Use of seismic attributes for sediment classification

Abstract

A study to understand the relationships between seismic attributes extracted from 2D high-resolution seismic data and the seafloor's sediments of the surveyed area. As seismic attributes are features highly influenced by the medium through which the seismic waves are propagated, the authors can assume that it would be possible to characterise the geological nature of the seafloor by using these attributes. Herein, a survey was performed on the continental margin of the South Shetland Islands in Antarctica, where both 2D high-resolution seismic data and sediment gravity cores samples were simultaneously acquired. A computational script was written to extract the seismic attributes from the data, which have been statistically analysed with clustering analyses, such as principal components analysis, dendrograms and k-means classification. The extracted seismic attributes are the amplitude, the instantaneous phase, the instantaneous frequency, the envelope, the time derivative of the envelope, the second derivative of the envelope and the acceleration of phase. Statistical evaluation showed that geological classification of the seafloor's sediments is possible by associating these attributes according to their coherence. The methodologies here developed seem to be appropriate for glacio-marine environment and coarse-to-medium silt sediment found in the study area and may be applied to other regions in the same geological conditions.