Comparison between Eulerian and Lagrangian models for simulating atmospheric conditions in urban environments

Abstract

Populations in urban areas have continued increasing in developing and developed countries worldwide. The ability to accurately simulate the dynamical and chemical atmospheric conditions in urban areas is essential when investigating the health impacts related to environmental conditions in urban areas. To simulate atmospheric conditions in urban areas one must consider how the airflow is affected by urban structures, vegetation, and mesoscale meteorology as well as local emitters and non-linear chemical reactions.  There are a variety of different models available to simulate atmospheric conditions in urban areas. These models differ in many ways such as numerical methods, spatial and temporal resolution, and initial and lateral boundary conditions. Here we compare an Eulerian (PALM) and a Lagrangian (GRAL) model to determine which model can best simulate atmospheric conditions in an urban environment. These numerical differences affect the computational resources required and therefore influence how each model can be used to study the urban environment. The initial and lateral boundary conditions used in the simulations are derived from the Weather Research and Forecasting (WRF) model coupled with Chemistry (WRF-Chem) and a library of horizontal wind fields for PALM and GRAL, respectively. The results from both models will be compared for a domain over Heidelberg, Germany over a two-week period against hourly surface observations of wind and air quality parameters.