Numerical experiments on assimilation of high-resolution radar thermodynamic data

The accuracy of weather forecast in Beijing is still far from satisfying the demands of people's routine life and disaster prevention and reduction, especially the quantitative forecast for local convective rainfall. To address the demands for improved 0-12 h short-term weather forecast in Beijing, especially the forecast of local convective rainfall in the summer, an initialization module that can be used in the Weather Research and Forecasting (WRF) model is developed based on Variational Doppler Radar Analysis System (VDRAS) in Institute of Urban Meteorology (IUM), China Meteorological Administration, Beijing. The radar data, which contain high temporal and spatial resolution three-dimensional thermodynamic characteristics, are assimilated into WRF model through the Four-Dimensional Data Assimilation (FDDA) method. Impacts of radar thermodynamic data on WRF model results are analyzed based on numerical simulation experiments of several rainfall cases. The results show that assimilation of the high-resolution radar thermodynamic data into WRF model can improve the simulation of the rainfall cases. The accuracies of simulated 2 m humidity, location, period and intensity of the rainfall are improved. The missing rates in the rainfall simulation also decrease with the application of data assimilation. Further analysis indicates that the assimilation of temperature and humidity is more important than the assimilation of wind for the improvement of the model results. Although the present study have shown that the assimilation of radar thermodynamic data into the WRF model can significantly improve the model results for the selected rainfall cases, more comprehensive and systematic investigation is needed to further study the effects of data assimilation in operational numerical model systems.