On the correction of RASS-temperature errors due to turbulence

Temperature profiles, measured by RASS, show a mean positive bias up to 1 K as compared to radiosoundings. Here the influence of turbulence on the apparent sound velocity is analyzed, and a correction, based on measured turbulence profiles, is suggested. This correction was applied to data of two similar UHF wind profilers with RASS extension. One data set was obtained in June 1992 during the ROSE II expriment in Alabama (USA) and the other was taken from routine measurements in July/August 1994 at the Meteorological Observatory Lindenberg (Germany). The correction is basically satisfying for the ROSE II data set while systematic differences remain for the Lindenberg data, which requires further study. In the ROSE II data an error remains in the lowest two range gates, which is still not understood. Certain approximations used for the signal analysis are not applicable to these lowest gates.