Iceberg Melt Rates Reveal Spatial Variations in Ocean Thermal Forcing Around Antarctica

Subsurface ocean warming has been implicated as the primary trigger for ice loss in many regions throughout Antarctica. While submarine melt rates beneath ice shelves can be estimated using changes in ice shelf surface elevation over time, submarine melting of the Antarctic Peninsula's grounded termini cannot be estimated using satellite observations. Coastal ocean observations are also incredibly sparse in space and time in most regions, limiting our ability to investigate feedbacks between changes in ocean properties and submarine melting.

Here we present an analysis of near-terminus remotely-sensed iceberg melt rates from 2011-2022 with respect to observed and modeled ocean water properties. We demonstrate that spatial patterns in iceberg submarine melt rates from 15 study sites - including sites near eight glacier termini spanning the Antarctic Peninsula, two sites in West Antarctica, and five sites in East Antarctica - generally agree with variations in regional subsurface ocean temperatures. However, there is no coherent relationship between melt rate and ocean thermal forcing across all sites likely due to sparse observational coverage of the iceberg and oceanographic datasets. The strongest relationship between melt rate and draft is at Thwaites Glacier: iceberg melt rates are nearly 50 m/yr for the near-terminus icebergs that penetrate ~600 m into the water column, dropping to ~10 m/yr as the icebergs move seaward and thin by ~400 m. We compared our observation-based melting estimates to those from Holland and Jenkins (1999) and Weeks and Campbell (1973) parameterizations constrained by SOSE model outputs, and found that neither parameterization captures the observed spatio-temporal variations in iceberg melting around Antarctica. We conclude that iceberg melt rates can be used as a tool to assess spatial variations in ocean thermal forcing as well as biases in ocean model outputs near glacier margins.