Simulation Of Squall Line Through Assimilating Doppler Radar Radial Velocity

The study of how the rational use of Doppler radar radial wind data for the numerical simulation provides a more accurate initial field, has important significance in the process of mesoscale weather numerical simulation and data assimilation. In this paper,using a new generation mesoscale forecasting model-WRF and its 3DVAR assimilation system, for the "090614" squall line, occurred in Jiangsu Province, we have been taken some sensitivity tests abort Doppler radar data assimilation and multiple microphysical parameterization scheme set of ensemble 3DVAR assimilation based on members. The major research works and conclusions are summarized as follows:(1) Based on strict quality control of the radar data, a series of sensitivity experiments have been designed according to the different combinations of scale factors and assimilation frequencies. It is shown that different combinations of scale factors and assimilation frequencies can improve the performances of initial conditions. Specifically, choosing appropriate scale factors has more distinct effects on improving the distribution characteristics of 700hPa wind, 850hPa temperature and 1500m height reflectivity in initial conditions, which can furthermore improve the short-term precipitation forecast. And based on this combination of scale factors, when assimilation frequency is reasonable, the distribution of reflectivity in initial conditions is improved in a greater degree and short-term rainfall prediction is more accurate.(2) Based on experimental results of selecting parameters of optimal scaling factor and frequencies of assimilation, we discussed and analyzed the configuration of the horizontal and vertical structure of the square lines. Further, it demonstrated that the cycle of assimilation, using radial velocity data of Doppler radar, can adjust the initial field effectively and has a crucial improvement to the prediction and simulation of squall lines.(3) In the precondition of considering the error of models, an ensemble of multi-microphysical parameterization 3DVAR scheme, which was used to do ensemble prediction of squall lines, was constructed through selecting parameters of optimal scaling factor and frequencies of assimilation. The discrete degree test of ensemble verified that the members of ensemble have strong predictability in the stage of occurrence, development of squall lines. The root mean square error test verified that it has a small prediction error and the ensemble prediction and observations are similar. The distribution of reflectivity in the height of 1500m and analysis of rainfall showed that the ensemble mean fields are similar to observation in the shape, position and distribution although the magnitude is slightly weak. Finally, the intensities of the ensemble members are closed to the observation, so it needs to consider both the ensemble members and ensemble mean in actual application.