HR: 0830h
AN: U41B-0019    [PDF]
TI: The IRIS Model: Building the Infrastructure for Seismology
AU: * Benson, R B
EM: rick@iris.washington.edu
AF: IRIS Data Management Center, 1408 NE 45th St, Suite 201, Seattle, WA 98105 United States
AU: Ahern, T K
EM: tim@iris.washington.edu
AF: IRIS Data Management Center, 1408 NE 45th St, Suite 201, Seattle, WA 98105 United States
AB: The IRIS Consortium began a global, cooperative development in the mid-1980's, with the goal of being able to provide the necessary infrastructure (system) that would bring expensive, globally distributed geophysical data into the arena of open access, and to make these data available through a streamlined system of accessing that has profoundly improved the science of seismology. Historically, geophysical data were generated by scientists studying a particular problem, were limited in both the spatial and the temporal scales, and were self-managed and in native formats that hindered exchange. Data had little chance of being used again. The IRIS Data Management Center, located in Seattle, WA and affiliated with the University of Washington, operates an NSF-funded facility that is charged with acquiring, archiving, and distributing over 30 years (and counting) of geophysical time series data. Even though this volume exceeds 50 terabytes of dual-sorted data, these data can be requested on any spatial or temporal time scale. If you are studying the movement of mantle convection cells over decadal time scales, or studying aftershocks of a local earthquake, you can access these data from your workstation, convert these into your analysis format, and begin the process of doing hard science without ever having to burden yourself with acquiring the data, or drafting a detailed proposal justifying your area of interest. You can proceed directly to doing science. In the last two years, the IRIS DMC has targeted the problem of handling real time data generated in the field, forwarding it to a unified disk buffer, thereby eliminating all heterogeneity, and providing open access to large volumes of data which can be immediately utilized, providing the back-end for decision making in human time scales, not scientific time scales. We have now pushed this to the final step, closing the data loop and providing robust utilities that enable information-to-application functionality, where applications can now access these disk buffers through what we call the "Data Handling Interface (DHI)", so you no longer have to cache data locally. This has shortened the timeline for doing high quality research considerably. It's a community-driven system, is focused directly with dealing with end users, and is truly, end to end. This poster will outline the standards and the methods for how cooperative initiatives are drafted and supported, and give an overview of the process that allows homogeneous, globally distributed tools to function, and in the end, illustrate that this has been successfully integrated into the science of seismology, creating the fabric of modern seismology.
UR: http://www.iris.washington.edu
DE: 7299 General or miscellaneous
SC: U
MN: 2003 Fall Meeting