Tropospheric scintillation has an impact on earth-space low availability and deep-space systems, but also on fade mitigation techniques that are developed for communications above 20 GHz. The classical prediction models [1] predict the cumulative distribution of scintillation variance using ground measured data, such as temperature and humidity. Scintillation being due to the presence of turbulence in the troposphere, it is realistic to expect that the accuracy of the prediction should be increased if height profiles of those parameters should be used. Abstract???