Halifax: 19th - 21st October, 2015

Ventilation, Interactions and Transports Across the Labrador Sea

Historical Analysis


  1. To use existing measurements, notably those collected as part of DFO's monitoring programs, to place the observations collected during our process study into a broader spatial and temporal context and document natural variability.
  2. To provide data for comparison with numerical simulating and testing the physical and biogeochemical models.
  3. To document decadal variability in water mass properties (and oxygen availability) from proxy-records in corals and sediments.


Incorporating historical observations of the Labrador Sea into VITALS will extend the time frame and spatial scope of the project. This research axis is important given the range of ocean and climate variations in the northwest North Atlantic observed at inter-annual to decadal time scales (e.g. van Aken et al. 2011; Yashayaev and Loder 2009) and reconstructed at centennial time scales from archival data (e.g. Andresen et al. 2012). The sedimentary archives show significant variations of temperatures, salinity and stratification, thus illustrating that modern deep convection is not a stable feature of the Labrador Sea (e.g., Hillaire-Marcel et al. 2001; de Vernal and Hillaire-Marcel 2006). Data indicate large changes in productivity, carbon fluxes and depth of the lysocline (e.g. de Vernal et al. 1992; Radi and de Vernal 2008), which further illustrate the need for accurate physical and biogeochemical models.

Different types of time-series data will be utilized to meet the VITAL objectives. Several observational data sets have been under-utilized and show great potential for testing hypotheses about Labrador Sea circulation, deep water formation, and biogeochemistry. In addition, sedimentary and coral archives provide a means to develop centennial scales time series with annual (coral) to decadal (sediment) resolutions for the documenting of natural variability and extremes. This work leverages enormous past investments in data collection to shed light on problems relevant to the CCAR program goals, namely understanding the ocean component of the Earth System and recent changes in the cold weather environment of the sub-polar North Atlantic. The major expenditures in this activity will be in the training of HQP in addition to some analytical costs.

The historical analysis will revolve around the evaluation of three hypotheses, each related to deep convection and the setting of physical and biogeochemical conditions in the Labrador Sea:

  1. Deep water formation is influenced by the transport of heat and freshwater from the boundary currents. Deep convection consequently feeds back on the circulation of the Labrador Sea.
  2. The Labrador Sea carbon budget varies in response to the formation of Labrador Sea Water (LSW) and the transport of water masses into the basin.
  3. The influence of different water masses entering the Labrador Sea is manifest in changing water column nutrient ratios, biological communities, and the depth and rate of particle export, as outlined below. These will also provide important benchmarks for comparison with the model output.


Direct velocity observations and altimetry: (Palter, Yashayaev, Ruddick): Profiling floats were first deployed in large numbers in the early 1990s as part of the Labrador Deep Sea Convection Experiment and Atlantic Circulation and Climate Experiment, with the total number of functional floats peaking in the late 1990s at about 200. Since 2010, the total number of Argo floats in the North Atlantic has matched the numbers from the 1990s. Thus, there exists a largely untapped opportunity to compare the properties and circulation between the two periods, as was done for the West Greenland Current system by Rykova et al. (2009). Such an analysis literally adds depth to recent evidence from surface (i.e. 15 m) drifters and satellite altimetry of a transition between a regime of suppressed subtropical-subpolar exchange in the 1990s to one of more vigorous exchange in the 2000s (Hakkinen and Rhines 2009).

In our preliminary analysis, maps of deep currents show the same qualitative features as the earlier float deployments, but with possible quantitative differences, such as a slowing of the boundary currents from the Reykjanes Ridge to the Grand Banks and a weakening of the interior anticyclonic recirculation gyre—a feature discovered only about a decade ago by Lavender et al. (2000). In addition to velocities derived from float displacements, we have access to several years of direct velocity measurements from ADCP instruments on shelf and interior moorings. Altimetry data gives complimentary information to the in situ sensors by allowing the calculation of geostrophic surface velocities since 1995. A synthesis of all such velocity work will help us evaluate Hypothesis 1, particularly the feedback between the strength of convection and the circulation. We will analyze the velocity records for correlations between the depth of convection and mean velocities and EKE, and evaluate variability in the presence/strength of the Labrador Sea’s anticyclonic interior recirculation.

Historical hydrographic stations and sections (Myers, Yashayev, Tremblay, Palter, Azetsu-Scott): BIO has been conducting an annual spring survey of the LS for the past two decades, as part of its ocean climate monitoring program (Yashayaev and Loder 2009). Systematic high-quality physical, chemical, and biological observations are collected along the AR7W line (initiated during the World Ocean Circulation Experiment) extending from the Labrador Shelf to the Greenland Shelf and from the sea surface to the sea floor. In addition, stations along the shelf have been occupied sporadically as far back as the 1940s, with systematic annual surveys, occupying sections across the Labrador and Newfoundland shelves since 1999, in DFO's AZMP. The proposed work will synthesize this data to help evaluate all three hypotheses, as further described below.

Historical data will be used to test the link in Hypothesis 1 between variability in freshwater transport and Labrador Sea Water formation. All synoptic sections will be used to calculate geostrophic velocities with reference velocities at the surface provided from altimetry or at depth from in situ sensors when possible. The section data and velocities will be combined to compute volume, heat, and freshwater transports, to expand on previous efforts to understand the variability of the currents bounding the Labrador Sea (e.g. Myers et al. 2009; Myers and Kulan, 2012). A focus will be on understanding the historical pathways of freshwater (connected to analyses of historical ?18O) to the Labrador Sea and their variability, and their links to stratification and convection. This historical analysis will also be used to validate the modelling hind-cast studies to ensure the model is able to simulate the evolution of the freshwater routes in the future under different climate change scenarios.

The nutrient, carbon and oxygen data collected on dozens of cruises in the last two decades will help us quantify aspects of the carbon budget, and test the link in Hypothesis 2 between the depth of convection and carbon storage. Hydrographic data collected in the interior LS and its shelves have yielded immense insight into the physical and spatial variability in the LS (e.g. Fratantoni and McCartney, 2010, Yashayaev and Loder 2009), but the chemical data have received considerably less attention. For example, no comprehensive analysis of the nutrient or carbon data collected along AR7W has yet been performed. We propose to use this data to evaluate inter-annual variability in the amount of biological carbon stored in the LSW. Our expectation is that for years during which LSW is not formed (or formed in a lighter density class at a slower rate), biological carbon will accumulate in the vestigial water mass, providing a short-term sink of carbon. Conversely, for years during which LSW is vigorously formed, greater deep mixing will lead to sinking with a larger amount of preformed nutrients, reflecting an inefficiency in the biological pump and a short-term decrease in ocean biological carbon storage. Though a year in which LSW formation is slowed or stopped should result in an increase in biological carbon storage, we hypothesize that such an effect will be partly compensated for by a reduction in the amount of carbon dissolved in the water column by physical processes. CO2 is more efficiently dissolved in water at low temperatures and high wind speeds, conditions which favour for LSW formation. Using the historical data, we plan to quantify the degree of compensation between these processes.

Finally, the historical archives will be helpful in evaluating Hypothesis 3, as nutrients provide useful tracers of water masses and influence biological communities, which, in turn, influence the depth of particle export and the strength of the biological carbon pump. Unexplained variability in nutrient ratios in the LSW layer has been observed far downstream of the Labrador Sea (Deutsch and Weber 2012). The proposed historical analysis will allow us to evaluate whether this variability can be set locally in the Labrador Sea or is a consequence of settling particles with variable N:P ratios along the LSW export pathways. Biological processes such as denitrification and nitrogen fixation are thought to be too slow to have a measurable impact locally in the interior Labrador Sea, but rather reflect the large-scale connectivity following the exposure of imported water masses to shallow shelves. In order to trace the influence of imported water masses on nutrient concentrations, we will compare the interior Labrador Sea nutrient field from AR7W sections with other historical data (e.g. ArcticNet and AZMP) upstream and downstream of the Labrador Sea. Particles formed higher up in Baffin Bay tend to have low Nitrogen:Phosphorus ratios, which we expect to influence nutrient concentrations in the Labrador Sea. Corals and sediments: (de Vernal, Edinger, Francois, Hillaire-Marcel, Kienast): The coral and sedimentary data archives will be used for the characterization of water mass properties on the shelf and of the interior Labrador Sea during different time intervals.

Covering the last century and the Medieval Warm Period (MWP, 10th-13th centuries), corals will permit us to resolve annual-decadal variations of ambient water properties from their isotopic and elemental contents. The occurrence of corals in Hudson Strait and along the Labrador Margins, from the slope to the rise, notably in Orphan Knoll and Flemish Pass areas, offer the possibility to document the chemical and physical variability of water masses and currents, the decadal variability of convection depth in the Labrador Sea, and changes in alkalinity and productivity. On longer time scales, the stability of deep convection in the Labrador Sea will be examined in relationship with different climate states, notably warm optimum of the recent past. Given that formation of deep-intermediate waters in the Labrador Sea started only after the thermal optimum of the present interglacial (e.g. de Vernal and Hillaire-Marcel 2006), we hypothesize that convection in the Labrador Sea will not survive climate warming unless there is a significant reduction of freshwater export through the Canadian Archipelago and Baffin Bay. Based on time series created from the analyses of sedimentary cores from the Labrador shelf to the deep sea, we aim to document the relationship between freshwater fluxes and LSW formation.

Finally, looking at the last interglacial (LIG), which represents the latest period with a climate as warm as the one projected for the next century in most IPCC models, analyses will be conducted to reconstruct Labrador Sea “breathing” under the “warm extreme” scenario. Most of the material (corals and sedimentary cores) is already available from several expeditions previously conducted in the Labrador Sea. Corals from Orphan Knoll and Flemish Cap were collected from the CCGS Hudson in 2010. Additionally data will be provided by two independent cruises (not part of VITALS) that are planned with respect to the sampling of deep coral communities (ROPOS 2013) and surface sediments (Polarstern 2016). The analytical work to reconstruct water mass properties from the skeletons of corals will be based on isotopic analyses for dating (210Pb, 226Ra, Th/U, 14C; e.g., Hillaire-Marcel 2009) to evaluate temperature and salinity (?18O, Mg/Ca), water mass origin (Nd), and alkalinity (11B) (Rollion-Bard et al. 2011). Cores from key shelf and deep LS sites with suitable sediment accumulation rates to resolve recent (~100 yrs), Holocene and LIG time series are available from international expeditions of the Integrated Ocean Drilling Program (IODP), and IMAGES (International Marine Global Change Studies) in addition to Canadian expeditions, most of them archived at the Atlantic Geoscience Center. Complementary cores will be collected at targeted sites along the Labrador margin during the planned 2015 Polarstern cruise.

The methodological approach will combine isotopic analyses, micropaleontological and sedimentological studies for the reconstruction of surface, intermediate and deep water conditions. Microfossil assemblages (mostly dinocysts) and biomarkers (alkenones, IP25) will be used to reconstruct sea-surface temperature, salinity, sea-ice and productivity, the calcareous shells of foraminiferal taxa representing different depths in the water column will be analyzed for their isotopic (?18O, ?13C) content in order to reconstruct density and water mass stratification for assessing convection. Excess 231Pa/230Th in sediments, decay-corrected to the time of deposition, will also be used to reconstruct past changes in deep convection (e.g. Luo et al. 2010). Calcium carbonate dissolution and alkalinity will be examined from foraminifer shell preservation (using scanning electron microscope observation and shell density measurements) and 11B analyses of deep corals. Ba/Ca data in foraminiferal calcite and coral aragonite will inform on productivity changes, whereas Nd in corals and foraminifer shells (Copard et al. 2012) and oxide coatings in sediment (e.g. Rutberg et al. 2000) will inform on water mass origins, distinguishing, in particular between waters originating from Baffin Bay and those reaching the Labrador Sea through Denmark Strait. Moreover, Nd and mineralogical data of the lithogenic fraction of the sediments will serve to constrain deep circulation patterns. These will be complemented by the analysis of siliciclastic grain size distributions (sortable silt parameter) as proxy of the degree to which deep-sea sediments have been reworked by currents, i.e. as paleo-current meters. The overall analyses, which will include elemental (Corg, CaCO3, N) measurements will also permit us to construct a carbon budget for the LS and to propose balances in terms of export rates.


Synthesis of several decades of observational data, with a focus on velocities, transports and freshwater pathways. Characterization of the relationship between convection and deep Labrador Sea biogeochemical properties. Records of oxygen availability and organic carbon fluxes on the Labrador Shelf going back through at least the last century. Compilation of proxy data for further comparison with model simulations. Assessment on the natural variability of the deep water convection in the Labrador Sea.