The PALM-4U model's performance with regard to different mesoscale initial and boundary conditions

Abstract

The microscale LES model PALM-4U is an open-source cutting-edge atmospheric numerical model used for urban environment modeling. Due to its ability to resolve the physical processes within the boundary layer and capture the effects of the small-scale processes within it, it’s becoming more popular within the urban boundary modeling community. In this experiment PALM-4U model’s response to different driving conditions was tested in order to see to what extent the microscale model results are determined by the driving conditions, and what mesoscale model provides the most optimal boundary conditions. Several different mesoscale models were used for creating the initial and boundary conditions for the PALM-4U simulations (e.g, ICON, WRF, ALADIN). Firstly, the mentioned models were compared among themselves, and their accuracy was tested against the measurements. All PALM-4U model simulations were performed on a real densely built urban environment located in the city of Prague, Czech Republic for a three-day episode during the year 2022. Apart from differing in the driving data, the simulations have identical configurations and are performed in 10 m resolution. The PALM-4U outputs were compared against measurements and among themselves. This work’s findings confirm that the microscale model is influenced to a large extent by the driving conditions and that the ICON mesoscale model has the best performance among the mesoscale models. In addition, we stress that the accuracy and precision of the mesoscale input data is of high importance for the microscale simulations due to the PALM-4U model’s high dependence on the driving conditions.