Jesse et al. (2025) Sub-shelf melt pattern and ice sheet mass loss governed by meltwater flow below ice shelves

Identification

• Journal: ˜The œcryosphere
• Year: 2025
• Date: 2025-09-16
• Authors: Franka Jesse, Erwin Lambert, R. S. W. van de Wal
• DOI: 10.5194/tc-19-3849-2025

Research Groups

• Institute for Marine and Atmospheric Research Utrecht, Utrecht University, the Netherlands
• R&D Weather and Climate Models, Royal Netherlands Meteorological Institute (KNMI), the Netherlands
• Department of Physical Geography, Utrecht University, the Netherlands

Short Summary

This study presents a novel coupled ice sheet–sub-shelf melt model (IMAU-ICE/LADDIE) to resolve 2D horizontal meltwater flow and compares its impact on Antarctic ice sheet evolution against traditional melt parameterizations. The findings reveal that resolving 2D meltwater flow introduces critical feedbacks, particularly along shear margins, leading to distinct transient volume loss and sensitivity to ocean warming not captured by simpler parameterizations.

Objective

• To examine the effects of 2D horizontal meltwater flow on an idealised ice-sheet–shelf system by coupling the ice sheet model IMAU-ICE with the sub-shelf melt model LADDIE.
• To analyze the feedbacks between ice shelf geometry and melt patterns, and their impact on transient volume loss and grounding line retreat, comparing these results with those obtained using three widely adopted sub-shelf melt parameterizations (Quadratic, PICO, Plume).

Study Configuration

• Spatial Scale: Idealised ice-sheet–shelf system (MISMIP+ geometry) with a horizontal resolution of 2 km, spanning the horizontal extent of the ice shelf.
• Temporal Scale: Simulations run for 1000 years after a 50 000-year spin-up to a steady state.

Methodology and Data

• Models used:
  • IMAU-ICE v2.0 (vertically integrated ice sheet model)
  • LADDIE v1.0 (one-layer Antarctic model for dynamical downscaling of ice–ocean exchanges, resolving 2D horizontal meltwater flow)
  • Sub-shelf melt parameterizations for comparison: Quadratic, PICO (Potsdam Ice-shelf Cavity mOdel), and Plume.
• Data sources:
  • Idealised MISMIP+ protocol geometry.
  • Prescribed, vertically varying, and horizontally uniform ambient ocean temperature and salinity profiles (hyperbolic tangent profiles resembling observations around Dronning Maud Land and the Amundsen Sea), constant over time.

Main Results

• The coupled IMAU-ICE/LADDIE model reveals a positive feedback mechanism: initial ice draft steepening near the grounding line strengthens westward meltwater flow, enhancing the western boundary current, leading to increased melt and progressive weakening of the western shear margin.
• This margin thinning causes dynamical detachment of grounded ice from floating ice, reducing buttressing and accelerating ice flow on centennial timescales.
• On millennial timescales, reduced thermal forcing near the deep grounding line weakens the meltwater flow, suppressing the western boundary channel. This allows the western margin ice shelf to thicken and reattach, reducing ice velocities and volume loss.
• Traditional parameterizations (Quadratic, Plume) fail to capture this positive feedback and margin weakening due to their simplified or absent representation of 2D meltwater flow.
• The PICO parameterization, assuming meltwater flow origin along the entire grounding line, shows persistent high melt rates along the margins, preventing later thickening and recovery.
• LADDIE experiments show distinct transient volume loss and grounding line retreat patterns compared to parameterizations, with different timing and magnitude of peak volume loss.
• LADDIE exhibits lower sensitivity to increased ocean temperatures compared to the parameterizations on timescales beyond 300 years, partly due to negative feedbacks from localized melt-through events disrupting the western boundary current.
• Over 1000 years in the moderate-melt scenario, LADDIE predicts the largest total volume above flotation (VAF) loss, double that of the Plume experiment and approximately 25% more than Quadratic and PICO.
• In the high-melt scenario, PICO predicts the largest VAF loss (over 10% greater than LADDIE) due to widespread and persistent melt-through of ice shelf margins.

Contributions

• Introduces a novel online coupling between the ice sheet model IMAU-ICE and the 2D sub-shelf melt model LADDIE, enabling a more physically advanced representation of sub-shelf melt patterns in ice sheet simulations.
• Demonstrates the critical role of 2D horizontal meltwater flow dynamics (including Coriolis deflection and topographic steering) in governing sub-shelf melt patterns, particularly along shear margins, and the resulting transient ice sheet response.
• Highlights significant limitations of widely adopted sub-shelf melt parameterizations in capturing key melt–geometry feedbacks, such as margin thinning/thickening and associated changes in buttressing, leading to divergent projections of ice sheet evolution.
• Provides process-based insights into the mechanisms driving transient volume loss and grounding line retreat, suggesting that melt along shear margins, influenced by 2D meltwater flow, is a more significant factor than melt near the grounding line alone.

Funding

• Utrecht University (Franka Jesse)
• Netherlands Organisation for Scientific Research (NWO) project HiRISE (grant no. OCENW.GROOT.2019.091) (Erwin Lambert)

Citation

@article{Jesse2025Subshelf,
  author = {Jesse, Franka and Lambert, Erwin and Wal, R. S. W. van de},
  title = {Sub-shelf melt pattern and ice sheet mass loss governed by meltwater flow below ice shelves},
  journal = {˜The œcryosphere},
  year = {2025},
  doi = {10.5194/tc-19-3849-2025},
  url = {https://doi.org/10.5194/tc-19-3849-2025}
}