Prior to the present study, no dedicated in situ measurement of the thickness distributions of fast ice and the sub-ice platelet layer, formed by supercooled Ice Shelf Water in the near-surface ocean, had been carried out in Terra Nova Bay. Previous studies have recognised the biological importance of the sub-ice platelet layer observed beneath fast ice beside the Campbell Ice Tongue. Furthermore, a recent airborne survey of fast ice in the western Ross Sea implied that smaller ice bodies (ice tongues and outlet glaciers) contribute to the formation of supercooled Ice Shelf Water. With the objective of inferring source regions and circulation of Ice Shelf Water, we measured fast ice and sub-ice platelet layer thickness distributions near the Campbell Ice Tongue in late spring of 2021, using drill hole surveys and high-resolution ground-based electromagnetic induction soundings. We observed thicker fast ice and sub-ice platelet layer near the ice tongue with very thick and narrow sub-ice platelet layer maxima identifying highly channelled outflow of supercooled Ice Shelf Water from beneath the ice tongue through ice mélange, subglacial formations, and grounded regions. We conclude that a significant volume of supercooled Ice Shelf Water is locally sourced from the Campbell Ice Tongue through basal melting and affirm that the icescape in north Terra Nova Bay results from a complex interplay of glacial morphology, bathymetry, polynya dynamics, and ocean circulation.

We present a 700 km airborne electromagnetic survey of late-spring fast ice and sub-ice platelet layer (SIPL) thickness distributions, from McMurdo Sound to Cape Adare, providing a first-time inventory of thickness close to its annual maximum. The overall modal consolidated ice (including snow) thickness was 1.9 m, less than its mean of 2.6±1.0 m. Our survey was partitioned into level and rough ice, and SIPL thickness was estimated under level ice. Although results show a prevalence of level ice, with a mode of 2.0 m and mean of 2.0±0.6 m, rough ice covered 41% of the transect by length, 50% by volume, with a mode of 3.3 m and mean of 3.2±1.2 m. The thickest 10% of rough ice was almost 6 m on average, and a 2 km segment in Moubray Bay had a thickness greater than 8 m, demonstrating the overwhelming influence of deformation against coastal features. The fast ice was thus significantly thicker than adjacent pack ice. The presence of a significant SIPL was observed in Silverfish Bay, offshore Hells Gate Ice Shelf, New Harbour, and Granite Harbour where the SIPL transect volume was a significant fraction (0.30) of the consolidated ice volume. The thickest 10% of SIPLs had an average thickness of nearly 3 m, and near Hells Gate Ice Shelf the SIPL was almost 10 m thick, implying vigorous heat loss to the ocean (~ 90Wm-2). We conclude that polynya-induced deformation and interaction with continental ice influence fast ice thickness in the western Ross Sea.