Nutthida Kitwiroon and Sean Beevers
Environmental Research Group (ERG), King's College London, London, United Kingdom
The meteorological fields derived from four alternative WRF PBL (ACM2) and land surface models (5-layer thermal diffusion, Noah, PX and RUC) were analysed across 169 UK surface meteorological monitoring sites and 2 radiosonde stations to establish the influence that LSM and PBL schemes have in predicting wind speed, temperature, relative humidity, atmospheric stability and using CMAQ on NOx, NO2 and O3 predictions. The model was run for January and July 2006, covering a wide range of atmospheric conditions, from sub zero temperatures to greater than 30°C.
The balance of surface heat and moisture fluxes, friction velocity, atmospheric stability and PBL height are strongly influenced by different combinations of PBL and LSM schemes. Of the four experiments, the ACM2-RUC is able to to predict the magnitude and diurnal profile of temperature and relative humidity well, compared with observations. It also predicts wind speeds well in summer, although is does not perform as well in winter. The modelled vertical profiles are in reasonable agreement with observations, with insignificant differences predicted between the vertical profiles of temperatures and wind speed above the surface layer.
It is difficult to evaluate the measures of stability parameters due to the lack of measurements, consequentially, their effect on vertical mixing of pollutants in CMAQ have been quantified through the use of the diffusion coefficient (Kz). Using Kz the ACM2-RUC model has tendency to produce the strongest mixing and this is reflected in its predictions of NOx.
An ethane sensitivity analysis of CMAQ was performed to disentangle the influence of meteorology and emissions on NOx predictions. The new diagnostic analysis suggested that NOx emissions played an important role in the model's performance in January and that dispersion and emissions were responsible for model biases in July.
Overall, there is no combined physics setting in WRF that delivers the best performance against measurements across all meteorological variables and geographic locations. The combined land surface model and PBL scheme that delivered the best NOx, NO2 and O3 performance was recommended for use with the CMAQ model for UK policy applications.
Key words: PBL, land surface model, meteorological and air quality prediction, model evaluation, WRF, CMAQ
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner