Answers are grounded in this paper and cite their sources.
Three-dimensional (3D) geological modeling supports the design, construction, and risk assessment of underground and surface engineering projects. Observations may be insufficient because of cost, access, depth, and technical constraints, but information sufficiency also depends on geological complexity, regional knowledge, and the engineering decision. This narrative review synthesizes 50 studies identified through Google Scholar searches covering publications from 1990 to 2026. Studies were selected for their methodological representativeness and relevance to geological reconstruction under limited observations, uncertainty analysis, or engineering applications. Methods are compared by representation, inference mechanism, uncertainty treatment, and validation design. The evidence shows that implicit representation does not preclude probabilistic inference, borehole-level separation does not establish transfer to another geological domain, and reported prediction accuracy alone does not establish engineering reliability. Published synthetic density experiments and engineering cases illustrate why investigation density, response sensitivity, and validation conditions must be reported together. The review connects geological realizations to numerical response and decision thresholds and provides separate recommendations for practitioners and researchers. Heterogeneous protocols and incomplete reporting preclude a pooled ranking of method families.
No takes yet. Share an insight, caveat, or question.
Zhou et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: