Spectral analysis of the lower Eocene Wilkins Peak Member, Green River Formation, Wyoming: Support for Milankovitch cyclicity

Abstract

This study is the first to employ spectral analysis to examine meter-scale sedimentary cyclicity in the Wilkins Peak Member of the lower Eocene Green River Formation of Wyoming. Generally regarded as the classic example for orbital forcing of lacustrine sediments at eccentricity and precession time scales, this long-standing interpretation was recently contested, with a much shorter duration (≤10 ky) inferred for the dominant cyclicity. Earlier work lacked adequate age control or spectral analysis or both. Our analysis is based upon an evaluation in the frequency domain of oil-yield values from four boreholes, accuracy estimation for suggested orbital interpretations, and comparison to independent geochronology. Cored intervals 266–364 m thick represent a span of 1.2–1.7 m.y., with temporal resolution of ∼3–5 ky (∼1 m) for oil-yield values. Variations in spectral power with depth within the original records are interpreted to reflect changes in the rate of sediment accumulation. These changes are corrected prior to testing the orbital forcing hypothesis by using two methods: 1) a minimal adjustment (three segments) accounting for the dominant changes of spectral frequency with depth; and 2) correlating the published definitions of precessional cycles in these records to a 21 ky cosine curve. Orbital age models resulting from the two tuning methods are compared to available chronology and the tuned records are tested for the expected spectral peaks from orbitally forced records. We conclude that the dominant cyclicity of the Wilkins Peak Member is orbitally forced. Orbital age models overlap 40Ar/39Ar ages and inferred periods include long and short eccentricity, weak obliquity and precession. Eccentricity is resolved in the analyzed records but the expected ∼95 and ∼125 ky periods are not resolved, controlling the range of possible tuning periods and the accuracy of orbital age models. Sub-Milankovitch variability exists and can be resolved to a minimum period of ∼3–5 ky by the analyzed records. However, it cannot be characterized fully with the available chronology or by the previously calculated mean cycle duration.