Optimal excitation of AMOC decadal variability: links to the subpolar ocean

Florian Sévellec and Alexey V. Fedorov

The variability of the Atlantic Meridional Overturning Circulation (AMOC) induced by perturbation in Sea Surface Temperature and Salinity (SST and SSS) is studied using a generalized stability analysis in a realistic ocean general circulation model. This method is an extension of the classical linear stability analysis to include transient growth. This powerful method has already been used in the context of the AMOC variability in a wide range of oceanic models, from Primitive Equation model with idealized configuration (flat bottom rectangular basin) to realistic Ocean GCM, as well as coupled GCM. These studies reveal two robust properties: (i) the roughly decadal timescale of the optimal transient growth and (ii) the north of the North Atlantic location of the optimal perturbation. However, the physical mechanism behind this transient growth is still in debate. In this study, the generalized stability analysis is derived from a maximization procedure based on Lagrangian multipliers in a non-autonomous context. This more general approach allows us to include constraint on the perturbations, unlike the typical generalized stability analysis, which is often described as the growth of perturbations (measure with a norm).
In this study, we obtained the optimal SST and SSS perturbations for two different measures of the AMOC (the meridional volume and heat transports) and for different surface boundary conditions (the flux and mixed boundary conditions). The structure of the optimal perturbations is characterized by anomalies in temperature or salinity localized in the northern Atlantic off the east coasts of Greenland and Canada centered south of the Denmark Strait. The maximum impact of the perturbations on the AMOC is reached after 7 to 9 yr. These are robust results independent on the type of perturbations, the measures used, the boundary conditions, or other additional constraints. The AMOC transient growth involves the following mechanism: after the initial positive surface perturbation of density reaches the deep ocean, it generates a cyclonic geostrophic flow that extracts a temperature anomaly (with a zonal gradient) from the mean temperature field (with a strong meridional gradient). In turn, the anomalous zonal temperature gradient induces, by thermal wind balance, a northward flow in the upper ocean and a southward flow in the deep ocean, thus strengthening the AMOC. The optimal perturbations not only generate the transient growth of the AMOC, but also excite a damped oscillatory eigenmode in the system with a period of about 24 yr. This oceanic mode corresponds to a large-scale baroclinic Rossby wave, transporting temperature and salinity anomaly westward and modifying the AMOC and the heat content of the North Atlantic. An idealized model, formulated to investigate this mechanism, highlights the nonnormality of the dynamics leading to the transient change including the role of deep ocean and convection. Simple estimates show that realistic changes in salinity or temperature in the upper ocean (such as changes due to the Great Salinity Anomaly) can induce AMOC variations via this mechanism on the order of several Sverdrups, or 10-20% of the mean meridional overturning.