Inversion of MT Data with Enhanced Structural Fidelity presented by Alessandra Giovanna Battaglini

Abstract

Oil and gas exploration in foothill areas is challenged by several factors. The accessibility is not straightforward due to steep terrains and, in some cases, also due to forest coverage; thus, making both exploration and development plans far more expensive. Satellite imagery and airborne acquisition techniques can be used to cover large areas in an economic manner and to gain a preliminary knowledge of the subsurface. Unfortunately, these techniques lack the resolution required to infer the geologic setting at depth and to build a structural model with adequate confidence; this is the time when it becomes necessary to step into ground exploration. Seismic imaging techniques are the main tools used to build an accurate description of the subsurface; however, in thrust-belt areas, the signal-to-noise ratio of the seismic reflections is severely reduced by the steep topography and by the almost vertical layering of the formations. The magnetotelluric method has been successfully used as a complementary tool in the exploration of foothill plays but its intrinsic non-uniqueness and its reduced structural fidelity, with respect to a seismic section, make its interpretation more complex. Since the solution of a non-unique inverse problem is heavily dependent on its starting point, a traditional method to tackle the non-uniqueness relies on the careful definition of the starting model, used as a tool for injecting the available a-priori information into the inverse problem. In this work we present an alternative approach to the exploitation of the a-priori knowledge that, instead of embedding it into the starting model, relies on the continuous assimilation of the structural/geological information into the resistivity volume. The method we propose for the inversion of the MT soundings tries to maximize, simultaneously, the data-fit and the structural fidelity: the structural fidelity is defined in least-square sense as the similarity the current resistivity model and the “guiding model” which expresses and synthesizes the a-priori knowledge in terms of shape and geometry of the structures. We discuss the advantages deriving from the application of this technique by mean of a synthetic example representative of a foothills play.

AAPG Datapages/Search and Discovery Article #90332 © 2018 AAPG International Conference and Exhibition, Cape Town, South Africa, November 4-11, 2018