Quantifying the historical surface mass balance of the Greenland ice sheet using both RACMO2.4p1 and in situ data, and its correlation with large scale atmospheric circulation

Abstract

Increase in surface melting is the main contributor to Greenland Ice Sheet (GrIS) mass loss, therefore its quantification is extremely relevant in assessing the future rise of sea levels. However, direct measurements of surface melt and runoff are difficult and present many gaps in time and space. Therefore, numerical modeling is essential to calculate melt and assess the current state of the GrIS Surface Mass Balance (SMB), which is the sum of the mass fluxes removing or accumulating mass at the ice-sheet surface. Besides this, the last decades have been marked by a shift in the atmospheric circulation over Greenland, with more negative North Atlantic Oscillation (NAO) phases and higher frequency blocking events during summer that increased the Greenland Blocking Index (GBI). These accompany the decrease in SMB, bringing relatively warm and moist air towards the GrIS. This study uses the latest output of the Regional Atmospheric Climate Model RACMO2.4p1 (5.5 km horizontal resolution) and observations from selected Automatic Weather Stations (AWSs) and annual stakes measurements to: (1) evaluate the model performance; (2) reconstruct the GrIS integrated Surface Mass Balance (SMB) and its components from 1945 to 2023; (3) relate the interannual variability of SMB and SMB components to GBI variability, highlighting the connections between summer melt and higher GBI values. The results show that (1) RACMO2.4p1 underestimates the net shortwave radiation (-3 W/m2) due to a delayed onset of the melting season. The SHF is also underestimated (-2 W/m2), hence leading overall to an underestimation of melt energy, especially in the low-lying stations (i.e. KAN_L, QAS_L, TAS_L), for which the SMB bias between modeled and observed annual SMB is higher (0.1 m w.e. yr−1 for stations above 1000 m a.s.l. vs. 1.7 m w.e. yr−1 below 1000 m a.s.l.). (2) The modeled GrIS integrated SMB between 1945 and 2023 is 491 Gt yr−1, the integrated runoff is 257 Gt yr−1 and the integrated accumulation is 822 Gt yr−1. The rate of integrated SMB decline accelerated from 1982, which marks the onset of a series of climatological periods characterized by the most rapid decreases in SMB. (3) Finally, the SMB is anti-correlated to the summer GBI (R2=-0.67) through the correlation between summer melt and summer GBI (R2=0.85). Both correlations show different spatial sensitivity to GBI, whereas the integrated SMB shows a decrease of 102 Gt/yr when the summer GBI increases by one-standarddeviation. Therefore, this study highlights the important relationship between SMB and summer GBI, and the necessity of realistically simulating regional circulation variability to model future GrIS SMB, as much as refining RACMO2.4p1 to better constrain the current state of the GrIS SMB.