Back to Top

December 22, 2008 Vol. 1, Issue 12


Dr. Charles Kennel, former NASA Associate Administrator for the Mission to Planet Earth, recalled the origins of the Earth Observing System Data and Information System.

[Editor’s Note: Dr Charles Kennel served as the NASA Associate Administrator for the Mission to Planet Earth from 1994 to 1996. In a recent interview with ASK Magazine Managing Editor Don Cohen, he recounted the differences of opinion within the earth science community about the design of the Earth Observing System Data and Information System (EOS-DIS). The full interview will appear in the Winter 2009 issue of ASK Magazine.]

Kennel: In 1988 when NASA scientists were designing the data management system for the EOS spacecraft, they understood that unprecedented volumes of information would be collected that would have to be integrated. They designed a system called EOS-DIS: the Earth Observing System Data and Information System. The idea was to have eight national data analysis centers that would get the data pertinent to the earth science disciplines down from the different spacecraft. They would all be connected one to the other with a data pipe that NASA would construct. In addition, NASA had the idea that they should do end-to-end data management from the moment that information leaves the spacecraft to error-correction and adding information about the spacecraft, conversion to physical variables and then conversion to disciplinary and multi-disciplinary products. For its day, it was an advanced conception. It still exists today and is still state of the art. What was done was substantial.

But I had a revolt among the scientists who were saying, “The internet is coming. What we really need is large distributed data pipes to all the research centers related to the program. NASA should focus on the Google function — the smart search engines.” We did try to put some extra money into the science data centers at different labs around the country, but we never succeeded in making a fully integrated, high capacity, Internet-based data analysis system. In the future, the means for doing that are going to be there.

The National Science Foundation has created a system call National LambdaRail. They’ve discovered a technique to put forty conversations at a time on a single optical pipe. We now have an experimental 40 GB/sec data pipe that connects most of the large research laboratories in the United States. So now the vast quantities of data that the earth observing system is collecting can be transmitted and used by a diverse collection of people, not all of them funded on this network by NASA. We hope that the next observing system will rely on a much more network-centric approach.

About the Author

Share With Your Colleagues