As the oceanic portion of a tectonic plate gets older, it also grows colder and denser. Eventually it sinks back into the interior of the Earth in a process called subduction.  Subduction is a critical process for the planet because it brings sediments, rocks, and – critically – water and volatiles (CO₂) from the surface to the deep Earth. Water cycling enhances melting of the mantle, giving rise to volcanism that returns elements to the Earth’s surface in a giant recycling process. Given that the cold descending plate takes a long time to heat up to the temperature of their surroundings, and that seismic waves propagate faster in colder materials, we can image the subducting plate by analyzing the time it takes for seismic waves to traverse the planet to derive seismic velocity at depth. Likewise, we can see where hotter material from the deep Earth rises in response to the sinking plate.

Temperature affects seismic velocity regardless of the direction in which the wave is traveling (i.e. velocity variations are isotropic). We know, however, that as a plate subducts, the surrounding material (Earth’s mantle) flows with it and around it, and that this aligns the minerals in such a way that creates anisotropy (i.e. the speed of the wave depends on its direction). Most commonly, analyses of seismic velocity assume isotropic material. As part of this project, our research group demonstrated that this incorrect assumption leads to changes in wave travel time arising from strong mineral alignment being misinterpreted as produced by temperature variations. Given that the seismic data alone are typically not sufficient to untangle the effects of isotropic and anisotropic velocity variations at this scale, we addressed the problem by incorporating estimates of anisotropy derived from additional observations and from numerical modeling into our analysis. By determining whether the assumed anisotropy fits the observations, we simultaneously explored the effectiveness of the modeling in predicting anisotropy and the potential for misidentification of anisotropic travel time effects as temperature variations. Applying these techniques to data from the western mediterranean we have determined that previously imaged low-velocity anomalies do not require the presence of melt if anisotropy is taken into account. We are also exploring how anisotropy may affect the imaging of high-velocity anomalies beneath Nevada. Ultimately this approach can lead to more accurate estimation of temperature variations, and this in turn helps us understand how much melt is produced and retained and where thermal upwellings are bringing deep material to shallower levels. Thermal structure can only provide a snapshot of the system at the present moment, but anisotropy represents a record of its history. Therefore, including anisotropy in our analysis opens a new window into the evolution of the subducting plate and the related flow of Earth materials.

Last Modified: 03/26/2019
Modified by: Maximiliano Bezada