Uiboupin & Laanemets (2009) showed that as much as 40% of the area of the Gulf of Finland can be under the influence of upwelling during extreme conditions. For example, in 2006 the cross-shore extent of upwelling in the Gulf of Finland was 25 km (1/3 of the width of the Gulf of Finland) and the alongshore extension was 360 km (Suursaar & Aps 2007). On the Polish coast upwelling has find more most often been found to take place off the Hel Peninsula in the Gulf of Gdańsk (see e.g. Matciak et al., 2001 and Myrberg et al., 2010). The potential maximum area of all upwelling on the Polish coast is 10 000 km2, which is ca 30% of
the Polish economic zone (Krężel et al. 2005). Statistical studies of upwelling have been carried out before. Myrberg & Andrejev (2003) determined an upwelling index based on the numerical calculation of vertical velocity
for a 10-year period (1979–1988). A similar study was carried out by Kowalewski & Ostrowski (2005) based on a 7-year experiment of calculated vertical velocities in the southern Baltic. The present paper extends the statistical investigation of Baltic Sea upwelling events based on the integrated use of observations and modelling to cover – for the first time – the entire sea area. For the years 1990–2009, weekly sea surface temperature (SST) maps based on NOAA/AVHRR satellite data were used to evaluate the properties of upwelling during the thermally stratified period from May to 3-oxoacyl-(acyl-carrier-protein) reductase September, that see more is to say, when upwelling is strong enough to raise the thermocline to the surface, thus producing an SST signal. To obtain an independent estimate, numerically simulated daily averaged SST maps were analysed for the same period. Furthermore, favourable and unfavourable wind conditions for upwelling were determined from the wind forcing used as model input. The structure of the paper is as follows: after this introduction, data and methods
are briefly described. Then the results of the statistical analysis are discussed for the period 1990–2009; they are also compared with previous studies. A trend analysis over the total period and for individual months is carried out for identified upwelling areas. Furthermore, for specific upwelling locations, 10-m winds are discriminated into upwelling-favourable and -unfavourable wind conditions, and the relation between upwelling and wind forcing is studied. The paper concludes with a discussion on potential changes in upwelling regions as a consequence of changing climate (wind) conditions. The analysis of upwelling regions and their occurrence is based on SST data with a horizontal resolution of about 1 km calculated from NOAA/AVHRR satellite data for the period 1990–2009. The accuracy of the satellite measurement (cloud detection has been carried out) in comparison with in situ data is about 0.5 °C (Siegel et al. 1994).