Tobias Borsdorff
Variability and trends of the vertical concentration distribution of carbon monoxide and water vapor derived from solar FTIR spectrometry as measured on the Zugspitze and in Garmisch
Supervisor: Prof. Dr. Wolfgang Seiler [Experimental physics I]
Date of oral examination: 12/10/2010
185 pages, german [original version ]

This work derives long-term time series of the trace gases CO and H2O from solar FTIR spectra as measured at stations located on the Zugspitze (47.4° N, 11.0° E, 2964 m a.s.l.) and in Garmisch (47.5° N, 11.1° E, 745 m a.s.l.). This provides new insights into the global transport of CO as well as on the spatial variability of trends of the column-integrated atmospheric water vapor. The derivation of a trace gas profile from an infrared spectrum is disturbed by strong cross-sensitivities to other trace gases. These interference errors are neglected by the classical error analysis of remote sensing. Current approaches to error analysis underestimate interference errors and are not applicable to all inversion methods. A new error analysis presented here allows the correct quantification of interference errors for each inversion method. A regularization strategy is then developed that minimizes the occurrence of interference errors. These new concepts are applicable for all trace gases and remote sensing techniques in the infrared. Existing methods for the inversion of CO profiles from FTIR spectra tend to dampen the profile concentration variability and even to displace variabilities to incorrect heights. A new regularization strategy proposed here eliminates these sensitivity problems, thus allowing one to derive a long-term time series of CO profiles. The measurements show that anthropogenic emissions cause a strong gradient in the CO concentration between the boundary layer and the free troposphere. With increasing altitude, the annual cycle of CO becomes more symmetric, and its maximum shifts towards the summer months. It is further demonstrated that wildfires originating in the northern and southern hemispheres affect the CO concentration in different heights. Long-term measurements of strato-mesospheric CO above mid-latitudes have been required for 20 years in order to improve modeling of the transport in the middle atmosphere. However, currently the longest time series comprises only 2 years of microwave measurements. In addition, it was not possible to derive long-term time series from FTIR measurements. In this thesis, a new regularization strategy allows the derivation of a long-term time series of strato-mesospheric CO from FTIR spectra and is applicable to all stations at mid-latitudes. The measurements show a significant seasonality and provide evidence that planetary waves shift CO enriched polar air towards the mid-latitudes. The first statistics of these transport events are presented. Trends in column-integrated atmospheric water vapor can provide information on whether global warming intensifies the natural greenhouse effect. Previous long-term time series based on radiosonde measurements are not homogeneous and therefore not useful for trend studies. A new regularization strategy allows the derivation of a long-term time series of the column-integrated atmospheric water vapor, which is homogeneous in contrast to the radiosonde time series. The measurements of the Zugspitze and Jungfraujoch stations show that trends in the column-integrated atmospheric water vapor are different even at horizontal scales of 250 km, thus supporting recent satellite measurements showing strong spatial water vapor variability over land.