The Most Parsimonious Ultralow‐Velocity Zone Distribution From Highly Anomalous SPdKS Waveforms

Abstract

For nearly 3 decades, features sitting on top of the core-mantle boundary (CMB) and referred to as ultralow-velocity zones (ULVZs) have been invoked to explain anomalous travel-time delays and the existence of a variety of pre-and post-cursor arrivals observed in seismic data. A clear definition of what constitutes a ULVZ is currently lacking; nevertheless, most papers invoking ULVZ occurrence state that they are regions of low S-and/or P-wave velocities (of the order of 10% or greater decreases) in thin regions (typically less than 40 km is stated) sitting on top of the CMB (

A Compositional Component to the Samoa Ultralow‐Velocity Zone Revealed Through 2‐ and 3‐D Waveform Modeling of SKS and SKKS Differential Travel‐Times and Amplitudes

“…We show probability contours of 0.4 and 0.5 from Thorne et al. (2021) in Figure 1 which are consistent with the ULVZ imaged here. However, our 2‐D modeling efforts are averaging the 3‐D structure.…”

“…If a single large ULVZ is responsible for our observations, then we infer a maximum width of 10.75° (650 km along the CMB) when modeling in 2-D. However, the data used in this study appear to obliquely cross the Samoa ULVZ inferred in Thorne et al (2021), so the actual ULVZ width may not be this large. This previous work inferred a long linear ULVZ stretched out in the North-South direction.…”

“…However, δA SKKS is still underestimated and the magnitude of the δT SKKS decrease is still not replicated. Both of these observations could be explained by the result of the different focusing observed in 2-D axi-symmetric versus 3-D synthetic seismograms for the SPdKS related phases (see supplemental animation in Thorne et al, 2021) as well as complex 3-D structure of the ULVZ.…”

“…The second model is selected in order to assess whether the long linear shape is necessary. The third model is chosen to see how the irregularity of the boundary affects the wavefield and to test whether the probability contours of Thorne et al (2021) may act as a reasonable proxy to the ULVZ shape. Thorne et al (2021) compared 2-D axi-symmetric synthetics to a cylindrical 3-D ULVZ.…”

“…The third model is chosen to see how the irregularity of the boundary affects the wavefield and to test whether the probability contours of Thorne et al (2021) may act as a reasonable proxy to the ULVZ shape. Thorne et al (2021) compared 2-D axi-symmetric synthetics to a cylindrical 3-D ULVZ. The 3-D cylindrical ULVZ strongly focused energy and predicted much more anomalous waveforms than the 2-D case.…”