Inversion of airborne time‐domain electromagnetic data to a 1D structure using lateral constraints

Abstract

We invert time‐domain airborne electromagnetic data in such a way as to obtain a model that varies slowly along the profile. This is achieved by modifying a typical one‐dimensional inversion algorithm to include lateral constraints. The lateral constraints are included as a roughness matrix that is solved simultaneously with the Jacobian matrix in an iterative eigenparameter inversion. In this case, multiple soundings along a line are all solved simultaneously. The lateral constraints can be applied to the resistivities and the thicknesses, both separately and together. We apply these techniques in two situations where airborne geophysical data are applied to near‐surface exploration. The first case is in a resistive environment where we are interested in quantifying a superficial conductive overburden. In this case, lateral constraints improve the geological image compared with those images obtained using unconstrained layered‐earth inversion. In the second case, we want to map the thickness of a resistive aquifer covering a saline layer. In this case, we show how varying the weights on the lateral constraints can change the image of the thickness of the aquifer. The presence of numerous cultural artefacts makes the inversion problematic. Application of a first‐difference constraint did the best job at removing culture but resulted in oversmoothing of the hydrogeology. The use of a second‐difference lateral constraint gave a good rendition of the hydrogeology but did not suppress the culture as well as the first‐difference constraint.

Regularization strategy for the layered inversion of airborne transient electromagnetic data: application to in‐loop data acquired over the basin of Franceville (Gabon)

“…Christensen (2002) developed a fast method called ‘One Pass Imaging’ which uses a regularization of the z ‐variability (in the vertical direction). Siemon et al (2009) and Vallée and Smith (2009) recently published other results obtained by regularization with horizontal constraints. In complement to all these approaches, in this paper we expose and compare three others methods to solve the 1D inverse problem of AEM data.…”

Theoretical Developments in Electromagnetic Induction Geophysics with Selected Applications in the Near Surface

“…The imaging algorithm is based on the system response to a pseudolayer half-space model in which a half-space model is overlain by a perfectly resistive layer that accounts for errors in the system altitude determination caused by factors such as altimeter errors and the occurrence of forest canopy. Vallée and Smith (2009) have performed a laterally constrained 1-D inversion of timedomain AEM data for hydrogeological applications. The lateral constraint restricts alongtrack model roughness such that only slowly varying 2-D conductivity models are permissible.…”

A parallel, scalable and memory efficient inversion code for very large‐scale airborne electromagnetics surveys

“…Recently, a promising low cost numerical strategy for full 2.5D modelling was published with application to CSEM systems (Streich, Becken and Ritter 2011). Constrained codes capable of utilizing the spatial coherence of a layered earth also exist (Santos 2004;Tartaras and Beamish 2005;Vallée and Smith 2009) but scalability issues severely limit the size of the problem that can be inverted at a time. However, we are unaware of any AEM implementations or applications in the literature.…”