This thesis explores the meteorology in an urban area during a winter period and how well a numerical model is able to represent it. The air quality in the cities is of large interest to the citizens and policymakers. However, effects of buildings and anthropogenic activity on the meteorology in the city are complex. Lack of meteorological observations makes it difficult to create good predictions of the dispersion of tracers and to evaluate the models used as well. Meteorological modeling with focus on the planetary boundary layer (PBL) over Bjørvika, Oslo is performed using the Weather Research and Forecast (WRF) model combined with the single-layer urban canopy model (SLUCM). The simulated meteorological parameters are evaluated by comparison with observational measurements of the temperature at six levels, and the wind speed and the wind direction at two levels in a tower crane mast in Bjørvika, from January 9 to March 1, 2018. The results show a large difference of the meteorology between the model and the observations. The wind speeds are overestimated and the temperatures are underestimated by WRF. Furthermore, the atmosphere in the model is more stable compared to the observations when comparing hourly values, indicating that the model parameters will not represent the local dispersion correctly. However, the summed distribution of the atmospheric stability in WRF was more similar to the observations, indicating that the parameters from WRF could work better for dispersion estimations of longer timescales.