Estimating Greenland ice sheet surface mass balance contribution to future sea level rise using the regional atmospheric climate model MAR

Abstract

Abstract. We report future projections of Surface Mass Balance (SMB) over the Greenland ice sheet (GrIS) obtained with the regional climate model MAR, forced by the outputs of three CMIP5 General Circulation Models (GCMs) when considering two different warming scenarios (RCP 4.5 and RCP 8.5). The GCMs selected in this study have been chosen according to their ability to simulate the current climate over Greenland. Our results indicate that in a warmer climate (i) the mass gained due to increased precipitation over GrIS does not compensate the mass lost through increased run-off; (ii) the surface melt increases non-linearly with rising temperatures due to the positive feedback between surface albedo and melt, associated with the expansion of bare ice zones which, in addition, decreases the ice sheet refreezing capacity; (iii) most of the precipitation is expected to fall as rainfall in summer, which further increases surface melt; (iv) no considerable change is expected on the length of the melting season, since heavier winter snowfall dampens the melt increase at the end of spring; (v) the increase of meltwater run-off versus temperature anomalies is dependent of the GCM-forced MAR ability to simulate the current climate; (vi) the MAR-simulated SMB changes can be approximated using the annual accumulated snowfall and summer 600 hPa temperature increase simulated by the forcing GCMs. In view of the large range in the CMIP5 future projections for the same future scenario, the GCM-based SMB approximations allow us to estimate what future projections are most likely within the CMIP5 multi-model ensemble. In 2100, the ensemble mean projects a sea level rise, resulting from a GrIS SMB decrease, estimated to be +4 ± 2 cm and +9 ± 4 cm for the RCP 4.5 and RCP 8.5 scenarios, respectively. The GrIS SMB should remain positive with respect to RCP 4.5 scenario and becomes negative around 2070 in the case of the RCP 8.5 scenario since a global warming >+3 °C is needed. However, these future projections do not consider the positive melt-elevation feedback because the ice sheet topography is fixed through the whole simulation. In this regard, the MAR simulations suggest a cumulative ice sheet thinning by 2100 of ~100–200 m in the ablation zone. This highlights the importance of coupling climate models to an ice sheet model to consider the future response of both surface processes and ice-dynamic changes, and their mutual feedbacks to rising temperatures.