Where did the ice reach the sea? The utility of coupled K-feldspar Rb-Sr, Ar-Ar, and Pb-isotope analysis applied to mid-Miocene ice-rafted debris in Antarctic marine sediment

The middle Miocene climate optimum (c. 14.2 to 13.8 Ma), a significant warm period, was followed by a series of step-wise global cooling and Antarctic ice-sheet expansion events visible in marine isotope records (e.g., Holbourn et al., 2013), the oldest of which is termed the mid-Miocene climate transition. Associated episodes of ice-sheet instability and iceberg calving are recorded by ice-rafted debris in mid- to high-latitude marine sediment, accessible via deep-sea sediment cores around the Antarctic margin. Paleo-ice sheet models indicate that step-wise ice-sheet growth in part reflects ice expansion across previously ice-free low-elevation regions (Gasson et al., 2016; Halberstadt et al., 2021). Such predictions are amenable to testing by detrital provenance analysis of ice-rafted debris. However, the small-volume and mineralogically impoverished samples which are typically recovered from distal marine sediment preclude use of conventional accessory heavy-mineral proxies: instead, use of a rock-forming mineral is necessitated.

Here, we present in-situ Rb-Sr, Ar-Ar, and Pb-isotope data from ice-rafted K-feldspar collected from mid-Miocene marine sediment in the Weddell Sea (Neofitu et al., 2024) and offshore Prydz Bay. Source regions for these depocenters respectively include the Recovery and Aurora sub-glacial basins, where ice-sheet embayment formation during warm periods is predicted. Our data suggest that the Wilkes and Aurora subglacial basins were free of marine-terminating ice during the middle Miocene climate optimum. During the transition, ice advanced to the coast across the Aurora sub-glacial basin, and both the Recovery and Aurora basins at least intermittently hosted marine-terminating ice during the subsequent cooling step.