Abstract: Communication's environment of the meteorological satellite is the physical layer of the atmosphere. In different weather conditions, there are varying degrees of attenuation, shadowing and multipath effects in meteorological satellite channels due to the impact of cloud and rainfall, which makes the receiving signals become unstable concerning to the atmosphere state, and result in inter-symbol interference, affecting the quality of receiving meteorological satellite data, prediction of weather phenomena.In order to study the influence of the physical layer of atmospheric space on meteorological satellite communications, the weather conditions affecting meteorological satellite transmission are classified into three cases: Clear sky, cloudy and rainy weather. There are thick clouds, covering clouds and rainfall in the rainy weather, which take the whole shadow block above the ground station receiving signals of the meteorological satellite communications. In this case, the signals received by ground station are only composed of multipath scattering signals without the line of sight, and the envelope probability density function (PDF) obeys the statistical characterizations of Rayleigh. There are thick clouds and covering clouds in the cloudy weather, which form part of the shadow block over a certain range spread above the ground station. In this case, the signals received by ground station may have two situations. One circumstance, the received signals are composed of the line of sight and a certain intensity of multipath scattering signals that are diffracted, refracted and scattered, and PDF obeys the statistical characterizations of Rice. The other circumstance, the received signals are composed of the line of sight obscured by clouds, and PDF obeys the statistical characterizations of Lognormal. There are a few thin clouds in the clear sky and good visibility in the high atmosphere layer, when the signals received by ground station are composed of very weak multipath scattering signals and the line of sight, and PDF obeys the statistical characterizations of Gauss.According to theoretical analysis, the simulation models of Rayleigh, Rice, Lognormal and Gauss probability distribution are established. Through the computer calculation, the results of the simulation models show that the signals received by ground station with different composition lead to different statistical characterizations because meteorological satellite signal pass through different physical state of the atmosphere. The multipath scattering signals both exist in the Rice channel and the Rayleigh channel, however, the line of sight only exists in the Rice channel. Gauss channel model and the Rice channel model have the same structure, but the received signals in both channels have different intensity of the multipath scattering components. That explains the cause why the variety of the received signals envelope brings out different statistical characterizations in different channels. The probability density curve of the simulation model and the theoretical model match quite well, verifying the correctness and validity of the theoretical analysis, providing a theoretical guidance to calculate the data error rate of the meteorological satellite communication.