Scientists have identified six previously unknown structures hidden deep within the Earth's interior, according to new research published in the Journal of Geophysical Research: Solid Earth. These structures lie approximately 2,900 kilometers beneath the surface, near the boundary between the lower mantle and the liquid outer core. The discovery was made using advanced seismic data analysis powered by artificial intelligence, which examined over two million seismic records collected between 1990 and 2024. The findings were achieved through the study of weak signals known as PKP precursors, waves that pass through the liquid outer core and scatter upon encountering small irregularities or variations in the deep Earth’s composition. These signals arrive at detectors slightly before the main seismic waves but are too faint to be easily detected without sophisticated tools. Traditionally, scientists manually searched for these signals among vast datasets, a process prone to human error and subjectivity. To overcome this challenge, researchers trained a deep learning model with signals already identified by experts. Errors in the system were reviewed and corrected by the team, and these corrections were fed back into the training process to improve accuracy. As a result, the AI tool analyzed more than two million seismic records from nearly 5,000 earthquakes and recognized 174,929 high-quality PKP precursor signals. This number exceeds previous totals by more than tenfold, enabling the creation of a much more detailed map of the deep mantle. The newly discovered structures appear to be subtle anomalies that subtly alter the path of seismic waves, suggesting they could be remnants of ancient tectonic plates or even fragments linked to the Moon’s formation. However, their exact composition remains undetermined, and further studies will be needed to confirm these hypotheses. The area where these structures are located is one of the most dynamic and complex regions of the planet. It marks the boundary between the solid lower mantle, composed of slowly deforming rocks, and the liquid outer core, primarily made up of iron and nickel. At this interface, there is a sharp contrast in temperature, density, and composition. Heat escaping from the core drives mantle convection and influences volcanic activity on the surface. The presence of these mysterious structures might affect how heat is transferred through this critical zone, potentially altering our understanding of Earth’s internal dynamics. This breakthrough highlights the growing role of artificial intelligence in planetary science. By automating the detection of subtle seismic patterns, the technology has enabled researchers to uncover features that would otherwise remain hidden. The ability to analyze such large volumes of data efficiently opens new avenues for exploring the Earth’s interior and other celestial bodies. Similar techniques could soon be applied to study Mars, where recent observations have revealed unusual geological formations that remain poorly understood. The implications of this discovery extend beyond pure scientific curiosity. Understanding the structure and composition of the deep Earth can provide insights into the planet’s history, including its early evolution and the processes that shaped its current state. As researchers continue to refine their models and expand their datasets, future studies may reveal even more secrets buried beneath the surface, offering a clearer picture of the forces that govern our world.
★
Keep the news honest.
ObjectiveNews is reader-funded and ad-free — we show you the bias instead of hiding it. Support independent journalism for €4/month.
Become a Supporter