The 10th Symposium on Global Change Studies

2A.24
ON THE NATURE OF THE NORTH ATLANTIC OSCILLATION AND ITS INTERDECADAL VARIABILITY

Thomas Jung, Institut fuer Meereskunde, Kiel, Germany; and E. Ruprecht

The North Atlantic Oscillation (NAO) has a strong impact upon the climate of the neighboring continents and projects onto the observed Northern Hemisphere winter warming during the last three decades (Hurrell 1996). Here we will discuss the spatio-temporal behaviour of the NAO and possible links with other climate system components using different in situ measurements and analyses. Wavelet analysis of the winter (JFM) NAO index (1865-1997) reveals, that no distinct mode of decadal variability is evident. There exists, however, an abrupt shift during the end of the last century from interannual towards lower-frequency variability which comes in parallel with increasing power on the interdecadal time scale. EOF analysis of anomalous local wavelet spectra shows that 46% of the total variance can be explained by this reddening.

Lag composite analyses (first done by Hurrell and McCartney) for the NAO index and SLP analyses (1900-1997) show that strong winter NAO anomalies typically disappear during the subsequent spring, summer and autumn but reappear during winter one year later. This recurrence is a strong indication for an external forcing of the atmosphere and seems to be connected with processes on the interdecadal time scale, because it vanishes if the same analysis is applied to a highpass-filtered (2-30 years) data. Epoch analyses of SST fields show that interdecadal changes in the upper layer of the Atlantic Ocean are associated with these NAO variations. These oceanic changes, which are even global in character, depend upon the analyzed period, i.e. they are primarily of one sign during the first part of this century and much more structured during the last four decades. For the latter period the propagation of SST anomalies along the Gulf Stream/North Atlantic Current extension (Sutton and Allen 1997) seems to be connected with changes of the NAO. In order to describe interdecadal variability (50-80 years) with sufficient confidence instrumental records are generally too short. Therefore paleoclimate data have to be analyzed to prove the unusual increase in interdecadal variability of the NAO during this century

The 10th Symposium on Global Change Studies