Assessment of the Weather Research and Forecasting model implementation in Cuba addressed to diagnostic air quality modeling

This paper evaluates the implementation of the Weather Research and Forecasting model, WRF, for its use as the meteorological pre–processor for diagnostic air quality modeling in Cuba. The implementation of the WRF involved two studies: the first one was aimed at defining which global meteorological data is more suited for Cuba; the second one consisted of an analysis of the results for long–term runs on two domains, with the specific objective of assessing the general performance of the model. The results of the model were compared with the observations of the National Weather Service surface stations. The comparisons showed good performance for temperature and acceptable performance for prediction of wind tendencies. On average, the wind speed is overestimated in the model and the wind direction deviations exceed 30 degrees for several of the meteorological stations. These deviations are related to nearby topography and the low–wind speed. Some additional studies must be conducted in order to clarify and reduce the wind deviations. The research concludes that the WRF output is able to provide realistic meteorological patterns for air quality models, which require high–resolution three–dimensional (3D) meteorological data. The WRF–fsl tool was developed to use WRF to feed the local models as AERMOD when upper air data is not available. This tool takes the WRF output and gets the upper air data, in the fsl radiosonde format. The WRF–fsl results were compared to other solution, which incorporates a surface data parameterization. The conclusion is that the efforts, to run WRF for long periods, are not justified with the improvement in the results for regulatory purposes. However, as the differences in convective mixing height could be significant, this solution would be very useful for other kind of studies.