Geostatistical Interpolation Methods for Cobalt in Indiana Coal Data: A Comparative Analysis of Spatial Interpolation Techniques

The increasing global demand for critical minerals driven by the transition to renewable energy technologies has heightened interest in unconventional mineral sources, including coal deposits. This study examines the spatial distribution of cobalt within Indiana coal fields using geostatistical interpolation techniques to improve predictability and resource assessment. Three interpolation methods Inverse Distance Weighting (IDW), Simple Kriging, and Ordinary Kriging were applied to geochemical data derived from the U.S. Geological Survey COALQUAL database. Model performance was evaluated using cross-validation statistics, including root-mean-square error and standardized error measures. The results indicate that kriging-based approaches outperform the deterministic IDW method, with Ordinary Kriging demonstrating the highest predictive accuracy, yielding lower root-mean-square errors (4.3052–4.7696) compared to IDW (4.8059). Semivariogram analysis revealed a nugget value of 0.360 and a major range of 3049.50 units, confirming the presence of meaningful spatial autocorrelation in cobalt concentrations across the study area. These findings establish Ordinary Kriging as the most reliable method for interpolating cobalt distribution in Indiana coal deposits and highlight the value of geostatistical modeling for enhancing mineral resource evaluation. The study provides methodological guidance for assessing trace elements in sedimentary environments and supports the strategic exploration of critical minerals from nonconventional sources relevant to clean energy technologies.