Impact of 3-D Earth structure on W-phase CMT parameters

Catalina Morales-Yáñez,Luis Rivera,Zacharie Duputel

doi:10.1093/gji/ggaa377

Catalina Morales-Yáñez, Luis Rivera + Show 1 more

Open Access

https://doi.org/10.1093/gji/ggaa377

Copy DOI

Abstract

SUMMARYWe investigate the impact of unmodelled 3-D structural heterogeneity on inverted W-phase source parameters. We generate a large data set of synthetic seismograms accounting for the Earths 3-D structure for 250 earthquakes globally distributed. The W-phase algorithm is then used to invert for earthquake CMT parameters, assuming a spherical Earth model. The impact of lateral heterogeneity is assessed by comparing inverted source parameters with those used to compute the 3-D synthetics. Results show that the 3-D structure mainly affects centroid location while the effect on the other source parameters remains small. Centroid mislocations present clear geographical patterns. In particular, W-phase solutions for earthquakes in South America are on average biased 17 km to the east of the actual centroid locations. This effect is significantly reduced using an azimuthally well balanced distribution of seismological stations. Source parameters are generally more impacted by mantle heterogeneity while the scalar moment of shallow earthquakes seems to be mainly impacted by the crustal structure. Shallow earthquakes present a variability of Mrθ and Mrϕ moment tensor elements, resulting both from the small amplitude and a larger uncertainty of the associated Green’s functions.

Highlights

The characterization of earthquake sources is based on different data types such as teleseismic body waves, surface waves, strong-motion waveforms, GNSS and interferometic synthetic aperture radar (InSAR) data (e.g. Delouis et al 2010; Polet & Thio 2011; Crowell et al 2012; Ekstrom et al 2012)
This phase can be described as the superposition of normal mode overtones, that have limited sensitivity to shallow lateral heterogeneity compared to fundamental mode surface waves (Kanamori 1993)
We evaluate the performance of the W-phase algorithm using a large data set of synthetic seismograms (∼64 000 waveforms) computed for 3-D Earth models

Summary

Introduction

The characterization of earthquake sources is based on different data types such as teleseismic body waves, surface waves, strong-motion waveforms, GNSS and InSAR data (e.g. Delouis et al 2010; Polet & Thio 2011; Crowell et al 2012; Ekstrom et al 2012). While different strategies have been developed to quantify such uncertainties (Yagi & Fukahata 2011; Duputel et al 2014; Hallo & Gallovic 2016), most applications rely on a simple 1-D (or spherical) Earth model (Hallo et al 2017; Gombert et al 2018). It is the case for W-phase CMT inversions for which 3-D effects are supposedly small but have not yet been fully quantified (Kanamori & Rivera 2008; Duputel et al 2016). This phase can be described as the superposition of normal mode overtones, that have limited sensitivity to shallow lateral heterogeneity compared to fundamental mode surface waves (Kanamori 1993)

Methods

Findings

Discussion

Conclusion