Determining Atlantic Overturning Circulation Variability from Observations at the Ocean Surface

Climate models demonstrate that the interdecadal variability of the Atlantic Meridional Overturning Circulation (AMOC), also known as the ocean conveyor, exerts an important control on European climate. This is due to the northward transport of large but variable, amounts of warm water near the ocean surface and the southward return of cooler water at depth. In addition to natural interdecadal variability, most of the models used in the latest IPCC assessments projected a substantial decline of the AMOC strength by 2100 in response to increasing greenhouse gases. The resultant effect on European climate of changes in the AMOC, motivate the need for constant real-time monitoring of its variability. Although monitoring of the AMOC, by the RAPID mooring array has taken place at 26N since 2004, the system suffers from two limitations that the proposed research will address. First, recent model results indicate the AMOC variability at 26N is quite poorly correlated to latitudes that are closer to and have a stronger influence on the British Isles. The array thus provides little insight into the AMOC variability at mid-high latitudes. The second limitation is that the brevity of the 26N time series means that the interannual variability cannot be placed in terms of the longer and possible stronger decadal variability as well any longer term trends. The proposal will address this issue by testing the viability of a method for monitoring the AMOC between 40 and 60N. Because the proposed method requires only observations at the ocean surface, it will also be possible to place any recent changes in AMOC strength in the context of a 50-year time series. Recent results from the PI and collaborators suggest that between 40N and 60N, the AMOC can be estimated from surface observations by an adaptation of the water mass transformation method. The theoretical basis for the method involves some simplifications and assumptions. Probably, as a consequence of these, the method at present has only been able to account for about 40% of the variability of the AMOC. In this research, we will assess the importance of the assumptions and see if by relaxing them, the method can be improved to the extent where it is reliably accounting for the majority of the variability in the AMOC. This part of the work will be carried out by analysing output from a high-resolution global ocean model that provides a realistic depiction of the North Atlantic. Using the model output it will be possible to calculate the AMOC estimated from the water transformation method and the actual AMOC between 40N and 60N. In addition we will be able to make the additional calculations necessary to assess the previous assumptions, a) that the z-coordinate system is appropriate for the method, b) that the water masses are in steady state and c) that the wind driven (Ekman) transports could be ignored. The goal of this analysis will be to determine the best version of the water mass transformation method for reliable monitoring of the mid-high latitude AMOC that theoretically can be achieved using real-time observations from the ocean surface. The extent to which the method can be extended to sub-tropical latitudes will also be investigated. Having determined the maximum amount of the AMOC variability that is theoretically possible to explain with the method, the second part of the research will assess the extent that this can be achieved in practice, given the present day capabilities in sustained ocean surface observations. Therefore using the best version of the water mass transformation method, the mid-latitude AMOC from 1958-present will be reconstructed using a suite of state-of-the-art air-sea flux data sets. Using the resultant ensemble of estimates will enable the PI to determine if there are common trends or variations in the AMOC across the range of flux products and will allow for a lower bound of uncertainty in the estimates to be calculated.