Ocean‐Driven and Topography‐Controlled Nonlinear Glacier Retreat During the Holocene: Southwestern Ross Sea, Antarctica

Abstract

Recent ice sheet mass loss in Antarctica has been attributed to an influx of warm ocean waters, which drove grounding‐line retreat and ice thinning. Episodic retreat and rapid thinning also occurred in the southwestern Ross Sea during the Holocene, which today accommodates cold ocean waters. We applied finite element ice‐flow modeling to investigate the roles of ocean temperature and bed topography in the deglaciation of this region. First, our experiments demonstrate that bed topography controlled the spatial pattern of grounding‐line retreat. Topographic pinning points limited the rate of ice loss until retreat progressed beyond a bathymetric threshold. Second, ocean thermal forcing determined the timing of this ice loss. Enhanced ocean‐driven melt is required during the Early‐to‐Mid Holocene to replicate geological records of deglaciation, possibly indicating that warm ocean waters were once present in this region. On multi‐centennial timescales, ocean temperature drove, while bed topography controlled, nonlinear rates of ice mass loss.

Holocene thinning of Darwin and Hatherton glaciers, Antarctica, and implications for grounding-line retreat in the Ross Sea

“…The Hatherton drift is a minimally weathered deposit, commonly underlain by relict ice and found within 50-100 m elevation of the modern margin of Hatherton Glacier. While Bockheim et al (1989), Storey et al (2010), Joy et al (2014, and King et al (2020) agree on the general characteristics of Hatherton drift, their interpretations of its significance dif-fer. Bockheim et al (1989) posited that Hatherton drift represents the last episode of thinning from the highstand represented by the Britannia deposits.…”

“…During the last glaciation, other outlet glaciers throughout the TAM thickened initially at their mouths in response to thickening grounded ice in the Ross Sea, and later upstream due to increased accumulation after the Antarctic LGM (Todd et al, 2010). Lateral moraines formed during this period have been used to interpret ice thicknesses of both the grounded ice sheet in the Ross Sea and the EAIS outlet glaciers (Bockheim et al, 1989;Bromley et al, 2010Bromley et al, , 2012Denton et al, 1989;Hall et al, 2015;Jones et al, 2015;Joy et al, 2014;King et al, 2020;Spector et al, 2017;. Bockheim et al (1989) mapped lateral moraines and drift sheets in ice-free valleys alongside Hatherton Glacier and dated these deposits based on weathering, soil characteristics, and radiocarbon ages of freeze-dried algae.…”

“…The chronology and correlation of these deposits along the Hatherton-to-Darwin glacier flow line remains uncertain, and the Britannia II, Britannia I and Hatherton drifts have each been identified with the last maximum glaciation in this sector of the TAM (Bockheim et al, 1989;Storey et al, 2010;Joy et al, 2014;King et al, 2020). In addition to determining which deposits represent the last local maximum, projection of the associated ice surface downstream from sites on Hatherton Glacier to the surface of the grounded ice in the Ross Sea has been complicated by the absence of comparable deposits on Diamond Hill at the Darwin Glacier mouth.…”

“…Although Bockheim et al (1989) tentatively attributed Britannia II drift to the LGM and Britannia I deposits to a Holocene stillstand or readvance, Joy et al (2014) have since dated Britannia II drift to the penultimate glacial maximum. Younger and fresher Britannia I drift occurs up to ∼ 350 m above the modern margin of Hatherton Glacier in Dubris and Bibra valleys, in Magnis Valley, and at Lake Wellman (Bockheim et al, 1989;Storey et al, 2010;Joy et al, 2014;King et al, 2020). At these sites, the deposit is a boulder-and cobble-rich ablation till, with clasts showing varying degrees of weathering.…”

“…The inference by Storey et al (2010) that Hatherton rather than Britannia I drift marks the LGM highstand of Hatherton Glacier led these authors to propose no more than ∼ 50 m of LGM thickening. Joy et al (2014) showed that the Britannia I drift in fact represents the LGM highstand, with several hundred meters of LGM thickening relative to present. They showed that Britannia II drift is a deposit of the penultimate glaciation and obtained Holocene ages from recessional positions in Britannia I drift.…”