Unsupervised characterization using high energy amplitude: 3D seismic case study

Abstract The computing aspects of an unsupervised clustering technique based on the correlation integral (CI) method were investigated. Two parameters were derived from the CI equation: the scaling exponent (SE) and its double derivative (DD). These parameters are suitable attributes for automated seismic data interpretation and imaging. This unsupervised clustering technique can highlight channels, gas chimneys, faults, progradation clinoforms, and other geological features. For the first time, 3D volumes of SE and DD were converted to SEGY format to use them for further interpretation with professional software. The new CI computational workflow decreased the calculation stride, which caused a dramatic increase in SE and DD image resolution. At the same time, there was a marginal increase in the calculation time. However, a more significant calculation stride was found suitable for wide-scale observations and a smaller stride for detailed analysis.