Submarine melting of Greenland's glaciers: What are the relevant ocean dynamics?

By Inga Koszalka, Thomas Haine (Johns Hopkins University) and Marcello Magaldi (ISMAR-CNR)


Submarine melting of Greenland's glaciers has emerged as a key term in the ice sheets mass balance and as a plausible trigger for their recent acceleration, which contributed to doubling Greenland's contribution to sea-level rise. Notwithstanding its importance, understanding of submarine melting is limited and it is presently absent or crudely parameterized in glacier, ice sheet and climate models. Greenland's tidewater glaciers end in ~600-800m deep, long fjords that connect the margins of the ice sheet to the shelf. The glaciers' termini are typically grounded several hundreds of meters below sea-level and, as such, are exposed to a thick portion of the fjords' water column. Recently collected data from two large glacial fjords in south-east Greenland (located in the Sermilik and Kangerdlugssuaq fjords) shows that these are filled year-round with cold, fresh waters of Arctic origin and warm, salty waters of subtropical origin advected into the fjords from the East Greenland Shelf (Straneo et. al., 2010, Nature Geoscience, 3). This means that submarine melt rates may depend on a suite of oceanic processes including externally forced fjord circulations, fjord/shelf exchange and the distribution of properties on the shelf. Yet, the details of how these processes may contribute to the submarine melt rate or affect its variability are presently unknown.
Fig.1. The study area with the location of the East Greenland Shelf marked with a star

Fig.2. Schematic of the currents in the upper layer superimposed on the modelled depth-averaged speed (m/s). Red (blue) stands for warm (cold) currents. From Magaldi et. al., (2011), J. Phys. Oceanogr., 41, p., 2308. The locations of the Sermilik and the Kangerdlugssuaq Fjords are shown.

Project goals

The project aims at identifying parameters and mechanisms which control the properties and circulation in the south-east Greenland fjords (see Fig.1-2) and thus the rate of submarine melting at the ocean/glacier interface. We focus on two glacier fjords: Sermilik and Kangerdlugssuaq (see Fig. 2). These two fjords have been in the focus of recent observational campaigns (e.g., Straneo et. al., 2010, Nature Geoscience, 3, Straneo et. al., 2011, Nature Geoscience, 4). The observations suggest that the warm and salty Subtropical Atlantic Water (STW) carried by the Irminger Current (IC) can flow onto the East Greenland Shelf (EGS) and into the fjords (Fig.2). This warm water inflow may significantly augment glacier melting in the fjords. This work aims at investigating the inflow of the warm Atlantic-origin water onto the shelf and into the mouths of the Sermilik and Kangerdlugssuaq fjords. The project has two phases. First, we will use an existing configuration of a high resolution regional ocean model with fields retrieved at at high sampling rate and supplemented by Lagrangian particle simulations to study the circulation on the East Greenland Shelf. By this, we aim at addressing these questions:

-- What are the pathways of Subtropical Atlantic Water (STW) on the continental shelf?

-- What processes control the inflow and variability of Subtropical Atlantic Water (STW) onto the continental shelf?

-- What controls the property distributions at the fjord mouths?

Second, we will run the circulation model for a 20-year period to investigate the interannual variability and to answer these questions:

-- How did environmental conditions before and after 1995 govern STW intrusions onto the shelf?

-- How might future atmospheric and oceanic conditions govern STW intrusions onto the shelf?

To perform these tasks, we employ a high resolution ocean model validated by observations, and a Lagrangian particle code.

High resolution circulation model of the Irminger Basin

We use the MIT General Circulation Model (MITgcm) configured for the Irminger Basin area. The configuration features:

-- Partial bottom cells and rescaled coordinate (for a better representation of flows in a complex bathymetry)

-- Highest spatial resolution to date of the Irminger Basin: dx~2km, dz=2-15m (210 layers), dt=30s

-- Simulation period: summer 2003 (7/1 - 9/1)

-- 3rd tracer advection with implicit diffusion; KPP for vertical diffusion, Leith horizontal viscosity

-- Forcing by NCEP fluxes & SeaWinds

-- We verified that the model transports and hydrography agree with the observations

For more details, see: I. Koszalka, T. W. N. Haine and M. G. Magaldi: Fates and travel times of Denmark Strait Overflow Water in the Irminger Basin, J. Phys. Oceanogr., 43, 2611–2628: doi

The distribution of the Subtropical Atlantic Water in the model

First we looked into the distribution of the Subtropical Atlantic Water (potential temperatures θ>=4 C) in the study area. The STW constantly occupies the surface and sub-surface layers of the Irminger Basin, with a prominent boundary flow of the Irminger Current along the Iceland Shelf and along the East Greenland Shelf (Fig. 3a). Occasionally, the STW flows onto the East Greenland Shelf and toward Sermilik Fjord (-38E, 65.5N) and Kangerdlugssuaq Fjord (-31E, 68N). This inflow is steered by the bathymetric depressions on the shelf: Sermilik Deep and  Kangerdlugssuaq Trough, respectively (Fig. 3b).
Fig.3. (a, left): Frequency of occurrence (in % of the 60-day simulation period) of the Atlantic Subtropical Water (STW), defined as waters of the potential temperature higher than (θ=4 C). Superimposed are depth-averaged currents at the depth range of 200-800m. (b, right): Bathymetry (depth of the sea bed) in the study area.

Particle code

We integrate the Lagrangian particles off-line with a code written in MATLAB. At each time step, the simulated velocity fields (u,v,w components) are linearly interpolated on current particle locations. The particles are then advanced with MATLAB ode23t routine (2nd order predictor, 3rd order corrector). The time step is 15 minutes, chosen after a sensitivity study. For the boundary conditions, the code identifies events of particles hitting the boundary (inevitable since the time step for the Lagrangian code is larger than for the model simulation) and resumes the time stepping with a particle put on the boundary with the velocity component normal to the wall equal zero.

We performed two particle deployments in the STW layer: one at the mouth of the Sermilik Fjord and one at the Kangerdlugssuaq Fjord. The particle deployments are summarized in Table 1. Particles were integrated backwards to study the pathways of the Atlantic Waters to the fjords' mouths.

Deployment Date Simulation length (days) No. particles
Sermilik 1 24-Aug-2003 54 38
Kanger 1 26-Aug-2003 56 28

Table 1. Summary of the particle deployments at the mouth of the Sermilik and Kangerdlugssuaq Fjords.

Trajectories of the Lagrangian particles

Trajectories of the backward-integrated particles deployed at the mouths of the Sermilik and Kangerdlugssuaq Trough are shown in Fig.4. The Animations of particles' positions are available here: MOVIES.

Fig.4. Trajectories of 38 backward-integrated Lagrangian particles deployed at the mouth of the Sermilik Fjord (red) and the 28 particles at the mouth of the Kangerdlugssuaq Fjord (dark blue). Note that the particles inflowing to the Sermilik Fjord (38 particles) tend to follow the branch of the Irminger Current that separates from the boundary current flowing along the continental slope onto the shelf at the Sermilik Deep. The particles arriving at the Kangerdlugssuaq Trough (28 particles) come either from the branch of the Irminger Current coming straight from the Irminger Basin or the branch recirculating in the Denmark Strait. Two particles of this latter set come from further north of the Denmark Strait with the dense overflow water (see also Fig. 5 below). The Kangerdlugssuaq particles reciculate in the Kangerdlugssuaq Trough before reaching the fjord mounth.

Water property transformation

From the times series of water mass properties (temperature, salinity) on particles we can identify the locations of the strongest transformation, shown in Figure 5. The strongest transformation occurs at the front between the Polar Waters and the Atlantic Waters along the shelf break and in the Sermilik Deep. For the particles released at the Sermilik Fjord, temperature-salinity (T-S) diagrams point to mixing between the Subpolar Atlantic Water in the Irminger current, the Recirculating Atlantic Water and the Polar Surface Water as the main process causing the transformation. For the particles released at the Kangerdlugssuaq Trough there is also evidence for mixing with the dense overflow (the Arctic Atlantic Water mass).
Fig.5. Locations of strongest transformation (θ> 0.5C) derived from one-day averaged time series of potential temperature on the trajectories.
Fig.6. (a, left): T-S diagrams with marked potential temperature-salinity values for the particles released at the mouth of Sermilik Fjord (blue dots) and the values at the end of the simulation (brown crosses), linked by straight red lines. (b,right): Same for the particles released at the mouth of Kangerdlugssuaq Fjord. Abbreviations for water masses (based on Rudels et. al., 2002, ICES J. of Mar. Sci., 59): AW (Atlantic Water, or STW), RAW (Recirculating Atlantic Water), PSWw (Polar Surface Water warm), AAW (Arctic Atlantic Water).


-- The Atlantic Waters arrive at Sermilik Fjord through a branch of the Irminger Current separating from the boundary flow at Sermilik Deep.

-- The Atlantic Waters arrive at Kangerdlugssuaq Fjord with either a branch of the Irminger Current coming directly from the Irminger Basin or one recirculating in the Denmark Strait. Few dense water overflow particles from the Denmark Strait are mixed into this inflow as well.

-- The strongest transformation occurs at the frontal zone between the Atlantic Waters and the Polar Surface Waters.


Inga Koszalka, Johns Hopkins University:,

Thomas Haine, Johns Hopkins University:,


This work is supported by NSF grant OCI-1129895. Particle simulations have been performed on the Johns Hopkins Data-Scope funded by OCI-1040114.

Header photo: East Greenland, March 2007, by T. Haine

Last updated: April 4,  2014