The Virtual Earthquake and seismology Research Community e-science environment in Europe (VERCE): a European Research Infrastructure project
Abstract
The nature of science in seismology is changing - new discoveries will emerge from statistical analysis and modeling (inversion, assimilation) of large amounts of data generated from dense observational and monitoring networks and from large-scale wave propagation simulations. In many cases our ability to acquire observational and synthetic data outpaces today our ability to process and analyze them. Addressing these challenges requires a new and holistic approach with important augmented societal applications in seismic hazard assessment and monitoring, and exploration geophysics. VERCE is a four-year FP7-INFRASTRUCTURE project, with a consortium of ten partners from seismology and computer science, and contributes to the e-science infrastructure of the European Plate Observatory System (EPOS), the ESFRI initiative of the solid Earth community in Europe. We report here the progress of VERCE toward a service-oriented architecture and a platform of services and tools - integrating European computing and data infrastructures with the distributed seismological data archives - in support for data-intensive analysis and modeling applications. Two prototype applications were selected within VERCE: a data-intensive analysis application based on seismic-noise correlation; and a data-intensive HPC wave simulation application. Both applications consist of multiple phases where ingestion interleaves with data processing and analysis, generating highly parallel and asynchronous data workflows together with massively parallel data access. VERCE efforts are devoted in particular to providing efficient scalable and transparent distributed data management and data transfer services, together with execution models that enable data processing and analysis computation to overlap with data transfer and I/O operations, thereby achieving high throughput under heavy asynchronous access to data. We discuss the current progress of VERCE in enabling the application prototypes, and in developing a scientific gateway that will provide them as a service to the seismology community. We will also discuss elements of the next phase plan and implementation roadmap.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFMGC13D1124V
- Keywords:
-
- 0525 COMPUTATIONAL GEOPHYSICS Data management;
- 1902 INFORMATICS Community modeling frameworks;
- 1916 INFORMATICS Data and information discovery;
- 7290 SEISMOLOGY Computational seismology