From the 'Community-wide evaluation of knowledge of quasi-coherent structures in turbulent boundary layers' [Kline and Robinson, 1989] and from results of Robinson's [1991] DNS-analysis, it appears beyond doubt that turbulence production close to the surface is an autonomous process which takes place independently of large scale processes in the outerlayer, although some slight interaction does occur. As shown in Kline and Robinson [1989] this statement is not only true for the 'canonical' boundary layer (the zero pressure boundary layer over a flat plate) but has probably very general validity. Thus, most of the turbulence in the boundary layer as well as most of the shearing stress is being produced as a result of this autonomous near-wall process. In the terminology of Townsend [1961] we may identify this as 'active' turbulence. Townsend also identified turbulence of boundary layer depth scale which does not produce shearing stress at the surface; he called it 'inactive turbulence'. Högström [1990] and [1992] has shown that 'inactive turbulence' plays an important role in the neutrally stratified atmospheric boundary layer (ABL).
From measurements in the ABL it is during 'normal' conditions impossible to distinguish between 'active turbulence' produced at the surface and 'active turbulence' produced in the upper part of the boundary layer and 'inactive turbulence also produced in the upper part of the BL. However, in the marine ABL during conditions when the waves are traveling faster than the wind (swell) there is a possibility to get some indications of the magnitude of the shearing stress produced aloft.
The basic conceptual idea for air-sea interaction during swell conditions is that the surface waves transport momentum upwards into the atmosphere, with the aid of pressure fluctuations induced by the waves. The result is a surface layer where the shearing stress is close to zero. That this actually occurs has been convincingly demonstrated in several studies [Harris, 1966, Lai and Shemdin, 1971, Makova, 1975, Smedman et al., 1994 and Smedman et al., 1998]. The depth of the atmospheric layer affected by this energy transfer was found to be appreciable and much deeper than the corresponding depth affected in the case of developing waves. Turbulence measurements taken at the island Oestergarnsholm in the Baltic Sea during four days with neutral stratification and developing swell conditions are analyzed. The data set comprises tower based profile and turbulence measurements at several levels together with measurements from a 3-D wave rider buoy deployed outside the island.
From spectral analysis of co- and quadrature spectra as well as the phase angle between the u and w components it is possible to distinguish between(i) small scale 'active turbulence' produced close to the surface, which decreases with increasing swell (wave age), (ii) large scale 'activeturbulence' probably produced in the upper part of the BL and (iii)'inactive turbulence' also produced high up in the BL, which does not contribute to the shearing stress at the surface.