| Summary: | Variability in the strength of the Northern Hemisphere stratospheric polar vortex is an important climate feature. Strong vortex conditions, as well as disruptions of the vortex, known as sudden stratospheric warmings (SSW), are associated with significant tropospheric and surface variations. This makes them an important feature for improving the skill of seasonal weather forecasts. This thesis addresses some ongoing research questions regarding the nature of variations in the polar vortex and their interactions with the wider climate system. First, multi-decadal variability of SSW events is examined in a 1000-yr pre-industrial simulation of a coupled global climate model. A wavelet spectral decomposition method shows that hiatus events (intervals of 5 years or more with no SSWs) and consecutive SSW events (extended intervals with at least one SSW in each year) vary on multi-decadal timescales with the most persistent spectral feature appearing at a period of 90 years. This long-term SSW variability is found to be associated with similar variations in amplitude and vertical coherence of the stratospheric quasi-biennial oscillation (QBO), likely via the well-known Holton-Tan link where the QBO has an in-season influence over the vortex. Interactions between multi-decadal variability in vortex strength and modes of surface and ocean variability are subsequently examined. Intervals that exhibit persistent anomalous vortex behaviour are found to lead to oscillatory responses in the Atlantic Meridional Overturning Circulation (AMOC). These AMOC responses peak in magnitude at lags of 2-3 years and ~17 years following the vortex anomalies. The vortex-AMOC interaction is characterised by non-stationary variations at periods of 30 and 50 years and involves a corresponding modulation of the North Atlantic Oscillation (NAO). Using the relationship between persistent vortex behaviour and the AMOC response established in the model, the contribution of the observed 8-year SSW hiatus interval in the 1990s to the recent ...
|
|---|