Variability in formation, properties, and transport of North Atlantic Deep Water
North Atlantic Deep Water is found in much of the deep Atlantic Ocean, and its formation
in the Labrador and Nordic Seas and subsequent southward export are a vital part of global
ocean circulation and Earth’s climate system. The overarching goal of this dissertation is to
better understand the processes controlling variability of North Atlantic Deep Water formation,
properties, and transport in the Atlantic Ocean.
Chapter 1 uses data from the central Labrador Sea during winter to estimate the uptake of oxygen
associated with deep convection in 2014–15. The results show that intense air-sea exchange
results in an uptake of 29.1 ± 3.8 mol m^−2 during the convective season, with much of the flux
being associated with injection of air bubbles. Chapter 2 looks at lateral fluxes of carbon, oxygen,
and nitrate from the Labrador Sea’s boundary current into the center of the basin during the
summertime productive season. Lateral fluxes are found to play an important role for the carbon
and nitrate budgets immediately below the mixed layer, with respiration rates underestimated by
up to 50% if they are ignored.
In chapter 3, gravity measurements from satellites are used to investigate variability in ocean
circulation. After trends in the data are validated using independent measurements, they are used
to study decadal circulation changes of North Atlantic Deep Water in the North Atlantic Ocean.
The analysis reveals a strengthening of the interior branch of North Atlantic Deep Water flow,
with transport increasing by 13.9 ± 3.7 Sv (1 Sv = 10^6 m^3 s^−1 ), balanced by a weaker southward
flow in the Deep Western Boundary Current.
A twenty-year record of mooring data is analyzed in chapter 4 to investigate changes in North
Atlantic Deep Water transport at 16 ◦ N. Multi-decadal variability is observed in the transport time
series, and is largely associated with density changes in the lower half of the North Atlantic Deep
Water layer, which in turn appear to be caused by changes in the source region. The data are also
compared to another transport time series at 26 ◦ N, and similarities and differences are discussed.