Examples of Improved Inversion of Different Airborne Electromagnetic Datasets Via Sharp Regularization

Large geophysical datasets are produced routinely during airborne surveys. The Spatially
Constrained Inversion (SCI) is capable of inverting these datasets in an efficient and effective
way by using a 1D forward modeling and, at the same time, enforcing smoothness constraints
between the model parameters. The smoothness constraints act both vertically within each 1D
model discretizing the investigated volume and laterally between the adjacent soundings. Even if
the traditional, smooth SCI has been proven to be very successful in reconstructing complex
structures, sometimes it generates results where the formation boundaries are blurred and poorly
match the real, abrupt changes in the underlying geology. Recently, to overcome this problem,
the original (smooth) SCI algorithm has been extended to include sharp boundary reconstruction
capabilities based on the Minimum Support regularization. By means of minimization of the
volume where, the spatial model variation is non-vanishing (i.e., the support of the variation),
sharp-SCI promotes the reconstruction of blocky solutions. In this paper, we apply the novel
sharp-SCI method to different types of airborne electromagnetic datasets and, by comparing the
models against other geophysical and geological evidences, demonstrate the improved
capabilities of in reconstructing sharp features.