We analyze a nearly 8-year record (2010-2018) of the superconducting gravimeter OSG-060 located at Djougou (Benin, West Africa). After tidal analysis removing all solid Earth and ocean loading tidal contributions and correcting for the long term instrumental drift and atmospheric loading, we obtain a gravity residual signal which is essentially a hydrological signal due to the monsoon. This signal is first compared to several global hydrology models (ERA, GLDAS, MERRA). Our superconducting gravimeter residual signal is also superimposed onto episodic absolute gravity measurements and to space gravimetry GRACE data. A further comparison is done using local hydrological data like soil moisture in the very superficial layer (0-1.2 m), water table depth and rainfall. The temporal evolution of the correlation coefficient between the gravity observation and both the soil moisture and the water table is well explained by the direct infiltration process of rain water together with the lateral transfer discharging the water table. Finally we compute the water storage changes (WSC) using a simulation based on the physicallybased Parflow-CLM numerical model of the catchment, which solves the water and energy budget from the impermeable bedrock to the top of the canopy layer using the 3D Richards equation for the water transfers in the ground, the kinematic wave equation for the surface runoff, and a land surface model (CLM) for the energy budget and evapotranspiration calculation. This model forced by rain is in agreement with evapotranspiration and stream flow data and leads to simulated water storage changes that nicely fit to the observed gravity signal. This study points out the important role played by surface gravity changes in terms of a reliable proxy for water storage changes occurring in small catchments.