Imaging the Shallow Subsurface Structure of the North Hikurangi Subduction Zone, New Zealand, Using 2‐D Full‐Waveform Inversion

The northern Hikurangi plate boundary fault hosts a range of seismic behaviors, of which the physical mechanisms controlling seismicity are poorly understood, but often related to high pore fluid pressures and conditionally stable frictional conditions. Using 2‐D marine seismic streamer data, we employ full‐waveform inversion (FWI) to obtain a high‐resolution 2‐D P wave velocity model across the Hikurangi margin down to depths of ~2 km. The validity of the FWI velocity model is investigated through comparison with the prestack depth‐migrated seismic reflection image, sonic well data, and the match between observed and synthetic waveforms. Our model reveals the shallow structure of the overriding plate, including the fault plumbing system above the zone of slow‐slip events to theoretical resolution of a half seismic wavelength. We find that the hanging walls of thrust faults often have substantially higher velocities than footwalls, consistent with higher compaction. In some cases, intrawedge faults identified from reflection data are associated with low‐velocity anomalies, which may suggest that they are high‐porosity zones acting as conduits for fluid flow. The continuity of velocity structure away from International Ocean Discovery Program drill site U1520 suggests that lithological variations in the incoming sedimentary stratigraphy observed at this site continue to the deformation front and are likely important in controlling seismic behavior. This investigation provides a high‐resolution insight into the shallow parts of subduction zones, which shows promise for the extension of modeling to 3‐D using a recently acquired, longer‐offset, seismic data set. Key Points Full‐waveform inversion was used to produce a high‐resolution P wave velocity model of the upper ~2 km of the north Hikurangi subduction zone Recovered velocities and gradients correlate well with drilled sonic logs, suggesting that the details we see in our FWI velocity models are real Velocity changes across faults may relate to compaction, and in some cases low velocities along faults may indicate conduits for fluid flow