Projecting the retreat of Narsap Sermia using a minimum observational inputs approach

Mass loss from marine terminating glaciers in Greenland is often simulated using physically-based models driven by multiple parameters such as atmospheric forcing, submarine melt, fjord topography and iceberg calving, each of which carry their own uncertainties. While these models may perform well, they are time intensive to set up, calibrate and validate, and the development of simpler though potentially informative observationally-based models has received less attention.

In this study, we develop a simple observationally derived approach to hindcast and project the future behaviour of Narsap Sermia, a marine-terminating glacier in Nuuk Fjord, south-west Greenland. This glacier has experienced significant retreat (~3.5km) between 2014 and 2024 and is approaching a significant overdeepening located ~7.5km from its current terminus position. Once it reaches this overdeepening, it has the potential to rapidly destabilise, with the next likely stable topographic configuration located ~21km further up-glacier. This will have implications for the safety of local and tourist activities and the operation of Greenland’s largest port at nearby Nuuk.

To achieve this, we have constrained terminus behaviour under different fjord conditions through analysis of terminus positions from satellite imagery. The link between terminus migration and fjord conditions has been compared with varying potential drivers (e.g. runoff, fjord surface temperature and air temperature), and is found to be most closely linked to the presence or absence of a proglacial ice mélange which in turn is linked to cumulative positive and negative degree days. Using calculated degree day thresholds, the model estimates dates of mélange formation and break-up, driving changes in the pattern of terminus migration by switching between observationally derived values of terminus change for rigid mélange and open water conditions. The model reproduces terminus migration at Narsap Sermia over the 2014-2024 period, achieving a mean absolute deviation of 243m for the entire period of observed retreat. Assuming current calving behaviour continues, we are able to project future mélange behaviour and terminus migration using bias corrected CMIP6 2m air temperature data for three climate scenarios. We use this approach to explore a range of scenarios projecting when Narsap Sermia will reach the overdeepening.