Uncertainty quantification of Greenland Ice Sheet mass balance via assessment of GRACE mascon solutions and variability of high-resolution surface mass balance forcing of an ice flow model

Accurate sea level projections hinge on the ability to precisely model how the Greenland and Antarctic ice sheets respond to climate change. In this study, we compare a state-of-the art ice flow model, the Ice Sheet System Model (ISSM), to GRACE mascon solutions over the Greenland ice sheet at large spatial scales (~300 km), and monthly temporal resolution. With the use of established ISSM capabilities, we investigate how errors and variations in surface mass balance (SMB) forcing propagate through the model resulting in uncertainties in ice discharge. We identify which SMB components contribute the most to model uncertainty, as well as quantify leakage errors due to land/ocean boundaries in the GRACE solution. Additional errors sourced in the ISSM spinup (e.g. errors in bedrock, lack of physics) are identified. Rigorous uncertainty quantification of the surface mass balance models, GRACE solutions, and the ISSM ice flow model is performed via comparison of average trends and estimated seasonal amplitudes.