Elucidating Earth's Interior Through Advanced Teleseismic Phase Picking
1. The validation with two teleseismic phases, PcP and PKiKP, demonstrate high detection accuracies and low picking errors.
2. Deep-learning picking of first peaks is more accurate than picking first breaks of teleseismic phases.
3. This proposed scheme would enhance mining of teleseismic phases and probing of Earth's interior structures and their dynamics.
How do scientists explore Earth's hidden interior — its crust, mantle, and core? The answer lies in earthquake waves. Like X-rays, these seismic waves travel through and reflect off internal structures, allowing scientists to visualize the planet's interior. By analyzing these waveforms, seismologists can create 'B-scans' of structure or 'CT scans' of physical properties.
Identifying these reflections in seismic data, however, is complex and time-consuming. Reflected phases can be distorted by structural discontinuities and local heterogeneities, making it easy to mistake noise for real signals. This challenge underscores the need for efficient and accurate automatic phase pickers to handle the vast volume of teleseismic earthquake data, whose phases often reflect off deep structures within the Earth.
A recent study published in KeAi's Artificial Intelligence in Geosciences (KeAi) introduces a new deep learning-based workflow to automatically detect and pick teleseismic phases with high efficiency and accuracy.
'To improve the workflow's performance, we divide it into three parts: phase preparation, detection, and picking,' explains leadingco-author Dr. Congcong Yuan, a postdoctoral researcher at Cornell University. 'We apply physical constraints during preparation to highlight potential signals. The detection step filters out low-quality data, so that the final picking step can determine arrival times more accurately and without bias.'
'This approach allows fast, reliable and robust teleseismic phase processing. 'It enables us to extract more meaningful data, helping us better understand the physics and dynamics deep inside the Earth,' adds Yuan.
'Yuan and colleagues are collaborating with another research team in teleseismic imaging to apply this method to specific tectonic regions. 'The more high-quality phase data we have, the more we can uncover about Earth's inner workings,' he says.
'For decades, seismologists have faced the tedious task of processing seismic data,' notes co-author Prof. Jie Zhang from the University of Science and Technology of China. 'With the rise of deep learning, seismology is reaching a turning point from semi-automatic workflows to truly autonomous systems.'
References
DOI
10.1016/j.aiig.2025.100108
Original Source URL
https://doi.org/10.1016/j.aiig.2025.100108
Lucy Wang
BioDesign Research
email us here
Legal Disclaimer:
EIN Presswire provides this news content 'as is' without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Gizmodo
5 days ago
- Gizmodo
Researchers Uncover Hurricane Sandy's Hidden Death Toll
More than a decade after Hurricane Sandy flooded the East Coast and killed over 200 people, a new study has revealed the storm's lasting impact on mortality risk among older adults. The researchers linked hurricane flood exposure to a 9% increase in risk of death for people over the age of 65 who continued living in the area after landfall. Experts estimate that Hurricane Sandy caused over $60 billion in damage after it made landfall in 2012. Record-breaking storm surge and catastrophic flooding inundated the coastlines of New York and New Jersey, destroying hundreds of thousands of homes. Over a decade later, affected communities have largely rebuilt, but the storm's invisible impact on senior mortality risk endures, according to the new study. The findings, published Wednesday, August 6, in the journal Frontiers in Public Health, underscore the need to consider how natural disasters impact public health over the long term, particularly within vulnerable populations, the researchers say. 'Hurricanes and associated flooding and damage impact the lives of those around them; fixing homes, dealing with the stress associated with this all can have long-lasting impacts, which may be seen in terms of mortality,' Arnab Ghosh, senior author and assistant professor of medicine at Cornell University, told Gizmodo in an email. 'We also know from other data from hurricanes that the environment changes in areas that are impacted. Socially, people move away; economically, businesses close. These factors can also limit social networks and friendships, all of which are particularly important in older adults,' Ghosh added. He and his colleagues split the areas that were most affected—New York State, New Jersey, Connecticut, and New York City—into 959 zip code tabulation areas. ZCTAs are generalized representations of zip codes that public health researchers use to group demographic and health information by location. Of the 959 ZTCAs, 454 had flooded during Hurricane Sandy, and 505 were within a 10-mile (16-kilometer) radius of flooded areas. The researchers then analyzed health data from nearly 300,000 residents aged 65 or older who were enrolled in Medicare and continued to reside in the same zip code area from 2013 to 2017. To isolate the effects of hurricane-related flooding, they controlled for age, gender, race, and living circumstances, as well as other demographic and socioeconomic factors. The analysis revealed that Medicare beneficiaries who stayed in flooding-impacted ZCTAs had a significantly higher risk—9% on average—of death from all causes compared to those living in areas unaffected by flooding. That increase persisted for up to 5 years after the hurricane hit. Localized increases in mortality risk varied by region, but not in ways the researchers expected. Medicare beneficiaries living in flooded parts of Connecticut and New York City showed the greatest increases in risk of death: 19% and 8%, respectively. Ghosh and his colleagues were surprised to see that seniors living in affluent Connecticut fared worse than those living in New York City, where socioeconomic and demographic characteristics may result in greater exposure to natural disasters. Further research will need to investigate the reasons behind this, but Ghosh suspects it may be partly related to lack of hurricane preparedness in Connecticut, as this state rarely sees tropical storms. What's more, hard-hit New Jersey and New York State did not show significant hurricane flooding-related effects on seniors' all-cause mortality. 'We were surprised by this finding,' Ghosh said. In New Jersey, this may be due to the fact that most coastal residences are single-family vacation homes, limiting the consequences of flood exposure, he explained. Ghosh hopes that this data will aid disaster relief and preparedness efforts at the local and federal levels. Organizations need to think about not just the short-term outcomes but also the long-term health problems that stem from hurricanes, he said. 'We certainly need more policies and infrastructure, particularly as hurricanes become more intense, [frequent], and dangerous.'
Forbes
5 days ago
- Forbes
Should Kids Use ChatGPT AI For School? Parents Are Divided
Two-thirds of American students now use AI for schoolwork multiple times a week. That's not a future prediction. It's already happening. According to a new survey from Preply, an online language learning platform, students aren't just experimenting with AI. They're integrating it into how they learn, study, and complete assignments. Thirty-six percent use AI both in class and at home. And across Arkansas, Mississippi, and Texas, student AI usage is outpacing national averages. If you're wondering whether students and parents are worried, 90 percent say they feel confident in their ability to use AI responsibly. This is the new classroom: powered by ChatGPT, Google Gemini, X's Grok and Microsoft Copilot, and focused on high-stakes subjects like English, math, and history. The question now isn't if students will use AI, but how well we're preparing them to think alongside it. The Cognitive Tradeoff Of AI No One Talks About While the benefits of AI in education are clear—on-demand help, instant feedback, and access to powerful research tools—there's a growing body of evidence suggesting that these gains might come at a hidden cost. Emerging neuroscience research suggests that relying on AI for language tasks may lead to reduced engagement in the prefrontal cortex — the brain region responsible for reasoning, attention, and decision-making. Supporting this concern, a study by Cornell University and Microsoft Research found that participants who received AI writing assistance performed worse in follow-up critical thinking tasks compared to those who completed the work independently. In short, when AI does the thinking, our brains may start checking out. This doesn't mean AI use is inherently harmful. But it does mean we need to be more intentional, especially in education. If students use AI to skip the struggle, they might also skip the learning. That's why how kids use AI matters just as much as whether they use it. When students rely on AI only to get quick answers, they miss the chance to build deeper thinking skills. But when they use it to challenge their thinking, test ideas, and reflect on what they know, AI can actually strengthen both memory and critical thinking. Parents can help their kids make that shift—by encouraging them to ask follow-up questions, explore alternative perspectives, and talk through what they learn. 'I'm obviously a big believer in AI, but I think we're underestimating just how deeply influential it could become, especially in education. These tools are emotionally engaging, subtly flattering, and incredibly persuasive. It's like the algorithmic influence of social media, but 100 times more powerful. I believe we'll start to see a divide where some parents will treat AI like they did Facebook, keeping their kids away from it entirely, while others will unintentionally enroll them in a real-time experiment with unknown outcomes," Matthew Graham, Managing Partner at Ryze Labs. AI Is Rewriting the Rules of Learning The role of a student is changing. From memorizing content to synthesizing it. From answering questions to asking better ones. AI has the potential to accelerate this evolution, but only if used wisely. The challenge is that many students, and schools, are still figuring it out on the fly. Some districts have embraced AI tools and incorporated digital literacy into the curriculum. Others have banned them outright. That inconsistency creates a new learning divide. One not based on income or geography, but on AI fluency and critical awareness. And AI literacy is not the same as AI usage. Just because a student knows how to prompt ChatGPT doesn't mean they know how to assess its output for bias, hallucination, or ethical issues. Or how to keep themselves safe. In chatting with Sofia Tavares, Chief Brand Officer at Preply, she told me that 'AI has become a widely used tool among students, with 80% reporting they rely on it for schoolwork, especially in language subjects such as English, linguistics, and history. While it offers convenience and quick access to information, over reliance on AI can result in a superficial understanding of the material. Tools like ChatGPT help generate ideas and clarify concepts, but they cannot replicate the emotional intelligence, adaptability, and encouragement that human educators provide. Meaningful learning occurs when students actively engage with challenging material, and this is most effective with human teachers who inspire students to ask questions and build confidence. For this reason, AI should be viewed as a supplement to, not a substitute for, skilled human instructors, like tutors.' Confidence Is High, But Are We Ready For AI The Preply survey reports that 90 percent of students and parents feel 'somewhat confident' or 'very confident' about responsible AI use. That optimism is encouraging, but we should also ask: confident in what, exactly? True readiness means knowing: And the deeper questions that every user should be asking: Without that foundation, confidence can lead to complacency. The Parental AI Divide Is the Next Digital Divide As AI becomes more embedded in learning, a new cultural rift is forming. Some parents are drawing hard lines, choosing to block tools like ChatGPT until their kids are 18—similar to early restrictions around Facebook or smartphones. Others take the opposite approach, giving their kids free rein with little oversight. Mickie Chandra, Executive Fellow at The Digital Economist and advocate for school communities, elaborates on the risks when there is little oversight. 'Aside from concerns around privacy and sharing of sensitive information, the ethical concerns are far less overt and understood by parents. A significant number of children who use ChatGPT report that it's like talking to a friend and have no issues with taking advice from them. With engagement as the driving force behind these apps, it's clear to see how children may become hooked. Consequences include developing a false sense of connection with ChatGPT, becoming overly reliant on a bot, and less interested in peer-to-peer interactions. Children do not possess the tools to discern safety or trust worthiness of their interactions with ChatGPT.' Neither extreme sets students up for success. Withholding access entirely may delay critical digital fluency. Unrestricted access without guidance can breed dependency, misinformation, or even academic dishonesty. The real opportunity lies in guided exposure. Parents can play a defining role not by banning or blindly allowing AI, but by helping kids navigate it—asking where the information came from, encouraging original thought, and modeling responsible use. From Chatbots to AI Tutors Right now, most students use chatbots as reactive tools, asking a question and getting a response. But that is changing quickly. Next-generation educational AI agents will be proactive. They'll remind students of deadlines, suggest study strategies, and flag gaps in understanding before a teacher ever intervenes. They won't just answer questions. They'll co-drive the learning journey. OpenAI's new 'study mode' in ChatGPT marks a major shift in how students engage with AI for learning. Rather than simply providing answers, this mode uses Socratic questioning to prompt students to think critically, reflect on their reasoning, and deepen their understanding of assignments. Early feedback describes it as a 24/7 personalized tutor—always available, endlessly patient, and focused on helping learners build real comprehension. If widely adopted, it could transform traditional study habits and redefine how educational support is delivered across classrooms and homes. Olga Magnusson, The Digital Economist Senior Executive Fellow, told me that 'We need to understand the impact of AI on the cognitive development of our children. The prefrontal cortex, the part of the brain responsible for reasoning, critical thinking, emotional regulation and other high cognitive functions, keeps developing until around 25 years of age. We still don't have a clear understanding and until we do, we need to ask ourselves, what is the price we are willing to pay for progress. Until then, education is our only tool.' And that opens up a bigger challenge: how do we ensure students don't outsource thinking itself? Rethinking the Roles With AI: Teachers, Parents, Policymakers The rise of AI in education doesn't replace teachers. It makes their role more essential than ever. Teachers will need to become facilitators of AI literacy, guiding students on how to think with AI, not just how to use it. Parents, too, have a new role. They're not just checking homework anymore. They're helping their kids navigate a world where AI can write the homework for them. That means having conversations about ethics, originality, and the long-term value of struggle. Policymakers have a responsibility to bridge the gap, ensuring all students have access to high-quality AI tools, and that national guidance exists for safe, responsible, and equitable implementation in schools. The Real Test For AI Is Ahead Every generation has its transformative technology. The calculator. The internet. YouTube. Now it's AI. But here's the difference. This time, the tool doesn't just give you answers. It can do your thinking for you if you let it. The question isn't whether students will use AI. They already are. The real test is whether we'll teach them to use it wisely, and make sure they don't lose the cognitive muscle that built our brightest ideas in the first place.
Fast Company
01-08-2025
- Fast Company
YouTube channels are being sold and repurposed to spread scams and disinformation, says new research
YouTubers dedicate their lives to building a following in hopes of creating and sustaining a livelihood. For top creators, the rewards are immense: MrBeast, the world's biggest YouTuber, is estimated to be worth $1 billion. It's no surprise, then, that YouTube channels are valuable assets, often bought and sold for significant sums. A new study published in the Cornell University archive arXiv reveals that 1 in every 400 YouTube channels has changed hands on third-party platforms, frequently undergoing complete transformations in the process. Researchers at Carnegie Mellon University analyzed Fameswap—a kind of eBay for social media profiles—cataloging 4,641 YouTube channels with a combined 823 million subscribers listed for sale between October 2024 and March 2025. They then tracked which of those channels sold, confirming more than $1 million in transactions during that six-month period. 'It was really remarkable,' says Alejandro Cuevas, the paper's lead author. A YouTube spokesperson said that selling accounts violates the platform's Terms of Service, and the company will take action if such activity is detected. 'If we detect that bad actors are creating channels with the intent to scam, mislead, spam, or defraud other users, we terminate those channels,' the spokesperson tells Fast Company. Given the high prices paid, new owners are eager to see a return on their investment. About one in four of the tracked channels—representing a combined 220 million followers—were completely overhauled within 30 days, with changes to their handle, title, and description that erased any connection to the original identity. The researchers found that these revamped channels continued to gain subscribers over the following 12 weeks, suggesting most viewers didn't notice the switch. However, many new owners weren't focused on producing quality content: 37% of repurposed channels later promoted material flagged by YouTube as potentially harmful—especially political disinformation, crypto schemes, and gambling ads. A follow-up analysis by the same researchers of 1.4 million randomly selected channels from analytics platform Social Blade revealed that around 0.25% showed similar patterns of transformation. The growing trade in YouTube channels reflects broader shifts in how we consume content, says Manoel Horta Ribeiro, a coauthor on the study. 'Part of the reason why this is so prevalent right now is because systematically, in these platforms, we've seen a decrease in the agency over what we consume,' he says. Knowing that a channel already has an audience and a veneer of credibility makes it easier to push new content into users' feeds. 'The shortform content paradigm facilitates this a lot, because in the past, you would search for channels directly, maybe by name, or be more aware of that,' says Cuevas. 'This just makes it more fertile ground for these type of thing.'
