Abstract

Firn provides pore space that can retain meltwater. Its depletion on the Antarctic ice shelves can lead to hydrofracturing. To understand firn behaviour, we model its densification, which is so far done on a relatively course grid of 27 km in the IMAU firn densification model (IMAU-FDM). The coarse resolution cannot capture firn air content (FAC) gradients in areas with complex topographies, overestimating FAC near the grounding lines. This study pioneers statistically downscaled forcing data to drive IMAU -FDM at a 2 km resolution on the ice shelves Larsen C, George VI, Roi Baudouin, and Thwaites. Near grounding lines on the ice shelves, the downscaled data shows increased snowmelt and reduced snowfall, leading to up to 10% lower FAC at high resolution. Overall, we found that the high-resolution FAC aligns better with the observations in areas with complex topographies. Surprisingly, the temporal resolution of the input data has a larger influence on the FAC than previously expected: the absence of a daily cycle in the downscaled input data increased the FAC by up to 20% compared to earlier simulations that included a daily cycle. During my PhD, I will further investigate the sensitivity of the model to forcings with different temporal resolution.