Treatment and assessment of the CO2-exchange at a complex forest site in Thuringia, Germany

Eddy covariance measurements were carried out at the Wetzstein site in Thuringia, Germany since December 2001. Soon after the start of the measurements night-time fluxes well above average CO2-fluxes measured in temperate forest ecosystems were detected which could not be explained by biological processes but were valid with respect to standard quality criteria. The Wetzstein site is part of the CarboEurope-IP flux-network and the CO2-exchange of this spruce forest is of general ecological interest as the site is typical for central European spruce forest ecosystems at mountainous elevation. Additional investigations were made in order to identify the causes for the large difference between the flux balance and the inventory based NEP. Specific weather patterns and micrometeorological situations were identified during which a decoupling of the flows above and below the canopy leads to additional CO2-effluxes at the tower site which are not part of the net ecosystem exchange (NEE) at night. Rejecting data from these periods and gap-filling thereafter results in yearly sums of NEE, GPP and TER which are in better agreement with the biometric measurements at the tower site and comparable to other spruce forest sites. In this process ecosystem respiration was determined not only from extrapolation of night-time data but also from flux partitioning based on day-time data using the hyperbolic light response function. It can be shown that flux measurements at this complex site need to be treated in a modified procedure compared to what is generally performed, namely extrapolating ecosystem respiration from night-time data. Using multiple data sources and applying a careful filtering of the data, confidence in the estimates of the carbon balance components increased.