VARIABILITY IN SEDIMENTARY PETROGRAPHY: IDENTIFYING INCONSISTENCIES TO PRODUCE A REPRESENTATIVE SAMPLE SET

Frequently in sedimentary studies, the petrography of an entire study area is determined by the analysis of a few field samples (n<5). Unfortunately, in the case that a location exhibits significant variability in lithology, grain size, or mineralogy, a small sample size may not be representative of the area as a whole. We attempt to determine the potential for petrographic variability, and examine the minimum sample size necessary to accurately characterize the petrography of small sampling areas (approximately 1 km²). Samples were collected and analyzed from three locations in the Permian Cutler Group, located in the Paradox Basin of southwestern Colorado and southeastern Utah. These locations are Onion Creek and Fisher Towers, Utah, and Castro Draw, Colorado. The 3 localities represent proximal to distal settings of an alluvial fan, and span a distance of approximately 34 km. Samples were collected from each subenvironment identified in the sampling areas. For each sample, the grain size, mineralogy, and cement were analyzed in both hand sample and in thin section. At Castro Draw, 13 samples were collected and analyzed from proximal alluvial fan facies: debris flows, sheet flows, and fluvial deposits. Average grain size of these samples ranges from 3.3 - 1 φ, and the composition ranges from lithic arkose to arkose. At Fisher Towers, 6 samples were collected and analyzed from debris flows and fluvial deposits (distal fan/braid plain setting). Average grain size of these samples ranges from 1.5-0.9 φ, and the composition ranges from subarkose to arkose. Both calcite and iron cements are identified in thin section, and the amount of each cement changes drastically between samples only a few meters apart. Similarly, at Onion Creek, 9 samples were collected and analyzed from subenvironments of distal alluvial fan/braid plain facies: debris flows and fluvial deposits. Average grain size of the samples ranges from 4.3 - 0 φ, and the composition ranges from lithic arkose to arkose. Once again, these samples contain highly varied amounts of calcite and iron cement. In each locality, a high degree of petrographic variability exists within the small geographic sample area. We propose that sample sizes larger than those commonly used (n<5) are necessary to achieve a truly representative sample set for petrographic analysis.