DETERMINATION OF HEAT AND WATER-VAPOR FLUXES ABOVE A SPRUCE FOREST BYEDDY-CORRELATION

Heat and water vapour fluxes have been measured by eddy correlation at three levels above a spruce forest (tree height 31 m) in central Germ any. A number of methodical problems are discussed, and necessary data corrections are presented. Stationarity of the fluxes is proven by co spectral analysis, confirming that averaging times of 10-30 min are ad equate to describe the transport processes. Sensible and latent heat f luxes do not vary with height in the layer from (z - d) = 6.5z0 to (z - d) = 11.7z0. Based on footprint estimates it is concluded that, in g eneral, the measured fluxes may be taken as representative of the spru ce canopy. Even in the worst-case situations, when the wind comes from a beech-spruce transition 200 m away, no height dependence of the flu xes is detected. On the selected days, which were mainly bright, a Bow en ratio near one is found during the hours of highest available energ y, whereas at night no evapotranspiration is observed. The turbulent d iffusion coefficients of heat and water vapour follow a distinct diurn al pattern, with a maximum of about 10 m2 s-1 shortly after noon. The heat diffusion coefficient tends to exceed that of water vapour by a f actor of about 1.2, showing daytime-dependent variations. Owing to som e unexplained deficits in the energy balance this value is not very ce rtain, but the results are consistent with some other published data. At least they demonstrate that uncritical application of the similarit y hypothesis (D(h) = D(v)) above high vegetation may lead to erroneous flux calculations.