Phase scintillation caused by propagation through solar wind, ionospheric, and tropospheric irregularities is a noise process for many spacecraft radio science experiments. In precision Doppler tracking observations, scintillation can be the dominant noise process. Scintillation statistics are necessary for experiment planning and in design of signal processing procedures. Here high-precision tracking data taken with operational spacecraft (Mars Observer, Galileo, and Mars Global Surveyor) and ground systems are used to produce temporal statistics of tropospheric and plasma phase scintillation. The variance of Doppler frequency fluctuations is approximately decomposed into two propagation processes. The first, associated with distributed scattering along the sight line in the solar wind, has a smooth spectrum. The second, associated principally with localized tropospheric scattering for X-band experiments, has a marked autocorrelation peak at the two-way light time between the Earth and the spacecraft (thus a cosine-squared modulation of the fluctuation power spectrum). For X-band data taken in the antisolar hemisphere the average noise levels of this process are in good agreement with average tropospheric noise levels determined independently from water vapor radiometer observations and radio interferometic data. The variance of the process having a smooth spectrum is consistent: with plasma noise levels determined independently from dual-frequency observations of the Viking spacecraft made at comparable Sun-Earth-spacecraft angles. The observations reported here are used to refine the propagation noise model for Doppler tracking of deep space probes. In particular, they can be used to predict propagation noise levels for high-precision X- and Ka-band tracking observations (e.g., atmosphere/ionosphere/ring occultations; celestial mechanics experiments, and gravitational wave experiments) to be done using the Cassini spacecraft.
