Résumé
A new high-resolution (5 km) gridded daily precipitation dataset for Tunisia between 1979 and 2015 is introduced. This product combines 960 rain gauges with the SAFRAN analysis to produce the precipitation gridded data. A validation approach on two different datasets reveals that the SAFRAN analysis outperforms other standard interpolation methods such as Inverse Distance, Nearest Neighbors, Ordinary Kriging or Residual Kriging with altitude. When compared to EOBS, a widely used gridded dataset over Europe, a strong negative bias in EOBS precipitation is found. However due to the aridity and the low density of rain gauges in south Tunisia, results in this region must be analyzed with care. The SAFRAN product could be useful for various purposes such as climate model evaluation, climate studies, hydrological modelling to support the planning and management of surface water resources in Tunisia.