Advanced Geological Interpretation by Directivity-driven Seismic Imaging
Advanced seismic interpretation technology applied to post-stack migrated images provides very valuable information about the subsurface geology. The target-oriented, directivity-driven EarthStudy 360 imaging and interpretation system can be considered an augmented “magnifying glass” for enhancing the imaging quality, resolution and reliability of specific areas of interest, especially of oil- and gas-producing reservoirs. It can also be an extremely important tool for better understanding and identifying potential reservoirs to store CO2, as part of the global effort to preserve the environment.
This lecture will begin with a focus on the high sensitivity of seismic images to the different directivity components of waves illuminating from different target geological objects. It will be followed by a brief review of the potential of well-known geometrical seismic attributes applied to 3D post-stack migrated images (e.g., coherency – directivity-driven semblance, gradient- and curvature-based attributes), to emphasize the different structural and stratigraphic characteristics of the subsurface geology.
The lecture will then cover the novel approach used by the EarthStudy 360 imaging system to provide directivity-driven seismic imaging amplitudes and geometrical attributes (ordered as directional gathers or volumes) directly from pre-stack seismic data, with higher resolution, accuracy and reliability. This is particularly relevant to reservoir regions which are characterized by complex stratigraphy broken by natural fracture systems, located within geological traps containing unconformities and conflicting dips.
Finally, we will describe a more general decomposition approach, principal component analysis (PCA), applied to the directivity-driven imaging data provided by EarthStudy 360. The high value principal components are associated with certain energy directions that contribute to the different geological objects of interest, such as smooth/rough curved surfaces, steep dip fault plans crossing the geological layers, buried rivers/channels, and even small-scale faults and fracture systems. Coherent events due to acquisition footprints or ambient noise can also be decomposed with this method. The high-fidelity decomposition provides an attractive platform for training and labeling the geological images, making it reliable input for Deep Learning interpretation and characterization. This innovative approach has great potential for the future.
Zvi Koren founded the geophysical program at Paradigm in 1990 and headed the team responsible for developing the GeoDepth system for velocity model building, seismic modeling and depth imaging. He also founded the EarthStudy 360 project, which is an innovative, full wavefield exploration and development system based on full-azimuth angle domain imaging and analysis.
Dr. Koren has won numerous research awards and fellowships, and has published over 100 articles in his fields of expertise, including seismic wave propagation, seismic imaging and the study of the structure of the Earth's crust, in particular geophysical modeling and inversion methods.
He holds a Ph.D. in Geophysics from the Department of Geophysics and Planetary Sciences at Tel Aviv University. He performed post-doctoral research at the Institute Physique du Globe, Paris University, and served as a researcher on a 3D modeling project at the Geophysics Department of Hamburg University.