On the Relative Roles of the Atmosphere and Ocean in the Atlantic Multidecadal Variability

Oluwayemi A. Garuba,Jian Lu,Hansi A. Singh,Fukai Liu,Phil Rasch
DOI: https://doi.org/10.1029/2018gl078882
IF: 5.2
2018-01-01
Geophysical Research Letters
Abstract:The relative roles of the ocean and atmosphere in driving the Atlantic multidecadal variability (AMV) are investigated by isolating anomalous sea surface temperature (SST) components forced by anomalous surface heat fluxes and ocean dynamics in fully and partially coupled simulations. The impact of the ocean dynamics-forced SST on air-sea interaction is disabled in the partially coupled simulation in order to isolate the atmosphere-forced variability. The atmosphere-forced AMV component shows weak but significant variability on interdecadal timescales (10- to 30-year periods), while the ocean-forced component exhibits a strong multidecadal variability (25- to 50-year periods). When coupled to the atmosphere, this ocean-forced variability weakens and is imprinted on the coupled surface heat fluxes that further act to damp the ocean-forced SST variability, causing a much weaker fully coupled AMV. Our results suggest that the AMV is largely driven by ocean circulation variability, but its power is also affected by the strength of air-sea coupling. Plain Language Summary The mechanism of the Atlantic multidecadal variability (AMV) has recently been debated. We investigate the relative roles of the ocean and atmosphere in this variability by applying an ocean temperature decomposition and partial coupling method to isolate the components of the AMV driven by the atmosphere and the ocean in a fully coupled model. We show that the ocean drives stronger multidecadal variability (greater than 30years), while the atmosphere drives weaker variability up to interdecadal timescales (10-30years). Surface interactions between the atmosphere and ocean decrease the ocean-driven variability, as surface heat fluxes act to damp the ocean anomalies. Our results imply that the strength of the surface coupling in global climate models might be one of the reasons why they simulate weaker AMV compared to what is found in observations.
What problem does this paper attempt to address?