Projections of Greenland Ice Sheet Surface Mass Balance Based on CMIP5 Climate Forcing

The contribution of the Greenland Ice Sheet to sea level rise is likely to accelerate in the coming century as atmospheric temperatures continue to rise, potentially triggering even larger surface melt rates. However, at present considerable uncertainty remains in projecting the contribution to sea level of the Greenland Ice Sheet both due to uncertainty in climate forcing and the limitations in the resolution of large-scale global climate models. Regional models provide a means of dynamically downscaling climate forcing, but remain computationally expensive. Here, we examine the feasibility of statistically downscaling coarse scale climate data to provide a higher resolution and more accurate estimate of Greenland Ice Sheet mass loss. We use ERA-interim data to assess model skill degradation associated with statistical downscaling. We find that statistical downscaling results in a relatively small (2%) greater mass loss compared with native resolution simulations based on ERA-interim data. However, statistical downscaling results in pronounced differences in regional patterns that more closely resemble regional climate model based projections, including a 20% increase in surface melt in the Southeastern Greenland and a 15% decrease in mass loss from Southwestern. We next used IPCC Representative Concentration Pathways (RCP8.5) climate scenarios from an ensemble of global climate models in our simulations to project the range of ice volume loss. Our results show that mass loss is likely to continue to accelerate throughout the next century resulting in as much as 1500 Gt mass loss by the end of the century. However, our preliminary results also show significant variability in predicted mass loss associated with individual climate models and this variability rivals or even exceeds uncertainty associated with different downscaling schemes, including dynamic downscaling. This suggests that that projections of Greenland mass balance should rely on a suite of climate models to fully characterize the uncertainty in climate forcing and how it translates to surface mass balance.