Using simulated maps to interpret the geochemistry and formation of the Blue Gem Coal Bed, Kentucky, USA
This study presents geostatistical simulations of coal-quality parameters, major oxides and trace metals for an area covering roughly 812 km2 of the Blue Gem coal bed in southeastern Kentucky, USA. The Blue Gem, characterized by low ash yield and low sulfur content, is an important economic resource. Past studies have characterized the Blue Gem's geochemistry, palynology and petrography and inferred a depositional setting of a planar peat deposit that transitioned to slightly domed later in its development. These studies have focused primarily on vertical geochemical trends within the coal bed. Simulated maps of chemical elements derived from 45 measured sample locations across the study area provide an opportunity to observe changes in the horizontal direction within the coal bed. As the Blue Gem coal bed shows significant vertical chemical trends, care was taken in this study to try to select samples from a single, middle portion of the coal. By revealing spatial distribution patterns of elements across the middle of the bed, associations between different components of the coal can be seen. The maps therefore help to provide a picture of the coal-forming peat bog at an instant in geologic time and allow the interpretation of a depositional setting in the horizontal direction. Results from this middle portion of the coal suggest an association of SiO2 with both K2O and TiO2 in different parts of the study area. Further, a pocket in the southeast of the study area shows elevated concentrations of elements attributable to observed carbonate-phase minerals (MgO, CaO, Ba and Sr) as well as elements commonly associated with sulfide-phase minerals (Cu, Mo and Ni). Areas of relatively high ash yield are observed in the north and south of the mapped area, in contrast to the low ash yields seen towards the east. Additionally, we present joint probability maps where multiple coal-quality parameters are plotted simultaneously on one figure. This application allows researchers to investigate the associations of more than two components in a straight-forward manner useful in guiding resource exploration.