An attempt is made to identify and investigate the structure, behaviour and synoptic characteristics of cyclogenesis over the Aegean Sea. A simple, realistic and easily used methodology of identification, in an analytical form, was established. This combines two types of criteria: synoptic and dynamic/thermodynamic. Six synoptic categories that could favour cyclogenesis over the Aegean Sea are distinguished according to the synoptic criteria. Grid-point values of geopotential height and temperature are used to calculate the dynamic/thermodynamic criteria, which are used to identify potential cases of cyclogenesis with respect to the thermodynamic prerequisites and the time continuance. It is demonstrated that cyclogenesis over the Aegean Sea is not as rare as was believed. The majority of cases result from the prevalence of a south-westerly flow, a long-wave trough and a closed system over the Aegean Sea. The cyclogenesis appears to occur primarily in the cold period of the year, mainly from October to May, and to have a small intensity and lifetime.

An attempt is made to find the dynamical and physical characteristics of the atmospheric circulation which contributed to the development of a great storm in the vicinity of Greece on 5 October 1989. The surface cyclogenesis could be considered as a consequence of an upper-tropospheric minor wave which formed upstream on the eastern flank of a blocking-type anticyclone which was centred over Britain and dominated the atmospheric circulation over western Europe. This minor wave formed three days before the initiation of the surface cyclogenesis and could be explained by the theory of ’a dynamically unstable ridge’. As the wave moved south-eastwards, an elongated trough formed in the upper troposphere. The southernmost part of this elongated trough was disrupted by dynamical processes, leading finally to the formation of a cut-off low. When the area of positive relative vorticity advection due to the low became superimposed upon a shallow frontal surface lying across southern Greece in the afternoon of 4 October 1989, a deep depression developed at the surface between Crete and Athens during the next 12 hours. To further elucidate the role played by the forcing conditions at the upper levels in the surface cyclogenesis, the potential vorticity has been analysed.

The prediction of propagation and motion of thunderstorm activity, especially when heavy rainfall accumulation amounts could occur, is a matter of great importance. The objective of this study is to examine the possibility of using the environmental wind data for the problem of assessing predictability of the propagation and movement of thunderstorms in northern Greece. The propagation of radar thunderstorm echoes (radar echo reflectivity maxima) and cell movement (fine radar echo patterns) were determined for the spring and summer thunderstorms in 1992 and 1993 (April to September). Radar data were used by examining and analyzing digitally recorded plan position indicator for the entire time that the radar was operated. Multicell and single-cell thunderstorms were identified and thunderstorm propagation and cell motion were calculated for each. Environmental winds at standard levels of 850, 700, and 500 hPa were considered and, additionally, a mean 0- to 6-km layer density-weighted wind was also examined. In northern Greece, winds and radar thunderstorm echoes rarely moved from the clockwise sector 070° to 220° (ENE - SW). The movement of thunderstorm cells was not only steered by the 700-hPa level wind but was also well represented by the wind at the 500-hPa level and by the average wind in the layer 0-6 km. Using standard levels, cell motion can be determined by adding 5° to 500-hPa wind direction and reducing the 500-hPa wind speed by 30%. With the use of the mean 0- to 6-km density-weighted wind, cell motion can be represented by adding 5° to the 0- to 6-km wind direction and increasing 35% the 0- to 6-km wind speed. The propagation of multicell and single-cell thunderstorm echoes was very similar, in spite of initial expectations, and was approximately equal to the cell motion, suggesting that the new cells grew on all sides of existing multicell thunderstorms. The majority of northern Greece thunderstorms do not propagate significantly and their motion is substantially translational and similar to cell motion. However, in cases of severe thunderstorms, propagation was indicated. Characteristic cases of storm propagation are presented and vertical wind shear is investigated as a particularly important factor in influencing storm structure and evolution and the resulting storm propagation.