Improving near-surface anomaly detection via a full-waveform inversion scheme integrated with seismic diffraction tomography

Near-surface seismic data often contain high levels of coherent and non-coherent noise, making seismic imaging a challenging task. Furthermore, the targets of interest are usually small in size and irregular in shape. Geophysicists often focus on different parts of seismic record to extract information related to the near-surface targets. Accordingly, various geophysical inversion schemes are often implemented. For example, traveltime tomography is a robust technique that focuses on the first arrival-time of refracted waves. Another conventional seismic method is based on dispersion analysis of the surface waves and outputs the shear wave velocity profile. Since these methods suffer from rather low resolution, reflection-based full waveform inversion scenarios have gained interest in recent years. These techniques are adapted from industry-scale seismology and are applied to near-surface data sets in order to enhance the resolution. Waveform inversion schemes are more sensitive to starting parameters relative to the other inversion methods. In this work, we present a brief overview of the role of diffraction in increasing the resolution of inversion-based techniques such as seismic traveltime tomography and full waveform inversion. Furthermore, we investigate a diffraction-based velocity model building technique and evaluate its potential to be integrated into full waveform inversion.