Meltwater effecting the ocean

Coastline in Northeast Greenland. (CC-BY Kirsten Meulenbroek)


Satellite observations and ocean models are needed to better understand changes in the North Atlantic. Since the last 25 years there has been an increased mass loss of Greenland ice. The direct impact on the surrounding ocean needs to be better understood. For this purpose, satellite data providing mass changes over land (e.g. GRACE) as well as changes in sea surface heights (altimeter satellites, e.g. Envisat, Cryosat-2), local observations and model experiments are available. Ocean model simulations can be used to calculate scenarios of how Greenland freshwater is distributed in the ocean. In Subproject 10, we are interested in how well Greenland melt signatures found in model simulations are represented in Earth observations or other models. 

Contacts: Dr. Sophie Stolzenberger & Prof. Dr. Jürgen Kusche

Fig. 1: The differences between the model experiments with and without Greenland freshwater for the upper 100 m are particularly noticeable in Baffin Bay and around Greenland. Differences for salinity (left) and temperature (right) are shown. Red indicates that it gets warmer/saltier due to the melt rates. (Stolzenberger et al., in prep.)

Figure 2: Simulated steric heights in Baffin Bay fit the ORAS5 ensemble well, both for year-to-year (top) and month-to-month (bottom) variations. The model experiments including Greenland freshwater flux (GF) and without (NGF) are shown. (Stolzenberger et al., in prep.)

What was the main question you aimed to answer during the first project phase? 

In GROCE-1, we have investigated the extent to which Greenlandic melt signatures can be found in geodetic and oceanographic data. The goal is to better understand changes in the North Atlantic and in the immediate vicinity of Greenland from simulations and various Earth observations.

Which methods do you use to answer your research question? 

First, in an integrated analysis, we analyzed the gravimetric mass balance of Greenland together with sea level variations in the North Atlantic to separate mass-related and steric contributions. We also used ocean model simulations (FESOM, see also TP2) and observational data to determine relevant changes in temperature and salinity that are responsible for the observed steric variations, among other factors. To analyze the influence of Greenland meltwater, we performed a model experiment each with and without corresponding freshwater flux.

What were your main results?

  • The effects of Greenland meltwater can be detected down to a depth of 100 m near the Greenland coast. The freshwater input shows a warming, less saline effect west of Greenland and a cooling of up to 1°C further south (Figure 1).
  • Baffin Bay west of Greenland is an area where ocean melt rates are particularly visible due to the current system. Simulated steric heights that include the Greenland freshwater flux fit well with the ORAS5 reanalysis especially in the 1990s and until 2013 (Figure 2). The model resolution must be appropriately high for this to realistically transport meltwater onward from the coast.

What are your goals for GROCE-2?

  1. In GROCE-2 we are dealing with how realistically the ocean components are represented in the CMIP6 climate models. For this purpose, we will investigate the influence of uncertainties arising from (regional) atmospheric forcing and model resolution.
  2. We will analyze the effect of atmospheric forcing on oceanic variables in sensitivity studies with a special focus on the northeast region of Greenland.
  3. In addition, we will determine altimetric-gravimetric mass balances with longer dated and updated data, including for periods before the GRACE era.