Elastic FWI of land W-VSP data: A case study in the Sichuan Basin of China

Abstract

Abstract Elastic full-waveform inversion (FWI) methods are expected to construct high-resolution subsurface elastic properties, which are of great importance for accurately characterizing and delineating hydrocarbon reservoirs. However, elastic FWI for land seismic data is challenging due to low signal-to-noise ratio data, complex near-surface environments, unknown source parameters, etc. In this study, elastic FWI is applied to multicomponent land walkaway vertical seismic profile data acquired in the Sichuan Basin of Southwest Oil and Gas Field in China. A series of methods and strategies are used to overcome the difficulties for elastic FWI. For example, the spectral element method with irregular mesh is used for forward modeling and inversion with complex topographic variations. The source wavelets are estimated shot by shot using the direct P waves. Vertical transverse isotropy wave-equation traveltime tomography with a velocity-based model parameterization is first conducted to invert for the low-wavenumber velocity and anisotropy structures. This step helps reduce the cycle-skipping problem and multiparameter coupling effects. Finally, reflected P waves and P-S converted waves are extracted from the shot gathers for constructing the detailed elastic impedance profiles, which provide valuable information for identifying the potential reservoir zones.