Interpretação sísmica avançada de canais em águas profundas usando análise multiatributo e aprendizado de máquina: estudo na Bacia do Ceará, Margem Equatorial Brasileira

Abstract

The Brazilian Equatorial Margin (BEM) has garnered scientific interest due to the discovery of oil and gas reserves in its African counterpart and the offshore basins of Guyana, Suriname, and French Guiana. Foz do Amazonas, Potiguar, and Ceará basins stand out as promising targets due to their geological similarities, including the presence of reservoirs formed by turbidite. The Ceará Basin, part of the BEM, is subdivided into the Mundaú, Icaraí, Acaraú, and Piauí-Camocim sub-basins. This study aims to investigate the evolution of a channel system formed between the Albian and Maastrichtian, during the drift phase of the Icaraí sub-basin, a region less explored compared to the Mundaú sub-basin. A seismic volume of 2700 km², provided by TGS and processed using Full Waveform Inversion (FWI), was utilized, consisting of depthmigrated data with 4240 inlines and 2922 crosslines, spaced 12.5 meters apart. The research employs machine learning methods to interpret the evolution of these channels and infer possible infilling lithologies, overcoming the scarcity of well data in the area. The method used in this study was based on multi-attribute analysis and Self-Organizing Maps (SOM) to indirectly interpret the lithologies within the channel. Two channels were identified, both showing stratigraphic control with variations in internal patterns reflecting changes in depositional or erosional processes. The width of the channels in the proximal zone is approximately 2 km, increasing to 4 km in the distal zone. To better describe this evolution, a hierarchical classification model was applied, segmenting the study into proximal, intermediate, and distal areas. The results show that the evolution of the channels occurs distinctly as they move away from the continental margin. Initially, both channels exhibit highly confined and erosive characteristics, indicating high-energy conditions. As they progress into more distal areas, the channels become wider, potentially forming submarine fans. This widening promotes the development of significant turbidite elements (levees, overbanks), with sediment dispersal to the east, accompanied by the deposition of materials transported by the two channels. Additionally, large mass transport complexes were identified in the system, associated with variations in channel energy levels. These complexes not only influence depositional dynamics but also induce directional changes in the channels throughout their evolution. The Icaraí subbasin, still largely underexplored and with no deep-water well records, showed significant potential for increased exploratory studies through the application of machine learning methods combined with high-resolution seismic data. This approach enabled advances inthe geomorphological and stratigraphic understanding of the region, contributing to a more comprehensive understanding of the dynamics and evolution of the Late Cretaceous interval of the basin.