Capability of satellite rainfall products in simulating streamflows in the Black Volta Basin

The study employed pairwise statistics (BIAS, R and NSE) to assess the potential of five different satellite rainfall products (CHIRPS, PERSIANN, TRMM, RFE, and ARC) in reproducing the hydrology of the Black Volta Basin. This was achieved by using the HEC-HMS model calibrated with observed streamflow data from three gauged stations (Lawra, Chache and Bui). The study showed that the model is robust and capable of reproducing daily streamflow in the study basin with marginal BIAS (− 21.1–0.32%), satisfactory Nash–Sutcliffe Efficiency (0.73–0.85), and a high Pearson correlation coefficient (0.88–0.94) values during the simulation period (1998–2008). All the satellite rainfall products produced satisfactory results in terms of simulating the daily streamflows in the study basin, but the temporal dynamics were best replicated by CHIRPS, PERSIANN, and RFE. The gauge and satellite rainfall data performed better in simulating high streamflows as compared to the low streamflows in the study basin. There was a general underestimation of low and high streamflows in the study basin, which could be attributed to the errors in the satellite products that were propagated to the streamflow simulated through the hydrologic (HEC-HMS) model. Overall, the RFE dataset was better in simulating extreme streamflows in the study basin, this is followed by CHIRPS, PERSIANN, TRMM, and ARC in that order.