2026 GeoGulf Transactions, Vol. 75

TITLE

Machine Learning Lithofacies Prediction: Revolutionizing Reservoir Characterization

AUTHOR(S)

Alvaro Chaveste, Marwa Hussein, Robin Dommisse, and Hal Green

EXTENDED ABSTRACT

Traditional seismic inversion methods for lithofacies identification face significant limitations, including low dimensionality, resolution constraints, and dependence on physical models that may not accurately represent subsurface complexity. We present a novel machine learning (ML) approach that overcomes these challenges by leveraging self-organizing maps (SOMs) and k–means clustering to directly predict lithofacies from seismic and well data without requiring adherence to predetermined physical models.

...

PAGE(S)

175-178

DOI

https://doi.org/10.62371/KLZS6351

Purchase Link via the Bureau of Economic Geology Bookstore is not yet available. Check back later for update.