Methods for Modeling Sea-level Rise in a 4°C World The authors developed sea-level scenarios using a combination of approaches, acknowledging the fact that both physicallybased numerical ice sheet modeling and semi-empirical methods have shortcomings, but also recognizing the need to provide ice sheet loss estimates to be able to estimate regional sea-level rise. They did not attempt to characterize the full range of uncertainties, either at the low or high end. Future contributions from groundwater mining are also not included in the projections, and could account for another 10 cm (Wada et al. 2012). The scenario construction is as follows. For the upper end of the sea-level scenario construction, the authors apply a semi-empirical sea-level rise model (Rahmstorf, Perrette, and Vermeer 2011; Schaeffer et al. 2012), giving a global estimate for specific emission scenarios leading to a 2°C or 4°C increase in global mean temperature by 2100. As the semi-empirical sea-level rise models do not separately calculate the individual terms giving rise to sea-level increases, further steps are needed to characterize plausible ice sheet contributions. The authors calculate the contribution from thermal sea-level rise and from mountain glaciers and icecaps and deduct this from the total global sea-level rise and assign this difference to the ice sheets, half to Greenland and the other half to Antarctica. The resulting contributions from the ice sheets are significantly above those estimated by most process based ice sheet models and approximates the ice sheet contribution that would arise, if the rates of acceleration of loss observed since 1992 continued unchanged throughout the 21st century. For the lower end of the scenario construction, the authors use as a starting point the calculated thermal sea level-rise and the contribution from mountain glaciers and ice caps. To this, they add a surface mass balance contribution from the Greenland ice sheet (GIS; excluding ice dynamics) and assume that the Antarctic ice sheet (AIS) is in balance over the 21st century. Most AIS models project that this ice sheet would lower sea-level rise in the 21st century as it does not warm sufficiently to lose more ice than it gains because of enhanced precipitation over this period. On the other hand, observations indicate that the ice sheet is losing ice at a slowly increasing rate close to that of the Greenland ice sheet
at present. Setting the AIS contribution to zero is, thus, a way of leaving open the possibility that short-term processes may have been at work over the last 20 years. This very low ice sheet contribution scenario approaches the levels of some process-based model projections, where the projected net uptake of ice by Antarctica is balanced by ice melting from Greenland over the 21st century. In the lower ice-sheet scenario (47 cm sea-level rise in the global mean), eastern Asian and northeastern American coasts both experience above-average sea-level rise, about 20 percent and 15 percent, respectively above the global mean (for example, –3 percent to +23 percent around New York City, 68 percent range). In the higher ice-sheet scenario (96 cm sea-level rise in the global mean), where ocean dynamic effects are relatively less significant, the eastern Asian coast clearly stands out as featuring the highest projected coastal sea-level rise of 20 percent above the global mean. In that scenario, sea-level rise is projected to be slightly below the global mean in northeast America, and 20 percent (5–33 percent, 68 percent range) below the global mean along the Dutch coast (Figure A1.1, Figure 32). It is important to note the likely weakening in the Atlantic Meridional Overturning Circulation (AMOC) with increasing warming could be exacerbated by rapid ice sheet melt from Greenland. That effect, which is not included in the authors’ projections, could potentially add another 10 cm to the local sea-level rise around New York City, as currently discussed in the scientific literature (Sallenger et al. 2012; Slangen et al. 2011; Stammer, Agarwal, Herrmann, Köhl and Mechoso 2011; Yin et al. 2009). Post-glacial adjustment would also add another 20 cm, albeit with large uncertainties (Slangen et al. 2011). 67