This thesis aimed to determine atmospheric temperature turbulence in complex terrain, with the aid of a passive microwave radiometer (MWR). Through the ability to measure atmospheric temperature and humidity proles within 1 Hz sample rate, the MWR can be used to potentially cover atmospheric turbulence at different heights. Because of missing references of MWR based turbulence studies in general, this thesis tried to ll the current gap of determining temperature turbulence structures in an alpine valley. The analyzed periods cover the years 2014 and 2015, mainly the summer months August and September. The goal was to use 1 h temperature time series, estimate temperature variances and compare the near-ground values of the MWR proles to a reference station (Campbell sonic anemometer) located next to the MWR measurement site. This approach is pragmatic because of the rarely available references in literature and the only option to compare turbulence describing parameters to a reliable reference station, since no additional proling with a LIDAR could be used.
The RPG-HATPRO (Radiometer Physics GmbH - Humidity And Temperature PROler) is a multichannel MWR, designed to measure atmospheric temperature and humidity proles within the troposphere. Two operational modes are available, a continuous measuring zenith pointing mode for humidity and temperature proles starring at 90 zenith angle and a boundary layer scan mode for more accurate temperature proling of approximately the lowest 500 m over ground, operating at different elevation angles.
An improved retrieval of Massaro (2013), adjusted to the local climatology, was used to relate the HATPRO measurements to atmospheric proles. Because for turbulence measurements high sample rates are mandatory to cover the full spectrum of turbulence, the zenith mode with a sample rate of 1 Hz is sufcient to potentially cover turbulence structures. Unfortunately the data available for this study is a combination of the two measurement modes. Therefore data gaps in the continuous zenith mode measurements occurred. An algorithm was designed and tested, delivering sufcient results to interpolate missing data. The comparison of temperature variances between the lowest HATPRO measurement (0 m height) and the sonic showed no correlation. Temporal variations of temperature variances could not be observed with the HATPRO.
By investigating the reasons for the HATPRO behavior, a noisy interfering signal lying on the HATPRO measurements could be found. Corrupting noise was thereafter ltered and temperature variances of the ltered HATPRO and sonic data were compared a second time. The results show a similar uncorrelated picture of the ltered HATPRO near-ground variances. In summary the HATPRO is neither able to cover the near-ground daily driven turbulence pattern, nor the right magnitude of variances.
Difculties of MWRs temperature proling in the rst several hundred meters, especially for complex stratication are well-known. They also appear for turbulence investigations as well. However special designed MWRs with limitations with respect to height (approximate maximum height 600 m) are sensitive enough, to depict daily driven near-ground turbulence. The hypothesis, that the HATPRO yields useful information on temperature turbulence has to be rejected.