Investigating near‐offset wave propagation in elastic media: A model study

When seismic energy is released in the subsurface, whether from an anthropogenic or natural event, a multi‐component wavefield is created containing body waves (compressional, P, and shear, S) and surface waves (Rayleigh and Love). Each component of the wavefield has its own characteristics of propagation, such as velocity, amplitude, frequency and particle motion relative to direction of travel. When a vertically oriented seismic source is initiated within an array of vertical‐component geophones, it is assumed that the direct arrivals comprise compressional (P‐) waves because we typically associate one with the other. The expectation that P‐waves would not be readily observed, given the perpendicularity of the source/geophone sensitivity to their particle motion, is valid for a simplified wave propagation model. However, real‐world near‐offset wavefield behaviour is more complex. Multiple modes are generated regardless of source orientation and interact in the near field before they have had enough time and/or distance to separate into different wavetrains. In this study, the near‐offset direct arrivals generated from a vertically oriented, impulsive source are investigated. The present work is a controlled synthetic demonstration investigating the development and modality of first arrivals in the near‐field environment. In the experiment presented here it was found that for a synthetic survey modelled on the surface, adjacent to a free‐surface boundary, the near‐offset first arrival contains a low‐velocity, large‐amplitude first arrival close to the assumed Rayleigh wave velocity. However, when the source and receivers are moved away from the free‐surface boundary, the first arrivals have a low‐velocity, large‐amplitude first arrival within near‐offsets travelling near the shear‐wave velocity. The implications of this near‐field effect are explored on inversion‐based techniques, such as travel time tomography.