PhD Thesis
Eddy-covariance measurements as a tool for ecological and hydrological studies on Tibetan Plateau
Tobias Biermann (03/2009-11/2014)
Support: Thomas Foken
The environment of the Tibetan Plateau (TP) is regarded as one of the most sensitive ecosystems of the world. However, investigations on eects caused by global climate change or anthropogenic activities are rare due to its quite remote location. This thesis deals with various aspects of the carbon and water cycle within ecological and hydrological studies that can be assessed by measurements of turbulent uxes with the eddy-covariance (EC) method. This includes ux measurements over the most common two vegetation types on the TP, Kobresia pygmaea pasture and alpine steppe. In particular these in situ measurements were used to investigate dierences in the uxes of grazed and ungrazed Kobresia pygmaea pasture at Kema in Naqu province and moist and dry alpine steppe within the Nam Co basin. Additionally for the rst time on the TP direct ux measurements were conducted over a lake surface in the same basin. Furthermore, these ux measurements were used to adapt and validate land-surface models to be applied on the TP. Within the experiments a special focus were measurements of latent heat ux during winter conditions on the TP, involving side-by-side measurements with a LI-COR 7500 (LI-COR Biosciences) and a Krypton Hygrometer KH20 (Campbell Sci. Ltd.). This comparison revealed that in general an application of both sensors for an estimation of turbulent uxes is possible but they can not be used for measurements of absolute humidity concentrations. Furthermore, this study showed the need for a thoroughly planned calibration procedure for gas analyzers at long-term stations. Carbon uxes measured over the Kobresia pygmaea pastures were used for the estimation of short term eects of grazing cessation and to gain a detailed look into the carbon cycle of this unique ecosystem. For this purpose the EC derived uxes were coupled with the results of a 13CO2 pulse labeling experiment. With this quite novel approach it is possible to on the one hand estimate a more suitable timing for the pulse labeling experiment and on the other hand to estimate absolute C turnover in dierent compartments of the soil-plant-atmosphere continuum. These results identi ed the unique root layer of Kobresia pygmaea as the most important part of the pasture ecosystem. This is of great importance for the further conducted study which estimated the eect of degradation on the carbon and water cycle within these pastures with an interdisciplinary approach, which combined plot- and ecosystem scale in situ measurements with land-surface-atmosphere models. A simulation of dierent stages of degradation of the Kobresia pygmaea pastures and also a vegetation shift to alpine steppe, showed that only an intact Kobresia pygmaea pastures acts as C sink for the observation period within the main vegetation growing season. Although evapotranspiration in general is not aected as strong as carbon exchange, a shift in the ratio between evaporation and transpiration has a feedback on convection development and precipitation which could be also shown with an atmospheric model.