Ring Structures of the Norilsk Metallogenic Region: Evidence from Neural Network Analysis of Areal Geophysical Data

AbstractA complex geophysical and mathematical model of the Norilsk metallogenic region, consisting of fifteen levels of hierarchy, was constructed using the method of automatic express interpretation of areal geophysical data developed by the authors. The model is based on the mathematical approach of fuzzy logic, including artificial neural networks without a teacher using the method of self-organizing Kohonen maps. The formation of groups of clusters that characterize the likely correlations of multidimensional geophysical data was substantiated and the relationships between them were analyzed by identification of correlations. Various geophysical transformants were analyzed using self-organizing Kohonen maps. A number of input representative indicators of their groups were determined, which allowed us to construct a geophysical–mathematical model in the form of a two-dimensional map of clusters using fuzzy logic tools. Terminological sets for each group of clusters were formed; the type of membership functions of previously unknown geological objects was set based on the newly interpreted data. The clusters calculated according to the results of the study reflect the northwestern fragment of the fields of basaltic magmatism of the trap formation, including two ring structures. The model ring structure in the western part of the studied area reflects the Bolgokhtokhskii granodiorite stock. The second ring structure, Pyasinskaya, which was previously unknown, indicates the presence of a dome-shaped object at a great depth, which is a likely source of root zones of intrusions of the Norilsk Formation. The experimental study confirmed the adequacy of the constructed model and the effectiveness of its application for the purpose of express analysis of geophysical data and for making decisions on geological prospecting tasks.