The definition and derivation of a 0.125º, 3-hourly atmospheric forcing dataset for the South American continent is described which is appropriate for use in a Land Data Assimilation System and which, because of the limited surface observational networks available in this region, uses remotely sensed data merged with surface observations as the basis for the precipitation and downward shortwave radiation fields; while other variables are model-calculated values from South American Regional Reanalysis (SARR). (i.e., air temperature, wind speed and specific humidity at 2m, surface pressure, etc). The aim of this paper is to present, the rainfall retrieval methodology based on a multi-satellite technique (TRMM Multi-satellite Precipitation Analysis – TMPA) and a combination of additive and multiplicative bias correction schemes in order to get the lowest bias when compared with the observed values. Inter-comparisons and cross-validations tests have been carried out for the control algorithm (TMPA real-time algorithm) and the proposed merging scheme for several regions in South American continent.