The Earth's mantle consists of the rock layer lying between the crust (7-70 km deep) and the core (2886 km deep) that makes up over 80% of the Earth's volume. The "transition zone" is one of the least understood regions of the mantle even though it is one of the most important in terms of understanding how plate tectonics operates in the deep Earth. The transition zone is defined by two abrupt jumps in seismic velocity (i.e. discontinuities) at approximately 410 and 660 km depth that are referred to as the 410 and 660 km discontinuities. Earthquakes produce shear waves that reflect off of these seismic jumps, and these SS phases provide global discontinuity depth maps. These seismic jumps result from an abrupt change in atomic structure due to high pressures at depth, i.e. a phase transition. Variations in the depth of these discontinuities indicate regional changes in temperature and/or composition, which provide a means of locating the source of warm volcanic regions and the accumulation of oceanic crust in the mantle due to plate tectonics.
A perplexing observation from global mapping of these seismic jumps is that the 410 km discontinuity is correlated with the 660 km discontinuity at very long wavelengths. This observation defies expectations since the phase changes responsible for these discontinuities have opposite responses to changes in temperature. It is well known that higher temperatures will move the 410 km seismic discontinuity to higher pressures (i.e. greater depth). Likewise, colder temperatures will move this phase change to lower pressures (i.e. shallower depths). This relationship is reversed for the 660 km discontinuity. Therefore, a simple thermal anomaly is expected to produce opposite topography of the 410 and 660 km discontinuities, not the observed positive correlation. The correlated topography could be due to a shift in dominance from olivine to the garnet system at high temperature. This study tests this hypothesis by examining whether the reflections change character at different frequencies. The current catalog of clean long-period SS arrivals will be used to establish a database of higher frequency arrivals and to compare the PI's recently expanded long-period stacks to their short-period equivalents. These observed long and short-period reflections will be compared to predictions for different 1D transition zone velocity models. The increased seismic activity in the Sumatran region along with the progress of the USArray TA network have resulted in a staggering increase in the number of SS traces available for analysis. These highest-quality stacks occur under the NW Pacific where slow shear wave velocities and depressed 410 km discontinuity topography indicate regions that are likely warmer than the average mantle. Thus, the highest quality data is co-located with the regions where it may be hot enough to change the dominant phase from the olivine system to garnet system.
This work will provide the greater solid Earth community with the specific seismic characteristics needed to distinguish between the competing effects of temperature, composition, and mineralogy on the transition zone discontinuities. If the character of the reflections indicates that there are two phase changes occurring around 660 km depth, this would favor the explanation that there is a change in mineralogy due to high temperatures. Likewise, if there is no indication of a second phase transition, then chemical heterogeneity may better explain the observations, and this information is critical to arguments regarding the spatial distribution of chemical heterogeneities in Earth's mantle. The results of this work are important to mantle convection calculations and should drive further mineral physics experimentation at these depths and temperatures.