An example of holistic inversion of time domain AEM data

Abstract

An inversion algorithm for time-domain airborne electromagnetic (AEM) data has been developed. A complete flight line of data are inverted in each inversion to simultaneously recover a 2D layered earth conductivity distribution and selected unmeasured elements of the AEM system geometry. The conductivity and thickness of each layer in the earth model and the system geometry, are all parameterised by separate along line 1D cubic B-splines. It is the coefficients of these splines that become the unknowns that we solve for in the inversion. This new method is motivated by the success of three separate concepts: the ?holistic approach? to the inversion and calibration of frequency-domain AEM data (Brodie and Sambridge, 2006); laterally constrained inversion (Auken et al., 2005; Vallée and Smith, 2008); and simultaneous estimation of subsurface conductivity and system geometry parameters (Lane et al., 2004; Sattel et al., 2004). The method could be described as a form of laterally constrained inversion with a spline model parameterisation and the additional capability of solving for system geometry.