Joshua C. Stachnik

Determination of New Zealand Ocean Bottom Seismometer Orientation via Rayleigh-Wave Polarization

Online material: Complete list of earthquakes used in this study. Three-component ground-motion recordings are critical to modern seismic analysis techniques such as receiver functions and body- and surface-wave polarization studies. Modern three-component seismometers typically resolve three axes of ground motion into one vertical and two orthogonal horizontal directions. The standard convention for installation is to orient the instrument such that one of the horizontal components is aligned with true north, taking into account the current local magnetic declination at the station. This practice can be difficult in the case of borehole instruments and practically impossible for ocean-bottom seismometers (OBS). Some deployments of OBS have used airgun shots to determine sensor orientation (Anderson et al. , 1987; Duennebier et al. , 1987); however, this is not an error- or cost-free procedure and is not always available. In these cases, determination of sensor orientation is necessary independent of human interaction with the physical instrument. The elliptical particle motion of Rayleigh waves is exploited here to calculate sensor orientation based on the statistical analysis of earthquakes recorded on OBS deployed around the south island of New Zealand. This robust method is computationally fast because synthetic seismograms are not generated and accurate source parameters are not necessary. The technique is compared with body-wave orientation results and full-waveform surface-wave orientation results. Application of this method to waveforms recorded on land stations and synthetic waveforms confirms its reliability. Surface-wave arrival azimuth is determined via polarization analysis of the Rayleigh wave recorded on a three-component sensor. Rayleigh waves exhibit retrograde elliptical particle motion, theoretically only observed on the vertical and radial components. Because it is more stable to quickly determine a linear relationship than to measure ellipticity, the polarization analysis is performed by cross correlating the vertical component with the Hilbert-transformed radial component. The 90° phase …

A joint Monte Carlo analysis of seafloor compliance, Rayleigh wave dispersion and receiver functions at ocean bottom seismic stations offshore New Zealand

Teleseismic body-wave imaging techniques such as receiver function analysis can be notoriously difficult to employ on ocean-bottom seismic data due largely to multiple reverberations within the water and low-velocity sediments. In lieu of suppressing this coherently scattered noise in ocean-bottom receiver functions, these site effects can be modeled in conjunction with shear velocity information from seafloor compliance and surface wave dispersion measurements to discern crustal structure. A novel technique to estimate 1-D crustal shear-velocity profiles from these data using Monte Carlo sampling is presented here. We find that seafloor compliance inversions and P-S conversions observed in the receiver functions provide complimentary constraints on sediment velocity and thickness. Incoherent noise in receiver functions from the MOANA ocean bottom seismic experiment limit the accuracy of the practical analysis at crustal scales, but synthetic recovery tests and comparison with independent unconstrained nonlinear optimization results affirm the utility of this technique in principle.

Lithospheric shear velocity structure of South Island, New Zealand, from amphibious Rayleigh wave tomography

We present a crust and mantle 3-D shear velocity model extending well offshore of New Zealands South Island, imaging the lithosphere beneath the South Island as well as the Campbell and Challenger Plateaus. Our model is constructed via linearized inversion of both teleseismic (18–70 s period) and ambient noise-based (8–25 s period) Rayleigh wave dispersion measurements. We augment an array of 4 land-based and 29 ocean bottom instruments deployed off the South Islands east and west coasts in 2009–2010 by the Marine Observations of Anisotropy Near Aotearoa experiment with 28 land-based seismometers from New Zealands permanent GeoNet array. Major features of our shear wave velocity (Vs) model include a low-velocity (Vs  50 km) beneath the central South Island exhibits strong spatial correlation with upper mantle earthquake hypocenters beneath the Alpine Fault. The ~400 km long low-velocity zone we image beneath eastern South Island and the inner Bounty Trough underlies Cenozoic volcanics and the locations of mantle-derived helium measurements, consistent with asthenospheric upwelling in the region.