Sensitivity study of the PALM model system to different driving conditions

The LES-based PALM model system 6.0 is a state-of-the-art atmospheric numerical model widely used among scientists for urban boundary layer and urban climate studies. Despite being subjected to many validation and sensitivity studies which tested the model’s accuracy and applicability to urban environments, a major step is needed to test its sensitivity to different driving conditions.

In this study, we performed a series of PALM model simulations for a given domain and two selected episodes during the year 2019. Each simulation lasts 72 h, and all of them were performed for an 8 km x 8 km domain in 10 m resolution encompassing a real built-up residential area in the southeast part of the city of Prague, Czech Republic. The simulation setups are identical apart from  the initial and boundary conditions imposed. For that purpose we utilized the mesoscale WRF model and created an ensemble consisting of several members with different parameterization schemes.

This study's findings are serving to better understand how different initial and boundary conditions affect the PALM model simulations. In addition, we present the process, complexity, and challenges one can encounter while trying to find the optimal set of initial and boundary conditions for a given PALM model simulation. Moreover, the performed simulations have shown that most of the variance in the ensemble comes from the driving conditions.