Constraining variability of the Atlantic Meridional Overturning Circulation at 25N from historical observations, 1980 to 2005

The Atlantic Meridional Overturning Circulation (MOC) plays an important role inmaintaining western Europe’s moderate climate. Although expected to weaken underincreasing greenhouse gas concentrations, the magnitude of the response remainsuncertain. Monitoring of the Atlantic MOC strength at 25ºN commenced in 2004 as partof the UK Rapid Climate Change programme. Prior to this, detection of such a slowingof the MOC and its associated meridional heat transport was inhibited by the temporallysparse observations of its strength. This thesis aims to extend the time series of AtlanticMOC at 25ºN back to 1980 using historical observations, thus constraining themagnitude of its recent interannual variability.Analysis of transatlantic hydrographic sections at 25ºN show the annual average MOCto have varied in strength by approximately 25% over the last 50 years, with weakeroverturning in 2004 than previously computed. Temporal resolution of the dataset isincreased through decomposition of the Atlantic MOC at this latitude into its Ekmancomponent computed from wind stress climatology, transport of the Florida Currentfrom cable observations and mid-ocean geostrophic transport from CTD or mooredtemperature and pressure derived boundary dynamic height profiles. The resulting timeseries constrain interannual fluctuations of the Ekman, Florida Straits and mid-oceanbaroclinic geostrophic transports at ± 1, ± 2, and ± 3 Sv respectively between 1980 and2005. In addition to interannual fluctuations in strength, evidence is found for a 2 to 3Sv weakening of the MOC between 1980 and 2005. The underlying changes ofincreased mid-ocean southward transport of thermocline waters and decreasedsouthward flow of lower North Atlantic Deep Water are consistent in sign, although ofreduced magnitude, with the transatlantic hydrographic sections observations. Thisthesis presents and discusses these findings along with the relative merits of differentdatasets for detecting such variability