Conference Agenda

Satellite image time series and derived products are increasingly available thanks to the launch of Earth Observation missions which aim at providing a coverage of the Earth every few days with high spatial resolution. The high revisit time of Copernicus (Sentinel-1, Sentinel-2) and Landsat satellites allow for the setup of systematic calculation of ground motion products, opening the way to science and operational monitoring capacities of geohazards. Many services are deployed in order to offer to users systematic or on-demand calculation of optical and InSAR time series products representing ground deformation.

Satellite-derived products and services (e.g. EGMS; EPOS satellite products; GEP, Comet and ARIA services, etc) for the processing of SAR and optical imagery allow accessing variables (displacement/velocity) time series over large areas and time periods. Analyzing and exploiting these datasets (stacks of interferograms, PSInSAR time series, optical derived ground motion, possibly organized in datacubes) necessitate the development of post-processing tools in order to combine the datasets and investigate the spatial and temporal behavior of the studied variables.

TimeSAT is a service allowing to classify ground motion displacement time series in specific behaviors/patterns, detect changes in the time series (increase, decrease, periodicity, …) and identify spatial clusters of homogeneous styles of ground motion. The service currently allows ingesting PSInSAR and SBAS InSAR time series and optical offset-tracking time series. It consists of:

A) a module for data pre-processing (advanced Savitzky-Golay filtering, data subset masking);

B) a module for time series classification, for which three processing workflows are possible: 1) the classification in pre-defined distinctive patterns (uncorrelated trend, linear trend, quadratic trend, bilinear trend) based on a sequence of conditional statistical tests, 2) the unsupervised classification using a combination of independent component analysis (ICA) and principal component analysis (PCA) to detect and classify specific patterns, and 3) the classification using deep Convolutional Neural Network (CNN) architecture using InceptionTime models.

C) a module for the spatial clustering of similar patterns to identify areas and sources of deformation.

A great advantage of TimeSAT is to allow the processing of time series non structured and unevenly distributed in time and in space. The workflow has been optimized and parallelised and is implemented on the Mésocentre/HPC infrastructure of the University of Strasbourg. Thanks to this parallelization and scaling of the code, the processing of about 1 million time series of 5 years period lasts less than 5 hours. The service is currently accessible on the Geohazards Exploitation Platform (GEP) and is part of the eo4alps-landslides application.

The objective of this work is to present the functions of the service through two use case applications, which are the analysis of a SqueeSAR massive dataset available for the Wallis and Vaud cantons in Switzerland, and the analysis of a SNAPPING Full Resolution massive dataset available for part of Slovenia.

The accessibility and availability of Sentinel-1 synthetic aperture radar (SAR) data and Sentinel-2 optical data have revolutionized remote sensing over the last decade. Yet, working with satellite-based SAR and optical data requires specialized training that can hinder broader use by earth scientists, engineers, and decision makers. The Observational Products for End-Users from Remote Sensing Analysis (OPERA) project at the Jet Propulsion Laboratory, with project partners from NASA Goddard Space Flight Center, U.S. Geological Survey, University of Maryland, University of Alaska Fairbanks, and Southern Methodist University, is removing these barriers by producing three analysis ready data products: (1) a near-global Surface Water Extent product suite from optical and SAR data, (2) a near-global Surface Disturbance product suite from optical data, and (3) a North America Displacement product suite from SAR data. The products are designed to meet the needs of U.S. federal agencies as identified by the Satellite Needs Working Group (an initiative of the U.S. Group on Earth Observations) and have broad applications. In addition to these three primary products, OPERA will produce two intermediate SAR products that allow for user-customized product generation: (1) a North America Coregistered Single-Look Complex (CSLC) stack product, and (2) a near-global Radiometric Terrain Corrected (RTC) product. Current data products are derived from various SAR and optical satellites including the ESA Sentinel-1, NASA/USGS Landsat 8, and ESA Sentinel-2 sensors. Future products will utilize data from NASA-ISRO NISAR and NASA SWOT.

In this presentation, we will present an overview of the project status and product information, including how to access the free and open OPERA data through the NASA Distributed Active Archive Centers (DAAC). We will showcase the Surface Water Extent and Surface Disturbance operational products. OPERA’s Surface Water Extent product provides critical data on variations in reservoirs, ponds, rivers, and wetland water surfaces that are useful for science, resource management, environmental protection, hazard mitigation and emergency response. OPERA’s Surface Disturbance product provides data that can be used to identify logging activities, urban expansion, and natural hazards such as landslides and lava flows. We will also introduce the intermediate level OPERA RTC and CSLC products, which will have their operational production release starting at the end of September 2023. OPERA’s RTC product consists of the radar backscatter normalized with respect to the topography and maps signals largely related to the physical properties of the ground scattering objects. Key application examples for RTC include mapping floods and water extent, fires, and landslides. OPERA’s CSLC product consists of SLC images that are precisely aligned or “coregistered” to a common grid and contain both amplitude and phase information of the complex radar return. Key application examples for CSLC include burst-wise interferograms or pixel offset tracking for measuring ground surface deformation for important geophysical phenomena such as earthquakes, volcanoes, groundwater change, and more. Lastly, we will show samples of the OPERA’s future Sentinel-1 Displacement products.

Synthetic Aperture Radar Interferometry (InSAR) techniques are nowadays playing an important role to reveal and analyze ground deformation phenomena, such as those induced by seismic events, volcanic eruptions and landslides, thanks to their capability to provide dense measurements over wide areas and at relatively low costs. This is particularly true thanks to the availability of huge and easily accessible SAR data archives, as those acquired by the Copernicus Sentinel-1 constellation. Indeed, Sentinel-1 routinely provides, since late 2014, C-band SAR data with a defined repeat-pass frequency (down to 6 days when both satellites have been available) at a rather global scale. Therefore, such a constant and reliable availability of data allowed us to move from single event analysis to monitoring tasks, particularly when addressing natural hazard prone areas.

In this work we present the activities that are carried out at the Institute for the Electromagnetic Sensing of Environment of Italian National Research Council (IREA-CNR) to support the national Department of Civil Protection (DPC) for volcanic and seismic areas monitoring with InSAR techniques.

First, we implemented an automatic service [1] that generates, if relevant, the InSAR co-seismic displacement maps once an earthquake occurs. The service queries the main publicly accessible earthquake catalogues (e.g. USGS and INGV) and, according to defined thresholds on magnitude, depth and expected ground deformation, retrieve all the Sentinel-1 data that cover the area of interest (from multiple track and passes) and process them to generate geocoded InSAR products (i.e. displacement maps, wrapped interferograms and spatial coherence). The processing lasts for one month after the main shock, thus ensuring that the phenomena are well imaged. Originally developed to monitor the Italian territory, the service has been extended to operate at global scale and the generated products constitute an archive (see Figure 1) that is made freely available to the scientific community through the European Plate Observing System Research Infrastructure (EPOS-RI) [2, 3].

Moreover, we developed a second service which is devoted to volcano ground displacement monitoring and is also based on Sentinel-1 data, although in this case the temporal evolution of the ground displacement is investigated. The designed system is once again fully automatic and the process is triggered by the availability of a new SAR data in the Sentinel-1 catalogues acquired from both ascending and descending passes, for every monitored volcano site. The data, per each orbit, are automatically ingested and then processed through the well-known Parallel Small BAseline Subset (P-SBAS) technique [4, 5] that allows generating the displacement time series and the corresponding mean displacement velocity maps relevant to the overall observation period. The computed Line of Sight (LOS) measurements are subsequently combined to retrieve the Vertical and East-West components of the deformation, which are straightforwardly understandable by the end user. This service is currently operative for the main active Italian volcanoes: Campi Flegrei caldera, Mt. Vesuvius, Ischia, Mt. Etna, Stromboli and Vulcano. Figure 2 provides an example of the products that are made available to DPC. While tailored for Italian volcanoes, the service can be easily extended to include other volcanic areas on Earth depending on computing resources disposal and data coverage.

Finally, thanks to the availability of an airborne platform which is equipped with a X-band and L-band SAR sensor, we implemented a pre-operative infrastructure referred to as the Multiband Interferometric and Polarimetric SAR (MIPS) system [6] that, in conjunction with the already mentioned spaceborne systems, allows us to provide further information on the areas under study. Due to its flexibility, this system is particularly suitable during emergency scenarios and, for instance, allowed us to understand the elevation changes and the associated large mass wasting and accumulation occurred during the 28 August 2019 paroxysm eruption at Stromboli volcano (see Figure 3).

This work is supported by the 2022-2024 CNR-IREA and Italian DPC agreement, as well as the H2020 EPOS-SP (GA 871121) and Geo-INQUIRE (GA 101058518) projects.

References

1. Monterroso et al. (2020) “A Global Archive of Coseismic DInSAR Products Obtained Through Unsupervised Sentinel-1 Data Processing,” Remote Sens., vol. 12, no. 3189, pp. 1–21. https://doi.org/10.3390/rs12193189
2. EPOS web site: https://www.epos-eu.org/
3. EPOS Data Portal: https://www.ics-c.epos-eu.org/
4. Casu et al. (2014) “SBAS-DInSAR Parallel Processing for Deformation Time Series Computation”, IEEE JSTARS, doi: 10.1109/JSTARS.2014.2322671
5. Manunta et al. (2019) “The Parallel SBAS Approach for Sentinel-1 Interferometric Wide Swath Deformation Time-Series Generation: Algorithm Description and Products Quality Assessment”, IEEE Trans. Geosci. Remote Sens., doi: 10.1109/TGRS.2019.2904912
6. Natale et. al. (2022) “The New Italian Airborne Multiband Interferometric and Polarimetric SAR (MIPS) System: First Flight Test Results, IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium, Kuala Lumpur, Malaysia, pp. 4506-4509, doi: 10.1109/IGARSS46834.2022.9884189

The SNAPPING service for the Copernicus Sentinel-1 mission has been operational on the Geohazards Exploitation Platform (GEP) since February 2021. The service offers GEP users on-demand access to a Persistent Scatterers Interferometry (PSI) chain. The service is meant to simplify the exploitation of EO data resources by combining fast data access, hosted processing and flexibility for users’ own data analysis. SNAPPING services generate average Line-of-Sight (LoS) motion rate maps and corresponding displacement time series at both reduced spatial (approx. 100 m) and full sensor resolutions. The conceptual twofold processing of the service separating the generation of the interferometric data stack (SNAPPING IFG) and the time series analysis (SNAPPING PSI) provides flexibility when regular updates of the solution are required, reducing in the meanwhile the consumption of resources and the corresponding processing time.
Although successfully utilized by numerous GEP users for both science and application projects, for the majority of cases processing is limited in terms of spatial extent. Herein, an effort has been made to demonstrate the underlying capabilities of platform-based solutions by showcasing nationwide SNAPPING processing of Greece.
A dedicated scheme based on SNAPPING PSI Med was developed to ensure coverage of entire land surfaces (including isolated islands), while minimizing propagation of error sources. The Greek territory (~132k sq.km) was thus splitted into 54 tiles of approx. 90 x 90 km, having spatial overlap not lower than 10 km. The totality of Copernicus Sentinel-1A archive over Greece in descending orbits was exploited covering the period between 04/2015 and 12/2021. With an observation period of approximately 7 years the millimeter accuracy of the obtained surface motion calculations is achieved. The input dataset consisted of more than 18k acquisitions, corresponding to 174-198 observation dates per tile.
Initial processing steps involved the manual selection of the acquisitions for each tile, preparation of platform input parameters and finally the supervised execution of tile-based processing on the GEP platform. Special attention has been made to ensure proper handling of regions affected by abrupt motion induced by major earthquakes. As an outcome of the activity, a total number of 4M point measurements were detected, showing surface motion for nation-wide Greece at medium resolution (Figure 1). The inter-comparison of the obtained results to other sources of wide-area interferometric measurements, such as the European Ground Motion Service (EGMS), underlines the consistency of independent solutions, while highlighting the differences between the various processing approaches. The obtained dataset is made publicly accessible via GEP, anticipating further exploitation in various research domains to improve our understanding of geohazard phenomena.

Protecting the population and their livelihood from natural hazards is one of the central tasks of Swiss state. Efficient prevention, preparation and intervention measures can be used to prevent or at least limit potential material damage and fatalities as a result of natural hazards. Warnings and alerts are particularly cost-effective instruments for reducing damage, as they allow emergency personnel and the population to take the prepared measures.

The Swiss Federal Office of Topography (swisstopo) therefore procures processed InSAR data to detect any changes in the terrain of the whole of Switzerland. The object of the service is the procurement of processed InSAR data for the entire surface of Switzerland. The data provided by the Sentinel-1 (S1) SAR satellite constellation, as part of the European Union’s Copernicus Earth observation programme, are processed as the data basis for the Swiss-wide monitoring of surface motion.

The service implementation includes the analysis of all the available historical (S1), from 2014 up to November 2020, followed by annual updates, at least up to 2023.

The area of interest is covering Switzerland and Liechtenstein, including a 5 km buffer, for a total surface of approximately 50’000 km². This area is covered by five different S1 tracks, two ascending and three descending, from October 2014 up to now. The approximate number of acquisition per track is about 300, characterized by a 6-day revisiting time, which is showing a regular sampling with no data gaps starting from November 2015.

The end-to-end workflow of the production chain includes the S1 Data Ingestion, the core processing and a final quality control step.

Southern Switzerland is characterized by prominent topography, as it includes more than the 13% of the Alps, comprising several peaks higher than 4’000 m above sea level. In fact, the Alps cover 60% of Switzerland. Therefore, a preliminary analysis has been addressed on the creation of layover and shadow maps, for each S1 relative orbit, to identify the portions of the study area where the combination of topography and the satellite acquisition geometry do not allow getting information from InSAR techniques.

Additionally, the vast mountainous areas are often affected by seasonal snow cover, which, in turn, is affecting S1 interferometric coherence over long periods, resulting in loss of data for parts of the year. To handle the periodical data decorrelation or misinterpretation of the data phase information during the snow period, a specific strategy to correctly threat these circumstances has been designed.

The Core Processing is responsible for the generation of all required products, operating on S1 and ancillary data. The deformation products are obtained exploiting a hybrid algorithm, which is integrating of both Small Baseline subset (SBAS) and Persistent Scatterers Interferometry (PSI) methods, in order to estimate the temporal deformation at both DS and point-like PS. In the following, the terms low-pass (LP) and high-pass (HP) are used to name the low spatial resolution and residual high spatial frequency components of signals related to both deformation and topography.

The role of the SBAS technique is twofold: on the one hand, it provides the LP deformation time series in correspondence of DS points and the LP DEM-residual topography; on the other hand, the SBAS estimates the residual atmospheric phase delay still affecting the interferometric data after the preliminary correction carried out by leveraging GACOS products.

The temporal displacement associated to PS points is obtained applying the PSI method to interferograms previously calibrated removing the LP topography, deformation and residual atmosphere estimated by the SBAS technique. This strategy connects the PSI and SBAS methods ensuring consistency of deformation results obtained at point-like and DS targets.

A key aspect considered in the framework of the project implementation is related to the estimation and corrections of atmospheric effects affecting the area, generally more evident over the mountainous areas.

An initial correction is applied to each interferogram through the Generic Atmospheric Correction Online Service for InSAR (GACOS), which utilizes the Iterative Tropospheric Decomposition model to separate stratified and turbulent signals from tropospheric total delays, and generate high spatial resolution zenith total delay maps to be used for correcting InSAR measurements. This atmospheric calibration is later refined by the data-driven atmospheric delay estimation in order to obtain atmospheric delay maps at a much higher spatial resolution than that achievable by using external GACOS data.

GNSS data provided by swisstopo, consisting in more than 200 points over Switzerland, are used for the products calibration and later for the result validation during the quality control procedure.

The generated products consist of:

• Line-of-Sight (LOS) surface deformation time series for ascending and descending datasets in SAR geometry;
• Line-of-Sight (LOS) surface deformation time series for ascending and descending datasets in map geometry;
• Combination and projection of deformation results to calculate vertical and east-west deformations.

The quality control (QC) procedures are divided into automatic QC and operator QC. The automatic QC include the analyses of point-wise indicators (coherence maps, precision maps, points density, deformation RMSE with respect to a smooth fitting model), quality indicators at sparse locations (comparison with GNSS data, consistency of stable targets) and other quality indicators (short-time interferogram variograms before and after atmospheric calibration, consistency of overlapping areas). The additional operator QC are focusing on a visual assessment of deformation maps reliability / realism leveraging also on a priori knowledge about the expected deformation behaviour.

The results of this service are delivered to swisstopo that manages the possibility of sharing the deformation maps through their national geo-portal.