CARRA-driven simulation of Greenland Ice Sheet surface mass balance at 2.5 km resolution

Projections of present and future ice mass loss of the Greenland Ice Sheet are important for assessing its contribution to future sea-level rise. Critical for the total mass balance is the surface mass balance (SMB) which can be estimated from models, and improving these models can help to further constrain the uncertainties in future projections.

In this project, we use the CARRA reanalysis dataset generated from the HARMONIE-AROME weather forecast system to force an SMB model. The CARRA dataset is remarkable for its 2.5 km horizontal resolution providing unprecedented spatial detail. This is particularly important at the ice-sheet margins where both accumulation and ablation processes are impacted by strong topographic gradients. For example, the greater spatial detail is expected to provide more realistic profiles of accumulation and drying of airmasses from the coast toward the interior, in turn improving the SMB simulation.

The SMB model utilizes a subsurface scheme that consists of columns with 32 layers in the vertical. Driven by the atmospheric input, the SMB model computes all the interactions between the atmosphere and subsurface layers, such as accumulation, melting, percolation, refreezing and runoff. Using this SMB model, we performed a CARRA-driven simulation over the period 1991-2020 on the 2.5 km CARRA grid.

Our initial results show the CARRA-driven SMB model yielding somewhat higher SMB values compared to other published SMB products. The ice sheet-wide totals of accumulation and melt are comparable to other products. However, the location of maximum melt contributions is shifted further towards the interior of the ice sheet in the CARRA-driven simulation. This allows for larger refreezing and contributes significantly to the high SMB seen in the CARRA-driven simulation. Here, we evaluate the SMB model output and driving fluxes against PROMICE data and satellite observations and provide a new updated assessment of Greenland ice sheet SMB.