We model mantle discontinuities using seismic data collected from earthquakes. Specifically we are looking for a low velocity zone (LVZ) above the 410 km discontinuity which would indicate the presence of silicate melt. The events were selected based on magnitude (>= 5.7) and depth (>= 100 km). Data was collected from all seismic stations within 60 degrees of the event. The earthquake events we used were grouped into three separate "paths" which allowed us to examine the backarc basin of Japan as well as the area between Japan and Hawaii.
For our purposes, the most useful portion of the seismic record is the ScS reverberative interval. This is the interval that occurs after the direct Love wave and before the arrival of the first SSS wave from the opposite direction. This interval is useful because it is dominated by high amplitude multiple ScS and sScS phases. Since we are only interested in the S wave portion of the record the data was filtered and only the SH-polarized portion was used for modeling.
The seismograms we collect contain time and amplitude information that we use to characterize mantle discontinuity structure. In order to create a model we need to know the mantle travel time, average attenuation, and crustal structure. We obtain this information from zeroth-order reverberations. Higher-order reverberations provide us with information about additional discontinuities that may be located within the mantle. First-order reverberations are used to locate the depths of these additional discontinuities.
Initially we assume that the Earth has a simple mantle structure that contains only two discontinuities, one at 410 km and one at 660 km. In order to compare the actual data with our synthetic data we use a reflection profile. This profile is created by converting amplitude and time into reflectivity and depth respectively. This is accomplished by adding discontinuities to the mantle model until the synthetic profile matches the data profile. Due to the complex manner in which seismic waves interact with discontinuities, a perfect match is not possible. The goal is to obtain a model that closely resembles the data.
The research conducted this summer yielded very interesting results. The two paths located in the western Pacific provided evidence of a discontinuities at ~80 km (G) and ~300 km (X). The path located in the backarc basin of Japan showed evidence for silicate melt at 337 km as well as a small discontinuity at 850 km. The evidence for silicate melt is what we were initially searching for, however the "X" discontinuity is some thing that warrants further studies. With more research we hope to better understand the "X" discontinuity found in the western Pacific.