Ultra-Resolution Nodal Seismic and Machine Learning for Geohazard Analysis

Shallow geohazards are found in areas where near surface formations such as carbonates or evaporites are subject to dissolution related to the circulation of water. We develop a workflow for rapid and detailed screening of the shallow subsurface in search of potential geohazards. Seismic acquisition is performed with dense nodal arrays providing a detailed spatial sampling of the wavefield. Automated data analysis involves transmission surface-consistent decomposition and corrections, followed by beam forming with virtual super gathers. Machine learning is utilized for different tasks of processing and inversion of multi-wave data. A collection of surface-consistent attributes and geophysical parameter distributions with depth is generated. The ultra-resolution seismic and the automated data analysis is successfully applied for the evaluation of geohazards at several drilling locations.