Correction: The original story stated that NOAA has "20 billion petabytes" of data. This has been changed to "20 petabytes" of data.

The National Oceanic and Atmospheric Administration has more than 20 petabytes of data stored at its Princeton, N.J. data center. Much of the data comes from a grid of sensors located throughout the world's oceans, and other data comes from sensors on bridges, airplanes and buildings. The sensors collect and transmit a continuous stream of information, making data management a key priority at the agency, reports J. Nicholas Hoover at InformationWeek.

Three strategies help ensure that the data can be used effectively, according to NOAA CIO Joe Klimavicz. The efforts include being sure that metadata is accurate, that data is published in standard format, and that the storage strategy is well planned-out. Some of the agency's sites use cutting-edge storage technologies, while others have legacy systems in place.

NOAA gathers 80 TB of science-related data every day, and by 2020 Klimavicz anticipates there will be 10 times that much coming in. One of the agency's goals is to make the resolution on its climate models twice as good, but each time the resolution is doubled it takes 16 times the computing power. High-performance computing remains an ongoing focus for the agency, which installed a 260 TFlop Cray XT6 supercomputer last year, and plans to deploy another supercomputer later this year.

For more:
- see J. Nicholas Hoover's article at InformationWeek

Related Articles:
DARPA plans exascale supercomputer
Data centers growing larger, not prepared for disasters, says industry group