Massive, systematic and automatic generation of Sentinel-1 deformation time series via the P-SBAS DInSAR processing chain

The Sentinel-1 (S1) constellation is a family of satellites designed to collect C-band SAR data in continuity with the previous ERS-1/2 and ENVISAT missions, within the framework of the Copernicus Programme of the European Union, with the aim to detect and analyze Earth's surface displacements. S1 is characterized by significant enhancements in terms of spatial coverage, revisit time, timeliness and service reliability. In particular, S1 Interferometric Wide Swath (IWS) scenes are collected through the innovative acquisition mode referred to as Terrain Observation by Progressive Scans (TOPS) [1], which allows a considerable improvement of the range coverage (of about 250 km) with respect to the conventional Stripmap mode, and, at the same time, a significant increase of the acquired SAR data size (around 10 times greater than ERS and ENVISAT scenes). Moreover, the constellation is nowadays made up of two twin sensors (Sentinel-1A and Sentinel-1B) that acquire images all around the world with a repeat pass of 6 days in most areas, thus leading to the creation of large SAR data archives, which can be exploited according to a "free and open" data distribution policy. Such characteristics require the development of innovative and appropriate solutions aimed at handling these huge SAR data archives, more and more increasing in terms of both temporal and spatial coverage, to effectively and routinely exploit the advanced DInSAR methodologies. In this work we present a strategy to perform massive, systematic and automatic analysis of S1 SAR data via the generation of deformation time series. In particular, we propose a solution, based on the Parallel version of the well-known SBAS algorithm (P-SBAS) [2][3], properly designed for processing S1 SAR data. Our solution takes advantage of the P-SBAS characteristics to run on distributed computing infrastructures (i.e., cluster, grid, cloud) by making use of both multi-core and multi-node programming techniques and exploiting an "ad - hoc" designed distributed storage [4]. Moreover, it strongly takes into account the data characteristics of the TOPS mode. Indeed, IWS scenes consist of series of bursts that can be considered as independent, separate acquisitions. This makes a large part of the processing inherently parallel at a burst granularity level; such a condition implies that the processing time can be significantly reduced when large computing resources are available. The developed S1 P-SBAS processing chain is exploited to generate mean deformation velocity maps and corresponding deformation time series of South Italy. In particular, we processed six S1 interferometric SAR data stacks, acquired along descending orbits (tracks 22, 124, 51) and spanning the time interval October 2014 - September 2016. Starting from these data (almost 300 slices), we generated 850 differential interferograms with 5 looks in azimuth and 20 in range directions, thus resulting in a pixel size of approximately 60 x 60 m. In Figure 1 we display the retrieved LoS mean deformation velocity map, geocoded and superimposed on an optical image of the area. The map reveals the presence of localized displacements associated, for instance, to the volcanic activity of the caldera of Campi Flegrei and Mt. Etna volcano. The achieved results demonstrate the capability of the presented processing chain to effectively deal with massive amount of data to generate advanced DInSAR products aimed at detecting displacements at very large spatial scale. Moreover, they show that S1 P-SBAS can be exploited to build up operational services for the easy and rapid generation of advanced interferometric products that can be very useful within risk management and natural hazard monitoring scenarios. Acknowledgments This work has been supported by the Ministry of Economic Development - DGS-UNMIG (Directorate-General for Safety of Mining and Energy Activities - National Mining Office for Hydrocarbons and Georesources), the Italian Department of Civil Protection, the European Union Horizon 2020 research and innovation programme under the EPOS-IP project (grant agreement No 676564), the ESA GEP (Geohazards Exploitation Platform) and I-AMICA (Infrastructure of High -55- Technology for Environmental and Climate Monitoring - PONa3_00363) projects. Sentinel-1 data are copyright of Copernicus (2016). The DEMs of the investigated zone were acquired through the SRTM archive. References [1] F. De Zan and A. M. Monti Guarnieri, "TOPSAR: Terrain Observation by Progressive Scans," IEEE Transactions on Geoscience and Remote Sensing, vol. 44, no. 9, pp. 2352-2360, Sept. 2006. [2] Berardino, P., Fornaro, G., Lanari, R., Sansosti, E., "A new Algorithm for Surface Deformation Monitoring based on Small Baseline Differential SAR Interferograms", IEEE Trans. Geo. Rem. Sens., 40, 11, pp. 2375-2383, 2002. [3] F. Casu, S. Elefante, P. Imperatore, I. Zinno, M. Manunta, C. D. Luca, and R. Lanari, "SBAS-DInSAR Parallel Processing for Deformation Time-Series Computation," Selected Topics in Applied Earth Observations and Remote Sensing, IEEE Journal, 2014. [4] Zinno, S. Elefante, L. Mossucca, C. De Luca, M. Manunta, O. Terzo, R. Lanari, and F. Casu, "A First Assessment of the P-SBAS DInSAR Algorithm Performances Within a Cloud Computing Environment," IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., pp. 1-12, 2015.