Modelling the influence of urban processes in St Petersburg on the near-surface atmospheric layer

The territory of a city is a surface with complex physical properties. Urbanized areas influence atmospheric processes in the near-surface layer, affecting the spatial and temporal distribution of gases and aerosols in a city. In general, numerical models, used for operational weather forecasts, consider the urban effect in a simplified way (one-layer models or by land-surface classification), which may not fully represent urban canopy processes. The aim of this study is to analyse the potential effect of St Petersburg, a Russian megacity, on the state and chemical composition of the near-surface atmosphere in the city and surrounding area using 3-D numerical modelling with urban effect parametrization.

Three numerical experiments were carried out using the high-resolution numerical model coupled with chemistry WRF-Chem. The experiments are as follows: CTRL (no St Petersburg as an urban class), URB (detailed urban effects by the Building Effect Parametrization and Local Climate Zones with 10 urban classes) and SimpURB (no explicit urban effect parametrization, with St Petersburg as a generic urban land-surface class). Simulations were performed for July 2021.

Compared with CTRL the URB simulation shows increases in near-surface air temperature (>2.5K), decreases in wind speed (>2m/s) and substantial changes in atmospheric composition including a reduction in NO2 mixing ratio (>30%). These impacts extend vertically above the city up to ~300-400 m and weaken with height. The vertical extent of the urban effect is most pronounced during nighttime, when the surrounding rural areas cool more rapidly. The urban representation also modifies atmospheric pollutant distributions, leading to a reduction in near-surface NO2 mixing ratio (by ~30%) and an increase in O3 compared to the CTRL case. In contrast, compared with SimpURB, the URB simulation exhibits higher NO2 mixing ratios, indicating strong sensitivity to the level of urban parameterisation. Overall, neglecting detailed building-resolved processes may lead to underestimation of NO2 and overestimation of O3 pollutant concentrations in simulations over St Petersburg. The strongest urban impacts are found in the western part of the city, covering the northern, central and southern municipal districts.