Characterization of a complex measuring site for flux measurements

Several long-term flux measuring programs, e. g. organised in FLUXNET, are based on measurements at very complex sites. These are not in agreement with the fundamental as-sumptions of eddy-covariance measurements like horizontal homogeneity, steady state and advection free conditions. On the other hand, the land use conditions in many places do not allow ideal measuring conditions. An investigation took place at the 'Weidenbrunnen' meas-uring site of the University of Bayreuth in the hilly region of the Fichtelgebirge at 780 m a.s.l. and over a 20 m high spruce forest. A station of the EUROFLUX programme is installed on a 30 m high tower. The investigation was made with stability dependant footprint modelling and investigations of normalised standard deviations of turbulent parameters. It was found that even over the non-homogeneous forest the normalised standard deviation of the vertical wind velocity fulfills usual parametrizations of this parameter. It can be assumed that eddy-covariance measurements of turbulent fluxes of scalars as well as relaxed eddy accumula-tion measurements are not influenced by the terrain. In contrast, influences of single high trees and clearings in the forest were found in the normalised standard derivation of the hori-zontal wind velocity, where a small influence of the mixing layer height was found. Conse-quently, the use of the K-approach at this site is restricted. Measurements according to the profile assumption, K-models (which should for other reasons not be used in the forest) and stability parametrizations using the standard deviation of the wind directions can only be used with local correction factors. This investigation can be a basis for a quality classification depending on the site and meteorological conditions.