Extended Data Fig. 1: Imaging approach: P-wave velocity models derived from traveltime tomography and elastic full waveform inversion of downward extrapolated (Synthetic Ocean Bottom Experiment) multichannel seismic data along profile 05CM-38, central Hikurangi convergent margin, New Zealand.
From: Physical conditions and frictional properties in the source region of a slow-slip event
a Traveltime tomography P-wave velocity structure (see Supplementary Movie 1); b Elastic full waveform inversion P-wave velocity structure derived after a from refraction and wide-angle reflection seismic energy ahead of the seafloor reflection arrival in downward extrapolated MCS data (see Supplementary Movie 2). c Elastic full waveform inversion P-wave velocity structure derived after a and b from the seismic reflection energy below the seafloor reflection and above the first seafloor multiple in downward extrapolated MCS data (see Supplementary Movie 3). Comparison of 1-D velocity models at a distance of 25 km (d), 34.55 km (e) and 63.5 km (f) from the deformation front. The seafloor streamer tomography model isolates the large wavelength velocity structure (green line), while the two following stages of elastic FWI incrementally improves the fine scale gradient structure (red and black lines). The dashed green lines mark + /− 10% of the seafloor tomography model. Here, we suggest that the error in the high-frequency component of the P-wave velocity model is <10% of the final recovered value (region bounded by the dashed green lines). The error in the low-frequency component (that is ~the green line) is expected to be smaller.
