Logo: to the web site of Uppsala University

The exchange of sensible and latent heat at the sea surface has been studied with the aid of a comprehensive data set from the marine site Östergarnsholm in the Baltic Sea, with additional data from another site in the Baltic Sea, Nässkär. The measurements include turbulent flux at about 10 m above the water surface, profile data of temperature and wind at several levels on towers, sea surface temperature and wave data. The Stanton number, CH was found to follow predictions from surface-renewal theory quite well for unstable conditions up to a wind speed of about 10 ms-1. For higher wind speed, the experimental data deviate to an increasing degree from the prediction based on surface-renewal theory, giving about 30% higher values at 14 ms-1and 50% higher values at 16 ms-1. For wind speed less than 10 ms-1, co-spectra of heat flux have the same shape as found over land with one peak at a fairly low frequency (~10-2 Hz). For increasing wind speed, a secondary maximum around 0.5 Hz is forming. The ratio between the spectral values of the high frequency peak and the low-frequency peak is found to be a strong function of significant wave height, Hs, being as high as 2 for Hs > 2.5 m. The interpretation is made that spray is the cause of the rapid increase of heat exchange with wind speed above c. 10 ms-1. In a plot of CH against stability, z/L, the data show a jump at neutral stratification from about 2 on the unstable side to 0.5 in stable air. For stable conditions, CH-data are widely scattered, It appears though as if the data have an approximate upper bound given by surface-renewal theory. It is demonstrated that at least some of the strong suppression of the flux of sensible heat during stable conditions and low wind speed can be explained as a shear sheltering effect caused by the presence of a wind maximum at a low height. For wind speed above 10 ms-1, the spray-mediated sensible heat flux is lowering the negative heat flux, thus giving a lower CH. The combination of those two mechanisms gives a low mean value of CH for the entire stable range and explains the ‘jump’ at neutral stratification. The exchange coefficient for humidity, CE behaves in much the same way as the Stanton number. Because of the failure of Monin-Obukhov theory over the sea, the conventional way to reduce CH and CE to their neutral values cannot be used.