Spatial variation in Late Ordovician glacioeustatic sea-level change

Abstract

Mass extinction of Late Ordovician marine fauna closely coincided with southern hemisphere glaciation. The sequence stratigraphic architecture of shallow marine deposits informs estimates of glacioeustatic sea-level change at sites both proximal and distal to the reconstructed Ordovician ice sheet(s) and contemporaneous changes in ice volume. A recent correlation framework for the stratigraphic architectures of one near and one far field Late Ordovician margin concluded that the Late Ordovician glaciation encompassed multiple long-term cycles of ice volume growth and retreat with superimposed higher frequency cycles. Here we posit that—similar to Cenozoic glacial cycles—glacial isostatic adjustment can preclude synchronous and similar magnitude (or directional) changes in Late Ordovician sea level between ice proximal and ice distal locations and, hence, distort a globally correlative sequence stratigraphy. We explored whether long-duration (i.e., million year) Late Ordovician glacial cycles should produce a globally coherent, eustatic record of sea-level change between ice proximal and ice distal margins using a gravitationally self-consistent theory that accounts for the deformational, gravitational and rotational perturbations to sea level on a viscoelastic Earth model. We adopted a Late Ordovician paleogeography and a synthetic continental ice-sheet distribution and volume informed by the areal extent of glaciogenic deposits and geochemical records, respectively. We demonstrate that modeled million year Late Ordovician glacial cycles produce sea-level histories on near and far field margins that differ from eustasy, and from one another, due primarily to elastic flexure and associated gravitational effects. While predicted far-field sea-level histories faithfully preserve the temporal structure of modeled glacioeustasy, their amplitude may differ from eustasy by as much as 30–40%. The impact of glacial isostatic adjustment is largest at the margins of glaciated continents, and these effects can be of the same order of magnitude as the eustatic, and even induce a local sea-level rise during an episode of ice growth and eustatic sea-level fall, and vice versa. In this regard, stratal surfaces of maximum regression and flooding expressed at near-field margins need not reflect global (‘eustatic’) trends in ice sheet growth and decay, respectively, and thus may not provide chronostratigraphic horizons for correlation with far-field sequence stratigraphic architectures.