How to apply ice shelf fracture mask (path C)

[talbrecht]

When using the prescribed hydrofracture mask (Path C in https://www.ismip.org/participants/focus-groups/ais-ice-shelf-collapse-calving), I wonder how and where it should be best (in a physically meaningful way) applied in the ice sheet model, as instructions say: "impose ice shelf removal where criteria met.":

• I guess it should not apply in grounded ice sheet regions, only to ice shelves?
• Hydrofracturing according to the mask starts in the Ross and Filchner Ice Shelves far from the margin near the grounding line, "where downslope winds likely enhance melt"". I would assume that hydrofracturing-induced Larsen-B style collapse would happen from the ice shelf margin and not somewhere near the grounding line (where maybe it could produce some mélange, but not a hole in the ice shelf)?

-> The decision if the mask applies to floating ice shelves or only at the ice shelf margins explains a doubling in ice sheet loss until 2300 in some of my sensitivity simulations.

Maybe the Fracture/Calving focus group has a suggestion?

[hseroussi]

• This collapse is indeed for ice shelves only, so it should be applied only on the floating ice (which might vary depending on your model).
• This patterns is rather similar to what we observed in ISMIP6 Antarctica 2300 projections, with conditions for ice shelf collapse starting form the grounding line and not from the ice front. It caused some very large tabular event in these simulations. There is some flexibility for the models on how to best apply the collapse in these case. One possibility is to combine the collapse conditions with a criteria on ice shelf stress conditions, so that ice shelf collapse when both the surface conditions and the ice shelf stress are combined to link to the collapse, following for example recommendations in Lai et al. (2020).