During the last 25 years, the Differential Synthetic Aperture Radar Interferometry (DInSAR) has played an important role for understanding the Earth's surface deformation and its dynamics. In particular, the large collections of SAR data acquired by a number of space-borne missions (ERS, ENVISAT, ALOS, RADARSAT, TerraSAR-X, COSMO-SkyMed) have pushed toward the development of advanced DInSAR techniques for monitoring the temporal evolution of the ground displacements with an high spatial density. Moreover, the advent of the Copernicus Sentinel-1 (S1) constellation is providing a further increase in the SAR data flow available to the Earth science community, due to its characteristics of global coverage strategy and free and open access data policy. Therefore, managing and storing such a huge amount of data, processing it in an effcient way and maximizing the available archives exploitation are becoming high priority issues. In this work we present some recent advances in the DInSAR field for dealing with the effective exploitation of the present and future SAR data archives. In particular, an efficient parallel SBAS implementation (namely P-SBAS) that takes benefit from high performance computing is proposed. Then, the P-SBAS migration to the emerging Cloud Computing paradigm is shown, together with extensive tests carried out in the Amazon's Elastic Cloud Compute (EC2) infrastructure. Finally, the integration of the P-SBAS processing chain within the ESA Geohazards Exploitation Platform (GEP), for setting up operational on-demand and systematic web tools, open to every user, aimed at automatically processing stacks of SAR data for the generation of SBAS displacement time series, is also illustrated. A number of experimental results obtained by using the ERS, ENVISAT and S1 data in areas characterized by volcanic, seismic and anthropogenic phenomena will be shown. This work is partially supported by: the DPC-CNR agreement, the EPOS-IP project and the ESA GEP project.
Unsupervised SBAS-DInSAR Processing of Space-borne SAR data for Earth Surface Displacement Time Series Generation
Casu F;De Luca C;Lanari R;Manunta M;Zinno;
2016
Abstract
During the last 25 years, the Differential Synthetic Aperture Radar Interferometry (DInSAR) has played an important role for understanding the Earth's surface deformation and its dynamics. In particular, the large collections of SAR data acquired by a number of space-borne missions (ERS, ENVISAT, ALOS, RADARSAT, TerraSAR-X, COSMO-SkyMed) have pushed toward the development of advanced DInSAR techniques for monitoring the temporal evolution of the ground displacements with an high spatial density. Moreover, the advent of the Copernicus Sentinel-1 (S1) constellation is providing a further increase in the SAR data flow available to the Earth science community, due to its characteristics of global coverage strategy and free and open access data policy. Therefore, managing and storing such a huge amount of data, processing it in an effcient way and maximizing the available archives exploitation are becoming high priority issues. In this work we present some recent advances in the DInSAR field for dealing with the effective exploitation of the present and future SAR data archives. In particular, an efficient parallel SBAS implementation (namely P-SBAS) that takes benefit from high performance computing is proposed. Then, the P-SBAS migration to the emerging Cloud Computing paradigm is shown, together with extensive tests carried out in the Amazon's Elastic Cloud Compute (EC2) infrastructure. Finally, the integration of the P-SBAS processing chain within the ESA Geohazards Exploitation Platform (GEP), for setting up operational on-demand and systematic web tools, open to every user, aimed at automatically processing stacks of SAR data for the generation of SBAS displacement time series, is also illustrated. A number of experimental results obtained by using the ERS, ENVISAT and S1 data in areas characterized by volcanic, seismic and anthropogenic phenomena will be shown. This work is partially supported by: the DPC-CNR agreement, the EPOS-IP project and the ESA GEP project.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
