retrieve surface deformation information relevant to tectonically active areas. However, this technique exhibits some limitations due the presence of possible decorrelation effects, phase unwrapping errors, and artefacts due to the temporal/spatial variability of the atmospheric conditions between the SAR acquisition pairs exploited to generate the interferograms and the corresponding deformation maps or time series. A challenging situation may arise when an earthquake event occurs and a co-seismic DInSAR deformation map is generated to quickly support risk management operations. In this scenario, having an automatic process reducing the uncertainties of the retrieved, DInSAR-based, surface deformation information could highly improve the quality of the products made available to the scientific community and of the service provided to the national disaster recovery authorities. We present in this work a solution, based on standard CNN architectures embedded in the DInSAR processing chain of the EPOSAR service, developed within the European Plate Observing System (EPOS) Research Infrastructure, to automatically identify co-seismic ground deformation patterns.

It is widely recognized that SAR images can suffer from the presence of RFI, especially at the low frequencies of the microwave spectrum. Such nuisance signals can significantly impair the quality of the SAR data with unwanted effects on the final products. This paper describes a processing strategy for the detection and cancellation of RFI signals corrupting airborne SAR raw-data. In addition, the performance of the proposed strategy is assessed by exploiting airborne L-band FMCW-SAR data.

In this work, we present the ground deformation analysis of the Italian Peninsula carried out via the Sentinel-1 (S-1) Parallel Small BAseline Subset (P-SBAS) approach. In particular, we generate surface deformation time-series and the corresponding mean deformation velocity maps by considering the overall archive of S-1 data acquired from both ascending and descending orbits during the March 2015 - December 2018 period. Moreover, we perform a large-scale comparison of the retrieved results with the deformation measurements of the available GPS stations deployed on the Italian territory and provided by the Nevada Geodetic Laboratory at the University of Nevada, Reno, USA (UNR-NGL). Finally, we properly combine the ascending and descending deformation time-series in order to compute the vertical and east-west components of the retrieved surface displacements. The achieved results demonstrate the capability of the P-SBAS processing chain to efficiently process large S-1 datasets to investigate ground deformation dynamics of large portions of the Earth surface.

We present an advanced differential synthetic aperture radar (SAR) interferometry (DInSAR) processing chain, based on the Parallel Small BAseline Subset (P-SBAS) technique, for the efficient generation of deformation time series from Sentinel-1 (S-1) interferometric wide (IW) swath SAR data sets. We first discuss an effective solution for the generation of high-quality interferograms, which properly accounts for the peculiarities of the terrain observation with progressive scans (TOPS) acquisition mode used to collect S-1 IW SAR data. These data characteristics are also properly accounted within the developed processing chain, taking full advantage from the burst partitioning. Indeed, such data structure represents a key element in the proposed P-SBAS implementation of the S-1 IW processing chain, whose migration into a cloud computing (CC) environment is also envisaged. An extensive experimental analysis, which allows us to assess the quality of the obtained interferometric products, is presented. To do this, we apply the developed S-1 IW P-SBAS processing chain to the overall archive acquired from descending orbits during the March 2015-April 2017 time span over the whole Italian territory, consisting in 2740 S-1 slices. In particular, the quality of the final results is assessed through a large-scale comparison with the GPS measurements relevant to nearly 500 stations. The mean standard deviation value of the differences between the DInSAR and the GPS time series (projected in the radar line of sight) is less than 0.5 cm, thus confirming the effectiveness of the implemented solution. Finally, a discussion about the performance achieved by migrating the developed processing chain within the Amazon Web Services CC environment is addressed, highlighting that a two-year data set relevant to a standard S-1 IW slice can be reliably processed in about 30 h.The presented results demonstrate the capability of the implemented P-SBAS approach to efficiently and effectively process large S-1 IW data sets relevant to extended portions of the earth surface, paving the way to the systematic generation of advanced DInSAR products to monitor ground displacements at a very wide spatial scale.

In this work we present a fully parallel SBAS-DInSAR processing chain for the effective and massive generation of Sentinel-1 (S-1) Interferometric Wide Swath (IWS) deformation time-series. The pursued strategy to develop such a parallel processing chain fully benefits from the data acquisition characteristics of the TOPS mode, consisting of a series of bursts that can be considered as separate images to be independently processed. Such a data structure fosters the intrinsic parallelization of the S-1 IWS interferometric data processing, which can automatically be carried out through the exploitation of high-performance distributed computing infrastructures, and of efficient algorithms for tackling the huge data flow provided by the S-1 constellation. In order to demonstrate the capability of the presented S-1 P-SBAS processing chain to deal with the huge amount of data acquired by the S-1 constellation, we process a very extended dataset acquired over Italy from which we compute the mean velocity maps and corresponding deformation time-series.

We present a simple solution for the phase-preserving focusing of Terrain Observation with Progressive Scan (TOPS) Sentinel-1 data. The proposed algorithm focuses each raw data burst independently. In particular, after a conventional range data compression, it is based on a modification of the two-step processing approach formerly developed to focus spotlight and hybrid stripmap/spotlight SAR data. The presented results relevant to a real Sentinel-1 data acquired over central Italy confirm the effectiveness of the proposed approach.

In this work we present an advanced interferometric processing chain, which is based on the DInSAR algorithm referred to as Parallel Small BAseline Subset (P-SBAS) approach, for the massive processing of SENTINEL-1 (S1) Interferometric Wide Swath (IWS) data. The P-SBAS S1 processing chain produces surface deformation time series, and the relevant mean velocity maps, in automatic and systematic way by efficiently exploiting Cloud Computing infrastructures, thus allowing us to perform DInSAR analyses at very large scale in reduced time frames. As experimental results, the overall mean deformation velocity map relevant to the Central and Southern Italy zone (from Lazio to Sicily), generated by processing in parallel about 300 S1 acquisitions within the Amazon Web Services Cloud Computing platform, is presented. Moreover, the displacement time series of some pixels located in volcanic deforming areas such as the Campi Flegrei Caldera (Napoli Bay area) and the Mt. Etna (Sicily) are shown.

In this work we present an efficient interferometric processing chain, based on the advanced DInSAR algorithm referred to as Parallel Small BAseline Subset (P-SBAS), for the generation of Sentinel-1A (S1A) Interferometric Wide Swath deformation time-series, which is able to exploit distributed computing architectures. The presented S1A P-SBAS processing chain has been successfully implemented within the ESA Geohazard Exploitation Platform to provide an on-demand automatic service for the unsupervised generation of P-SBAS displacement time-series. To give an idea of the effectiveness of the presented S1A processing chain, as a preliminary result we show a 12-days interferometric analysis at continental scale, carried out by exploiting 150 S1A interferometric pairs acquired over Europe for an overall covered area of about 7,500,000 km2.

This work is aimed at describing the development of an efficient interferometric processing chain, based on the well-known advanced Differential Interferometric Synthetic Aperture Radar (DInSAR) algorithm referred to as Small BAseline Subset (SBAS) technique, for the generation of Sentinel-1A (S1-A) Interferometric Wide Swath (IWS) deformation time-series. Due to the TOPS mode characterizing the IWS acquisitions, the existing SBAS processing chains was properly adapted with new procedures for efficiently handling the S1-A data. The developed SBAS-DInSAR chain has been tested on both S1-A and TOPS RadarSAT-2 interferometric dataset, clearly demonstrating the capability of the developed SBAS-DInSAR processing chain to effectively investigate land subsidence phenomena affecting large areas.

The aim of this paper is to demonstrate the applicability of sparse representation to filter out the noise from SAR interferograms. More specifically, we consider an innovative approach for interferometric phase denoising, based on K-SVD technique applied to SAR interferograms, for estimating the noise-free underlying structure of the phase information. To examine the effectiveness of th is approach, the method has been successfully tested both on simulated and real interferograms. In the latter case, the proposed approach has been evaluated using ENVISAT SAR data acquired over the area of Napoli (Italy). The good results achieved applying the proposed method demonstrate its effectiveness for denoising SAR interferograms.

The atmospheric phase screen is one of the main error sources that affect the precise phase measurements in many fields of earth remote sensing. The atmospheric effects can be mitigated if a precise knowledge of the power spectral density of the process is available and if same sample observations can be retrieved on a sparse grid. At smaller scales, where interactions are no longer isotropic, the behaviour is not easily predicted by the ultimate dissipative behaviour of turbulence eddies. We start by assuming that the propagation of the electromagnetic wave in the lower atmosphere can be represented by a ray travelling in a layered medium where the refractive index is constant along each layer. In a turbulent atmosphere, the interaction among eddies induces a diffusion process that propagates with different rates in both vertical and horizontal direction with the final effect of ruling the number of effective layers in the atmosphere. In this way, the overall path travelled by the electromagnetic wave is governed by the accumulated number of such effective layers whose interactions play a primary role in our model. A good model for the interactions among different layers is the linear interaction model. The power spectrum of the process can be found by solving a differential equation with given initial conditions. It can be demonstrated that an asymptotic power law decay is found under binomial competitive interactions and that, at a smaller scale, the behaviour observed in the observed data is naturally induced by the interaction process itself. The model predictions have been tested using samples of the atmospheric phase screen extracted from Synthetic Aperture Radar interferograms. To this end, the model parameters have been estimated from the data set and the predicted spectrum has been compared with the measured one.

The main limiting factors for deformation mea- surements using repeat-pass satellite radar interferometry are temporal decorrelation of the scattering characteristics of the earth and atmospheric delay phase contributions to the interfer- ometric phase. Ferretti et al.(1) showed that using a multitude of radar acquisitions over the same site—a time series approach— reflections can be identified with a stable phase behavior in time, thus allowing deformation behavior to be estimated. The model of observation equations consists of m phase observations for a specific pixel and n unknown parameters describing surface deformation, elevation, and trend. Atmospheric delay is an important error source in these observations, but since it is temporally uncorrelated (while spatially correlated) it can be approximated per pixel per interferometric combination. This approximation requires a heuristic decision on which part of the temporal behavior of the interferometric phase is due to unmodelled deformation (e.g., non-linear deformation if the model estimates only linear deformation), and which part is due to uncorrelated atmospheric signal. Second, the residues attributed to atmosphere for all selected points within a single interferogram are expected to show spatial correlation, following a specific power law behavior (Hanssen, 2001). This results in a second decision on dividing atmospheric contribution and phase noise. Although the assumptions on which these two decisions are based are reasonable and results of previous studies show estimations of deformation and topography which are very likely, there is no independent means of control for the approach followed. In this paper, we will investigate the atmospheric signal estimated from a stack of 70 radar images acquired over Berlin, Germany. The estimated signal will be statistically parameterized and physically compared with meteorological data such as visual, infrared, and water vapor images from meteorological satellites and synoptic data. We will draw conclusions on the likelihood of the assumptions underlying the isolation of atmospheric signal, resulting in an increased reliability of the estimated parameters.