SEG 2014 Workshop: Hybrid strategies for modelling gravity gradient data
October 30, 2014
Colorado Convention Center
Denver, Colorado, USA | Website
Mark your calendar for the post conference SEG Gravity and Magnetics Workshop, where Gerry Connard will present the abstract, Hybrid strategies for modelling gravity gradient data, co-authored with Rob Ellis and Tom Popowski.
Thursday, October 30, 2014
W-12 - Gravity Gradiometry in Exploration Workflow
1:30pm - 6:00pm
3D modelling of gravity, gravity gradient, and magnetic data has emerged as a cost-effective strategy to improve imaging and interpretations in challenging seismic environments such as sub-salt (O’Brian et. al., 2005; Fichler et. al., 2007; Alvers et. al., 2014). However, to be useful, the inherent ambiguity of potential field data require strict integration with 3D seismic interpretations and other constraining data, giving rise to an extremely challenging inverse problem. The use of gravity gradient data is encouraged by studies that show including selected gravity gradient data in the modelling process can improve the resolution of the modelling (e.g., Zhdanov, et. al., 2004).
Modelling and inversion of potential field data has focused traditionally on two distinct model types: volume defined models (meshes, 3D voxel grids); and surface defined models (layers or surface boundaries). The interpreter would choose which type of model is most appropriate for their project and work within that framework. However, highly constrained inversion problems such as the sub-salt problem and advanced mineral exploration problems require a more flexible approach, one that incorporates the advantages of both volumes and surfaces. To this end, we have developed a 3D “Hybrid” modelling and inversion strategy that takes advantage of the two different ways of representing model components and also takes advantage of the strength of both space-domain and frequency-domain calculation algorithms. Hybrid models by definition contain both volumes and surfaces as core elements.
In this brief abstract we describe one application of hybrid models: we illustrate the hybrid model inversion for solving the base of salt problem using the SEAM density model (Pangman, 2007). We call our implementation of this hybrid model inversion strategy VALEM for Voxel Assisted Layer Enhanced Modelling.