Inferences of Source Lithologies for Chicxulub Microtektites Using a Bayesian Approach

Abstract

Abstract The Chicxulub impact on the Yucatán Peninsula at the Cretaceous‐Paleogene boundary (KPB) was a likely contributor to the end‐Cretaceous mass extinction. Glassy objects produced by quenched melt from the impact were distributed over a large region centered on the Caribbean basin and have long been known to preserve compositional information that could allow for more robust constraints on the pre‐impact target lithologies. The Chicxulub‐derived glasses are generally altered in most localities, but a recently discovered deposit at Gorgonilla Island, Colombia, yields a large percentage of unaltered glassy microtektites up to ∼2 mm (There is inconsistent usage of these terms in the literature. We use the term microtektites herein to distinguish them from e.g., the tektites of the Australasian strewn field, which reach sizes many orders of magnitude larger) in dimension. Backscattered electron and X‐ray imaging reveal that most of the Gorgonilla microtektites are composed of intricate mixtures of diverse compositions juxtaposed in schlieren textures. Electron probe microanalysis (EPMA) of 521 analyses in 90 unaltered microtektites encompassing the observed range of colors and shapes, reveals a continuous range of compositions. The EPMA data reveal a binary mixture of granitic and Ca‐S‐rich components as has been previously inferred for these and other KPB microtektites. We formulate a reversible jump Bayesian model and infer that a six‐component endmember mixture of granitoid (∼70%), limestone (∼14%), mafic igneous (∼10%), shale (∼4%), anhydrite (∼0.3%), and halite (∼0.1%) can adequately explain the observed data set. From our model‐derived lithology budget, we make conservative estimates of lower bounds for the volatile budgets of sulfur (1.8−0.8+3.9 Gt), CO2 (144−77.5+136 Gt), and H2O (297−157+511 Gt). Combining our model estimates of target lithology with impact simulation may allow for more accurate predictions of the subsequent climate effects.

Keywords

• Gorgonilla microtektites
• Bayesian analysis
• Chicxulub impact
• source lithologies
• Markov Chain Monte Carlo
• climate and recovery