Wegener, Jörg (2006) Moden nordatlantischer Klimavariabilität. (Diploma thesis), Christian-Albrechts-Universität, Kiel, Germany, 103 pp.

Preview

Text JWegener_Dipl_2006.pdf - Accepted Version Available under License German copyright act UrhG. Download (4MB)

Abstract

According to Rowan T. Sutton and Miles R. Allen (hereafter: SA) there is a chance to enhance the predictability of decadal climate variability over the northern European sector. This enhancement supposedly originates from the horizontal advection of SST anomalies along the Gulf Stream and North Atlantic Current (NAC). SA suggest that fluctuations in SST which arise from advection might be predictable, so would be their impact on the atmosphere. The described advection mechanism is subject of this work, analyzing mainly NCEP/NCAR and COADS datasets of various meteorological and oceanographic parameters. This issue is investigated by means of statistical tools like EOF- and timeseries analysis, especially Dietmar Dommenget’s rotated DEOFs: a technique which fits a spatial first-order autoregressive ("AR(1)") process as a null hypothesis to the data. These analyses should clarify whether a primary mode of advection exists or whether it is solely the atmospheric imprint on the ocean which generates a sham advection. The findings of SA could partly be confirmed, although a few remarks about the statistical reliability of their analyses are made. Applying an AR(1)-model, which divides its driving noise into a NAO and a non-NAO part, it is shown that the former locally contributes more than 40% to the explained variance of SST variability. The maximum contributions align with the major areas of the NAC, where SA suggest the propagation of SST anomalies. It is shown that this signal is strongly influenced by the atmospheric forcing and that latter generates an advection-like pattern.

Actions (login required)

View Item