First Published: GEO ExPro, September 2023

 

Abstract

Full Wave Inversion (FWI) delivers added value and detailed velocities enabling direct extraction of seismic images for improved interpretability and decision-making. In this example from offshore Liberia, FWI Imaging was used to address the shallow turbiditic setting that generates signal absorption and scattering, diminishing the image resolution and amplitude continuity of traditional seismic imaging. We would like to thank TGS management and NOCAL, National Oil Company of Liberia, for permission to show the seismic data used here.

The evolution of FWI Imaging technology

FWI is a sophisticated model-building technique. This method has traversed a remarkable journey, from its nascent stages in the 1980s to its near failure around 2000, battling with limited computational capabilities. With the exponential growth in computational pow­er over the years, FWI has experienced significant advancement and a place as a vital model-building technology in current workflows.

Figure: Original Kirchhoff PSDM line (no deghosting applied) from 2010 overlaid with the tomography-based velocity model from 2010 (the mass transit complexes in the shallow section generate signal absorption and scattering); b) Latest 2023 hi-res DM FWI(40Hz) derived velocity model and associated FWI Image (enhanced image resolution with increased S/N, better illumination, and improved amplitude consistency).

About 15 years ago, work on FWI increased dramatically. Development efforts were focused on employing acoustic FWI as a complementary velocity model building (VMB) tool. Over time, it has aided in generating better-focused images of subsurface features by utilizing more waveform information recorded by seismic receivers (phase, amplitude, and different types of waves) than traditional methods such as tomogra­phy. Recently FWI’s capabilities have expanded even further, transforming it into a technique capable of generating interpretable models that could start to be cross-validated with data from well measurements. In addition, the acoustic approximation is being changed to elastic FWI with a better theoretical agreement with the nature of field seismic data. All this is bringing new levels of confidence to FWI-derived models.

Read the full article here.