Item

Abstract

In this thesis, different processes that might contribute to the generation of decadal
climate variability were investigated using general circulation models (GCMs) of
the atmosphere and the ocean. First, the sensitivity of the atmospheric circulation
to decadal changes in the underlying sea surface temperatures (SSTs) was esti-
mated from an ensemble of six integrations of the Hadley Centre atmospheric GCM
HadAMl, all forced by observed SSTs and sea-ice extents for the period 1949-93.
Using a novel approach to estimate the 'true' SST-forced atmospheric response in
the presence of spatially correlated internal atmospheric variability, the decadal at-
mospheric variability was studied over the North Atlantic and North Pacific regions.
After filtering out the atmospheric circulation changes associated with the El Niño -
Southern Oscillation (ENSO) phenomenon, the dominant mode of forced variability
over the North Atlantic exhibits a meridional dipole in the mean sea level pressure
(MSLP) field and is related to a tripole in the anomalous North Atlantic SSTs. Over
large parts of the North Atlantic region, however, the atmospheric response is not
consistent enough to provide feedbacks to the underlying ocean that could cause
self-sustained decadal oscillations. Over the North Pacific the atmospheric response
is dominated by ENSO. In addition to the ENSO-related response an independent
decadal atmospheric signal was detected. It consistently involves iarge-scaie wind
stress curl anomalies over the North Pacific region. The effect of such wind stress
curl anomalies on the ocean was studied in the second part of this thesis using the
Hamburg Ocean Primitive Equation model (HOPE). It is shown how the adjust-
ment of the North Pacific gyre circulation to large-scale wind stress curl anomalies
determines the decadal timescale and how it may be exploited for predictions of
decadal upper-ocean temperature changes in the central North Pacific. The HOPE
model was also used to investigate a mechanism for the generation of decadal cli-
mate variability in the tropical Pacific which relies on subduction of midlatitudinal
North Pacific SST anomalies and their equatorward propagation within the oceanic
thermocline. It is demonstrated that such a mechanism is unlikely to cause decadal
climate variability in the tropical Pacific.