Sensitivity of the LES model PALM in the urban environment: a case study in Prague

Numerical models using the large-eddy simulation technique (LES) are particularly useful in the urban environment in which their high resolution allows for capturing small-scale features of built-up areas such as buildings, roads, pavements or urban greenery. However, good model performance is not only the result of an accurate representation of the relevant physical processes. It also largely depends on the availability of proper input data describing the urban setup, namely the building and land-surface properties. In this study an LES-based modeling system PALM 6.0 featuring an improved urban surface model (USM) was used to assess sensitivity with respect to land-surface and building properties in a densely built-up residential area in Prague, Czech Republic, particularly in the vicinity of a typical crossroads. Two types of scenario simulations were employed. First, a set of synthetic scenarios changing surface and material parameters such as albedo or emissivity, by which the sensitivity of the model simulations to potentially erroneous input data was tested. These showed the highest sensitivity to the correct setting of surface parameters used in radiation balance equations. Second, a set of urbanistic scenarios was designed to assess the limits of effects of commonly considered urban-heat-island mitigation measures such as greening of the streets or altering surface materials. In this case, urban greenery is confirmed to be the most effective measure, especially when considering both physical and biophysical temperature indicators. On the other hand, analysis of air quality, specifically with respect to PM2.5 dispersion, confirmed the opposite behavior to that of thermal indicators; i.e. improved thermal comfort brings deterioration of PM2.5 concentrations.