Abstract

SUMMARYSpectra of whole Earth oscillations or normal modes provide important constraints on Earth’s large-scale structure. The most convenient way to include normal mode constraints in global tomographic models is by using splitting functions or structure coefficients, which describe how the frequency of a specific mode varies regionally. Splitting functions constrain 3-D variations in velocity, density structure and boundary topography. They may also constrain anisotropy, especially when combining information from spheroidal modes, which are mainly sensitive to P–SV structure, with toroidal modes, mainly sensitive to SH structure. Spheroidal modes have been measured extensively, but toroidal modes have proven to be much more difficult and as a result only a limited number of toroidal modes have been measured so far. Here, we expand previous splitting function studies, by focusing specifically on toroidal-mode overtone observations. We present splitting function measurements for 19 self-coupled toroidal modes of which 13 modes have not been measured before. They are derived from radial and transverse horizontal component normal mode spectra up to 5 mHz for 91 events with MW ≥ 7.4 from the years 1983–2018. Our data include the Tohoku event of 2011 (9.1MW), the Okhotsk event of 2013 (8.3MW) and the Fiji Island event from 2018 (8.2MW). Our measurements provide new constraints on upper- and lower-mantle shear wave velocity structure and in combination with existing spheroidal mode measurements can be used in future inversions for anisotropic mantle structure. Our new splitting function coefficient data set will be available online.

Highlights

  • Studying the splitting of whole Earth oscillations enables us to recover large-scale averages of the Earth’s P- and S-wave velocity, density, attenuation and anisotropy

  • Normal modes are most measured with the splitting function approach (i.e. Woodhouse & Giardini 1985) using leastsquares inversion of normal mode spectra

  • Splitting functions have mainly been measured for spheroidal modes which are mainly sensitive to P-wave and vertically polarized S-wave velocity (P–SV) and are able to constrain velocity and density structure in the mantle and core (e.g. Giardini et al 1988; Roult et al 1990; He & Tromp 1996; Resovsky & Ritzwoller 1998; Durek & Romanowicz 1999; Deuss et al 2010, 2013; Koelemeijer et al 2013)

Read more

Summary

Introduction

Studying the splitting of whole Earth oscillations enables us to recover large-scale averages of the Earth’s P- and S-wave velocity, density, attenuation and anisotropy. They provide strong constraints on 3-D variations in both isotropic and anisotropic velocity structure of the Earth’s mantle and are the only seismic data type directly sensitive to density (Ishii & Tromp 1999; Kuo & Romanowicz 2002; Resovsky & Trampert 2003; Trampert et al 2004; Koelemeijer et al 2017). Splitting functions have mainly been measured for spheroidal modes which are mainly sensitive to P-wave and vertically polarized S-wave velocity (P–SV) and are able to constrain velocity and density structure in the mantle and core Splitting functions of more than 180 spheroidal modes in the frequency range up to 10 mHz have been measured far

Results
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call