Coastal-urban meteorology: A sensitivity study using the WRF-urban model

Abstract

The air quality of a city can be assessed by applying a computational modelling approach. However, one of the most critical issues faced while running air quality models is based on the meteorological model setup, which can significantly affect simulations over a coastal urban area with complex terrain. Thus, this study evaluates the performance of the Weather Research and Forecasting (WRF) model using urban canopy model schemes considering three different model heights (4, 10, and 20 m), and updated land use and cover data (ESA-CCI). The one-year simulation results, based on the observed surface parameters (wind speed and direction, temperature, humidity, and precipitation) over the Metropolitan Region of Vitória (MRV), Southeast of Brazil, showed that the WRF model systematically overpredicted temperature and wind speed, and underestimated humidity, especially during the autumn and winter seasons. However, a significant improvement was obtained using the building effect parameterisation (BEP) formulation to emulate meteorological variables over urban surfaces, especially for wind speed. For temperature and relative humidity, the use of BEP did not show appreciable improvement because the BULK scheme (the control experiment) had a similar performance. The land-use dataset changes showed minimal impact on the representation of urban surfaces; however, significant differences were observed in natural/vegetated areas. The different land covers impacted the surface meteorological parameters, such as temperature and precipitation. However, due to the low coverage of monitoring stations in the countryside area, it was insufficient to determine whether the performance of the WRF model is favoured by the land-use update.