Use of footprint modelling for the characterization of complex measurement sites

The adoption of the eddy covariance technique to estimate surface exchange is based on the assumption that certain meteorological conditions are valid. The most important of these are horizontal homogeneity, steady-state, and non-advective conditions. As such conditions are often violated under complex terrain conditions, e.g. at flux monitoring sites over forests, this study aims to evaluate the influence of surface heterogeneity in order to allow a correct interpretation of the measurement results.

An approach has been developed which combines existing quality assessment tools for eddy covariance measurements with footprint modelling. For the determination of the spatial context of the measurements, the Thomson (1987) Lagrangian stochastic trajectory model of Langevin type (e.g. Wilson et al., 1996) is used in the version as parameterised by Rannik et al. (2003). In a pre-processing step, effective roughness lengths are determined with a flux aggregation model (Hasager and Jensen, 1999). This way, the land use features of the surrounding terrain can be included in the analysis. The approach enables to determine the dominating flux data quality for different wind sectors and varying meteorological conditions, so that the most suitable situations for the collection of high-quality data sets can be identified. Additionally, the flux contribution of the different land use types present in the footprint area is calculated.

The results can be visualised in two-dimensional graphs, which show the spatial distribution of the quality of different fluxes. These graphs help to identify terrain influences affecting the flux data quality, such as dominating obstacles in the fetch, slopes biasing the wind field, or even flow distortion or misalignment of the sensor itself. The evaluation is especially useful for checking to what extent the measured fluxes at a site are representative for a specific type of land use.