Evidence for fluids at the shallow Hikurangi Margin, New Zealand, imaged by offshore seismic attenuation

Abstract:

In this study we use data from offshore ocean-bottom seismometers (OBS) for 3D P- and S-wave local earthquake attenuation tomography. The Northern Hikurangi Subduction Margin, New Zealand, has been the focus of interdisciplinary geophysical studies to better characterize active plate margins. Understanding the causes behind shallow slow-slip events (SSEs), which may slip to the trench, is a primary motivation behind these collocated studies. Many researchers propose that high pore fluid pressures decrease the effective stress on the plate interface, leading to SSEs, and seismic reflection, high Vp/Vs, and numerical modeling support the presence of pockets of fluids along the interface and in the accretionary prism. Seismic attenuation tomography can determine whether fluids and fractures, detected offshore in the upper plate and Hikurangi Plateau up to ~10 km depth, are clear at a broader scale up to ~25-35 km in depth. We use data from the Hikurangi Ocean Bottom Investigation of Tremor and Slow Slip (HOBITSS) 2014-2015 seismic deployment, co-temporal and collocated with a shallow SSE, to produce images over the slow-slip area and close to the trench. P- and S-wave measurements of t* from 180 earthquakes > M_L 2.0 are inverted for Qp and Qs (the inverse of attenuation). The spectra from which t* is measured are corrected for the average site effect for each station. This is especially important for OBS stations, where thick sediment underlying the instruments causes prominent peaks in resonance. Near the trench, we find anomalously low Qp and Qs (1-50) throughout the heavily faulted accretionary prism and down-dip of a subducting seamount. The low Q anomaly down-dip of the seamount is spatially coincident with Vp/Vs > 1.85 and a high seismic reflectivity zone on the plate interface. The low to intermediate Q in the subducting slab (Qp 50-200, Qs 50-300) is likely due to the basaltic composition of the large igneous province (the Hikurangi Plateau). This is supported by experimental Q studies of basalts and Qp studies in Costa Rica (e.g. Christeson et al., 2000). Over the ~80 km scale of the array, high attenuation (in conjunction with high Vp/Vs) at interface and upper plate depths is broader than individual seamounts, and points to heterogeneous fluid patterns possibly related to the margin-scale properties of the incoming plate.