A Pattern Classification Distribution Method for Geostatistical Modeling Evaluation and Uncertainty Quantification

Geological models are essential components in various applications. To generate reliable realizations, the geostatistical method focuses on reproducing spatial structures from training images (TIs). Moreover, uncertainty plays an important role Earth systems. It is beneficial for creating ensemble of stochastic realizations with high diversity. In this work, we applied a pattern classification distribution (PCD) to quantitatively evaluate modeling. First, proposed correlation-driven template capture geological patterns. According dependency TI, region growing and elbow-point detection were launched create adaptive template. Second, combination clustering was suggested characterize realizations. Aiming at simplifying parameter specification, program employed hierarchical decision tree categorize structures. Third, designed stacking framework develop multi-grid analysis. The contribution each grid calculated based morphological characteristics TI. Our extensively examined by channel model, 2D nonstationary flume system, subglacial bed topographic Antarctica, 3D sandstone models. We activated programs produce experimental results indicated that PCD capable addressing multiple categories, continuous variables, high-dimensional