Identifying marine transgressive-regressive depositional cycles in distal shelf mudstones from the Western Interior Seaway of North America using high-resolution (500 um) chemostratigraphy of the Upper Cretaceous Mancos Shale-Datasets
Description
High-resolution elemental data collected from an Itrax micro-X-ray fluorescence core scanner can be used to develop chemostratigraphic profiles of visually homogeneous mudstones that can potentially be correlated with near-shore facies with much higher accuracy than with lithology and well-log data alone. Our data, which was collected at 0.5-mm sampling resolution from 92 metres of the Upper Cretaceous Mancos Shale from North America, showed transgressive-regressive marine cycles based on siliciclastic input, terrestrial vs marine-dominant sedimentation, and relative redox and organic matter conditions in the sediment at sub-Milankovitch (i.e., millennial scale) frequency; this has not previously been documented in the literature over this length of core. By clustering the data using a hierarchical clustering algorithm known as a Self-Organising Map, we were able to also create detailed chemofacies for every sampling interval to compare the overall elemental signatures in the sediments and describe the depositional environment (prodelta, mudbelt, shelf) and relative proximity to shoreline. The elemental data are presented in total counts of fluoresced X-rays for each element. As the total number of fluoresced X-rays will be dependent on the energy of the incident X-rays as well as the exposure time, both of which are user-defined parameters, the data are semi-quantitative and do not provide information on the absolute amount (e.g., wt.%, ppm, ppb) of any given element. Each folder contains the individual elemental data files, based on core depth in feet below the surface, as tab-separated .txt files (e.g., C15_6557-6561.txt) as well as the accompanying core image (e.g., C15_6557-6561.tif) that is collected by the Itrax at the same time as the elemental data allow for direct comparison. Additionally, the full dataframe containing all intervals (MancosCore_final.csv), the row-wise centred dataframe used in the Self-Organising Map (MancosZ_final.csv), and the clusters (MancosClusters_final.csv) are also included as comma-separated values. For further explanation of the clustering, the reader is referred to the manuscript of the same title in Palaeogeography, Palaeoclimatology, Palaeoecology.
Files
Steps to reproduce
The chemostratigraphic profiles can be produced by simply plotting an element, or ratio (e.g., Fe/Ca), by Depth. Due to the high number of observations, the trends can be seen better using a moving average or smoother; we generally used a 39-point moving average which is equivalent to 2 cm of core length. To generate the clusters, we used the R package SOMbrero, with a 40 by 40 initial tesselation and a seed value of 5376; this seed must be used to obtain the exact same clustering dendrogram.