Assimilating the proper amount of water vapor into a numerical weather prediction (NWP) model is essential in accurately forecasting a heavy rainfall. Radar data assimilation can effectively initialize the three-dimensional structure, intensity, and movement of precipitation fields to an NWP at a high resolution (±250 m). However, the in-cloud water vapor amount estimated from radar reflectivity is empirical and assumes that the air is saturated when the reflectivity exceeds a certain threshold. Previous studies show that this assumption tends to overpredict the rainfall intensity in the early hours of the prediction. The purpose of this study is to reduce the initial value error associated with the amount of water vapor in radar reflectivity by introducing advanced remote sensing data. The ongoing research shows that errors can be largely solved by assimilating satellite all-sky radiances and global positioning system radio occultation (GPSRO) refractivity to enhance the moisture analysis during the cycling period. The impacts of assimilating moisture variables from satellite all-sky radiances and GPSRO refractivity in addition to hydrometeor variables from radar reflectivity generate proper amounts of moisture and hydrometeors at all levels of the initial state. Additionally, the assimilation of satellite atmospheric motion vectors (AMVs) improves wind information and the atmospheric dynamics driving the moisture field which, in turn, increase the accuracy of the moisture convergence and fluxes at the core of the convection. As a result, the accuracy of the timing and intensity of a heavy rainfall prediction is improved, and the hourly and accumulated forecast errors are reduced.