Determining Seismic Stratigraphic Features Using Dip-Guided Symmetry Attribute
Objectives Symmetry is a fundamental attribute in human visual perception as well as many natural structures. As described herein, a symmetry analysis may be used as a new seismic attribute to measure the degree of chaotic in seismic data. Symmetry is a new type of seismic volumetric attribute which is sensitive to the seismic amplitude variations and therefore, strongly correlates with discontinuities and curvatures associated with imaged geologic surfaces. The 3-D symmetry attribute can be used as a tool to identify faults, fractures, channels and other stratigraphic features. Procedures We first calculate the dip attribute with 2-D log-Gabor filter array, then use the dip attribute as a guide to build a 3-D window for the degree of symmetry evaluation. For each sample in the volume we evaluate the 180-degree rotational symmetry in its neighborhood. The neighborhood is defined by a 3-D cylindrical window which is corrected by the dip angle. The 180 degree rotational symmetry is then evaluated by comparing pairs of traces opposite to the window center and an overall degree of symmetry is derived for the whole window. In our implementation GPU computing technologies are used to optimize performance. Results Symmetry attribute provides a sharp stratigraphic image. Compared with other existing attribute calculation techniques, such as correlation-based, semblance-based, variance-based, eigenstructure-based, gradient structure tensor-based, and least-square-based a symmetry attribute calculation more effectively shows subtle local symmetry variations in a circular pattern of adjacent seismic wave forms. Conclusions Symmetry attribute is a new seismic attribute which provides higher resolution imaging of many subtle stratigraphic features that may be missed using other types of attribute analysis. An attribute based on the symmetry data can be used to provide a high quality seismic attribute suitable for fault, channel and geo-body picking.
AAPG Datapages/Search and Discovery Article #90189 © 2014 AAPG Annual Convention and Exhibition, Houston, Texas, USA, April 6–9, 2014