Monitoring the integrated deep meridional flow in the tropical North Atlantic

Meridional transport of heat is accomplished by fundamentally dierent mechanisms in the atmosphere and the ocean. While in the atmosphere eddies exhibit a dominant role, the largest fraction of northern hemisphere poleward heat transport in the ocean is related to the Atlantic meridional overturning circulation (MOC). The evolution of the MOC and its impact on climate have been subject to intensive theoretical and numerical studies, however continuous measurements of MOC variability have not been carried out. In this study results from an observational pilot project to monitor uctuations of the deep southward branch of the MOC across a latitude circle in the tropical North Atlantic are presented. Within the framework of the Meridional Overturning Variability Experiment (MOVE) a four year long time series of deep meridional volume transport uctuations has been recorded. The backbone of the experiment design is an end point measurement method, which makes use of the fact that the deep ocean ow eld is to rst order in geostrophic balance: Fluctuations of deep zonally integrated meridional transports in the western trough1 of the Atlantic are ef- ciently monitored by continuous moored measurements of the evolution of the zonal density and bottom pressure dierence between the eastern and western end point of the section. One main aspect of this study comprises data calibration and processing as well as a thorough technical performance assessment of the dierent measurement components of the monitoring array. It is found that two components (density and current meter measurements) provide robust estimates of transport uctuations. As a consequence of sensor characteristics and data processing the third element (bottom pressure) is found to suppress low frequency variability. Simulations suggest that changes in the deployment scheme might help to overcome these problems to a large extent. Bottom pressure uctuations derived from space-borne gravity eld measurements at 16N deviate substantially from the in-situ observations and thus do not provide robust estimates of the evolution of deep transports. For the interpretation of the observed mean and time variable velocities and volume transports and the verication of the monitoring design comparisons to independent observational data and numerical model output have been carried out and spectral analysis as well as basic theoretical aspects of uid dynamics have been applied. Since only the western trough of the Atlantic is covered by the array, westward propagating Rossby waves from the eastern trough represent a major source of noise, which may mask the MOC signal. An extension of the zonal integration scale from the western boundary from 400 to 1000 km leads to a substantial suppression of the wave signal, thus conrming the monitoring strategy. The best estimate of mean southward transport of North Atlantic Deep Water is 14.9 +/- 3.0 Sv, its inter-annual variability amounts to 2.4 Sv. A verication of the experiment design using model simulations attest the transport signal observed by MOVE to be moderately representative for MOC and meridional heat transport uctuations at 16N at inter-annual time scales. An eastward extension of the array into the eastern trough might lead to a drastic increase in the signal-to-noise ratio. However it is found that only at longer than decadal time scales coherent MOC uctuation over the entire meridional extent of the Atlantic can be found. To separate locally and remotely forced MOC uctuations on shorter time scales, it is suggested to operate two end point monitoring systems simultaneously at dierent latitude circles of the North Atlantic. Additional monitoring elements specically designed to quantify the impact of dierent mechanisms responsible for MOC uctuations should be added.