What is our pick? Assessment of satellite and reanalysis precipitation datasets over Italy

Italy. Gridded precipitation datasets, satellite and reanalysis based, offer the potential to provide precipitation estimates with quasi-global coverage, extending to areas not accessible by ground-based instruments. Even though their quality has improved over time, their performances vary across the globe, since their accuracy is related to the orography and climate of the area under analysis. This variability together with the increasing availability of freely accessible gridded datasets often lead scholars and hydrologists to assess the suitability of precipitation products for specific analyses. In this paper, we address this issue by evaluating the performance of six widely used gridded precipitation products (CERRA-Land, ERA5-Land, CHIRPS, CMORPH, IMERG, PERSIANN-CCS-CDR) over Italy. We benchmark them against the national rain gauge network dataset (SCIA) by computing continuous and categorical performance metrics across different temporal scales (daily, monthly, yearly). Our findings are presented at various spatial scales − from watershed to Köppen-Geiger climate zones − to support future hydrological studies with diverse objectives. The two reanalysis datasets (i.e., ERA5-Land and CERRA-Land) are the best performing across Italy, closely followed by IMERG satellite-based product. CMORPH and PERSIANN-CCS-CDR resulted the least performing. Although all the gridded datasets considered showed to have difficulties in detecting precipitation over mountainous areas, they proved to be a valuable alternative to conventional rain gauge networks, especially at longer temporal scales. • Evaluation of six precipitation datasets across Italy using the gauge-based SCIA dataset as ground-truth; • Reanalysis datasets perform best, with ERA5-Land and CERRA-Land outperforming satellite-based products; • Underestimation in mountainous areas observed across all datasets, improving at coarser temporal scales; • Reanalysis datasets recommended as the most reliable alternative in data-scarce regions.