This website uses features that your browser doesn't support. Please upgrade to a recent version of your browser.

Gatekeepers of Antarctica















Ice shelves, which are floating masses of ice that encircle approximately 75% of Antarctica's coastline, act as gatekeepers and play a crucial role in global sea level rise. They serve as barriers that prevent grounded ice from flowing into the ocean, thereby slowing the rate of sea level rise.

Ice shelf collapse



Breaking up of ice shelves does not directly cause sea level rise since they are already floating in the ocean. However, these ice shelves act as a barrier that holds back land ice, and when they break, it causes the land ice to flow into the ocean at a faster rate, leading to an accelerated rise in sea level.

Illustration of ice speed up due to ice shelf break up.



An ice shelf can disintegrate when melt ponds form on its surface and grow, percolating through surface fractures and causing its collapse.

A prominent example of this process is the collapse of the Larsen B Ice Shelf on the Antarctic Peninsula in the austral summer of 2002, where the formation of meltwater ponds due to a warm summer led to its sudden disintegration.



Thresholds for meltwater ponding

The collapse of Larsen B has led to the suggestion that a -5° C temperature limit exists for ice shelf viability as higher temperatures lead to increased melt ponding that can result in desintegration. However, it is currently uncertain if this -5° C threshold applies to other Antarctic ice shelves.

Therefore, we use another approach based on the fact that melt ponding occurs when the ratio of melt over accumulation (MoA) exceeds 0.7, meaning that melting is too extensive or snow accumulation is too limited to prevent the formation of ponds. This threshold has been found to accurately predict the viability limit for ice shelves in the eastern Antarctic Peninsula, such as Larsen B, for temperatures that are typically below -5° C.

Illustration of temperature thresholds and snowmelt, accumulation and MOA showing that Larsen B had a MOA above 0.7

Modelling meltwater ponding



Subsequently we use two present-day and three future high-resolution Antarctic climate simulations to predict warming thresholds for Antarctic ice shelf melt pond formation based on the melt over accumulation (MOA) ratio.

Illustration of the temperature thresholds for meltwater ponding across Antarctica

The predicted warming thresholds align well with observed melt pond volumes in recent Sentinel-2 satellite imagery: the regions where the MOA threshold is already exceeded, as shown in yellow, match the areas where high melt pond volumes were observed by satellite, as shown in red.

Furthermore, it is found that the warming thresholds for melt pond formation are highly variable across the spatial domain and are primarily determined by the amount of snow accumulation.

Our research confirms that the -5° C temperature threshold applies to relatively wet ice shelves, however, cold and dry ice shelves such as Amery, Ross and Filchner-Ronne are found to be more susceptible to melt pond formation than previously believed, with threshold temperatures significantly lower than -15° C.

MOA threshold temperature as a function of accumulation and melt rate showing that many ice shelves reach ponding before -5° C



Future ponding in a warming climate

Illustration of the remaining warming (ΔT) needed to reach MOA=0.7

The latest climate models project that towards the end of this century these thresholds can be reached on many ice shelves, even on cold ice shelves and under moderate warming scenarios.

This implies that even with average global warming those temperatures could be reached on many ice shels, including the Ross Ice Shelf, by the end of this century.

Melt pond formation under CMIP6 future warming (different colors). Ice shelf average threshold temperature (TT) to reach MoA=0.7 (solid lines) and spatial variability (dashed lines)