In meteorological, as well as air quality, modelling, input data plays an important role in the accuracy of the results, next to the model configuration. There are many sources of meteorological data available, both global and regional, and they differ not only by spatial and temporal resolution, but also by the number of observations included in the reanalysis and method of data assimilation used. In this study, the performance of the weather research and forecasting (WRF) model with two global reanalyses (ERA-Interim and NCEP FNL) used as input datasets has been assessed for a period of high tropospheric ozone concentrations. Both WRF model runs are in good agreement with observations, with IOA statistic ranging from 0.78 for wind speed to 0.98 for surface pressure. The ERA-Interim simulation showed better results for surface pressure, temperature and wind speed, while the performance of both datasets for parameters related to atmospheric moisture (e.g., dew point temperature) was comparable.
Keywords: tropospheric ozone, meteorological modelling, weather research and forecasting, WRF, atmospheric reanalysis, model evaluation