The increasing ice loss from Antarctica since the early 21st century, primarily from the Amundsen Sea Embayment (ASE) of the West Antarctic Ice Sheet (WAIS), is a significant contributor to global sea-level rise. This ice mass loss is attributed to two main processes: ocean-driven melting of ice shelves due to upwelling of warm Circumpolar Deep Water (CDW), and atmospheric-driven surface melting of glaciers along the coast. Understanding the relative contributions of these processes is crucial for accurate sea-level rise predictions. Previous research has highlighted the role of warm CDW upwelling in ASE ice mass loss. Additionally, increased precipitation and warming over West Antarctica, linked to higher tropical Pacific Ocean temperatures, have been observed. Atmospheric rivers, transporting significant water vapor, contribute to summer surface melt. However, significant uncertainty remains regarding the long-term stability of the ice sheet, future sea-level contributions, and the dominant driving mechanisms. To address these uncertainties, long-term records of past ice-sheet change are needed to validate current ice-sheet models. While the Amundsen sector of the WAIS reached its modern limits by the Early Holocene, the mechanisms driving subsequent early-to-mid-Holocene thinning remain unclear. This study utilizes beryllium (Be) isotopes as a proxy for glacial processes to reconstruct the history of the Cosgrove Ice Shelf (CIS) in the eastern ASE over the last 10.3 kyr, aiming to improve our understanding of past and future ice sheet behavior.

Existing research demonstrates that the Amundsen Sea sector of the West Antarctic Ice Sheet (WAIS) has experienced significant ice mass loss since the early 21st century, primarily driven by ocean-driven melting of ice shelves through the upwelling of warm Circumpolar Deep Water (CDW). Studies have also linked increased precipitation and warming over West Antarctica to higher tropical Pacific Ocean temperatures. Atmospheric rivers play a significant role in surface melt. However, the relative importance of oceanic and atmospheric processes in driving ice mass loss remains unclear, emphasizing the need for long-term records of past ice-sheet change to validate existing ice-sheet models and improve predictions of future sea-level rise. Earlier studies have shown that the Amundsen Sea reached its modern limits in the early Holocene, yet the mechanisms behind subsequent ice stream thinning during the early-to-mid-Holocene remain poorly understood.

This study utilizes beryllium (Be) isotopes as a proxy to reconstruct the history of the Cosgrove Ice Shelf (CIS) in the eastern Amundsen Sea Embayment (ASE) over the last 10.3 kyr. Three Kasten cores (KC-15, KC-16, KC-17) were collected from Ferrero Bay during the IB Oden OSO-0910 expedition. The reactive 10Be and 9Be abundances and 10Be/9Be ratios were measured in marine sediments. The analysis included reactive ⁹Be abundance ([⁹Be]reactive), ¹⁰Be abundance ([¹⁰Be]reactive), and ¹⁰Be/⁹Be ratios. These Be isotope records were supported by previously published multi-proxy analyses of total organic carbon (TOC), total nitrogen (TN), microfossil abundance and assemblages (diatoms and foraminifera), and magnetic susceptibility (MS). A new age-depth model was created for KC-15 using radiocarbon dating and Bayesian analysis, incorporating analytical and depth uncertainties. The researchers correlated reactive ¹⁰Be and ¹⁰Be/⁹Be ratios with TOC, TN, and MS to interpret glacial history. The researchers also considered factors influencing Be isotope systematics in glacial environments, such as the disparate sources of ¹⁰Be (cosmogenic production) and ⁹Be (chemical weathering). The study distinguished between Be sources from coastal regions (low ¹⁰Be/⁹Be ratios) and open marine environments (high ¹⁰Be/⁹Be ratios), considering factors like meltwater input from icebergs, basal and/or surface melting of the CIS, and scavenging efficiency.

Analysis of beryllium isotope data from the collected sediment cores revealed three key periods in the Holocene glacial history of Ferrero Bay and the CIS: (1) Early Holocene (10.3–9.8 kyr BP): low ¹⁰Be/⁹Be ratios indicated proximity to the grounding line; (2) Early-to-Mid Holocene transition (9.8–5.9 kyr BP): an increase in ¹⁰Be/⁹Be ratios suggested grounding line retreat and the development of a more distal environmental setting. This period saw significant increases in [¹⁰Be]reactive and ¹⁰Be/⁹Be, alongside TOC and TN, and a decrease in magnetic susceptibility, reflecting a transition to fine-grained sediments; (3) Late Holocene (5.9 kyr BP–present): relatively constant ¹⁰Be/⁹Be ratios and lithological changes indicated a stable ice shelf in Ferrero Bay, followed by a collapse and establishment of open marine conditions. The strong correlation between [¹⁰Be]reactive and ¹⁰Be/⁹Be ratios in all three cores supports the interpretation that the rise in [¹⁰Be]reactive is the main factor driving variations in the ¹⁰Be/⁹Be ratios. The similar trend of reactive Be with TOC and TN further emphasizes the influence of this relationship. The negative correlation between magnetic susceptibility and Be isotopes indicates a lithological shift from coarse-grained to fine-grained sediments. The researchers determined that the increase in ¹⁰Be/⁹Be ratios from 9.8 kyr BP to 5.9 kyr BP is more likely associated with meltwater originating from subglacial sources rather than ocean currents. The study suggests that the early-to-mid-Holocene retreat of the CIS was not primarily driven by the upwelling of CDW, although some influence may be attributed to CDW incursion along the west Antarctic Peninsula. The findings are consistent with the widespread melting event observed in other parts of the Amundsen Sea sector between 9 and 6 kyr BP.

The study's findings indicate that the Holocene melting and retreat of the Cosgrove Ice Shelf (CIS) in the Amundsen Sea Embayment (ASE) was predominantly driven by atmospheric warming and increased precipitation, rather than solely by ocean-driven processes. The observed increase in ¹⁰Be/⁹Be ratios correlates with atmospheric changes associated with tropical Pacific warming, specifically the generation of atmospheric Rossby waves that influenced atmospheric circulation over West Antarctica. This led to warmer air advection, increased precipitation, and enhanced surface melting. Increased accumulation rates and intensified atmospheric rivers also contributed to surface melting and potential ice shelf instability. While CDW upwelling played a role, the dominant driver seems to have been atmospheric forcing. The results are consistent with previous findings indicating widespread thinning of other glaciers in the ASE during this period. The study highlights the significant role of atmospheric processes in driving West Antarctic ice sheet change, impacting future sea-level rise predictions. The findings show the successful use of Be isotopes as a valuable proxy for reconstructing past glacial dynamics and meltwater input.

This study successfully used beryllium isotopes and other proxies to reconstruct the Holocene history of the Cosgrove Ice Shelf, revealing that a major melting event between 9 and 6 kyr BP was primarily driven by atmospheric warming and increased precipitation linked to tropical Pacific warming. This emphasizes the critical role of atmospheric processes in shaping West Antarctic ice sheet dynamics. The findings are significant for validating ice-sheet models and refining sea-level projections, highlighting the need to consider both atmospheric and oceanic forcing mechanisms in future predictions. Future research should focus on further constraining Be abundances in sea ice, ice sheets, and shelf waters to better understand the intricacies of the Be isotope system in glacimarine sediments.

While the study provides valuable insights into past ice sheet dynamics, there are some limitations. The interpretation of Be isotopes relies on several assumptions about sources and processes influencing Be isotopes in glacimarine sediments. Uncertainty remains regarding the relative contributions of different meltwater sources (e.g., iceberg calving vs. basal melting). The study's focus on the Cosgrove Ice Shelf may not be fully representative of the entire ASE. Higher temporal resolution data would be beneficial for understanding millennial-scale variability during the Late Holocene.