The Atlantic Meridional Overturning Circulation (AMOC) consists of an upper limb with net northward transport between the surface and approximately 1200 m depth, and a lower limb of denser, colder, fresher waters returning southward between 1200 m and 5000 m. The AMOC is responsible for most of the meridional transport of heat and carbon by the mid-latitude NH ocean and associated with the produc- tion of about half of the global ocean’s deep waters in the northern North Atlantic. Coupled climate models find that a slowdown of the AMOC in the next decades is very likely, though with uncertain magni- tude (Section 18.104.22.168). Observations of the AMOC are directed toward detecting possible long-term changes in its amplitude, its northward energy transport, and in the ocean’s capacity to absorb excess heat and greenhouse gases, as well as characterizing short-term variability and its relationship to changes in forcing.
Presently, variability in the full AMOC and meridional heat flux are being estimated on the basis of direct observations at 26.5°N by the RAPID/MOCHA array (Cunningham et al., 2007; Kanzow et al., 2007; Johns et al., 2011). The array showed a mean AMOC magnitude of 18 ± 1.0 Sv (±1 standard deviation of annual means) between April 2004 and April 2009, with 10-day values ranging from 3 to 32 Sv (McCarthy et al., 2012). Earlier estimates of AMOC strength from five shipboard expeditions over 47 years at 24°N (Bryden et al., 2005) were in the range of variability seen by RAPID/MOCHA. For the 1-year period 1 April 2009 to 31 March 2010, the AMOC mean strength decreased to 12.8 Sv. This decrease was manifest in a shift of southward interi- or transport from the deep layers to the upper 1000 m. Although the AMOC weakening in 2009/2010 was large, it subsequently rebounded and with the large year-to-year changes no trend is detected in the Observations targeting one limb of the AMOC include Willis (2010) at 41°N combining velocities from Argo drift trajectories, Argo tempera- ture/salinity profiles, and satellite altimeter data (Figure 3.11b). Here the upper limb AMOC magnitude is 15.5 Sv ± 2.4 from 2002 to 2009 (Figure 3.11b). This study suggests an increase in the AMOC strength by about 2.6 Sv from 1993 to 2010, though with low confidence because it is based on SSH alone in the pre-Argo interval of 1993–2001. At 16°N, geostrophic array-based estimates of the southward transport of the AMOC’s lower limb, in the depth range 1100 to 4700 m, have been made continuously since 2000 (Kanzow et al., 2008). These are the longest continuous measurements of the southward flow of NADW in the western basin. Whereas the period 2000 to mid-2009 suggested a downward trend (Send et al., 2011), the updated time series (Figure 3.11b) has no apparent trend. In the South Atlantic at 35°S, estimates of the AMOC upper limb were made using 27 high-resolution XBT tran- sects (2002–2011) and Argo float data (Garzoli et al., 2013). The upper- limb AMOC magnitude was 18.1 Sv ± 2.3 (1 standard deviation based on cruise values), consistent with the NH estimates.
The continuous AMOC estimates at 16°N, 26.5°N and 41°N have time series of length 11, 7, and 9 years respectively (Figure 3.11b). All show a substantial variability of ~3 to 5 Sv for 3-month low-pass time series, with a peak-to-peak interannual variability of 5 Sv. The short- ness of these time series and the relatively large interannual variability emerging in them suggests that trend estimates be treated cautiously, and no trends are seen at 95% confidence in any of the time series. Continuous time series of AMOC components, longer than those of the complete system at 26.5°N, have been obtained using moored instrumentation. These include the inflow into the Arctic through Fram Strait (since 1997, Schauer and Beszczynska-Möller, 2009) and through the Barents Sea (since 1997, Ingvaldsen et al., 2004; Mauritzen et al., 2011), dense inflows across sills between Greenland and Scotland (since 1999 and 1995 respectively, Olsen et al., 2008; Jochumsen et al., 2012) and North Atlantic Deep Water carried southward within the Deep Western Boundary Current at 53°N (since 1997, Fischer et al., 2010) and at 39°N (Line W, since 2004, Toole et al., 2011). The longest time series of observations of ocean transport in the world (dropsonde and cable voltage measurements in the Florida Straits), extend from the mid-1960s to the present (Meinen et al., 2010), with small decad- al variability of about 1 Sv and no evidence of a multi-decadal trend (Figure 3.11a). Similarly, none of the other direct, continuous transport estimates of single components of the AMOC exhibit long-term trends at 95% significance.
Indirect estimates of the annual average AMOC strength and variability can be made (Grist et al., 2009; Josey et al., 2009) from diapycnal trans- ports driven by air–sea fluxes (NCEP-NCAR reanalysis fields from 1960 to 2007) or by inverse techniques (Lumpkin and Speer, 2007). Decadal fluctuations of up to 2 Sv are seen, but no trend. Consistent with Grist et al. (2009), the sea level index of the strength of the AMOC, based on several coherent western boundary tide gauge records between 39°N and 43°N at the American coast (Bingham and Hughes, 2009) shows no long-term trend from 1960 to 2007.
In summary, measurements of the AMOC and of circulation elements contributing to it, at various latitudes and covering different time periods, agree that the range of interannual variability is 5 Sv (Figure 3.11b). These estimates do not have trends, in either the subtropical or the subpolar gyre. However, the observational record of AMOC varia- bility is short, and there is insufficient evidence to support a finding of change in the transport of the AMOC. updated time-series (Figure 3.11b).