Pollard et al say:
"The equivalent eustatic sea level rise reaches 5 m after ∼200 yr and 17 m after ∼3000 yr (Fig. 4, red curve), similar in magnitude to albeit uncertain proxy estimates of past sea-level variations mentioned above. About 3 mesl comes from West Antarctica, and the remaining ∼14 mesl comes from East Antarctic basins."
And:
"To investigate the impact of the cliff-failure and melt-driven hydrofracture mechanisms, the ice-sheet model is run forward in time, forced by climate representative of past warm periods. Simulations are started from a previous spin-up of modern Antarctica using observed climatology. An instantaneous change to a warmer climate is applied, broadly representative of a warm Pliocene period. The past warm atmospheric climate is obtained from the RegCM3 Regional Climate Model (Pal et al., 2007) applied over Antarctica with some physical adaptations for polar regions, and with 400 ppmv CO2 and an orbit yielding particularly strong austral summers (DeConto et al., 2012). Detailed simulation of ocean warming beneath Antarctic ice shelves is currently not feasible on these time scales, so a simple uniform increment of View the MathML source is added to modern observed ocean temperatures, broadly consistent with circum-Antarctic warming in Pliocene paleo-oceanic reconstructions (Dowsett et al., 2009)."
And:
"For simplicity, this paper uses step-function climate forcing representative of generalized warming episodes during the late Cenozoic. A natural next step will be to use time-dependent forcing to model specific warm events or periods of the past and compare with available data, such as warm Pliocene intervals ∼5–3 Ma, MIS-31 at ∼1.08 Ma, and strong Pleistocene interglacials (Naish et al., 2009, Raymo and Mitrovica, 2012 and O'Leary et al., 2013). Another important step will be the use of regional ocean models to resolve different oceanic responses in different Antarctic embayments (Hellmer et al., 2012)...
The main aim of adding hydrofracturing and cliff failure was to produce total Antarctic retreat consistent with albeit poorly constrained past sea-level data, and no effort was made to adjust the rate of retreat. The time scale that emerges for West Antarctic collapse (∼3 m contribution to global sea-level rise within O(100) years after a step-function warming) is an order of magnitude faster than previous estimates for the next century, which range from ∼0.1 to 0.6 m by 2100 AD (Pfeffer et al., 2008, Levermann et al., 2014 and Joughin et al., 2014). The modeling approaches in Pfeffer et al. and Levermann et al. are very different, and our study is not directly applicable to the future because of our step-function climate change, Pliocene-like climate, and homogeneous ocean warming. But even so, our predicted WAIS retreat rates are much faster than might be expected from the previous work."
So what does this mean?
My understanding: on the one hand it's unlikely that WAIS-collapse will start as fast as in this simulation, because we're not at the Pliocene climate yet. On the other hand, the forcing now and in the future will probably be much stronger than during the Pliocene, except maybe in the strongest mitigation cases. So once collapse does fully set in, say in the second part of this century, it could be very fast, with 3m of SLR within 100 years from WAIS alone. With contributions from GIS of possibly up to 2 meter/century, according to Applegate et al 2014 (see folder What's new in Greenland?), we could get 5-6 meter in 100 years, say from 2100-2200, for a total of maybe 7-8 meter by 2200. That has a 0.1-0.5% chance in Kopp et al 2014, but this may now seem an under-estimate, as Kopp et al already thought possible.
Any other views?