Abstract

The study of subsoil, by nature inaccessible to direct observation, relies on sediment cores analysis, representing a fundamental information source. Sedimentary facies, in particular, i.e. sediment bodies or packages of strata formed in specific depositional environments, is essential for a wide range of scientific applications such as climate change studies, engineering geology, land subsidence calculation, and reservoir characterization. High-resolution facies analysis requires specific skills and training that can be a limitation to a proper understanding of the subsoil. This dataset consists of a robust database of digital images of sedimentary cores acquired in the Po Plain and the Adriatic coastal plain of Marche, Abruzzo, and Apulia regions (Italy). This database has been used to perform semantic segmentation of sedimentary facies using conventional neural networks, identifying six target classes that reflect a wide spectrum of continental to shallow-marine depositional environments.