Wave equation dispersion inversion of surface waves recorded on irregular topography

Abstract

Significant topographic variations will strongly influence the amplitudes and phases of propagating surface waves. Such effects should be taken into account, otherwise the S-velocity model inverted from the Rayleigh dispersion curves will contain significant inaccuracies. We now show that the recently developed wave-equation dispersion inversion (WD) method naturally takes into account the effects of topography to give accurate S-velocity tomograms. Application of topographic WD to demonstrates that WD can accurately invert dispersion curves from seismic data recorded over variable topography. We also apply this method to field data recorded on the crest of mountainous terrain and find with higher resolution than the standard WD tomogram.

Imaging the Fault Damage Zone of the San Jacinto Fault Near Anza With Ambient Noise Tomography Using a Dense Nodal Array

“…The Vs profile reveals the damage structure of the Clark fault, including a~200-m-wide asymmetric low-velocity zone. Compared to previous results at Blackburn Saddle (Li et al, 2019) to the northwest, we find the damage zone width increases to the northwest from Anza, reflecting the increase in fault geometrical complexity in this direction. Combined with previous geophysical, geological, and numerical results, our findings reinforce the idea that large San Jacinto earthquakes nucleate to the southeast and propagate to the northwest.…”

“…The relative narrowness of the damage zone here can be explained by the simplicity of the Anza segment of the Clark fault; large earthquakes tend to be confined to a single slip surface here (e.g., Rockwell et al, ) resulting in a relatively straight and geometrically simple segment (Sharp, ). This is in contrast to previous studies of the San Jacinto fault on the Hemet step over to the northeast (Li et al, ; Share et al, ; Share et al, ) and the Trifurcation to the southeast (Hillers et al, ; Qin et al, ; Mordret et al, ); both sites feature wider and more complicated damage zones and are located in regions with multiple major fault strands. Together with paleoseismological work showing that the Anza segment of the Clark fault experienced the largest coseismic slip during past M > 7.5 earthquakes (Rockwell et al, ; Rockwell et al, ; Salisbury et al, ) and numerical results examining the interaction of damage zone structures with earthquake ruptures (Dieterich & Smith, ; Fang & Dunham, ; Xu & Ben‐Zion, ), our results support the interpretation that large earthquakes do not nucleate within the Anza segment and ruptures propagate smoothly at relatively high rupture velocity.…”

“…The San Jacinto fault is the most seismically active component of the Southern California plate boundary (Hauksson et al, ) and has produced at least 21 major earthquakes (Mw > 7) in the last 4,000 years according to paleoseismic studies at Hog Lake (Figure b; Rockwell et al, ; Rockwell et al, ). The significant hazard potential and complexity of the San Jacinto fault have motivated many geological and geomorphic observations (Dor, Rockwell, & Ben‐Zion, ; Wechsler et al, ), regional‐scale tomography with local body wave and ambient noise data (Allam et al, ; Allam & Ben‐Zion, ; Fang et al, ; Zigone et al, ), bimaterial interface and damage zone imaging with small‐aperture (up to ~2 km) linear nodal arrays (Lewis et al, ; Li & Vernon, ; Li et al, ; Qin et al, ; Qiu et al, ; Share et al, ; Share et al, ; Yang & Zhu, ; Yang et al, ; Zigone et al, ), and a rectangular single‐component nodal array (Ben‐Zion et al, ; Hillers et al, ; Roux et al, ; Qin et al, ; Mordret et al, ).…”

Wave‐Equation Dispersion Inversion of Guided P Waves in a Waveguide of Arbitrary Geometry

“…These dispersion curves are concatenated together to get a pseudo‐section of the phase‐velocity distribution as a function of frequency (Figure e). We use the multioffset WDG strategy to invert for the P ‐velocity model using the conjugate‐gradient method (J. Li, Lin, et al, ). Figure f shows the WDG tomogram after 25 iterations, and the predicted dispersion curves shown in Figure a (black solid lines) resemble the observed values (black dashed lines).…”

“…The shot spacing is 5 m, the vertical‐component receiver spacing is 2.5 m, and there are 240 traces per shot gather to give a total of 120 shot gathers. Inverting the Love waves gives a low‐ S ‐velocity tomogram in the middle of site 2 (J. Li, Lin, et al, ).…”