This thesis presents results from an AGCM sensitivity study in which the response in the Northern Hemisphere storm tracks to an imposed SST anomaly is investigated. The study was motivated by observational studies cited in the IPCC Fourth Assessment Report indicating that the storm tracks have shifted northward during the second half of the 20th century, a shift which may be related to global warming.
To perform sensitivity studies, the NCAR CAM3 model was applied using the data ocean model with an imposed 2 K SST anomaly in all oceanic grid points north of 45◦S. Additionally, the sensitivity to longitudinal and latitudinal variations in the SST anomaly domain was investigated by heating high-latitudes and low-latitudes, only high-latitudes and only low-latitudes in the Atlantic and Arctic Ocean in three different runs. To investigate the importance of a potential reduction in the ice cover, CAM3 was run without ice in the Northern Hemisphere. The storm tracks were represented in terms of bandpass variance using the bandpass filter method and cyclone count using the CCI method developed by Rasmus E. Benestad at the Norwegian Meteorological Institute.
Warming the ocean by 2 K in all oceanic grid points north of 45◦S yields, in terms of bandpass variance, a northeastward shift in the North-Atlantic storm track and no latitudinal shift in the Pacific storm track, with corresponding changes in atmospheric baroclinicity and the mean circulation. The zonally averaged Eady parameter shifts upward and northward in response to an increased upper-level temperature gradient and a decreased lower-level temperature gradient, consistent with the findings of Yin (2005). As Yin (2005) performed a climate scenario study using a 15 member ensemble of coupled GCMs, while this study investigates the effects in a uncoupled AGCM where the only forcing is a positive SST anomaly, indications are that oceanic heating is the main driver of the observed storm track changes.
Variations in the SST anomaly domain reveals that low-latitude heating is the primary driver of the observed storm track changes. Removing the ice cover yields changes of comparable magnitudes to highlatitude heating, as is therefore not as important as low-latitude heating.