Better Mask

[bradlipovsky]

We currently use the BedMachine Antarctica mask. Does this create systematic biases? But we have altimetry data! Shouldn't we just use the data itself to know where the ice shelf is?

[ram21ARC]

We discussed this with reference to the grounding line, but could equally apply for rifts at the calving front of advancing ice shelves (for example, a seaward rift wall could advance past a defined calving front, and we wouldn't be able to tell whether it had calved or not). Maybe not so relevant on the short timescale

[ram21ARC]

We could buffer (eg 5km/a function of flow speed (relevant to frontal advance and GL retreat)) around the ice shelf mask, so long as the detector is robust, we won't find any rifts on the grounded ice, and we won't find any rifts beyond the calving front because we won't find a second wall.