Articles | Volume 10, issue 1
Research article
15 Jan 2016
Research article |  | 15 Jan 2016

SeaRISE experiments revisited: potential sources of spread in multi-model projections of the Greenland ice sheet

F. Saito, A. Abe-Ouchi, K. Takahashi, and H. Blatter

Abstract. The present paper revisits the future surface-climate experiments on the Greenland ice sheet proposed by the Sea-level Response to Ice Sheet Evolution (SeaRISE; Bindschadler et al., 2013) study. The projections of the different SeaRISE participants show dispersion, which has not been examined in detail to date. A series of sensitivity experiments are conducted and analyzed using the ice-sheet model for integrated Earth-system studies (IcIES) by replacing one or more formulations of the model parameters with those adopted in other model(s). The results show that large potential sources of the dispersion among the projections of the different SeaRISE participants are differences in the initialization methods and in the surface mass balance methods, and both aspects have almost equal impact on the results. The treatment of ice-sheet margins in the simulation has a secondary impact on the dispersion. We conclude that spinning up the model using fixed topography through the spin-up period while the temperature is allowed to evolve according to the surface temperature history is the preferred representation, at least for the experiment configuration examined in the present paper. A benchmark model experimental setup that most of the numerical models can perform is proposed for future intercomparison projects, in order to evaluate the uncertainties relating to pure ice-sheet model flow characteristics.

Short summary
This article, as the title denotes, is a follow-up study of an ice-sheet intercomparison project SeaRISE, which focuses on the response of the Greenland ice sheet to future global warming. The projections of the different SeaRISE prticipants show diversion, which has not been examined in detail to date. This study detects the main sources of the diversion by a number of sensitivity experiments and shows the importance of initialization methods as well as climate forcing methods.