Timescales and regions of the sensitivity of Atlantic meridional volume and heat transport: Toward observing system design

A dual (adjoint) model is used to explore elements of the oceanic state influencing the meridional volume and heat transports (MVT and MHT) in the sub-tropical North Atlantic so as to understand their variability and to provide the elements of useful observational program design. Focus is on the effect of temperature (and salinity) perturbations. On short timescales (months), as expected, the greatest sensitivities are to local disturbances, but as the timescales extend back to a decade and longer, the region of influence expands to occupy much of the Atlantic basin and significant areas of the global ocean, although the influence of any specific point or small area tends to be quite weak. The propagation of information in the dual solution is a clear manifestation of oceanic teleconnections. It takes place through identifiable “dual” Kelvin, Rossby, and continental shelf-waves with an interpretable physics, in particular in terms of dual expressions of barotropic and baroclinic adjustment processes. Among the notable features are the relatively fast timescales of influence (albeit weak in amplitude) between 26°N and the tropical Pacific and Indian Ocean, the absence of dominance of the sub-polar North Atlantic, significant connections to the Agulhas leakage region in the southeast Atlantic on timescales of 5–10 years, and the marked sensitivity propagation of Doppler-shifted Rossby waves in the Southern Ocean on timescales of a decade and beyond. Regional, as well as time-dependent, differences between MVT and MHT sensitivities highlight the lack of a simple correspondence between their variability. Some implications for observing systems for the purpose of climate science are discussed.National Oceanographic Partnership Program (U.S.) (‘‘Estimating the Circulation and Climate of the Ocean’’ (ECCO) and the ‘‘Atlantic MOC Observing System Studies Using Adjoint Models’’ projects)National Science Foundation (U.S.) (NSF Collaboration in Mathematics and Geoscience (CMG) project ‘‘Uncertainty Quanti- fication in Geophysical State Estimation’’