Installing a mode-matching mirror and suspension into a vacuum chamber during LIGO construction.
Image Courtesy LIGO Laboratory |
Albert Einstein predicted the existence of gravitational waves in 1916, and for the past three decades a succession of experiments has attempted to directly detect the waves with no success. Now using the Laser Interferometer Gravitational Wave Observatory and grid computing, the 400 scientists and engineers of the LIGO Scientific Collaboration hope to finally detect them.
"Gravity waves are fluctuations in the space-time continuum created whenever large masses move," said Kent Blackburn, senior scientist at the California Institute of Technology. "Even such catastrophic events as colliding black holes produce barely detectable waves, and these events are what the LSC searches for with the world's two most sensitive gravitational wave detectors."
Many sensors monitor the detectors and their environment so that a gravity wave can be distinguished from an earthquake or a dropped hammer. All this information adds up to a lot of data—the detectors have produced 400 terabytes of data since they started operating in 2002. To determine if a signal is a gravitational wave, the data are constantly being compared against tens of thousands of known signals. Grid computing provides the computational power to run these constant comparisons and a way to replicate and locate data.
"Until now, the primary use for the LIGO Data Grid has been data replication and discovery," said Scott Koranda, staff scientist at the University of Wisconsin-Milwaukee. "We use grid tools to ensure that all nine LDG computing sites have a copy of the interesting data, and researchers at the other 36 LSC institutions use the LDG to find the data they need."
LDG developers are now testing the Virtual Data System tool Pegasus, which creates the grid workflow necessary for a user to execute an analysis. Pegasus will also automatically find and transfer the necessary data to available computing resources on the LDG or the Open Science Grid.
"With Pegasus, the scientist will only need to worry about what analysis they want to run, not where the data is or where the computing will be done," explained Koranda. "We hope to have it running by the fall, in time for LIGO's first yearlong science run at full sensitivity."
Learn more at the LIGO Web site.
—Katie Yurkewicz
e-mail this article
|
