Love Wave Tomography with Mode Separation Analysis in Northeastern America

Northeastern America has been extensively investigated using surface waves. However, Love waves suffer from the mode interference problem, causing large uncertainties in phase velocity measurements and radial anisotropy models. In this study, we adopt a newly developed stacking-stripping method to extract Love wave fundamental-mode waveforms and use them to obtain phase velocities in the region. The data are the transverse components of 114 distant earthquakes recorded by 220 USArray Transportable Array stations. We apply a 10 mHz bandwidth filter to the data at a central frequency of 0.02 Hz because the dispersion analyses show that higher-mode interference is severe around this frequency. The stacking-stripping method effectively removes the effects of higher mode interference and improves the data quality. We then apply a two-plane-wave inversion method to extract fundamental-mode waveforms to construct a phase velocity map. Compared with the results without mode separation analysis, this inversion reduces data misfits and phase velocity uncertainties. The phase velocity map with mode separation analysis is consistent with that from the original data. Also, the strength of the well-known slow anomaly in New England is increased after removing mode interference. This experiment suggests that applying the stacking-stripping method in Love wave data processing helps improve phase velocity measurements and radial anisotropy models.