Atmosphere Ocean Science Colloquium
Ocean-atmosphere dynamics of decadal climate variability
Speaker: Robert Jnglin Wills, University of Washington
Date: Wednesday, April 21, 2021, 3:30 p.m.
Decadal sea-surface temperatures (SST) variability, such as Atlantic Multidecadal Variability (AMV) and Pacific Decadal Variability (PDV), has wide ranging impacts on regional temperature and rainfall, sea ice extent, global temperature, and radiative feedbacks. However, the physical mechanisms of this variability remain unclear: some studies suggest that this variability can be explained by the persistence of atmospherically forced temperature anomalies in the ocean mixed layer; others suggest that ocean circulation changes play an important dynamical role. To clarify the mechanisms of decadal climate variability, we have developed a novel statistical method called low-frequency component analysis (LFCA), which identifies the slowest changing patterns of climate anomalies, and we use it to identify patterns of decadal variability in Atlantic and Pacific SSTs in observations and climate models. This allows clearer insight into the mechanisms of decadal variability by separating it from higher frequency variability and from externally forced climate change. We find that ocean circulation changes are required to explain the relationship between decadal SST anomalies and air-sea heat fluxes and identify the relevant ocean-atmosphere dynamics. We show that anomalies in the Atlantic Meridional Overturning Circulation (AMOC) and North Pacific subpolar gyre represent integrated responses to stochastic buoyancy and wind forcing and establish their fundamental roles in the AMV and PDV, respectively. While these modes of variability show power spectra consistent with red noise, the reddening occurs due to ocean dynamics at longer timescales than would be expected from mixed-layer thermal inertia, leading to persistent SST anomalies that can be utilized for prediction on multi-year timescales. The atmospheric response to these persistent SST anomalies remains an important source of uncertainty, and I will briefly discuss work in progress to address this uncertainty.