Curation and Analysis of Global Sedimentary Geochemical Data to Inform Earth History

Abstract

Large datasets increasingly provide critical insights into crustal and surface processes on Earth. These data come in the form of published and contributed observations, which often include associated metadata. Even in the best-case scenario of a carefully curated dataset, it may be nontrivial to extract meaningful analyses from such compilations, and choices made with respect to filtering, resampling, and averaging can affect the resulting trends and any interpretation(s) thereof. As a result, a thorough understanding of how to digest, process, and analyze large data compilations is required. Here, we present a generalizable workflow developed using the Sedimentary Geochemistry and Paleoenvironments Project database. We demonstrate the effects of filtering and weighted resampling on Al 2 O 3 and U contents, two representative geochemical components of interest in sedimentary geochemistry (one major and one trace element, respectively). Through our analyses, we highlight several methodological challenges in a "bigger data" approach to Earth science. We suggest that, with slight modifications to our workflow, researchers can confidently use large collections of observations to gain new insights into processes that have shaped Earth's crustal and surface environments. 1 Supplemental Material: table of valid lithologies; map depicting sample locations; crossplot illustrating analytical uncertainty; flowchart of the proposed workflow; histograms showing the effects of progressive filtering, the distribution of spatial and age scales, and proximity and probability values; and results of sensitivity tests.

Pyrite mega-analysis reveals modes of anoxia through geological time

“…Last, from the Mesoproterozoic to the present day, we compared the pyrite-derived fraction of ferruginous versus euxinic conditions to the classic bulk Fe speciation and TOC/P redox proxies recorded in the Sedimentary Geochemistry and Paleoenvironments (SGP) phase 1 database ( Fig. 5 ) ( 24 , 25 ).…”

“…We explored Phanerozoic spatiotemporal redox variation by combination of the xDD text-mining outputs, k NN pyrite–type predictions, and the bulk redox proxies ( Fig. 6 ) ( 24 , 25 ). Spatial analysis shows that the xDD, pyrite trace element, and SGP datasets are biased toward the flanks of Laurentia and Pangaea ( Fig.…”

“…Resampling was implemented with replacement ( n = 1,000,000) and where scale age = 10 Ma and scale spatial = 0.5, the same parameters as adopted by Mehra et al . ( 25 ). Resampling was considered particularly important to account for the spatially clustered nature of the dataset.…”

“…The advantage of considering pyrite morphology as a means to reconstruct basinal redox history is that the presence and nature of pyrite in ancient marine sediments have been extensively documented over many years. Here, we provide a new perspective on marine anoxia in the geological record by deploying machine-reading techniques over a large library of scientific publications, which we integrated with geochemical analyses of coeval shales (19)(20)(21)(22)(23)(24)(25). We searched for phrases related to the type of pyrite observed in sediments, including pyrite framboids, concretions, and nodules.…”

The Sedimentary Geochemistry and Paleoenvironments Project

“…age estimates can also be stored, and indeed, are critical for the type of re-weighted bootstrap analyses employed by many SGP working groups (Mehra et al, 2021).…”