[ Roll & Rock ] [ Main Menu ]
23052
From: ryan, [DNS_Address]
Subject: an AI algorithm that correctly predicted 70% of earthquakes a week bef
URL: https://news.utexas.edu/2023/10/05/ai-driven-earthquake-forecasting-shows-promise-in-trials/
|
+ Researchers at UT Austin have developed an AI algorithm that correctly predicted 70% of earthquakes a week before they happened during a seven-month trial in China. This map shows the location in China of the AI-predicted earthquakes (blue dots) joined by a red line to where each actual earthquake happened (red dots). The numbers indicate the week the earthquake occurred. During the 30-week trial, the UT Austin-developed AI missed only one earthquake. Credit: Yangkang Chen/Jackson School of Geosciences. AI-Driven Earthquake Forecasting Shows Promise in Trials Two color orange horizontal divider Photo of a large break in the road, one side is about a foot higher than the other. A road sign reads 'Shoulder Closed' and there are road workers in the background. Damage after an earthquake hit New Zealand in 2011. Researchers are working to forecast earthquakes with AI. Credit: Flickr/Martin Fluff. AUSTIN, Texas — A new attempt to predict earthquakes with the aid of artificial intelligence has raised hopes that the technology could one day be used to limit earthquakes’ impact on lives and economies. Developed by researchers at The University of Texas at Austin, the AI algorithm correctly predicted 70% of earthquakes a week before they happened during a seven-month trial in China. The AI was trained to detect statistical bumps in real-time seismic data that researchers had paired with previous earthquakes. The outcome was a weekly forecast in which the AI successfully predicted 14 earthquakes within about 200 miles of where it estimated they would happen and at almost exactly the calculated strength. It missed one earthquake and gave eight false warnings. It’s not yet known if the same approach will work at other locations, but the effort is a milestone in research for AI-driven earthquake forecasting. “Predicting earthquakes is the holy grail,” said Sergey Fomel, a professor in UT’s Bureau of Economic Geology and a member of the research team. “We’re not yet close to making predictions for anywhere in the world, but what we achieved tells us that what we thought was an impossible problem is solvable in principle.” Map with an inset showing a larger map of China. Red and blue dots are scattered around. Each blue dot is connected to a red one with a number next to it. Some dot pairs are close together, others are hundreds of kilometers apart. The trial was part of an international competition held in China in which the UT-developed AI came first out of 600 other designs. UT’s entry was led by bureau seismologist and the AI’s lead developer, Yangkang Chen. Findings from the trial are published in the journal Bulletin of the Seismological Society of America. “You don’t see earthquakes coming,” said Alexandros Savvaidis, a senior research scientist who leads the bureau’s Texas Seismological Network Program (TexNet) — the state’s seismic network. “It’s a matter of milliseconds, and the only thing you can control is how prepared you are. Even with 70%, that’s a huge result and could help minimize economic and human losses and has the potential to dramatically improve earthquake preparedness worldwide.” The researchers said that their method had succeeded by following a relatively simple machine learning approach. The AI was given a set of statistical features based on the team’s knowledge of earthquake physics, then told to train itself on a five-year database of seismic recordings. Once trained, the AI gave its forecast by listening for signs of incoming earthquakes among the background rumblings in the Earth. “We are very proud of this team and its first-place finish in this prestigious competition,” said Scott Tinker, the bureau’s director. “Of course, it’s not just location and magnitude, but timing that matters as well. Earthquake prediction is an intractable problem, and we can’t overstate the difficulty.” The researchers are confident that in places with robust seismic tracking networks such as California, Italy, Japan, Greece, Turkey and Texas, the AI could improve its success rate and narrow its predictions to within a few tens of miles. One of the next steps is to test the AI in Texas since the state experiences a high rate of minor- and some moderate-magnitude earthquakes. The bureau’s TexNet hosts 300 seismic stations and more than six years of continuous records, which makes it an ideal location to verify the method. Eventually, the researchers want to integrate the system with physics-based models, which could be important where data is poor, or places such as Cascadia, where the last major earthquake happened hundreds of years before seismographs. The research was supported by TexNet, the Texas Consortium for Computational Seismology and Zhejiang University. The bureau is a research unit of the Jackson School of Geosciences. |
23062
From: shatterbrain, [DNS_Address]
Subject: Re: an AI algorithm that correctly predicted 70% of earthquakes a...
| Me thinks strange animal and pet behavior still has a better track record.😉 |
Responses:
None
23058
From: blindhog 6th sense, [DNS_Address]
Subject: I'm Confused, So Just Asking:
|
The 4th paragraph down says "the AI algorithm correctly predicted 70% of earthquakes a week before they happened during a seven-month trial in China." Then the 5th paragraph says "The outcome was a weekly forecast in which the AI successfully predicted 14 earthquakes within about 200 miles of where it estimated they would happen and at almost exactly the calculated strength. It missed one earthquake and gave eight false warnings." Ok, there 14 they supposedly got correct, 1 they missed, and 8 were false. 14+1+8=23. 14 correct/accurate out of 23 is only 60.87% correct/accurate, so how do they come up with 70% accuracy? IMO, a 200 mile radius and 60.87% accuracy might be a hard sell to shut down and evacuate an entire city. |
Responses:
None
23054
From: EQF, [DNS_Address]
Subject: Re: an AI algorithm that correctly predicted 70% of earthquakes a...
|
Posted by EQF on November 5, 2023 It appears that you have been visiting the phys.org Web site. I was planning to post a note myself regarding that report. I have also been planning to contact people associated with the report and have put together a list of 10 E-mail addresses to use to contact them. That E-mail is scheduled to be sent out tomorrow. The subject matter of the report represents in my opinion a major breakthrough if it is accurate. Here are two addresses that I know of associated with that report: https://phys.org/news/2023-10-ai-driven-earthquake-trials.html https://pubs.geoscienceworld.org/ssa/bssa/article-abstract/doi/10.1785/0120230031/627949/Earthquake-Forecasting-Using-Big-Data-and?redirectedFrom=fulltext There is something that I often refer to as "Geologists' Disease." It is the fact that they are almost universally missing an important word in their vocabularies. That word is "URGENT." My theory is that scientists generally move at around the same speed as what they study. In my case I am an experienced and highly successful disaster mitigation professional. Disasters don't wait around for anyone's permission to occur. Geologists generally study ROCKS. Question: How fast do rocks move unless there is an earthquake, a landslide, or a volcano eruption? Answer: Not too fast. Hence, the geologists don't seem to me to be in much of a hurry to learn how to predict earthquakes. Comments that I plan to send those researchers. 1. I have often recommended that researchers should test their forecasting programs in Japan rather than anywhere else. Japan has frequent earthquakes that are associated with every type of fault zone that exists. They can also be extremely powerful and destructive. In contrast, California and China don't have nearly as many earthquakes in a given area. 2. Japan has highly advanced communications systems. That is important if you want to collect, process, and circulate information regarding earthquake precursors and forecasts. 3. Some earthquakes appear to be almost totally random events. It could be almost impossible to accurately predict them. 4. Some earthquakes occur soon after a powerful earthquake occurs in a nearby or even distant area. This likelyl involves the rapid transfer of strain from one fault zone to another after the first earthquake occurs. It also makes those earthquakes difficult to predict in a timely manner. 5. Some earthquakes are reportedly associated with powerful solar storms. The interaction mechanism has not been established as far as I am aware. One theory proposes that this is associated with fault zone heating as the result of solar storm electric currents in the Earth's crust. The rock layers expand causing pressure to build within them. The additional abrupt pressure causes the fault zone to fracture. I have an alternate theory for those relationships. 6. My own earthquake forecasting computer program and data are likely just as valuable as theirs. 7. They reportedly came in first place with their AI programs in a contest in the People's Republic of China. With humor intended, that isn't fair. Roger and I got there first. His and my data and computer programs were discussed quite a few years ago in detail by one of my earthquake forecasting colleagues at a disaster management conference in Bejing, China. I didn't know about that recent competition. Otherwise I would have submitted an entry. I would be willing to bet that my Earthquake Forecasting computer program would have come in first place, or at worst, second place. In keeping with Mr. Bopp's wishes regarding subject matter posted to this present forum, I posted my ENERGY ISLANDS Web page note to the following Web page instead of here: http://www.earthboppin.net/talkshop/science/messages/7345.html These are personal opinions. Regards to all, EQF |
Responses:
None
23053
From: Skywise, [DNS_Address]
Subject: Re: an AI algorithm that correctly predicted 70% of earthquakes a...
|
This is very interesting. Too bad the paper is behind a paywall. Something I would like to know is how big an area the predictions were for, as from the map alone it appears some of the claimed hits were hundreds of miles from the predicted location. And not even on the same fault systems. Not saying AI couldn't lead to some predictive capability by 'seeing' patterns and connections that our minds haven't noticed. Brian |
Responses:
None
[ Roll & Rock ] [ Main Menu ]