In the last two decades, the radio occultation (RO) technique has proved to be a reliabletool for climate monitoring. The acquisition of such measurements is one of the main goals of the mini-satellite mission CHAllenging Minisatellite Payload (CHAMP). Since the climatologies from RO dataare composed of discrete measurements, the entire spatio-temporal variability of the atmospherecannot be captured and variations in both time and location occur. In orderto investigate the representativeness of climatologies, "true" profiles at the location anddate of the occultation event were compared with the "true" mean field. The resultingquantity is called sampling error, which can be regarded as the representativness of RO climatologies.The investigation of the long-term stability of temperature sampling errorsin the CHAMP RO climate record 2002 to 2008 lies in the focus of this thesis. Former investigationsof sampling errors by Steiner et al. (2007) yielded a significant negative trendconsidering the years 2001 to 2006. Potential reasons for this effect may be orbit driftsof the involved satellites. In order to figure this out,the event distribution of CHAMP RO events was simulated for different assumptions oforbit development for CHAMP and GPS satellites.By means of the respective event distribution, the 10 zonal, monthly averaged samplingerror profiles were calculated. In addition the profiles were weighted with a weightingfunction in order to calculate the sampling error for the lower stratosphere. The graphical presentation of the results displays the largest values at high latitudes above15 km, where certain months reveal top values of about 2 K. The analysis of trends did notreveal any significant trend for any of the simulation runs. Hence,no appreciable influence of varying satellite orbits on the long-term stability of samplingerrors could be detected.