Hydrological models are viewed as powerful tools that have a major importance for managing water resources and predicting flows. It should be specified that the meteorological parameter rainfall is the main input in these models. In the current study, data from only one rainfall station are available over the analysis domain, which cannot represent the entire Hammam Boughrara watershed of Algeria. The precipitation data remotely detected by the tropical rainfall measuring mission (TRMM) provide good spatial coverage in the watershed and can be used to fill in the missing data. The use of raw TRMM data gives poor results from the simulated flow rates with a Nash–Sutcliffe efficiency NSE equal to 0.34 for the validation period that ranges from year 2000 to 2005; this is mainly due to uncertainties in the TRMM data. For this reason, it was deemed necessary to carry out a performance test of the model. The results obtained give an unsatisfactory percent bias (PBIAS) of − 46.24%, which suggests the need to perform a correction to decrease the PBIAS of satellite precipitation. For this, two methods were used: the linear regression method and the multiplicative method. These two techniques can only be applied if there is at least one rainfall measurement station available in the watershed. The obtained results are very satisfactory since the PBIAS reaches − 0.62% for the linear regression method and − 11.58% for the multiplicative method. In addition, the use of corrected TRMMs gives also very good results with a Nash–Sutcliffe efficiency that ranges from 0.74 to 0.88 for both validation and calibration periods. Overall, the current study is supportive to estimate the satellite-based rainfall, one of the very sensitive to measure the meteorological parameter, in northwestern Algeria.