Lorenzo Colli/Courtesy
A UH team of geoscientists have been awarded 2.5 million hours worth of research on a central processing unit system called Stampede2.
A news release by the University mentions the supercomputer as one of the most powerful in the world. Additionally, the award resources equal a monetary value of more than $2 million, which is synonymous to one year of continuous research.
Stampede2, awarded by the National Science Foundation, is a flagship supercomputer located at the University of Texas at Austin at the Texas Advanced Computing Center.
“The reason why we need this is that we have this computer model and we are really crunching the numbers day in and day out,” said UH College of Natural Sciences and Mathematics assistant professor Lorenzo Colli.
“So if we ran this model on one of our laptops, it would need to run for 2.5 million hours before it would get through the conclusions,” Colli added.
Stampede2 will assist in this research by reenacting mantle convection, such as the churning of hot ductile rocks in the Earth’s interior.
“The ultimate goal, the light at the end of the tunnel, but it is a long tunnel, will be to be better prepared to forecast earthquakes,” Colli said. “But it’s not that with this research we will be able to predict earthquakes one by one.”
Colli is investigating how the mantle has formed in the past, how it is forming right now and hopes to be able to forecast mantle formations in the future.
The team is also looking into the remains of a lost ocean they believe to be sinking toward the center of the Earth.
“We can build a plate reconstruction that we think fits where this ‘lost ocean’ is positioned and how long the remains have been there,” said Department of Earth and Atmospheric Sciences doctoral student Spencer Fuston to UH.
The research Colli and his team are looking at examines the circulation of rocks within the earth mantle instead of looking at air in the atmosphere, and the supercomputer does that by running numerical models.
“The research will also allow us to improve our models and better understand how the Earth’s interior behaves,” Colli said.
“In the long term, the data will help us in dealing with earthquake forecasts and assessing the risk of earthquakes.”
