Tomography is a powerful seismic technique for constraining the three-dimensional (3-D) structure of the Earth's interior [e.g., Romanowicz, 2003; Thurber and Ritsema, 2005]. Models of 3-D seismic structure take a central place in the research of Earth scientists from a wide range of disciplines.
The traveltime (of a body- or surface wave) is the most popular data type in tomography. Traveltimes are available in seismological catalogs, or they can be measured in a relatively straightforward manner. Moreover, traveltimes can be related to velocity heterogeneity in a linear fashion, simplifying the tomographic inverse problem. Millions of traveltimes constitute modern data sets. However, do tomographic models explain other attributes of wave propagation and how do simplified wave theories compromise model accuracy? These are the questions we are addressing in our research.
We are analyzing 3-D Spectral-Element Method (SEM) wave simulations [Komatitsch and Tromp, 1999; Komatitsch et al., 2002] to investigate how 3-D seismic models can be improved using observables that are traditionally not used in tomographic inversions.
(1) Body wave amplitudes We have demonstrated [Ritsema et al., 2002; Komatitsch et al., 2002] that the pattern (but not the amplitude) of SS/S amplitude ratios can be explained by shear velocity heterogeneity in the mantle for model S20RTS. We are investigating whether these body wave amplitude ratios can be used to constrain the (lateral) gradients of shear velocity in the mantle, which we suspect to be underestimated by tomographic models.
(2) Frequency-dependent body wave traveltimes We are observing a systematic variation of body wave traveltimes between 20-100 mHz. While such a variation is expected on theoretical grounds [Dahlen et al., 2000], we are using SEM simulations to investigate whether this signal can be used to constrain the amplitude and spectrum of heterogeneity in the mantle.
(3) Resolution of "small-scale" heterogeneity We are analyzing SEM waveforms to conduct new resolution tests that explore whether "small-scale" (i.e., < 200 km) structures produce observable traveltime or waveform variations. We are particularly interested in constraining the mantle structure of Iceland, a region overlying an anomalous mantle upwelling.
Broader Impact: The research is supporting the training of a Geophysics student in seismological data analysis and advanced scientific computing. The research enhances the P.I.'s collaboration with specialists in global wave propagation theories and geodynamics.
