Three-dimensional inversion of airborne data with applications for detecting elongated subvertical bodies overlapped by an inhomogeneous conductive layer with topography

We propose the approach to 3D inversion of airborne electromagnetic data, which is intended for discovering subvertical bodies overlapped by essentially inhomogeneous conductive layers. The approach is based on the geometric inversion in which a geoelectrical medium is parameterized with the use of block structures. During the inversion, the coordinates of the borders between the blocks and the rows of the blocks as well as resistivities inside them are determined. In order to solve the forward problem of the airborne electromagnetic survey, we use the non-conforming optimized mesh with the hexahedral cells, which enables us to reduce the number of degrees of freedom and smoothly approximate the curved borders of a geological medium. For a more reliable discovery of subvertical objects, we propose to carry out 3D inversions at several rotations of block structures relative to the flight lines. The workability of this approach is demonstrated using the data which are synthesized for complex geoelectrical models with topography, inhomogeneous overlapping layers and target subvertical bodies oriented differently relative to the flight lines. The results of this investigation show that, in some way or other, the elongated subvertical object is discovered and its orientation (the direction of its long side) is defined at different rotations of block structures used in 3D inversions. However, the most accurate recovery of the subvertical object length is achieved when the direction of its long side almost coincides with the direction of one of the block structures axes. Thus, the block structures rotations allow not only more reliably discovering a target object in complex geoelectrical conditions, but also more exactly defining its orientation and length.