AstroGrid3; Bristol component

By the end of 2007, the AstroGrid project will have completed development of all the core infrastructure software components needed for an operational Virtual Observatory (VO) system, and will have had experience setting up and running a prototype working system. Using both the AstroGrid software product, and the skills of key data centre staff, we propose to construct a fully operational and sustainable VO service for UK astronomers. This will involve several areas of work. (i) Setting up core VO services at each data centre site - Registry, Job Execution Service, MySpace, and Community Authentication. (ii) Publishing an agreed list of specific prime data services and applications, to VO standards and with VO interfaces. (iii) Working with the astronomy community to optimise science return from key missions, using the VO - including user guides, helpdesk, and training workshops. (iv) Working with system administrators and tools writers across all UK departments, providing training workshops, automated and reliable auto-install versions of any components that they wish to deploy, and easy methods for tools in any language to interact with VO services. (v) Maintaining and upgrading the core software product to engineering standards. New functionality will be added, and re-engineering will be necessary as the standards emerging from the IVOA and the rest of the e-Science world mature. (vi) Keeping abreast of the changing technology landscape, assessing new technologies and constructing prototypes of possible new infrastructure components. The service will be run for the benefit of all astronomers in the UK, but its delivery is centred on a handful of key centres. The consortium consists of institutions which were involved in constructing the AstroGrid product, and which are closely involved in the delivery in just about all of the key databases expected over the next decade - such as XMM, WFCAM, VISTA, SDO, Planck, GAIA, ALMA, e-MERLIN, and WFMOS. Ensuring that these databases are available through VO interfaces is the main aim of this project. Huge investment has been made in building each of these facilities or missions, and substantial investment has been or will soon be made in the basis data management for these missions - pipelining and archiving the data. Integrating all these datasets into the VO infrastructure, and setting up and operating core VO services such as data discovery, workflow, community storage, single sign on, and tool interoperability, will add substantially improved science for relatively little money. Most of the cost is in the staff required to install, maintain, and improve services, with a small hardware investment for running core services and improved networking.