Increased glacier runoff enhances the penetration of warm Atlantic water into a large Greenland fjord

Abstract. The retreat and acceleration of Greenland's marine-terminating outlet glaciers have been linked to ocean warming. However the mechanisms which control the transmission of this warming along fjords towards the glaciers remain poorly understood. The aim of this paper is to elucidate observed changes in water properties in Kangerdlugssuaq Fjord (KF), East Greenland using the Bergen Ocean Model (BOM). Model outputs are compared with observed potential temperature, salinity and velocity data to determine the principal controls on heat transport within KF and to estimate resulting submarine ice front melt rates of Kangerdlugssuaq Glacier (KG). The BOM includes wind, tidal and glacier runoff forcing and is able to replicate observed temperature and salinity profiles. Model results describe a robust four-layer estuarine flow, consisting of two distinct circulations. The shallow circulation (0–~ 60 m) is forced by surface wind stress and to a lesser extent supraglacial runoff, while the intermediate circulation (~ 60–500 m) is driven by runoff discharged into the fjord subglacially. Atlantic Water (AW) and warm Polar Surface Water (PSWw) are drawn into the fjord by the intermediate and shallow circulation cells respectively, in a pattern consistent with observations, and AW reaches KG over a single summer. Along-fjord heat transport towards KG increases significantly with both glacier runoff and coastal water temperature. A doubling of glacier runoff produces a 29% (48%) amplification of mean annual (summer) heat transport towards the KG terminus, increasing estimated mean annual (summer) submarine melt rates from 211 to 273 (842 to 1244) m yr^–1. In contrast, heat transport towards KG in the surface ~ 60 m of the fjord decreases with rising glacier runoff because the enhanced down-fjord component of the intermediate circulation interferes with the up-fjord part of the shallow circulation. Thus, as ice sheet runoff increases, KG's dynamic response to oceanic forcing will likely be driven primarily by enhanced submarine ice front melting and consequent undercutting rather than through diminished buttressing from seasonal sea ice and ice mélange. Our model shows, in agreement with observations, that maximum submarine melt rates occur when AW and PSWw are present at the fjord mouth and, crucially, glacier runoff is also high. Rising ice sheet runoff therefore increases the sensitivity of KG (and other Greenland marine-terminating glaciers) to ocean warming.