Diploma Thesis
Einsatzvorbereitung einer REA-Anlage über Wiesenflächen
Jörg Hübner (03/2009-06/2010)
Support: Thomas Foken
Turbulent fluxes of some trace gases as well as their isotope ratios cannot be determined directly by the eddy-covariance (EC) method due to a lack of fast analyzers. Adequate tools are eddy-accumulation (EA) methods that combine the eddy-covariance method with an eddy-accumulator, which collects sample air into different reservoirs, depending on the vertical wind. Afterwards the samples can be analyzed in the laboratory. The aim of this thesis is to reinitialize a relaxed eddy-accumulation (REA) complex after long idle time and prepare it for a field measurement campaign. After some modifications (installation of larger foliage balloon reservoirs and a new pressure sensor), performance tests have been conducted to proof that the REA complex still meets the requirements of 13C flux measurements. Installation of a pressure sensor closed a gap in the systems performance security. Now leakage can be detected by observing inner system pressure. New foliage balloons have no influence on sample air and δ13C values differ only marginally from gas bottle samples. Beneath technical preparation of the REA complex, simulations of the hyperbolic relaxed eddy-accumulation (HREA) method have been conducted for the grassland site „Voitsumra“ in the Fichtelgebirge Mountains (Northeastern Bavaria, Germany) for eight golden days during 2009-08-31 and 2009-09-27. The HREA simulation provides evidence for the determination of an appropriate hyperbolic deadband size Hh and of the coefficient b for the HREA measurements in summer 2010. Therefore, the scalar of interest (CO2) and two other proxy scalars (temperature and water vapour) have been simulated. A hyperbolic deadband size of Hh = 1.0 turned out to be the best criterion to achieve sufficient concentration differences as well as a balance between up- and downstream events. This is essential for accurate determination of 13C fluxes. The simulation results show, at high scalar similarity between scalar of interest and proxy scalar, a good conformity between simulated HREA CO2 flux and determined EC CO2 flux during the first five golden days. For this period coefficients b from 0.16 to 0.27 can be found. Due to fluctuations of b during the day, a variable b was used for flux alignment. Decreasing scalar similarity causes overestimation of the simulated flux by 20 % for water vapour and up to 40 % for sensible heat flux. After 2009-09-21 downward CO2 fluxes collapsed due to mowing of the meadow, whereby minor upward EC CO2 fluxes caused a positive net EC CO2 flux. Compared to the low EC CO2 flux, the HREA simulation errors were relatively big and so the simulated flux could not be determined correctly. The result was a negative HREA flux, which causes negative coefficients b, when compared to the EC flux during simulation. The mowing had lesser influence on the heat fluxes and associated proxy scalars and thereby caused a significant decease of the scalar similarity to the scalar of interest (CO2). Usage of those proxy scalars to determine the HREA CO2 flux causes significant overestimation, compared to the real CO2 flux.