BAYESIAN INFERENCE OF DEPOSITIONAL STATISTICS FROM ANDRILL AND-2A, A MIOCENE ANTARCTIC GLACIOMARINE SEDIMENT CORE

Recent work in quantitative stratigraphy has advanced theory about fundamental length and time scales linking depositional process to stratigraphic completeness. In particular, the so-called compensation length (CLS) and time scales (CTS) of sedimentation have been applied to experimental and geologic fluvial settings, where they are linked to channel depth and avulsion rate, respectively. I apply these concepts, which are relevant for any depositional setting, to orbitally forced grounding-line proximal glaciomarine sedimentation. I hypothesize that the vertical amplitude of grounding line motion, which is the locus of glaciomarine sedimentation, sets the CLS. For orbitally forced ice sheets, this quantity should be on the order of dozens to hundreds of meters.

Using a novel Bayesian age modelling approach, I explicitly infer posterior distributions of hiatus duration at bedding contacts in the core. From hiatus durations, I probabilistically constrain the 95% CI of the CTS for AND-2A to 170–460 ky. With the CTS and the long-term core accumulation rate, I constrain the 95% CI of the CLS to 30–81 m, which is consistent with independent estimates of early Miocene glacioeustasy. This work provides a probabilistic framework for estimating depositional statistics from a single stratigraphic section and proposes a model for the fundamental length and time scales of orbitally forced glaciomarine sedimentation. This model makes quantitative predictions for the CLS in deeper time glaciomarine stratigraphies, where the role of orbital forcing is often debated.