Application of the gradient structure tensor analysis method to the prediction of channel sand from 3D seismic data

Li Yi,Zhong Guangfa,Song Jisheng,Wang Ping,Wu Xiaohua
DOI: https://doi.org/10.3787/j.issn.1000-0976.2011.03.011
2011-01-01
Abstract:Gradient structure tensor analysis is a new seismic attribute analysis method introduced into seismic interpretation in recent years.It can be used to identify sedimentary topographic units with different lamination features in seismic data volume.On seismic profile,channel fill patterns mainly include onlap fill,lateral aggradations,composite fill,and fills in chaotic and blanket reflection.As to lamination on images,the former 3 types of fills show curved layered lamination,while the later 2 types of fills show chaotic lamination.Through a case study of the Lower Jurassic reservoir of fluvial facies in the northern Sichuan Basin,gradient structure tensor analysis is performed on 3D seismic data to obtain the anisotropy parameters reflecting the lamination features of channel deposits.The strata slice is adopted to describe the geomorphic unit and spatial distribution of the channels with satisfactory results.In comparison with the conventional methods for obtaining coherent and geometry attributes,this method has a strong noise-proof ability and is more sensitive to sedimentary topographic units such as channels.
What problem does this paper attempt to address?