An Inversion Toolbox for Frequency- and Time-domain Airborne Electromagnetic Data from Surveys in Ireland

Summary We have developed a highly flexible toolbox using the Python language for 1-D inversion of AEM data along the flight lines. The computational core is derived from the well-tested AMIRA code suite developed by CSIRO (Australia) and the AMIRA consortium. Different inversion methods were implemented as (i) constrained Tikhonov-type inversion including optimal regularisation methods, (ii) Bayesian MAP inversion in parameter and data space, and (iii) Full Bayesian inversion with Markov Chain Monte Carlo techniques. In addition we have included different methods for improving the spatial consistency as moving windows in the data and parameter domain, and PCA filtering. This contribution will introduce the toolbox, and present synthetic and field studies from Ireland (Tellus and legacy data).