The central role of the Atlantic meridional overturning circulation in the Bjerknes compensation

Poleward heat (energy) transport plays a major role in shaping the Earth's climate. Its oceanic and atmospheric components carry heat from low to high latitudes thus reducing the equator-to-pole temperature contrast. In quasi-equilibrium climate states, changes in the top-of-the-atmosphere (TOA) energy fluxes and ocean heat content remain small. In such conditions, anomalies in oceanic and atmospheric heat transport must have the same magnitude but opposite signs. This phenomenon is known as the Bjerknes compensation (BJC). The BJC hypothesis is of high importance in climate, since it imposes a strong constraint on climate variability at sufficiently long timescales (on sub-decadal and shorter timescales TOA flux variations may become comparable to other heat budget terms, interfering with BJC). However, to which extent BJC operates in the climate system and the key mechanisms of the compensation remain poorly understood. Here we analyze BJC in the IPSL-CM6A-LR climate model, focusing on its timescale dependence, its links to the Atlantic Meridional Overturning Circulation (AMOC), and the connection to Intertropical Convergence Zone (ITCZ) shifts. We show that BJC occurs in the model at both multi-decadal and centennial timescales, but is stronger on centennial timescales than decadal. In both cases BJC is initiated by variations in ocean heat transport induced by AMOC variability that are partially or fully compensated by atmospheric heat transport. For decadal timescales, we find two regions of a strong BJC at latitudes associated with the storm track region and the marginal ice zone in the Northern Hemisphere. Finally, on centennial timescales we observe a Bjerknes-like interbasin compensation between the Atlantic and Indo-Pacific heat transports, which is also related to strong centennial AMOC fluctuations and involves Southern Ocean zonal heat transport.