Ice-rafted feldspar grains in recent shelf sediments of West Antarctica: provenance pathways via lead isotope compositions

EGU 2024 oral presentation
Published:

Thomas Arney, Claus-Dieter Hillenbrand, J. Andy Milton, Christine Siddoway, Gavin Foster, Paul Wilson, Julia S. Wellner, and Steven M. Bohaty

DOI: 10.5194/egusphere-egu24-18206

Abstract

Knowledge of the past dynamics of the Antarctic ice sheets is essential for better understanding their present and future stability and, importantly, the resulting effects on global mean sea level rise in a warming world. The petrology and geochemistry of iceberg-rafted debris (IRD; >150 μm) produced by these ice sheets can provide vital information about past ice sheet extent, but the characteristics of geological sources are poorly constrained in ice-covered Antarctica. Lead isotope ratios (controlled by protolith age) in West Antarctic basement and derived sedimentary rocks (mostly <500 Ma) have previously been assumed to be largely homogenous, and therefore useful only for distinguishing West Antarctic IRD from older (0.5-3 Ga) East Antarctic IRD in Southern Ocean sediments. Laser ablation analysis of individual mineral grains, however, avoids the averaging effects of bulk mineral separate, sediment, or rock analyses and reveals the full variation of isotopic signatures in a sample of detrital grains.

Here, we present a survey of lead isotope ratios determined in ~600 iceberg-rafted feldspar grains from 26 seafloor surface sediment samples from the West Antarctic continental shelf. Machine-learning clustering of the lead-isotope data reveals at least two major populations, including a previously unknown population of grains which are less radiogenic than the main cluster. These less radiogenic grains are only found in two areas: near the ice-shelf front of Thwaites Glacier and near the eastern edge of the Ross Ice Shelf. The Thwaites signal is not present at sites on the middle and outer continental shelf, suggesting an offshore dilution effect. Supervised clustering of the main group reveals additional subdivisions in Pb-isotope space that are geographically restricted to: (i) the wider Amundsen Sea, (ii) the Bellingshausen Sea, and (iii) Sulzberger Bay at the boundary between the Amundsen and Ross seas. These subdivisions will be further investigated using 87Rb-87Sr dating of a subset of the feldspar grains used for Pb-isotope analysis.

Together, these new data provide a novel IRD provenance tool, allowing tracing of offshore IRD back to either Thwaites Glacier or the eastern Ross Ice Shelf source areas. Given the observed offshore dilution effect, detection of this signal in sediment cores from the outer shelf or deep-sea would indicate a significantly increased supply of detritus sourced from Thwaites Glacier or the Ross Ice Shelf, both important iceberg outlets of the West Antarctic Ice Sheet.

Take-home messages

Methods

  • >600 ice-rafted K-feldspar sand grains (150–600 μm) from 24 marine sediment core-tops on the West Antarctic continental shelf in the Pacific sector of the Southern Ocean
  • Laser-ablation multicollector ICP-MS
  • 6-component Gaussian mixture model fit to 208Pb/207Pb vs 206Pb/207Pb data

Results

  • Five distinct clusters in 208Pb/207Pb vs 206Pb/207Pb
  • Four clusters are geographically restricted and useful for provenance
  • One cluster is restricted to only two core sites: near the ice-shelf front of Thwaites Glacier and near the eastern edge of the Ross Ice Shelf.

Conclusions

  • Pb isotopes in feldspars can be used to distinguish different West Antarctic IRD sources.
  • New proxy for Thwaites Glacier dynamics, based on an anomalous less radiogenic cluster of lead isotopes in ice-rafted K-feldspar sand grains.

Future work

  • Confirm the Pb-Pb clusters and attempt to link to sources with other geochemical systems:
    • date the same grains used for the Pb analyses with the new in-situ Rb-Sr mass-shift laser ablation method
    • analyse elemental chemistry (including REE)
  • Investigate the glacial-interglacial variability of these clusters downcore in a slope record which covers the Plio-Pleistocene.

If you have any comments or questions, please don’t hesitate to get in touch!