HR: 17:20h
AN: IN24A-06    [Abstracts]
TI: A Federated Model for Distributed Space Science Data Sets
AU: * Rankin, R
EM: rankin@phys.ualberta.ca
AF: University of Alberta, Department of Physics, University of Alberta, Edmonton, AB T6G 2J1 Canada
AU: Pessoa, A
EM: akil.pessoa@ualberta.ca
AF: University of Alberta, Academic ICT, University of Alberta, Edmonton, AB T6G 2J1 Canada
AU: Simon, J
EM: James.Simon@sun.com
AF: Sun Microsystems Inc., 901 San Antonio Road, Palo Alto, Palo Alto, CA 94303 United States
AB: Providing access to globally distributed space science data sets is difficult at present. Significant problems exist in terms of security, rights-access, data format and cross-platform variability. Ideally, a virtual observatory should present users with a common software framework that allows scientific data to be retrieved from various remote sites, as well as allowing users to receive and process data from new sites without compromising the integrity of the application framework. Data repositories must also allow users to store and process scientific data by applying a standard set of Algorithms and user-supplied tools. Systems must also be able to deliver data and data products to the respective scientific community in desirable formats, as well as providing statistics of useage and access to data providers. In this paper, we present experience in developing a grid-based Space Sciences Data Portal (SSDP) for Canadian space science data sets collected under the banner of Canadian Geospace Monitoring. We describe the software framework under development, and illustrate its planned expansion through three phases. The third phase brings the SSDPinto a federated computing model, and involves the development of a Resource Broker that greatly expands the accessbility of SSDP data sets, allowing them to be integrated into a global framework. This phase allows processing on grids that are distributed globally, facilitating the execution of complex Algorithms that can be applied to data in a collaborative global setting.
DE: 9810 New fields (not classifiable under other headings)
DE: 9820 Techniques applicable in three or more fields
SC: Earth and Space Science Informatics [IN]
MN: Fall Meeting 2005