Abstract

Distinguishing different lithostratigraphic formations that are in contact with each other can be problematic when they are of like facies, appearance, and provenance. For example, formations of like facies that are vertically separated by a subtle unconformable surface are difficult to distinguish, especially if multiple candidate surfaces exist. Similarly, time-equivalent formations of like facies that were separated into distinct subbasins by a topographic divide may also be challenging to distinguish. In this study, formations of this nature were discriminated using statistical discriminant analysis of X-ray fluorescence (XRF) elemental data that included both mobile and immobile elements. This discrimination has important consequences for stratigraphic analysis. For example, accurate placement of an unconformable surface within a stratal succession is important because it affects the interpreted thickness of units and sequences in contact with that surface. Unit thickness influences the interpretation of basin subsidence and its timing, depositional style of basin fill, paleotopography, and diagenesis.

When developing discriminant models, data must be collected from outcrop or well cores where formations bounding the unconformity or topographic divide are known. The models next establish and high-grade key elemental differences between the known formations. The models can then be used to classify samples from unknown formations to one of the known formations. In our study, samples from three lithostratigraphic formations were used to build two models: one grouped by facies trends, and one grouped by geographic location or tectonic feature. Results demonstrate that discriminant analysis of elemental XRF data can be used to distinguish these formations both vertically and laterally. Our models relied on both mobile and immobile elements to make the discrimination, implying that as long as sandstone formations of similar provenance have experienced different diagenetic histories (i.e., cements and alteration products) they can be distinguished in the rock record.

This study applied the methodology of discriminant function analysis to identify the J-1 unconformity on Utahs Colorado Plateau. The J-1 unconformity is a regional unconformity that separates the Early and Middle Jurassic. Three formations of similar facies, appearance, and provenance are all in contact with this unconformity. The Early Jurassic Navajo Sandstone underlies the Middle Jurassic Page Sandstone in the eastern part of the study area, and the Middle Jurassic Temple Cap Formation in the western part of the study area. In much of the central part of the study area the stratigraphic position of the J-1 unconformity is unknown and the overlying formation is less well established. The J-1 unconformity appears to have been a baffle or barrier to fluid flow in much of this area. Also, paleotopography on the Navajo Sandstone appears to have separated the overlying formations into two subbasins. Thus, all three formations had different diagenetic histories that facilitated the discriminant modeling. The methods used in this study can be applied to similar sedimentary packages that were likewise partitioned vertically by fluid-flow divides such as unconformities and hiatuses, or laterally by topographic, hydrodynamic, structural divides, or lateral facies changes.