Interpretation of airborne electromagnetic data with neural networks

Abstract

Artificial Neural networks (ANN) are used in the interpretation of airborne electromagnetic (AEM) data with horizontally layered earth structures. The problem is separated into three different tasks. At first an ANN is trained to interpret the AEM data with a homogeneous half-space, then a two layer model. Since the error of the ANN for the 2-layer model increases when data represents a half-space response, a separate ANN is trained to classify data as representative of a homogeneous half-space (HHS) or a 2-layer model (2LHS). The use of ANN classification for a 2-D (dyke) model is also discussed in this paper.