Kerstin Jochumsen

On the origin and propagation of Denmark Strait overflow water anomalies in the Irminger Basin

Denmark Strait Overflow Water (DSOW) supplies the densest contribution to North Atlantic Deep Water and is monitored at several locations in the subpolar North Atlantic. Hydrographic (temperature and salinity) and velocity time series from three multiple-mooring arrays at the Denmark Strait sill, at 180 km downstream (south of Dohrn Bank) and at a further 320 km downstream on the east Greenland continental slope near Tasiilaq (formerly Angmagssalik), were analyzed to quantify the variability and track anomalies in DSOW in the period 2007-2012. No long-term trends were detected in the time series, while variability on time scales from interannual to weekly was present at all moorings. The hydrographic time series from different moorings within each mooring array showed coherent signals, while the velocity fluctuations were only weakly correlated. Lagged correlations of anomalies between the arrays revealed a propagation from the sill of Denmark Strait to the Angmagssalik array in potential temperature with an average propagation time of 13 days, while the correlations in salinity were low. Entrainment of warm and saline Atlantic Water and fresher water from the East Greenland Current (via the East Greenland Spill Jet) can explain the whole range of hydrographic changes in the DSOW measured downstream of the sill. Changes in the entrained water masses and in the mixing ratio can thus strongly influence the salinity variability of DSOW. Fresh anomalies found in downstream measurements of DSOW within the Deep Western Boundary Current can therefore not be attributed to Arctic climate variability in a straightforward way

On the hydrography of Denmark Strait

Using 111 shipboard hydrographic sections across Denmark Strait occupied between 1990 and 2012, we characterize the mean conditions at the sill, quantify the water mass constituents, and describe the dominant features of the Denmark Strait Overflow Water (DSOW). The mean vertical sections of temperature, salinity, and density reveal the presence of circulation components found upstream of the sill, in particular the shelfbreak East Greenland Current (EGC) and the separated EGC. These correspond to hydrographic fronts consistent with surface-intensified southward flow. Deeper in the water column the isopycnals slope oppositely, indicative of bottom-intensified flow of DSOW. An end-member analysis indicates that the deepest part of Denmark Strait is dominated by Arctic-Origin Water with only small amounts of Atlantic-Origin Water. On the western side of the strait, the overflow water is a mixture of both constituents, with a contribution from Polar Surface Water. Weakly stratified “boluses” of dense water are present in 41% of the occupations, revealing that this is a common configuration of DSOW. The bolus water is primarily Arctic-Origin Water and constitutes the densest portion of the overflow. The boluses have become warmer and saltier over the 22 year record, which can be explained by changes in end-member properties and their relative contributions to bolus composition.

On the propagation and decay of North Brazil Current rings

Near the western boundary of the tropical North Atlantic, where the North Brazil Current (NBC) retroflects into the North Equatorial Countercurrent, large anticyclonic rings are shed. After separating from the retroflection region, the so-called NBC rings travel northwestward along the Brazilian coast, until they reach the island chain of the Lesser Antilles and disintegrate. These rings contribute substantially to the upper limb return flow of the Atlantic Meridional Overturning Circulation by carrying South Atlantic Water into the northern subtropical gyre. Their relevance for the northward transport of South Atlantic Water depends on the frequency of their generation as well as on their horizontal and vertical structure. The ring shedding and propagation and the complex interaction of the rings with the Lesser Antilles are investigated in the inline equation Family of Linked Atlantic Model Experiments (FLAME) model. The ring properties simulated in FLAME reach the upper limit of the observed rings in diameter and agree with recent observations on seasonal variability, which indicates a maximum shedding during the first half of the year. When the rings reach the shallow topography of the Lesser Antilles, they are trapped by the island triangle of St. Lucia, Barbados and Tobago and interact with the island chain. The model provides a resolution that is capable of resolving the complex topographic conditions at the islands and illuminates various possible fates for the water contained in the rings. It also reproduces laboratory experiments that indicate that both cyclones and anticyclones are formed after a ring passes through a topographic gap. Trajectories of artificial floats, which were inserted into the modeled velocity field, are used to investigate the pathways of the ring cores and their fate after they encounter the Lesser Antilles. The majority of the floats entered the Caribbean, while the northward Atlantic pathway was found to be of minor importance. No prominent pathway was found east of Barbados, where a ring could avoid the interaction with the islands and migrate toward the northern Lesser Antilles undisturbed.

Enhanced turbulence driven by mesoscale motions and flow-topography interaction in the Denmark Strait Overflow plume

The Denmark Strait Overflow (DSO) contributes roughly half to the total volume transport of the Nordic overflows. The overflow increases its volume by entraining ambient water as it descends into the subpolar North Atlantic, feeding into the deep branch of the Atlantic Meridional Overturning Circulation. In June 2012, a multiplatform experiment was carried out in the DSO plume on the continental slope off Greenland (180 km downstream of the sill in Denmark Strait), to observe the variability associated with the entrainment of ambient waters into the DSO plume. In this study, we report on two high-dissipation events captured by an autonomous underwater vehicle (AUV) by horizontal profiling in the interfacial layer between the DSO plume and the ambient water. Strong dissipation of turbulent kinetic energy of O( math formula) W kg−1 was associated with enhanced small-scale temperature variance at wavelengths between 0.05 and 500 m as deduced from a fast-response thermistor. Isotherm displacement slope spectra reveal a wave number-dependence characteristic of turbulence in the inertial-convective subrange ( math formula) at wavelengths between 0.14 and 100 m. The first event captured by the AUV was transient, and occurred near the edge of a bottom-intensified energetic eddy. Our observations imply that both horizontal advection of warm water and vertical mixing of it into the plume are eddy-driven and go hand in hand in entraining ambient water into the DSO plume. The second event was found to be a stationary feature on the upstream side of a topographic elevation located in the plume pathway. Flow-topography interaction is suggested to drive the intense mixing at this site.

Interannual transport variability of Upper Labrador Sea Water at Flemish Cap

The transport of Upper Labrador Sea Water (ULSW) at Flemish Cap (47°N/45°W) is investigated in the period 1960–2009 using the output from an 8 km resolution numerical ocean model. The average model transport of ULSW decreases southward from 6.7 Sv at 53°N to 4.5 Sv at 45°N due to interior pathways. The largest fraction of the total ULSW volume transport goes around Flemish Cap within the Deep Western Boundary Current (DWBC, 72%) but a significant part goes through Flemish Pass (20%). At interannual timescales, the variability at Flemish Pass shows a distinct behavior when compared to the variability in the DWBC and to the upstream fluctuations. A running correlation method is applied to obtain the connection of the transport variability at Flemish Pass with several quantities, representative for different physical mechanisms: (1) the North Atlantic Oscillation index, (2) the Ekman transport, (3) the rate of ULSW formation in the Labrador Sea, (4) the position of the North Atlantic Current (NAC) relative to the slope and (5) the averaged transport in the subpolar gyre. Weakened or strengthened transport of ULSW through Flemish Pass coincides with changes of the atmospheric forcing or with changes of the NAC‘s position. Strong meandering of the NAC close to the DWBC reduces the transport off Flemish Cap, and the ULSW flow is “redirected” into the Flemish Pass, enhancing the transport there. In contrast, the transport variability in the DWBC is mainly caused by upstream fluctuations and changes according to the rate of ULSW formation.

Variability of Labrador Sea Water transported through Flemish Pass during 1993–2013

Flemish Pass, located at the western subpolar margin, is a passage (sill depth 1200 m) that is constrained by the Grand Banks and the underwater plateau Flemish Cap. In addition to the Deep Western Boundary Current (DWBC) pathway offshore of Flemish Cap, Flemish Pass represents another southward transport pathway for two modes of Labrador Sea Water (LSW), the lightest component of North Atlantic Deep Water carried with the DWBC. This pathway avoids potential stirring regions east of Flemish Cap and deflection into the interior North Atlantic. Ship-based velocity measurements between 2009 and 2013 at 47°N in Flemish Pass and in the DWBC east of Flemish Cap revealed a considerable southward transport of Upper LSW through Flemish Pass (15–27%, −1.0 to −1.5 Sv). About 98% of the denser Deep LSW were carried around Flemish Cap as Flemish Pass is too shallow for considerable transport of Deep LSW. Hydrographic time series from ship-based measurements show a significant warming of 0.3°C/decade and a salinification of 0.03/decade of the Upper LSW in Flemish Pass between 1993 and 2013. Almost identical trends were found for the evolution in the Labrador Sea and in the DWBC east of Flemish Cap. This indicates that the long-term hydrographic variability of Upper LSW in Flemish Pass as well as in the DWBC at 47°N is dominated by changes in the Labrador Sea, which are advected southward. Fifty years of numerical ocean model simulations in Flemish Pass suggest that these trends are part of a multidecadal cycle.