POSSIBILITIES OF THE SOM METHOD APPLICATION FOR PETROPHYSICAL RECONSTRUCTIONS OF SEDIMENTATION BY THE EXAMPLE OF THE IRTYSH SEDIMENTARY BASIN

The Irtysh sedimentary basin represents a complex geological system characterized by significant lithological heterogeneity and insufficiently studied stratigraphic structure. This study aims to refine the lithological composition of the section, stratigraphic subdivision of deposits, and assess the reservoir properties of rocks based on an integrated analysis of geophysical, petrophysical, and petrographic data. For the first time in this region, the Self-Organizing Maps (SOM) method was applied — a neural network-based machine learning algorithm. The method was used for automatic classification of geophysical and petrophysical parameters, enabling the identification of lithofacies complexes, refinement of stratigraphic boundaries, and detection of distribution patterns in the rock succession. The analysis resulted in the identification of 10 lithofacies complexes, detailed characterization of Paleozoic deposits, and delineation of the crystalline basement. Petrographic examination of core samples revealed well-developed secondary porosity caused by tectonic and diagenetic processes, which significantly affects the reservoir potential of rocks. The results refine the stratigraphic framework of the basin, supplement the regional geological database, and can be used to forecast potential reservoirs, evaluate reservoir properties, and assess the prospects of productive horizons as potential CO2 storage reservoirs.