Ice-sheet mass-balance redistribution is non-trivial. Even if the ice-sheet is assumed to have 0 net mass-balance change, regional discharge will vary. Ie part of the ice-sheet can be growing, another part shrinking resulting in spatially heterogenous discharge to the oceans. Combined drainage/ice-sheet/isostasy modelling can resolve this and create a self-consistent routing map.

A related issue is iceberg versus meltwater discharge. Glaciological models approximately resolve this (we lack well constrained calving models, so emphasis on the “approximate”). Is it worth considering different treatment of freshwater versus ice-berg discharge wrt when mixing can occur in the ocean?

Topography and coastlines need to be a part of the self-consistent ice-sheet/topography package.

A vegetation map for 21ka reconstructed from a combination of model output and data analyses would be very useful for modeling groups that will not be using interactive vegetation in their PI and 20th century runs. We could query groups on the best format. For CCSM, a pft vegetation map would be best.

We would like to add in this Table (21 ka boundary conditions) the new line “PERMAFROST DISTRIBUTION”. Apart from, we can discuss the scale of Arctic ocean regression during the 21 ka BP as at that time in Yamal, Gydan and Timyr Peninsulas marine terraces were formed. It means that ocean level was high as today. We have the detailed maps of stable isotope distribution in permafrost at this time so we have good instrument for winter temperature reconstruction.

Professors Alexander Kislov and Yurij Vasil’chuk (Moscow State University)

As example, we present two maps, containing the mean January temperature reconstructed based on the signal of the δ18O in ice wedges Professors Alexander Kislov and Yurij Vasil’chuk (Moscow State University)

(unfortunately, maps, probably, can not be represented!)

Fig.1. Sibiria winter palaeotemperatures, reconstructed by δ18O in ice wedges formed 30-25 (a) and 22-14 (b) ka BP.