We use cookies on this website. By continuing to use this site without changing your cookie settings, you agree that you are happy to accept our privacy policy and for us to access our cookies on your device.
Researchers at the University of Texas are using artificial intelligence to better forecast earthquakes. From Austin, the Texas capital, Deana Mitchell has our story in this week’s episode of LogOn.
Content TypePackage
LanguageEnglish
Transcript/Script((PLAYBOOK SLUG: Log On: Earthquake Forecasting
HEADLINE: US Scientists Using AI to Forecast Earthquakes
TEASER:
PUBLISHED AT: 02/06/2024 AT 9:45AM
BYLINE: Deana Mitchell
CONTRIBUTOR:
DATELINE: Austin, Texas
VIDEOGRAPHER:
VIDEO EDITOR: Deana Mitchell
ASSIGNING EDITOR: Stearns
SCRIPT EDITORS: Stearns, Reifenrath
VIDEO SOURCE (S): Reuters, VOA
PLATFORMS (mark with X): WEB __ TV _X_ RADIO _X_
TRT: 1:58
VID APPROVED BY: MAS
TYPE: TV/R
EDITOR NOTES: ))
((INTRO)) [[Researchers at the University of Texas are using artificial intelligence to better forecast earthquakes. From Austin, the Texas capital, Deana Mitchell has our story in this week’s episode of LogOn.]]
((NARRATOR))
Devastating earthquakes strike all over the world without warning.
((NARRATOR))
Sergey Fomel is a professor of geophysics at the University of Texas at Austin.
((Sergey Fomel, UT Austin Geophysics Professor))
“Predicting earthquakes is the holy grail of seismology. People have been trying to do it for more than 100 years.’
((NARRATOR))
A team at UT developed a data-driven model for forecasting natural destructive earthquakes using artificial intelligence and big data analysis
and applied it to a real-time earthquake prediction competition that lasted two years.
The team leader, Yangkang Chen, an assistant professor at UT's Bureau of Economic Geology, says they were able to forecast earthquakes with an accuracy of 70 percent including magnitude, location, and occurrence, winning the competition.
((Yangkang Chen, UT Bureau of Economic Geology))
“Because of the competition, we had a very good dataset–and testbed most importantly, that's a real-time testbed.”
((NARRATOR))
Chen said the AI model takes input and output predictions and finds patterns, something too difficult for humans to do on their own.
The university team works with TexNet, the Texas seismological network, where Alexandros Savvaidis is a principal investigator.
((Alexandros Savvaidis, TexNet Principal Investigator))
“We could identify which of the parameters of the ground motion or the electromagnetic sensor system is changing prior to the earthquakes.And because of that, we can see when in the future those parameters are changing again. And when they change, then we expect to have an earthquake.”
((NARRATOR))
The team said it has predicted some earthquakes a week in advance, but most were much closer to the quake.
((Sergey Fomel, UT Austin Geophysics Professor))
“The most accurate prediction may come just immediately before the earthquake, but for early warning systems, it is sufficient to have just a little bit of time. “
((NARRATOR))
The team is using more real-time test beds in various geographical locations with the hope of forecasting earthquakes further in advance.
((Deana Mitchell, VOA News, Austin, Texas))
NewsML Media TopicsArts, Culture, Entertainment and Media
Rights“Rights are not granted for this content on YouTube. Posting of this content on YouTube is strictly prohibited.”