Predominant-Mode Inversion of Surface Waves: Inherently Addressing Inconspicuous Low Frequency Mode Jumps

TL;DR

This paper introduces a predominant-mode inversion method for surface waves that automatically identifies the dominant mode at each frequency, improving shear-wave velocity estimates at sites with impedance contrasts.

Contribution

It presents a novel inversion framework that does not require explicit mode indexing, effectively addressing low-frequency mode jumps in surface-wave data.

Findings

Accurately recovers velocity contrasts and interface depths in synthetic models.

Outperforms fundamental-mode inversion by avoiding overestimation of Vs.

Shows strong agreement with downhole logs in real data applications.

Abstract

Inversion of Rayleigh-wave dispersion data is particularly challenging at sites with strong impedance contrasts, where modal energy often transitions smoothly from the fundamental to higher modes at low frequencies. Analysts may misinterpret this transition as a continuation of the fundamental mode, leading to an overestimation of shear-wave velocity (Vs) in deeper layers and/or a misinterpretation of bedrock depth. Although effective-mode inversion can theoretically account for such behavior, it requires precise source-receiver geometry and cannot be applied when target dispersion data are formed by combining multiple active shots with passive-array recordings that have unknown source locations. This study introduces a predominant-mode inversion framework that addresses low-frequency mode jumps by automatically identifying, at each frequency, the Rayleigh-wave mode with the maximum vertical surface amplitude. This enables inversion without explicit mode indexing or assumptions about fundamental-mode dominance. The predominant-mode forward model is derived and implemented using the thin-layer method. The forward model is integrated into a particle-swarm-optimization global search algorithm and applied to invert three synthetic models exhibiting low-frequency mode jumps, using multiple layering parameterizations. Across all cases, the predominant-mode method accurately recovers major velocity contrasts and interface depths, whereas fundamental-mode inversions consistently overestimate Vs and mislocate deeper boundaries. The method is further validated using real field data from active and passive surface-wave measurements at the I15 Downhole Array Site. Inverted Vs profiles show strong agreement with downhole PS logs and with empirical transfer functions. Overall, the predominant-mode framework provides a robust approach for surface-wave inversion at sites with strong impedance contrasts.