Analysis of turbulence at 60 m agl largely confirms previous findings of a relatively well-defined relationship between Monin-Obhukov stability and integral statistics such as normalized standard deviations. The site is surprisingly homogeneous as evident from a lack of a dependence of turbulence on wind direction. In addition, the neutral limit analysis for normalized standard deviations resulted in a values insignificantly different from the commonly used values of 2.3, 1.8, and 1.2 for u, v, and w, respectively. Slight differences to previous data, summarized by Roth (2000) were observed for the temperature gradients and heat flux data, consistent with a tendency towards higher values for σT/T* with increasing observation height. In contrast to previous work, obtained at observational heights (zS), between 1.5 to 3 times the average roughness element height (zH), which may have affected results due to roughness sublayer influence, our measurements are conducted at zS/zH = 7.5, and turbulence measurements at zS of 40 versus 60 m agl confirmed that the roughness sublayer was avoided.
Surface displacement height and roughness length were evaluated for comparisons to standard air quality model input data. While we confirmed that rule-of thumb values related to the major roughness elements are appropriate under the local conditions, the underlying distributions were not symmetric. First, building parameters appear to have a more limited influence on roughness parameters compared to the mature trees in the area. The latter, although far from presenting a closed canopy, appear to determine the displacement height distribution's mode. The trees' influence is surprisingly well described by a recently introduced tree height distribution method to determine displacement height for complex forest morphologies, when including building height distribution. More so than displacement heights, roughness length values distribution is highly skewed towards higher values, with median and mode around 1 m, a value commonly used in models for urban areas. Wind speed and friction velocity display a strong influence on these values as the footprint area of the tower is similar with wind direction but inhomogeneous especially regarding tree density with distance from the tower.
Carbon dioxide and air pollutant fluxes are strongly modulated by car traffic in the area, with the most prominent flux peaks during morning rush-hours. CO2 fluxes are additionally modulated by photosynthesis, with net ecosystem exchange close to neutral during summer weekends under non-drought conditions. The CO to NOx flux ratio is consistent with recent tailpipe emission re-evaluations, but emissions of selected hydrocarbons appear to differ substantially from recent emission inventories, likely influenced by larger than expected evaporative emissions. In addition, unexpectedly high isoprene tailpipe emissions were discovered.
Flux tower operations will resume summer 2010 and collaborations are encouraged. New particle flux measurements will be added to the setup. Previous data will be made available on our project page as more of it becomes quality assured. Manuscripts are in preparation.