Intrinsic ocean variability partly randomizes the mean seasonal cycle of sea level 

Oceanic nonlinearities drive random intrinsic sea level variations over the global ocean, which locally compete with forced sea level variations that are paced by atmospheric and astronomical drivers. This study utilizes a global ocean/sea-ice 50-member ensemble simulation to characterize the sea level mean seasonal cycle (computed over 1993-2015) and partition its forced and intrinsic components. The model faithfully represents many features of the observed sea level mean seasonal cycle. We show that the mean seasonal cycle of sea level is most stochastic in the Southern Ocean, in western boundary currents, and along +/-20° latitudes, and remains partly random up to 10°x10° scales in these regions. Forced and intrinsic components mostly have a steric origin but with deeper signals involved for the intrinsic term. Our study thus demonstrates that ocean nonlinearities give a marked stochastic flavor to the sea level seasonal cycle averaged over 23 years and illustrates the usefulness of eddying ocean ensemble simulations for adequately interpreting observations.