The behavior of least-squares extended reverse time migration for vertical seismic profiling data

Vertical seismic profiling (VSP) survey differentiates itself from surface seismic survey by deploying receivers inside the borehole, while the sources are excited on the surface. Due to this distinctive acquisition geometry, special attentions should be paid in processing and imaging the VSP data. Considering the fact that velocity errors are inevitable for field data application, it is better to utilize least-squares reverse time migration with subsurface offset extension (LSERTM) because of its good property of preserving data information in the extended domain. LSERTM has been actively studied for surface seismic data in academic research, but it is never reported with application to VSP data. We fill this gap by applying LSERTM to VSP data with a simple layered model, and analyze the behavior of the subsurface offset common image gathers (SOCIGs) and the angle domain common image gathers (ADCIGs). Numerical examples demonstrate that there are significant differences in the subsurface offset behavior between the VSP survey and the surface seismic survey. We conclude that the strategies functioning well in surface seismic processing workflow may not be suitable for direct application to VSP data.