Abstract

This study introduces a global seismic interpretation workflow that combines Relative Geological Time (RGT) modelling with machine learning to build high-resolution, stratigraphically consistent frameworks. Unlike traditional approaches that rely on limited horizons, this method utilizes all seismic samples through amplitude-driven and automationassisted processing. It supports fault and salt body extraction, waveform classification, and subseismic stratal slicing. Applications span diverse basins and depositional settings, from siliciclastic to carbonate systems, and from passive margins to complex tectonic regimes. Results are compiled into a catalogue showcasing geological expressions emphasized by advanced stratigraphic and structural attribute generation. This integrated approach enhances early reconnaissance, stratigraphic trap analysis, and resource evaluation across complex, data-limited environments