Master Thesis
Carbon dioxide exchange above Kobresia meadows
Thomas Leipold (08/2010-04/2011)
Support: Thomas Foken, Johannes Lüers
To quantify land use effects on the carbon sequestration potential for high mountainous grassland ecosystems, two Eddy Covariance measurement towers, were installed at an ungrazed and a grazed Kobresia meadow in the summer 2010 on the Tibet plateau to investigate their net ecosystem exchange (NEE). Carbon flux measurements in general contains periods with no data. This is caused by the measurement method itself where necessary prerequisites have to be fulfilled for the correct application and in additional instrumentation errors. In this study a gap filling strategy based on regression analysis with the Lloyd-Taylor function for the night time NEE and the Michaelis-Menton function for the day time NEE was investigated to calculate the missing data. To improve the parametrisation of these functions an additional data treatment with site specific thresholds and a statistical based spike detection was applied. The meteorological environment conditions with extreme fluctuations in global radiation and a very dry vegetation period in July with 40 mm precipitation results in extreme low and high scattered NEE fluxes and caused problems for the parametrisation of the day time NEE, whereas the applied method for the night time flux parametrisation can be assumed as appropriate and the spike detection process may regarded as method to improve the quality of the input data the parametrisation and the gap filling is based on. Due to the unsuitable gap filling an underestimating of the results for the NEE with »-6 g C m-² for the ungrazed and »-4 g C m-² for the grazed site during the measurement period and the resulting carbon sink potential is obvious. The small difference between the land use types indicates that a grazing effect, excluding a low grazing intensity, can not be assumed. This result is quite opposite to the expected difference in land use management and was maybe caused by the dry situation. To provide a reliable gap filling strategy for further investigations in ecosystems with comparable environmental conditions the applied Michaelis-Menton light response regression, based on temperature classification have to be verified. Therefore, investigations with temperature ranges, different to the in this study applied 3K classification may improve the parametrisation. Additionally not temperature dependent gap filling approaches, based on wind or the photosynthetic photon flux density, should be investigated. This study showed that a lot of attention have paid using gap filling to calculate seasonal sums of the NEE to avoid wrong interpretation of the carbon sequestration potential for ecosystems with high NEE flux rate and an extreme fluctuating solar radiation during the day cycle.