Evaluating the urbanized ICON atmospheric model over the Bologna case study area

Urbanized regional climate models bridge the gap between local urban dynamics and larger atmospheric processes, covering metropolitan areas and their surroundings. These models are essential for understanding the interactions between urban areas and mesoscale dynamics, including the urban heat island’s regional impacts and extreme weather events. Therefore, km-scale models coupled with urban parameterizations are capable of capturing the complex interaction processes between atmosphere and urban land cover, having a key role for assessing heat stress and developing strategies for urban climate adaptation and mitigation.

To address this need, the bulk urban canopy parameterization, TERRA_URB (TU), was developed for the multi-layer land surface scheme of the COSMO regional atmospheric model. This parameterization resulted effective in capturing the key characteristics of urban areas, accurately reproducing prominent urban meteorological features across a range of European and global cities. As part of the transition from the COSMO model to the Icosahedral Nonhydrostatic (ICON) Weather and Climate regional model, TU was successfully ported into ICON. In this work, we present results of the TU parameterization, coupled into the ICON model, for high-resolution (2km) climate simulations driven by ERA5 reanalysis. These simulations focus on modelling extreme temperatures and urban heat islands in Bologna, a key case study in the CARMINE project (Climate Resilient Development Pathways in Metropolitan Regions of Europe, https://carmine-project.eu/). Bologna was selected due to its vulnerability to extreme heat and heatwave events, with a particular focus on enhancing urban climate resilience for vulnerable populations. The promising results of the newly urbanized ICON model, compared against high-quality observational data, contribute to advancing climate simulations for urban modeling applications, representing one of the first uses of the ICON model coupled with TU at the climate timescale, and could also serve as input data for urban models at hectometric scales.