Greenland Ice Sheet Ocean Interaction
Prof. Jürgen Kusche
University of Bonn, Institute for Geodesy and Geoinformation, Bonn, Germany
Local Project ID:
HPC Platform used:
JUWELS/JURECA at JSC
The Greenland Ice sheet has been losing mass to the ocean over the last two decades, as detected with gravimetric satellite missions (Gravity Recovery and Climate Experiment, GRACE, and its successor GRACE-FO), and other observations. Several studies derived information on the individual components of Greenland freshwater flux (solid ice discharge, runoff, snowmelt) from in-situ measurements, satellite-based observations and regional atmospheric climate models. Between 2010 and 2018, the reconstructed ice mass loss of the Greenland ice sheet is estimated to about 280-290 Gt/year (1Gt corresponds to 1 mio tonnes and thus about the mass of a cube with 1 km side length filled with water), and a rapid mass loss event followed in 2019. This accelerated melting places the Greenland ice sheet as one of the main contributors to global sea level rise during the last two and a half decades and has caused a cumulative global mean sea level rise of about 10-11 mm since the 1990s. Future projections indicate that the Greenland ice sheet contribution to global mean sea level rise could reach 13 cm (on average, relative to the baseline of 1995–2014) by the end of the 21st century.
Increased Greenland ice sheet melting has an impact on global mean and regional sea level rise and the ocean circulation. The regional sea level rise and, occasionally sea level fall is however far from being uniform. In this study, we explore whether Greenland melting signatures found in numerical ocean model simulations are visible in observations from radar altimetry, satellite gravimetry and Argo floats (buoys that dive to about 2000 m depth and measure vertical profiles of temperature and salinity). We have included realistic Greenland freshwater flux in the global Finite-Element-Sea ice-Ocean Model (FESOM) for the years 1993–2016, whereas a reference run is computed by excluding Greenland freshwater input. Our experiments are performed on a low resolution (ca. 24 km) and a high resolution (ca. 6 km) eddy-permitting mesh. For comparison with the model experiments, we use direct observational data, such as Argo floats, various satellite observations, and reanalyses (ocean model simulations that integrate data sets). We find that surface freshwater flux maps into distinct signatures or patterns in temperature and salinity down to about 100 m in the surroundings of Greenland. The simulated melting signatures are particularly visible in steric heights (sea level ride due to volumentric water expansion) in Baffin Bay and Davis Strait. Here, we find an improvement of the mean square error of up to 30% when including Greenland freshwater flux. In other words, simulations that include the melting effect fit much better to real observations of the ocean. The picture is mixed hower; for the Nordic part of the Nordic Seas, we find no improvement when including the melting. We compare steric heights with reanalysis data and a new setup of an inversion method that integrates data from gravimetric and altimetric satellites and that takes into account self-gravitation effects (the changes to the equilibrium water surface due the reduced gravitational attraction of the ice sheet) and the elastic uplift of the Earth’s crust under a melting ice sheet.
Our study was part of the collaborative research projects GROCE (Greenland Ice Sheet Ocean Interaction) and GROCE-2 funded by the BMBF (Bundesministerium für Bildung und Forschung, Federal Ministry for Education and Research). The aim of GROCE/GROCE-2 was to obtain an in-depth knowledge of changing processes and dynamics in Greenland ice sheet considering changing climate conditions and the interaction with the surrounding ocean. One focus was also to assess the degree of realism of patterns of ice-sheet mass loss from Greenland as predicted by state of the art IPCC-type climate models.
Study relevant changes in temperature, salinity, and sea level via oceanographic modeling requires HPC access and the use of sophisticated numerical models; we had to perform several several global simulations on the JEWELS system with the Finite Element Sea-Ice Ocean Model (FESOM) developed at the Alfred-Wegener Institute (AWI), which allows to increase the resolution locally i.e. in the vicinity of Greenland.
The animation shows the effect of recent accelerated Greenland ice sheet and glacier melt on the North Atlantic ocean salinity and temperature in the uppermost layers. Adding freshwater disturbes the ocean circulation and the spatial distribution of heat and salt; shown are anomalies with respect to an unperturbed ocean. Temperature increases (red color in the two bottom figures) lead the ocean water to expand and thus to sea level rise. Salinity changes have a smaller effect and average out globally. The two figures in th eleft column were computed with a low-resolution ocean model configuration, the two figures in the right column with a high-resolution (eddy-permitting) configuration.
Stolzenberger, S., Rietbroek, R., Wekerle, C., Uebbing, B., & Kusche, J. (2022). Simulated signatures of Greenland melting in the North Atlantic: A model comparison with Argo floats, satellite observations, and ocean reanalysis. Journal of Geophysical Research: Oceans, 127, e2022JC018528. https://doi. org/10.1029/2022JC018528