Diploma Thesis
Quality assessment of Eddy Covariance measurements above Tibetan Plateau grasslands
Stefan Metzger (10/2006-04/2007)
Support: Thomas Foken, Tiina Markkanen
The Institute of Tibetan Plateau Research, Chinese Academy of Sciences, investigates the turbulent fluxes of sensible and latent heat as well as CO2 under typical grazing conditions of the Tibetan grassland at altitudes up to 4745 m above sea level. For the two sites under investigation, Eddy Covariance flux measurements are set up at 3 m height above ground in addition to nearby PBL-, Wind Profiler- and Radio Acoustic Sounding System observations. Besides the Campbell CSAT-3 sonic anemometer and the LICOR-7500 infrared gas analyzer, instrumentation consists of radiation instruments for shortwave (Kipp & Zonen CM3) and longwave (Kipp & Zonen CG3) components along with soil heat flux measurements (Hukseflux HFP01). The datasets available almost cover a span of one year and therewith allow, besides assessment of the diurnal cycles, statements on the annual cycles including phenomena such as the Asian monsoon. After a high resolution land use classification based on satellite- imagery and ground verification a subsequent quality analysis including tests on stationarity, integral turbulence characteristics and footprint modeling has revealed spatial and temporal structures in the quality of the Eddy Covariance measurements. Vertical wind analysis revealed that up to 1/4 of the measurements do not fulfill assumptions necessary for a physically correct data processing. Despite this fact, flux measurements in general fulfill the quality test criteria to an extent that ≈50 % of the results can be regarded as suitable for fundamental research. Further, footprint analysis confirms high representativity of the flux measurements for the grassland component and revealed sectors of compromised data quality. Part of the latter could be interpreted as related to instrument set up, for what measures are proposed to improve data quality. Remaining indications could not be entirely explained due to local topography. Investigations of the energy balance revealed a non-closure of more than 30 % of the net radiation. Primary reasons are believed to be the regional topography leading to advection and unconsidered turbulent fluxes due to wavelengths longer than the averaging interval. Supplemented findings of Wind Profiler and Radio Acoustic Sounding System observations include the detection of two wind shears at 500 m and 1500 m above ground and the phenomenon of glacier wind with a reversed compensatory flow reaching up to 2000 m above ground. Together with the results from Eddy Covariance quality assessment, this gives way to the hypothesis of mesoscale flow patterns that contribute organized structures of long wavelength to the boundary layer affecting representativity of Eddy Covariance results.