Dominant Higher Surface-wave Modes and Possible Inversion Pitfalls

Dominant higher modes of surface waves are known to be generated at sites with large stiffness contrasts and/or reversals between layers. At many engineering sites, quite often all modal energy is transferred to successively higher modes with increasing frequency. These ‘jumps’ may occur at low (20 Hz) to high (60 Hz) frequencies, and, more than once over the recorded bandwidth, but the common features are that the jumps are in amounts exceeding 50% of the phase velocity, and dispersion does not return to the fundamental mode at high frequency. A nonlinear geophone spacing helps to resolve these modes more uniquely. Full-wavefield synthetic seismogram modelling reinforces the effect to be due to a steep, nonlinear stiffness gradient in the uppermost 1–2 meters of soil. If a high-velocity substrate at depth is not present, the ‘effective’ phase velocity of the higher modes can exceed that of the maximum shear-wave velocity of the model, where leaky modes persist. When this happens, conventional modal dispersion modelling cannot be applied in the inversion. If an artificial stiff layer is added at depth, jumps across several modal boundaries can occur, which is also generated in a Gibson half-space, both at very low and high frequencies. This is a major pitfall, where mode-misidentification, especially at low frequency, can lead to errors in the estimated shear wave velocity models of over 50%. Guided P-waves also manifest as large dispersion discontinuities, but usually at higher frequency and phase velocities, with buried explosive sources. Although Poisson9s ratio has a strong influence of the generation of—and frequencies of transitions to—dominant higher modes, the higher-mode phase velocities themselves are relatively independent of shallow P-wave velocity.