A new study lead by Matt Kohn and published Oct. 31 in the Proceedings of the National Academy of Sciences shows that the temperatures associated with the earth’s subduction zones have been historically miscalculated, which has major implications for our understanding of how the planet’s deadliest earthquakes and volcanic arcs are generated.
“Our research shows there is a disparity between rocks examined today and our models of how they formed,” Kohn explained. He specializes in understanding the temperature structure of subduction zones.
Subduction zones are where one of Earth’s tectonic plates dives down beneath another. This creates a large proportion of Earth’s deadliest earthquakes and volcanoes.
“With a few exceptions, most major earthquakes and volcanic eruptions are associated with a subduction zone,” he said.
By reinterpreting heat loss from Earth’s surface and creating new models of subduction zones, Kohn and Boise State graduate student Buchanan Kerswell; César R. Ranero, a research professor with the Spanish National Research Council; and Adrian E. Castro and Frank S. Spear, researchers with the Department of Earth and Environmental Sciences at the Rensselaer Polytechnic Institute, were able to prove that past, commonly referenced thermal-mechanical models of subduction zones are two times too cold. Rather, pressure-temperature conditions determined from exhumed metamorphic rocks are better indicators of subduction zone temperatures.
Accurately inferring subduction zone temperatures is crucial for predicting how rocks deform and melt, which defines patterns of arc volcanoes and earthquakes.
“There was a huge disparity between the metamorphic rocks we see today and the models that predicted how they were formed,” Kohn said. “We were trying to understand where these differences come from.”
To create a new, more accurate model, the team compiled surface heat flow data from subduction zones worldwide and were able to show that the rate of heat loss is higher than can be explained for frictionless sliding, which is often assumed for modeling. Incorporating friction into models, as required by heat flow, reproduces the rock record.
“One particular earthquake, one particular volcano – they don’t reflect the long-term behavior of subduction. Rather our work looks at the sum of all the earthquakes and sum of all volcanoes that might exist in a subduction zone over millions of years,” Kohn said. “This is like the Google Earth view of subduction zones.”
The team’s research was funded by a $4 million collaborative grant through the National Science Foundation.