AI reconstructs hidden history of Earth's mantle flow from surface clues
The mantle, a rocky layer accounting for more than 80% of Earth's volume, circulates at rates of only a few centimeters per year. This circulation drives plate tectonics and influences major geological phenomena such as earthquakes and volcanic activity. Despite its fundamental importance, the history of mantle circulation remains poorly understood because direct observations of the deep Earth…
The Earth's mantle, comprising more than 80% of its volume, is a rocky layer that circulates at a rate of only a few centimeters per year. This slow circulation drives plate tectonics and influences major geological events such as earthquakes and volcanic activity. Due to the difficulty of directly observing the deep Earth, scientists primarily rely on geological records of surface movements and geophysical imaging techniques, like seismic observations, to understand mantle structure and behavior. However, reconstructing the historical flow of the mantle remains a significant challenge.
In a recent study, researchers from the University of Tsukuba developed an AI model using a physics-informed neural network. This innovative model was trained to fit observational data while also adhering to the physical equations governing heat transport and fluid flow in the mantle. To evaluate the model, the researchers first created computer simulations of two-dimensional mantle thermal convection. These simulation results served as a reference solution against which the AI reconstruction was assessed.
To test the model, synthetic observations representing near-surface mantle motion and a snapshot of the mantle's current temperature distribution were provided. Remarkably, the model successfully reconstructed hidden features such as past temperatures and deep-mantle flow with high accuracy, despite not receiving direct information about these aspects.
The study's findings are published in the Journal of Geophysical Research: Machine Learning and Computation, highlighting the model's potential as a powerful tool for uncovering how Earth's deep interior has evolved over time.
The researchers emphasize that combining different types of geophysical information is crucial for accurately reconstructing realistic mantle convection histories. With further development and application to real geophysical data, this physics-informed neural network approach may prove invaluable in revealing the complex history of Earth's deep interior.
Written by urgent.news from Phys.org's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
