Self-organizing maps for geoscientific data analysis: geological interpretation of multidimensional geophysical data

Data interpretation is a common task in geoscientific disciplines. Interpretation difficulties occur especially if the data that have to be interpreted are of arbitrary dimension. This paper describes the application of a statistical method, called self-organizing mapping (SOM), to interpret multidimensional, non-linear, and highly noised geophysical data for purposes of geological prediction. The underlying theory is explained, and the method is applied to a six-dimensional seismic data set. Results of SOM classifications can be represented as two-dimensional images, called feature maps. Feature maps illustrate the complexity and demonstrate interrelations between single features or clusters of the complete feature space. SOM images can be visually described and easily interpreted. The advantage is that the SOM method considers interdependencies between all geophysical features at each instance. An application example of an automated geological interpretation based on the geophysical data is shown.