Dynamical Downscaling of ERA5 with COSMO-CLM at 2.2 km over European cities

The hourly high-resolution dataset (i.e., at 0.02°, ~2.2 km) for European cities, referred to as ERA5@2km, is derived by dynamically downscaling ERA5 reanalysis data to a convection-permitting scale for the past thirty years (1989–2018). The downscaling is performed using the Regional Climate Model COSMO-CLM (Rockel et al., 2008), with the TERRA-URB module (Wouters et al., 2016) activated to account for urban parameterizations. The precipitation dataset is designed to provide data suitable for applications such as pluvial flooding risk analysis, where hazards (e.g., inundated areas, water levels) and risks (e.g., estimated damages) are assessed. Conversely, the temperature dataset offers spatially detailed information that can be used to study urban heat island effects, as well as the occurrence and magnitude of heatwaves. Additionally, the dataset can support the training of neural networks and emulators. The overarching goal is to aid the Disaster Risk Reduction community involved in urban risk assessments by providing a robust basis for impact analysis at the city scale, focusing on extreme hourly precipitation and temperatures, while meeting the expected spatial and temporal requirements.