Comparing melt models to reduce sea-level rise uncertainties

Research update on Antarctic ice shelf melt

Sea-level rise poses a growing threat to low-lying delta regions. Coastal communities are therefore challenged with designing adaptation strategies based on highly uncertain projections. The major source of uncertainty stems from our limited physical understanding of the Antarctic ice sheet, and in particular the response of this ice sheet to a warming ocean. In this project, funded by the Water, Climate, and Future Deltas hub, we compare different existing models that quantify ice shelf melt from ocean temperatures.

Approximately seventy percent of the Antarctic ice sheet is surrounded by ice shelves. These floating extensions reduce the ice flow and hence play an important role in keeping the ice where it is. Yet the ice shelves rest on an increasingly warming ocean, which induces melt, an acceleration of the ice flow, and ultimately sea-level rise at the global scale. How much the Antarctic ice sheet will contribute to future sea-level rise depends critically on the amount of ocean-induced ice shelf melt. However, climate models, ocean models and ice sheet models all cannot resolve the fine-scale processes governing ice shelf melt. Rather, this is estimated by simple equations of highly idealized models.

Graphs with lines that display the global mean sea level in time
The IPCC’s latest global mean sea-level projections. This global mean projection is thought to be representative of future sea-level rise at low-lying deltas worldwide (Oppenheimer et al, 2020).
Graph with 4 lines that represent melt rate for different models
Preliminary results for an idealised ice shelf for four different melt models and two linear equations.

In this study, we compare the ice shelf melt-sensitivity of different models to ocean warming. Our preliminary results indicate that, at first sight, models are generally consistent and a warming ocean leads to increased melt. However, melt estimates vary by nearly an order of magnitude under the same forcing. In addition, models disagree on the linearity between melt and ocean temperatures, with some models capturing a positive feedback which can cause an acceleration in ice shelf melt and ultimately Antarctic mass loss.

Example experiment with the ‘Sheet’ model, illustrating the complex flow and melt patterns below the Pine Island ice shelf.

Besides these ice-shelf average melt rates, spatial differences in melt are crucial for the stability of the ice shelves. We find that the two-dimensional flow of melt water below the ice shelf, which is only simulated by the ‘Sheet’ model, strongly constrains the spatial melt fields. This model is therefore considered most promising to realistically estimate melt rates.

In the remainder of the project, realistic experiments will be done on the existing ice shelves of the Antarctic ice sheet. This project will be continued within a larger project to apply the sheet model to realistic ocean simulations, in order to provide ice sheet models with realistic forcing to project future sea-level rise.