Model performance evaluation and computational efficiency of chemical mechanisms implemented in the microscale urban climate model PALM-4U

Large Eddy Simulation (LES) models can capture the inherent unsteadiness of atmospheric turbulence and therefore, permit to resolve local scale urban processes. A new highly efficient state-of-the-art microscale urban climate model - PALM-4U, has been developed that is able to simulate a variety of urban features such as the urban heat island, ventilation in street canyons, and pollution hotspots etc. PALM-4U is based on the well-established LES model PALM (Maronga et al., 2015). Among others, a fully coupled 'online' chemistry model has been implemented into PALM-4U which includes both gas-phase and aerosol chemistry. The model utilizes Kinetic PreProcessor KPP, Version 2.2.3 (Damian et al., 2002), to generate code for the numerical integration of the gas phase chemistry. The user may choose amongst several gas phase chemistry mechanisms of different complexity. Aerosol processes are described by the sectional aerosol model SALSA (Kokkola et al., 2008). In this work, we present model results for an area in central Berlin at a resolution of 10 meter. The model considers emissions from street canyons and selected point sources. Results show transport, gas-phase reactions and removal of atmospheric pollutants. The model performance has been evaluated by comparing modelled concentrations with the observed field data from Berlin, Germany. Three different chemical mechanisms, namely, CBM4, SMOG (a reduced chemistry mechanism, which includes only O₃, NO, NO₂, a highly simplified VOC chemistry and a small number of products) and photostationary equilibrium have been evaluated and the model performance statistics is presented. The computational efficiency of PALM-4U for chemistry simulations has been evaluated by benchmarking and scaling of chemical mechanisms over the available Intel cluster and HPC systems.