Utilization of multimode surface wave dispersion for characterizing roadbed structure

Abstract

Accurate measurement of shear (S)-wave velocity in a roadbed subsurface with a velocity reversal is crucial as well as challenging. A high frequency Rayleigh wave survey was carried out on a roadbed site from Henan, China to determine S-wave velocities to approximately 15 m depth. The phase shift approach was used to directly construct a high-resolution image of multimode dispersion curves in the frequency–velocity ( f– v) domain from a multi-channel raw shot gather with a high signal-to-noise (S/N) ratio. The measured fundamental and two higher-mode dispersion curves were then inverted by genetic algorithms to reconstruct the fine structure of the tested roadbed subsurface. Our case study demonstrates that an inversion performed with only fundamental mode data with a limited resolution and a high degree of error may yield an unrealistic model. A better agreement with direct borehole measurements, however, is obtained when the second-mode data are inverted simultaneously with the fundamental mode data, which clearly reveals a 1-m thick softer soil layer between two stiff soil layers in the tested roadbed. A higher accuracy S-wave velocity profile is achieved by simultaneously incorporating the three modes of data into the inversion process. This representative case study highlights advantages of fully exploiting intrinsic multimodal properties of Rayleigh waves and greatly enhances our confidence of accurately imaging and characterizing a complex subsurface.