Long-term surface energy balance of the western Greenland ice sheet and the role of large-scale circulation variability

Abstract. We present the surface energy balance (SEB) of the west Greenland ice sheet (GrIS), using an energy balance model forced with hourly observations from nine automatic weather stations (AWS) along two transects: the K-transect with seven AWS in the southwest and the T-transect with two AWS in the northwest. Modeled and observed surface temperatures for non-melting conditions agree well, with RMSEs of 1.1–1.6 K, while reasonable agreement is found between modeled and observed 10-day cumulative ice melt. Absorbed shortwave radiation (Snet) is the main energy source for melting (M), followed by the sensible heat flux (Qh). The multi-year average seasonal cycle of SEB components show that Snet and M peak in July at all AWS. The turbulent fluxes of sensible (Qh) and latent heat (Ql) decrease significantly with elevation, and the latter becomes negative at higher elevations, partly offsetting Qh. Average June, July, August (JJA) albedo values are  0.7 for the higher stations. The near-surface climate variables and surface energy fluxes from reanalysis products ERA-interim, ERA5 and the regional climate model RACMO2.3 were compared to the AWS values. The newer ERA5 product only significantly improves on ERA-interim for albedo. The regional model RACMO2.3, which has higher resolution (5.5 km) and a dedicated snow/ice module, unsurprisingly outperforms the re-analyses for (near-) surface climate variables, but the reanalyses are indispensable to detect dependencies of west Greenland climate and melt on large-scale circulation variability. We correlate ERA5 with the AWS data to show a significant positive correlation of western GrIS summer surface temperature and melt with the Greenland Blocking Index (GBI), and weaker and opposite correlations with the North Atlantic Oscillation (NAO). This analysis may further help to explain melting patterns in the western GrIS from the perspective of circulation anomalies.