Sill Cube: An automated approach for the interpretation of magmatic sill complexes on seismic reflection data

Abstract

The use of conventional interpretation techniques like manual interpretation, seismic attribute analysis and opacity rendering or geobody extraction have demonstrated significant success in interpreting magmatic sills from seismic reflection data. Nonetheless, the downsides of seismic imaging and interpretation of magmatic sills are that they are rarely drilled, or too thin to be detected. We present an automated approach for interpreting magmatic sills from seismic reflection data by using a newly designed meta-attribute, called as the Sill Cube (SC). The SC meta-attribute is generated by amalgamating a number of different attributes that are trained over the interpreter's knowledge of the sill networks through a supervised scheme of neural learning. The approach has shown its efficiency by delivering enhanced images of magmatic sills from seismic data in three frontier areas i.e., the Kora prospect (offshore Taranaki Basin, New Zealand), the Utgard and the Tulipan prospects (both offshore Vøring Basin, Norway). The SC meta-attribute has completely captured the saucer-shape geometry of most of the sills. Additionally, the meta-attribute has brought out different structural elements such as the limbs, bridges, junctions and fingers and highlighted the overlying forced folds. The interpreted sills cover areas of 6–16 km2 in the Kora prospect, 105–1140 km2 in the Utgard prospect and 15–60 km2 in the Tulipan prospect. Such an interpretational approach does not only honour the interpreter's knowledge of sills networks but also add values in understanding the tectonic architecture.