A151-0017
Time-Varying Probability Distributions of Southern Ocean Surface Winds: Leading Empirical Modes Linked to SAM and the Annual Cycle
Abstract:
In this study, time-varying pentad probability distributions from ERA5 and a suite of scatterometer products of SO surface winds and stress are used in a singular value decomposition to derive a linearly independent, empirical, reduced-order basis. The leading order modes of wind and stress (72% of the total variance) are highly correlated with a standard SAM index (r =0.82) and show how SAM drives extra-tropical cyclone intensity and, in turn, extreme winds. Southerly and less westerly winds are more frequent during short and distinct negative SAM phases, and strong westerly winds that occur during positive SAM are insufficient to drive mixed-layer ventilation. The second mode (16% of the total variance) describes seasonal changes in the wind variance that are uncorrelated with the first mode.
The strong link between leading modes of atmospheric variability and surface stress suggests that empirical modes of surface distributions serve as a novel pathway to study atmosphere-ocean interaction across scales and to link extreme short-term events to their large-scale drivers.