Abstract Brazilian pre-salt carbonates are notoriously heterogeneous and heavily overprinted by multi-phase diagenesis, making reservoir-quality prediction away from well control particularly challenging. The spatially continuous constraints available for subsurface characterization are typically limited to inversion-derived geophysical attributes, most commonly acoustic impedance (IP) and, where pre-stack data quality permits, VP/VS. Using two wells from the Atapu field (Santos Basin) as a well-log–scale proxy for geophysical attributes available away from wells, this study evaluates the incremental information content of different attribute combinations for supervised facies classification. Core, thin-section, and well-log data were integrated into three reservoir-quality–driven classes, and Random Forest models were trained using four feature scenarios: IP-only, IP + VP/VS, IP + deep resistivity (RESD), and full-log suite. Model performance was assessed through blind-well testing, and attribute value was quantified using an information-theoretic framework based on entropy reduction. Adding VP/VS to IP yields only modest information gain: relative to the IP-only baseline, facies entropy reduces moderately, whereas combining IP with deep resistivity produces substantially larger gain, approaching the full-log-suite reference. The incremental value of VP/VS over IP is measurable but limited, less than that provided by attributes directly linked to pore connectivity. Quantifying information gain, rather than classification accuracy alone, provides a robust framework for evaluating geophysical attribute combinations away from well control. The workflow offers a practical template to assess the incremental value of elastic and non-elastic attributes for seismic-driven reservoir-characterization strategies in heterogeneous carbonate settings.
Petersohn et al. (Mon,) studied this question.