Natascha Kljun

Ideally a flux tower should be installed on a homogeneous and flat terrain. The surface should be physically homogeneous (same forest height and thermal properties) as well as be covered by same tree species, or in the case of the mixed forest, the distribution of the different species should be even (“well-mixed”). The fetch, the outreach of the homogeneous surface, should be longer than the extension of source area of the measurement (footprint). However, many sites are not homogeneous enough in all directions from the tower. In the case of an inhomogeneous surface, knowledge of both the source area and strength is needed to interpret the measured signal. Note that inhomogeneity modifies the footprint by modifying the turbulent flow field. Thus, strictly speaking, any method not accounting for heterogeneities is useless for source area estimation. Namely, either the footprint model is fundamentally wrong because of the implicit assumption of homogeneity or, in the case of the fully homogeneous case, the outcome is trivial and no estimation is needed. Nevertheless, footprint models based on the assumption of horizontally homogeneous turbulence field serve as first approximation for evaluation of source contribution to measured flux in real observation conditions. An alternative is to take the flow inhomogeneity into account in footprint estimation by models capable of simulating such flow fields (cf. Sect. 8.4.1).