‘Earth’s evil twin’

As a third-year Ph.D. student, Madeleine Kerr faced a critical decision. At the University of California San Diego, she models mysterious features on Venus’ surface. But she and her advisor, Dave Stegman, realized that her simulations couldn’t answer their central question: What’s going on in the mantle?

She could have settled for respectable but anticlimactic results and wrapped up her doctoral degree. Rebooting would require learning new software and might not yield an answer. Stegman “made a Moby Dick reference, something like ‘Let’s go after the biggest fish,’” Kerr recalls. “I was all in.”

The pivot paid off. The Department of Energy Computational Science Graduate Fellowship (DOE CSGF) recipient is running simulations that appear to explain how the Venus mantle’s convection system produces the varied features on its surface.

This month, Kerr published her results in the Proceedings of the National Academy of Sciences.

“I love Venus,” she says. “It’s Earth’s evil twin.” The two planets are about the same size, but the differences are stark. “At the surface of Venus, it’s like the pressure of being a kilometer deep in the ocean,” Kerr says. “And it’s hot enough to melt lead.”

Venus’ surface is also quite different from Earth’s in that it lacks tectonic plates, which collide to form Earth’s mountains and allow heat to escape through the cracks. Instead, Venus’ lithosphere, the crust and outer mantle, is a solid shell called a stagnant lid.

As a result, volcanic highlands, likely made when hot plumes rose from the core-mantle boundary, scar the surface. Resembling Hawaii and Iceland, they are thousands of kilometers across, and Texas and California easily could fit inside the largest one.

Regions between the highlands are peppered with smaller features called coronae, 60 to 1,000 kilometers across. These crown-shaped lava domes are bordered by mountainous volcanoes. “The question is, how can one convective system create two different scales of upwellings?” Kerr says.

In Kerr’s early Ph.D. research, she developed a 2D numerical model of the mantle, using a geodynamic code called StagYY, which simulates the bulk mineral behavior as temperature and pressure change. She modeled hot plumes rising from the core-mantle boundary and cooler material peeling off the lithosphere’s interior. Kerr and Stegman thought some of these hot plumes created the smaller coronae, whereas cooler material dripped into and blocked other plumes, eventually building them into the super plumes that formed the highlands.

In her model, which ran on the National Energy Research Scientific Computing Center’s Perlmutter supercomputer, Kerr replaced the simplifications in other researchers’ conceptualization with detailed physical parameters. But to her surprise, the model’s realistic values, especially for viscosity, choked off corona formation. Her research seemed to have its own stagnant lid.

Kerr and Stegman formed their new plan based on a conversation with Juliane Dannberg, now at Germany’s Kiel University. Dannberg had simulated early Earth’s uppermost mantle, when our planet was about as hot as Venus is now. But she adapted parts of a different code, ASPECT, to include an unconventional entropy-based method for tracking how heat and pressure changes can alter minerals’ crystal structures. Those shifts cause bulk materials to expand or contract in different circumstances. Kerr and Stegman wondered if a similar simulation could reveal more about the hot mantle under Venus’ stagnant lid.

Continuing on Perlmutter, Kerr used ASPECT to simulate Venus’ mantle. She credits her DOE CSGF practicum at the National Renewable Energy Laboratory for her rapid progress working with new code. There, with Marc Day in the High Performance Algorithms and Complex Fluids group, she helped develop simulations to visualize hydrogen combustion. “Figuring out how to compile a new code and debug all of the different errors and flags — that learning process was very important for the most exciting part of my Ph.D.,” she says.

In the new simulations, a boundary layer formed within the mantle at depths from 400 to 700 kilometers, between the cooler material above and the hotter material below. The researchers realized the layer might not just slow the flow of material but block it, which could account for the different surface features. Those early findings suggested that their new model might succeed.

Kerr and Stegman shared these results with Suzanne Smrekar, who leads the VERITAS mission to Venus at NASA’s Jet Propulsion Laboratory. Smrekar soon agreed to lend her knowledge and insights, which Kerr credits with having informed much of the group’s progress.

Inside rocky planets like Earth and Venus, pressure at greater depths tends to force minerals to form increasingly compact structures. But those trends aren’t uniform, and variations in mineral content and structure, temperature and pressure create a complex stew of layers that alter the movement of the mantle’s molten rock. The boundary layer that Kerr and Stegman observed resulted from a unique transition zone, 100 kilometers thick, where three mineral phases interact and behave unlike the surrounding layers. This so-called multiphase assemblage shows a property called negative thermal expansivity, forming a less dense mixture as it’s cooled and denser one as it’s heated.

That feature within this transition layer affects rock behavior at its boundaries above and below. As cooler, less dense blobs of rock from above bump into the transition zone, they become both warmer and denser than the surrounding material, which produces equally opposing forces that lock them into place in the boundary layer. The opposite effect traps warm, dense material that rises and bumps into the transition zone from below. This trapping mechanism defines the boundary layer.

In keeping with the long history of naming the planet’s features with a feminine spin, the team dubbed the boundary layer the “glass ceiling.” The researchers saw how the glass ceiling could create the planet’s surface features. In the model, cool mantle material dripped off the lithosphere and pooled on top of the boundary layer. Eventually, its weight broke through the barrier in avalanches that displaced hot material below. Hot material welled up through the breach and created coronae at the surface. At the same time, the highlands formed as massive, buoyant plumes from the core-mantle boundary below punched through the glass ceiling, allowing the plumes’ hot centers to continue to the surface.

The group continues to pursue answers to other Venus questions, such as how the lithosphere conducts heat to the surface.

After her Ph.D., Kerr hopes to find a position that will allow her to develop a 3D model. As she notes, “The first full Venus model that can actually express the surface of today — that’s the goal.”

Editor’s note: A version of this article is originally appeared in DEIXIS: The DOE CSGF Annual.