2004 Denver Annual Meeting (November 7–10, 2004)

Paper No. 7
Presentation Time: 9:50 AM


SNYDER, Walter S., Division of Earth Sciences, National Sci Foundation, 4201 Wilson Blvd, Rm. 785, Arlington, VA 22230, wsnyder@nsf.gov

CHRONOS is a successful parts of the larger “geoinformatics” system (GI) (cyberinfrastructure applied to the Earth sciences). Based on CHRONOS, other operating GI projects, and numerous workshops, the salient features of the system can be summarized. The most fundamental point is that GI must be a bottom-up, community-driven effort. The funding and construction of the system must be driven by the needs and perspectives of the domain science, but in collaboration with computer scientists. During the initial building and maturing of GI, the implementation and enhancement of current technologies is more important that pursuing “new” or “cutting-edge” CS or IT approaches. GI must be a system - a system that encompasses multiple database sites, real-time observatories, computational modeling centers, and that provides seamless access to the necessary computational resources. It must be easy to use in terms of getting data into and out of the system, and provide the tools necessary for researchers to analyze these data. The community is very adamant that the system will be “distributed”, that there will be multiple centers, hubs, nodes, etc. versus just a few designated mega-sites.

GI will require long-term support, also a very strong point made by the community. It is not simply a task of designing a few unsophisticated databases, quickly loading them up with data, and - bingo - you are done. It will be version 10 before the system is stable. Whereas we tend to think of "infrastructure" in terms of brick-and-mortar facilities, ships, and equipment - GI is also infrastructure - it is a platform for doing science. It will become (is becoming) the basic platform for conducting all future Earth science research. It really is that simple.

It is important to reemphasize that GI is not only about databases, rather it is about a systems approach to cyberinfrastructure. It must become one large, interoperable machine that needs all its parts working together to function properly. Anything short of that risks failure. Several projects, such as CHRONOS, are underway that allows GI to begin to function as a system – one that now is incomplete, but that become a complete system as budgets permit. Most importantly, the success or failure of the system will be the responsibility of both the community and relevant funding agencies.