Large-scale circulation in the northern North Atlantic plays a crucial role in the global climate by influencing ocean storage of atmospheric heat and carbon. Temperature and salinity changes in this region can have important consequences on ocean circulation due to density stratification at sites of deep water formation. Such influences can involve feedback mechanisms related to the Atlantic Meridional Overturning Circulation, which has been shown to influence the hydrography of the northern North Atlantic on decadal timescales. Current expectations are that through increasing sea-ice melting, river discharge, an intensifying hydrological cycle and glacial melt anomalies, future climate change could disrupt North Atlantic circulation patterns with cascading effects on carbon cycling and global climate. These interactions were investigated through circulation changes associated with salinity and freshwater variability, as well as variability in temperature and heat content. Recent changes in phytoplankton concentration and biological productivity in the Labrador Sea were also examined as part of this study. Spatial and temporal patterns of salinity in the North Atlantic were examined with the help of objective analysis and reanalysis salinity products using Argo observations of the recent decade (2005 to 2015). An overall freshening trend was evident, but with clear regional differences, particularly between the western subpolar gyre and the central North Atlantic. In general, the western subpolar region exhibited high interannual variability in surface salinity compared to the central North Atlantic. The western subpolar region also revealed a seasonal pattern of salinity fluctuation related to sea ice retreat and accretion, but with some years (i.e., 2008, 2012 and 2015) showing unusually large and negative salinity anomalies which were not present in the central or eastern North Atlantic. To understand the dominant factors influencing salinity and freshwater in the northern North Atlantic, budgets for liquid freshwater content over the northern North Atlantic were derived using a state-of-the-art ocean state estimate (ECCOv4). Here the subpolar North Atlantic (between $\sim$45\oN and the Greenland Scotland ridge) is distinguished from the Nordic Seas (north of the Greenland Scotland ridge). In a separate investigation ECCOv4 was used to describe global ocean heat budgets at varying spatial and temporal resolutions. This analysis showed that anomalies in temperature tendency are driven by atmospheric forcing at short time scales, while advection is the principle term at long time scales. ECCOv4 budget analysis was then used to investigate mechanisms behind interannual freshwater content variability in the northern North Atlantic over the time period 1992-2015. From the mid-1990s to the mid-2000s warming and salinification occurred in the subpolar North Atlantic. Consistent with the upper layer analysis with Argo-observations, ECCOv4 confirmed an overall freshening since about 2005. This freshening occurs simultaneously with an overall cooling in the subpolar North Atlantic. Advective convergence has been identified as the dominant driver of liquid freshwater content and ocean heat content variability in the subpolar North Atlantic, with liquid freshwater and heat content being anti-correlated. Consistent with the global heat analysis in ECCOv4, our results revealed that forcing is only important for establishing anomalies over shorter time scales (i.e., seasonal to interannual), but advective convergence becomes more important at longer (i.e., decadal) scales. Advection is the dominant term due to changes across the southern boundary on the decadal time scale, while exchanges with the Arctic Ocean have minor impact. Changes in freshwater and heat content in the subpolar North Atlantic due to advection occur through anomalies in the circulation itself, and not by the advection of anomalies in either liquid freshwater or heat content. In contrast to the subpolar North Atlantic, in the Nordic Seas interannual changes in liquid freshwater content are predominantly driven by forcing due to sea ice melting, which is in turn strongly correlated with Arctic sea ice export through Fram Strait. The overall concurrent warming and salinification followed by cooling and freshening in the subpolar North Atlantic suggests a relationship with changes in northward heat and salt transport through the Atlantic Meridional Overturning Circulation. This is consistent with decadal variability in deep convection in the Labrador Sea. It is evident that another consequence of changes in the Labrador Sea deep convection is the potential effects on nutrient availability and thus biological productivity. The Labrador Sea has become more productive in recent years, with mean chlorophyll-a concentrations closely correlated with silicate concentrations in the upper waters, which in turn are strongly correlated with wintertime convection depth. Thus annual production in the Labrador Sea appears to be influenced by the extent of deep winter mixing, thereby linking the Atlantic Meridional Overturning Circulation and deep convection to nutrient availability and ocean productivity in the subpolar North Atlantic.

The Atlantic Meridional Overturning Circulation (AMOC) is thought to exert considerable influence over the climate via heat redistribution and carbon storage. Its repeated variations along with the regional and global climate during the last glacial cycle suggest that the state of the AMOC may be roughly divided into “warm,” “cold,” and “off” modes. The three modes correspond to the vigorous deepwater formation in the subpolar North Atlantic, a reduced deepwater formation, and the widespread disruption of the AMOC, respectively. Questions remain about the cause and response of AMOC perturbations in each of the three modes.Reconstruction of the burial flux of ice-rafted debris can resolve questions about the timing and rates of ice sheet calving, which may have been responsible for the “off” mode of the AMOC, given the association of freshwater forcing with AMOC strength. The first chapter quantified the flux of ice-rafted debris in a pair of cores collected from sites in the western North Atlantic. The results show higher ice-rafted debris flux during all Heinrich events and that the western North Atlantic fluxes were higher than the east. The data demonstrate that the Laurentide Ice Sheet played a role in all Heinrich events. A catastrophic last interglacial Laurentide outburst (LILO) event some 125,000 years ago (125 ka) may have contributed to abrupt climate change during the Eemian, when the AMOC was in the “warm” mode. The LILO event was previously proposed to be an analog of the Holocene 8.2 ka event. The second chapter investigated the age and chemical compositions of a layer of red sediments deposited across much of the Northwest Atlantic at 125 ka. The results provide strong support for the occurrence of the LILO event that was analogous to the 8.2 ka event in provenance, timing, and delivery. Little is known about the zonal (east/west) characteristics of the AMOC when in the “cold” mode during the Last Glacial Maximum. Authigenic uranium preserved in sediments is a sensitive redox tracer and can shed light on bottom water oxygen, carbon storage, and water mass distributions. In the third chapter, new and published authigenic uranium data were used to reconstruct deep ocean oxygenation. The compilation shows that lower-than-Holocene oxygen and correspondingly greater respired carbon storage were persistent features of the LGM in the deep North Atlantic. The eastern basin was substantially less well oxygenated than the west. A farther advance and greater infilling in the east of deep waters originating from the Southern Ocean may have caused the zonal difference. Alternatively, deep waters originating from the subpolar North Atlantic may have increased in their residence time in the eastern transect. Questions remain about the flux of freshwater necessary to induce the AMOC to enter the “off” mode. Existing estimates do not agree on the freshwater fluxes associated with Heinrich events. The fourth chapter uses compiled 230Thxs-based mass fluxes in the North Atlantic during the last glacial period to calculate the surge mass fluxes as a measure of the rate of ice-rafted debris deposition. The surge mass fluxes were then converted into freshwater fluxes. Freshwater fluxes for an arbitrarily defined 2000-year period and total freshwater volumes between 20° and 70° N were as high as 0.11 Sv and 6.9 × 1015 m³ during Heinrich event 4 and as low as 0.0012 Sv and 7.6 × 1013 m³ during Heinrich event 3. The relatively low freshwater fluxes we reconstructed for Heinrich events might suggest potentially a high sensitivity of the Atlantic Meridional Overturning Circulation to freshwater perturbations, although the freshwater volumes are in line with previous reconstructions. Our project represents the first time an attempt made to reconstruct the freshwater fluxes and volumes during all Heinrich events of the last glacial period.